UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Metoclopramide kinetics in sheep : maternal-fetal disposition, fetal pharmacodynamics and a comparison… Riggs, Kenneth Wayne 1988

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1989_A1 R53.pdf [ 12.73MB ]
Metadata
JSON: 831-1.0098178.json
JSON-LD: 831-1.0098178-ld.json
RDF/XML (Pretty): 831-1.0098178-rdf.xml
RDF/JSON: 831-1.0098178-rdf.json
Turtle: 831-1.0098178-turtle.txt
N-Triples: 831-1.0098178-rdf-ntriples.txt
Original Record: 831-1.0098178-source.json
Full Text
831-1.0098178-fulltext.txt
Citation
831-1.0098178.ris

Full Text

METOCLOPRAMIDE KINETICS IN SHEEP: ... .MATERNAL-FETAL DISPOSITION, FETAL PHARMACODYNAMICS A ' 'AND A COMPARISON BETWEEN PREGNANT AND NONPREGNANT EWES By KENNETH WAYNE RIGGS B.Sc. (Pharm), The University of British Columbia, 1971 M.Sc, The University of British Columbia, 1983 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (Faculty of Pharmaceutical Sciences) (Division of Pharmaceutics) We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA December, 1988 © KENNETH WAYNE RIGGS, 1988 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of Pharmaceutical Sciences The University of British Columbia Vancouver, Canada Date March 8, 1989 DE-6 (2/88) i i ABSTRACT Metoclopramide (MCP) is a potent antiemetic and gastric motility modifier, widely used to treat nausea and vomiting and to promote gastric emptying. This thesis reports the modification of a previously developed capillary column electron-capture detection gas-liquid chromatographic assay for MCP and its application to the study of MCP pharmacokinetics in nonpregnant ewes, pregnant ewes and fetal sheep. A switch from the split to splitless mode of sample injection allows sample injection to be automated and increases sensitivity, with a lower limit of 2 ng/mL. Metoclopramide was administered to nonpregnant and pregnant ewes by intravenous (i.v.) bolus injections on a cross-over basis. Transfer of the drug to the fetus was rapid and the ratio of fetal to maternal area under the plasma concentration-time curves averaged 0.74 indicating significant fetal exposure. Plasma concentrations in maternal and fetal plasma were best described by a biexponential equation in most animals with rapid distribution and elimination phases. The terminal elimination half-lives in maternal and fetal plasma averaged 71.3 and 86.8 min, respectively. Total body clearance and volume of distribution averaged 3.5 L/h/kg and 5.8 L/kg in the pregnant ewes, and 4.7 L/h/kg and 7.2 L/kg in nonpregnant animals. The only significant difference in pharmacokinetic parameters was total body clearance, which was 24% higher in nonpregnant animals. MCP followed dose-independent kinetics in both the pregnant and nonpregnant ewe over the 4-8-fold dose range studied. The placental and nonplacental clearances of MCP were studied in pregnant sheep using a two compartment open model. MCP was administered to the ewe and fetus as an initial i.v. bolus followed by constant rate infusion to steady-state. The paired infusions were separated by 48-72 i i i hours. A mean fetal to maternal steady-state concentration ratio of 0.57 was obtained following maternal drug administration. Protein binding of MCP at steady-state in maternal and fetal plasma averaged 49.5% and 39.5%. Total clearance of metoclopramide from the fetal lamb averaged 7.9 L/h/kg and was significantly greater than that in the ewe, 2.9 L/h/kg. Similarly, the placental clearance of drug from the fetus to the ewe (6.2 L/h/kg) was significantly greater than that from the maternal to the fetal compartment, 4.3 L/h/kg. Nonplacental clearance of MCP from the fetal compartment (1.7 L/h/kg) accounts for 20.5% of total fetal clearance, while nonplacental clearance from the maternal compartment (2.8 L/h/kg) contributes to 95.8% of the total clearance from the ewe. A marked accumulation MCP in fetal tracheal fluid occurred following both maternal and fetal dosing, with tracheal fluid concentrations exceeding those in fetal plasma by =15-fold. Significant accumulation also occurred in amniotic f luid. MCP persisted in both tracheal and amniotic fluids for more than 28 hours. Recirculation of MCP via fetal swallowing of these fluids as well as diffusion across fetal membranes may, in part, explain the longer half- l i fe of MCP in fetal plasma. Transfer of MCP from the amniotic cavity was rapid, with significant concentrations in umbilical venous (UV) and fetal arterial (FA) plasma 5 min after intra-amniotic drug injection. MCP was not detected in maternal plasma for a further 5-15 min and concentrations were =13% of those in fetal plasma. MCP concentration in tracheal fluid again exceeded fetal plasma levels by 15-fold. MCP concentrations were higher in UV than FA plasma suggesting uptake of drug from the amniotic fluid by the chorioallantoic membranes. Amniotic fluid would appear to serve as a reservoir for the fetus, with drug entering the fetal circulation via fetal swallowing and diffusion across the chorioallantois. i v Maternal or fetal drug administration had no effect on fetal cardiovascular parameters and, except for a small (-1.8 mm Hg), transient decline in P02 following maternal bolus dosing, fetal blood gases were unaltered. Amniotic pressure, breathing-like and electrocortical activities were also unaffected. V TABLE OF CONTENTS Page ABSTRACT i i LIST OF TABLES vi i i LIST OF FIGURES xi LIST OF SCHEMES xiv LIST OF ABBREVIATIONS xv ACKNOWLEDGEMENTS xviii 1. INTRODUCTION 1 1.1 Pharmacology, Clinical Uses and Side Effects 1 1.2 Use in Pregnancy 5 1.3 Factors Affecting Maternal-Fetal Drug Transfer and Fetal Drug Exposure 7 1.4 Metoclopramide Pharmacokinetics 9 1.5 Assay Methods 12 1.6 Rationale 15 1.7 Objectives 17 2. EXPERIMENTAL 19 2.1 Materials and Supplies 19 2.1.1 Chemicals 19 2.1.2 Reagents 19 2.1.3 Solvents 19 2.1.4 Gases 20 2.1.5 Supplies for Sheep Studies 20 2.2 Equipment 20 2.3 Stock and Reagent Solutions 21 2.4 Sample Extraction and Derivative Formation 22 2.5 Standard Curve Preparation 25 2.6 GLC 25 2.6.1 General Chromatographic Conditions for the Split!ess Mode of Sample Introduction 26 2.6.2 Optimization of GLC Conditions for the Splitless Injection Mode 26 2.7 Pharmacokinetic Studies in Pregnant and Nonpregnant Sheep 27 2.7.1 Animal Handling 27 2.7.2 Surgical Preparation 27 2.7.3 Experimental Protocol 2.7.3.1 Intravenous Bolus Studies 2.7.3.2 Intravenous Infusion to Steady-State 2.7.3.3 Intra-Amniotic Bolus Experiments 2.8 Recording Procedures and Blood Gas Analysis 2.9 Plasma Protein Binding at Steady-State 2.10 Quantitative Plasma, Amniotic and Tracheal Fluid Analysis 2.11 Data Analysis 2.11.1 Computer Fitting 2.11.2 Calculation of Pharmacokinetic Parameters 2.11.3 Fetal Arterial Pressure, Fetal Heart Rate, Amniotic (Intrauterine) Pressure 2.11.4 Fetal Breathing-Like and Electrocortical Activities 2.11.5 Statistical Tests RESULTS 3.1 GLC Assay Modifications 3.2 Selection of Weighting Factor for Computer Data Analysis 3.3 Metoclopramide Pharmacokinetics Following I.V. Bolus Dosing to the Ewe 3.4 Metoclopramide Pharmacokinetics Following Maternal and Fetal Infusions to Steady-State 3.5 Metoclopramide Placental (Transplacental) and Nonplacental Clearances 3.6 MCP Accumulation in Fetal Tracheal and Amniotic Fluids 3.7 Metoclopramide Pharmacokinetics Following Intra-Amniotic Drug Administration 3.8 Physiological Effects of Metoclopramide on the Fetal Lamb 3.8.1 Fetal Arterial Blood Gases 3.8.2 Fetal Arterial Pressure, Fetal Heart Rate and Amniotic Pressure (Intrauterine Pressure) 3.9 Behavioural Effects of Metoclopramide on the Fetal Lamb 3.9.1 Fetal Breathing and Electrocortical Activities 4. DISCUSSION 4.1 Metoclopramide Assay Modification and Application vi i 4.2 Selection of Weighting Factor for Computer Data Analysis 136 4.3 Metoclopramide Pharmacokinetics in Chronically Instrumented Pregnant and Nonpregnant Sheep 139 4.3.1 Maternal and Fetal Drug Concentration Ratios 140 4.3.2 Metoclopramide Pharmacokinetics in the Pregnant Ewe 146 4.3.3 Metoclopramide Pharmacokinetics in the Fetus 148 4.3.4 Comparison of Pharmacokinetic Parameters in Pregnant and Nonpregnant Ewes 152 4.3.5 Metoclopramide Pharmacokinetics in the Nonpregnant Ewe 154 4.4 Metoclopramide Pharmacokinetics in the Chronically Instrumented Ewe and Fetus Following Infusions to Steady-State. 155 4.4.1 Maternal MCP Infusions 155 4.4.1.1 Fetal Exposure to Metoclopramide 155 4.4.1.2 Maternal Metoclopramide Pharmacokinetics 157 4.4.1.3 Metoclopramide Pharmacokinetics in the Fetus 160 4.4.2 Fetal MCP infusions 161 4.4.2.1 Metoclopramide Pharmacokinetics in the Ewe and Fetus 161 4.5 Maternal and Fetal Metoclopramide Plasma Protein Binding at Steady-State. 166 4.6 Metoclopramide Placental (Transplacental) and Nonplacental Clearances Following Maternal and Fetal Infusions to Steady-State 169 4.7 Metoclopramide Accumulation in Fetal Tracheal (Lung) and Amniotic Fluids Following Maternal or Fetal Administration 184 4.8 Pharmacokinetics of Metoclopramide in Plasma, Amniotic and Tracheal Fluids Following Intra-Amniotic Drug Administration 196 4.9 Physiological and Behavioural Effects of Metoclopramide on the Fetal Lamb Following Maternal or Fetal Dosing. 206 4.9.1 Fetal Blood Gases, Heart Rate, Arterial Pressure and Amniotic Pressure 207 4.9.2 Fetal Breathing-Like and Electrocortical Activities 208 5. SUMMARY AND CONCLUSIONS 212 6. REFERENCES 220 7. APPENDIX 237 VI 11 LIST OF TABLES Page 1. Comparison of GC operating conditions for the split and splitless modes of sample injection. 49 2. Weighting factor determination for computer fitting of MCP i.v. bolus concentration versus time data. 52 3. Weighting factor determination for computer fitting of MCP infusion concentration versus time data. 53 4. Distribution and disposition half-lives and compartmental rate constants in the pregnant ewe following i.v. bolus doses. 56 5. Pharmacokinetic parameters obtained in the pregnant ewe following i.v. bolus doses. 57 6. Pharmacokinetic parameters obtained in the fetal lamb following maternal i.v. bolus doses. 62 7. Distribution and disposition half-lives and compartmental rate constants in the nonpregnant ewe following i.v. bolus crossover 67 doses. 8. Pharmacokinetic parameters obtained in the nonpregnant ewe following i.v. bolus doses on a crossover basis. 68 9. Maternal and fetal weights and gestational age of the animals used for the maternal (MI) and fetal (FI) MCP infusion 71 experiments. 10. Steady-state metoclopramide plasma protein binding (Bm,Bf), plasma concentration (Cm, Cf) and concentration ratio (fetus/mother) following maternal infusion. 74 11. Maternal distribution and disposition half-lives and compartmental transfer rate constants following constant rate infusion of metoclopramide to the ewe. 75 12. Maternal pharmacokinetic parameters obtained following constant rate infusion of metoclopramide to the ewe. 76 13. Post-infusion estimates of fetal arterial plasma terminal elimination rate constant and half- l i fe following maternal constant rate infusion. 79 14. Steady-state metoclopramide plasma protein binding (Bm,Bf), plasma concentration (Cm', C f ) and concentration ratio (fetus/mother) following fetal infusion. 81 15. Fetal distribution and disposition half-lives and compartmental transfer rate constants following constant rate infusion of metoclopramide to the fetal lamb. 83 16. Pharmacokinetic parameters in the fetal lamb following constant IX rate infusion of metoclopramide to the fetus. 85 17. Post-infusion estimates of maternal arterial plasma terminal elimination rate constant and half- l i fe following fetal constant rate infusion. 86 18. Measurement of metoclopramide adsorption to ultrafiltration membrane. (Mean + SD). 88 19. Placental and nonplacental clearances of metoclopramide (total drug concentration) in maternal and fetal sheep. 91 20. Total clearance of metoclopramide (total drug concentration) and percent of nonplacental contribution to total drug 92 elimination in the ewe and fetus. 21. Placental and nonplacental clearances of metoclopramide (unbound drug concentration) in maternal and fetal sheep. 94 22. Total clearance of metoclopramide (unbound drug concentration) and percent of nonplacental contribution to total drug elimination in the ewe and fetus. 95 23. Values for the ratio of drug concentration in amniotic (AMN/FA) and tracheal (TR/FA) fluids to that in fetal arterial plasma and estimated elimination half-lives in tracheal f luid. 97 24. Maternal and fetal weights and gestational age of the animals used for the 10 mg MCP intra-amniotic bolus experiments. 104 25. Pharmacokinetic parameters obtained in maternal (MA) and fetal (FA) arterial and umbilical venous (UV) plasma, amniotic (AMN) and tracheal fluid (TR) following a 10 mg intra-amniotic injection of MCP. 106 26. Effect of metoclopramide on fetal arterial pH, Po2 and Pco2 following maternal intravenous bolus dosing. [Mean (±SD)]. 109 27. Effect of metoclopramide maternal or fetal infusion on fetal arterial pH, Po2, Pco2, 02 content (02C), and hematocrit (HCT). I l l [Mean (±SD)]. 28. Maternal and fetal weights and gestational age of the animals used for the control maternal and fetal normal saline infusions. 113 29. Effect of control maternal or fetal infusion of normal saline on fetal arterial pH, Po2, Pco2, 02 content (02C), and hematocrit (HCT). [Mean (±SD)]. 114 30. Effect of intra-amniotic drug administration on fetal arterial pH, Po2, Pco2, 02 content (02C), and hematocrit (HCT). 115 [Mean (±SD)]. 31. Mean (±SE) values for amniotic pressure (Amn P), fetal heart rate (HR) and fetal arterial pressure (AP) calculated over 30 minute intervals before, during and following infusion of X metoclopramide to the ewe. (n = 10). 119 32. Mean (±SE) values for amniotic pressure (Amn P), fetal heart rate (HR) and fetal arterial pressure (AP) calculated over 30 minute intervals before, during and following infusion of metoclopramide to the fetus, (n = 9). 120 33. Mean (±SE) values for fetal heart rate (HR) and arterial pressure (AP) calculated over 30 minute intervals before, during and following infusion of normal saline to the ewe or fetus. 121 34. Incidence of low and high voltage in electrocortical episodes (ECoG) and breathing activity (BR) in the fetus following a 40 mg i.v. bolus of metoclopramide to the ewe. 124 35. Incidence of low and high voltage in electrocortical episodes (ECoG) and breathing activity (BR) in the fetal lamb following maternal MCP infusion. 125 36. Incidence of low and high voltage in electrocortical episodes (ECoG) and breathing activity (BR) in the fetal lamb following infusion of metoclopramide to the fetus. 127 37. Incidence of low and high voltage in electrocortical episodes (ECoG) and breathing activity (BR) in the fetal lamb following infusion of normal saline. (Control study). 129 38. Data for comparison of mean placental (transplacental) and nonplacental clearances of various drugs in the ewe and fetus employing the model proposed by Szeto et al., 1982b. (Total drug concentration). Also included is the percentage of drug bound to maternal (m) and fetal (f) plasma proteins. 177 39. Data for comparison of mean total clearance and percent of nonplacental contribution to total drug elimination for various drugs in the ewe and fetus employing the model by Szeto et al., 1982b. (Total drug concentration). 178 xi LIST OF FIGURES Page 1. MCP and MAP recovery as a function of purge activation time. Triplicate 2 pi injections of separate samples containing MCP 16 ng/mL or MAP 80 ng/mL. 46 2. Representative chromatograms of the heptafluorobutyryl derivatives of MCP (1) and the internal standard, MAP (2) in maternal arterial plasma (MA), fetal arterial plasma (FA), tracheal fluid (TR) and amniotic fluid (AMN). Samples were taken at 45 min into a maternal steady-state infusion and are presented with their superimposed control (blank). 47 3. Coefficient of variation study showing a typical standard curve of sheep plasma extracts obtained by plotting the area ratio of the heptafluorobutyryl (HFB) derivatives of MCP/MAP versus MCP concentration, n = 4, duplicate injections; Y = 0.018X - 0.001, r = 0.999. 51 4. Representative semilogarithmic plots of metoclopramide concentration versus time profiles in maternal (MA) and fetal (FA) arterial plasma, tracheal (TR) and amniotic (AMN) fluids following a 40 mg i.v. bolus to a pregnant ewe. 55 5. Comparison of pharmacokinetic parameters obtained in pregnant and nonpregnant ewes and fetuses following i.v. bolus crossover doses of 10, 20 and 40 mg to the ewe. The data are presented as the overall mean and 1 SD. Key: (a), significantly different from maternal t i o (p = 0.0167); (b) significantly different from pregnant ewe CL5 (p = 0.0487). 64 6. Representative semilogarithmic plots of metoclopramide concentration verus time profiles in arterial plasma following 10, 20, 40 and 80 mg crossover i.v. bolus doses to a nonpregnant ewe. 66 7. Representative semilogarithmic plots of the maternal (MA) and fetal (FA) arterial plasma, tracheal (TR) and amniotic (AMN) fluid metoclopramide concentration versus time profiles obtained following a 15 mg i.v. bolus loading dose and a constant rate infusion of 0.21 mg/min for 90 min to the ewe. 72 8. Representative semilogarithmic plots of the maternal (MA) and fetal (FA) arterial plasma, tracheal (TR) and amniotic (AMN) fluid metoclopramide concentration versus time profiles obtained following a 5 mg i.v. bolus loading dose and a constant rate infusion of 0.07 mg/min for 90 min to the fetus. 80 9. Representative semilogarithmic plots of the maternal (MA) and fetal (FA) arterial plasma, tracheal (TR) and amniotic (AMN) fluid metoclopramide concentration versus time profiles obtained following a 15 mg i.v. bolus loading dose and a constant rate infusion of 0.21 mg/min for 90 min to the ewe. 100 xi i 10. Representative semi logarithmic plots of the maternal (MA) and fetal (FA) arterial plasma, tracheal (TR) and amniotic (AMN) fluid metoclopramide concentration versus time profiles obtained following a 5 mg i.v. bolus loading dose and a constant rate infusion of 0.07 mg/min for 90 min to the fetus. 101 11. Representative semilogarithmic plots of the maternal arterial (MA), fetal arterial (FA) and umbilical venous (UV) plasma, tracheal (TR) and amniotic (AMN) fluid metoclopramide concentration versus time profiles obtained following a 10 mg intra-amniotic bolus injection. 105 12. Representative control recordings of IUP (intra-uterine pressure), FAP (fetal arterial pressure), HR (fetal heart rate), ITP (intra-tracheal pressure), ECoG (fetal electrocortical activity), E0G (fetal electroocular activity). 117 13. a) A diagrammatic representation of the anatomic relationships between the fetal lamb and the amniotic and allantoic fluid compartments (Modified from Fig. 1, Szeto et al., 1979). b) A diagrammatic illustration of the major elements contributing to drug disposition and elimination within the ovine conceptus. 203 APPENDIX FIGURES 1. a) Total ion current mass chromatogram and b) electron impact mass spectrum of the HFB-derivative of MCP. 241 2. The postulated fragmentation pattern of the HFB-derivative of MCP based on the most prominent fragment ions obtained with electron impact GC-MS. 242 3. a) Total ion current mass chromatogram and b) electron impact mass spectrum of the HFB-derivative of di-deethylated MCP. 243 4. The postulated fragmentation pattern of the HFB-derivative of di-deethylated MCP based on the most prominent fragment ions obtained with electron impact GC-MS. 244 5. Chemical ionization mass spectra of the HFB-derivative of MCP. a) positive-ion. b) negative-ion. 245 6. Chemical ionization mass spectra of the HFB-derivative of di-deethylated MCP. a) positive-ion. b) negative-ion. 246 7. a) Total ion current mass chromatogram for a plasma extract obtained from a pregnant ewe receiving a 40 mg i.v. bolus of MCP. b) electron impact mass spectrum of the peak with a retention time of 11.69 min corresponding to the HFB-derivative of MCP. 247 8. The postulated fragmentation pattern of the HFB-derivative of the mono-deethylated metabolite of MCP based on the most prominent fragment ions obtained with electron impact GC-MS. 248 a) Total ion current mass chromatogram for a urine extract obtained from a pregnant ewe receiving a 40 mg i.v. bolus of MCP. b) electron impact mass spectrum of the peak with a retention time of 11.53 min corresponding to the HFB-derivative of the mono-deethylated metabolite of MCP. LIST OF SCHEMES 1. Metoclopramide extraction procedure. 2. Two compartment open model. 3. Two compartment open model used for the determination of MCP placental (transplacental) and nonplacental clearances. xiv Page 23 59 90 XV LIST OF ABBREVIATIONS AAG alphaj-acid glycoprotein AMN amniotic Amn P amniotic pressure ANOVA analysis of variance AP arterial pressure AUC area under the plasma concentration versus time curve AUMC area under the f irst moment curve Bf percent of MCP bound to fetal plasma protein Bm percent of MCP bound to maternal plasma protein b.p. boiling point BR fetal breathing-like activity C.V. coefficient of variation Cf fetal steady-state MCP plasma concentration CLff total clearance from the fetal compartment CLfm placental (transplacental) clearance of drug from the fetal to maternal compartment CLf 0 nonplacental clearance of drug from the fetal compartment CLmf placental (transplacental) clearance of drug from the maternal to fetal compartment CLm m total clearance from the maternal compartment CL m o nonplacental clearance of drug from the maternal compartment CLS systemic or total body clearance Cm maternal steady-state MCP plasma concentration ECD electron capture detector xvi ECoG electrocortical (electrocorticographic) activity ELS extended least squares Eq. equation ER extraction ratio FA fetal arterial FI fetal infusion Fig. figure g acceleration due to gravity GC gas chromatography GC-MS gas chromatography mass spectrometry GLC gas-liquid chromatography GLC-ECD gas-liquid chromatography with electron capture detection HCT hematocrit HFB heptafluorobutyryl HFBA heptafluorobutyric anhydride HPLC high pressure liquid chromatography HR heart rate i .d. internal diameter i.v. intravenous IPL isolated perfused lung ko infusion rate MA maternal arterial MAP maprotiline MCP metoclopramide MCP.HCL metoclopramide hydrochloride MI maternal infusion n number of samples (animals) xvi i n/a not applicable p .s . i . pounds per square inch PTFE polytetraf1uoroethylene Qum umbilical blood flow r correlation coefficient r 2 coefficient of determination SD standard deviation SE standard error ^ mean residence time TLC thin layer chromatography TR tracheal h half- l i fe uv umbilical venous ^area apparent volume of distribution calculated from AUC v ss volume of distribution at steady-state WLS weighted least squares xv i i i ACKNOWLEDGEMENTS I would like to sincerely thank Dr. James Axel son and Dr. Dan Rurak for their supervision, encouragement and friendship during the course of this project. Without either of you it would not have been possible to carry out these studies. Your assistance and guidance in the preparation of this thesis is also very much appreciated. Special thanks to both of you as well for your financial support. I would also like to extend my appreciation to Sandy Taylor, Sun Dong Yoo and Eddie Kwan for their endless help with all aspects of the experimental studies and for their friendship. Similarly, sincere thanks to Barbara McErlane for her assistance with the assays and her encouragement. Thanks to my committee members Dr. Frank Abbott, Dr. Jim Orr and Dr. Marc Levine for their suggestions and assistance during the course of this work. Special thanks to Grace Chan for preparation of the mass spectral figures and her help with the plotting routines. And last, but not least, special thanks to my wife, Sheila for her suggestions, grammatical corrections and proofing of this document. This project was made possible by grants received from the Medical Research Council of Canada. This thesis is dedicated to wonderful wife, Sheila, whose constant understand encouragement, unending patience and support have carried me through my Ph.D. program my dear Mom and brother, the memory of my Dad, and to the newest joy in my l i fe , my son, Trevor 1 1. INTRODUCTION 1.1 Pharmacology, Clinical Uses and Side Effects Metoclopramide (MCP), is the 2-methoxy, 5-chloro analog of procainamide, also known chemically as 4-amino-5-chloro-2-methoxy-N(2-diethyl aminoethyl) benzamide. In spite of their close structural similarities, there is a large difference in pharmacodynamics of the two compounds. Metoclopramide is a potent antiemetic and promotes gastric motility, while procainamide is known for its local anaesthetic and antiarrhythmic properties. Metoclopramide's pharmacological actions, indications for use, side effects and mechanisms of action have been the subject of several comprehensive reviews (Pinder et al., 1976; Smith and Salter, 1980; Ponte and Nappi, 1981; Schulze-Delrieu, 1981; Albibi and McCallum, 1983; Harrington et al., 1983; Shaughnessy, 1985; Desmond and Watson, 1986). Although MCP is an analog of procainamide, it has negligible antiarrhythmic or local anaesthetic properties (Harrington et al., 1983; Desmond and Watson, 1986). The pharmacological effects of MCP are most evident in the gastrointestinal tract, where a pronounced increase in gastric motility is seen in both animals and man after either oral or i.v. administration (Pinder et al., 1976; Schulze-Delrieu, 1981). Clinically, MCP is used to facilitate diagnostic procedures (radiologic, endoscopic) of the upper gastrointestinal tract and small Metoclopramide 2 intestine and to treat a variety of functional and organic gastrointestinal disorders such as stasis and reflux esophagitis. The effects of MCP include an increase in the resting tone of the lower esophageal sphincter, an increase in the amplitude of gastric contractions, relaxation of the pyloric sphincter and duodenal bulb during gastric contractions and an increase in peristalsis of the proximal small intestine. This results in accelerated gastric emptying and shortened transit time through the small bowel (Schulze-Delrieu, 1981; Albibi and McCallum, 1983; Harrington et al., 1983). In contrast, MCP has l i t t le effect on the lower small bowel or colon (Desmond and Watson, 1986). The precise mechanisms of MCP's gastrointestinal effects are not fully understood, but are thought to be related to its ability to antagonize the inhibitory neurotransmitter dopamine, to augment acetylcholine release and sensitize the muscarinic receptors of the gastrointestinal smooth muscle, and a direct action on smooth muscle (Harrington et al., 1983, Desmond and Watson 1986). Metoclopramide is also widely used as an antiemetic in the treatment of nausea and vomiting of varying etiologies, for example following surgery (Harrington et al., 1983), cancer chemotherapy (Harrington et al., 1983; Sailer et al., 1985; Desmond and Watson , 1986; Joss et al., 1986; Taylor et al., 1987), in uremia (Desmond and Watson, 1986), following radiation therapy (Pinder et al., 1986) and during pregnancy (Harrington et al., 1983). The antiemetic effects of MCP are thought to be mediated through both peripheral (gastrointestinal) and central sites. Metoclopramide is believed to raise the threshold of the chemoreceptor trigger zone through antagonism of dopamine receptors, and to decrease the sensitivity of visceral nerves which transmit afferent impulses from the gastrointestinal tract to the emetic centre in the lateral reticular formation (Pinder et al., 1976). The drug is also used in the treatment of gastroparesis 3 associated with diabetes and vagotomy (Desmond and Watson, 1986), in migraine therapy to enhance gastric emptying thereby promoting absorption of oral antimigraine medications (Shaughnessy, 1985; Desmond and Watson, 1986), to promote lactation by stimulation of prolactin release (Harrington et al. , 1983) and appears to be of benefit in the treatment of orthostatic hypotension, vertigo, hiccups, feeding intolerance in infants and anxiety nervosa (Shaughnessy, 1985; Harrington et al., 1983). The drug has also been used prior to induction of general anaesthesia for elective and emergency surgery (Solanki et al., 1984; Shaughnessy, 1985; Gipson et a / . , 1986) as well as for Caesarian section (Bylsma-Howel1 et al., 1983; Cohen et al., 1984; Shaughnessy, 1985), where its ability to increase lower esophageal sphincter tone and accelerate gastric emptying reduce the risk of aspiration of stomach contents. More recently it has been shown to be of benefit in preventing nausea and vomiting associated with epidural anaesthesia during elective Caesarian section (Chestnut et al., 1987). While MCP is structurally related to procainamide, animal studies have not demonstrated any significant effects on blood pressure or intracardiac conduction. There are, however, occasional reports of hypotension during general anesthesia, hypertensive crisis in patients with pheochromocytoma and cardiac arrhythmias in man (Harrington et a l . , 1983). A decrease in renal plasma flow of -20% has been reported in oncology patients receiving high drug doses (1-2.5 mg/kg) (Israel et al., 1986) as has a decrease in hepatic blood flow in the rat (Tam et al., 1981) at doses of >25 mg/kg. The effects of MCP on gastrointestinal motility may alter either the rate or the extent of absorption of a number of coadministered drugs. An increased rate of absorption has been reported for acetaminophen, aspirin, lithium, diazepam, metoprolol, propranolol, ethanol (Desmond and Watson, 4 1986) and cyclosporin (Wadhwa et al., 1987), while a decrease in the absorption and plasma concentrations of cimetidine (Gugler et al., 1981), digoxin (Desmond and Watson, 1986; Kirch et al., 1986) and quinidine (Yeun et al. , 1987) have been observed. Metoclopramide also stimulates the release of various hormones in man and in animals. Prolactin release has been observed in the rat (Harrington et al., 1983), healthy adults (Harrington et al., 1983), pregnant women (Arvela et al., 1983; Harrington et al., 1983), newborns (Ruppert et al., 1983), children (Harrington et al. , 1983) and in the pregnant ewe (Fitzgerald and Cunningham, 1982). This effect is considered to be mediated by MCP's ability to antagonize the dopamine-mediated inhibition of prolactin secretion by the hypothalamus or pituitary. Increases in plasma aldosterone have also been reported in rats (Harrington et a7., 1983) and man (Harrington et al., 1983; Sommers et al., 1988) apparently as a result of inhibition of central dopamine receptors (Albibi and McCallum, 1983). More recent information, however, suggests that the aldosterone response to MCP is mediated by acetylcholine release from post-gang!ionic cholinergic nerve terminals within the adrenal cortex (Sommers et al., 1988). Stimulation of vasopressin secretion, by an as yet unknown mechanism, has also been reported in healthy human volunteers (Norbiato et al., 1986). The effect of MCP on the secretion of growth hormone (Harrington et al., 1983) and thyrotropin (Ruppert et al., 1986) remains controversial, with some studies reporting no change and others a stimulated release. Metoclopramide has been reported to readily cross the blood-brain barrier (Smith and Salter, 1980), entering the central nervous system to antagonize the'effects of dopamine (Harrington et al., 1983; Desmond and Watson, 1986). When administered in the usual therapeutic doses (30-40 mg daily; maximum 0.5 mg/kg/day) MCP causes few adverse reactions, although it can induce biochemical and behavioural changes characteristic of neuroleptics (e.g. chlorpromazine, haloperidol) (Grimes et a7., 1982; Harrington et al., 1983). Side effects such as lassitude, drowsiness, transient feelings of agitation and anxiety have been reported to occur in -5-10% of patients (Robinson, 1973; Pinder et al., 1976; Smith and Salter, 1980; Harrington et al., 1983; Desmond and Watson, 1986). These effects are normally mild, transient and reversible upon withdrawl of the drug. Alarming extrapyramidal reactions (motor restlessness, dystonias, and Parkinsonian-1ike symptoms of tremor, rigidity and akinesia) have also been reported in =1% of patients at therapeutic dose levels. Their incidence is more common after high doses, with long-term use and in the elderly (Harrington et al., 1983; Desmond and Watson, 1986). Infants and children also appear to be more susceptible to the neurological effects of MCP (Casteels-van Daele et al., 1970; Low and Goel 1980; Bateman et a7., 1983; Harrington et al., 1983) with an increased incidence of extrapyramidal reactions, particularly acute dystonias. It should be noted, however, that these effects are generally rare and have occurred in this population largely because the recommended maximum dose (usual 0.1 mg/kg/day; maximum 0.5 mg/kg/day) had been exceeded (Pinder et al., 1976; Reynolds, 1978; Low and Goel, 1980). Side effects also appear to be more common in patients with renal impairment (Bateman and Gokal, 1980; Lehmann et al., 1985). 1.2 Use in Pregnancy Metoclopramide has been shown to significantly reduce the emesis of early pregnancy (Singh and Lean, 1970) as well as during labour (McGarry, 1971; Vella et ah, 1985) when compared to a placebo. There have been no reports of congenital malformations, or drug-related effects on fetal size and weight in the offspring of mice, rats or rabbits following oral or i.v. 6 administration during various stages of gestation (Pinder et al., 1976). Similarly, Singh and Lean, 1970, in their study of MCP's effectiveness in controlling the emesis of early pregnancy, found no increase in the incidence of abnormalities in the babies of the 120 women receiving 30 mg of MCP daily, either orally or i .v. , over a seven day period at various times between <12-19 weeks of gestation. There have been no subsequent reports of malformations either in laboratory models or in the clinical literature (Harrington et al., 1983), but because its effects on the human fetus have not been definitely established it is not recommended for use in the f irst trimester (Desmond and Watson, 1986). Metoclopramide has been found to significantly increase the gastric emptying rate of women in labour (Howard and Sharp, 1973) and to increase the tone of the lower esophageal sphincter (Brock-Utne et al., 1978), which, like gastric emptying, is impaired during pregnancy and labour. These positive actions, as well as the ability to significantly diminish the frequency of vomiting during labour have led to its more frequent use in obstetrical anaesthesia (elective, emergency) in North America (Bylsma-Howell et al., 1983; Cohen et al., 1984), Europe and Japan (Schulze-Delrieu, 1981). Metoclopramide has been shown to be effective in reducing the risk factors associated with aspiration pneumonitis in patients undergoing general anaesthesia (Shaughnessy, 1985). In the study of elective Caesarian patients by Bylsma-Howel1 et al., 1983, MCP significantly reduced gastric volume when compared to patients receiving a normal saline placebo. In contrast, Cohen et al., 1984, in an essentially identical study, found no significant differences in gastric volumes between treated and untreated groups. The number of patients "at risk", however, decreased significantly when given metoclopramide. 7 Investigations have demonstrated that MCP readily crosses the placenta in humans (Riggs, 1982; Arvela et al., 1983; Bylsma-Howell et al., 1983; Cohen et al., 1984) and in a preliminary study in sheep (Riggs, 1982). Studies at term in human pregnancy have shown no effect on the newborn, based upon Apgar assessment at delivery and extensive neurological and adaptive capacity testing conducted for 24 hours after birth. Effects of the drug on the fetal lamb in utero were not examined in the preliminary report, but have been extensively assessed in the current work. 1.3 Factors Affecting Maternal-Fetal Drug Transfer and Fetal Drug Exposure The placenta is lipoprotein in nature and as a result the mechanisms for placental drug transfer are essentially the same as for any other biological membrane. Passive diffusion is considered to be the most important mode of transport for drugs, and, for l ipid soluble molecules at least, there is no significant barrier to transfer of compounds with a molecular weight of less than 600 (Reynolds, 1979). Other mechanisms such as active transport, pinocytosis and passage through membrane pores are felt to be generally unimportant for drugs although important for some biological materials such as amino acids and immunoproteins. Virtually all commonly used drugs will cross the placenta, although the rate and extent of transfer will vary depending upon their physico-chemical properties. Because of this, considerable attention has been devoted to the study of the mechanisms involved. In addition, those factors (pharmacokinetic, placental) governing the rate and extent of transfer which ultimately determine the drug's concentration in the mother and subsequently in the fetus/neonate have been thoroughly reviewed (Mirkin and Singh, 1976; Reynolds, 1979; Brock-Utne et al., 1980; Krauer et al., 8 1980; Ward et al., 1980; Dilts, 1981; Waddell and Marlowe, 1981; Yurth, 1982; Bogaert and Thiery, 1983; Mihaly and Morgan, 1984; Mucklow, 1986; Mitani et al., 1987); these complex factors may be briefly summarized as follows: 1. Maternal plasma drug concentrations are dependent upon (i) the total dose, site and mode of drug administration, (ii) distribution, which may be affected by hemodynamics, tissue affinity and the extent of protein binding, ( i i i ) the rate and degree of drug metabolism and (iv) the rate of excretion (intact drug, metabolites). 2. Placental drug transfer is a function of (i) the physico-chemical properties of the drug such as its l ipid solubility, molecular weight, degree of ionization and protein binding, (ii) the concentration gradient of free drug across the placenta, ( i i i ) maternal and fetal placental blood flows, (iv) the maternal-fetal blood pH gradient, (v) the stage of placental development, which affects its thickness, area available for exchange and the degree of perfusion, (vi) possible placental metabolism (while drug metabolism has been demonstrated with in vitro placental tissue homogenates (Juchau, 1976) its significance in vivo is unknown), (vii) differences in the extent of maternal and fetal protein binding and (viii) the duration of exposure to the drug in the maternal circulation {viz., single, multiple or continuous dosing). 3. Fetal drug concentrations are affected by (i) the degree of distribution which is a function of the fetal circulation pattern, tissue affinity and protein binding, (ii) possible hepatic metabolism, ( i i i ) renal excretion into amniotic fluid and subsequent recirculation as a result of fetal swallowing, (iv) umbilical cord blood flow (e.g. possible compression during labour), (v) dilution of the drug in the fetal circulation resulting in a delay in equilibration between fetal tissues and blood. 9 1.4 Metoclopramide Pharmacokinetics There are few reports regarding the distribution of MCP into specific organ systems. In mice the highest concentrations were found in the intestinal mucosa, l iver, biliary tracts and salivary glands. Smaller amounts were present in the central nervous system, heart, thymus, suprarenal glands, and in fat and bone marrow. In the central nervous system MCP was localized in the area postrema, which contains the chemoreceptor trigger zone for vomiting in man (Harrington et al., 1983). Accumulation in lung tissue has also been reported in a study employing the isolated perfused rat lung (Okumura et al., 1978). Metoclopramide also readily enters breast milk, achieving concentrations similar to or higher than those in maternal plasma (Kauppila et al., 1983). Animal Studies - Studies with MCP in animal models (rabbit, rat, dog) have shown that the drug is well absorbed, extensively metabolized and rapidly excreted in the species studied (Bakke and Segura, 1976; Bateman et al., 1980; Tarn et al., 1981), with partial metabolism by sulphate and glucuronide conjugation (rabbit, dog) (Arita et al., 1970; Cowan et al., 1976; Bateman et a7., 1980), O-demethylation, N-de-ethylation and amide hydrolysis (rabbit, rat, dog) (Arita et al., 1970; Bakke and Segura, 1976; Cowan et al., 1976). Peak plasma concentrations were observed between 30-120 min after oral administration, suggesting rapid absorption from the gastrointestinal tract. The elimination half- l i fe (t^) in early reports averaged 20, 28, and 36 min in the rat, rabbit and dog, respectively, following i.v. injection (Bakke and Segura, 1976). In a more recent report (Bateman et al., 1980), an elimination ti, of -120 min has been observed in the dog 10 following oral or i.v. dosing. Similarly Tarn and Axelson, 1978, report a half- l i fe of 50 min in the rat with a dose dependent increase at doses in excess of 15 mg/kg (Tarn et al., 1981). Further studies in the rat have shown that MCP undergoes saturable first-pass metabolism at doses below 1 mg/kg (Kapil et al., 1982) and unusual dose-dependent kinetics at doses above 15 mg/kg, apparently as a result of altered hepatic blood flow (Tarn et al., 1981). Human studies - There has been considerable study of the pharmacokinetics and bioavailability of metoclopramide in normal, healthy volunteers. Metoclopramide is rapidly absorbed from the gastrointestinal tract, with peak plasma levels occurring =1 hour after oral administration, and undergoes significant first-pass metabolism. Reports of mean bioavailability range from =50-76% (Graffner et al., 1979; Bateman et al., 1980; Ross-Lee et al., 1981; Wright et al., 1988), with considerable intersubject variability (=30-100%) (Bateman, 1983). As expected for a basic 1ipid-soluble drug, MCP is widely distributed in man with a volume of distribution ranging from =2.2-4.0 L/kg (Bateman, 1983; Harrington et al., 1983; Wright et al., 1988). Metoclopramide is =40% bound to plasma proteins, primarily alphaj-acid glycoprotein (Webb et al., 1986), in both normal and uremic patients. Binding to albumin is limited (Webb et al., 1986) and this interaction has been reported to be weak (Pagnini and Di Carlo, 1972; Denisoff and Molle, 1978; Gourley et al., 1982). The elimination half- l i fe ranges from =2.6-6.3 hours (Bateman et al., 1983; Harrington et al., 1983), considerably longer than those reported for the rabbit, rat and dog as discussed above. Following i.v. administration, =80% of the dose is recovered in the urine in 24 hours (Teng et al., 1977), with =20% as unchanged drug and the balance as metabolites (Bateman et al., 1980; Bateman, 1983). The drug is metabolized by conjugation to inactive 11 N^-sulfate («30-40% of an i.v. or oral dose) and N^-glucuronide (<5%) metabolites (Teng et al., 1977; Bateman et al., 1980). Renal clearance of MCP is low (~ 1.7-2.6 mL/min/kg), accounting for -20-30% of total body clearance (-6.2-11.61 mL/min/kg) (Bateman et al., 1983, Wright et al., 1988). The total body clearance of MCP approaches hepatic blood flow, suggesting that its clearance is probably limited by liver blood flow, rather than by hepatic metabolic capacity (Harrington et al., 1983; Hellstern et a7., 1987). The liver is thought to be the major site of metoclopramide elimination (Desmond and Watson, 1986). Claims of dose dependency have been made following doses ranging from 5-20 mg (Graffner et a7., 1979; Bateman, 1983), based upon increases in elimination half- l i fe with increasing dose. These reports appear to be artifactual and the result of inadequate assay sensitivity which failed to allow plasma sampling of sufficient duration to accurately estimate the true elimination half - l i fe . It has been shown that truncation of the sampling interval may lead to incorrect estimation of biological half- l i fe (Gibaldi and Weintraub, 1971). More recent reports have demonstrated linear kinetics in the range of 5-20 mg (Wright et a7., 1988) and 20-100 mg (Wright et a7., 1984). There are also several recent reports of linear kinetics with the high i.v. doses of MCP used to control nausea and vomiting in patients undergoing chemotherapy, for example Taylor and Bateman, 1983 (10-500 mg), Taylor et al., 1984 (3-7 mg/kg), Sailer et al., 1985 (7-14 mg/kg) and Havsteen et al., 1986 (5-10 mg/kg). Studies in patients with renal failure have shown that the total body clearance of MCP decreases ~2-4-fold (Bateman et al., 1981; Wright et a7., 1988a) with a similar increase in terminal elimination half - l i fe. No significant change in volume of distribution occurred compared to that in normal volunteers. These findings are unexpected since only a small 12 percentage (=20%) of an administered dose is excreted in urine as intact drug. This suggests that the renal elimination pathway for MCP would be relatively unimportant. A similar disproportionate increase in MCP tk and reduction in total body clearance has been observed in rats with experimental renal dysfunction (Tarn et al., 1981a), despite the fact that only about 20% of the administered MCP dose is excreted as intact drug in this species as well. It was postulated that significant extrahepatic metabolism or diminished hepatic metabolism secondary to renal failure might account for these observations (Tarn et al., 1981a). More recent studies, however, with in vitro l iver, kidney and lung tissue homogenates have ruled out extrahepatic metabolism (Kapil et al., 1984) and diminished hepatic metabolism secondary to renal failure seems more likely (Bateman et al., 1981, Kapil et al., 1984; Wright et al., 1988a). The presence of an unidentified substance in the plasma of uremic patients which may inhibit MCP's metabolism has been suggested (Wright et a l . , 1988a). Hemodialysis has been found to be ineffective in removing MCP from the body (Wright et al., 1988a). The bioavailability of orally administered MCP has been reported to be increased significantly in patients with cirrhosis of the liver compared to patients with normal liver function (Hellstern et al., 1987). These authors suggest that this increase is a consequence of intra- and extrahepatic shunting of blood away from hepatocytes, resulting in higher fractions of unmetabolized drug reaching the systemic circulation. 1.5 Assay Methods A number of analytical methods have been developed over the past 20-25 years for the measurement of MCP in biological fluids obtained from humans and animals, including thin layer chromatography (TLC), gas-liquid 13 chromatography (GLC), GLC-mass spectrometry and high pressure liquid chromatography (HPLC). Early methods employing TLC (Arita et al., 1970; Bakke and Segura, 1976; Huizing et al., 1979) lack specificity, so interference from structurally related metabolites limits their use. Although a lower limit of sensitivity of =20 ng/mL is possible (Huizing et al., 1979) this is inadequate for good pharmacokinetic studies. Extraction of fluid volumes ranging from 1-5 mL are also required making these methods unsuitable for experiments in small animals such as fetal sheep. Many HPLC assay methods have been developed (Teng et al., 1977; Graffner et al., 1979; Bateman et al., 1981; Taylor et al., 1984; Havsteen et al., 1986; Slordal et al., 1986; Beckett et al., 1987; De Jong et al., 1987; Takahashi et al., 1987), and while they offer ease of sample preparation compared to those for GLC, sensitivity remains a problem. Several provide sensitivity in the desired range of 5-10 ng/mL (Teng et al., 1977; Graffner et al., 1979; Bateman et al., 1981; Taylor et al., 1984; Beckett et al., 1987), but this requires the extraction of 2-5 mL of plasma, precluding their use in investigations involving extensive serial sampling in small animals or infants. The remaining HPLC methods require the extraction of only 1 mL of plasma and provide sensitivity limits ranging from 2-10 ng/mL using reconstitution volumes of between 40-500 ill and injections volumes of 20-500 /zL. Even a 1 mL plasma volume, however, is s t i l l a limiting factor for detailed pharmacokinetic studies in small animals or infants. The HPLC method developed by Slordal et al., 1986, requires the extraction of only 0.2 mL of serum, but the lower limit of detection is 50 ng/mL (50 p,l injection volume). While this level of sensitivity would be adequate for monitoring serum concentrations during high dose MCP therapy [e.g. in chemotherapy) it is not sufficient for most single dose pharmacokinetic studies. The selected ion monitoring GLC-mass spectrometric method 14 developed by Bateman et al., 1978 provides a sensitivity limit of 15 ng/mL and has been used to study the pharmacokinetics of MCP in humans and dogs. No details of the sample volumes extracted were provided. Several packed column GLC assay techniques employing electron capture detection (ECD) have also been developed for pharmacokinetic studies in rats (Tarn and Axelson, 1978) and in humans (Tarn et al., 1979; Ross-Lee et al., 1980). The methods for human plasma analysis all provide good sensitivity allowing quantitation of MCP to a concentration of 5 ng/mL following the extraction of 0.5-1.0 mL of plasma. The technique reported by Tarn and Axelson, 1978, is sufficiently sensitive to allow serial blood sampling (0.1-0.2 mL) for kinetic studies in rats receiving i.v. doses of 5-25 mg/kg (Tarn and Axelson, 1978; Tarn et a7., 1981). Although sensitive, these packed column methods demonstrate considerable potential for interference from endogenous substances when used for trace level analysis as well as from other drugs in clinical studies. More recently, a GLC-ECD method employing highly selective fused si l ica capillary columns and split sample injection has been reported (Riggs et a7., 1983). This technique allows quantitation to a lower limit of 4 ng/mL following the extraction of small plasma volumes (human, sheep) of 0.25-0.5 mL. A packed column GLC assay employing nitrogen phosphorus detection has also been developed to study the kinetics of MCP in cancer patients receiving high dose infusions (Sailer et al., 1985). Plasma volumes of 1 mL are required again, making this technique unsuitable for studies requiring extensive serial blood sampling in infants or small animals such as fetal sheep. A radioimmunoassay with a lower limit of sensitivity of 1 ng/mL has also been developed for MCP concentration determinations in serum (De Vil l iers et a7 . , 1987). While sensitive, the potential for cross-reactivity with MCP metabolites exists 15 although the authors found negligible cross-reactivity with the structurally related drug lidocaine. In summary, while GLC assay methods require considerably more time for sample preparation compared to HPLC techniques, those employing electron-capture detection st i l l provide the best sensitivity with small sample volumes. This makes them the most suitable for pharmacokinetic studies in those instances where only small biological fluid volumes are available. 1.6 Rationale Experimental access to the piacental-fetal unit in humans is restricted for technical and ethical reasons. Because of this, most of the available data on human placental drug transfer is based on single-point plasma concentration determinations. Paired maternal and "fetal" (umbilical cord) blood samples are drawn at birth and from a large number of different patients, maternal and infant concentration versus time profiles are constructed. However, the considerable intersubject variability as well as the artefacts introduced by data averaging severely limit the pharmacokinetic interpretations which can be drawn from such composite profiles (Krauer and Krauer, 1977; Levy, 1981; Waddel and Marlowe, 1981). A number of animals have been used to study the placental transfer of biochemical compounds and/or drugs, including rats, guinea pigs, pigs, goats, monkeys and sheep. Experiments involving serial blood sampling are, however, limited to larger animals such as monkeys, goats and sheep. The size of these animals allows chronic implantation of catheters into various fetal and maternal blood vessels and the attachment of monitoring devices 16 to measure various physiological parameters such as fetal heart rate, blood flow and pressure, electrocortical and muscle activities. Sheep are the most commonly used chronic preparation (model) for investigations into fetal biochemistry, and physiology (Comline and Silver, 1974; Szeto et al., 1978; Van Petten et al., 1978). They are of moderate size, are relatively easy to handle surgically {viz., fetal size), are docile and generally stand quietly during blood sampling and monitoring. Although used largely for studies of in utero fetal physiology (e.g., oxygen consumption, glucose metabolism) they have been recognized, and are now commonly used, as a model for pharmacological investigations (Van Petten et al. 1978). A few examples of drugs for which pharmacokinetics and/or fetal effects have been studied in this model include: lidocaine (Morishima et al., 1979; Bloedow et al., 1980), methadone (Szeto, 1983), cimetidine (Ching et al., 1985), ethanol (Patrick et al., 1985), acetaminophen (Wang et al., 1986), and diphenhydramine (Yoo et al., 1986). There are clear differences in placental structure and permeability between sheep and man, but for lipophilic molecules these differences are of less importance than for hydrophilic molecules. As a model, sheep provide valuable information on the pharmacokinetics of drugs in the ewe as well as in the fetus in utero, but with all animal models, extrapolations to the human situation must be made cautiously. As previously mentioned (Section 1.2), the placental transfer of MCP has been reported in humans with no apparent adverse effects on the neonate. There is , however, controversy that the neurobehavioural tests (Neurologic and Adaptive Capacity Score; Amiel-Tison et al., 1982) used to assess the infant at birth may not be sensitive to the subtle effects of drugs (Tronick, 1982). By monitoring fetal electrocortical, electroocular and breathing-like activities, a more precise assessment of fetal 17 behavioural effects can be obtained. Hence, the chronically catheterized pregnant sheep with implanted fetal electrocortical and electroocular electrodes would seem to be an acceptable model to monitor the fetal behavioural state, and may provide data relevant to human pregnancy. While the fetal lamb is more mature at birth than is the human infant, both species show similar fetal behavioural states (breathing-like movements, limb and body movements, patterns of high and low voltage electrocortical activity, eye movements) towards the end of gestation (de Vries et al., 1982, 1985; Nijhuis et al., 1982) 1.7 Objectives 1. To modify the existing fused si l ica capillary gas-liquid electron-capture chromatographic assay to permit automatic sample injection and to apply the assay to MCP measurement in maternal and fetal plasma, amniotic and tracheal fluids. 2. To study the kinetics of MCP placental transfer in the pregnant ewe and fetus following maternal i.v. bolus dosing. 3. To examine the pharmacokinetics of MCP in the ewe and fetus following separate maternal and fetal infusions to steady-state. 4. To study MCP dose linearity in pregnant and nonpregnant sheep over a 4- and 8-fold dose range, respectively, and to compare pharmacokinetic parameters between these two groups of animals. 5. To assess the placental and nonplacental clearances of MCP following separate maternal and fetal drug infusions to steady-state using a two compartment open model. 6. To examine the pharmacokinetics of MCP in the ewe and fetus following intra-amniotic drug injection. 18 7. To assess the physiological effects of MCP on fetal heart rate, blood pressure, arterial pH and blood gases (P02, PCO2, C^-content). 8. To examine the effect of MCP on fetal behavioural states by monitoring fetal breathing-like movements and electrocortical activity. 19 2. EXPERIMENTAL 2.1 Materials and Supplies 2.1.1 Chemicals Metoclopramide (4-amino-5-chloro-2-methoxy-N-(2-diethyl aminoethyl) benzamide monohydrochloride monohydrate) (MCP.HCL.h^O; Lot No. F058), deethylated metoclopramide (4-amino-5-chloro-2-methoxy-N-(aminoethyl) benzamide and metoclopramide monohydrochloride (MCP.HCL) 5mg/mL injectable (ReglanR; Lot No. 84637) (A.H. Robins Canada Inc. Montreal Que.); maprotiline (N-methyl-9,10-ethanoanthracene-9(10H) propanamide monohydrochloride) (MAP.HCL; Lot Al1663096472-0) (Ciba Pharmaceuticals, Mississauga, Ont.); thiopental sodium 1 g/vial (Pentothal^; Abbott Laboratories, Montreal Que.); halothane (Fluothane^; Ayerst Laboratories, Montreal, Que.); ampicillin injectable 250 mg/vial (Penbritin^ Ayerst Laboratories, Montreal, Que.); atropine sulfate injection 0.6 mg/mL (Glaxo Laboratories, Montreal, Que.); heparin 1000 units/mL (Organon Canada Ltd., West H i l l , Ont.). 2.1.2 Reagents ACS reagent grade sodium hydroxide (Fisher Scientific Co., Fair Lawn, NJ, U.S.A.); ACS reagent grade hydrochloric acid 37% (American Scientific and Chemical, Seattle, WA, U.S.A.); ammonia strong solution 27% (Mallinckrodt Inc., St. Louis, MI, U.S.A.); heptafluorobutyric anhydride and triethylamine sequanal grade (Pierce Chemical Co., Rockford, IL, U.S.A.). ACS reagent grade sodium chloride, potassium dihydrogen orthophosphate (monobasic) and disodium hydrogen orthophosphate (dibasic) (BDH Chemicals, Toronto, Ont.). 2.1.3 Solvents 20 Benzene and toluene (distilled in glass) (Caledon Laboratories Inc., Georgetown, Ont.); deionized water was produced in the laboratory using a Milli-RO Water System (Mi 11ipore Corp., Bedford, MA., U.S.A.). 2.1.4 Gases Nitrogen U.S.P. (Union Carbide Canada Ltd., Toronto, Ont.); ultra high purity (UHP) hydrogen and argon/methane (95:5) (Matheson Gas Products Canada Ltd., Edmonton, Alta.). 2.1.5 Supplies for Sheep Studies Needles and plastic disposable Luer-lok^ syringes for drug administration and sample collection (Becton-Dickinson Canada, Mississauga, Ont.); heparinized Vacutainer^ tubes (Vacutainer Systems, Rutherford, NJ, U.S.A.); 15 mL Pyrex^ disposable glass culture tubes (Corning Glass Works, Corning, NY, U.S.A.); polytetrafluoroethylene (PTFE) lined screw caps (Canlab, Vancouver, B.C.); silicone rubber tubing for catheter preparation (Dow Corning, Midland, MI, U.S.A.); PTFE-coated stainless steel wire for electrode preparation (Cooper Corp., Chatsworth, CA, U.S.A.); disposable Centrifree1^ Micropartition System with YMT ultrafiltration membrane (Amicon Division, W.R. Grace and Co., Danvers, MA, U.S.A.) for metoclopramide plasma protein binding determinations. 2.2 Equipment A model 5840A Hewlett-Packard gas chromatograph equipped with a 5 3 Ni electron capture detector (ECD), a packed column compatible model 18835B capillary inlet system and a model 18850A integrator terminal (Hewlett-Packard Co., Avondale, PA, U.S.A.); 25 m x 0.31 mm I.D. cross-linked 5 % phenylmethylsilicone (Ultra 2; 0.52 nm film thickness) fused si l ica capillary column (Hewlett-Packard Co., Palo Alto, CA, U.S.A.); vortex-type mixer (Vortex-Genie^, Fisher Scientific Industries, Springfield, MA, U.S.A.); incubation oven (Isotemp^ model 350, Fisher Scientific Industries, Springfield MA, U.S.A.); IEC model 2K centrifuge (Damon/IEC Division, Needham Hts., MA, U.S.A.); rotating-type tube mixer (Labquake^ model 415-110, Lab Industries, Berkley, CA, U.S.A.); infusion pump (Harvard model 944, Harvard Apparatus, Mill is , MA, U.S.A.); model J2-21 refrigerated centrifuge with JA-20 fixed angle (35°) rotor (Beckman, Schiller Park, IL, U.S.A.). 2.3 Stock and Reagent Solutions Metoclopramide.HCL.H2O was accurately weighed and dissolved in deionized water using serial dilution to a final concentration of =0.04 /xg/mL (=11.82 mg of MCP.HCL.H20 is equivalent to =10 mg of MCP free base). The internal standard, maprotiline.HCL, was accurately weighed and dissolved in deionized water using serial dilution to a final concentration of =0.4 /zg/mL (=11.32 mg of MAP.HCL is equivalent to =10 mg of MAP free base). Metoclopramide and maprotiline stock solutions were stored at 4 °C for up to two months. Sodium hydroxide (NaOH) IN and 5N solutions were prepared by dissolving NaOH pellets in deionized water. Hydrochloric acid 5N was prepared by diluting ACS reagent grade concentrated hydrochloric acid (37%) in deionized water. Ammonium hydroxide 4% was prepared by diluting Ammonia Solution Strong (27%) in deionized water. Triethylamine 0.0125M was prepared by diluting triethylamine with toluene. Four or five pellets of NaOH were added to the solution. 22 Isotonic, pH 7.4 phosphate buffer was prepared by accurately weighing and dissolving 2.28 g of potassium dihydrogen orthophosphate (KH2PO4) in deionized water to a volume of 250 mL (Solution A) and 9.47 g of disodium hydrogen orthophosphate (Na2HP04) to a volume of 1000 mL (Solution B). The final buffer solution was prepared by combining solutions A and B. To this, 5.5 g of sodium chloride was added for isotonicity adjustment. If required, pH was adjusted to 7.4 by the dropwise addition of 5N NaOH or phosphoric acid. 2.4 Sample Extraction and Derivative Formation The procedure for sample extraction and derivative formation remains the same as that published by Riggs et al., 1983 with the exception of the final reconstitution volume which has been increased from 0.2 mL to 0.8 mL. The method is outlined in Scheme 1. All glassware used in the preparation of stock solutions and for extractions was prewashed with detergent in an automatic dishwasher and then soaked in chromic acid for a minimum of six hours. Thereafter it was thoroughly rinsed with tap water for =5 hours using a mechanical rinser, and finally with deionized water before being dried. Vials for use in the autosampler (1 mL) were boiled in a 5% aqueous detergent solution (Extran^ 300, BDH Chemicals, Toronto, Ont.) for =30-45 min, thoroughly rinsed with tap water and soaked for several hours in deionized water prior to being dried. All extractions were carried out in 15 mL glass culture tubes equipped with PTFE-lined screw caps. Plasma (0.05-0.5 mL), tracheal (0.01-0.2 mL) or amniotic fluid (0.05-0.5 mL) obtained from MCP treated ewes or fetal lambs was added to clean tubes containing 0.5 mL IN NaOH (pH =14) and 0.2 mL of MAP internal standard solution (=0.4 //g/mL). The aqueous phase was then adjusted to a final volume of 2.2 mL with deionized water. Six mL 23 Blank plasma, tracheal or amniotic fluid spiked with MCP standard Discard aqueous layer Discard organic layer' Discard organic wash' Discard aqueous layer4 Plasma, tracheal or amniotic fluid (0.01-0.5 mL) IN NaOH, 0.5 mL MAP, 0.2 mL Water q.s. 2.2 mL Benzene, 6 mL Extract, 20 min Centrifuge, 2 x 5 min Organic layer IN HCL, 2 mL Extract, 20 min Centrifuge, 10 min Aqueous layer Wash with benzene ( 2 x 4 mL) 5N NaOH, 0.5 mL Benzene, 6 mL Extract, 20 min Centrifuge, 10 min Organic layer Evaporate to dryness under N2 at 40° C Dried extract Discard aqueous layer 0.0125M TEA in toluene, 0.8 mL HFBA, 20 nl 1 hour at 55° C Cool to room temperature Water, 0.5 mL, vortex 10 sec 4% NH^ OH, 0.5 mL, vortex 10 sec Centrifuge, 1 min Inject into GC (2 /iL) Scheme 1. Metoclopramide extraction procedure. 24 of benzene was added, the tubes capped, and metoclopramide and maprotiline extracted into the organic layer by shaking for 20 min on a rotary shaker (LabquakeR model 415-110, Lab Industries, Berkley, CA, U.S.A.)- Following centrifugation at 2300 g for 5 min the tubes were removed and shaken lightly to break any emulsion that may have formed during the extraction process. Following a further 5 min centrifugation the organic layer was transferred to clean tubes containing 2 mL of IN HCL using a disposable Pasteur pipet, and metoclopramide and maprotiline back-extracted by shaking for 20 min. The samples were centrifuged at 2300 g for 5 min and the organic layer subsequently aspirated and discarded (water vacuum aspirator). The remaining aqueous layer was washed with two 4 mL aliquots of benzene which were also aspirated and discarded. The aqueous layer was then alkalinized by adding 0.5 mL of 5N NaOH (pH =14) and the samples re-extracted for 20 min following the addition of 5 mL of benzene. Following centrifugation at 2300 g for 5 min, the organic layer was transferred to a clean, dry tube (using a disposable Pasteur pipet) and taken to dryness under a gentle stream of nitrogen in a 40 °C water bath. The nitrogen dried residue was reconstituted with 0.8 mL of 0.0125M triethylamine in toluene. A 20 fil volume of heptafluorobutyric anhydride (HFBA) was added, the tube capped, the sample vortex-mixed (Vortex-Genie^, Fisher Scientific Industries, Springfield, MA, U.S.A.) and placed in an oven at 55 °C for 60 min. After cooling to room temperature, the excess derivatizing agent (HFBA) was removed by hydrolysis with 0.5 mL of deionized water (vortex-mix for 10 sec) and the generated HFB-acid neutralized with 0.5 mL of 4% ammonium hydroxide (vortex-mix for 10 sec). Following centrifugation at 2300 g for 1 min, the organic layer was immediately transferred to a clean autosampler vial which was capped with a PTFE-lined seal. The samples were generally stored at -20 "C overnight, prior to injection, as this process 25 removed some endogenous peaks resulting in cleaner chromatograms. Volumes of 2 juL were then automatically injected into the gas chromatograph for GLC-ECD analysis. 2.5 Standard Curve Preparation Volumes of 0.05, 0.1, 0.2, 0.4, 0.6, 0.8 and 1.0 mL of MCP.HCL stock solution (=0.04 jug/mL) were pipetted into 15 mL Pyrex^ culture tubes containing 0.5 mL IN NaOH, 0.2 mL of internal standard, MAP.HCL (0.4 /jg/mL), and, depending on the samples to be analyzed, 0.5 mL of blank sheep plasma, 0.2 mL of blank tracheal fluid or 0.5 mL of blank amniotic fluid. The aqueous phase was adjusted to a final volume of 2.2 mL with deionized water and the standards were then extracted and derivatized as described in Section 2.4. A calibration (standard) curve for MCP was constructed by plotting the area ratios of the heptafluorobutyryl (HFB) derivatives of MCP/MAP against the known concentrations of MCP free base. 2.6 GLC The assay method used for MCP concentration determinations in plasma, amniotic and tracheal fluids is based on the electron-capture detection capillary GLC technique published by Riggs et al., 1983. The previous assay has been modified by switching from the split to splitless mode of sample introduction to allow automatic sample injection. The same type of capillary column and stationary phase are being used, however columns with a film thickness of 0.52 nm are currently employed as those with a 0.15 [im film thickness became unavailable, due to manufacturer quality control problems, during the course of this work. An open (unpacked) splitless fused si l ica injection port insert (liner) is currently used, replacing the Jennings^ tube or packed si l ica insert employed in the previous assay. 26 2.6.1 General Chromatographic Conditions for the Splitless Mode of Sample Introduction The operating conditions for routine plasma, amniotic or tracheal fluid sample analysis were: injection temperature 260 °C; initial column temperature 205 °C for 0.81 min, followed by an increase at the rate of 4 °C/min to a temperature of 240 °C for 3.5 min, followed by a second increase at 15 °C/min to a final temperature of 255 °C for 1 min; detector (ECD) temperature 350 °C; carrier gas (Hydrogen UHP) flow =2.3 mL/min; inlet pressure =10.5 p . s . i . ; inlet flow rate 30 mL/min; septum purge flow rate 1.5 mL/min; make up gas (Argon/Methane, 95:5) flow rate 60 mL/min; purge valve activated at 0.1 min, autosampler cycling time 1.25 min with sample injected at 1.25 min, purge valve closed at 1.75 min to yield an inlet purge time of 30 sec following sample injection; chart speed 0.4 cm/min; slope sensitivity 0.15-0.4; attenuation 2 .^ 2.6.2 Optimization of GLC Conditions for the Splitless Injection Mode Sample reconstitution volumes of 0.2, 0.4, 0.6 and 0.8 mL were examined for the presence of interfering endogenous sample components. The following instrument parameters were varied and tested for optimal separation and quantitation of MCP: 1. Inlet purge activation time: 5, 10, 20, 30, 40, 50 and 60 sec. 2. Injection port temperature: 220, 230, 240, 250, 260 and 270 °C. 3. Initial column temperature: 180, 190, 195, 200, 205 and 210 °C. 4. Column temperature programming rate: 2, 4, 5, 6, 7, 8, 10, 12 and 15 °C. 27 2.7 Pharmacokinetic Studies in Pregnant and Nonpregnant Sheep 2.7.1 Animal Handling Pregnant and nonpregnant ewes (Dorset, Suffolk, or crossbred) were brought into the unit =1 week prior to surgery to allow them to become accustomed to the environment. Ewes were housed singly or in groups of 2 or 3 in large adjacent pens. The-animals were provided a standard diet and water ad libitum. 2.7.2 Surgical Preparation Twenty two time-dated pregnant ewes underwent surgery at 115-130 days gestation (term =145 days). The ewe was fasted overnight and received a 3 mg i.v. bolus injection of atropine sulfate (Glaxo Laboratories, Montreal, Que.) =15 min prior to induction of anaesthesia with pentothal (15 mg/kg, i.v. bolus; Abbott Laboratories, Montreal, Que.). Following tracheal intubation, anaesthesia was maintained with 1.0-2.5% halothane (Ayerst Laboratories, Montreal, Que.) in 50% oxygen. Aseptic techniques were employed throughout surgery. The maternal abdomen was opened by a midline incision and access gained to the fetal hindlimbs, and head and neck, through two small incisions in the uterine wall in areas free from placentomes and major blood vessels. Sterile silicone rubber catheters (Dow Corning, Midland MI, U.S.A.), f i l led with heparinized 0.9% saline (12 units/mL), were implanted into the fetal femoral artery, lateral tarsal vein, trachea, and amniotic cavity. In four preparations a second amniotic catheter was placed into the amniotic cavity. Both amniotic catheters were sutured to the fetal skin, one in the 28 area of the neck and the other to a hindlimb. In three fetuses a catheter was also implanted in the common umbilical vein. The tracheal catheter (2.2 mm o.d.) was inserted through a small incision between adjacent tracheal rings, 1-2 cm below the larynx, and advanced 4-5 cm; it did not impair the free flow of fluid out of the lung via the upper airway. In ten fetuses electrodes of PTFE-coated stainless steel wire (Cooper Corp., Chatsworth, CA, U.S.A.) were implanted biparietally on the dura to record the fetal electroencephalograph and through the orbital ridge of the zygomatic bone of each eye for electroocular recordings. Following catheterization of the fetus, the uterine incisions were closed in two layers. Maternal catheters were then implanted in a jugular or femoral vein and a femoral artery. All catheters were tunneled subcutaneously, and exteriorized through a small incision in the flank of the ewe where they were stored in a cloth pouch on the ewe's flank when not in use. The maternal abdominal incision was closed in layers. Amniotic fluid lost during surgery was replaced with sterile irrigation saline via the amniotic catheter. Immediately following surgery ampicillin (500 mg; Ayerst Laboratories, Montreal, Que.) was administered to the fetus via the tarsal venous catheter, intramuscularly to the ewe, and into the amniotic fluid via the amniotic catheter. Ampicillin was also administered prophylactically to the ewe for 3 days following surgery and daily into the amniotic fluid for the duration of the preparation. All vascular catheters were kept patent by flushing daily with 2 mL of sterile 0.9% saline containing 12 units of heparin/mL. Five nonpregnant ewes also underwent general anaesthesia and aseptic surgery for the implantation of catheters in a jugular or femoral vein and a femoral artery. 29 Following surgery, the ewes were kept in holding pens with other sheep and given free access to food and water. All animals were allowed to recover for a minimum of 3 days prior to experimentation. On experimental days the ewe was placed in a monitoring cart adjacent to the holding pen in full view of companion ewes and with free access to food and water. 2.7.3 Experimental Protocol For each experiment, a volume of blood, amniotic or tracheal fluid equal to three or four times the catheter dead volume was withdrawn prior to each sample. The samples were then withdrawn and the respective initial volume of blood, amniotic or tracheal fluid was reinfused. Finally, the catheters were f i l led with heparinized saline (12 units/mL). 2.7.3.1 Intravenous Bolus Studies Pregnant Ewes - Metoclopramide hydrochloride (Reglan^ injectable, 5mg/mL; A.H. Robins, Montreal, Que.) was diluted to 10 mL with 0.9% saline for injection and administered, as a bolus, via the maternal venous catheter over a period of 1 min. The maternal catheter was then flushed with 5 mL of heparinized 0.9% saline (12 units/mL). Five ewes (Nos. 56, 59, 68, 624, 712) received a 10, 20 and 40 mg dose on a crossover basis. At least 48 hours was allowed to elapse between successive experiments. Four additional ewes (Nos. 121, 127, 138, 140) received a single 40 mg i.v. bolus dose. Blood samples for metoclopramide determination were simultaneously collected from the maternal (2.5 mL) and fetal (1.5 mL) arterial catheters at 5 min before and 1, 3, 5, 10, 15, 20, 30, 45, 60, 90, 120, 150, 180, 210, 240, 270, 300, and 360 min following maternal dosing. Fetal arterial blood (1.0 mL) was also collected at -5, 1, 5, 10, and 30 30 min for fetal blood gas (Po2, Pco2) and pH measurements. Total fetal blood volume was replaced with an equal volume of maternal blood at intervals following each 10-12 mL cumulative blood sampling and at the end of the experiment. The collected blood was transferred to a heparinized Vacutainer^ (Becton Dickinson, Rutherford, NJ, U.S.A.) and immediately centrifuged at 3500 g for 10 min. Amniotic and tracheal fluid samples (1.5 mL each) were also collected in the four ewes receiving the single 40 mg dose. Plasma, amniotic and tracheal fluid samples were stored at -20 °C in p Pyrex* disposable glass culture tubes (Corning Glass Works, Corning, NY, U.S.A.) with PTFE-lined screw caps until the time of assay. Gestational ages in the ewes at the time of the experiments averaged 125.0 ± 7.0, 127.8 ± 7.2, and 132.0 ± 6.5 days for the 10, 20 and 40-mg doses, respectively (range 117-141 days). Maternal body weight averaged 71.5 ± 7.2 kg (range 61.2-80.9 kg). Nonpregnant Ewes - Metoclopramide hydrochloride was administered to these animals under the same experimental conditions used for the pregnant animal studies. Doses of 10, 20, 40 and 80 mg were administered on a four-way crossover basis, with a minimum of 48 hours between successive experiments. The sampling times, sample volume, and plasma storage conditions were identical to those used for the pregnant ewes. Body weight in the nonpregnant ewes averaged 57.9 ± 8.3 kg (range 50.0-69.0 kg). 2.7.3.2 Intravenous Infusion to Steady-State The maternal volume of distribution (V a r e a ) and terminal elimination rate constant (/?) obtained from the i.v. bolus studies in the pregnant ewes were used to calculate the maternal i.v. bolus loading dose (LD) and infusion rate (ko) for the infusion experiments utilizing the following equations (Gibaldi and Perrier, 1975): 31 LD = C C C .V area (1) C s s = ko/(V area (2) where C s s equals the desired maternal steady-state concentration of the drug. A maternal steady-state MCP plasma concentration of 40 ng/mL was chosen. The fetal loading dose and infusion rate were one-third that used in the ewe. Paired maternal and fetal infusions were performed in nine chronically instrumented ewes. The maternal and fetal infusions were separated by 48-72 hours. Maternal MCP Infusions - A 15 mg i.v. bolus loading dose of metoclopramide hydrochloride (Reglan^ injectable, 5 mg/mL; A.H. Robins, Montreal, Que.), diluted to 5 mL with 0.9% saline for injection, was administered to the ewe via the maternal femoral venous catheter over 1 min. This was followed immediately by a continuous infusion at a rate of 0.21 mg/min (0.17 mL/min) for 90 min employing a Harvard^ infusion pump (Model 944, Harvard Apparatus, Mill is MA, U.S.A.). Samples for metoclopramide determination were simultaneously collected from the maternal (2.5 mL) and fetal (1.5 mL) arterial, amniotic (1.5 mL) and tracheal (1.5 mL) catheters at -5, 5, 15, 30, 45, 60, 75, 90, 95, 100, 105, 110, 120, 135, 150, 180, 210, 240, 270, 300, 330, and 390 min. In two experiments post-infusion sampling was continued for 28 hours with additional fluid samples collected at 6, 7, 8, 10, 12, 20, 22 and 28 hours. In one animal (Ewe 130), blood (1.5 mL) was also withdrawn from an umbilical venous catheter at -5, 5, 15, 30, 45, 60, 75 and 90 min. Fetal arterial blood (1.0 mL) was also collected at -5, 5 and 30 min for blood gas (P02, PCO2, 02-content), pH and packed cell volume measurements. As in the i.v. bolus experiments fetal blood volume was replaced with an equal volume of maternal blood at intervals during and at the end of the study. 32 The collected blood was transferred to a heparinized Vacutainer^ (Becton Dickinson, Rutherford, NJ, U.S.A.) and immediately centrifuged at 3500 g for 10 min. Plasma, amniotic and tracheal fluid samples were stored at -20 °C in Pyrex^ disposable glass culture tubes (Corning Glass Works, Corning, NY, U.S.A.) with PTFE-lined screw caps until analysis. Fetal MCP Infusions - A 5 mg i.v. bolus loading dose of metoclopramide hydrochloride (Reglan^ injectable, 5 mg/mL; A.H. Robins, Montreal, Que.), diluted to 3 mL with 0.9% saline for injection, was administered to the fetus via the tarsal venous catheter over 1 min. This was immediately followed by the initiation of a continuous infusion at a rate of 0.07 mg/min (0.17 mL/min) for 90 min. In one experiment, post-infusion sampling was continued for 12 hours and in another for 28 hours with additional samples collected at 6, 7, 8, 10, 12, 16, 20, 24 and 28 hours. With the exception of the post-infusion sampling times used in these two experiments the samples collected, sampling times and storage conditions were the same as those for the maternal infusion experiments. Umbilical venous blood was also collected in one of the fetal infusion studies (Ewe 130); sampling times were the same as those used for the paired maternal infusion experiment. Control Normal Saline Infusions - Sterile sodium chloride for injection was administered as a 10 mL i.v. bolus over 1 min to the ewe via the maternal femoral venous catheter. Immediately following the bolus, sodium chloride for injection was continuously infused at a rate of 0.17 mL/min for 90 min. Similarly, the fetus received a 5 mL i.v. bolus (over 1 min) and continuous infusion (0.17 mL/min for 90 min) of sodium chloride for injection via the tarsal venous catheter. Fetal arterial blood (1.0 mL) was collected at -5, 5 and 30 min for blood gas (P02, PCO2, 0 2 content), pH and packed cell volume measurements. 33 2.7.3.3 Intra-Amniotic Bolus Experiments A 10 mg dose of metoclopramide hydrochloride (Reglan^ injectable, 5 mg/mL; A.H. Robins, Montreal, Que.) was diluted to 10 mL with 0.9% saline for injection and administered, as a bolus, into one of the amniotic cavity catheters over a period of 1 min. In an attempt to optimize mixing, amniotic fluid was withdrawn and reinjected (using the s t i l l attached dosing syringe) four times. Following this the catheter was flushed with 10 mL of heparinized saline (12 units/mL), capped and stored in the cloth pouch on the ewe's flank. Samples for metoclopramide determination were simultaneously withdrawn from the maternal (2.5 mL) and fetal (1.5 mL) arterial, amniotic (1.5 mL) and tracheal (1.5 mL) catheters at -5, 5, 10, 15, 20, 30, 45, and 60 min, 1.5, 2, 2.5, 3, 3.5, 4, 5, 6, 8, 10, 12, 20, 24 and 28 h. In two of the animals (Ewes 237, 287) umbilical venous samples (1.5 mL) were also collected at these times. The amniotic fluid samples were withdrawn from the second catheter which had been implanted in the amniotic cavity and which had not been used for drug injection. Fetal arterial blood (1.0 mL) was also collected at -5, 30 and 60 min for blood gas (P02, Pco2> O2 content), pH and packed cell volume measurements. As in the i.v. dosing experiments fetal blood volume was replaced with an equal volume of maternal blood at intervals during and at the end of the study. Collected blood samples were transferred to heparinized Vacutainers^ (Becton Dickinson, Rutherford, NJ, U.S.A.) and immediately centrifuged at 3500 g for 10 min. Plasma, amniotic and tracheal fluid samples were stored at -20 °C in Pyrex^ disposable glass culture tubes (Corning Glass Works, Corning, NY, U.S.A.) with PTFE-lined screw caps until analysis. 2.8 Recording Procedures and Blood Gas Analysis 34 For 1 hour before, during and for 1 hour following drug or control sodium chloride administration, fetal arterial, tracheal, and amniotic pressures were continuously recorded on a Beckman R612 recorder (Beckman, Schiller Park, IL, U.S.A.) using strain-gauge manometers (Statham model P23Db, Gould Inc., Oxnard, CA, U.S.A). Fetal heart rate was measured from the arterial pulse pressure using a cardiotachometer (Model 9857, Sensormedics Corp., Anaheim, CA, U.S.A.). The recordings were analyzed to provide estimates of arterial and intrauterine pressures and heart rate at 1 min intervals. Mean fetal arterial pressure was calculated and corrected for changes in intrauterine pressure, measured from the amniotic catheter, by subtraction using the following equation: Mean Corrected Pressure = [(SP - DP)/3] + DP - AP (3) where SP, DP and AP, in mm Hg, represent systolic, diastolic and amniotic (intrauterine) pressure, respectively. Episodes of fetal breathing activity were identified from the tracheal pressure recordings. In those fetuses equipped with suitable electrodes, fetal electrocortical (ECoG) and electroocular (EOG) activity were also recorded simultaneously using a type 9805A AC/DC coupler (Sensormedics Corp., Anaheim, CA, U.S.A.) with the low and high frequency cutoff frequencies set to approximately 0.5 and 50 Hz, respectively. Recordings of fetal electrocortical activity were divided into periods of low, intermediate and high voltage amplitude, with voltage ranges of 25-50, 80-110 and 125-175 /zV, respectively. The normal recording speed was 0.1 mm/sec, but occasionally speeds of 5-10 mm/sec were employed to obtain qualitative information on the frequency characteristics of the ECoG activity. Episodes of rapid eye movements were identified from the electroocular recording. Fetal arterial P02, PCO2, and pH were measured with an IL micro 13 blood gas analyzer (Instrumentation Laboratory Corp., Lexington, MA, 35 U.S.A.) set at a temperature of 39.5 °C. Blood oxygen content (C^C) was measured using a Lex-G^-con-K oxygen analyzer (Lexington Instruments, Waltham, MA, U.S.A.). Packed cell volume (hematocrit) was determined using a micro-capillary centrifuge (International Equipment Co., Needham, MA, U.S.A.). 2.9 Plasma Protein Binding at Steady-State The protein binding of MCP in maternal and fetal plasma obtained from the maternal and fetal infusion experiments was determined using ultrafiltration. The degree of MCP binding to maternal and fetal plasma following both maternal and fetal infusions were separately measured. Equal volumes of the 60, 75 and 90 min plasma remaining from previous MCP assays were pooled and volumes of 0.5-1.0 mL transferred to the sample reservoir of the ultrafiltration device (Centrifree^ Micropartition System, Amicon Division, W.R. Grace and Co., Danvers, MA, U.S.A.). The sample reservoir was immediately capped to prevent sample evaporation and pH change due to loss of CO2. The filtration devices were placed in a fixed angle rotor (35°) and centrifuged (Model J2-21 refrigerated centrifuge with a JA-20 rotor, Beckman, Schiller Park, IL, U.S.A.)) at 1900 g for 30 min. The rotor had been equilibrated to 25 °C by a blank 30 min run prior to spinning the samples. Aliquots of plasma and ultrafiltrate were analyzed for total and unbound drug concentrations, respectively, using the extraction and derivative formation procedure in Section 2.4. Assays were carried out in duplicate immediately following ultrafiltration. The fraction of MCP bound to proteins (B) was calculated as follows: B(%) = (Cp - C u ) /C p x 100 (4) where Cp is the concentration of MCP in plasma (total drug) and C u is the concentration of MCP in the ultrafiltrate. 36 Adsorption of MCP to the f i l ter membrane was examined by the ultrafiltration of MCP stock solutions. Serial dilutions containing 5, 50 or 500 ng/mL of MCP equivalent to MCP free base were prepared by dissolving MCP.HCL.H2O in isotonic pH 7.4 phosphate buffer. Volumes of 1.0 mL were transferred to the reservoir of the ultrafiltration device and centrifuged in the same fashion as the plasma samples. Ultrafiltration was performed in triplicate at each concentration. Metoclopramide concentrations were determined (in duplicate) using 0.05-0.8 mL aliquots of the prepared stock solutions and their corresponding ultrafiltrate. The regular extraction and derivative formation procedure described in Section 2.4 was used. 2.10 Quantitative Plasma, Amniotic and Tracheal Fluid Analysis Duplicate plasma (0.05-0.5 mL), tracheal (0.01-200 mL) or amniotic fluid (0.05-0.5 mL) samples obtained from MCP treated ewes or fetal lambs were extracted and derivatized as described in section 2.4. Each sample was automatically injected (2 [il) into the gas chromatograph (GC) in duplicate. Quantitative estimation of MCP was made by fitting the area ratios of the HFB-derivatives of MCP/MAP to the standard curve regression line (area ratio of MCP/MAP versus MCP concentration equivalent to MCP free base). Standard curve samples were extracted, derivatized and chromatographed the same day as the experimental samples. 2.11 Data Analysis 2.11.1 Computer Fitting Intravenous Bolus Studies - The data obtained from analysis of the biological fluid samples were plotted manually to obtain initial kinetic parameter estimates. The concentration versus time data were then analyzed using the decision making computer program AUTOAN (Sedman and Wagner, 1976) 37 to predict an appropriate kinetic model and initial estimates of intercepts and distribution and elimination rate constants. The AUTOAN generated estimates were subsequently f i t to the selected pharmacokinetic model using the nonlinear least-squares computer program NONLIN (Metzler et a7., 1974) to provide the final estimates of intercepts and rate constants used in further calculations of pharmacokinetic parameters. Steady-State Infusion Experiments - Again, the data obtained from biological fluid analysis were plotted manually to obtain initial kinetic parameter estimates. A decision making program which would accommodate simultaneous injection of an initial i.v. bolus loading dose and initiation of a constant rate infusion was not available. The init ial estimates of intercepts, rate constants and volume terms from manual plotting were therefore used in the nonlinear least-squares computer program N0NLIN84 (Metzler and Weiner, 1984). Maternal MCP plasma data obtained from the maternal infusion experiments were tested for best f i t using both one and two compartment i.v. bolus plus infusion models written for us by Dr. D.L. Weiner of Statistical Consultants, Inc., Edgewood, KY, U.S.A., suppliers of the N0NLIN84 computer modeling package. Fetal MCP plasma concentration data obtained from the fetal infusions was treated in the same fashion. Based upon the goodness of f i t by an examination of the observed and computer calculated concentration versus time plots, correlation coefficient of observed versus calculated MCP concentrations, residuals and minimum sum of squared residuals, a one or two compartment model was selected. The resulting computer generated kinetic estimates (rate constants, half-lives, distribution volumes) were then used for further pharmacokinetic parameter calculations. Post-infusion fetal MCP plasma concentration data obtained following a maternal infusion were analyzed using the decision making program AUTOAN (Sedman and Wagner, 1976) to 38 predict the kinetic model and initial estimates of rate constants and intercepts. The AUTOAN generated estimates were subsequently f i t to the appropriate model in the N0NLIN84 library with the latter program supplying final estimates of rate constants and half-lives. Maternal MCP post-infusion concentration versus time data obtained following a fetal infusion were similarly treated. Estimates of an apparent terminal elimination half- l i fe (t^) for MCP in tracheal and amniotic fluids were calculated from the slope of the linear regression line for the terminal elimination phase of the drug concentration versus time plot after cessation of the infusion, using the following standard equations (Gibaldi and Perrier, 1975, 1982): Slope = -K/2.303 (5) tk = 0.693/K (6) where K is the apparent f irst order elimination rate constant. Selection of Weighting Factor - Data points for individual subjects were weighted for computer analysis using the squared reciprocal of drug concentration based on the following equation proposed by Albert et al., 1974: In a 2 = In a + n' ln~C (7) This equation describes the relationship between concentration and variance, and can be used to determine a suitable weighting factor for fitting individual subject concentrations to a pharmacokinetic model based on the mean plasma concentrations of the group of subjects to which the individuals belong. The term a 2 is the variance corresponding to the mean concentration, C, for a group of subjects at each sampling time following injection, and a and n' are constants. A plot of In o versus In C yields a straight line with the intercept In a and a slope n'. If n' = 1, then an appropriate weighting factor is the reciprocal of plasma drug 39 concentration; i f n' = 2, the squared reciprocal of the plasma drug concentration can be used. Using this equation, the slope of the linear regression line of In as a function of In C was determined for the MCP plasma concentrations obtained in each group of animals. 2.11.2 Calculation of Pharmacokinetic Parameters Except where indicated, formulae used in kinetic parameter calculations where obtained from Gibaldi and Perrier, 1975, 1982. Metoclopramide concentrations and pharmacokinetic data are all expressed in terms of metoclopramide free base. The areas under the concentration versus time curves in the ewe and fetus were calculated using the following equation: AUCQ = AUCQ + AUC" (8) where t represents the time the last sample was taken. The first term, t OO AUC0, was calculated using the linear trapezoidal rule and AUCt was determined from the relationship AUC" = Ct/0 (9) where C T is the concentration of the last sample and /J is the terminal elimination rate constant. The area under the f irst moment of the plasma drug concentration curve (AUMC™), that is , the area under the curve of the product of time and drug concentration C from time zero to infinity, was calculated as follows: AUMC™ = AUMCg + AUMC* (10) where t represents the time when the last sample was taken. The f irst term AUMCQ was calculated using the linear trapezoidal rule and the second term using the equation AUMC* = t(Ct//3) + Ct/02 (11) where C t is the concentration of drug in the last sample and ft is the terminal elimination rate constant. Total body (systemic) clearance (CLS) and apparent volume of distribution (V a r e a ) were calculated from CLS = dose/AUCg (12) Varea = dose/jSAUCg (13) For the infusion experiments, dose is the total of the initial i.v. bolus and the infused drug. In the infusion studies, CLS was also calculated according to: CLS = ko/C s s (14) where ko is the infusion rate and C s s the steady-state MCP plasma concentration. The volume of the central compartment (Vc) was determined in the i.v. bolus experiments using: V c = dose/(A + B) (15) where A and B are the zero time intercepts of the distribution and elimination phases for a two compartment model. In the infusion studies Vc was determined from (Gibaldi, 1969): Vc = kok21/[a/3(R + S)] (16) where ko is the infusion rate, k 2 j the intercompartmental transfer constant for drug transfer between the central and peripheral compartments, a and j8 the respective distribution and elimination rate constants and R and S the respective intercepts of the distribution and elimination phases describing a post-infusion biexponential equation. For the i.v. bolus experiments, volume of distribution at steady-state (V s s) was calculated using both compartmental: V s s = V c [ (k 2 1 + k 1 2 ) /k 2 1 ] (17) where k 2j and k 1 2 are the intercompartmental transfer rate constants between the central and peripheral compartments describing a two compartment open model, and noncompartmental methods: V s s = dose[(AUMC 0)/(AUC 0T 2] (18) V s s was also determined in the infusion studies employing equation 17 in the case of the compartmental method and the following equation (Perrier and Mayersohn, 1982) in the case of the noncompartmental calculation: V s s = (dose/AUCo)tb (19) where dose is the sum of the i.v. bolus loading dose and the amount of drug infused and t b is the mean transit (residence) time for a drug in the body and is calculated from the relationship: t b = (AUMCQ/AUCQ) - koT2/2(koT + i.v. bolus dose) (20) where ko is the infusion rate and T the infusion time. Placental (Transplacental) and Nonplacental Clearances - The placental and nonplacental clearances of metoclopramide from the ewe and fetus were calculated at steady-state employing the method of Szeto et al., 1982. The maternal-fetal unit is represented by a general two compartment open model. Solving the simultaneous rate equations describing the change in amount of drug with time in the fetal and maternal compartments at steady-state the following clearance equations are obtained: CLm m = ko/[Cm - C f (C m ' / (C f ' ) ] (21) C L f f = ko' / [C f ' - Cm ' (C f /(Cm)] (22) C Lmf = C L f f ( c f / ( c m) (23) C Lfm = C Lmm(cm'/(C f') (24) CLmo = CLmm " C Lmf ( 2 5) C L f 0 = C L f f - CL f m (26) where ko and ko' represent the infusion rate in the mother and fetus, respectively; CL m m , the total clearance of drug from the maternal compartment; CLff, the total clearance from the fetal compartment; CLmf, clearance of drug from the maternal compartment to the fetal compartment; CLfm, clearance of drug from the fetal compartment to the maternal compartment; CL m o , nonplacental clearance of drug from the maternal compartment; CLf 0 , nonplacental clearance of drug from the fetal compartment; Cm and Cf are the respective average steady-state drug concentrations in the mother and fetus during drug infusion into the ewe; C m ' and Cf' are the respective average steady-state concentrations in the ewe and fetus during infusion into the fetus. The steady-state metoclopramide concentrations used in the calculation of all clearance values are the average of the 60, 75 and 90 min infusion experiment samples. Clearance values for both total and unbound drug were calculated using the above equations. 2.11.3 Fetal arterial pressure, fetal heart rate, amniotic (intrauterine) pressure Fetal heart rate, fetal arterial pressure and intrauterine pressure were determined and plotted at 1 min intervals from polygraph recordings during a 60 min control, dosing and 60 min post dosing interval. For the maternal and fetal infusions, this 1 min data was further analyzed for statistical evaluation (ANOVA) by averaging over 30 min intervals during a 60 min control, 90 min infusion and 60 min post-infusion period. 2.11.4 Fetal Breathing-Like and Electrocortical Activities Fetal breathing-like activity was identified from the tracheal pressure recordings. Fetal breathing episodes are presented as the 43 percentage of breathing-like activity over the 60 min control and 60 min post dosing (bolus or infusion) time interval. For the infusion experiments the percentage over the 90 min infusion period was also calculated. Fetal electrocortical tracings were analyzed for periods of low and high voltage activity and each expressed as a percentage of the 60 min control, 90 min infusion and 60 min post-infusion time intervals. 2.11.5 Statistical Tests Statistical evaluations were performed on the various pharmacokinetic, physiologic and behavioral parameters using either the Student's t-test (paired or unpaired) or one-way ANOVA. The level of significance chosen was p < 0.05. Except where otherwise indicated, values in the text and tables are presented as the mean ± one standard deviation (SD). 44 3. RESULTS 3.1 GLC Assay Modifications The electron capture detection capillary GLC assay previously developed in our laboratory for the analysis of MCP (Riggs et al., 1983) was further modified for use in the current studies. The primary modification involved a change from the split to splitless mode of sample injection. Various operating parameters related to this injection technique were therefore optimized. An open (unpacked) fused si l ica injection port liner (insert) was used, replacing the JenningsR tube or glass wool packed si l ica inserts previously employed. Both the solvent effect and cold trapping methods of reconcentrating MCP and the internal standard, MAP, at the head of the column were examined. Initial column temperatures 10-30 °C below the boiling point of the solvent (toluene, b.p. 110 °C) are recommended in order to achieve a good solvent effect (Freeman, 1981). Initial column temperatures ranging from 80-100 °C were tested, but resulted in very complex chromatograms with incomplete separation of both MCP and MAP from endogenous sample components. Column bleed was also a problem, resulting in considerable baseline drift and poor peak integration. For cold trapping, an initial column temperature low enough to allow condensation of the solute(s) at the head of the column is used. Initial column temperatures from 180-210 °C were examined. Temperatures below 195 °C resulted in the cold trapping of considerable endogenous sample components in addition to MCP and MAP, resulting in complex chromatography and interference. Temperatures in excess of 205 °C resulted in a decrease in both MCP and MAP area counts and peak heights. An initial column temperature of 205 °C was adopted as it provided good solute recoveries and the trapping of minimal endogenous sample components. 45 The effect of injection port temperature on MCP and MAP response was also studied, with temperatures ranging from 220-270 °C being examined. Increasing area counts for MCP and MAP were observed up to 260 °C and decreased again once this temperature was exceeded. An injection port temperature of 260 °C was selected as optimal, while 220 °C had been used in the previous assay. Column temperature programming rates ranging from 2-15 °C/min were also evaluated. A rate of 4 °C/min to a maximum column temperature of 240 °C resulted in the complete resolution of MCP and MAP from each other as well as from endogenous sample components. A relatively short run time (=13 min) was also obtained. A second column temperature increase at a rate of 15 °C/min for 1 min was added at the end of the run to purge nonvolatile components from the column following each sample injection. In the splitless configuration, excess solvent must be removed (purged) from the inlet in order to minimize solvent tai l ing. The time the inlet purge is activated can also affect the quantitative transfer of solutes onto the column. Purge activation times ranging from 5-60 sec were examined and are presented in Fig. 1. No significant increase in MCP or MAP recovery (measured in terms of absolute area counts) occurred after 30 sec, while some sample loss resulted with activation times less than 20 sec. Solvent tailing was also minimal (Fig. 2). Based on these findings an inlet purge activation time of 30 sec was chosen for subsequent assays. A significant increase in sensitivity accompanied the switch from the split to splitless mode of sample injection necessitating an increase in the final reconstitution volume. Final reconstitution volumes of 0.2, 0.4, 0.6 and 0.8 mL were examined. Interfering component peaks were frequently observed with sample extracts reconstituted to volumes of 0.2 and 0.4 mL and in some instances with a volume of 0.6 mL, particularly for plasma to -\-> c 13 o o o (D < 9000-8000-7000-6000-5000 -4000-3000 -2000-1000-0 -0 10 20 T -A-i 30 40 Purge activation time (sec) 50 j MAP 60 F i g . 1. MCP and MAP recovery as a fun c t i o n of purge a c t i v a t i o n time. T r i p l i c a t e 2 pi i n j e c t i o n s o f separate samples co n t a i n i n g MCP 16 ng/mL or MAP 80 ng/mL. MCP -Fa CTl MA 1 vJUL FA Www J TR 1 MCP 2 MAP Fig. 2. Representative chromatograms of the heptafluorobutyryl derivatives of MCP (1) and the internal standard, MAP (2) in maternal arterial plasma (MA), fetal arterial plasma (FA), tracheal fluid (TR) and amniotic fluid (AMN). Samples were taken at 45 min into a maternal steady-state infusion and are presented with their superimposed control (blank). extracts. A final reconstitution volume of 0.8 mL was selected for the assay as it resulted in no interference from extraneous compounds and st i l l provided good sensitivity. The increased sensitivity gained through use of the splitless mode allowed the addition of a 2 ng/mL concentration point to the standard curve, down from the previous low of 4 ng/mL. Other than an increase in the final reconstitution volume the sample extraction and derivative formation procedure remains the same as in the previous assay. The final difference between the previous assay and the current method involves the use of columns with a film thickness of 0.52 nm. This change was necessary as columns with a phase thickness of 0.15 fim became unavailable due to manufacturer quality control problems early in the course of this project. The column phase, internal diameter and length remain unchanged. No change in chromatography other than an increase in retention time occurred. A comparison of the differences and similarities in GC operating conditions for the split and splitless modes of sample injection is presented in Table 1. In addition to providing an increase in sensitivity, the splitless mode of sample introduction was found to be compatible with automatic sample injection providing good injection to injection and sample to sample reproducibility. Representative chromatograms of fetal and maternal arterial plasma, tracheal and amniotic fluids obtained from MCP dosed ewes along with their superimposed controls using the optimized purge activation time (30 sec), injection port temperature (260 °C), initial column temperature (205 °C), column programming temperature (4 °C/min) and final reconstitution volume (0.8 mL) are presented in Fig. 2. Retention times for MCP and MAP are =8 and =10 min, respectively. No interference from endogenous components was observed. Calibration (standard) curves were obtained by analyzing blank maternal or fetal arterial plasma, amniotic or 49 Table 1. Comparison of GC operating conditions for the split and splitless modes of sample injection. Parameter Split Splitless Injection temperature 220 °C 260 °C Detector temperature 350 °C same Inlet pressure 10.5 p .s . i . same Inlet flow rate 30 mL/min same Column flow rate =1 mL/min =2 mL/min (h^, carrier gas) Septum purge flow rate 1.5 mL/min same Make-up gas flow rate 60 mL/min same (Argon/methane, 95:5) Purge activation time n/a 30 sec Column (oven) temperature Isothermal, 235 °C Initial 205 °C for 0.81 min, then 4 °C/min to 240 °C for 4 min, then 15 °C/min to 255 °C for 1 min. Column (5% phenylmethyl- 25m x 0.31mm i.d. same silicone) -film thickness 0.15 nm 0.52 nm Run time =6 min =13 min MCP retention time =3 min =8 min MAP retention time =5 min =11 min Injection volume 2 nl same Automatic sample injection no yes 50 tracheal fluids spiked with varying amounts of MCP (2, 4, 8, 16, 24, 32 and 40 ng/mL) and plotting the area ratio of the HFB derivatives of MCP/MAP against the indicated MCP concentrations. The data for a representative standard curve (in plasma) obtained during a coefficient of variation study are presented in Fig. 3. The line of best f i t through the data points was obtained from linear regression analysis. A linear electron capture detector response was observed over the 2-40 ng/mL concentration range studied. Coefficients of determination, r ,^ were > 0.98 for all regression lines used for MCP quantitation. The coefficient of variation obtained for each data point is reported in Fig. 3 with an overall average of 6.4%. The HFB derivatives of MCP and MAP were found to be stable at room temperature for at least 72 hours and for more than 3 weeks when stored at -20 °C as verified by repeat injections showing no significant decline in peak areas over the period of storage. Structural confirmation of MCP in a standard solution, plasma and urine was carried out by gas-chromatography mass-spectrometry (electron impact, positive and negative ion chemical ionization) and is presented in the Appendix. The qualitative identification of MCP metabolites in plasma and urine samples obtained from pregnant ewes is also presented in the Appendix. 3.2 Selection of Weighting Factor for Computer Data Analysis Plots of In oL versus In C were made for each group of animals based on the relationship between variance and mean plasma concentration for a group of subjects (Eq. 7) proposed by Albert et al., 1974. The slopes (n') of the plots were determined using linear regression analysis and are presented along with their correlation coefficients in Tables 2 and 3 for the i.v. bolus and infusion experiments, respectively. In the pregnant ewe i.v. bolus crossover studies, n' values of 1.79, 1.70, and 2.02 were 0.800-1 0.700-0.600-0.500-0.400-0.300-0.200-0.100-0.000 SamDle3 A .R. b C.V.C 2.0 0.0249 ± 0.0030 12.0 4.0 0.0594 ± 0.0038 6.4 8.0 0.1519 + 0.0070 4.6 16.1 0.3053 ± 0.0171 5.6 24.1 0.4252 + 0.0209 4.9 32.2 0.5740 + 0.0512 8.9 40.2 0.7058 + 0.0189 2.7 a) b) c) MCP concentration, ng/mL Mean area ratio ± SD Coefficient of variation, % MCP concentration (ng /mL) Fig. 3. Coefficient of variation study showing a typical standard curve of sheep plasma extracts obtained by plotting the area ratio of the heptafluorobutyryl (HFB) derivatives of MCP/MAP versus MCP concentration, n = 4, duplicate injections; Y = 0.018X - 0.001, r = 0.999. Table 2. Weighting factor determination for computer fitting of MCP i.v. bolus concentration versus time data. 3 1. Pregnant ewes; maternal i.v. bolus crossover doses, n = 5. Dose (mg) n' MA r FA n' r 10 20 40 1.79 1.70 2.02 0.909 0.926 0.776 1.64 0. 2.34 0, .882 .983 2. Pregnant ewes; n = 4. single maternal 40 mg i.v. bolus dose. Dose (mg) n' MA r FA n' r 40 1.57 0.891 1.40 0. .798 3. Nonpregnant ewes; i.v. bolus crossover doses, n = 5. Dose (mg) n' r 10 1.91 0.958 20 1.70 0.986 40 1.77 0.985 80 2.28 0.977 a) Linear regression slope (n') and correlaton coefficient (r) values obtained from plots of In variance versus In mean MCP plasma concentration in the ewe (MA) or fetus (FA). Table 3. Weighting factor determination for computer fitting of MCP infusion concentration versus time data. 3 1. Maternal i.v. bolus (15 mg) plus constant rate infusion (0.21 mg/min x 90 min). n = 8. MA n' FA r n' r 1.81 0.973 2.17 0.982 2. Fetal i.v. bolus (5 mg) (0.07 mg/min x 90 min). plus constant rate infusion n = 8. MA n' FA r n' r 1.95 0.939 1.64 0.992 a) Linear regression slope (n') and correlaton coefficient (r) values obtained from plots of In variance versus In mean MCP plasma concentration in the ewe (MA) or fetus (FA). 54 obtained for maternal plasma at the 10, 20 and 40 mg doses, respectively. Slopes of 1.64 and 2.34 were determined in fetal plasma for the 10 and 20 mg maternal doses. In the four ewes receiving the single 40 mg bolus dose only, n' values of 1.57 and 1.40 were obtained for the ewe and fetus, respectively. In the nonpregnant ewes, respective slopes of 1.91, 1.70, 1.77 and 2.28 were calculated for the 10, 20, 40 and 80 mg i.v. bolus doses. Respective values of 1.81 and 2.17 were determined for the ewe and fetus for the maternal infusion experiments and 1.95 and 1.64 following the fetal infusions. Correlation coefficients were good ranging from 0.776 to 0.992. Since the n' values were near or approached two, the squared reciprocal of plasma drug concentration was used as the weighting factor in the nonlinear least-squares regression analysis (AUT0AN, N0NLIN74, N0NLIN84). This weighting factor also provided the best agreement between computer generated and hand calculated pharmacokinetic constants. 3.3 Metoclopramide Pharmacokinetics Following I.V. Bolus Dosing to the Ewe Pregnant Ewes - A representative semilogarithmic plot of the maternal and fetal plasma concentration versus time profiles obtained following a 40 mg maternal i.v. bolus is illustrated in Fig. 4. The pharmacokinetic parameters calculated for the ewe from the crossover (n = 5) and single 40 mg (n = 4) i.v. bolus experiments are presented in Tables 4 and 5. In the majority of the experiments, the maternal metoclopramide concentration versus time curves were described by the following biexponential equation: C = Ae" a t + Be"^ (27) representing a two compartment open model, where C is the plasma concentration at time t, and a and /3 are rate constants for the distribution and elimination phases, respectively. A and B are the Fig. 4. Representative semilogarithmic plots of metoclopramide concentration versus time profiles in maternal (MA) and fetal (FA) arterial plasma, tracheal (TR) and amniotic (AMN) fluids following a 40 mg i.v. bolus to a pregnant ewe. cn Table 4. Distribution and disposition half-lives and compartmental rate constants in the pregnant ewe following i.v. bolus doses. Dose Ewe a 0 Ha H/3 k10 k 1 2 k21 mg h"1 h"1 min min h"1 h' 1 h"1 10 56 6.908 0.694 6.0 59.9 1.845 3.159 2.598 59 3.344 0.571 12.4 72.8 1.266 1.140 1.509 68 12.179 0.840 3.4 49.5 3.709 6.551 2.759 624 25.646 0.852 1.6 48.8 3.123 16.378 6.997 712 2.818 0.331 14.7 125.7 1.187 1.176 0.785 Mean 10.179 0.658 7.6 71.3 2.226 5.681 2.930 ±SD 9.418 0.216 5.7 31.9 1.135 6.373 2.413 20 56 4.905 0.748 8.5 55.6 2.142 1.798 1.712 59a - 0.585 - 71.1 - - -68 11.740 1.059 3.5 39.3 3.544 5.748 3.507 624 2.834 0.466 14.7 89.2 1.639 0.855 0.806 712 3.781 0.663 11.0 62.7 1.674 1.273 1.497 Mean 5.815 0.704 9.4 63.6 2.250 2.419 1.881 ±SD 4.040 0.224 4.7 18.5 0.893 2.253 1.151 40 56 11.535 0.512 3.6 81.3 1.918 7.052 3.076 59a - 0.824 - 50.5 - - -68 6.823 0.393 6.1 105.8 1.493 3.927 1.795 624a - 0.852 - 48.8 - - -712a - 1.007 - 41.3 - -Mean 9.179 0.717 4.8 65.5 1.706 5.489 2.436 +SD - 0.255 - 27.2 - - -40 121 4.572 0.438 9.1 95.0 1.819 2.091 1.100 127 2.977 0.414 14.0 100.4 0.957 1.146 1.290 138 7.1.65 0.660 5.8 62.9 2.421 3.450 1.954 140 3.500 0.443 11.9 93.8 1.137 1.441 1.365 Mean 4.554 0.489 10.2 88.0 1.584 2.032 1.427 ±SD 1.863 0.115 3.5 17.0 0.671 1.024 0.369 a) Delay in the time to peak plasma concentration 57 Table 5 . Pharmacokinetic parameters obtained in the pregnant ewe following i.v. bolus doses. Dose mg Ewe Weight kg A U C ; ng.h/ mL A U M C ™ ng.h 2 / mL C L S L / h / kg V c L/kg V A  vss L/kg V B  vss L/kg ^area L/kg 10 56 61 .4 36 .1 14 .6 3 . 8 2.1 4 . 7 4 .4 5 .5 59 79 .5 40 .1 5 4 . 8 2 .6 2.1 3 . 7 3 . 6 4 . 6 68 70 .5 37 .2 33 .2 3 . 2 0 . 9 3 .1 3 . 4 3 . 8 624 8 0 . 9 49 .2 4 2 . 8 2.1 0 . 8 2 .8 2 . 7 2 .5 712 7 3 . 6 54 .6 115 .8 2.1 1.8 4 . 6 4 . 5 6 .4 Mean 73 .2 43 .4 52 .2 2 .8 1.6 3 . 8 3 . 7 4 . 6 +SD 7 . 8 8 .1 3 8 . 5 0 .7 0 .6 0 . 8 0 . 7 1.5 20 56 55 .4 5 1 . 9 5 .0 2 .4 4 . 9 4 . 7 6 . 7 5 9 C 6 0 . 9 - 3 . 5 - - - 6 .0 68 4 1 . 8 3 0 . 8 5 .7 1.7 4 .4 4 .2 5 .4 624 5 9 . 7 73 .2 3 . 5 2 .3 4 . 6 4 .3 7 .5 712 51 .5 57 .1 4 . 5 2 .8 5 .3 4 . 9 6 .7 Mean 5 3 . 9 53 .2 4 .4 2 .3 4 . 8 4 . 5 6 .5 ±SD 7 .7 17 .5 1.0 0 .5 0 .3 0 .3 0 .8 40 56 178.9 302 .8 3 .1 1.7 5 . 5 5 .2 6 .0 5 9 C 136.0 - 3 .1 - - - 3 . 8 68 211 .9 453 .3 2 .3 1.6 5.1 4 . 8 5 .8 6 2 4 C 111 .5 - 3 . 7 - - _ 4 .4 7 1 2 C 123.3 - 3 . 7 - - - 3 . 7 Mean 152.3 3 7 8 . 0 3 . 2 1.6 5 .3 5 .0 4 . 7 ±SD 4 1 . 9 - 0 .6 - - - 1.1 40 121 66 .4 149.6 237 .6 3 . 4 2 .0 5 .7 5 .4 7 .8 127 7 7 . 0 153.2 3 0 0 . 8 2 . 9 3 . 0 5 .7 5 .6 6 .9 138 6 1 . 4 102.7 116.4 5 .4 2 .2 6 .2 6.1 8 .1 140 7 2 . 7 184 .6 305 .2 2 .5 2 .5 5 .1 4 .2 5 .7 Mean 6 9 . 4 147.5 240 .0 3 . 5 2 .4 5 .7 5 .3 7.1 ±SD 6 .9 3 3 . 8 8 8 . 0 1.3 0 .4 0 .5 0 . 8 1.1 a) V s s = V c [ ( k 2 1 + k 1 2 ) /k 2 1 ] b) V S S = dose[(AUMCo )/(AUCQ ) 2] c) Delay in the time to peak plasma concentration v 58 respective intercepts for the a and /J phases. Values for A, a, B and /3 obtained from NONLIN (Metzler et al., 1974) were used to calculate the pharmacokinetic parameters for the model shown in Scheme 2. In three of the ewes involved in the crossover studies, however, Ewe 59 at the 20 and 40 mg doses and Ewes 624 and 712 at the 40 mg dose, there was an average delay in the time to maternal peak plasma concentration of 20.0 ± 8.2 min. While it is possible that extravascular administration of the drug occurred in these three ewes, there was no obvious swelling around the catheter site as would be expected following a 10 mL injection volume. Since the animals are not sacrificed following experimentation verification of catheter position was not possible. The plasma profile in the fetal lambs of these ewes followed a similar pattern with an average delay in the time to peak plasma concentration of 45.0 + 12.2 min. Pharmacokinetic parameters in these three ewes were subjected to a one-way ANOVA with no significant differences being observed in the apparent elimination rate constant, tj,^, CLS or V a r e a when compared to these parameters in the other ewes at all doses, and therefore have been included in Tables 4 and 5. To check for the further occurrence of a delay in the time to reach peak plasma concentration, 40 mg doses were administered to four additional ewes. These animals demonstrated a normal biphasic exponential decline in their plasma MCP concentration versus time profile. Similarly, there were no significant differences in the distribution (a) or terminal elimination (/i) rate constants, t^a, t^ , CL S , AUC, AUMC, V c , V s s , the apparent f irst order elimination rate constant (k^) or the intercompartmental distribution rate constant (k2]) between those ewes receiving the 40 mg bolus dose in the crossover experiments and those receiving the single 40 mg i.v. bolus. In spite of a significant difference in V a r e a (p = 0.0134) and the intercompartmental distribution rate constant, k 2^ (P = 0.0481), the 59 (ko + ) a D r Central compartment v c k12 Peripheral compartment K21 k10 where D represents the i.v. bolus dose of MCP.HCL, V c is the apparent volume of the central compartment, V p is the volume of the peripheral compartment, kj 2 and k 2j are the apparent first order intercompartmental distribution rate constants for MCP and kig is the apparent first order MCP elimination rate constant from the central compartment. a) For the infusion experiments, the model includes constant i.v. input (ko) in addition to an initial i.v. bolus loading dose. Scheme 2. Two compartment open model. 60 pharmacokinetic data from these single 40 mg dose animals have been included in Tables 4 and 5, and all 40 mg data were grouped for further statistical comparisons. The difference in V a r e a between the two groups appears to be due to generally lower values of /3 and AUC for those ewes receiving the single 40 mg bolus dose (Eq. 13). In those ewes in which MCP followed a biexponential plasma profile, distribution of metoclopramide was fairly rapid, with an overall ti. f t of 8.4 ± 4.5 min. There were no significant differences in this parameter for the three doses. Elimination of MCP from the ewe was rapid, with an average apparent terminal t ^ of 71.3 ± 31.9, 63.6 ± 18.5 and 75.5 ± 24.9 min at the 10, 20 and 40 mg doses, respectively (Table 4). Again, there were no significant differences among the three metoclopramide doses. Maternal CLS averaged 2.8 ± 0.7, 4.4 + 1.0 and 3.3 ± 0.9 L/h/kg at the 10, 20 and 40 mg doses, respectively (Table 5). Although CLS tended to be higher at the 20 and 40 mg doses, there were no statistical differences among the three doses. Similarly, V a r e a showed no significant differences at the three doses, averaging 4.6 ± 1.5, 6.5 ± 0.8 and 5.8 ± 1.6 L/kg at the 10, 20 and 40 mg doses, respectively (Table 5). As with CL S , V a r e a estimates for the 20 and 40 mg doses tended to be higher. The area under the metoclopramide plasma concentration versus time curve showed a linear correlation with the increase in "dose (r = 0.884), however the area for the 20 mg dose was less than what would be expected and accounts for the generally larger values for V a r e a and CLS observed at this dose level. The apparent volume of the central compartment, V c averaged 1.6 ± 0.6, 2.3 ± 0.5 and 2.2 ± 0.5 L/kg for the 10, 20 and 40 mg doses, respectively (Table 5). Although it tended to be higher at the 20 and 40 mg doses there were no significant differences among the three doses. The apparent volume of distribution at steady-state, V s s , relating the amount of MCP in the body to the concentration in plasma at steady-61 state was determined by both compartmental (Eq. 17) and noncompartmental (Eq. 18) methods. Average values for the 10, 20 and 40 mg i.v. bolus doses, using the compartmental calculation were 3.8 ± 0.8, 4.8 ± 0.3 and 5.5 + 0.4 L/kg, respectively, with the 40 mg estimate being significantly larger than that for either the 10 or 20 mg doses (p = 0.0018 and 0.0210, respectively). There was no significant difference between the 10 and 20 mg V s s estimates. Values of 3.7 ± 0.7, 4.5 ± 0.3 and 5.2 ± 0.7 L/kg were determined for the 10, 20 and 40 mg bolus doses, respectively, using the noncompartmental determination. Like the compartmental calculation, the volume obtained for the 40 mg dose was significantly larger than that for the 10 or 20 mg doses (p = 0.0015 and 0.0193, respectively) and there was no difference between the 10 and 20 mg values. The two methods of V s s determination were compared, showing good agreement, with the noncompartmental estimates averaging 6.2 + 4.9% lower than the compartmental values. The apparent f irst order elimination rate constant, kjQ, and the intercompartmental distribution rate constants, k 1 2 and k 2 i , were calculated from the computer generated values obtained for A, B, a and /J using standard equations (k 2 1 = Aa + ByS/(A + B); k 1 0 = a/3/k21; k 1 2 = a + /? - k 2 1 - k 1 0 ; Gibaldi and Perrier, 1982) and are presented in Table 4. Overall average values of 1.992 ± 0.860, 3.812 ± 4.047 and 2.183 ± 1.560 h" * were calculated for k^, kj 2 and k2^, respectively with no significant differences among the three doses. The number of fetuses present (1-3; Table 6) had no apparent effect on pharmacokinetic parameters between the ewes. Transfer of metoclopramide from the ewe to the fetus was observed to be rapid with peak fetal plasma concentrations observed in the 1 min sample (Fig. 4) for most studies in which a biexponential profile was observed in maternal plasma. Plasma data in these fetuses were also described by Eq. 62 Table 6. Pharmacokinetic parameters obtained in the fetal lamb following maternal i.v. bolus doses. Dose mg Ewea G.A. b days A < ng.h/ mL AUC f/ AUCm a h- 1 h- 1 Ha min min 10 56(1) 128 31.5 0.87 9.691 0.436 4.2 95.4 59(2) 118 34.6 0.86 37.740 0.483 1.1 86.0 68(3) 131 15.5 0.42 20.592 0.824 2.0 50.5 624(2)c 131 - - - - - -712(1) 117 38.2 0.70 20.448 0.513 2.0 81.0 Mean 125 29.9 0.71 22.118 0.564 2.3 78.2 ±SD 7 10.0 0.21 11.599 0.176 1.3 19.4 20 131 32.6 0.59 3.957 0.498 10.5 85.0 59d 120 37.0 0.61 - 0.748 - 55.6 68 134 37.4 0.90 11.908 0.488 3.5 85.1 624e 134 41.1 0.69 - 0.427 - 97.3 712 120 63.7 1.24 6.572 0.562 6.3 73.9 Mean 128 42.4 0.81 7.479 0.543 6.8 79.4 +SD 7 12.3 0.27 4.053 0.124 3.5 15.7 40 5 6 H d P 135 145.3 0.81 _ 0.293 _ 142.1 5 9 d,e 125 81.9 0.60 - 0.973 - 42.7 139 83.1 0.39 18.826 0.454 2.2 91.6 624 d ' e 141 61.5 0.55 - 0.789 - 52.7 712d 123 104.3 0.85 - 0.612 - 67.9 Mean 133 95.2 0.64 18.826 0.624 2.2 79.4 +SD 8 31.8 0.19 - 0.268 - 39.6 40 121(2) 132 114.2 0.76 7.975 0.446 5.2 93.3 127(3) 132 150.4 0.98 10.513 0.291 4.0 142.7 138(1) 136 71.9 0.70 5.302 0.487 7.8 85.3 140(2) 125 150.5 0.82 2.963 0.310 14.0 134.1 Mean 132 121.8 0.82 6.688 0.384 7.8 113.8 +SD 6 37.3 0.12 3.270 0.098 4.5 28.8 a) Number in parentheses indicates the number of fetuses. b) Gestational age at time of experiment. c) Fetal plasma concentration not available after 30 min, parameters not calculated. d) Delay in time to peak plasma concentration. e) Unable to computer f i t fetal concentration-time curve data, terminal elimination half- l i fe calculated from linear regression of the terminal elimination phase. 63 27. In two of these fetal lambs (Ewe 624, 20 mg dose; Ewe 56, 40 mg dose), however, peak plasma concentration occurred 5 min after maternal dosing. Identification of a distribution phase in these two fetuses was, therefore, not possible. The number of fetuses present and their gestational age at the time of the experiment are given in Table 6, as are the distribution and elimination rate constants and their half-lives, determined for MCP in fetal plasma. The overall distribution half- l i fe in the fetus averaged 5.2 + 3.9 min and showed no significant differences between doses. The overall elimination half- l i fe in the fetus averaged 86.8 ± 29.3 min with no significant differences between doses. Fetal elimination half - l i fe, however, was significantly higher than the mean value (72.5 ± 24.6 min) in the ewe (p = 0.0167), Fig. 5. As with maternal parameters, the comparable fetal kinetic constants appeared unaffected by the number of fetuses present. The overall ratio of fetal to maternal AUC averaged 0.74 ± 0.20 and was not significantly different at the three doses. In one instance, at the 20 mg dose (Ewe 712), this ratio exceeded one. The number of fetuses had no consistent effect on the fetal/maternal AUC ratio. Fetal to maternal concentration ratios were also determined at each time point in the 19 i.v. bolus experiments, with an overall mean ratio of 0.70 ± 0.43. In 13 of the 19 experiments the fetal arterial concentration exceeded that in the ewe, occurring between 1-2^ hours following maternal dosing in 12 cases and by 15 min in one (e.g. Figs. 4, 7 and 9). The fetal/maternal concentration ratio following this point of crossover averaged 1.25 ± 0.19. In the remaining 6 experiments the fetal to maternal concentration ratio also increased in a continuous manner, but a crossover pattern was not observed by the end of the sampling period. The replacement of fetal blood volume with an equal amount of maternal blood at intervals during the experiment did not result in any 120-i 100-80-60-40-20-20-i min 15-10-5-T I fT 1 • s NONPREGNANT PREGNANT FETUS 1 CLS varea |8 a t1 /2 a L/h/kg L/kg h~1 h"1 min Fig. 5. Comparison of pharmacokinetic parameters obtained in pregnant and nonpregnant ewes and fetuses following i.v. bolus crossover doses of 10, 20 and 40 mg to the ewe. The data are presented as the overall mean and 1 SD. Key: (a), significantly different from maternal tj,o (p = 0.0167); (b) significantly different from pregnant ewe Cl_s (p =z6.0487). CTl 65 marked aberrations in the fetal concentration-time curves {e.g. Fig. 4). The volume replaced at each interval represents =2-3% of total fetal blood volume. Nonpregnant Ewes - Representative plasma concentration versus time curves obtained in one of the nonpregnant ewes following 10, 20, 40 and 80 mg bolus crossover doses are illustrated in Fig. 6. Plasma profiles in all animals were best described by Eq. 27. Estimates for A, a, B and /J obtained from NONLIN (Metzler et al., 1974) were used to calculate the pharmacokinetic parameters for the model shown in Scheme 2 and are given in Tables 7 and 8. Distribution of metoclopramide in these animals was also rapid with an overall t\a of 6.0 ± 2.8 min. There were no significant changes in this parameter with increasing doses. Elimination of metoclopramide was rapid, with an average apparent terminal t ^ of 69.0 ± 22.1, 71.3 ± 27.7, 60.3 ± 13.7 and 63.3 ± 14.1 min at the 10, 20, 40 and 80 mg doses, respectively (Table 7). There was no significant difference in the elimination half- l i fe over the 8-fold dose range examined in this study. Total body clearance, CL S, averaged 3.9 + 1.1, 4.7 + 1.9, 5.1 + 1.3 and 5.2 ± 0.8 L/h/kg at the 10, 20, 40 and 80 mg doses, respectively (Table 8). Although CLS tended to be somewhat higher at the 40 and 80 mg doses, there were no significant differences among the four dose levels. Similarly, V a r e a was not significantly different at the four doses, averaging 6.1 ± 0.9, 7.7 ± 3.0, 7.3 + 1.8 and 7.8 ± 1.5 L/kg at the 10, 20, 40 and 80 mg doses, respectively. Again there was a tendency for this parameter to be slightly higher at the 40 and 80 mg doses. The area under the metoclopramide concentration versus time curve was linearly correlated with dose (r = 0.997). The apparent volume of the central compartment, V c , averaged 1.8 ± 0.5, 1.9 ± 0.7, 1.8 ± 0.3 and 2.1 ± 0.3 L/kg for the 10, 20, 40 and 80 mg bolus doses, respectively (Table 8). There were no •j _l 1 l 1 ( 1 ( I I 1 I 1 ( 0 60 120 180 240 300 360 Time (min) Fig. 6. Representative semilogarithmic plots of metoclopramide concentration verus time profiles in arterial plasma following 10, 20, 40 and 80 mg crossover i.v. bolus doses to a nonpregnant ewe. 67 Table 7. Distribution and disposition half-lives and compartmental rate constants in the nonpregnant ewe following i.v. bolus crossover doses. Dose Ewe a lhct hp k10 k12 k21 mg h"1 h"1 min min h"1 h"1 h"1 10 50 10.544 0.640 3.9 65.0 1.626 5.408 4.151 62 10.099 0.783 4.1 53.1 3.633 5.073 2.175 64 9.766 0.655 4.3 63.4 2.966 5.297 2.158 574 3.564 0.386 11.7 107.6 1.105 1.599 1.246 638 9.996 0.743 4.2 56.0 2.743 5.289 2.707 Mean 8.794 0.641 5.6 69.0 2.415 4.533 2.487 +SD 2.937 0.154 3.4 22.1 1.029 1.645 1.068 20 50 9.778 0.770 4.2 54.0 1.842 4.621 4.085 62 9.489 0.743 4.4X 56.0 3.792 4.581 1.859 64 4.853 0.694 8.6 59.9 1.958 1.868 1.721 574 7.004 0.346 5.9 120.1 2.694 3.757 0.900 638 8.620 0.625 4.8 66.5 3.348 4.288 1.609 Mean 7.949 0.636 5.6 71.3 2.727 3.823 2.035 ±SD 2.040 0.171 1.8 27.7 0.851 1.146 1.204 40 50 5.428 0.751 7.7 55.3 2.729 1.956 1.495 62 6.518 0.813 6.4 51.2 2.432 2.720 2.178 64 12.780 0.789 3.2 52.7 4.498 6.830 2.241 574 4.500 0.492 9.2 84.5 2.133 1.821 1.039 638 7.971 0.719 5.2 57.8 3.108 3.737 1.845 Mean 7.439 0.713 6.3 60.3 2.980 3.413 1.760 ±SD 3.253 0.128 2.3 13.7 0.922 2.056 0.501 80 50 10.973 0.794 3.8 52.4 2.469 5.770 3.527 62 5.705 0.646 7.2 64.4 2.439 2.400 1.511 64 8.933 0.736 4.6 56.5 2.910 4.500 2.259 574 3.151 0.476 13.2 87.3 1.868 0.956 0.803 638 11.805 0.740 3.5 56.2 2.875 6.631 3.040 Mean 8.113 0.678 6.5 63.3 2.512 4.051 2.228 +SD 3.637 0.125 4.0 14.1 0.422 2.351 1.106 6 8 Table 8. Pharmacokinetic parameters obtained in the nonpregnant ewe following i.v. bolus doses on a crossover basis. Dose Ewe Weight AUC° 0 AUMCQ C L S V c V A  ¥ss V B  vss ^area mg kg ng.h/ ng.h 2 / L / h / L/kg L/kg L/kg L/kg mL mL kg 1 0 5 0 6 9 . 0 3 9 . 4 5 4 . 9 3.1 1.9 4.5 4.3 4.9 6 2 5 9 . 1 3 3 . 5 2 9 . 0 4.3 1.2 4.1 3.7 5.4 6 4 4 9 . 6 4 0 . 9 4 5 . 2 4.2 1.4 5.0 4.6 6.4 5 7 4 6 1 . 9 5 2 . 5 1 0 5 . 1 2 . 6 2.4 5.4 5.2 6.7 6 3 8 5 0 . 0 3 1 . 7 3 3 . 0 5.3 2.0 6.0 5 . 5 7.2 Mean 5 7 . 9 3 9 . 6 5 3 . 4 3 . 9 1.8 5.0 4.7 6.1 ±SD 8.2 8.2 3 0 . 6 1.1 0.5 0.8 0.7 0.9 2 0 5 0 7 4 . 7 8 5 . 8 3 . 3 1.8 3 . 9 3.8 4.3 6 2 6 4 . 7 5 7 . 3 4.4 1.2 4.2 3 . 9 6.0 6 4 6 2 . 5 6 4 . 5 5.5 2.9 ' 6.0 5.6 7.9 5 7 4 9 4 . 1 1 7 4 . 4 2.9 1.1 5.7 5.4 8.4 6 3 8 4 4 . 3 4 7 . 6 7.6 2.3 8 . 6 8.2 1 2 . 2 Mean 6 8 . 1 8 5 . 9 4.7 1.9 5.7 5.4 7.7 ±SD 1 8 . 2 5 1 . 4 1.9 0.7 1.9 1.8 3 . 0 4 0 5 0 1 0 5 . 8 8 8 . 1 4.6 1.7 4.0 3 . 9 6.2 6 2 1 3 9 . 0 1 2 9 . 0 4.1 1.7 3 . 9 3 . 8 5.1 6 4 1 1 6 . 3 1 0 1 . 7 5.9 1.4 5.5 5.1 7.4 5 7 4 1 3 9 . 2 1 8 9 . 3 3 . 9 2.1 5.7 5.3 8.0 6 3 8 9 6 . 3 9 1 . 7 7.0 2 . 3 6.9 6.5 9.8 Mean 1 1 9 . 3 1 1 9 . 9 5.1 1.8 5.2 4.9 7.3 ±SD 1 9 . 4 4 2 . 0 1.3 0.3 1.2 1.1 1.8 8 0 5 0 2 1 6 . 6 2 2 6 . 7 4.5 1.9 5.0 4.7 5.7 6 2 2 2 6 . 4 2 3 6 . 9 5.1 2.1 5 . 5 5.3 7.8 6 4 2 5 9 . 8 2 6 4 . 9 5.2 1.8 5.5 5.4 7.1 5 7 4 2 4 1 . 4 2 8 5 . 3 4.5 2 . 5 5.5 5.4 9.5 6 3 8 2 0 4 . 7 2 2 3 . 1 6.6 2.3 7.5 7.2 8.9 Mean 2 2 9 . 8 2 4 7 . 4 5.2 2.1 5.8 5.6 7.8 ±SD 2 1 . 5 2 6 . 8 0.8 0.3 1.0 0.9 1.5 a) V SS V c [ ( k 2 1 + k 1 2 ) /k 2 1 ] b) V SS dose[(AUMCo ) / (AUCQ )2] 69 significant differences among the four doses. The apparent volume of distribution at steady state, V s s , was calculated using both compartmental (Eq. 17) and noncompartmental (Eq. 18) methods. Average estimates of 5.0 ± 0.8, 5.7 ± 1.9, 5.2 ± 1.2 and 5.8 ± 1.0 L/kg were obtained for the compartmental calculation for the 10, 20, 40 and 80 mg doses, respectively. There were no significant differences among the doses. Values of 4.7 ± 0.7, 5.4 ± 1.8, 4.9 + 1.1 and 5.6 ± 0.9 L/kg were obtained for the 10, 20, 40 and 80 mg crossover doses, respectively using the noncompartmental determination. Again, there were no significant differences among the four doses. A comparison of the V s s estimates showed the noncompartmental values to be an average of 5.5 ± 2.2% lower. The computer generated values obtained for A, B, a and 0 in the nonpregnant ewes were also used to calculate the apparent f irst order elimination rate constant, kjQ, and the intercompartmental distribution rate constants, k 1 2 and k 2 1 , using standard equations (k 2 1 = Aa + B/J/(A + B); k 1 Q = a £ / k 2 1 ; k 1 2 = a + 0 - k 2 1 - k 1 0 ; Gibaldi and Perrier, 1982) and are presented in Table 7. Overall average values of 2.636 ± 0.805, 3.955 ± 1.753 and 2.127 ± 0.965 h"1 were determined. There were no significant differences in these three parameters with increasing doses. Pharmacokinetic parameters for the pregnant and nonpregnant ewes were compared at the 10, 20 and 40 mg i.v. bolus doses, as illustrated in Fig. 5. No significant differences were observed in distribution (a) or terminal elimination (£) rate constants, t%a, t^ , AUC, V a r e a , V c , V s s , kjQ, k i 2 or k2^. Total body clearance, however, was found to be significantly different (p = 0.0487), with the nonpregnant animals (4.6 ± 1.4 L/h/kg) having an average clearance value which was =24% higher than in the pregnant ewes (3.5 ± 1.0 L/h/kg; Table 5). 70 3.4 Metoclopramide Pharmacokinetics Following Maternal and Fetal Infusions to Steady-State Details on the animals used for the infusion experiments are given in Table 9. Maternal body weight averaged 77.8 ± 16.6 kg and estimated fetal body weight 2.18 ± 0.86 kg. Gestational ages averaged 129.6 + 5.4 and 131.0 ± 5.7 days for the maternal and fetal infusions, respectively. Maternal Infusions - In the absence of a computer simulation program to choose the appropriate model and thus provide initial estimates of rate constants, intercepts and volume terms for nonlinear least-squares regression analysis, they were obtained from manual plotting. These initial hand-calculated estimates were then supplied to one and two compartment bolus plus constant infusion models in the N0NLIN84 library, with the program supplying final estimates for rate constants, half-lives and volume terms. Based upon the goodness of f i t of observed and calculated concentration versus time data, correlation coefficients and minimum sum of squared residuals, the maternal plasma concentration profiles, in all cases (n = 8), were best described by a two compartment open model (Scheme 2) representing the following biexponential equation (Gibaldi and Perrier, 1975): C = ko/(V ck 1 0) + Re" a t + S e - ^ (28) where R = (aXo - ko)(k 1 0 - 0)/(V ck l o)(a - 0) and S = (/?Xo - ko)(a -k 1 0)/(V ck 1 0)(<* - /}). Xo, represents the i.v. bolus loading dose and ko, V c , a, /} and kjQ are as previously defined. A representative semi logarithmic plot of the maternal and fetal MCP plasma concentration versus time profiles obtained following an init ial maternal i.v. bolus loading dose and simultaneous infusion are shown in Fig. 7. Metoclopramide reached average steady-state concentrations of 50.0 ± 20.2 ng/mL in the ewe (Cm) and 27.1 ± 8.6 ng/mL in 71 Table 9. Maternal and fetal weights and gestational age of the animals used for the maternal (MI) and fetal (FI) MCP infusion experiments. Ewe No. of Maternal Fetal Gestational Age fetuses weight weight MI FI (kg) (kg)a (days) 60 1 55.5 1.22 120.0 122.0 62 1 60.0 1.95 129.0 132.0 98 1 61.8 2.66 136.0 138.0 114 2 65.4 2.31 132.0 135.0 128 2 88.6 2.08 126.0 129.0 130 1 96.4 3.25 135.0 133.0 137 3 90.0 1.37 128.0 130.0 139 1 93.2 3.57 135.0 137.0 202 4 89.1 1.20 125.0 123.0 Mean 77.8 2.18 129.6 131.0 ±SD 16.6 0.86 5.4 5.7 a) Estimated fetal body weight, [Log Experimental weight = log birth weight - (0.0153)(No. of days between birth and experiment)]. Gresham et al., 1972. 1 0 0 0 q 1 0 0 -2 1 0 , A M A * F A • A M N Infusion 0 6 0 1 2 0 1 8 0 2 4 0 3 0 0 3 6 0 Time (min) Fig. 7. Representative semilogarithmic plots of the maternal (MA) and fetal (FA) arterial plasma, tracheal (TR) and amniotic (AMN) fluid metoclopramide concentration versus time profiles obtained following a 15 mg i.v. bolus loading dose and a constant rate infusion of 0.21 mg/min for 90 min to the ewe. the fetus (Cf), with a mean fetal to maternal concentration ratio (Cf/Cm) of 0.57 ± 0.14 (C.V. = 24.6%) (Table 10). The average coefficient of variation for the maternal and fetal steady-state concentrations was 7.0 ± 3.0% and 6.4 ± 3.0%, respectively. Steady-state concentrations were attained 15 min after initiation of the infusion in one of the eight ewes and by 30 min in seven. The time to reach steady-state concentrations in fetal plasma was more variable being achieved by 15 min in three cases, 30 min in three and 45 min in two. The pharmacokinetic parameters obtained in the ewe are presented in Tables 11 and 12. Distribution of MCP was rapid with an average t^a of 4.6 ± 1.5 min, approximately one-half the overall estimate obtained in the pregnant ewes involved in the i.v. bolus studies (8.4 + 4.5 min; Table 4). Elimination of MCP from the ewe was rapid with an average t^j of 64.7 ± 15.3 min in good agreement with the overall value of 71.3 ± 24.6 min (Table 4) obtained for the ewes in the i.v. bolus experiments. The computer generated values obtained for R, S, a and ft in the ewes were used to calculate the apparent f irst order elimination rate constant, k 1 Q , and the intercompartmental distribution rate constants, k 1 2 and k 2 j , using standard equations (kjg = Ra + S/3/(R + S); k 2 i = ajS/kin; k 1 2 = a + p - kiQ - k 2 i ; Gibaldi, 1969). The apparent f irst order elimination rate constant, k l f J, averaged 4.458 ± 1.350 h"1 approximately double the estimate obtained for the i.v. bolus studies (Table 4). The intercompartmental distribution rate constants, kj 2 and k 2 1 , averaged 4.941 ± 2.801 and 1.562 ± 0.517 h" 1, respectively, and are quite similar to those obtained in the bolus experiments (Table 4). The volume of the central compartment (Eq. 16), V c , averaged 1.1 + 0.3 L/kg somewhat lower than the overall average of 2.0 ± 0.6 L/kg for the ewes receiving the i.v. bolus doses. The volume of distribution at steady-state was also calculated for the infusion experiments using both compartmental (Eq. 17) and 74 Table 10. Steady-state metoclopramide plasma protein binding (Bm, Bf), plasma concentration (Cm, Cf) and concentration ratio (fetus/mother) following maternal infusion. Ewe Bm B f Cm C f Cf/C„ No. % % ng/mL ng/mL 60 35.8 30.6 94.3 (59.4) 42.3 (29.0) 0.45 (0.49) 62 47.7 ( - ) b 61.3 (31.7) 33.8 (20.6) 0.55 (0.65) 98 57.3 ( ") 63.6 (27.2) 31.7 (14.5) 0.50 (0.53) 114 51.2 ( -) 30.6 (14.5) 22.7 (10.4) 0.74 (0.72) 128 48.6 41.4 40.9 (19.8) 33.6 (19.4) 0.82 (0.98) 130 38.0 34.4 37.1 (23.1) 21.0 (13.9) 0.57 (0.60) 137 57.6 40.8 44.6 (18.4) 24.4 (14.9) 0.55 (0.81) 139 51.2 49.7 45.5 (20.5) 16.9 (8.7) 0.37 (0.42) 202 47.2 29.7 32.2 (15.8) 17.9 (12.9) 0.56 (0.82) Mean ±SD 48.3C 7.5 37.8 7.6 50.0 20.2 (25.6) (13.8) 27.1 8.6 (16.1) (6.2) 0.57 0.14 (0.67) (0.18) a) Numbers in parentheses indicate values for unbound drug, calculated using the average binding from the maternal and fetal infusions. b) ( - ) Insufficient plasma for determination. c) Significantly different from Bf (p = 0.0097). 75 Table 11. Maternal distribution and disposition half-lives and compartmental transfer rate constants following constant rate infusion of metoclopramide to the ewe. Ewe a /* k10 k12 k21 h"1 h"1 min min h ' 1 h- 1 h' 1 60 8.305 0.694 5.0 59.9 2.963 4.090 1.946 62 17.660 0.770 2.4 54.0 6.260 9.994 2.172 98 16.320 0.765 2.5 54.4 6.760 8.480 1.847 114 9.310 0.537 4.5 77.4 4.590 4.168 1.089 128 7.267 0.501 5.7 83.1 3.841 2.979 0.947 137 9.481 0.682 4.4 61.0 3.691 4.720 1.752 139 6.177 0.490 6.7 84.9 3.587 2.236 0.843 202 7.760 0.970 5.4 42.8 3.969 2.864 1.898 Mean 10.285 0.676 4.6 64.7 4.458 4.941 1.562 ±SD 4.288 0.164 1.5 15.3 1.350 2.801 0.517 76 Table 12. Maternal pharmacokinetic parameters obtained following constant rate infusion of metoclopramide to the ewe. Ewe A l< AUMCQ" C L S A C L S B V c vss V D  vss V E "area ng.h/ ng.h 2 / h L / h / L / h / L/kg L/kg L/kg L/kg mL mL kg kg 60 231.0 326.3 1.15 2.0 2.2 1.1 3.5 2.6 3.2 62 160.7 233.3 1.03 2.9 3.0 0.7 3.9 3.1 3.9 98 169.3 233.6 0.96 2.7 2.7 0.6 3.3 2.6 3.6 114 89.9 135.4 1.22 2.7 3.5 1.3 6.1 4.3 6.5 128 111.1 181.3 1.21 2.9 2.9 1.1 4.7 3.5 5.8 137 117.5 176.9 1.09 2.6 2.7 1.1 3.9 2.9 4.0 139 117.7 174.3 1.14 2.5 2.7 1.2 4.2 3.1 5.6 202 76.3 87.1 0.72 3.7 4.2 1.6 3.9 3.1 4.3 Mean 134.2 193.5 1.07 2.8 3.0 1.1 4.5 3.1 4.6 ±SD 50.3 72.0 0.16 0.5 0.6 0.3 1.4 0.5 1.2 a) CLS = ko/C s s b) CLS = 2dose/AUCo c) V ss = V c [ (k 2 1 + k 1 2 ) /k 2 1 ] d) v ss = [Sdose/AUCg ]t b e) ^area = Sdose/jSAUCg 77 noncompartmental (Eq. 19) methods. Respective estimates of 4.5 ± 1.4 and 3.1 ± 0.3 L/kg were obtained. Unlike the close agreement seen in the i.v. bolus experiments, the noncompartmental value for the infusions is an average of 32.4 ± 5.7% lower. The compartmental value (4.5 + 1.4 L/kg) however, is very similar to both the compartmental (4.8 ± 1.0 L/kg) and noncompartmental (4.5 + 0.9 L/kg) overall average estimates determined in the i.v. bolus dosed ewes (Table 5). V a r e a averaged 4.6 ± 1.2 L/kg and is quite similar to the average of 5.6 ± 1.5 L/kg determined for the ewes receiving i.v. bolus doses of MCP (Table 5)., Total body clearance, CL S , was calculated using both equations 12 and 14 with averages of 3.0 ± 0.6 and 2.8 ± 0.5 L/h/kg, respectively (Table 12). Agreement of the two methods of calculation is good with an average difference of 6.7%. Comparison of the clearance values determined by Eq. 12 for the infusion and i.v. bolus experiments illustrates good agreement with an overall value of 3.5 ± 1.5 L/h/kg for the ewes receiving the i.v. bolus doses of MCP (Table 5). Post-infusion estimates for terminal elimination rate constants and half-lives for MCP in fetal plasma following the maternal infusions are presented in Table 13. AUTOAN was used to choose the model and to supply initial estimates of rate constants and intercepts for subsequent nonlinear least-squares regression using the selected model in the N0NLIN84 library. Final estimates for the rate constants and intercepts were generated by the N0NLIN84 program. In six of the eight experiments MCP followed a monoexponential decline in fetal plasma and was biexponential (Eq. 27) in the remaining two. This is in contrast to the i.v. bolus studies where, with the exception of two cases in which there was a delay in the time to peak plasma concentration, fetal concentration profiles were biexponential in those instances in which a biexponential process was observed in the ewe. The elimination rate constant and half- l i fe in the fetal lamb 78 averaged 0.355 ± 0.079 h"1 and 122.9 ± 31.0 min, respectively (Table 13). Comparing maternal and fetal terminal elimination half-lives, the latter (122.9 + 31.0 min) was significantly longer (p = 0.0005) than that obtained for MCP in the ewe (64.7 + 15.3 min; Table 11). The apparent terminal half- l i fe in the infusion experiments also tended to be considerably longer than the overall average estimate (86.8 ± 29.3 min) calculated for the fetus following maternal i.v. bolus dosing (Table 6). Fetal Infusions - A representative plot illustrating the MCP maternal and fetal plasma concentration versus time profiles obtained following an initial i.v. bolus loading dose and constant rate infusion in the fetus are illustrated in Fig. 8. Average steady-state MCP plasma concentrations of 253.7 ± 92.1 and 13.8 ± 4.5 ng/mL were attained in the fetus (Cf') and ewe (Cm ' ) , respectively (Table 14). The average coefficient of variation for the maternal and fetal steady-state MCP concentrations was 13.1 ± 11.9% and 6.0 ± 4.6%, respectively. Fetal steady-state concentrations were reached by 30 min after beginning the infusion in six of the eight experiments and by 45 min in two. Maternal steady-state plasma concentrations were reached by 15 min in one of the ewes, 30 min in one and 45 min in the remaining six. In the manner described for the maternal infusions, MCP fetal infusion concentration versus time data were manually plotted and hand-calculated estimates of rate constants, intercepts and volume terms used for one and two compartment i.v. bolus plus constant rate infusion models in the N0NLIN84 library, with the NONLIN 84 program supplying final estimates for rate constants, half-lives and volume terms. In all eight cases, fetal MCP plasma concentration versus time profiles followed a biexponential (Eq. 28) decline following termination of the infusion (two compartment open model; Scheme 2). This is in contrast to the fetal, post-infusion profiles obtained following maternal infusions, which were 79 Table 13. Post-infusion estimates of fetal arterial plasma terminal elimination rate constant and half- l i fe following maternal constant rate infusion. Ewe Model3 Elimination Rate h Constant (h~*) (min) 60 1C 0.468 88.9 62 1C 0.430 96.6 98 1C 0.335 124.1 114 1C 0.263 158.4 128 2C 0.376 110.6 137 1C 0.371 112.0 139 2C 0.231 179.8 202 1C 0.369 112.6 Mean 0.355 122.9 ±SD 0.079 31.0 a) AUTOAN was used to select the model and to provide initial estimates of intercepts and rate constants. 1C, data best described by 1 compartment; 2C, data best described by 2 compartments. Final parameters were subsequently obtained from N0NLIN84. 10000i E 1000: CP • c Q) • mid 100, o t • CL O O O •*-> 10 : 0) 2 * M A • F A * TR D A M N 240 300 360 420 Time (min) Fig. 8. Representative semi logarithmic plots of the maternal (MA) and fetal (FA) arterial plasma, tracheal (TR) and amniotic (AMN) fluid metoclopramide concentration versus time profiles obtained following a 5 mg i.v. bolus loading dose and a constant rate infusion of 0.07 mg/min for 90 min to the fetus. 81 Table 14. Steady-state metoclopramide plasma protein binding (B„,Bf), plasma concentration (C m ' , Cf') and concentration ratio (fetus/mother) following fetal infusion. Ewe Bm B f C m ' Cf' C f / C m No. % % ng/mL ng/mL 60 38.1 32.0 23.1 (14.5) 219.7 (150.9) 9.5 (10.4) 62 48.7 39.1 17.7 (9.2) 203.6 (123.9) 11.5 (13.5) 98 ( - ) b 54.1 15.5 (6.6) 235.2 (107.8) 15.2 (16.3) 114 54.1 54.2 13.1 (6.2) 235.9 (108.1) 18.0 (17.4) 128 54.7 43.2 12.2 (5.9) 318.6 (183.7) 26.1 (31.1) 130 37.3 33.9 11.3 (7.0) 191.5 (126.1) 16.9 (17.9) 137 59.9 36.2 13.9 (5.7) 458.1 (281.8) 32.9 (49.1) 139 58.6 47.2 8.2 (3.7) 139.7 (72.0) 17.0 (19.4) 202 54.8 25.5 9.6 (4.7) 281.1 (203.4) 29.4 (43.4) Mean 50.8C 40.6 13.8 (7.1) 253.7 (150.9) 19.6 (24.3) ±SD 8.7 9.9 4.5 (3.2) 92.1 (63.5) 8.0 (13.7) a) Numbers in parentheses indicate values for unbound drug, calculated using the average binding from the maternal and fetal infusions. b) ( - ) Insufficient plasma for determination. c) Significantly different from Bf (p = 0.0196). biexponential in only two of eight cases (Table 13). The pharmacokinetic parameters calculated for MCP in fetal arterial plasma are provided in Tables 15 and 16. The distribution half - l i fe, t^ , averaged 12.1 ± 3.1 min and is approximately three times greater than that observed for the ewe following maternal infusions (4.6 ± 1.5 min; Table 11). The elimination half- l i fe of MCP from fetal plasma, t^ , averaged 115.9 ± 44.3 min and is almost double that calculated for MCP in maternal plasma (64.7 ± 15.3 min) during maternal drug infusion (Table 11). It also tends to be somewhat higher than the overall fetal half- l i fe calculated following maternal i.v. bolus dosing (86.8 ± 29.3 min; Table 6). The apparent f irst order elimination rate constant, kig, and the intercompartmental distribution rate constants, k i 2 and k 2 i , were calculated from the computer generated values obtained for R, S, a and p using standard equations (kjQ = Ra + Sj3/(R + S); k 2 1 = ap7k10; k 1 2 = a + p - k 1 0 - k 2 1 ; Gibaldi, 1969). The apparent f irst order elimination constant, k^, and the intercompartmental distribution rate constants, k 1 2 and k 2 1 , averaged 1.862 ± 0.614, 1.368 ± 0.401 and 0.820 ± 0.354 h" 1, respectively (Table 15). Compared to the estimates calculated for MCP in maternal plasma following maternal infusions (Table 11), fetal values are about 50% lower. The volume of distribution for the central compartment, V c , averaged 5.6 ± 1 . 0 L/kg and is considerably larger than that for MCP in maternal plasma obtained for the maternal infusions (1.1 ± 0.3 L/kg; Table 12). Fetal volume of distribution at steady-state, V s s , was also calculated using compartmental (Eq. 17) and noncompartmental (Eq. 19) methods, with average respective values of 15.9 ± 5.2 and 11.4 ± 3.5 L/kg (Table 16). Comparing the two methods of calculation, the noncompartmental estimate is an average of 38.9 ± 5.5% lower than the compartmental estimation. The apparent volume of distribution of MCP in the fetal lamb, V a r p a , averaged 17.9 ± 5.8 L/kg 83 Table 15. Fetal distribution and disposition half-lives and compartmental transfer rate constants following constant rate infusion of metoclopramide to the fetal lamb. Ewe a 0 H<* k10 k12 k21 h"1 h"1 min min h"1 h- 1 h- 1 60 5.309 0.875 7.8 47.5 2.955 1.657 1.572 62 4.018 0.562 10.3 74.0 2.717 1.032 0.832 98 3.674 0.448 11.3 92.9 1.560 1.507 1.054 114 3.522 0.240 11.8 173.2 1.447 1.730 0.584 128 2.256 0.313 18.4 132.7 1.386 0.673 0.510 137 2.949 0.413 14.1 100.7 1.471 1.064 0.828 139 3.769 0.256 11.0 162.5 1.621 1.808 0.595 202 3.501 0.289 11.9 144.0 1.738 1.470 0.582 Mean ±SD 3.625 0.876 0.424 0.212 12.1 3.1 115.9 44.3 1.862 0.614 1.368 0.401 0.820 0.354 84 (Table 16). As with the volume of the central compartment this value is approximately four times greater than that seen for MCP in the ewe (4.6 ± 1.2 L/kg; Table 12) for the maternal infusions. Fetal total body clearance, C l s , was also determined using both equations 12 and 14 with averages of 7.2 ± 3.3 and 7.9 ± 2.8 L/h/kg, respectively, (Table 16). Agreement between the two methods of calculation is good with an average difference of 9.7%. Clearance of MCP from fetal plasma is approximately double that obtained in the ewe (=3.0 L/h/kg; Table 12) during the maternal infusion experiments. The post-infusion estimates of the values of the terminal elimination rate constant and half- l i fe for MCP in maternal plasma following the fetal drug infusions are contained in Table 17. Again, AUTOAN was employed to choose the model and to supply init ial intercepts and rate constants for subsequent nonlinear least-squares regression analysis employing the selected model in the N0NLIN84 library. Final estimates of intercepts and rate constants were generated by the N0NLIN84 program. The decline of MCP in maternal plasma followed a monoexponential process in five of the eight experiments and was biexponential in three. This is in contrast to the maternal infusion experiments where it was biexponential in all cases. The terminal elimination rate constant and half- l i fe averaged 0.492 ± 0.098 h"1 and 87.7 ± 19.0 min, respectively. Comparing the estimates of the fetal and maternal terminal elimination half- l i fe obtained from the fetal infusion experiments, that in the fetal lamb tends to be longer (115.9 ± 44.3 min; Table 15) than for the ewe (87.7 ± 19.0 min) although they are not significantly different. Maternal elimination half- l i fe values obtained from the maternal and fetal infusions were also compared averaging 64.7 ± 15.3 (Table 11) and 87.7 ± 19.0 min (Table 17), respectively. The estimate calculated following the fetal infusions is significantly higher (p = 0.0174) than 85 Table 16. Pharmacokinetic parameters in the fetal lamb following constant rate infusion of metoclopramide to the fetus. Ewe A < AUMCQ tb C L sa CL s b V c V c  vss V d  vss V e "area ng.h/ mL ng.h 2 / mL h L/h/ kg L/h/ kg L/kg L/kg L/kg L/kg 60 572.4 697.8 0.80 13.3 13.7 6.9 14.1 11.0 15.6 62 479.4 646.8 0.93 8.9 10.2 5.7 12.7 9.5 18.2 98 713.0 1483.9 1.66 5.7 5.0 4.8 11.8 8.4 11.3 114 877.5 2890.3 2.88 6.5 4.7 4.8 18.9 13.6 19.6 128 1000.0 2161.0 1.74 5.4 4.6 5.0 11.5 8.0 14.7 137 1323.9 2763.6 1.67 5.7 5.3 5.3 12.2 8.8 12.8 139 485.1 1469.8 2.61 7.1 5.5 5.1 20.6 14.4 21.6 202 933.7 2342.8 2.09 10.5 8.5 7.3 25.7 17.8 29.5 Mean ±SD 798.1 292.7 1807.0 869.7 1.80 0.73 7.9 2.8 7.2 3.3 5.6 1.0 15.9 5.2 11.4 3.5 17.9 5.8 a) CLS = ko/C s s b) CLS = Sdose/AUCg c) V vss V c [ (k 2 1 + k 1 2 ) /k 2 1 ] d) ^ss [Sdose/AUCo ]t b e) ^area = 2dose//3AUCQ 86 Table 17. Post-infusion estimates of maternal arterial plasma terminal elimination rate constant and half- l i fe following fetal constant rate infusion. Ewe Model3 Elimination Rate Constant (h -*) (min) 60 IC 0.573 72.6 62 IC 0.524 79.3 98 IC 0.369 112.7 114 2C 0.365 113.9 128 IC 0.606 68.6 137 2C 0.536 77.5 139 IC 0.563 73.8 202 2C 0.402 103.6 Mean 0.492 87.7 ±SD 0.098 19.0 a) AUT0AN was used to select the model and to provide initial estimates of intercepts and rate constants. IC, data best described by 1 compartment; 2C, data best described by 2 compartments. Final parameters were subsequently obtained from N0NLIN84. 87 that observed for the maternal infusions. The elimination half-lives for MCP in fetal plasma obtained from the fetal and maternal infusions were similarly compared, averaging 115.9 ± 44.3 (Table 15) and 122.9 ± 31.0 min (Table 13), respectively. Although the fetal half- l i fe tended to be slightly longer following the maternal infusions it was not significantly different. As with the maternal i.v. bolus experiments, the replacement of fetal blood volume with an equal amount of maternal blood at intervals during the maternal or fetal infusion studies did not result in any marked aberrations in the fetal concentration-time curves (e.g. Figs. 7, 8, 9 and 10). The volume replaced at each interval represents =2-3% of total fetal blood volume. 3.5 Metoclopramide Placental (Transplacental) and Nonplacental Clearances Protein Binding - The extent of MCP binding to maternal and fetal plasma at steady-state (Eq. 4) was determined by ultrafiltration (Centrifree* Micropartition System, Amicon Division, W.R. Grace and Co., Danvers, MA, U.S.A.). Binding percentages thus obtained were subsequently used to correct the clearance values for unbound drug. Stock solutions containing 5, 50 or 500 ng/mL of MCP.HCL.r^O in pH 7.4 isotonic phosphate buffer (equivalent to MCP free base) were filtered in order to check for adsorption of MCP to the f i l ter membrane (Table 18). MCP recoveries averaged 99.4 ± 1.1, 102.5 ± 3.1 and 104.2 ± 1.1% for the 5, 50 and 500 ng/mL concentrations, respectively, indicating no significant membrane uptake. For the maternal infusions, MCP binding averaged 48.3 ± 7.5 and 37.8 ± 7.6% in maternal (Bm) and fetal (Bf) plasma, respectively (Table 10). MCP was 50.8 ± 8.7% bound in maternal plasma and 40.6 ± 9.9% bound in Table 18. Measurement of metoclopramide adsorption to ultrafi ltrati membrane.3 (Mean ± SD). MCP measured prior MCP measured in Recovery to ultrafiltration ultrafiltrate ng/mL ng/mL % 5.1 ± 0.2 50.2 ± 2.6 496.8 ± 5.4 5.1 ± 0.1 51.4 ± 2.6 518.9 ± 10.7 99.4 ± 1.1 102.5 ± 3.1 104.2 ± 1.1 a) n = 3; duplicate injections. fetal plasma following fetal drug infusions (Table 14). The extent of MCP plasma binding was found to be significantly lower (p = 0.0012) in the fetus. The two binding values determined for maternal plasma from the maternal and fetal infusions were averaged to calculate unbound drug concentration as they were not significantly different. Neither were the two fetal MCP plasma binding estimations significantly different following the fetal and maternal infusions and they were similarly averaged for unbound drug calculation. In those instances where there was insufficient plasma for both binding measurements a single maternal or fetal infusion value was used (Tables 10 and 14). Unbound steady-state drug concentrations averaged 25.6 + 13.8 and 16.1 ± 6.2 ng/mL in maternal (Cm) and fetal (Cf) plasma, respectively, following the maternal MCP infusions (Table 10). The fetal to maternal concentration ratio (Cf/Cm) for unbound drug, averaged 0.67 ± 0.18. For the fetal drug infusions, mean steady-state unbound drug concentrations of 7.1 ± 3.2 ng/mL in maternal (Cm')and 150.9 ± 63.5 ng/mL in fetal plasma (Cf') were obtained (Table 14). The fetal to maternal unbound drug concentration ratio (Cf'/Cm') averaged 24.3 ± 13.7. Placental and Nonplacental Clearances - The placental, nonplacental and total clearances of MCP at steady-state following the paired maternal and fetal infusions were calculated employing the method of Szeto et a7., 1982 (Eqs. 21-26). The maternal-fetal unit is represented by the two compartment model illustrated in Scheme 3. The results for the placental and nonplacental clearances determined for total MCP concentration (bound plus free) in nine animals are presented in Table 19. Total clearance and the percent of nonplacental contribution to total drug elimination in the ewe and fetus are contained in Table 20. Total clearance of MCP from the ewe (CLmm) averaged 2.9 ± 0.4 L/h/kg (Table 20) and compares favorably with Maternal Compartment vm d-mf Fetal Compartment ' CLfm cl-mo Vm = apparent volume of the maternal compartment Vf = apparent volume of the fetal compartment CLmf = placental clearance from the maternal to fetal compartment = kmfV CLfm = placental clearance from the fetal to maternal compartment = kfmV CL m o = nonplacental clearance from the maternal compartment = km oVm CLf 0 = nonplacental clearance from the fetal compartment = kf 0Vf CL m m = total clearance from maternal compartment = CL m o + CLmf = (kmf + km o)Vm CLff = total clearance from fetal compartment = CLf 0 + CLfm • (kfm + k fo) v f Scheme 3. Two compartment open model used for the determination of MCP placental (transplacental) and nonplacental clearances. 91 Table 19. Placental and nonplacental clearances of metoclopramide (total drug concentration) in maternal and fetal sheep.3 Ewe Placental Nonplacental C Lmf b C L fm b C LmoC C L f o b L/h L/h L/h L/h 60 7.6 (6.2) 12.5 (10.2) 111.1 (2.0) 4.5 (3.7) 62 10.1 (5.2) 15.9 (8.1) 172.7 (2.9) 2.5 (1.3) 98 7.8 (2.9) 11.4 (4.3) 165.6 (2.7) 4.2 (1.6) 114 11.7 (5.0) 10.1 (4.3) 170.1 (2.6) 5.6 (2.4) 128 9.4 (4.5) 10.3 (4.9) 259.5 (2.9) 1.2 (0.6) 130 10.9 (3.3) 17.6 (5.4) 286.4 (3.0) 1.6 (0.5) 137 4.3 (3.1) 7.4 (5.4) 238.7 (2.6) 0.5 (0.4) 139 9.7 (2.7) 14.1 (3.9) 230.0 (2.5) 11.9 (3.3) 202 7.2 (6.0) 11.5 (9.6) 330.4 (3.7) 1.4 (1.2) Mean ±SD 8.7 2.3 (4.3) (1.3) 12.3 3.1 (6.2)d (2.4) 218.3 69.0 (2.8)e (0.5) 3.7 3.5 (1.7) (1.2) a) CLmf = Clearance from maternal to fetal compartment. CLfm = Clearance from fetal to maternal compartment. CL m o = Nonplacental clearance from maternal compartment. CLf 0 = Nonplacental clearance from fetal compartment. b) Numbers in parentheses normalized to estimated fetal body weight (L/h/kg). c) Numbers in parentheses normalized to maternal body weight (L/h/kg). d) Significantly different from CL mr (p = 0.0022 for units of L/h; p = 0.0036 for units of L/h/kg). e) Significantly different from CLf0 (p = 0.0341). 92 Table 20. Total clearance of metoclopramide (total drug concentration) and percent of nonplacental contribution to total drug elimination in the ewe and fetus. 3 Ewe Total Clearance Nonplacental Contribution CL m m b C L f f c Maternal Fetal L/h L/h % % 60 118.7 (2.1) 17.0 (13.9) 93.6 26.6 62 182.8 (3.0) 18.3 (9.4) 94.5 13.4 98 173.4 (2.8) 15.6 (5.9) 95.5 26.9 114 181.7 (2.8) 15.7 (6.7) 93.6 35.8 128 269.0 (3.0) 11.5 (5.5) 96.5 10.7 130 297.3 (3.1) 19.2 (5.9) 96.3 8.5 137 243.0 (2.7) 7.9 (5.7) 98.2 6.3 139 239.7 (2.6) 26.0 (7.3) 96.0 45.7 202 337.5 (3.8) 12.9 (10.8) 97.9 10.9 Mean ±SD 227.0 68.9 (2.9) (0.4) 16.0 5.1 (7.9)d (2.9) 95.8 1.7 20.5 13.8 a) CL m m = Total clearance from maternal compartment = CL m o + CLmf. CLff = Total clearance from fetal compartment = CLf 0 + CLfm. b) Numbers in parentheses normalized to maternal body weight (L/h/kg). c) Numbers in parentheses normalized to estimated fetal body weight (L/h/kg). d) Significantly different from CLm m on a weight corrected basis (p = 0.0010). 93 the overall average total body clearance (CLS) obtained for the ewes in the i.v. bolus experiments (3.5 ± 1.0 L/h/kg; Table 5). Total clearance of MCP from the fetal lamb (Clff) averaged 7.9 ± 2.9 L/h/kg and is significantly greater (p = 0.0010) than that in the ewe on a weight corrected basis (Table 20). Both maternal and fetal total clearances showed a positive correlation with body weight with correlation coefficients of 0.878 and 0.723, respectively. Fetal placental clearance (CLfm; clearance of MCP from the fetal to maternal compartment) averaged 6.2 ± 2.4 L/h/kg while maternal placental clearance (CLmf; clearance of MCP from the maternal to fetal compartment) averaged 4.3 ± 1.3 L/h/kg (Table 19). The placental clearance of MCP from the fetal to the maternal (CLfm) compartment was significantly higher (p = 0.0022 for units of L/h; p = 0.0036 for units of L/h/kg) than that from the maternal to the fetal compartment (CLmf). While CLmf did not correlate with maternal weight (r = -0.100) it did show a . slight, positive correlation with fetal weight (r = 0.420). The nonplacental clearance from the fetal compartment (CLf0) averaged 1.7 ± 1.2 L/h/kg (Table 19) and accounts for 20.5 ± 13.8% of total fetal clearance (Table 20). The mean nonplacental clearance from the maternal compartment (CLmo) was 2.8 + 0.5 L/h/kg (Table 19). Nonplacental clearance from the maternal compartment contributes to 95.8 ± 1.7% of the total drug clearance (CLmm) from the ewe (Table 20) and is significantly larger (p = 0.0341) than CLf 0. The clearances determined for unbound drug are reported in Tables 21 and 22. Total clearance of MCP from the maternal compartment (CLmm) averaged 5.9 + 1.2 L/h/kg (Table 22) and is significantly lower (p = 0.0008) than the average total clearance from the fetal compartment (CLff = 13.3 ± 3.6 L/h/kg; Table 22). The fetal placental clearance (CL f m) for unbound drug averaged 10.4 ± 2.7 L/h/kg (Table 21) while that for the ewe 94 Table 21. Placental and nonplacental clearances of metoclopramide (unbound drug concentration) in maternal and fetal sheep.3 Ewe Placental C Lmf b C L fm b L/h L/h 60 12 1 (9.9) 18 1 (14.9) 176 2 (3.2) 6 6 (5.4) 62d 19 5 (10.0) 26 1 (13.4) 333 3 (5.5) 4 0 (2.1) 9 8 de 18 2 (6.8) 24 9 (9.4) 387 6 (6.3) 9 2 (3.4) 114d 24 6 (10.5) 22 0 (9.4) 359 1 (5.5) 12 3 (5.2) 128 19 5 (9.4) 17 8 (8.6) 536 5 (6.0) 2 1 (1.0) 130 17 5 (5.4) 26 7 (8.2) 459 5 (4.8) 2 5 (0.8) 137 10 4 (7.6) 12 0 (8.8) 578 9 (6.4) 0 8 (0.6) 139 21 5 (6.0) 27 3 (7.7) 509 6 (5.4) 23 1 (6.5) 202 14 6 (12.2) 15 9 (13.3) 674 2 (7.6) 1 9 (1.6) Mean 17 5 (8.6) 21 2 (10.4) f 446 1 (5.6)9 6 9 (3.0) ±SD 4. 5 (2.3) 5 5 (2.7) 149 4 (1.2) 7. 1 (2.2) a) CLmf = Clearance from maternal to fetal compartment. CLfm = Clearance from fetal to maternal compartment. CL m o = Nonplacental clearance from maternal compartment. CLf 0 = Nonplacental clearance from fetal compartment. b) Numbers in parentheses normalized to estimated fetal body weight (L/h/kg). c) Numbers in parentheses normalized to maternal body weight (L/h/kg). d) Binding in fetal plasma obtained from fetal infusion only. e) Binding in maternal plasma obtained from maternal infusion only. f) Significantly different from CLmf (p = 0.0158 for units of L/h; p = 0.0145 for units of L/h/kg). g) Significantly different from CLf0 (p = 0.0154). Nonplacental C LmoC C L f o b L/h L/h 95 Table 22. Total clearance of metoclopramide (unbound drug concentration) and percent of nonplacental contribution to total drug elimination in the ewe and fetus. a Ewe Total Clearance Nonplacental Contribution CL m m b C L f f c Maternal Fetal L/h L/h % % 60 188.3 (3.4) 24.7 (20.3) 93.6 26.6 62d 352.9 (5.9) 30.1 (15.4) 94.5 13.4 98 d e 405.8 (6.6) 34.1 (12.8) 95.5 26.9 114d 383.8 (5.9) 34.3 (14.6) 93.6 35.8 128 556.0 (6.3) 20.0 (9.6) 96.5 10.6 130 477.0 (4.9) 29.2 (9.0) 96.3 8.6 137 589.3 (6.5) 12.8 (9.4) 98.2 6.4 139 531.0 (5.7) 50.5 (14.1) 96.0 45.8 202 688.8 (7.7) 17.8 (14.9) 97.9 10.8 Mean ±SD 463.6 149.0 (5.9) (1.2) 28.2 11.2 (13.3) f (3.6) 95.8 1.7 20.5 13.8 a) CLm m = Total clearance from maternal compartment = CL m o + CLmf. CLff = Total clearance from fetal compartment = CLf0 + CLfm. b) Numbers in parentheses normalized to maternal body weight (L/h/kg). c) Numbers in parentheses normalized to estimated fetal body weight (L/h/kg). d) Binding in fetal plasma obtained from fetal infusion only. e) Binding in maternal plasma obtained from maternal infusion only. f) Significantly different from CLm m when normalized for body weight (p = 0.0008). 96 (CLmf) averaged 8.6 + 2.3 L/h/kg (Table 21), with CLfm being significantly greater than CLmf (p = 0.0158 for units of L/h; p = 0.0145 for units of L/h/kg). As observed for total drug concentration, CLmf did not correlate with maternal body weight (r = - 0.085). Again however, it did correlate positively with estimated fetal body weight (r = 0.642), showing an improved correlation value over that obtained for total drug concentration (r = 0.420). The nonplacental clearance of MCP from the fetal compartment (CL f o) averaged 3.0 ± 2.2 L/h/kg (Table 21) and accounts for 20.5 ± 13.8% of the total drug clearance. The mean nonplacental clearance from the maternal compartment (CLmo) was 5.6 ± 1 . 2 L/h/kg (Table 21). CL m o contributes to 95.8 + 1.7% of the total drug clearance from the ewe and is significantly larger than CLf0 (p = 0.0154). 3.6 MCP Accumulation in Fetal Tracheal and Amniotic Fluids Maternal 40 mg I.V. Bolus Studies - In addition to fetal and maternal arterial plasma, fetal tracheal and amniotic fluids were collected in the four ewes receiving a single 40 mg i.v. bolus dose of metoclopramide. Representative concentration versus time profiles are illustrated in Fig. 4. Peak concentrations were reached in tracheal fluid by 20 min in one of the fetuses, by 30 min in two and by 60 min in the remaining one. Drug concentrations in tracheal fluid were an average of 15.1 ± 1.4 times higher than were found in fetal plasma throughout the period of study (Table 23). A similar ratio was obtained with the maternal plasma MCP concentrations. Within 1.5 to 3 hours after dosing, MCP tracheal fluid concentrations followed an exponential decline, with an estimated average elimination half- l i fe 127.8 ± 32.6 min (Table 23) similar to that found in fetal plasma (113.8 ± 28.8 min; Table 6). The tracheal half- l i fe was calculated using linear regression analysis and Eqs. 5 and 6. MCP concentrations increased 97 Table 23. Values for the ratio of drug concentration in amniotic (AMN/FA) and tracheal (TR/FA) fluids to that in fetal arterial plasma and estimated elimination half-lives in tracheal f luid. Experiment Ewe AMN/FA TR/FA Tracheal fluid half- l i fe (min) Maternal 40mg 121 1.7 15.8 102.7 i.v. bolus 127 4.2 14.1 165.7 138 5.0 16.8 98.7 140 2.9 13.8 144.3 Mean 3.5 15.1 127.8 ±SD 1.5 1.4 32.6 Maternal 60 2.9 12.5 75.8 Infusion 62 1.7 9.5 67.0 98 1.7 10.1 118.4 114 2.5 14.9 123.9 128 3.1 139 - 16.3 116.6 202 5.4 16.5 223.8 Mean 2.7 13.3 120.9 ±SD 1.3 3.1 55.8 Fetal 60 4.9 20.6 57.0 Infusion 62 4.8 15.2 64.4 98 2.5 11.9 106.3 114 2.9 14.6 115.9 128 3.2 139 - 15.6 96.5 202 5.5 17.6 86.8 Mean 3.6 15.9 87.8 ±SD 1.4 2.9 23.3 98 more gradually in amniotic fluid (Fig. 4) reaching a plateau between 90 and 210 min, with the exception of one ewe where this occurred by 45 min. In three of the four ewes amniotic concentrations were decreasing toward the end of the sampling period (3-4.5 hours) but in one instance (Ewe 140) they were s t i l l at plateau levels at the end of the experiment (6 hours). The ratio of MCP concentrations in amniotic fluid to that in fetal plasma were also calculated, averaging 3.5 ± 1.5 for the four i.v. bolus experiments (Table 23) and is significantly different from 1.0 (p = 0.0213). Maternal and Fetal Steady-State Infusions - Tracheal and amniotic fluids were also collected for drug determination in six of the nine paired maternal and fetal infusion studies. For both infusion studies, MCP concentrations in tracheal fluid were at least 10- to 15-fold higher than in fetal plasma (Table 23). Representative concentration versus time profiles for MCP in tracheal and amniotic fluids following an initial i.v. bolus loading dose and constant rate infusion for 90 min are illustrated in Fig. 7 for a maternal infusion and in Fig. 8 for one of the fetal infusions. With the maternal infusions MCP concentrations in tracheal fluid reached their peak by 45 to 120 min following initiation of dosing, while with the fetal infusions this occurred between 15 and 45 min. For a period from 5 to 60 min after stopping the infusion, relatively large periodic fluctuations in MCP tracheal concentrations were observed for all maternal infusions (Fig. 7). Similar fluctuations were seen with the fetal infusions between 75 min into the infusion and 15 min post-infusion, but they were not as large as those observed in the maternal infusions, and occurred in only 3 cases. Throughout the time course of the experiments, however, drug concentrations in tracheal fluid remained at least 10- to 15-fold higher than either the fetal or maternal plasma concentrations. Estimated elimination half-lives were calculated for MCP in tracheal fluid during the infusion experiments using linear regression and Eqs. 5 and 6, with mean values of 120.9 ± 55.8 and 87.8 ± 23.3 min for the maternal and fetal infusions, respectively (Table 23). The average half-lives, determined from the maternal infusions, for MCP in both tracheal fluid (120.9 + 55.8 min) and fetal plasma (122.9 ± 31 min; Table 13) were very similar, as was seen in the 40 mg maternal i.v. bolus studies. The mean tracheal fluid half- l i fe value (87.8 ± 23.3 min) obtained for the fetal infusions is , however, somewhat lower, though not significantly so, than that calculated in fetal plasma (115.9 ± 44.3 min; Table 15) during the fetal infusions . These tracheal fluid half- l i fe estimates were obtained from the usual 5 to 6 hour post-infusion sampling protocol and may be underestimates. This is suggested by the results from one paired infusion study (Ewe 202) where sampling was continued for 28 hours post-infusion. For the maternal infusion, MCP tracheal fluid concentrations followed an apparent triexponential decline with an estimated terminal elimination half- l i fe of 12.6 hours (Fig. 9). For the fetal infusion a triexponential decline was also seen (Fig. 10) with an estimated terminal half- l i fe of 15.2 hours. The half- l i fe estimates obtained from these two experiments are very much longer than those obtained from the usual 5-6 hour sampling protocol. Similar to the i.v. bolus experiments, MCP concentrations in amniotic fluid increased gradually, reaching plateau levels between 120 to 150 min after starting the maternal infusion. Between 210 and 330 min drug concentrations in amniotic and tracheal fluids were very similar. For the fetal infusions, plateau levels in amniotic fluid were attained between 60 and 210 min. By 240 to 330 min MCP concentrations in tracheal and amniotic fluids were similar. The highest concentration ultimately achieved in amniotic fluid following the maternal infusions was considerably higher 1 I I 1 1 1 1 1 I 1 1 1 • 1 0 5 10 15 20 25 30 Time (hr) Fig. 9. Representative semilogarithmic plots of the maternal (MA) and fetal (FA) arterial plasma, tracheal (TR) and amniotic (AMN) fluid metoclopramide concentration versus time profiles obtained following a 15 mg i.v. bolus loading dose and a constant rate infusion of 0.21 mg/min for 90 min to the ewe. "1 • 1 • I • 1 1 1 • 1 ' 1 0 5 10 15 20 25 30 Time (hr) Fig. 10. Representative semilogarithmic plots of the maternal (MA) and fetal (FA) arterial plasma, tracheal (TR) and amniotic (AMN) fluid metoclopramide concentration versus time profiles obtained following a 5 mg i.v. bolus loading dose and a constant rate infusion of 0.07 mg/min for 90 min to the fetus. 102 than the highest fetal plasma MCP concentration observed in the f irst 15 min of the experiments (mean ratio 1.88 ± 0.67). For the fetal infusions on the other hand, plateau amniotic concentrations and initial fetal arterial plasma concentrations were similar with a mean ratio of 0.73 ± 0.34. Concentration ratios for MCP in amniotic fluid to those in fetal plasma throughout the time course of the experiments were also determined, with average values of 2.7 ± 1.3 and 3.6 ± 1.4 for the maternal and fetal infusions, respectively (Table 23). Both ratios were significantly different from 1.0 (p = 0.0197 and 0.0021, respectively). In two of the maternal infusions where sampling was continued for 28 hours post-infusion (Ewes 128 and 202), MCP concentrations amniotic fluid followed a monoexponential decline between 2 and 28 hours (e.g. Ewe 202; Fig. 9). In these two instances an estimated apparent terminal half- l i fe was calculated using linear regression and Eqs. 5 and 6, values of 6.8 and 8.7 hours were obtained. A monoexponential decline between 1 and 28 hours post-infusion was also observed in the one fetal infusion (Ewe 202) where post-infusion sampling was continued for 28 hours (Fig. 10). An estimated half- l i fe of 6.1 hours was similarly calculated. Post-infusion sampling was continued to 12 hours in Ewe 128 for the fetal infusion experiment. In this case a monoexponential decline was observed between 1 and 12 hours post-infusion and an estimated half- l i fe of 4.4 hours calculated. For the fetal infusion in which post-infusion sampling was continued for 28 hours (Ewe 202), the concentration of MCP in fetal plasma (=3.6 ng/mL) approached the limits of the assay (=2 ng/mL) in the sample taken at 12 hours post-infusion. When the next sample was taken at 20 hours after stopping the infusion, the concentration had increased by about 3-fold, with further small increases in the 22 and 28 hour samples (Fig. 10). This pattern was not seen in either of the two maternal infusions (Ewes 128 and 103 202) where the same sampling protocol was used. In the latter two experiments, MCP concentrations in fetal plasma were below the assay limit 7 hours after the infusion was stopped. 3.7 Metoclopramide Pharmacokinetics Following Intra-Amniotic Drug Administration Details on the four animals used for the intra-amniotic bolus experiments are presented in Table 24. Maternal body weight averaged 79.0 ± 4.1 kg and estimated fetal body weight 2.36 ± 0.72 kg. Gestational age at the time of the experiments ranged from 126.0 to 138.0 days, with a mean of 131.7 ± 5.1 days. Representative metoclopramide concentration versus time profiles in maternal arterial (MA), fetal arterial (FA) and umbilical venous (UV) plasma, tracheal (TR) and amniotic (AMN) fluids obtained following a 10 mg intra-amniotic bolus in one of the ewes are illustrated in Fig. 11. The pharmacokinetic parameters calculated for MCP in these fluids are presented in Table 25. Transfer of MCP from amniotic fluid to the fetal lamb was rapid with measurable concentrations in fetal arterial and umbilical venous plasma in the initial 5 min sample. Metoclopramide was detected in maternal plasma and tracheal fluid within 10-20 and 5-10 min of drug injection, respectively. Peak drug concentrations (Cmax) (Table 25) were reached between 5-120 min (mean 71.7 + 59.6), 90-150 min (mean 120.0 ± 34.6), and 120-150 min (mean 140.0 ± 17.3) in tracheal and fetal and maternal plasma, respectively. In the two fetuses (Ewes 237, 287) in which umbilical venous samples were drawn, peak concentrations occurred 30-45 min earlier than in the fetal artery. Umbilical venous MCP concentrations were consistently higher than those in the fetal artery throughout the f irst 1.5 h following intra-amniotic drug administration. Although there were short 104 Table 24. Maternal and fetal weights and gestational age of the animals used for the 10 mg MCP intra-amniotic bolus experiments. Ewe No. of Maternal Fetal Gestational fetuses weight weight Age (kg) (kg)a (days) 237 1 78.0 2.46 133.0 279 2 74.5 2.05 130.0 283 1 79.1 3.32 138.0 287 2 84.5 1.63 126.0 Mean 79.0 2.36 131.7 +SD 4.1 0.72 5.1 a) Estimated fetal body weight, [Log Experimental weight = log birth weight - (0.0153)(No. of days between birth and experiment)]. (Gresham et a7., 1972). 100000 F oAMN • TR AUV A FA • MA 10 15 20 TIME, (hr) 25 30 F i g . 11. Representative semi l o g a r i t h m i c p l o t s of the maternal a r t e r i a l (MA), f e t a l a r t e r i a l (FA) and u m b i l i c a l venous (UV) plasma, t r a c h e a l (TR) and amniotic (AMN) f l u i d metoclopramide c o n c e n t r a t i o n versus time p r o f i l e s obtained f o l l o w i n g a 10 mg i n t r a - a m n i o t i c bolus i n j e c t i o n . 106 Table 25. Pharmacokinetic parameters obtained in maternal (MA) and fetal (FA) arterial and umbilical venous (UV) plasma, amniotic (AMN) and tracheal fluid (TR) following a 10 mg intra-amniotic injection of MCP. 1. Peak drug concentration (Cmax) and the time to reach peak concentration (Tmax) in maternal and fetal arterial plasma and fetal tracheal and amniotic fluids. Cmax (ng/mL) Tmax (min) Ewe AMN TR FA UV MA AMN TR FA MA 237 9546.3 - 29.9 31.1 - 5.0 - 150. ,0 -279 32274.5 810. 6 49.6 - 6.7 5.0 120. .0 150. .0 150. .0 283 9460.2 10074. .7 66.1 - 6.7 10.0 5, .0 90. .0 120. .0 287 27087.5 1648. 4 92.1 125.5 9.0 10.0 90. .0 90. ,0 150. .0 Mean 19592.1 4177. 9 59.4 78.3 7.5 7.5 71. .7 120, .0 140. .0 ±SD 11840.6 5123. 9 26.3 - 1.3 2.9 59. .6 34. .6 17. .3 2. Apparent terminal elimination half- l i fe of MCP in amniotic f luid, tracheal fluid and plasma and ratios of MCP concentration in maternal to fetal arterial plasma (MA/FA), umbilical venous to fetal arterial plasma (UV/FA) and tracheal fluid to fetal arterial plasma (TR/FA). Ewe MA/FA TR/FA UV/FA AMN TR FA UV MA 237 - - 1.3 6.6 - 9.0 8.6 -279 0.14 16.1 - 4.2 3.5 3.0 - 3.2 283 0.10 15.6 - 4.7 3.7 6.0 - 5.4 287 0.14 14.2 1.3 2.3 4.9 2.4 7.2 4.2 Mean +SD 0.13 0.02 15.3 1.0 1.3 4.5 1.8 4.0 0.8 5.1 3.0 7.9 4.3 1.1 a) Slope obtained from linear regression of terminal elimination phase. Apparent terminal elimination half- l i fe calculated using Eqs. 5 and 6. 107 periods of arterial crossover after this initial interval, MCP concentrations were significantly higher (p = 0.0185) in umbilical venous than fetal arterial plasma with a UV/FA concentration ratio of 1.3 (Table 25). Drug concentrations in fetal arterial plasma were also much higher than in maternal plasma, with a MA/FA concentration ratio of 0.13 ± 0.02 (Table 25). Metoclopramide also accumulated in fetal lung fluid following this route of administration, with a TR/FA concentration ratio of 15.3 + 1.0 (Table 25) which is essentially identical to the ratios obtained following the i.v. bolus and infusion experiments (Table 23). The apparent terminal elimination half- l i fe of MCP in amniotic, tracheal and fetal and maternal arterial plasma, averaged 4.5 + 1.8, 4.0 ± 0.8, 5.1 ± 3.0 and 4.3 ± 1.1 h, respectively (Table 25). The fetal arterial plasma half- l i fe in these experiments is =2.5-3.5 times larger than the values obtained following maternal i.v. bolus dosing (=1.5 h; Table 6) and the paired steady-state infusions (=2 h; Tables 13 and 15). Similarly, the maternal elimination half- l i fe following intra-amniotic injection is =2.5-4-fold greater than the estimates calculated from the maternal i.v. bolus and paired steady-state infusion studies (=1-1.5 h; Tables 4, 11 and 17). Compared with the average half- l i fe estimates (=7 h; Section 3.6) obtained from the two maternal and one fetal infusion where sampling was continued for 28 h after the infusion was stopped, the mean value for MCP in amniotic fluid following the intra-amniotic studies is =1.5 times smaller. Similarly, the mean estimate for MCP in tracheal f luid, in the current studies, is =3.5 times smaller than the average estimate (=14 h; Section 3.6) calculated following the single maternal and fetal infusions in which sampling was continued for 28 h after drug administration. The half- l i fe of the drug in umbilical venous plasma was very similar for the two animals with an approximate value of 8 h (Table 25). In one instance (Ewe 237) 108 umbilical venous and arterial concentrations followed a parallel decline through out the 28 h sampling interval, with very good agreement between their elimination half-lives (8.6 and 9.0 h, respectively). In the second fetus (Ewe 287), umbilical venous concentrations followed an apparent biexponential decay (Fig. 11). In this case arterial and umbilical venous concentrations declined in parallel up to 12 hours following intra-amniotic drug injection, with respective half-lives of 2.7 and 2.4 h. Thereafter, metoclopramide was not detectable in arterial plasma, but persisted in the umbilical vein until 24 h with an apparent terminal elimination half- l i fe of 7.2 h, approximately 3-fold larger than the corresponding arterial value (2.4 h) (Table 25). Metoclopramide was not detected in maternal plasma 10 h after injection, but was st i l l present in fetal arterial and umbilical venous plasma until 28 hours in two fetuses. In all cases, the drug was s t i l l measurable in tracheal and amniotic fluid in the final 28 h samples. Final (28 h) MCP concentrations in amniotic fluid, tracheal fluid and fetal arterial and umbilical venous plasma ranged between =20-300 ng/mL, =4-16 ng/mL and =2-6 ng/mL, respectively. 3.8 Physiological Effects of Metoclopramide on the Fetal Lamb 3.8.1 Fetal Arterial Blood Gases Maternal MCP I.V. Bolus Experiments - The effect of MCP on fetal arterial pH, Po2 and Pco2 at 1, 5, 10 and 30 min for the maternal 10, 20 and 40 mg i.v. bolus dose studies are presented in Table 26. Fetal arterial pH in the control period (-5 min) was in the normal range for fetal lambs averaging 7.399 ± 0.023, 7.428 ± 0.022 and 7.389 ± 0.034 for the 10, 20 and 40 mg doses, respectively. With the exception of the 5 min sample at the 40 mg dose, there was a small decline in fetal arterial pH at 1, 5, 10 and 30 min following maternal drug administration. This change, Table 26. Effect of metoclopramide on fetal arterial pH, P02 and PCO2 following maternal intravenous bolus dosing. [Mean (±SD)]. Dose Time PH Po2 Pco2 mg (min) (mm Hg) (mm Hg) 10 -5 7.399 20.1 53.0 (n=8)a (0.023) (3.5) (3.2) 1 7.396 18.6 50.7 (0.024) (4.7) (2.6) 5 7.394 19.6 51.3 (0.017) (3.8) (3.4) 10 7.389 20.4 52.2 (0.025) (4.5) (3.2) 30 7.390 20.6 52.4 (0.023) (5.1) (3.6) 20 -5 7.428 17.4 50.3 (n=6)a (0.022) (3.1) (2.4) 1 7.405 15.9 50.5 (0.013) (1.7) (3.1) 5 7.412 17.2 50.8 (0.019) (2.2) (3.7) 10 7.405 17.1 50.7 (0.012) (3.2) (3.5) 30 7.404 16.9 50.0 (0.011) (2.7) (3.5) 40 -5 7.389 17.1 49.9 (n=8)a (0.034) (1.9) (3.8) 1 7.383 15.7 50.1 (0.033) (2.4) (3.7) 5 7.400 17.1 50.7 (0.008) (2.6) (3.8) 10 7.387 16.6 49.3 (0.037) (2.5) (3.3) 30 7.379 15.8 49.4 (0.022) (2.4) (3.9) a) Number of animals (n). however, was not significant for any of the three doses. Fetal arterial Po2 was also in the normal range for fetal lambs during the control period, with mean values of 20.1 ± 3.5, 17.4 ± 3.1 and 17.1 ± 1 .9 mm Hg following the 10 , 20 and 40 mg maternal doses. In all fetuses there was a statistically significant (p = 0.0453), small decline (=1.3-1.5 mm Hg) from control Po2 at 1 min. By 5 min following drug administration, Po2 had returned to control values, although a small decline was also observed at 30 min with the 40 mg dose (Table 26). None of these changes were significant. Similarly, fetal arterial Pco2 was in the normal range for fetal lambs averaging 53.0 ± 3.2, 50.3 ± 2.4 and 49.9 ± 3.8 mm Hg, respectively, following the administration of the 10 , 20 and 40 mg i.v. bolus doses to the ewe. A small but transient decline (=2 mm Hg) from control Pco2 was seen at 1 and 5 min following the 10 mg dose with a return to control values by 10 min. This pattern was not observed following either the 20 or 40 mg maternal i.v. bolus doses. There were no significant changes from control Pco2 with any of the three bolus doses. Maternal and Fetal MCP Infusion Studies - A reduced sampling protocol was used in these experiments since no marked effects of MCP on fetal blood gases were noted following maternal bolus dosing. The effects of either a maternal or fetal MCP infusion on fetal arterial pH, P02 , PCO2, oxygen content (O2C) and hematocrit are presented in Table 27. As with the i.v. bolus studies, fetal arterial pH, Po2 and PCO2 during the control period were in the normal range for fetal lambs averaging 7.393 + 0.030, 15.7 ± 4.4 mm Hg, and 52.3 ± 7.2 mm Hg for the maternal drug infusions and 7.404 ± 0.035, 15.6 ± 4.1 mm Hg and 52.2 + 5.3 mm Hg for the fetal infusions. There were no consistent or significant changes from control in the 5 and 30 min samples for any of the measured blood gas parameters (pH, P02 , PCO2, 0 2C). Although Pco2 decreased by =3 mm Hg at 5 and 30 min after starting I l l Table 27 . Effect of metoclopramide maternal or fetal infusion on fetal arterial pH, Po 2, Pco2, 0 2 content (02C), and hematocrit (HCT). [Mean (±SD)]. Time PH Po2 Pco2 o2cb HCT (min) (mm Hg) (mm Hg) (%) (%) Maternal -5 7.393 15.7 52.3 7.0 35.2 Infusion (0.030) (4.4) (7.2) (1.9) (3.1) (n=10)a 5 7.400 16.4 49.3 6.5 34.6 (0.035) (3.5) (6.3) (1.7) (3.5) 30 7.402 16.5 49.0 6.8 33.6 (0.034) (3.3) (4.2) (1.6) (3.3) Fetal -5 7.404 15.6 52.2 6.5 35.0 Infusion (0.035) (4.1) (5.3) (1.4) (3.8) (n=8)a 5 7.396 17.0 51.0 6.9 34.2 (0.034) (4.0) (6.9) (1.5) (3.2) 30 7.414 16.1 49.2 6.3 32.8 (0.027) (2.5) (4.8) (1.0) (3.5) a) Number of animals (n). b) n=8 for maternal infusions; n=6 for fetal infusions. 112 the maternal infusion, this change was not significantly different from control values. A similar decrease was observed in the 30 min sample for the fetal infusions, again it was not significantly different from control. Fetal hematocrit was ~Z% lower than control values at 30 min for both the maternal and fetal infusions. This was not, however, a significant change (Table 27). Control Normal Saline Infusions - Normal saline infusions were administered as a control to one ewe and three fetuses in order to determine i f there was an infusion effect on fetal arterial blood gases. Maternal weight, fetal weight and fetal gestational age at the time of the experiments are presented in Table 28 and are very similar to those for the drug infusion studies (Table 9). The results of these experiments on fetal arterial pH, P02, PCO2, O2C and hematocrit are contained in Table 29. All control values (-5 min) were in the normal range for fetal lambs and were not different from those found prior to the MCP infusions. In the single maternal infusion, pH decreased from control at 5 and 30 min following initiation of the infusion. A decrease in P02 and O2C was also observed, particularly at 5 min. These declines were, however, within the range seen for the maternal MCP infusion experiments (Table 27). For the fetal infusions (n = 3), there were no consistent or significant changes from control in any of the measured parameters. Intra-Amniotic Bolus Experiments - The effect of the intra-amniotic injection of MCP on fetal arterial pH, P02, PCO2, oxygen content (O2C) and hematocrit are presented in (Table 30). Like all previous experiments, fetal arterial pH, Po2 and PCO2 during the control period (-5 min) were in the normal range for fetal lambs averaging 7.312 ± 0.070, 21.7 + 3.2 mm Hg and 50.9 + 5.3 mm Hg, respectively. Again, there were no significant changes from control in any of the measured blood gas parameters (pH, P02, 113 Table 28. Maternal and fetal weights and gestational age of the animals used for the control maternal and fetal normal saline infusions. 1. Fetal normal saline infusion. n = 3. Ewe No. of Maternal Fetal Gestational fetuses weight weight Age (kg) (kg)a (days) 213 3 86.8 2.30 136.0 215 3 86.4 1.70 129.0 482 1 68.2 2.52 139.0 Mean 80.5 2.17 134.7 ±SD 10.6 0.42 5.1 2. Maternal normal saline infusion, n = 1. Ewe No. of Maternal Fetal Gestational fetuses weight weight Age (kg) (kg)a (days) 135 2 70.0 1.73 128.0 a) Estimated fetal body weight, [Log Experimental weight = log birth weight - (0.0153)(No. of days between birth and experiment)]. (Gresham et a)., 1972). 114 Table 29. Effect of control maternal or fetal infusion of normal saline on fetal arterial pH, P02, Pco 2, O2 content (O2C), and hematocrit (HCT). [Mean (±SD)] • Time PH Po2 Pco2 o 2 c HCT (min) (mm Hg) (mm Hg) (%) (%) Maternal -5 7.386 16.5 49.9 6.2 31.4 Infusion (n=D a 5 7.350 14.8 51.4 4.8 34.2 30 7.362 16.1 49.8 5.3 31.7 Fetal -5 7.373 21.3 45.4 8.9 38.1 Infusion (0.010) (1.5) (2.0) (0.9) (3.5) (n=3)a 5 7.369 22.7 47.2 9.1 38.0 (0.007) (1.5) (2.5) (0.8) (4.1) 30 7.364 20.0 46.2 8.4 37.8 (0.008) (2.0) (3.2) (0.4) (3.4) a) Number of animals (n). 115 Table 30. Effect of intra-amniotic drug administration on fetal arterial pH, P02, PCO2, O2 content (02C), and hematocrit (HCT). [Mean (±SD)]. Time PH Po2 Pco2 o2c HCT (min) (mm Hg) (mm Hg) (%) (%) (n=3)a -5 7.312 21.7 50.9 7.0 33.5 (0.070) 3.2 5.3 1.6 5.3 30 7.323 23.3 49.2 7.3 32.5 (0.072) 3.8 4.4 2.1 5.2 60 7.314 22.3 49.8 7.1 32.6 (0.057) 2.5 2.8 1.5 4.9 a) Number of animals (n). 116 P c o 2 , O2C) at either 30 or 60 min following intra-amniotic drug administration. Similarly, there was no significant change in fetal hematocrit. 3.8.2 Fetal Arterial Pressure, Fetal Heart Rate and Amniotic Pressure (Intrauterine Pressure) A polygraph trace representing =1 hour control recordings of fetal heart rate (HR), fetal arterial pressure (FAP), intrauterine pressure (IUP) and intratracheal pressure (ITP) is shown in Fig. 12. Maternal MCP I.V. Bolus Studies - Polygraph recordings of fetal heart rate, fetal arterial pressure and intrauterine pressure obtained during the control, dosing and post-dosing periods were analyzed at 1 min intervals. Fetal arterial pressure was corrected for intrauterine pressure using Eq. 3. The one min pressure and heart rate values obtained for the 10, 20 and 40 mg bolus doses were graphed and showed no marked change from control during or following drug administration. Maternal arterial pressure and heart rate were also monitored in the ewes involved in the 10, 20 and 40 mg i.v. bolus crossover experiments; no changes were observed. Maternal and Fetal MCP Infusion Studies - Fetal heart rate, fetal arterial pressure and amniotic pressure for the infusion experiments were also recorded and analyzed at 1 min intervals during the control, infusion and post-infusion periods. This data was graphed at one minute intervals. No obvious trends or changes from control were observed in the plots for any of these three variables. Fetal heart rate (HR), arterial pressure (AP), and intrauterine pressure (Amn P) were, however, further analyzed by averaging the 1 min data over 30 min intervals during the 60 min control, 90 min infusion and 60 min post-infusion periods for statistical HR 240 /min 1 i n f o i ft—' f»wf >^  » 1 / ^ ^ - ^ W ^ 1" i1 Wi>»mr' i| i i » i . i . ^ | ' t r i ^ r r < S T ^ E C O G R m>miM*taK&!^ EOG 100 /iV Fig. 12. Representative control recordings of IUP (intra-uterine pressure), FAP (fetal arterial pressure), HR (fetal heart rate), ITP (intra-tracheal pressure), ECoG (fetal electrocortical activity), EOG (fetal electroocular activity). 118 comparisons. This data is contained in Table 31 for the maternal MCP infusions and in Table 32 for the fetal infusions. For the maternal MCP infusions, fetal heart rate averaged 161 ± 2 and 163 + 4 beat/min over the two 30 min intervals (-60, -30 min) in the control period, with no significant changes during or after the infusion (Table 31). Average control values of 51.6 ± 2.4 and 51.8 ± 1.9 mm Hg were calculated for fetal arterial pressure, and again, there were no significant differences during the infusion or post-infusion intervals. While amniotic pressure tended to decrease from control values during the infusion and post-infusion periods (=2-3 mm Hg), there were no significant differences. During the two control intervals for the MCP fetal infusions, fetal heart rate averaged 160 ± 3 and 164 ± 6 beats/min (Table 32). There were no significant changes during or following the infusion. Similarly, fetal arterial pressure averaged 44.6 + 1.6 and 45.8 ± 1.6 mm Hg, with no significant changes during or after MCP infusion. In contrast to the maternal infusions, intrauterine pressure increased slightly (=2 mm Hg) during and following the infusion (Table 32), but this change was not statistically significant. Control Normal Saline Infusions - The effect of control infusions of normal saline to the ewe (n = 1) or fetus (n = 3) on fetal arterial pressure and heart rate were also studied at 1 min intervals. Plots of this data showed no obvious changes or trends. Once again, this 1 min data was further analyzed by averaging over 30 min intervals during the 60 min control, 90 min infusion and 60 min post-infusion periods for statistical comparisons and are provided in Table 33. For the control fetal infusion experiments, fetal heart rate average 142 ± 4 and 143 ± 4 beat/min for the two 30 min intervals in the one hour control period. No significant Table 31. Mean (±SE) values for amniotic pressure (Amn P), fetal heart rate (HR) and fetal arterial pressure (AP) calculated over 30 minute intervals before, during and following infusion of metoclopramide to the ewe. (n = 10). Control Infusion Post-infusion -60 -30 30 60 90 120 150 HR 161 163 167 164 169 168 165 (/min) ±2 ±4 ±4 ±4 ±6 ±7 ±5 AP 51.6 51.8 51.2 52.4 52.5 52.9 53.0 (mm Hg) ±2.4 ±1.9 ±1.8 ±1.7 ±1.6 ±2.4 ±2.1 Amn P 17.2 16.2 14.4 13.9 11.5 12.6 13.6 (mm Hg) ±1.9 ±2.7 ±2.9 ±1.5 ±2.7 ±3.2 ±1.8 120 Table 32. Mean (±SE) values for amniotic pressure (Amn P), fetal heart rate (HR) and fetal arterial pressure (AP) calculated over 30 minute intervals before, during and following infusion of metoclopramide to the fetus, (n = 9). Control Infusion Post- infusion -60 -30 30 60 90 120 150 HR 160 164 164 156 159 162 155 (/min) ±3 ±6 ±7 ±4 ±5 ±5 +3 AP 44.6 45.8 45.6 46.4 45.6 44.1 45.1 (mm Hg) ±1.6 ±1.6 ±0.9 ±2.4 ±1.3 ±1.7 ±2.3 Amn P 14.7 14.6 16.2 16.1 16.6 16.9 15.5 (mm Hg) ±1.5 ±1.8 ±1.8 ±1.7 ±2.4 ±2.1 ±3.2 Table 33. Mean (±SE) values for fetal heart rate (HR) and arterial pressure (AP) calculated over 30 minute intervals before, during and following infusion of normal saline to the ewe or fetus. 1. Fetal control normal saline infusion, n = 3. Control Infusion Post-infusion -60 -30 30 60 90 120 150 HR 142 143 145 142 143 141 141 (/min) ±4 ±4 ±6 ±4 ±9 ±7 ±8 AP 56.3 56.1 56.7 53.7 54.3 52.5 53.2 (mm Hg) ±5.1 ±5.2 ±6.7 ±8.0 ±6.9 ±5.7 ±6.9 2. Maternal control normal saline infusion. n = 1. Control Infusion Post-infusion -60 -30 30 60 90 120 150 HR 136 130 138 139 138 138 132 (/min) AP 44.7 45.3 43.7 44.4 44.0 44.2 42.6 (mm Hg) 122 changes were seen during or following the saline infusion. The mean fetal heart rate was =20 beat/min lower than that observed in both the maternal (Table 31) and fetal (Table 32) drug infusions and is likely the result of the smaller number of animals involved in the control saline infusion experiments. Fetal arterial pressure during the two 30 min control intervals averaged 56.3 ± 5.1 and 56.1 ± 5.2 mm Hg and decreased by =2-3 mm Hg during the 60 and 90 min intervals of the infusion period as well as in the 120 and 150 min post-infusion intervals. This decline, however, was not significant. The results of the control maternal infusion of saline on fetal heart rate and arterial pressure in the one ewe studied are given in Table 33. No marked changes were observed in either variable. 3.9 Behavioral Effects of Metoclopramide on the Fetal Lamb 3.9.1 Fetal Breathing and Electrocortical Activities A polygraph trace representing =1 hour recordings of fetal electrocortical (ECoG) activity, electroocular (EoG) activity and breathing-like activity is presented in Fig. 12. Fetal breathing activity may be identified as negative deflections in the intra-tracheal pressure (ITP) trace and occurs during periods of low voltage electrocortical activity. Similarly, rapid eye movements occur during low voltage ECoG activity and appear as an increase in the voltage pattern of the EoG recording. MCP Single 40 mg Maternal I.V. Bolus Dose Experiments - Fetal electrocortical (ECoG) and breathing-like (BR) activities were monitored for 60 min before and for 60 min after maternal i.v. bolus drug administration (n = 4). In the control period, all fetal lambs exhibited regular alterations between low and high voltage electrocortical activity. 123 The percentage of time spent in low and high voltage ECoG patterns during the control period averaged 67.8 ± 5.4 and 32.2 ± 5.4, respectively (Table 34). No significant alterations were observed following the maternal bolus, with the fetus spending an average 63.7 + 6.3 and 36.4 ± 6.3% of the time in low and high voltage activities, respectively. Fetal breathing activity was associated with low voltage ECoG activity and occurred 36.2 + 15.4% of the time during the control period. Although breathing activity tended to decrease during the post bolus period, averaging 25.4 ± 14.5% (Table 34), this decline was not significantly different. Fetal breathing activity was also examined in 17 maternal i.v. bolus experiments, involving 10, 20 or 40 mg doses of MCP. No consistent effects were observed. In 11 of the experiments, brief bursts of activity lasting 1 to 6 min occurred within 1 to 3 min of MCP administration. Maternal and Fetal MCP Infusion Studies - The effect of MCP on fetal electrocortical and breathing activities was also examined in the maternal and fetal infusions. Electrocortical activity was recorded for 60 min before, during the 90 min infusion period and for 60 min post-infusion in five of the nine paired infusion experiments. Fetal breathing activity (BR) was determined over these same intervals in eight of the paired infusion studies. For the maternal drug infusions, all fetuses exhibited regular alterations between low and high voltage ECoG activity during the control period. The percentage of time spent in low and high ECoG voltage patterns averaged 66.9 ± 12.0 and 33.1 ± 12.0, respectively (Table 35). No significant changes were seen during the infusion period with low and high voltage activities averaging 64.7 ± 12.0 and 35.3 ± 12.0%, or during the post-infusion interval where respective averages of 70.1 ± 13.4 and 29.9 ± 13.4% were calculated. Similarly, fetal breathing activity was not significantly affected, occurring during low voltage ECoG periods an 124 Table 34. Incidence of low and high voltage in electrocortical episodes (ECoG) and breathing activity (BR) in the fetus following a 40 mg i.v. bolus of metoclopramide to the ewe. % ECoG Activity %BR Ewe Low High Prior to Bolus 121 70.9 29.1 53.3 127 65.8 34.2 45.3 138 73.2 26.8 23.0 140 61.2 38.8 23.3 Mean 67.8 32.2 36.2 +SD 5.4 5.4 15.4 Post-Bolus 121 63.9 36.1 28.3 127 64.2 35.8 23.4 138 71.0 29.0 42.6 140 55.5 44.5 7.4 Mean 63.7 36.4 25.4 ±SD 6.3 6.3 14.5 125 Table 35. Incidence of low and high voltage in electrocortical episodes (ECoG) and breathing activity (BR) in the fetal lamb following maternal MCP infusion. % ECoG Activity %BR Ewe Low High Prior to 60a 34.6 Infusion 62a - - 57.4 98a - - 33.2 114 77.8 22.3 76.7 128 59.7 ,40.3 28.6 130 78.5 21.5 73.6 1 3 7 h 68.3 31.7 -139b - - 27.2 202 50.4 49.6 14.9 Mean 66.9 33.1 43.3 ±SD 12.0 12.0 23.0 During 60 34.7 Infusion 62 - - 64.7 98 - - 31.9 114 60.6 39.4 24.3 128 82.7 17.3 64.8 130 53.3 46.7 13.2 137 70.6 29.4 -139 - - 28.4 202 56.2 43.8 23.6 Mean 64.7 35.3 35.7 ±SD 12.0 12.0 19.0 Post- 60 8.7 Infusion 62 - - 64.2 98 - - 38.6 114 68.7 31.3 59.9 128 90.8 9.2 72.3 130 57.1 42.9 30.6 137 73.8 26.2 -139 - - 13.1 202 60.0 40.0 12.5 Mean 70.1 29.9 37.5 ±SD 13.4 13.4 25.4 a) Electrodes not implanted in these fetuses. b) Electrodes not functional. 126 average 43.3 ± 23.0, 35.7 ± 19.0 and 37.5 ± 25.4% of the time for the control, infusion and post-infusion periods, respectively (Table 35). As seen with the maternal infusions, all fetuses displayed a regular alteration between low and high voltage ECoG activity during the control period in the fetal infusion experiments. The fetuses spent an average of 74.7 ± 13.9% of the time in low voltage activity during the control period, 69.7 + 17.5% in the infusion interval and 76.1 ± 15.4% in the post-infusion period (Table 36). There were no significant changes in either voltage pattern during or after the infusion. One fetus (Ewe 137) spent 100% of the time in low voltage activity during the infusion period with a return to control values after the infusion was stopped. Another fetus (Ewe 202) spent 100% of the post-infusion period in low voltage activity. Mean high voltage activity increased from a control value of 25.3 ± 13.9% to 30.5 ± 17.6% during the infusion period and then returned to a near control value of 23.9 ± 15.4% in the post-infusion interval. This increase was not significantly different. Although fetal breathing activity tended to decrease from a control value of 39.1 ± 13.5% to 22.1 ± 16.1 and 26.8 ± 14.3%, respectively, during the infusion and post-infusion periods, these changes were not statistically significant (Table 36). As observed in the maternal infusion and i.v. bolus dosing studies, fetal breathing activity occurred during periods of low voltage ECoG activity. There were no significant differences in low and high voltage ECoG or breathing activities between the maternal (Table 35) or fetal (Table 36) MCP infusions. Control Normal Saline Infusions - Fetal electrocortical and breathing-like activities were also recorded for a 60 min control, 90 min infusion and 60 min post-infusion period during control fetal and maternal infusions of normal saline. The results of these studies are provided in 127 Table 36. Incidence of low and high voltage in electrocortical episodes (ECoG) and breathing activity (BR) in the fetal lamb following infusion of metoclopramide to the fetus. % ECoG Activity %BR Ewe Low High Prior to 60a 43.6 Infusion 62a - - 46.0 98a - - 41.1 114 89.0 11.0 -128 68.0 32.0 -130 60.8 39.2 27.5 1 3 7 K 65.3 34.7 20.7 139b - - 32.7 202 90.2 9.8 61.8 Mean 74.7 25.3 39.1 +SD 13.9 13.9 13.5 During 60 _ _ 31.3 Infusion 62 - - 50.7 98 - - 28.0 114 62.3 38.7 -128 58.1 41.9 -130 59.2 40.8 20.1 137 100.0 0.0 11.7 139 - - 3.9 202 69.1 30.9 8.9 Mean 69.7 30.5 22.1 ±SD 17.5 17.6 16.1 Post- 60 49.6 Infusion 62 - - 42.0 98 - - 17.4 114 65.8 34.2 -128 83.2 16.8 -130 67.7 32.3 20.2 137 63.9 36.1 8.4 139 - - 23.7 202 100.0 0.0 26.3 Mean 76.1 23.9 26.8 ±SD 15.4 15.4 14.3 a) Electrodes not implanted in these fetuses. b) Electrodes not functional. Table 37. Again, all fetuses exhibited regular alterations between low and high voltage ECoG activity during the control period. For the three fetal infusions low voltage activity averaged 59.8 ± 10.9% for the control period. While a small decrease was seen during the infusion (58.7 ± 6.6%) and post-infusion (53.3 ± 2.9%) intervals, these changes were not statistically significant. High voltage activity averaged 41.1 ± 10.9, 41.3 ± 6.6 and 46.7 ± 2.9% during the control, infusion and post-infusion periods, respectively. Once again, there were no significant differences. Fetal breathing activity in these infusions was also associated with low voltage ECoG patterns. Of the 60 min control period, the fetal lambs spent an average of 35.0 ± 12.6% of the time in breathing-like movements (Table 37). Breathing activity decreased to 24.5 ± 6.9% during the infusion and to 19.5 ± 9.5% in the post-infusion period, but these decreases were not significant. A similar decrease (=13-17%) was also seen during the MCP fetal drug infusions (Table 36). The electrocortical and breathing activities calculated during the single maternal control saline infusion are presented in Table 37. With the exception of the control values, both low and high voltage ECoG activities during the infusion and post-infusion periods are close to the range of values seen in the maternal MCP infusions (Table 35). The percentage of breathing activity for the saline infusion is also very similar to that observed during the the maternal drug infusion experiments (Table 35). 129 Table 37. Incidence of low and high voltage in electrocortical episodes (ECoG) and breathing activity (BR) in the fetal lamb following infusion of normal saline. (Control study). 1. Fetal normal saline infusion, n = 3. % ECoG Activity %BR Ewe Low High Prior to 213 70.0 30.0 48.3 Infusion 215 48.3 51.7 23.3 482 58.3 41.7 33.3 Mean 59.8 41.1 35.0 +SD 10.9 10.9 12.6 During 213 61.8 38.2 16.8 Infusion 215 51.1 48.9 30.0 482 63.3 36.7 26.7 Mean 58.7 41.3 24.5 +SD 6.6 6.6 6.9 Post- 213 55.0 45.0 16.7 Infusion 215 50.0 50.0 30.0 482 55.0 45.0 11.7 Mean 53.3 46.7 19.5 ±SD 2.9 2.9 9.5 2. Maternal normal saline infusion . n = 1. % ECoG Activity %BR Ewe Low High Prior to 135 46.7 53.3 38.3 Infusion During 135 66.7 33.3 37.2 Infusion Post- 135 55.0 45.0 35.0 Infusion 130 4. DISCUSSION 4.1 Metoclopramide Assay Modification and Application The MCP assay employing fused si l ica capillary columns and electron capture detection previously developed in our laboratory (Riggs et a?., 1983) was further modified for our current studies of the pharmacokinetics of MCP in biological fluids obtained from pregnant and nonpregnant ewes and fetal sheep. This modified method has also recently been used to study the disposition of metoclopramide in plasma obtained from normal and uremic humans (Wright et.al., 1987, 1988, 1988a). The split mode of sample injection was used in the original assay. While generally not considered suitable for trace analysis as a considerable proportion of the sample is vented, it did provide good reproducibility and sensitivity when used with manual sample injection. Attempts to automate sample injection, however, resulted in extremely variable results with poor injection to injection and sample to sample reproducibility. This may have been the result of sample discrimination associated with inlet flow and needle residence time in the injection port during the automatic injection cycle. One of the early goals of the current project then, was to modify the existing assay to allow for automatic sample injection in order to more readily handle the large numbers of samples to be collected and analyzed in the i.v. bolus and infusion studies. The primary modification involved a switch from the split to splitless mode of sample injection. Various operating parameters related to this injection technique were, therefore, optimized. With both methods the sample is flash vaporized, but when using the splitless technique it must be reconcentrated at the head of the column prior to chromatographic development. An open or unpacked injection port liner is generally used for the splitless mode of sample injection as transfer of the sample to the 131 head of the column is relatively slow. The additional heat capacity, in the form of packing and/or a special liner configuration, which is often necessary to provide adequate sample vaporization and mixing with the split technique is not usually necessary for the splitless method. The nature of the splitless injection, in fact, requires minimal mixing (i.e. dilution) of the sample with carrier gas (Freeman, 1981). For the current studies an open fused si l ica injection port liner was used, replacing the Jennings^ tube or glass wool packed si l ica inserts previously employed. Reconcentration of the solute(s) at the head of the column may be achieved either by cold trapping or by using a solvent effect (Jennings, 1980; Freeman, 1981). Both techniques of recondensation were tested. An init ial column temperature 10-30 °C below the boiling point of the solvent (toluene, b.p. 110 °C) is generally recommended when using the solvent effect. Initial column temperatures from 80-100 °C were tried, but resulted in very complex chromatograms with multiple peaks and protracted solvent fronts. Neither MCP nor the internal standard, MAP, were completely resolved from endogenous sample components. The rapid rate of column temperature programming used in conjunction with the solvent effect also caused considerable column bleed resulting in significant baseline drift and inconsistent peak integration. For cold trapping, an initial column temperature low enough to allow for recondensation of the solute(s), vaporized in the injection port, at the head of the column is used. In general, compounds with boiling points about 150 °C higher than the column temperature will undergo cold trapping, while those with boiling points below this may be reconcentrated using a solvent effect (Freeman, 1981). Temperatures ranging from 180-210 °C were examined with 205 °C being selected for routine use. Temperatures below 195 °C resulted in the cold trapping of considerable endogenous sample components, in addition to MCP 132 and MAP, resulting in complex chromatography and interference, while temperatures in excess of 205 °C resulted in a decrease in both MCP and MAP area counts and peak heights. An initial column temperature of 205 °C was adopted as it provided good solute recoveries and the trapping of minimal endogenous components. Since peak shape and response of the sample are, in part, dependent upon the temperature of the injection port, the effect of varying this operating parameter was also studied. Unlike split injection, sample vaporization can take place over a longer period of time in the splitless mode. Lower injection temperatures may, therefore, be possible which is advantageous for the analysis of thermally labile compounds. For compounds with polar substituents, however, increasing injection port temperatures may result in higher yields of solute on the column (de Zeeuw et al., 1984). Temperatures ranging from 220-270 °C were examined. Increasing area counts for MCP and MAP were observed up to 260 °C and decreased again once this temperature was exceeded. An injection port temperature of 260 °C was selected as optimal, while 220 °C had been used in the previous assay. Column temperature can be a major factor affecting the efficiency of an assay as it has a direct effect on peak shape and analysis time. Increasing the column temperature, for example, results in a reduced analysis time but also decreases resolution. The thermal stability of the column is also an important consideration. It is necessary then, to weigh all of these factors (resolution, column thermal stability, analysis time) and to choose, i f possible, a temperature which optimizes each of them. In the current studies optimal resolution and analysis time were achieved using column temperature programming. Temperature programming rates ranging from 2-15 °C/min were evaluated. A rate of 4 °C/min to a maximum 133 column temperature of 240 °C resulted in the complete resolution of MCP and MAP from each other as well as from endogenous sample components. A relatively short run time (=13 min) was also obtained. A second column temperature increase at a rate of 15 °C/min for 1 min was added at the end of the run to purge nonvolatile components from the column following each sample injection. In the splitless configuration, flow of carrier gas through the inlet is the same as that through the column. This relatively slow rate of flow (generally 0.5-15 mL/min) allows sufficient time for the solvent to diffuse throughout the inlet contributing to a solvent ta i l . Peaks of interest eluting close to the solvent front may, therefore, be obscured. By backflushing (purging) the inlet at a predetermined time following sample injection with a large volume of carrier gas, the solvent which has diffused throughout the inlet will be vented, and solvent tailing minimized. Properly timed, mainly solvent is removed although small amounts of solute(s) may also be lost during the inlet purge. The amount of time selected before activating the inlet purge depends on the solvent used, column pressure and flow and the boiling points of the solutes of interest (Freeman, 1981). Purge activation time can also affect the quantitative transfer of solutes onto the column with premature activation resulting in solute loss in addition to excess solvent removal. Activation times ranging from 5-60 sec were examined (Fig. 1). No significant increase in MCP or MAP recovery (measured in terms of absolute area counts) occurred after 30 sec, while some sample loss resulted with activation times less than 20 sec. Based on these findings an inlet purge activation time of 30 sec was chosen for subsequent assays. Minimal solvent tailing was observed (Fig. 2). 134 A significant increase in sensitivity accompanied the switch from the split to splitless mode of sample injection necessitating an increase in the final reconstitution volume. The 0.2 mL sample reconstitution volume used in the previous assay resulted in some interference from endogenous compounds when used with the splitless assay. Sporadic interference from endogenous components was also observed with volumes of 0.4 and 0.6 mL, particularly for plasma extracts. A final reconstitution volume of 0.8 mL was selected for the assay as it resulted in no interference from extraneous compounds and s t i l l provided good sensitivity. The increased sensitivity gained through use of the splitless mode of sample injection has allowed the addition of a 2 ng/mL concentration point to the standard curve, down from the previous lower limit of 4 ng/mL. Other than an increase in the final reconstitution volume the sample extraction and derivative formation procedure remains the same as in the previous assay. The final difference between the previous assay and the current method involves the use of columns with a film thickness of 0.52 zzm. This change was necessary as columns with a phase thickness of 0.15 pi became temporarily unavailable, due to manufacturer quality control problems, early in the course of this project. The column phase, internal diameter and length remain unchanged. Other than an increase in retention time no change in chromatography occurred. The differences and similarities in GC operating conditions for the split and splitless assay methods are summarized in Table 1. Unlike the previous assay, the splitless mode of sample introduction has been found to be compatible with automatic sample injection providing good injection to injection and sample to sample reproducibility. The modified assay has been applied to the analysis of metoclopramide in maternal and fetal arterial plasma, tracheal and amniotic fluids 135 following both maternal and fetal drug administration. The method has been found to be specific and reproducible and offers improved sensitivity over the previous split injection assay. The use of the splitless injection technique coupled with a double extraction procedure and increased final sample reconstitution volume results in complete separation of MCP and MAP from each other as well as from extraneous sample components (Fig. 2). Linearity over a 2-40 ng/mL concentration range was observed for metoclopramide with standard curves having a coefficient of determination (r 2) of at least 0.98 (Fig. 3). The method has been found to be reliable with an overall average coefficient of variation of 6.4% (Fig. 3). Although an average coefficient of variation of 12.0% was obtained for the 2 ng/mL concentration point, this was considered acceptable, and quantitation to this lower limit was routinely used. Part of this variability is the apparent result of greater duplicate injection to injection area ratio differences than were obtained with split injections. Some general observations about the routine use of this assay include the following requirements: First, it was necessary to clean the electron capture detector following each =1000-1200 sample injections. At around this time linearity was lost, init ial ly at the top end of the standard curve range (viz., 32 and 40 ng/mL) and subsequently proceeding downward over a period of several days. Disassembling and washing the electron capture cell with a solution of ammonium bifluoride generally restored a linear response. Thermally cleaning the cell at temperatures 20-25 °C above the normal 350 °C operating temperature was generally ineffective. In some cases the cell had to be returned to the manufacturer (Hewlett Packard, Toronto, Ont.) for an abrasive cleaning in order to obtain a linear response over the desired concentration range. This loss of linearity appeared to be peculiar to MCP, since injection of the HFB-136 derivative of the antiarrhythmic agent, propafenone, at such times resulted in a normal response. Whether this loss of linearity is due to contamination of the cell with endogenous sample components or stripped column phase or both is not clear. Second, the column lifetime for MCP was approximately three months of continuous use (=5000 injections), at which time there was a marked decrease in MCP response. Interestingly, although these columns would no longer respond in a normal fashion for MCP, they could often s t i l l be used for several months by others in the laboratory for continued analysis {e.g. diphenhydramine). 4.2 Selection of Weighting Factor for Computer Data Analysis There has, it seems, always been considerable controversy over the selection of a suitable weighting factor in nonlinear least squares analysis of pharmacokinetic data. In many published pharmacokinetic studies it would appear that this weighting factor is often arbitrarily and subjectively chosen, frequently not stated and, i f so, usually is not justified. The choice of a regression model for the analysis of data by least squares methods involves two models (Sheiner and Beal, 1985), one structural (pharmacokinetic) and the other a variance model which governs the selection of weights. The observed value usually deviates from that predicted by the pharmacokinetic model by a random component given by the variance model. This statistical error accounts for biological intra-individual variability, misspecification of the pharmacokinetic model and analytical error. Recognizing that each observation in a set of pharmacokinetic data is not generally known with the same degree of certainty (variance), the use of weighted regression analysis has become commonplace (Boxenbaum et al., 1974). Various weighting algorithms are used in pharmacokinetic data analysis but most commonly the use of integer 137 powers of the reciprocal of the observed value is used as the weight: WTi = 1/Ci a , where WTi = the weight for the ith observation, Ci = the ith observation value and a = 0 (unweighted), 1 or 2. According to Peck et al., 1984, the weights in weighted least squares (WLS) analysis should ideally be nonrandom quantities that are proportional to the reciprocal of the variances of the measured drug concentrations. The problem in WLS analysis then, is how to choose an appropriate weighting factor (i.e. a). The use of unweighted data can only be justified i f the analyst believes that all observations are known with equal certainty. Usually, however, the observation variance is contaminated by multiple sources of random variation due to model misspecification, intra-individual variability and analytical error such that a different weighting factor is required to adequately describe the data. The problems and controversy associated with the choice of a suitable weighting factor have been extensively addressed by Peck et al., 1984, 1984a and Sheiner and Beal, 1985. Comparison of several least squares regression methods were made by these authors with the emphasis placed on the commonly used WLS technique and extended least squares (ELS) nonlinear regression, a rather recent alternative suggested for the computer analysis of pharmacokinetic data. Both methods were applied to simulated data sets and the results compared. In all instances ELS gave more accurate estimates of the pharmacokinetic parameters when compared to WLS, employing with the latter, equal weights, 1/C and 1/C 2. The problems associated with the variance models commonly used in WLS, the "a priori" assignment of a weighting factor to be applied to the variance model by the analyst, the limiting of this factor to an integer (generally 0, 1 or 2), the dependence of the variance model on observed values and their effects on pharmacokinetic parameter estimates are presented and discussed. Extended least squares analysis on the other hand does not 138 require that the weights be chosen, instead the analyst specifies a variance model to describe the random variability (unknown variances) in the pharmacokinetic data. Using ELS the parameters of the variance model and the pharmacokinetic parameters are estimated directly from the data alone during the computer run. Separate ("a priori") studies are not required to determine the (unknown) variances associated with the data. The use of ELS, then, may allow pharmacokinetic data to be f i t under a wider class of variance models than is now employed in WIS and may permit greater accuracy and precision in the estimation of pharmacokinetic parameters. Unfortunately, we do not have access to this new method of pharmacokinetic data analysis and, further, it has yet to be extensively applied to data obtained from actual kinetic studies. Knowing the controversy and problems associated with the choice of a suitable weighting factor in WLS (e.g. NONLIN) we decided to calculate the variance associated with the plasma concentration data obtained from our studies in pregnant and nonpregnant sheep. According to Pedersen, 1977, i f there is a functional relationship between the variances and the independent or dependent variables then the data may be weighted appropriately. Employing the method of Albert et al. , 1974 we were able to demonstrate a relationship between variance and concentration described by Eq. 7 (Tables 2 and 3). Since the calculated slope values were near or approached two, a weighting factor of 1/C2 was used in our nonlinear least squares data analysis (AUTOAN, N0NLIN74, N0NLIN84) for both the i.v. bolus and infusion studies. This method of determining a suitable weighting factor has also been used by Bloedow et al., 1980, in a study of lidocaine pharmacokinetics in pregnant and nonpregnant sheep following i.v. bolus dosing. The guidelines of Boxenbaum et al., 1974, for the nonlinear least squares analysis of pharmacokinetic data were also applied i.e. an 139 examination of minimum sum of squared residuals, residual versus calculated concentration (and time) plots (NONLIN 84), correlation between the observed and calculated concentration points and visual inspection of the scatter (fit) of observed data points about the theoretical curve. Accordingly the data was weighted equally (unweighted), 1/C and 1/C 2. A weighting factor of 1/C2 provided the best f i t of observed and model calculated concentrations supporting the method of Albert et al., 1974, that was used to demonstrate a functional relationship between variance and the dependent variable, concentration. A weighting factor of 1/C2 also provided the best agreement between computer generated pharmacokinetic parameters and those calculated by hand. 4.3 Metoclopramide Pharmacokinetics in Chronically Instrumented Pregnant and Nonpregnant Sheep The pharmacokinetics of MCP have been studied in the chronically instrumented pregnant ewe and fetus and in nonpregnant animals. Metoclopramide was administered to pregnant and nonpregnant ewes by i.v. bolus dosing over a 4- and 8-fold dose range, respectively. Paired maternal and fetal infusions, separated by 48-72 hours, were also conducted in several animals. Arterial blood from the ewes and fetus as well as amniotic and fetal tracheal fluid were serially collected for MCP analysis and pharmacokinetic study. Kinetic parameters were compared between the pregnant and nonpregnant ewes, with only total body clearance being significantly different (=24% higher) in the nonpregnant animals. Transfer of drug across the placenta from the ewe to the fetus following maternal dosing, or in the reverse direction following fetal administration, was rapid with significant concentrations in plasma in the f irst sampling interval. Metoclopramide was found to accumulate in fetal tracheal fluid 140 to concentrations =15-fold greater than in fetal plasma following either maternal or fetal drug administration. Accumulation and persistence in amniotic fluid also occurred, with an apparent elimination half- l i fe of ~7 hours compared to a value of -1-2 hours in maternal or fetal plasma. Recirculation of MCP may occur within the fetus via swallowing of amniotic and/or tracheal f luid, and via drug reabsorption across the pulmonary epithelium, amniotic and allantoic membranes, contributing to the persistence of the drug in the fetal lamb. To investigate this possibility, MCP was administered by intra-amniotic bolus injection. Metoclopramide appeared rapidly in the fetal circulation, accumulating in fetal lung fluid to the same degree as was observed following intravascular drug administration. Appearance of the drug in maternal blood was, however, delayed. The results of these studies indicate that MCP undergoes preferential and significant recirculation in the fetus as a consequence of fetal swallowing and uptake into the fetal circulation via the chorioallantoic membranes. 4.3.1 Maternal and Fetal Drug Concentration Ratios This study appears to be the f irst detailing the pharmacokinetics of metoclopramide involving serial sampling from mother and fetus, in utero, and also from nonpregnant animals (Riggs et al., 1988). The placental transfer of metoclopramide has been previously reported in man (Riggs, 1982; Arvela et al., 1983; Bylsma-Howel1 et al., 1983), but these studies were all based upon a single point blood sample from the umbilical cord and a maternal vein at the time of delivery. The fetal to maternal concentration ratio obtained from single point studies does provide an indication of the extent of fetal drug exposure, but the results are highly dependent on the time of sampling as illustrated in Fig. 4. Samples drawn 141 prior to the time that fetal plasma concentrations exceed maternal would yield a fetal/maternal ratio of less than one. This would lead to an entirely different interpretation of the extent of fetal drug exposure than i f the samples where taken after the point of crossover, at which time the ratio would exceed one. Since ethical and technical reasons preclude serial sampling in clinical human tr ials, predictions of the extent of placental transfer and fetal drug exposure are often based on a composite of data obtained from different subjects at different times and often under different conditions (Krauer et al., 1980). A better indicator of the extent of fetal drug exposure is the use of the ratio of the area under the fetal to maternal drug concentration versus time curve (Levy and Hayton, 1973; Mihaly and Morgan, 1984). The use of the chronically instrumented pregnant sheep model allows extensive serial blood sampling, permits the calculation of this ratio and thus provides a measure of the relative extent of fetal drug exposure throughout the experimental period. In the present study, the transfer of metoclopramide from the ewe to the fetal lamb was rapid, with significant concentrations in fetal plasma in the f irst sample (1 min after maternal drug administration; Fig. 4). In all but two instances, where a biexponential profile was obtained in maternal plasma, peak fetal concentrations were also reached at this time. The overall fetal to maternal AUC ratio obtained in the 10, 20 and 40 mg maternal i.v. bolus experiments was 0.74 ± 0.20 and indicates significant fetal exposure to metoclopramide (Table 6). The extent of fetal drug exposure is dependent upon maternal and fetal pharmacokinetic and placental factors. Each of these factors has been extensively reviewed (Krauer et al., 1980; Bogaert and Thiery, 1983; Mihaly and Morgan, 1984). For example, the increase in maternal body water and fat which generally occurs during pregnancy may result in a larger 142 volume of distribution for the drug in the mother. The pregnancy related decrease in plasma albumin as well as an increase in the concentration of endogenous binding displacers (e.g. free fatty acids) could potentially increase the amount of free drug available for diffusion across the placenta. The pharmacokinetics of the drug will determine whether or not a change in the amount of drug bound in maternal plasma will alter maternal, and subsequently fetal, unbound drug concentrations. A potential change in unbound concentration will only occur if the drug undergoes flow-dependent, non-restrictive elimination (Wilkinson and Shand, 1975; D'Arcy and McElnay, 1982; Mitani et al., 1987). Changes in maternal hepatic and/or renal clearances may also result in increased or decreased drug concentrations available for transfer to the fetus (Mihaly and Morgan, 1984). Similarly, the degree of fetal plasma protein binding, effect of a generally lower plasma pH and the ability of the fetus to metabolize and/or excrete drugs may also affect the drug levels reached in the fetal compartment (Mihaly and Morgan, 1984). The possibility of placental drug metabolism, processes of active transport and the physicochemical properties (e.g. l ipid solubility, molecular weight, pKa) of the drug itself may additionally play a role in altering the degree of fetal drug exposure. The distribution of umbilical blood flow in the fetus may also have an effect on fetal drug exposure. Approximately 55% of umbilical venous blood bypasses the fetal liver to enter the fetal systemic circulation via the ductus venosus, while =45% is distributed to the fetal liver (Rudolph, 1985). Should first-pass metabolism of the drug occur, then the amount of unchanged drug to which the fetal tissues are exposed will be reduced. The extent of placental transfer and fetal exposure to a number of other drugs has been investigated in pregnant sheep, in acute or chronically instrumented preparations, following bolus drug dosing or 143 infusion to steady-state. These include verapamil (Murad et a7., 1985), meperidine (Szeto et al., 1980), methadone and morphine (Szeto et al., 1982), propranolol (Mihaly et al., 1982), cimetidine (Ching et al., 1985), indomethacin (Anderson et al., 1980), bupivicaine (Kennedy et a7., 1986), lidocaine (Morishima et al. , 1979; Bloedow et a7., 1980; Friesen et al., 1986), midazolam (Vree et a7., 1984), acetaminophen (Wang et a7., 1986), and diphenhydramine (Yoo et a7., 1986). Only in the latter case was the fetal/maternal AUC ratio for the drug calculated, with an average value of 0.85. In the other studies, fetal/maternal drug concentration ratios were reported; these averaged =0.15 and =0.35 for midazolam and verapamil, respectively, following bolus dosing, and =0.04 for cimetidine, =0.13 for morphine, =0.14 for methadone, =0.28 for indomethacin, =0.35 for bupivicaine and =0.77 for acetaminophen at steady-state following constant rate infusions. For both metoclopramide and diphenhydramine, the average fetal/maternal drug concentration ratios after bolus dosing (0.70 ± 0.43 and 0.90 ± 0.25, respectively) agree quite well with the fetal/maternal AUC ratios. As previously mentioned, the apparent differences in the extent of fetal exposure to the various therapeutic agents noted above could result from differences in placental permeability, or from maternal or fetal pharmacokinetic factors. Some of these factors are illustrated in the following three examples. In the acute maternal i.v. bolus study of verapamil (Murad et al., 1985) fetal hepatic extraction and a higher degree of maternal protein binding were suggested as possible reasons for the low fetal/maternal ratio (=0.35) obtained. Although a higher fetal/maternal ratio might be expected due to ion trapping of verapamil (pKa 8.5) in the fetal circulation, this can be limited by a high degree of protein binding in maternal plasma. In the study of cimetidine (Ching et a7., 1985) a 144 maternal to fetal plasma concentration gradient of 25:1 was maintained throughout a continuous six day maternal i.v. infusion of the drug. Differences in maternal and fetal plasma binding were minimal, suggesting that differential protein binding was not involved in the observed concentration gradient. Fetal elimination of cimetidine was also reported to be negligible. The authors postulate that active transport of cimetidine from the fetus to the ewe or significant placental drug metabolism may account for the observed maternal and fetal plasma concentration difference. The maternal and fetal disposition of bupivacaine (Kennedy et al., 1986) was also studied during a one hour maternal infusion to steady-state. Fetal to maternal ratios were considerably less than unity throughout the five hour period of study. Neither the ewe nor the fetus metabolized bupivacaine, but renal excretion was found to be an important route of fetal drug elimination. These authors concluded that the major factor responsible for the low fetal/maternal ratio (=0.35) was related to differences in maternal (=85%) and fetal (=40%) protein binding. This results in a greater concentration of free drug in the fetus and a net transfer back to the ewe when the infusion is stopped. While it appears from these three studies that some of the differences in fetal/maternal concentration ratios for various drugs may be due to placental as well as maternal or fetal pharmacokinetic factors, it is also possible that some of the differences in apparent fetal drug exposure are the result of differences in the method of drug administration. We base this suggestion on results that have been obtained with metoclopramide and also diphenhydramine (Riggs et al., 1988). When these drugs are infused to steady-state in pregnant ewes, the fetal/maternal concentration ratios at steady-state (0.57 ± 0.14 [Table 10] 145 and 0.19 ± 0.04 [unpublished data], respectively) are much lower than the values obtained after bolus injection. Gibaldi, 1969a, has shown this mathematically for a two compartment open model with elimination occurring from the central compartment as illustrated in Scheme 2. In our situation, the central compartment is represented by the ewe, with the fetus as the peripheral compartment. During infusion equilibrium the ratio of the amount of drug in the peripheral compartment (Xp) to that in the central compartment (Xc) is equal to k 1 2 / k 2 1 ' w n i l e f ° r a n i - v - bolus the following relationship exists: Xp/Xc = ^22/(^21 - /*) • Consequently, this ratio will always be greater than that obtained during the infusion. This difference in the fetal/maternal concentration ratios, following bolus versus infusion administration, illustrates that the comparison of the extent of fetal drug exposure to different drugs may only be appropriate when a similar drug administration protocol is employed. Fetal to maternal concentration ratios were also determined at each time point in the 19 i.v. bolus experiments with an overall ratio of 0.70 + 0.43. The fetal/maternal concentration ratios increased continuously over the period of study, with fetal concentrations exceeding maternal before the end of the sampling period in 13 of the experiments. In these 13 animals, the fetal/maternal concentration ratio averaged 1.25 ± 0.19 following the point of crossover (between 1 and Z\ hours after maternal dosing). These findings are consistent with the computer simulations conducted by Levy and Hayton, ,1973, in which the mother and fetus were each represented as a single compartment in the two compartment system shown in Scheme 2, with the fetal compartment being pharmacokinetically "deep" or slowly accessible. A similar situation is presented by Waddell and Marlowe, 1981, where the fetal compartment is considered to be large, resulting in a considerable delay in the time for maternal and fetal concentrations to equilibrate. Following equilibration, fetal concentrations exceed maternal and remain higher as the fraction of the fetal compartment cleared by transfer to the mother is small relative to the size of the fetal compartment. Fetal/maternal ratios greater than one may also be obtained as a result of a pH gradient which would trap the drug in the fetal compartment (Waddell and Marlowe, 1981). Similarly, more extensive binding of the drug to fetal than to maternal plasma (Levy, 1981; Waddell and Marlowe, 1981) may result in a ratio exceeding one, although this does not appear to be the case with MCP which is =39% bound in fetal plasma and =49% bound in maternal plasma at steady-state (Tables 10 and 14). 4.3.2 Metoclopramide Pharmacokinetics in the Pregnant Ewe Metoclopramide maternal concentration-time curves followed a biexponential decline (Fig. 4) similar to that seen in man (Graffner et al., 1979; Bateman et al., 1980; Ross-Lee et al., 1981). The volume of distribution at steady-state was calculated using both compartmental (Eq. 17) and noncompartmental (Eq. 18) methods and show good agreement (Table 5), with the noncompartmental values averaging =6.2% lower than the compartmental estimates. The close agreement of these two methods of calculation lends support to the premise that the data follow biexponential kinetics (Eq. 27) and are adequately described by the two compartment model illustrated in Scheme 2. While the fetus may be part of the peripheral compartment illustrated in Scheme 2, more complex models encompassing separate peripheral compartments in the ewe and/or the fetus with differing degrees of drug accessibility may also be involved (Levy and Hayton, 1973; Krauer et al., 1980). According to Levy et al., 1969, it may be "virtually impossible, in most instances, to distinguish between" a two-or-more-compartment system "on the basis of plasma concentrations alone", unless 147 perhaps the drug displays very prominent distribution and elimination phases and an intensive blood sampling protocol is used. The area under the maternal plasma concentration-time curve showed a good correlation with increasing dose indicating that the drug followed dose-independent kinetics over the 4-fold dose range (10-40 mg) studied in the pregnant ewes. This is further supported by a lack of significant changes in other pharmacokinetic parameters such as /3, t^ , CL S , V a r e a , kjg, k 2^ or k2^ with increasing dose. A further test of linearity involves a comparison of the averaged pharmacokinetic parameters obtained at each dose. It has been suggested that dose-related trends in the magnitude of averaged pharmacokinetic parameters may provide strong evidence of nonlinear kinetics (Wagner, 1973). Consistent patterns were not observed in the current studies for most parameters (Tables 4 and 5), however, the estimate for V $ s was found to be significantly larger at the 40 mg dose than at either the 10 or 20 mg doses, and although not significantly different k^g, kj2 and k2j tended to decrease with increasing dose. While these differences may be indicative of possible nonlinear kinetics it is also very possible, at least in the current studies, that the variations in V s s , k 1 0 , kj 2 and k2j are due to the considerable intrasubject variability seen in the individual pharmacokinetic parameters obtained at the various doses on different experimental days. Recent studies by Runciman et al., 1984, 1984a, using a chronic sheep preparation, discuss and illustrate the effects of day to day variations in regional blood flows (e.g. renal, hepatic) on drug disposition. Linearity of MCP disposition kinetics have also been reported in healthy male volunteers by Wright et al., 1984, at doses between 20 and 100 mg, and more recently in our laboratory at doses between 5 and 20 mg (Wright et al., 1988). Dose-independent kinetics have also been observed in cancer chemotherapy patients receiving high-dose 148 (greater than 0.5 mg/kg/h) MCP therapy (Taylor et al., 1984; Bryson et al., 1985). The apparent terminal elimination half- l i fe of MCP in pregnant sheep is considerably shorter (1.2 h) than that in man (=4.3 h) (Bateman, 1983). The volume of distribution in the pregnant ewe (5.7 L/kg) is higher than that in man (=2.9 L/kg), as is total body clearance (3.5 versus =0.5 L/h/kg) (Bateman, 1983). Similar values for elimination half- l i fe (1.1 h), and s t i l l higher values for volume of distribution (6.9 L/kg) and total body clearance (4.5 L/h/kg) were obtained in the nonpregnant ewes (Tables 7 and 8). Other studies also report a more rapid clearance of drugs such as diphenhydramine (Yoo et al., 1986), meperidine (Szeto et al., 1978), lidocaine (Bloedow et al., 1980), and etidocaine (Pedersen et al., 1982) compared with man. This more rapid clearance may be due in part to a higher hepatic blood flow in sheep (=0.5-3.0 L/min) (Altman and Dittner, 1972) compared to man (0.5-1.5 L/min) (Altman and Dittner, 1972). Given the much higher clearances of metoclopramide and other drugs in sheep, though, it seems likely that other factors are involved. 4.3.3 Metoclopramide Pharmacokinetics in the Fetus The elimination half- l i fe of metoclopramide in the fetus was significantly longer than in the ewe (Tables 4 and 6). This may be due to the fetal compartment being slowly accessible or "deep" (Levy and Hayton, 1973; Levy, 1981) resulting in a slow release of drug from the fetal tissues and circulation (Anderson et al., 1980a) during the terminal elimination phase in the mother. Fetal elimination half- l i fe can also be prolonged as a result of extensive drug binding in fetal plasma (Waddell and Marlowe, 1981) although this does not appear to be the case with metoclopramide since binding in both maternal and fetal plasma is relatively low (=49% versus =39%, respectively, at steady-state; Tables 10 149 and 14). The existence of a pH gradient could also cause the drug to be trapped in the fetal compartment (Waddell and Marlowe, 1981), but this seems unlikely since the pKa of metoclopramide is high (9.3) relative to the pH of maternal and fetal plasma, resulting in virtually complete ionization of the drug at physiological pH. We have also found metoclopramide levels in fetal lung fluid that are 10-15-fold higher than in fetal plasma (Table 23) (Riggs et a7., 1986, 1987). This is associated with high drug concentrations in amniotic fluid (Table 23), with measurable amounts of the drug persisting for up to 28 h after drug infusion to steady-state (Figs. 9 and 10). A study by Harding et a7., 1984, demonstrated that during the last third of gestation, fetal lambs swallow an average of 98-577 mL of mixed tracheal and amniotic fluids in a 24 h period with volumes greater than 1000 mL also being observed. Given that these swallowing bouts occur regularly (on average every 2.3 h; Harding et a7., 1984a), and that the concentrations of metoclopramide observed in fetal tracheal and amniotic fluids are high, it is also possible that fetal half- l i fe could be prolonged as a result of recirculation of the drug via fetal ingestion of these fluids (Harding et a7., 1984) and absorption from the gastrointestinal tract. The volume of fluid within the fetal lung, between 125 and 145 days gestation, is =30 mL/kg of fetal body weight (Olver and Strang, 1974). In our experiments this would represent a total lung volume of =60 mL based on an =2 kg average fetal weight (Tables 9 and 28). Since MCP appears to move from fetal vessels across the pulmonary epithelium to concentrate in lung liquid, this fluid reservoir may represent another "deep" compartment from which drug is slowly released back into the fetal circulation, contributing to the longer half- l i fe observed in the fetal lamb. It is also possible 150 that MCP binds to lung tissue and is slowly released into the fetal circulation prolonging the half- l i fe. Amniotic fluid also represents a very deep compartment (Reynolds, 1981) from which as previously mentioned, MCP could undergo recirculation within the fetus via fetal swallowing and, may also reenter the fetal circulation via diffusion across the chorioallantoic vasculature. The volume of amniotic fluid at 115-125 days gestation has been estimated to be =1000 mL (Tomoda et al., 1985). The long half- l i fe observed for MCP in amniotic fluid (=7 h) may also be indicative of the depth of this compartment. Another potential drug reservoir in sheep is the allantoic cavity, with a fluid volume =450 mL at 115 days gestation increasing to =750 mL at term (Mellor and Slater, 1971). Again, drug could concentrate in this fluid compartment and reenter the fetal circulation by diffusion across the vascularized chorioallantoic membrane. The presence of significant concentrations of metoclopramide in amniotic and allantoic fluids may also indicate the ability of the fetus to eliminate metoclopramide renally as fetal urine enters amniotic fluid via the urethra and allantoic fluid via the urachus. As the fetal bladder was not catheterized the degree of urinary excretion of metoclopramide by the fetus remains to be evaluated. The ability of the fetal liver to metabolize MCP could also have an effect on the drug's elimination half- l i fe. A study by Dvorchik et al., 1986, examined the ability of hepatic microsomes obtained from pregnant ewes and their fetuses, between 120-128 days gestation, to metabolize (dealkylate, hydroxylate, glucuronidate) a number of drugs. While considerably lower than maternal microsomal activity, hepatic microsomal fractions from the fetal lamb demonstrated the ability to demethylate methadone and meperidine and to form a glucuronide with morphine. Unlike 151 maternal fractions, fetal microsomes displayed no detectable activity for hexobarbital or benzo[a]-pyrene hydroxylation. Metoclopramide undergoes 0-demethylation and N-deethylation in the rat, rabbit and dog (Arita et al., 1970; Bakke and Segura, 1976; Cowan et a7., 1976) with mono-deethylated MCP as a major metabolite. Conjugation with glucuronic acid and sulphate has also been reported in the dog and rabbit (Arita et a7., 1970; Cowan et a7., 1976; Bateman et a7., 1978) as well as in man (Teng et a7., 1977; Bateman et a7., 1980). In a qual itative study of urine collected from one of the ewe's receiving a i.v. bolus dose of MCP, mono-deethylated MCP appears to be a major urinary metabolite (See Appendix). A peak eluting just prior to MCP in plasma extracts from this ewe was also identified as the mono-deethyl ated metabolite of MCP. In some instances in fetal plasma, a very small peak with the same retention time as the mono-deethylated compound in maternal plasma was also observed, however, insufficient fetal plasma was available for GC mass spectrometric study. While the presence of this peak may indicate the ability of the fetus to deethylate MCP it may well be the result of placental transfer of this compound from the ewe. The significance of these qualitative findings with regard to possible hepatic metabolism of MCP by the fetal liver and the degree to which this may reduce fetal drug exposure or alter fetal half - l i fe, however, remains to be studied. While in vitro metabolic activity for the fetal liver has been demonstrated in a number of animal species as well as man it is generally felt , at least in animals, to be very low compared to adult values (Waddell and Marlowe, 1976; Juchau et a7.,1980; Pelkonen, 1980; ). As always, caution must be exercised in extrapolating the results of in vitro metabolic studies to the clearance of drug in the intact preparation. There appears to be only a single study examining the possible hepatic metabolism (uptake) of drugs by the ovine fetal liver in vivo. 152 Mihaly et al., 1982, studied the maternal and fetal hepatic extraction of propranolol in an acute, anesthetized sheep preparation during maternal drug infusion. Fetal arterial, umbilical and hepatic vessels were sampled. Extraction by the right lobe of the fetal liver was negligible, while =33% of the drug was extracted by the left lobe. Compared with the ewe, where only 3-4% of the drug escaped hepatic uptake, fetal hepatic extraction was low. These authors comment that there is evidence that general anesthesia may affect the distribution of umbilical venous blood within the fetal l iver. Considering this fact and the acute nature of these experiments, these findings also need to be interpreted with caution. It would certainly appear that the most important route of drug elimination in the fetus is via transfer back across the placenta to the maternal circulation (Reynolds, 1979; Levy, 1981). 4.3.4 Comparison of Pharmacokinetic Parameters in Pregnant and Nonpregnant Ewes The comparison of pharmacokinetic parameters between pregnant (Tables 4 and 5) and nonpregnant ewes (Tables 7 and 8) at the 10, 20 and 40 mg doses indicated a significant difference only in total body clearance. Clearance of metoclopramide by the nonpregnant animals is =24% higher than in the pregnant animals. This difference was not associated with consistently higher values of volume of distribution ( V a r e a ) , the terminal elimination rate constant (/J), or AUC. In humans, following the i.v. administration of metoclopramide, approximately 80% of the dose is recovered in the urine in 24 h (Desmond and Watson, 1986), with about 20% as unchanged drug and the balance as metabolites (Bateman et a l . , 1980; Bateman, 1983). The renal clearance of metoclopramide is low (2.6 mL/min/kg) accounting for approximately 20% of total body clearance (11.61 153 mL/min/kg) (Bateman et al., 1980; Harrington et al., 1983). The total body clearance of metoclopramide approximates hepatic blood flow, suggesting that its clearance is probably limited by liver blood flow rather than by hepatic metabolic capacity (Harrington et al., 1983). Hepatic metabolism is thought to be the major route of metoclopramide elimination (Desmond and Watson, 1986). While renal blood flow is increased in human pregnancy (Krauer et al., 1980; Mihaly and Morgan, 1984) this does not appear to be the case in sheep (Rosenfeld, 1977). Rosenfeld, 1977, compared organ and tissue blood flows in nonpregnant sheep to those in pregnant ewes in the last third of gestation (130-140 days) using a radiolabelled microsphere technique. No significant changes in either renal weight or blood flow were observed. Hepatic arterial blood flow, however, was significantly lower in pregnant animals (9.44 mL/min) than in the nonpregnant ewes (16.30 mL/min); liver weights were not significantly different. There were no measurements of hepato-portal or intestinal blood flow in these animals. If the major route of metoclopramide elimination in sheep is via the liver, then it is possible that the significantly lower total body clearance observed in the pregnant ewes could be due, in part, to a decrease in liver blood flow, assuming that portal blood flow is also decreased during pregnancy. Changes in hepatic enzyme activity or capacity could also be involved but may not be important in this case since the hepatic clearance of metoclopramide appears to be flow-dependent (Harrington et al., 1983). In pregnant sheep, the volume of distribution of lidocaine is significantly higher and the total body clearance unaltered when compared to nonpregnant animals, resulting in a significantly increased half- l i fe (Bloedow et a l . , 1980). Other studies showing a decreased drug clearance during pregnancy include: phenytoin (Chou and Levy, 1984) and antipyrine (Stock, 1984) in rats, phenytoin in the rhesus monkey (Stock, 1984) and theophylline 154 (Gardner et al., 1987) and antipyrine (Chiba et al., 1982) in man; all apparently due to altered hepatic metabolism. Comparison of total body clearance as a percentage of cardiac output in the pregnant and nonpregnant ewe was also made for the 10, 20 and 40 mg bolus doses. The overall average total body clearance of MCP from the pregnant ewes was 3.5 ± 1.0 L/h/kg and accounts for =39% of cardiac output (=148 mL/min/kg (=8.9 L/h/kg); Rosenfeld, 1977). Mean total body clearance in the nonpregnant ewes on the other hand, was 4.6 ± 1 . 4 L/h/kg and accounts for =105% of cardiac output (=74 mL/min/kg (=4.4 L/h/kg); Rosenfeld, 1977). The phenomenon of drug clearance in excess of cardiac output following i.v. administration has been discussed by Collins and Dedrick, 1982, with the observation that the lung is the only organ other than blood which can account for clearance in excess of cardiac output. 4.3.5 Metoclopramide Pharmacokinetics in the Nonpregnant Ewe As observed in the pregnant ewes, metoclopramide plasma concentrations also followed a biexponential decline (Fig. 6) in the nonpregnant animals. The volume of distribution at steady-state, V s s , was also calculated using both compartmental (Eq. 17) and noncompartmental (Eq. 18) methods and, like the values for the pregnant ewes, showed good agreement (Table 8), with the noncompartmental estimate averaging =5.5% lower. The close agreement of these two V $ s estimates again supports the assumption that concentration data follow biexponential kinetics (Eq. 27) and are satisfactorily represented by the two compartment model illustrated in Scheme 2. Linear or dose-independent kinetics where also observed in the nonpregnant ewes over the 8-fold dose range studied (10-80 mg) with a good correlation between AUC and dose. A lack of significant changes in 155 other pharmacokinetic parameters (a, /5, t^a, t^ , CL S , V s s , V a r e a , k^ o» and k 2i) also supports dose independent kinetics over this dose range. 4.4 Metoclopramide Pharmacokinetics in the Chronically Instrumented Ewe and Fetus Following Infusions to Steady-State. 4.4.1 Maternal MCP Infusions 4.4.1.1 Fetal Exposure to Metoclopramide As observed in the i.v. bolus experiments, transfer of MCP to the fetal lamb was rapid with significant concentrations in the f irst sample (5 min) in all cases (e.g. Fig. 7). The extent of fetal drug exposure was also assessed in these experiments, with an average fetal to maternal concentration ratio of 0.57 ± 0.14 (for total drug concentration) at steady-state (Table 10). This was considerably lower than the overall average ratio obtained in the i.v. bolus studies (0.70 ± 0.43). A steady-state fetal to maternal concentration ratio of 0.67 ± 0.18 was determined for unbound drug (Table 10). Since unbound drug is felt to be responsible for pharmacological effect, the determination of fetal to maternal concentration ratios for total drug will lead to entirely different conclusions regarding the extent of fetal drug exposure than ratios determined using unbound drug concentrations. Fetal to maternal MCP concentration ratios increased over time, reaching constant values during steady-state and then increased continuously following termination of the infusion. Again, these observations are consistent with the computer simulations conducted by Levy and Hayton, 1973, where the ewe and fetus were each represented as a single compartment of a two compartment model (Scheme 2, where the ewe is represented by the central compartment and the fetus by the peripheral compartment), with the fetus being slowly accessible or "deep". As illustrated in Fig. 7, once the infusion is 156 stopped, a redistribution of drug occurs between the maternal and fetal compartments. At some time after stopping, the infusion the amount of drug in the fetal compartment becomes greater than that in the maternal compartment, and fetal to maternal concentration ratios exceed one. Following termination of the infusion the fetal to maternal concentration ratio averaged 1.57 ± 0.40, with fetal MCP concentrations exceeding maternal between 5 to 90 min post-infusion in all 8 experiments (e.g. Fig. 7). The index of relative fetal drug exposure was also calculated for the maternal MCP infusion experiments with an average AUCf/AUCm of 0.91 + 0.27 which is considerably higher than that determined in the i.v. bolus studies (0.74 + 0.20). The maternal and fetal pharmacokinetic and placental factors governing the extent of fetal drug exposure have been previously discussed for the i.v. bolus experiments (Section 4.3.1) and apply equally well here. As illustrated in Fig. 7, maternal MCP concentrations decrease following the initial i.v. bolus loading dose and then stabilize once the rate of infusion is equal to maternal plus placental clearance. Fetal concentrations rise and stabilize at a lower steady-state concentration once the rate of loss from the fetal circulation is equal to the rate of placental transfer of MCP from the ewe. The lower steady-state concentration of MCP in fetal plasma occurs partially because fetal tissues are clearing MCP from the fetal circulation in a manner similar to that seen in studies with acetyl salicylic acid by Anderson et al., 1980a, 1980b. This removal from the fetal circulation may be due to metabolism, excretion (e.g. renal, lung) as well as tissue uptake. The l ipid insoluble, ionized form of MCP (pka 9.3) predominates at physiological pH and may also act as a barrier to placental transfer, in part explaining the -2-fold concentration difference between the ewe and fetus at steady-state. 157 Another possible reason for the concentration difference between the ewe and fetus at steady-state involves protein binding. In a recent report, Hill and Abramson, 1988, illustrated that protein concentrations are a major determinant of the difference in the degree of drug binding to maternal and fetal plasma. Using literature values for human adult and fetal protein concentrations, of both albumin and ai-acid glycoprotein, as well as the percentage of maternal plasma protein binding, they were able to predict both the degree of binding in fetal plasma and a fetal to maternal steady-state concentration ratio. In both cases the correlation between predicted and known literature values was good, leading the authors to conclude that plasma protein binding is the primary factor responsible for establishing fetal to maternal steady-state drug concentration ratios. 4.4.1.2 Maternal Metoclopramide Pharmacokinetics Maternal MCP concentration versus time profiles followed a biexponential decline after stopping the infusion in all 8 ewes, as illustrated in the representative semilogarithmic plot in Fig. 7. The two compartment open model described by this process is shown in Scheme 2 with the ewe represented by central compartment and the fetus by the peripheral compartment. As previously mentioned for the i.v. bolus experiments (Section 4.3.2), more complex models, including separate maternal and/or fetal tissue, and placental compartments, may be involved. This degree of detail is not apparent from the plasma data, even in the two maternal infusion studies, where post-infusion sampling was continued for 28 hours (e.g. Fig. 9). The apparent triexponential decay of MCP in fetal tracheal fluid (Fig. 9) may be indicative of a third compartment. This is purely speculative, since tracheal fluid samples were only available in one of 158 these two experiments. Further studies employing a prolonged sampling period are required to confirm this single observation. The volume of distribution at steady-state, V s s , was also determined by both compartmental (Eq. 17) and noncompartmental (Eq. 19) methods for the maternal infusions. Unlike the close agreement obtained in the i.v. bolus studies, the noncompartmental estimate was an average 32.4 ± 5.7% lower than the compartmental value. The reasons for this could be twofold. One, the theoretical factor for mean residence time (t^) in the noncompartmental equation (Eq. 19) derived by Perrier and Mayersohn, 1982, for a simultaneous i.v. bolus loading dose and constant rate infusion, may not adequately estimate steady-state volume of distribution when applied to a practical problem. Two, it may be difficult to accurately determine pharmacokinetic parameters following i.v. infusion for those drugs which do not display prominent two compartment characteristics following rapid i.v. bolus injection (Gibaldi and Perrier, 1982). Gibaldi and Perrier, 1982, state that the larger the ratio for the zero-time intercepts A/B following i.v. bolus dosing, the easier it is to discern two compartment characteristics for a drug. As the intercept, A, for the a (distributive) phase approaches zero, the ratio of A/B also approaches zero and the plasma concentration versus time curve becomes monoexponential {i.e. a one compartment model). This may also occur i f A is very large, as the plasma concentrations in the distributive phase may decline in an apparent monoexponential fashion over several orders of magnitude, such that the terminal elimination (/}) phase is not observed prior to reaching the assay sensitivity limit for the drug. These authors provided examples of A/B ratio values for two drugs described by a two compartment model. The distribution and elimination rate constants, a and /J were the same for both drugs. In the f irst case, an A/B ratio of 0.3 resulted in an apparent 159 monoexponential plasma concentration versus time profile. In the second instance, an A/B ratio of 300, resulted in well defined distribution and terminal elimination phases. When a drug is administered by i.v. bolus injection the ratio of A/B is equal to [a - )/(k2j - 0) while the corresponding R/S ratio following infusion of a drug to steady-state is equal to [(a - V,2\)/{^2\ ' 0)](0/ a ) (Gibaldi and Perrier, 1982). Since /J is by definition smaller than a, then the ratio of R/S will always be smaller than the ratio of A/B. Unless the ratio of A/B is large, then the ability to distinguish the biexponential characteristics of the drug following an infusion is usually diminished. The A/B ratio was calculated for MCP using the computer generated A and B intercepts obtained for the 10, 20 and 40 mg i.v. bolus dose experiments, with an overall average value of 3.5 ± 1.2. This ratio is certainly not very large and by comparing Fig. 4, representative of i.v. bolus dosing, with Fig. 7 for one of the infusion experiments, it can be seen that the biexponential decay is more clearly defined following i.v. bolus dosing. This can also be seen by comparing the distribution half-lives (tj, f t), with averages of 8.4 ± 4.5 and 4.6 ± 1.5 min for the i.v. bolus and infusion experiments, respectively (Tables 4 and 11). On this basis then, it is possible that the computer generated estimates of the post-infusion R and S intercepts and compartmental rate constants are not as accurate as the rate constants and A and B intercepts obtained for the i.v. bolus experiments. Significant differences in these parameter estimates and their use in the compartmental equation (Eq. 17) for V s s estimation then, could result in the poor agreement we obtained. A comparison of overall average values for a, /J, kj 2 and k 2j for the bolus experiments (Table 4) shows relatively good agreement with those calculated for the infusions (Table 11). The estimate of k^ Q for the bolus studies is approximately one-half the maternal infusion value. It is interesting to note, however, that the average compartmental V s s estimate obtained for the maternal infusions (4.5 ± 1.4 L/kg), agrees very well with both the compartmental (4.8 ± 1 . 0 L/kg) and noncompartmental (4.5 ± 0.9 L/kg) values determined for the ewe in the i.v. bolus studies (Table 5). We also attempted to use the method of residuals (Gibaldi and Perrier, 1982) to determine the post-infusion R and S intercepts and a and /J by hand, and to use these values to calculate compartmental rate constants (kin, kj2> ^2^), V c (Eq. 16) and subsequently V s s (Eq. 17) for comparison with the computer generated estimates. Although it was possible to determine /} and the /3-phase intercept, S, we were unable to reliably estimate a and the corresponding R intercept as no samples had been drawn prior to 5 min, by which time the distribution phase was virtually complete (tj,ft, 4.6 ± 1.5 min). A more intensive sampling time protocol (e.g. 1, 2, 4, 6, 8, 10 min, etc.) may have solved this problem. Such a protocol would have resulted in truncation of sampling at later, more important time points due to sample volume constraints, particularly in the fetus. Agreement of the overall average estimates for CL S , V a r e a and t ^ obtained following bolus dosing (Tables 4 and 5) and infusion to steady-state (Tables 11 and 12) is also good. 4.4.1.3 Metoclopramide Pharmacokinetics in the Fetus Unlike maternal post-infusion concentration profiles, fetal plasma concentration versus time curves were biexponential in only two cases and monoexponential in the remaining six (Table 13). The observation of a monoexponential process in six of the fetal lambs may be indicative of the depth of the fetal compartment resulting in a slow redistribution of MCP from fetal tissues such that a distribution phase was not evident. Also, 161 as the f irst sample was taken 5 min after the infusion was stopped a short distribution phase in these fetuses may have been missed. A more intensive blood sampling protocol such as that used in the i.v. bolus experiments (1, 3, 6, 10 min, etc.) may have overcome this difficulty but was not used in the infusion experiments in order to reduce the amount of fetal blood sampled. As observed in the maternal i.v. bolus dosing experiments fetal elimination half- l i fe was significantly longer (122.9 ± 31.0 min) than that in the ewe (64.7 + 15.3 min). As previously discussed for the i.v. bolus experiments (section 4.3.3) a deep fetal compartment from which MCP is slowly released probably accounts for the longer half- l i fe in the fetal lamb. Other factors that may prolong fetal elimination half-l ife such as extensive binding to fetal plasma protein, ion trapping in the fetal circulation, drug recirculation via ingestion of tracheal and amniotic fluids and reabsorption across the pulmonary epithelium and chorioallantoic membranes have also been previously discussed and apply equally well here. 4.4.2 Fetal MCP infusions 4.4.2.1 Metoclopramide Pharmacokinetics in the Ewe and Fetus Transfer of MCP across the placenta to the ewe was also rapid with peak concentration in the f irst sample (5 min) in six of the 8 experiments (e.g. Fig. 8). In the remaining two ewes peak MCP concentrations were reached at 15 and 30 min. Fetal concentrations decreased from their peak concentration, at 5 min, following the initial i.v. bolus loading dose and reached steady-state concentrations (between 30-45 min) once the rate of infusion was equal to fetal elimination plus transfer across the placenta to the ewe. Maternal MCP plasma concentrations followed a similar profile in most cases and reached steady-state concentrations between 15 and 45 minutes. The much larger absolute volume of distribution in the ewe is 162 probably the most important factor responsible for the lower maternal steady-state drug concentrations. Although maternal elimination (e.g. renal, metabolic) and tissue uptake are also likely involved. The marginally lower plasma pH in the fetus (=0.1 unit) results in almost complete ionization of MCP (pKa 9.3) at physiological pH. The predominance of this l ipid insoluble form may also act as a barrier to placental transfer, explaining in part the =20-fold steady-state concentration difference between the fetus and ewe (Table 14). Fetal MCP concentration versus time profiles followed a biexponential decay after cessation of the infusion in all 8 fetal lambs (e.g. Fig. 8). The two compartment model representing this process is illustrated in Scheme 2 with the fetus comprising the central compartment and the ewe the peripheral compartment. As previously stated, more complex models may be involved but this was not apparent from the plasma data. In the single experiment where sampling was continued for 28 hours post-infusion (Fig. 10), MCP fetal plasma concentrations approached the limit of the assay (=2 ng/mL) 12 hours after stopping the infusion. When the next sample was taken at 20 hours post-infusion, there was a "reappearance" of drug with plasma concentrations remaining at =10-12 ng/mL until the end of the experiment. This single observation remains to be confirmed by similar studies involving prolonged plasma sampling, but may represent the release of MCP into the fetal circulation from a very deep fetal tissue compartment. Recirculation via fetal swallowing of amniotic and tracheal fluids, in which the drug concentrates to a significant extent (Table 23), and/or reabsorption across the pulmonary epithelium and chorioallantoic membranes could also be, in part, responsible. As previously observed in one maternal infusion (Fig. 9), a triexponential concentration versus time profile for MCP in tracheal fluid was obtained in one of the fetal infusion experiments (Fig. 10). This may suggest that another compartment is involved. Once again confirmation of this observation requires further experimentation. The steady-state volume of distribution for MCP in the fetal lamb was also estimated using compartmental (Eq. 17) and noncompartmental (Eq. 19) methods. As in the ewe, in the maternal infusion experiments, the fetal noncompartmental estimate (11.4 ± 3.5 L/kg) averaged 38.9 ± 5.5% lower than the compartmental value (15.9 ± 5.2 L/kg) (Table 16). As previously discussed (Section 4.4.1.2), this difference may be due to deficiencies in the factor for mean residence time (t^) in the noncompartmental equation (Eq. 19) or to error in the estimation of a and the corresponding distribution phase intercept R, due to insufficient sampling during the 10 to 15 min drug distribution period (t^a = 12.1 ± 3.1 min). Again, because more samples had not been taken prior to 5 min, it was not possible to reliably estimate a and R by hand, using residuals, to compare with the computer generated results. Unlike the ewe, we have not conducted i.v. bolus experiments in the fetus so are unable to make any Vss comparisons. The terminal elimination half- l i fe (t^) was significantly longer in the fetus than in the ewe (64.7 ± 15.3 min) following the fetal drug infusions, averaging 115.9 ± 44.3 min (Table 15). A very similar post-infusion t^g estimate (122.9 ± 31.0 min) was also obtained following the maternal infusions (Table 13). As already mentioned in the i.v. bolus (Section 4.3.3) and maternal infusion experiments, extensive protein binding and/or ion trapping in fetal plasma may contribute to a prolonged half- l i fe. These factors are probably not that important given the low degree of maternal (=49%) and fetal (=39%) binding and almost complete ionization of MCP in both maternal and fetal plasma at physiological pH. Fetal recirculation of MCP via ingestion of drug containing tracheal and 164 amniotic fluids (Table 23), in addition to possible reabsorption across the pulmonary epithelium and chorioallantoic vasculature may also be partly responsible for this observation. Even more likely is the possibility that the fetus is pharmacokinetically "deep", resulting in the slow release of MCP to clearing organs (e.g. placenta, l iver, kidney). By dosing the fetus directly we were able to obtain an estimate of the volume (depth) of the fetal compartment(s). Average estimates of 5.6 ± 1.0, 15.9 ± 5.2 and 17.9 ± 5.8 L/kg were calculated for Vc, Vss and Varea, respectively (Table 16). Corresponding mean estimates in the ewe, for the maternal infusions, were 1.1 ± 0.3, 4.5 ± 1.4 and 4.6 ± 1.2 L/kg (Table 12). In absolute terms fetal volumes are very small when compared with maternal (Average fetal weight, 2.18 ± 0.86 kg; average maternal weight, 77.8 ± 16.6 kg; Table 9). Compared on a weight corrected basis though, the depth (or size) of the fetal compartment(s) is considerably greater (=3-4 fold) than that in the ewe. Given the low degree of protein binding in the fetus it would appear that MCP is extensively distributed within the fetal lamb. Due to this apparent larger volume of distribution, a smaller fraction of drug is available to the clearing organ(s) per unit time, resulting in a slower rate of elimination. This is illustrated by the following equation, which shows that elimination half- l i fe is a function of both volume of distribution and clearance: tkp = [(0.693)(V a r e a)]/CL. Since the volume of distribution at steady-state is equal to the sum of the volumes of the central (Vc) and peripheral (Vp) compartments, we can also obtain an estimate of the volume of distribution of MCP in the peripheral compartment. Using the compartmentally derived value for Vss an average Vp of 10.3 L/kg was calculated, considerably larger than the central volume (5.6 L/kg). Although we have suggested that the ewe is represented by the peripheral (tissue) compartment in Scheme 2, a "deeper" fetal distribution 165 compartment, which is part of or separate from the peripheral compartment, appears very likely. The fact that kjg and the intercompartmental transfer rate constants, k^ and k 2 i» obtained during the fetal infusions (Table 15) are approximately one-half those obtained for the ewe in the maternal infusions (Table 11), would also seem to indicate that the fetus is pharmacokinetically "deep". Similarly, a mean peripheral volume (Vp) of 3.4 L/kg was calculated for the ewe in the maternal infusion studies and is also much larger than the maternal central volume (1.1 L/kg), particularly when considered in absolute terms. For the maternal infusions, we have presumed that the fetus is represented by the peripheral compartment in Scheme 2. Since maternal and fetal elimination half-lives are quite different it would appear that this peripheral compartment is composed of at least two parts with varying "depths". A shallow tissue area from which the drug redistributes quite rapidly accounts for the shorter maternal half- l i fe (Table 11), and a deeper tissue distribution compartment which equilibrates at a rate independent of maternal tissue results in a longer fetal half-l i fe (Table 13). Fetal total body clearance, CL S , (7.2 ± 3.3 L/h/kg) obtained for the fetal infusions is also considerably larger (=2.5 fold) than that determined for the ewe (3.0 + 0.6 L/h/kg) following the maternal infusions on a weight corrected basis. Considering the large plasma MCP concentration gradient between the ewe and fetus (=20 fold at steady-state; Table 14) and the large maternal-fetal mass difference, transfer across the placenta from the fetus to the ewe probably accounts for most of this difference rather than fetal renal and/or hepatic elimination. 166 4.5 Maternal and Fetal Metoclopramide Plasma Protein Binding at Steady-State. Protein binding of MCP in maternal and fetal plasma, obtained at steady-state from the infusion experiments, was determined using ultrafiltration. This method provides the advantages of ease and speed of accomplishment compared with equilibrium dialysis. No significant membrane uptake was observed following the filtration of MCP stock solutions prepared in isotonic phosphate buffer (Table 18). Ultrafiltrate volumes were approximately 30-40% of original plasma volumes (0.5-1.0 mL), but the sensitivity of the assay allowed duplicate sample measurements in spite of this limitation. Because a centrifuge which would maintain higher temperatures was not available binding experiments were conducted at 25 °C. Ideally ultrafiltration should have been carried out at the body temperature of sheep (39 °C and 39.5 °C for the ewe and fetus, respectively). As protein binding generally decreases with increasing temperature (Kwong, 1985), our results may be somewhat overstated. The binding of quinidine for example, was =6% lower at 37 °C than at 25 °C using equilibrium dialysis (Kwong, 1985). A decrease in the interaction between MCP and bovine serum albumin has also been reported when the temperature was increased from 4 °C to 38 °C during equilibrium dialysis (Pagnini and Di Carlo, 1972). Average binding of MCP to maternal plasma at steady-state, was =49% and to fetal plasma =39% (Tables 10 and 14), with no significant differences between the two determinations following the maternal or fetal infusions. Fetal MCP steady-state plasma concentrations were =10-fold higher during the fetal infusions than during the maternal drug infusions. The percentage of MCP bound in these experiments is very similar to that reported in man by Webb et al., 1986 (=40%; free fraction, 0.60 ± 167 0.04) following equilibrium dialysis. Like most basic lipophilic drugs (Piafsky and Borga, 1977; Piafsky, 1980; Paxton, 1983), MCP was found to be primarily bound to alphaj-acid glycoprotein (AAG). A small degree of binding to albumin was also seen but this interaction is reported to be very weak (Pagnini and Di Carlo, 1972; Denisoff and Molle, 1978; Gourley et al., 1982). These findings are consistent with reports for other basic drugs, most of which appear to bind both to albumin and AAG, with AAG as the major binding site. Examples include meperidine (Nation, 1981) and propranolol (Krauer et al., 1986). A lower percentage of binding to fetal (umbilical cord) than to maternal plasma has been reported in humans for lidocaine (Wood and Wood, 1981), meperidine (Naton, 1981) and propranolol (Wood and Wood, 1981; Krauer et al., 1986). This was associated with a lower concentration of AAG in fetal plasma which is =30-40% of the maternal concentration (Wood and Wood, 1981; Hill and Abramson, 1988). Assuming that MCP also binds to AAG in sheep, then the lower degree of binding in fetal plasma may also be the result of lower AAG concentrations, but this remains to be established. In general, however, total protein concentrations are lower in the fetus (=49 mg/mL; Nathanielsz et al., 1980) than in the ewe (=60 mg/mL; Altman and Dittman, 1972). The protein binding of several drugs to maternal and fetal plasma has been reported in sheep, including meperidine (Szeto et al., 1978), morphine and methadone (Szeto et al., 1982), cimetidine (Ching et al., 1985), indomethacin (Anderson et al., 1980) and bupivacaine (Kennedy et al., 1986). With the exception of the bupivacaine study, where albumin and total protein concentrations were determined by electrophoresis, protein concentrations have not generally been measured. For bupivacaine (Kennedy et al., 1986), fetal binding was =50% of maternal. This appeared to be due to the difference in total protein concentration, 168 with fetal levels =50% lower than maternal, although this correlation was not made. These authors concluded that the difference in maternal and fetal binding was responsible for the low fetal to maternal concentration ratios they observed. The specific protein to which these drugs are bound in sheep was not determined in these studies. Different degrees of binding in maternal and fetal plasma were also reported for meperidine and methadone with the percentage of drug bound in fetal plasma lower in both cases. These reports are in contrast to most human studies, where a correlation of differential binding to maternal and fetal protein concentrations and the specific binding protein is often made. This would appear to be due to the fact that purified forms of ovine albumin and AAG are not currently available. Similarly, monospecific antisera which would allow protein concentration determinations by nephelometry or radial immunodiffusion, such as those available for the previously mentioned human studies (e.g. lidocaine, propranolol; Wood and Wood, 1981), are also not available. It should be noted that in addition to protein concentration, differential binding in maternal and fetal plasma may be the result of variant forms of protein with different binding affinities and/or the presence of endogenous displacing agents such as free fatty acids and bilirubin (Nation, 1981; Szeto et al., 1982a; Hill and Abramson, 1988). The amount of AAG in maternal and fetal sheep plasma was measured indirectly by Hill et al., 1986, by determining the concentration of N-acetylneuraminic acid (NANA) which comprises 11% by weight of AAG. The binding of the basic lipophilic drugs propranolol, methadone and lidocaine in spiked maternal and fetal plasma was measured using equilibrium dialysis. While a good correlation between the percentage of drug bound and AAG content, as measured by NANA concentration, was reported for the 169 ewe this was not the case for fetal plasma. Fetal NANA concentrations were =19-fold higher than maternal, but this was not associated with an increase in the bound/free concentration ratios of these drugs which were significantly lower than maternal in all cases. The authors postulate that a variant form of fetal AAG or the presence of other glycoproteins in fetal plasma may be responsible for the elevated concentrations of NANA they observed. They conclude that the discrepancy between the maternal and fetal binding of these drugs is due to a functional deficiency of the glycoprotein(s) present in fetal plasma. 4.6 Metoclopramide Placental (Transplacental) and Nonplacental Clearances Following Maternal and Fetal Infusions to Steady-State Until quite recently the measurement of placental clearance of many compounds has been made by employing the Fick principle (Meschia et a7., 1967). According to Meschia et a7., 1967, placental clearance is defined as the rate of diffusion across the placenta divided by the transplacental arterial concentration difference. The Fick principle appears to have been most often used to study placental clearance of compounds from fetus to mother following fetal administration but is also applicable to clearance from mother to fetus following maternal administration. This technique is only suitable for the study of compounds which are not metabolized by the placenta, are cleared rapidly by the placenta and are not bound to plasma proteins. They must also produce a measurable arteriovenous concentration difference across the placenta. Furthermore, there must be no fetal metabolism or elimination. Because of these requirements the Fick principle has been primarily used to study relatively inert substances such antipyrine, urea, tritiated water (Meschia et a7., 1967), various polar nonelectrolytes (Boyd et a7., 1975) and glucose analogues (Stacey et a7., 170 1978). It is not useful for studying the placental clearance of most drugs in the maternal-fetal unit as many are protein bound and for others transfer is diffusion-limited. A method suitable for the measurement of placental clearance of diffusion-limited substances has since been proposed by Anderson et a?., 1980, 1980b. Using steady-state maternal and fetal drug concentrations and obtaining a value for total drug clearance from the terminal phase of a plot of log concentration versus time, a placental and tissue clearance may be calculated. The method is independent of arteriovenous concentration differences, as maternal and fetal steady-state concentrations are used in the equation, and it may also be used for compounds which are metabolized and/or protein bound. One limitation is that the bidirectional clearances of drug across the placenta are assumed to be equal. A more recent method has been proposed by Szeto et a7., 1982b which may be used to calculate both the maternal and fetal placental as well as nonplacental clearances of any compound at steady-state. The maternal-fetal unit is represented by the two compartment model shown in Scheme 3, with the assumption that drug disposition in both the ewe and fetus are adequately described by a one compartment system. This assumption is valid provided both compartments are at steady-state. At this time, even in more complex systems, drug distribution between the central and peripheral compartments is complete and both the ewe and fetus can be pharmacokinetically treated as homogeneous units. Unlike previous techniques, no assumptions regarding the direction or mechanism of drug transfer are made. Nonplacental elimination (biotransformation, renal elimination) and distribution of drug between the ewe and fetus are assumed to be f irst order processes. Constant rate infusion of drug to the mother results in two simultaneous rate equations describing the change in amount of drug with time in the maternal and fetal compartments at steady-state. 171 Similarly, two rate equations are obtained following fetal steady-state infusion. Solution of these simultaneous equations results in expressions (Eqs. 21-26) for the calculation of bidirectional clearance of drug across the placenta as well as clearance of the drug from the maternal and fetal compartments by nonplacental routes. Using these equations, the respective maternal and fetal infusion rates and experimentally determined steady-state maternal and fetal drug concentrations, the placental and nonplacental clearances of MCP have been determined. Calculated clearance values based on total drug concentration are contained in Tables 19 and 20 while those for unbound drug concentration are presented in Tables 21 and 22. Steady-state plasma concentrations of MCP were achieved in the ewe and fetus between 15-45 min after the bolus loading dose and initiation of the infusion as illustrated in the representative plots in Figs. 7 and 8. For the maternal infusions, the ratio of fetal to maternal steady-state MCP concentration (Cf/Cm) ranged from 0.37 to 0.82 with a mean of 0.57 ± 0.14 (Table 10). Similarly, the steady-state ratio for unbound MCP ranged from 0.42 to 0.98 with an average of 0.67 ± 0.18. While a steady-state was obtained in the ewe and fetus during the infusion period equilibrium between the ewe and fetus was not achieved as indicated by a ratio of less than one. As was discussed earlier (Section 4.3.1), differences in plasma protein binding and a pH gradient between maternal and fetal plasma have been proposed to account for such observations, but these would not appear to be important for MCP considering its degree of ionization (pKa 9.3) and similar binding to both maternal (=49%) and fetal (=39%) plasma proteins. Rather, it would appear that the lower steady-state concentration of drug in the fetus can be accounted for by fetal elimination of MCP. 172 Using compartmental analysis, Riggs, 1963, has shown that for a two compartment system such as that illustrated in Scheme 3, an inability to attain a state of equilibrium between the mother and fetus can only occur i f there is elimination of drug from the fetal compartment. Similarly Szeto et al., 1982b, have applied this analysis in their model for placental and nonplacental drug clearance. From Eqs. 23 and 26 it can be seen that following maternal drug administration the ratio of fetal to maternal drug concentration at steady-state is a function of the placental clearances (CLmf, CLfm) of drug across the placenta and fetal nonplacental clearance (CLf 0): Cf/Cm = C L m f / ( C L f m + CL f 0 ) (29) According to Eq. 29 only when CLf0 is equal to zero is it possible for the maternal and fetal compartments to achieve equilibrium (i.e. Clmf = CLfm; where there is no net transfer of drug between the ewe and fetus) and for maternal and fetal concentrations to be equal. A ratio of less than one suggests that the fetus has the ability to eliminate the drug while ratios greater than one imply active transfer of the drug from the maternal to the fetal compartment (Szeto et al., 1982b). The extent of fetal drug exposure then, will be a function of the bidirectional rates of drug transfer across the placenta as well as fetal elimination by nonplacental routes (e.g. renal, hepatic, lung). The contribution of each of these clearances to fetal drug exposure will be dependent upon maternal and fetal pharmacokinetic and placental factors as previously mentioned in Section 4.3.1. From our application of the Szeto model it was determined that the fetus has considerable ability to eliminate MCP by nonplacental mechanisms. Nonplacental clearance of MCP by the fetus (CLf0) ranged from 0.6-6.5 L/h/kg (unbound drug; Table 21) and accounts for =6-46% (mean = 20.5 ± 173 13.8%) of the overall elimination of MCP from the fetus. This activity was present as early as 120 days gestation (Ewe 60) but there was only a modest correlation with gestational age (r = 0.525) suggesting that the ability of the fetus to eliminate MCP did not change over the range of gestational ages studied (120-138 days; Table 9). It would appear then, that individual differences in the functional capacity to eliminate MCP are responsible for the considerable intersubject variability seen in this parameter. Similarly, there was no correlation (r = 0.399) between gestational age and fetal nonplacental contribution (Table 22) to total fetal drug clearance (CLff). Specifically which fetal nonplacental routes are involved in the elimination of MCP remain to more fully evaluated. As discussed in Section 4.7, the excretion of MCP in fetal lung fluid appears to represent a significant route of elimination. We have also observed considerable accumulation of the drug in amniotic fluid (Section 4.7). Since both fetal lung fluid and urine are responsible for amniotic fluid volume it is likely that MCP is also eliminated by the kidney. This, however, remains to be determined. The contribution of the fetal l iver, i f any, also remains to be evaluated, although in vitro studies with ovine fetal hepatic microsomes, previously discussed in Section 4.3.3, have shown some metabolic activity (dealkylation, hydroxylation, glucuronidation). Maternal nonplacental clearance (CLm o = 5.6 +1.2 L/h/kg for unbound drug; Table 21) is significantly greater than (=46%) fetal nonplacental clearance (CL f o = 3.0 + 2.2 L/h/kg; Table 21) indicating that MCP is much more efficiently eliminated by maternal nonplacental routes. When expressed as a fraction of cardiac output, CLf0 is even lower than CL m o , i.e. =9.5% of cardiac output (combined ventricular output =527 mL/min/kg (=31.6 L/h/kg); Rudolph and Heymann, 1970) is cleared in the fetus per unit time compared 174 to =62.9% of the cardiac output (=148 mL/min/kg (=8.9 L/h/kg); Rosenfeld, 1977) in the ewe. The placental (transplacental) clearance of MCP from the fetus to the ewe (CLfm) tended to increase with increasing gestational age (r = 0.701; for unbound drug) as did clearance of MCP from the ewe to the fetus (CLmf; r = 0.625; Table 21). The transfer of a drug across the placenta is determined by its physicochemical properties such as molecular weight, degree of ionization, l ipid solubility and the extent of binding to plasma proteins. Metoclopramide is a low molecular weight drug (=300) and is l ipid soluble so its transfer across the placenta is likely limited by placental blood flow rather than by anatomic properties of the placenta such as membrane thickness, area and permeability. Between 120-140 days gestation the placenta receives =42% of fetal cardiac output (combined ventricular output =527 mL/min/kg (=31.6 L/h/kg); Rudolph and Heymann, 1970) resulting in a greater degree of placental perfusion as fetal weight increases with gestational age (r = 0.851). Perfusion of the placenta on the maternal side also increases with gestational age receiving =16% of maternal cardiac output (=148 mL/min/kg (=8.9 L/h/kg); Rosenfeld, 1977) at 130-140 days. The placental clearance of MCP from the fetus to the ewe (CLfm = 21.2 ± 5.5 L/h) is significantly greater (=17%) than clearance of drug from the ewe to the fetus (CLmf = 17.5 ± 4.5 L/h) (Table 21). Since the degree of ionization of MCP in both maternal and fetal plasma is very similar (=98%) it would appear that the difference in the placental clearances is likely due to other factors. For example, the lower extent of plasma protein binding in the fetus (=39%) compared to the ewe (=49%) will result in higher fetal concentrations of free drug being available for diffusion. Furthermore, a greater proportion of cardiac output is 175 delivered to the placenta by the fetus as compared to the ewe leading to greater clearance. As would generally be expected, clearance of MCP across the placenta from the fetus to the ewe appears to be the major route of fetal drug elimination accounting for =80% of the total fetal clearance (CLff) while nonplacental clearance is only responsible for =20% (Table 22). Maternal placental clearance (CLmf) accounts for only about 5% of the total MCP clearance from the ewe with nonplacental routes (e.g. renal, hepatic) playing the major role (=95%; Table 22). To assess whether MCP is flow-limited, we have modified the basic equation for determining placental clearance by the Fick principle (Meschia et al., 1967) and made the numerator equal to the fetal steady-state metoclopramide concentration during a fetal drug infusion times CLfm. Using this modification, MCP's placental clearance averages 6.7 + 2.7 L/h/kg (for total drug concentration). This compares favorably with the published placental clearance values for antipyrine, a standard for flow-limited compounds, which range from =4.5-7.2 L/h/kg (Meschia et al., 1967; Walker, 1977; Owens et al., 1986). This strongly suggests that the placental transfer of metoclopramide in sheep is flow-limited and that its rate of transfer is largely determined by the rate of delivery of drug to the exchange surface. Total clearance of MCP by the fetus (CLff = 13.3 ± 3.6 L/h/kg for unbound drug; Table 22) was found to be significantly greater (=56%) than by the ewe (CLm m = 5.9 ± 1.2 L/h/kg) on a weight corrected basis. While total MCP clearance from the maternal and fetal compartments is the sum of placental and nonplacental clearances (Eqs. 25 and 26), this =2-fold difference would appear to be largely due to the significantly greater placental clearance of MCP from the fetus to the ewe (CLfm) rather than to fetal elimination by nonplacental routes (CLf0) (renal, hepatic, lung). 176 Total clearance of MCP by the fetus represents =42% of fetal cardiac output while that in ewe accounts for =66% of the maternal cardiac output. The total maternal and fetal clearance values calculated by the Szeto model (Table 20; total drug concentration) agree well with the model independent methods used for total body clearance (CLS) estimation (Eqs. 12 and 14) as shown in Tables 12 and 16. The placental and nonplacental clearances of acetaminophen (Wang et al., 1986), diphenhydramine (Yoo et al., 1987), methadone (Szeto et al., 1982) and morphine (Szeto et al., 1982b) have also been determined in chronically instrumented pregnant sheep following steady-state maternal and fetal infusions. The values from these reports have been converted from mL/min to L/h, to be consistent with our data presentation for MCP, and the summarized means are presented in Tables 38 and 39. The percentage of drug bound to maternal and fetal plasma proteins is also included in Table 38. The transfer of both acetaminophen and morphine was postulated to be diffusion-limited. Acetaminophen is a weak organic acid (pKa 9.5) with moderate l ipid and water solubility (Wang et al., 1986) while morphine (pKa 9.8) demonstrates poor water solubility and only moderate lipid solubility. Both compounds bind to maternal and fetal plasma proteins to a similar extent with respective values of =10% and <5%. The bidirectional placental clearances (CLfm, CLmf) of acetaminophen were approximately equal (=3.5 L/h; Table 38) and tended to increase with gestational age (r = 0.724). As gestation progresses a decrease in membrane thickness and an increase in placental surface area may result in increasing membrane permeability (Wang et al., 1986) and may explain the correlation with this diffusion-limited compound. The placental clearances of morphine on the other hand, show only a modest correlation (r = 0.572) with increasing gestational age. Unlike acetaminophen, the placental clearance of morphine 177 Table 38. Data for comparison of mean placental (transplacental) and nonplacental clearances of various drugs in the ewe and fetus employing the model proposed by Szeto et al., 1982b. (Total drug concentration). Also included is the percentage of drug bound to maternal (m) and fetal (f) plasma proteins. Drug Placental Nonplacental C Lmf a C L fm a C Lmo b C L f o a L/h L/h L/h L/h Acetaminophen0 3.6 ± 1.0 3.5 ± 0.8 54.8 ± 21.9 1.3 ± 0.6 (n = 12) (1.9 ± 0.4) (1.8 ± 0.4) (0.9 ± 0.3) (0.6 ± 0.2) % Bound: m =10%, f =10% Diphenhydramined (2.6 ± 1.7) (6.6 ± 3.6) (2.6 ± 0.6) (5.6 ± 2.5) (n = 6) % Bound: m =86%, f =72% Methadone6 5.6 ± 3.8 18.8 ± 11.2 104.4 ± 42.5 9.5 ± 3.1 (n = 4) % Bound: m =75%, f =39% Metoclopramide 8.7 ± 2.3 12.3 ± 3.1 218.3 ± 69.0 3.7 ± 3.5 (n = 9) (4.3 ± 1.3) (6.2 ± 2.4) (2.8 ± 0.5) (1.7 ± 1.2) % Bound: m =49%, f =39% Morphinef 1.5 ± 0.5 3.5 ± 1.1 167.8 ± 39.3 7.5 ± 2.4 (n = 7) % Bound: m <5%, f <5% a) Numbers in parentheses normalized to estimated fetal body weight (L/h/kg). b) Numbers in parentheses normalized to maternal body weight (L/h/kg). c) Data obtained from Wang et al., 1986. d) Data obtained from Yoo et al., 1987. e) Data obtained from Szeto et al., 1982b. f) Data obtained from Szeto et al., 1982. 178 Table 39. Data for comparison of mean total clearance and percent of nonplacental contribution to total drug elimination for various drugs in the ewe and fetus employing the model by Szeto et a7., 1982b. (Total drug concentration). Drug Total Clearance Nonplacental Contribution C L m m a C L f f b Maternal Fetal L/h L/h % % Acetaminophenc (n = 12) Diphenhydramine (n = 6) Methadone6 (n = 4) Metoclopramide (n = 9) Morphine1 (n = 7) 58.4 + 21.6 (0.9 + 0.3) (2.7 + 0.7) 110.0 + 46.2 227.0 + 68.9 (2.9 + 0.4) 169.3 + 39.4 4.9 ± 1.3 (2.5 ± 0.5) (12.2 ± 6.0) 28.2 ± 12.5 16.0 ± 5.1 (7.9 ± 2.9) 11.0 ± 2.8 92.2 ± 5.8 =97.8 - -95.2 ± 1.1 95.8 ± 1.7 99.1 ± 0.3 26.6 ± 6.9 46.3 ± 4.5 37.3 ± 14.7 20.5 ± 13.8 67.4 ± 10.4 a) Numbers in parentheses normalized to maternal body weight (L/h/kg). b) Numbers in parentheses normalized to estimated fetal body weight (L/h/kg). c) Data obtained from Wang et al., 1986. d) Data obtained from Yoo et al., 1987. e) Data obtained from Szeto et al., 1982b. f) Data obtained from Szeto et al., 1982. 179 from the fetus to the ewe (CLfm) is approximately twice that in the reverse direction (CLmf) suggesting some perfusion limitation as a result of the higher percentage of cardiac output on the fetal side of the placenta. Compared to MCP the transplacental clearances of both acetaminophen and morphine are considerably smaller (=2-6-fold; Table 38). The higher transplacental clearances of MCP compared to these other two drugs would appear to be related to its l ipid solubility resulting in flow-limited transfer. In their studies with acetaminophen Wang et al., 1986, also calculated a placental clearance (CLp) using the following model independent method based on the Fick principle: ER = (FA - UV)/FA (30) CLp = (ER)(Qum) (31) where ER is the extraction ratio of the drug across the placenta on the fetal side, FA and UV are the respective concentrations of drug in the fetal femoral artery and umbilical vein following a fetal steady-state infusion and Qum is the umbilical blood flow rate. The placental clearances calculated using the extraction ratio method were very similar to those obtained using the two compartment model (i.e. CLfm, CLmf) proposed by Szeto et al., 1982b, lending support to its validity. Unlike the Szeto model the clearance determined by use of the extraction ratio includes any metabolism by the placenta in addition to the transfer clearance of drug across the placenta. For the Szeto model, however, drug metabolism occurring in the placenta is included in the nonplacental clearance terms, CL m o and CLf0 (Wang et al., 1986). Since the model dependent and model independent placental clearance values were very similar, placental metabolism of acetaminophen was postulated to be negligible. 180 In one of our experiments (Ewe 130) an umbilical venous catheter was also implanted allowing the collection of umbilical venous samples. Using Eq. 30 an ER of 0.30 ± 0.05 was calculated for MCP at steady-state following the fetal infusion. This value is about two and one-half times greater than that reported for acetaminophen (0.12). In this particular experiment we did not have an accurate estimate of umbilical blood flow but by using an average value of =220 mL/min/kg (13.2 L/h/kg) (Rudolph and Heymann, 1970; Edelstone, 1980) and a fetal weight of 3.25 kg (Ewe 130, Table 9) an estimated CLp of =4.0 L/h/kg was calculated (Eq. 31). Although this estimate approximates the values obtained from the model dependent method (CLfm = 5.4 L/h/kg; Table 19) further study is required to substantiate this single observation and to determine i f there is placental metabolism of MCP. In one maternal infusion experiment the umbilical vein was also catheterized. This allowed the calculation of an extraction ratio (ER) for the fetus based upon the following relationship: ER = (UV - FA)/UV, where UV represents the umbilical venous concentration of MCP and FA the arterial concentration. An average extraction of 0.24 ± 0.04 was determined during steady-state representing considerable fetal uptake of MCP, but further experimentation is required to verify this single observation. Like MCP, diphenhydramine and methadone are weakly basic (pKa's =9.0-9.3), have similar molecular weights (=260-310), are lipophilic and are bound to a greater extent in maternal than in fetal plasma (Table 38). Unlike morphine and acetaminophen their placental clearance, like that of MCP, is probably limited by placental blood flow rather than by the anatomic properties of the placental membrane. Diphenhydramine and methadone are structurally quite similar with both molecules possessing two unsubstituted phenyl rings and comparable alkylamine side chains. While 181 the data for diphenhydramine is reported on a weight corrected basis (Yoo et al., 1987) i f we assume an average fetal weight of 2.5 kg, which is quite common for the period of 127-135 days gestation over which these animals were studied, then the placental and total fetal clearances are very similar to those for methadone. On this basis CLmf, CLfm and CLff average =6.5, 16.5 and 30.5 L/h, respectively, compared to values of =5.6, 18.8 and 28.2 L/h for methadone. The small variations in CLmf and CLfm are most probably related to subtle differences in their physicochemical properties (e.g. degree of ionization, l ipid solubility, protein binding). Methadone is =39% bound in fetal plasma while the corresponding value for diphenhydramine is =72%. The higher CLfm of methadone (=18.8 L/h) then, may well be due to more free drug being available for diffusion on the fetal side of the placenta. For MCP and diphenhydramine the fetal transplacental clearance (CLfm) values are very similar averaging =6 L/h/kg while CLmf is =1.5 times greater for MCP than for diphenhydramine. Compared to methadone the CLmf for MCP is =1.5-fold larger, while CLfm is =1.5 times smaller (Tables 38 and 39). Again, these differences in transplacental clearance values are likely related to differences in the amount of free drug available for diffusion, 1ipophilicity and the degree of drug ionization in maternal and fetal plasma. If we again assume a fetal weight of =2.5 kg and convert the weight normalized values for diphenhydramine in Tables 38 and 39 to units of L/h, then in general, the placental (CLmf, CLfm) and total fetal (CLff) clearances of MCP, diphenhydramine and methadone are considerably larger than those for morphine and acetaminophen, probably because of their higher l ipid solubility. The larger CLfm values for these three weakly basic drugs appears to be largely responsible for their higher total fetal clearance (CLff = CLfm + CLf0) compared to acetaminophen and morphine. Fetal nonplacental clearances (CLf0) differ considerably among these five drugs with the contribution to total fetal clearance ranging from =21-67% (Table 39). Nonplacental clearance represents removal of the drug by nonplacental mechanisms such as biotransformation in the liver and perhaps other organs as well (e.g. lung, kidney), renal elimination and possible uptake by various body tissues. The considerable variability observed in this parameter in the fetus then, would appear to represent differences in the functional maturity of these elimination pathways to remove (metabolize, excrete) these various substrates. All of these drugs are subject to oxidation reactions (dealkylation) as well as to conjugation (glucuronide, sulfate). The renal clearance of meperidine (Szeto et al., 1978, 1980), lidocaine (Morishima et al., 1979), methadone (Szeto et al., 1982b) and cimetidine (Mihaly et al., 1983) has been reported in the chronically catheterized fetal lamb. This may also be a significant route of elimination for both MCP and diphenhydramine considering their accumulation in amniotic fluid (Riggs et al., 1987) as fetal urine is a major contributor to amniotic fluid volume. For the most part biotransformation has been studied using in vitro enzyme preparations, although as previously discussed (Section 4.3.3) the in vivo hepatic extraction of propranolol has been reported in the anesthetized, acutely catheterized fetal lamb (Mihaly et al., 1982). The ability of the fetal lamb to form sulfate and glucuronide conjugates with acetaminophen in vivo has also been reported (Wang et al., 1986), although fetal activity was low compared to that in the adult, with the glucuronides eliminated in fetal urine. No in vitro glucuronidation activity was detected in placental enzyme preparations, and previous studies involving separate fetal and maternal i.v. bolus injections of both acetaminophen conjugates (Wang et al., 1985) showed that they do not undergo placental transfer. Generally 183 in vitro fetal hepatic enzyme preparations demonstrate low activity, with the onset of various pathways a function of gestational age and species (Juchau et al., 1980; Pelkonen, 1980). Because there is l i t t le or no in vitro data available from fetal sheep and because chronically catheterized pregnant sheep preparations are now commonly used for studies of maternal-fetal pharmacokinetics, Dvorchik et al., 1986, assessed the ability of hepatic microsomes from nearterm fetal lambs to catalyze the biotransformation of several substrates including methadone and morphine. Although enzyme activity was low compared to adult levels, fetal microsomes catalyzed the N-demethylation of methadone and the glucuronidation of morphine. The significance of these findings on the in vivo nonplacental clearances of MCP, diphenhydramine and morphine, however, remains to be established. Specifically which pathways are involved in the nonplacental clearance of MCP in sheep remain to be clarified as its metabolism shows considerable species variation, with conjugation (primarily sulfate) being the major route in humans (Teng et al., 1977; Bateman et al., 1980) while in rats and dogs (Arita et al., 1970; Bakke and Segura, 1976; Cowan et al., 1976) dealkylation to form the mono-deethylated metabolite is the major metabolic pathway. Binding of drugs to tissue as well as possible accumulation and excretion in fetal lung f luid, as has been shown for MCP and diphenhydramine (Riggs et al., 1987), likely also plays a role in the nonplacental clearance of these compounds and may account for some of the variability seen between different drugs. Considerable uptake of methadone in the isolated perfused rabbit lung has also been reported (Wilson et al.-, 1976) while the f irst pass uptake of morphine in the human lung following i.v. bolus injection is very low (<4%) (Roerig et al., 1987). 184 The maternal and fetal drug infusions in all of these studies have been separated by no more than 24-72 hours in order to minimize temporal factors such as changes in uterine and umbilical blood flows which could influence placental drug diffusion, particularly for those drugs whose transfer is flow limited. To overcome this potential problem it would be best to use simultaneous infusions of stable isotope-labelled and unlabelled drug to the ewe and fetus respectively (or vice versa) and to subsequently measure maternal and fetal drug concentration by selected ion monitoring. This technique would be particularly useful for drugs with long half-lives or which persist in amniotic fluid as much longer "washout" periods may be required, increasing the probability of temporal effects on both the placental and nonplacental clearances. This methodology has been used to study the pharmacokinetics of theophylline in the chronically catheterized pregnant ewe following i.v. bolus dosing (Brazier et al., 1984). Terminal elimination half-lives averaged =12 and 6 hours in the ewe and fetus respectively. Labelled and unlabelled drug accumulated in amniotic fluid but its persistence was not reported. 4.7 Metoclopramide Accumulation in Fetal Tracheal (Lung) and Amniotic Fluids Following Maternal or Fetal Administration A marked accumulation of MCP in fetal lung fluid was observed following the maternal i.v. bolus dosing experiments as well as in the maternal and fetal infusion studies. Concentrations in fetal tracheal fluid were at least 10-15-fold higher than in those in fetal arterial plasma throughout the sampling period. There are numerous reports of the pharmacokinetic behavior of drugs in the fetus, conducted either in man or animal species. Most studies have concentrated on measurement of drug concentrations in the maternal and 185 fetal vascular compartments, with fewer investigations also examining drug levels in fetal urine and/or amniotic f luid. One potential route of drug excretion from the fetus that does not appear to have been studied is via fluid secreted from the fetal lung. Such fluid production occurs at a rate of =4.5 mL/kg/hr in the fetal lamb (Harding et a7., 1984b), with the fluid being either secreted into the amniotic cavity or swallowed (Harding et al., 1984b). The fluid appears to be formed primarily as a consequence of active chloride transport from plasma across the pulmonary epithelium (Strang, 1977). Sodium follows passively down the electrical gradient and there is a net flow of water to balance the osmotic force of sodium chloride. In some respects the fluid resembles an ultrafiltrate of plasma, particularly in regard to low protein concentrations. The plasma/lung fluid concentration ratios for several electrolytes, however, differ substantially from those expected in a true ultrafiltrate, with higher concentrations of chloride and potassium, and lower concentrations of calcium and bicarbonate in lung fluid than in plasma (Olver and Strang, 1974; Mescher et al., 1975). In adults, metabolism, excretion and binding of many drugs, particularly amine compounds, by the lung are important processes in total body clearance of these agents, in part because the lung receives all of the cardiac output (Roth, 1979; Collins and Dedrick, 1982; Roth 1984; Bend et al., 1985; Benford and Bridges, 1986). In the sheep fetus, however, the percentage of combined ventricular output distributed to the lung is low (=6%), as is pulmonary blood flow (=28 mL/min/kg) (Anderson et al., 1981). In spite of this, studies in our laboratory demonstrate a marked accumulation of two amine drugs, metoclopramide and diphenhydramine, in fetal lung fluid (Riggs et al., 1987). The tracheal/fetal plasma concentration ratios for MCP averaged 15.1 ± 1.4, 13.3 ± 3.1 and 15.9 + 2.9 186 following maternal i.v. bolus dosing, and maternal and fetal infusions, respectively (Table 23). The overall average ratio for diphenhydramine was 3.0 ± 0.6 following these same three modes of drug administration. We do not yet have any data to explain this difference in the ratio, but the higher plasma protein binding in the ewe for diphenhydramine (=86%, unpublished data), as compared with MCP (=49%), may be involved. The cell types (more than forty) in the adult lung that are responsible for accumulation and metabolism have not been identified for all drugs, but pulmonary endothelial cells lining the small vessels and capillaries in the lung are known to be important in the uptake of many endogenous amines, such as serotonin and norepinephrine (Bend et al., 1985; Roth, 1985). Several of these endogenous compounds are taken into the lung by saturable, carrier-mediated transport processes (e.g. serotonin, norepinephrine), in addition to passive diffusion. It has been suggested that some of the basic amine drugs (e.g. amphetamine, propranolol) are also removed from the circulation by these transport systems and accumulate in endothelial cells in concentrations exceeding those in the circulation (Anderson et al., 1974; Philpot et al., 1977; Bend et a7., 1985; Benford and Bridges, 1986). We do not yet have any data to suggest specific mechanisms in the fetal lamb, neither are we aware of any reports of accumulation of other drugs in fetal tracheal f luid. It would appear that basic amine drugs (i.e. pKa > 8) which are amphiphilic in nature (compounds containing both a large hydrophobic region and a group ionized at physiological pH) accumulate and persist in lung tissue to a greater degree than drugs which lack these properties (Philpot et a7., 1977; Benford and Bridges, 1986). In addition to the charge on the nitrogen, protonation of the nitrogen atom also seems to be important for the uptake and accumulation of these drugs. Differences in the time course 187 of uptake and release of various basic by the lung appear to be related to differences in their physicochemical properties (Philpot, 1977). The mechanisms responsible for the accumulation and uptake of basic amines have been studied primarily in isolated perfused lung (IPL) preparations. Using an IPL preparation (rabbit) Philpot et al., 1977, observed that the rate of uptake of methadone and imipramine, both basic amphiphilic amines, was equal to the rate at which they were supplied to the lung. This suggests that these compounds are not removed from the circulation by an active transport system, but diffuse into the lung and accumulate as a result of binding. In contrast, the uptake of amphetamine, which while lipophilic is unionized, was saturable, suggesting a carrier-mediated transport process. It was further suggested that because of its structural similarity to norepinephrine, amphetamine may accumulate in lung tissue by the norepinephrine transport system. The steady-state accumulation of various basic amines (imipramine, methadone, amphetamine) and nonbasic amines (imidazole, promazine, aniline) was also examined. The accumulation of the nonbasic amines was linearly related to drug concentration in the perfusate, with lung tissue to blood concentration ratios close to one. Accumulation of the basic amines on the other hand, was curvilinear and was related to dose, with very high tissue-to-blood concentration ratios at lower perfusion concentrations. The results of these experiments suggest that there are specific mechanisms for the accumulation of basic, but not of nonbasic amine compounds in the lung (Philpot et al., 1977). The steady-state accumulation of the basic amine drugs consisted of a saturable and nonsaturable component apparently related to processes of drug diffusion and binding within the lung. These components were further studied by following the efflux of these drugs into drug-free perfusate with respect 188 to time. The decay curve of imipramine for example, was triexponential suggesting that at least three distinct "pools" of this drug exist in lung tissue. The f irst two phases had very short half-lives of 18 and 58 sec and were felt to correspond to the diffusion component of drug accumulation, representing diffusion into the ce l l . The third phase, with its half- l i fe of =8 min, was felt to be due to intracellular binding forming a more persistent pool, and is thought to be responsible for the saturable component of drug accumulation. A persistent pool was not observed for amphetamine. These authors state that irreversible binding of imipramine and methadone to lung tissue was not responsible for the persistent pool of drug. Rather they speculate that it is the result of the formation of a drug-surfactant complex. Surfactant, which consists of =90% lipids and =10% protein, is produced and stored intracellular^ by Type II pneumonocytes in the alveolar wall. More than 85% of the lipids are phospholipids with phosphatidylcholine the major component (70-85%) followed by phosphatidylglycerol (Haagsman and van Golde, 1985). The net negative charge present on the alveolar lining has been attributed to the presence of phospholipids and sialic acid residues within glycoproteins associated with surfactant (Faraggiana et al., 1986). The lipophilic nature of the basic amine drugs and the existence of a positive charge on the nitrogen atom at physiologic pH suggest that an interaction between amphiphilic phospholipid and sial ic acid residues in glycoprotein is possible. This, then, could result in a reservoir (or persistent pool) of drug within as well as on the surface of the lung. This would appear to explain why compounds such as amphetamine which are not ionized, do not form a persistent drug pool within the lung (Philpot et al., 1977). The basic amine metoclopramide is also an amphiphilic compound, having both a hydrophobic region and an ionizable nitrogen (pKa 9.3) at physiological pH. 189 It is conceivable then, that MCP accumulates in the fetal lamb lung by similar processes. It is hypothesized that movement of MCP across the pulmonary epithelium from the pulmonary vessels, as well as release from tissue and surfactant and glycoprotein complexes may be responsible then, for the high concentrations of drug observed in tracheal f luid. The lower pH of tracheal fluid (=6.23; Mescher et al., 1975) compared to plasma and the volume of fluid within the fetal lung (=30 mL/kg; Olver and Strang, 1974) could also result in drug accumulation due to ion trapping. Although this may be the case for some drugs it is likely not important for MCP (pka 9.3) as it is almost completely ionized in plasma at physiological pH. The uptake and accumulation of a number of basic amine, neutral and acidic drugs was also examined by Okumura et al., 1978, using an isolated perfused rat lung preparation. The basic amines, including diphenhydramine, MCP and procainamide showed significant accumulation in lung tissue while neutral and acidic drugs showed l i t t le or no uptake from the perfusion medium. We also have preliminary data from our own studies with the chronically catheterized fetal lamb, suggesting no significant accumulation of the neutral steroids estradiol and progesterone in fetal tracheal f luid. Tracheal fluid/fetal arterial plasma concentration ratios for these two compounds were 0.40 and 0.22, respectively (Rurak DW and Leung PK, unpublished observations). As previously mentioned (Philpott et al., 1977), Okumura et al., 1978, found that l ipid solubility as well as a positive charge on the molecule were essential for drug uptake by the lung, with the degree of protonation the more important of the two factors (Anderson et al., 1974; Okumura et al., 1978). Clearance of the basic amine drugs from the perfusion medium was init ial ly rapid and then attained a steady-state indicating saturation. Like Philpot et al., 1977, these investigators suggested that this pattern was due to diffusion and 190 subsequent binding to lung tissue. The endothelial cell was suggested as the probable site. Similarly, clearance of these compounds, as a function of perfusion concentration, was also found to be dose-dependent and was attributed to binding of the drug to the lung cell or cellular components. Accumulation of basic amines in the lung was found to be reversible, as indicated by the displacement of an accumulated amine into the perfusate, following the addition of a second more lipophilic basic amine (Anderson et al., 1974; Yoshida et al., 1987). The addition of various metabolic inhibitors (e.g. ouabain) did not affect the accumulation of these basic amine drugs in the lung (Anderson et al., 1974; Yoshida et al., 1987). In order to clarify the mechanism by which basic drugs accumulate in the lung Yoshida et al., 1987, examined the binding selectivity of three basic amine drugs (quinine, imipramine and MCP) to subcellular fractions of homogenized isolated perfused rat lung. Mitochondria were found to be the major binding site for all three drugs. Imipramine showed the greatest degree of accumulation followed by quinine and MCP corresponding to decreasing l ipid solubility. Accumulation was dose-dependent. The lack of accumulation in the microsomal and cytosol fractions, with their higher l ipid content, led these workers to suggest that these basic amines do not bind to l ip id, but to other lung components such as protein and that the mitochondria may serve as a reservoir for basic drugs. The observation of an apparent triexponential decay for MCP in tracheal fluid (Figs. 9 and 10) following the maternal and fetal infusions with 28 hour post-infusion sampling, may very well be consistent with the pattern of efflux observed by Philpot et al., 1977, for imipramine in their isolated perfused lung experiments. The binding of MCP to mitochondria and possibly to surfactant in the Type II pneumonocytes as well as to surfactant present in the tracheal fluid itself (Mescher et al., 1975) may 191 result in reservoirs ("pools") of drug within and on the surface of the lung. The diffusion and release of MCP from pools of varying "depths" into tracheal f luid, then, could well be responsible for the triexponential post-infusion concentration versus time profiles we obtained in these two infusion studies. Estimated half-lives for the %, a and /3 phases following the maternal infusion were 3.1, 3.7 and 12.6 hours, respectively. Similarly, approximate values of 1.4, 4.8 and 15.1 hours were calculated following the fetal infusion. Even with the low rate of pulmonary blood flow in the fetus, the delivery of these drugs to the lung via the pulmonary circulation would appear to be more than sufficient to account for the levels observed in tracheal f luid. We base this argument on rough calculations done for an experiment involving maternal infusion of MCP to a steady-state. We have assumed a fetal weight of 2 kg, a lung fluid volume of 30 mL/kg (Harding et al., 1984), a tracheal fluid excretion rate of 4.5 mL/hr/kg (Harding et al., 1984), and a lung blood flow of 56 mL/min (Anderson et al., 1981). With these values and the measured MCP concentrations in fetal plasma and tracheal f luid, it can be calculated that during the 90 min infusion period =13% of the MCP delivered to the lungs would have to be taken up to account for the accumulation of the drug in tracheal f luid. It also appears that the rate of pulmonary uptake of MCP would not be constant over the infusion period. Rather it would be greatest during the f irst =5 min after the infusion is started, when extraction would have to be =74% and would thereafter decline progressively as the MCP concentrations in tracheal fluid reached a plateau level. After cessation of the infusion, it would appear that further uptake of the drug by the lung would be very low. The results of a preliminary experiment in one fetus where the pulmonary artery and vein of the left lung were catheterized indicates an extraction of =12% 192 at steady-state, in good agreement with our calculated estimate (=13%). Further studies are needed, however, to confirm this single observation. Whatever the mechanism involved in MCP accumulation in tracheal f luid, it appears to operate to create approximately a 15-fold gradient in drug concentration between this fluid and fetal plasma. A large proportion of the total amount of MCP taken up by the lung would be in the initial phase of drug administration to create the gradient. Once this is achieved, further uptake of MCP is required only to replace that lost via tracheal fluid excretion from the airway. This pattern of MCP uptake is very similar to the experiments by Okumura et al., 1978, with the isolated perfused rat lung. In their studies the uptake of the basic amine drugs including MCP, by the lung was maximal in the f irst 5 min of perfusion, as indicated by a rapid decline in drug concentration in the perfusate. Following this initial rapid uptake, there was a much more gradual decline in drug concentration in the perfusate with attainment of steady-state drug concentrations between 15 to 30 min. The accumulation of MCP in fetal lung fluid could have several potential consequences. First, tracheal fluid could be a significant route of drug elimination from the fetus. For example (Ewe 62, Fig. 7), during the f irst 120 min after the start of the maternal infusion, =3.2 /ig of intact drug was excreted in tracheal f luid, assuming a rate of tracheal fluid production of 4.5 mL/h/kg. This compares with a calculated total renal excretion of 3.6 /jg, assuming a glomerular fi ltration rate of 1 mL/min/kg (Gresham et al., 1972; Lumbers, 1983) and no significant renal tubular resorption of MCP. Although in vitro studies of the fetal liver have demonstrated the relatively early ontogenic appearance of oxidative, hydrolytic and conjugative enzyme systems in man and other primates (Juchau et al., 1980) as well as in sheep (Dvorchik et al., 1986), in vivo data 193 from fetal lambs indicate a low hepatic extraction of propranolol (Mihaly et al., 1982). Hence it may be that excretion of intact drug via lung fluid and urine is more important in overall fetal drug elimination. Second, since fetal lung fluid either enters the amniotic fluid compartment or is swallowed, a high drug concentration in tracheal fluid could contribute significantly to amniotic fluid levels. Assuming a lung fluid production rate of 4.5 mL/h/kg and a fetal weight of 2-3 kg near term, then =200-300 mL of fluid could enter the amniotic cavity per day, although as previously mentioned some fluid is also swallowed directly (Harding et al., 1984, 1984b). Alternatively, fetal ingestion of tracheal fluid could be an important route for recycling of drug within the fetus. Reabsorption of the drug in the gastrointestinal tract then, may at least in part, be responsible for the significantly longer terminal elimination half- l i fe of MCP in fetal plasma compared with that in the ewe (Tables 11, 13, 15 and 17). Diffusion of MCP from fetal lung fluid (=30 mL/kg) across the pulmonary epithelium back into the fetal circulation may also contribute to the longer fetal half- l i fe. These two factors have been previously discussed in Section 4.3.3. Finally, it would appear that at least some of the cellular elements within the lung could be exposed to very high drug concentrations. This would not be predicted by estimates of relative fetal drug exposure based on fetal/maternal plasma concentration ratios. The administration of various amphiphilic amine drugs to animals over prolonged periods or at high doses has been reported to result in drug-induced pulmonary phospholipidosis (Philpot et al., 1977; Bend et al., 1985; Haagsman and van Golde, 1985; Benford and Bridges, 1986). Similarly, the persistence of many of these compounds in lung tissue has also resulted in pulmonary phosphol ipidosis. The condition is characterized by the accumulation of phospholipids in a number of pulmonary and bronchiolar 194 cells. A drug-phospholipid complex is formed which is resistant to the enyzmes responsible for normal phospholipid catabolism. Provided secondary cell damage has not occurred, the process can be reversed by withdrawing the drug (Benford and Bridges, 1986). We have also observed considerable accumulation of MCP in amniotic fluid with average amniotic/fetal plasma concentration ratios of 3.5 ± 1.5 and 3.4 ± 1.4 following maternal i.v. bolus dosing and infusions to steady-state, respectively (Table 23). The appearance of MCP in amniotic fluid was rapid with measurable concentrations at 6 min following maternal i.v. bolus dosing, at 5 min for the fetal infusions and between 5-15 min for the maternal infusions. As mentioned above the flow of fetal lung fluid, with its high MCP concentrations, into the amniotic cavity likely contributes significantly to the accumulation of drug in this compartment. In humans =80% of an i.v. dose of MCP is recovered in the urine in 24 hours (Desmond and Watson, 1986) both as unchanged drug (=20%) and metabolites. From a qualitative study of urine collected from one of the ewes, during an i.v. bolus experiment, it would appear that renal excretion of MCP is also an important route of drug elimination in adult sheep (See Appendix). If this is also true in the fetal lamb then, the other major source of MCP in amniotic fluid is likely via elimination in fetal urine, although this remains to be established, since the fetal bladder and urachus were not catheterized. Fetal urine is produced at a rate of =0.14 mL/min/kg (Gresham et al., 1972) and enters both the amniotic and allantoic cavities. A fetus weighing 2-3 kg near term will produce =400-600 mL of urine in a 24 hour period. Urine volumes in excess of 1000 mL/day have also been reported (Brace, 1986; Tomoda et al., 1987). Another possible source of drug in amniotic fluid may be diffusion of MCP across the chorioallantoic membranes into the allantoic cavity with subsequent diffusion across the 195 allantoic and amniotic membranes into the amniotic sac. Transfer across the umbilical cord directly into amniotic fluid may also occur, however, the contribution of this route is probably minor. The persistence of MCP in amniotic fluid (Figs. 9 and 10) and subsequent reentry into the fetal circulation by fetal swallowing or reabsorption across the chorioallantoic membranes and umbilical cord vessels may also, in part, contribute to the longer fetal than maternal MCP plasma elimination half - l i fe. This has been previously discussed in Section 4.3.3. The pH of amniotic fluid (range 7.00-7.50; Mellor and Slater, 1971) may also result in ion trapping of some drugs and their persistence in this compartment. This would not seem to be important for MCP (pKa 9.3), however, considering that it is almost totally ionized (=98%) in maternal and fetal plasma. The accumulation of many other drugs in amniotic fluid has also been reported both in humans and in animals (Seeds, 1980; Reynolds, 1981) following maternal drug administration. Examples in humans include penicillin and ampicillin (Reynolds, 1981), meperidine (Szeto et al., 1978a), ethanol (Brien et al., 1983) and metoprolol (Lindeberg et al., 1987) with amniotic fluid samples being collected at the time of delivery or termination of pregnancy. In chronically instrumented pregnant goats, maternal administration (i.v. bolus, and/or infusion) of chlorpromazine, pentobarbital or phenylbutazone resulted in the accumulation and persistence of these drugs both in amniotic fluid and plasma (maternal and fetal) for up to 48 hours after dosing (Boulos et al., 1971). There are also several other examples of drug accumulation and persistence in amniotic fluid of chronically catheterized pregnant sheep, following maternal dosing, including meperidine (Szeto et al., 1978, 1979), lidocaine (Morishima et al., 1979), cimetidine (Mihaly et al., 1983.) and ethanol (Brien et al., 1985). 196 4.8 Pharmacokinetics of Metoclopramide in Plasma, Amniotic and Tracheal Fluids Following Intra-Amniotic Drug Administration In the previous section (4.7) the accumulation of MCP in tracheal fluid of the fetal lamb following maternal and/or fetal dosing was discussed. A potential consequence of this accumulation is drug recirculation within the fetus, as a result of either reuptake of the drug directly from the lung or the fetus swallowing lung and/or amniotic fluid. To examine this possibility we have administered MCP directly, as a bolus (10 mg), into the amniotic cavity followed by serial sampling of amniotic and fetal tracheal fluids, and fetal and maternal arterial and umbilical venous blood in chronically catheterized fetal lambs of 126-138 days gestation (Table 24). Transfer of MCP from the amniotic cavity to the fetus was rapid with significant concentrations in umbilical venous and fetal arterial plasma 5 min after intra-amniotic drug injection (e.g. Fig. 11). In contrast, MCP was not detected in maternal plasma until between 10 and 20 min following drug administration. Drug concentrations in the ewe remained lower than those in fetal plasma throughout the experimental period, with an average maternal to fetal arterial plasma concentration ratio of 0.13 ± 0.02 (Table 25). Maximum concentrations in fetal plasma (FA, UV) were =8-10 times higher than those measured in the ewe (Table 25). The concentrations of MCP in tracheal fluid were also much higher than those in fetal plasma following this route of administration, with an average TR/FA concentration ratio of 15.3 ± 1.0 (Table 25). These findings are consistent with our previous observations of MCP accumulation and excretion (Table 23 and Section 4.7) in fetal lung fluid and suggest that a portion of the drug taken up by the fetus from amniotic fluid would be returned to amniotic 197 compartment via the lung. The net result of this recirculation appears to be a much longer persistence of MCP in fetal plasma and amniotic and tracheal fluids than in the ewe, with measurable concentrations in these fluids at 28 hours while MCP was not detected in maternal plasma after 10 h. The higher concentrations of MCP in umbilical venous plasma when compared to femoral arterial plasma indicates uptake of metoclopramide from the amniotic fluid by the chorioallantoic membranes. These membranes are in contact with the amniotic membrane (which is poorly vascularized) and receive their blood supply from the umbilical circulation, i.e. the vessels which also supply the placenta (Mellor and Slater, 1974; Brace, 1986). Substances taken up by the membrane vasculature (intercotyledonary chorion) then, would be returned to the fetus via the umbilical vein. The chorioallantoic membranes are well vascularized and receive -6% of the total umbilical blood flow (Makowski et a7., 1968). Using the metoclopramide concentration versus time data obtained from these experiments in conjunction with other kinetic data obtained for MCP from the infusion studies, as well as published values for total umbilical and chorioallantoic blood flows, we have made some estimates as to the routes by which the drug leaves the amniotic fluid compartment. First, the rate at which the fetus swallows amniotic fluid was determined by following the disappearance of ^Cr-1 abel 1 ed fetal red blood cells, as described by Tomoda et a7., 1985. Briefly the red cells are incubated with a known concentration of label and subsequently injected into the amniotic cavity from which their only route of removal is by fetal swallowing. Samples of amniotic fluid are taken over time and a plot of log isotope counts per min versus time made. The amniotic fluid volume is determined by dividing the injected counts by the intercept obtained from the log isotope counts per min versus time plot. An average value of 698.5 ± 145 mL was calculated. 198 The rate of disappearance of the label is then determined from the slope of the regression line as a percentage of the total amniotic fluid volume and expressed as percent per hour, with an average value of 5.3 ± 2.9 %/h being obtained. The rate of fetal swallowing is calculated as the product of amniotic fluid volume times the disappearance rate of the isotope label; an average estimate of 845.5 ± 434.8 mL/day was calculated. Similarly, the disappearance rate of MCP from amniotic fluid was calculated as a percentage of the initial dose. An average value of 16.2 ± 6.6 %/h was obtained which when compared to the estimate for fetal swallowing (i.e. 5.3 ± 2.9 %/h), is =3 times greater and indicates that the drug leaves amniotic fluid by routes in addition to fetal swallowing. Further comparison of the two values indicates that fetal swallowing could account for =33% of the total loss of MCP from amniotic f luid. As previously mentioned, concentrations of MCP were significantly higher in umbilical venous than in fetal arterial plasma (Table 25), indicating uptake of the drug by the chorioallantoic membranes. From the following series of calculations this would appear to be a significant route for MCP loss from amniotic f luid. First, the uptake of the drug via this route was calculated (Eq. 32) and averaged =2,296 ng/min. fetal uptake = (UV-FA)(Qum)(W) = 2,296 ng/min (32) where (UV-FA) is the average MCP concentration difference between umbilical venous and fetal arterial plasma (5.1 ng/mL); Qum is the umbilical flow rate (=220 mL/min/kg; Rudolph and Heymann, 1970; Edelstone, 1980); W is the average estimated fetal weight (Fetuses 237, 287) at the time of the experiments (2.05 kg). By rearranging Eq. 32, correcting Qum for the percentage of blood flow to the membranes (=6%; Makowski et al., 1968) and substitution of the average umbilical arterial concentration supplying the membranes (i.e. [FA] =32.9 ng/mL for fetuses 237 and 287) we can determine 199 the concentration of MCP in blood leaving the chorioallantoic membranes (i.e. let UV = the membrane drug concentration). A concentration value of =118 ng/mL was obtained which is =3.6 times higher than the concentration of MCP in the umbilical arterial blood (i.e. 32.9 ng/mL) supplying the membranes. Total uptake of MCP from the chorioallantoic membranes (2.5 mg) was then calculated by multiplying the value from Eq. 32 by the average time period over which paired umbilical venous and fetal arterial plasma samples were available (18 h) and the result compared to the total amount of MCP eliminated (cleared) by the fetus (7.0 mg) determined from Eq. 33. Total amount of MCP eliminated = (AUC f a)(CL f f) = 7.0 mg (33) where AUCfa is the average area under the fetal arterial MCP concentration time curve for fetuses 237 and 287 (433.7 ng.h/mL); CLff is the average total fetal clearance of MCP obtained from the two-compartment open model used in the calculation of placental and nonplacental clearances (Section 4.7) (131.7 mL/min/kg; Table 20). The comparison of these two values indicates that =36% (i.e. [2.5 mg/7.0 mg] X 100) of the drug eliminated (cleared) by the fetus is received from the chorioallantoic membranes. Also, from the total amount of MCP injected into the amniotic fluid (10 mg, equivalent to 8.5 mg of MCP base), the total apparent uptake of MCP via the chorioallantoic membranes (2.5 mg) and the concentration of MCP in amniotic fluid at the end of the sampling period, we have calculated that =30% of the injected drug left the amniotic fluid via this route (Amount injected = 8.5 mg; amount remaining in amniotic fluid at the end of the sampling period = [average concentration for fetuses 237 and 287 = 306.5 ng/mL][average amniotic fluid volume of 1000 mL] = 0.3 mg. i.e. 2.5 mg/(8.5 mg - 0.3 mg) x 100 =30%). Finally, by using the average area under the plasma concentration versus time curve values for the fetus (AUCfa = 433.5 ng.h/mL) and the ewe (AUCma =46.0 ng.h./mL) and the average values 200 for fetal placental clearance (CLfm = 205.0 mL/min) and maternal nonplacental clearance (CLm o = 3,638.3 mL/min) previously obtained from the infusion studies (Section 4.7; Table 19), we can estimate (Eq. 34) the percentage of drug that actually reached the ewe via the fetus and was eliminated nonplacentally by the ewe was =53%. elimination from ewe = (CLfm)(AUCfa)/(CLmo)(AUCma) x 100 (34) The percentage of drug transferred from amniotic fluid directly to the ewe (by diffusion across the chorioallantoic membranes and uterine wall) can be calculated as follows: direct transfer = [(Cl m o x AUCma) - (CL f m x AUC f a)] total drug loss from amniotic fluid (35) = 0.578 x 100 = 57.8%. Adding the estimated percentages of the injected dose removed from amniotic fluid by fetal swallowing (=33%), by fetal uptake via the fetal membranes (=30%) and by direct transfer to the ewe (=58%) accounts for =121% of the injected dose. Given the assumptions and approximations involved in some of the calculations this seems to be a reasonable degree of agreement. It is important to note that some of the estimates made above could have been affected by recirculation of the drug from the fetus back to amniotic fluid via the kidney or lung. We feel, however, that these processes would have l i t t le impact on our estimates, because of the relatively low concentrations of MCP in tracheal fluid (Fig. 11; Table 25), and presumably present in fetal urine, compared to those in amniotic fluid. In summary, these data suggest preferential uptake of MCP by the fetus from amniotic fluid with the chorioallantoic membranes being a significant element in this process. Similar findings have been made in our laboratory with the antihistamine, diphenhydramine (Yoo et al., 1988) where drug concentrations in the fetus are substantially higher than in the 201 ewe following intra-amniotic injection. In contrast, Szeto et al., 1978, have found the opposite situation with meperidine. Peak plasma concentrations occurred in the ewe 15 min after intra-amniotic administration, while peak concentrations in the fetus were delayed until =75 min and were only -9% of those in maternal plasma. This study appears to involve only a single animal, and it was suggested that the slow appearance of meperidine in the fetal circulation may have been the result of fetal swallowing (which is sporadic) or due to transfer of drug across the placenta from the ewe. It was further suggested that, at least in this case, the amniotic fluid serves as a source of drug for the ewe, rather than being a reservoir of drug for the fetus. We do not have an explanation for this apparent difference, but it points out the need for studies on other drugs following this route of administration. The persistence of vasopressin in the fetal circulation has also been reported (Ervin et a7., 1986) in the chronically catheterized fetal lamb (n = 7) following intra-amniotic drug injection. Fetal swallowing and subsequent gastrointestinal reabsorption was suggested as the route of recirculation. These appear to be the only other studies in sheep involving this route of drug administration in which fetal and/or maternal drug concentrations have been determined. A study involving the injection of prednisolone and a sulfonamide into the amniotic cavity of women in late pregnancy (several hours prior to delivery) has also been reported (Yamaguchi et a7., 1973). Both compounds disappeared slowly from amniotic fluid and concentrations were higher in umbilical venous (cord) than in maternal blood at the time of delivery. It was suggested that the drugs entered the fetus via fetal swallowing and were subsequently transferred to the mother across the piacenta. The anatomic relationships between the fetal lamb and the amniotic and allantoic fluid compartments and potential routes of exchange are illustrated in Fig. 13a. The amniotic membrane surrounds the fetus and contains =1.0 L of fluid in late gestation (Tomoda et al., 1985, 1987). Surrounding the amniotic compartment, although not completely, is the allantoic cavity bounded on inner and outer surfaces by the chorioallantoic membrane. Fluid enters the amniotic compartment via the lungs (=150-450 mL/day; Harding et al., 1984; Brace, 1986; Tomoda et al., 1987) and from the fetal kidneys via the urethra while fluid can leave (in bulk) by fetal swallowing which occurs at a rate of =800-1000 mL/day (Harding et al., 1984; Tomoda et al., 1985, 1987). Fluid enters the allantoic cavity from the kidneys by the urachus, but in contrast to the amniotic compartment there is no route for the exit of fluid in bulk (i.e. via fetal swallowing) from the allantoic space. Diffusional movement of solvent and solutes is possible between allantoic and amniotic fluids and from the former compartment to the ewe and, such exchange has been demonstrated (Mellor and Slater 1971, 1974; Brace, 1986,). The ionic compositions of the two fluids are, however, distinctly different and remain relatively constant with only gradual progressive changes, for at least the last three months of gestation (Mellor and Slater, 1971, 1974). Figure 13b diagrammatically illustrates the major elements that we think contribute to drug disposition and elimination within the conceptus. The arrows linking the various boxes indicate potential uni- or bi-directional routes of exchange. Those routes for which we have obtained at least some data are indicated by solid lines, while those for which information is s t i l l required are denoted by dotted 1ines. Numbers 1-3 indicate routes of drug exchange between allantoic fluid on one hand and amniotic f luid, fetus and mother on the other. The 203 b . Fig. 13. a) A diagrammatic representation of the anatomic relationships between the fetal lamb and the amniotic and allantoic fluid compartments (Modified from Fig. 1, Szeto et al., 1979). b) A diagrammatic illustration of the major elements contributing to drug disposition and elimination within the ovine conceptus. 204 appearance of meperidine in allantoic fluid has been reported in the chronic sheep preparation after maternal i.v. bolus dosing (Szeto et al., 1979). The drug was detectable in allantoic fluid as early as 3 min after maternal administration and the concentrations increased rapidly, while measurable concentrations were not detected in amniotic fluid until about 30 min. The time course for the appearance of drug in these two fluids suggests that meperidine init ial ly diffuses across the chorioallantoic membranes (via the fetal circulation) into the allantoic cavity and subsequently diffuses across the allantoic and amniotic membranes to enter the amniotic compartment. These results support our findings for MCP where as previously discussed, umbilical concentrations exceed fetal arterial, and the latter exceed maternal, suggesting uptake of MCP from amniotic fluid by these membranes. This appears to be the only study in which allantoic drug concentrations have been measured and while, as previously mentioned, there are other reports involving intra-amniotic drug injection there do not appear to be any studies in which drug has been injected into the allantoic cavity. Injections of MCP into allantoic fluid are planned and should provide information on the rapidity and degree of drug uptake by the fetus and ewe from this fluid compartment, as well as the time required for drug to diffuse into amniotic f luid. Route number 4 denotes drug excretion in urine to the amniotic and allantoic cavities. Daily urine production by the fetal lamb in late gestation ranges from =600-1200 mL (Wlodek and Patrick, 1986; Tomoda et al., 1987) with =50% of the volume going to each of the amniotic and allantoic compartments. Urinary excretion has been demonstrated in the chronically catheterized fetal lamb for a number drugs, including meperidine (Szeto et al., 1979), lidocaine (Morishima et al., 1979), methadone (Szeto et al., 1982b), cimetidine (Mihaly et al., 1983), 205 acetaminophen conjugates (Wang et al., 1986) and ethanol (Clarke et al., 1987). Fetal elimination by this route, however, is very low compared to that in the ewe. In the studies with ethanol (Clarke et al., 1987), the role of urinary excretion by the fetal lamb and its contribution to amniotic fluid ethanol concentrations were assessed. Ethanol was administered to the chronically instrumented ewe by a constant rate infusion. All fetal routes of urine entry (bladder, urachus) into the amniotic or allantoic cavities were catheterized and the urine collected throughout the period of study. In spite of this, significant concentrations of ethanol were measured in the amniotic f luid. Comparison of amniotic fluid ethanol concentrations from the catheterized fetuses to controls, in which these catheters were absent, showed no significant differences in peak concentrations, or in the amniotic or blood concentration versus time profiles. Ethanol was measured in the fetal urine, but these findings indicate that while fetal urinary excretion of ethanol occurred, it did not represent a major route of entry into the amniotic f luid. Rather, it would appear that transfer of ethanol into the amniotic fluid occurred primarily by diffusion across the chorioallantoic membranes. This would be consistent with the significant uptake of MCP from amniotic fluid by these membranes, as discussed above. Similar experiments are planned for metoclopramide in order to assess the degree of fetal urinary excretion, and the importance of this route for the appearance of drug in both the amniotic and allantoic fluid compartments. Routes numbered 5-6 indicate gastrointestinal absorption of drug ingested by the fetus and hepatic "first-pass" metabolism of the absorbed drug. There is l i t t le doubt that drug(s) in amniotic fluid are ingested by the fetus through swallowing of this fluid, but the extent of uptake is uncertain given the irregular nature of these episodes as well as the 206 volume swallowed (Harding et al., 1984a). The gastrointestinal absorption of a number of nutrients has been demonstrated including amino acids and glucose following their administration to the chronically instrumented fetal lamb through a duodenal catheter (Charlton, 1984). Similar experiments are planned to determine the extent of this process for MCP. In summary, our experiments indicate significant recirculation of MCP from amniotic fluid to the fetus and ewe, with much higher concentrations and a longer persistence of drug in the fetus. It appears then, that drugs could enter the fetus via fetal swallowing, uptake from the lung (Section 4.7) and through uptake by the chorioallantois, with delivery to the fetus via the umbilical vein. These findings are relevant to drug administration and disposition in human pregnancy and suggest a method of preferential drug administration to the fetus by injection into amniotic f luid. 4.9 Physiological and Behavioural Effects of Metoclopramide on the Fetal Lamb Following Maternal or Fetal Dosing. It is well known that drugs administered to pregnant patients may exert deleterious effects on the fetus and newborn. The teratogenic potential of drugs and the associated structural abnormalities which can occur, in the f irst trimester of pregnancy have been well studied. Similarly, the renal and hepatic immaturity of the newborn is well known (Morselli, 1976) and drugs received prior to delivery, as a result of placental transfer, may persist in the neonate and produce adverse effects after birth. The developing central nervous system (CNS) appears to be particularly susceptible to toxic drug effects, partly because of its extended period of development. While serious CNS morphologic anomalies are usually the result of drug use in early pregnancy during the period of organogenesis (Dobbing, 1976), the use of drugs during the second and third trimesters can result in subtle alterations in CNS function (Hutchings, 1978). It is believed that neurotransmitters are important in controlling cell proliferation in the nervous system. There is , therefore, concern that compounds which affect neurotransmitter activity in the CNS may interfere with brain cell differentiation and proliferation, resulting in permanent change in later functional capacity but without obvious structural malformations (Beeley, 1986). These subtle abnormalities are termed "behavioural teratogenic effects" and involve deficits in CNS activities such as motor function, learning and memory. Most studies of behavioural teratology have involved research for functional impairments after birth. Two areas that have received less attention are fetal CNS function and behaviour during the actual period of drug exposure and the relationship between fetal plasma drug concentrations and observed effects (Kimmel, 1981). With the development of large animal models, such as the chronically catheterized pregnant sheep, a variety of physiological, endocrine and behavioural parameters can be monitored in the undisturbed fetus in utero over the last third of gestation (Nathanielsz, 1980). A number of neurological and behavioural parameters have been recorded in these preparations, including electrocortical (ECoG) activity, eye movements, electromyographic (EMG) activity in limb, neck and trunk muscles, limb and body movements and breathing-like activity (Dawes et al. 1972; Natale et al., 1981; Rigatto et al., 1982; Harding, 1984; Harding et al., 1984a). 4.9.1 Fetal Blood Gases, Heart Rate, Arterial Pressure and Amniotic Pressure Using the chronically catheterized pregnant ewe preparation we have examined some of the physiological and behavioural effects of MCP on the fetal lamb in utero. The maternal administration of MCP by i.v. bolus had no marked effect on fetal arterial pH or blood gases (Table 26). The small, but significant decline (=1.3-1.5 mm Hg) in fetal Po2 at 1 min after dosing is not associated with changes in umbilical blood flow, and so may be due to a transient decline in uterine blood flow (Harding et al., 1981). Similarly, the exposure of the fetal lamb to sustained concentrations of MCP following either maternal or fetal constant rate infusions (for 1.5 hours) resulted in no significant changes in fetal pH or blood gases (Table 27). The transient hypoxemia observed in the i.v. bolus experiments was not seen during the infusions, although in this case the f irst sample was taken 5 min after the initiation of dosing. Control infusions of normal saline had no effect on fetal pH or blood gases (Table 29). Maternal i.v. bolus dosing had no significant effect on fetal arterial pressure or heart rate. Intrauterine (amniotic) pressure was also unaltered. The heart rate and pressure data were more fully analyzed in the infusion studies as the fetus was exposed to a longer dosing interval, again there were no significant changes from control fetal heart rate, fetal arterial pressure or amniotic pressure values (Tables 31 and 32). Control infusions of normal saline also had no effect on these parameters (Table 33). The intra-amniotic injection of MCP also did not affect fetal arterial pH or blood gases (Table 30). 4.9.2 Fetal Breathing-Like and Electrocortical Activities The effect of MCP on fetal behavioural activity (electrocorticogram, breathing movements) was also assessed following the maternal i.v. bolus studies and the steady-state maternal and fetal drug infusions. The fetuses involved in the majority of the maternal i.v. bolus experiments were not equipped with electrocortical (ECoG) electrodes. Analysis of breathing activity in these animals showed no consistent or persistent effects, although there were brief bursts of activity, within 1 to 3 min of MCP administration, in 11 of the 17 experiments. Electrocortical activity was assessed in 4 fetuses following a single maternal 40 mg i.v. bolus of MCP. There were no significant changes in the pattern or duration of low and high voltage activities (Table 34). Similarly the frequency and pattern of breathing movements was not altered in these 4 fetal lambs, occurring during periods of low voltage electrocortical activity (Table 34). The exposure of the fetus to sustained steady-state concentrations of MCP during the maternal infusions as well as direct infusion of the drug to the fetal lamb also had no significant effects on the fetal breathing or electrocortical activities (Tables 35 and 36). Low and high voltage ECoG activities were unaltered as was the frequency and occurrence of fetal breathing movements. Likewise, the control infusions of normal saline did not alter these fetal behavioural states (Table 37). Neither have we observed any adverse drug effects in infants during a previous study of MCP placental transfer in 23 patients undergoing general anaesthesia for elective Caesarian section for healthy term pregnancies (Bylsma-Howell et al., 1983). Metoclopramide (10 mg), or an equivalent volume of normal saline for injection, was administered to the mother by i.v. bolus prior to induction of anaesthesia. Heart rate and systolic blood pressure were measured in the newborn one hour after delivery. Apgar scores were assessed at 1 and 5 min after birth and the infants were also subjected to an extensive set of neurological and adaptive capacity tests (NACS Neurobehavioural test; Amiel-Tison et al., 1982) at 2, 4, 6 and 24 hours following delivery. There were no significant differences in Apgar scores, cardiovascular parameters or neurobehavioural scores between the treated and untreated groups of neonates. In a similar study Cohen et al., 210 1984, also found no differences in Apgar or NACS scores between infants born to mothers receiving either MCP or a saline placebo. Although we observed no effect of MCP on the physiologic parameters measured (i.e. fetal heart rate and arterial pressure, arterial pH and blood gases), there are reports that it stimulates the release of various hormones, both in man and in animals. Prolactin release has been observed in the rat (Harrington et al., 1983), in healthy adults (Harrington et al., 1983), pregnant women (Arvela et al., 1983, Harrington et al., 1983), in newborns (Ruppert et a7., 1986) and children (Harrington et a7., 1983) apparently as a result of blockade of dopamine-mediated inhibition in the pituitary. Prolactin release has also been reported in the pregnant ewe following i.v. bolus dosing (10 mg) (Fitzgerald and Cunningham, 1982). Although there are as yet no published reports of prolactin release in the ovine fetus following maternal or fetal dosing, personal communication with Rl Stark, M.D. (Columbia University) indicates that this also occurs. Considering the central dopamine antagonistic effects of MCP this would seem likely since prolactin secretion has been shown to be under dopaminergic control in the fetal lamb during the last third of gestation (Gluckman et a7., 1979). Prolactin has been shown to reduce water transfer across the amnion (Holt and Perks, 1975; Leontic et a7., 1979) and to increase the secretion rate of fetal lung fluid, at least when placed directly in the lungs (Cassin and Perks, 1982). There is also evidence that prolactin may be important in fetal pulmonary maturation (Schel1enberg et a7., 1988). Metoclopramide, then via its action on fetal prolactin levels could have several effects, particularly i f the drug was administered for a prolonged period. Increases in plasma aldosterone have also been reported in rats (Harrington et a7., 1983) and man (Harrington et a7., 1983; Sommers et a7., 1988) apparently as a result of inhibition of 211 central dopamine receptors (Albibi and McCallum, 1983). More recent information, however, suggests that the aldosterone response to MCP is mediated by acetylcholine release from post-ganglionic cholinergic nerve terminals within the adrenal cortex (Sommers et al., 1988). Stimulation of vasopressin secretion, by an as yet unknown mechanism, has also been reported in healthy human volunteers following a single 10 mg i.v. bolus dose (Norbiato et al., 1986). The effect of MCP on the secretion of growth hormone (Harrington et al., 1983) and thyrotropin (Ruppert et al., 1986) remains controversial, with some studies reporting no change and others a stimulated release. The extent of changes in endocrine function in the ovine fetus as a result of maternal and fetal MCP administration remains to be evaluated. 212 5. SUMMARY AND CONCLUSIONS The previously developed capillary GLC-ECD assay for MCP in human and ovine plasma (Riggs et al., 1983) has been modified and applied to study the pharmacokinetics of MCP in plasma obtained from nonpregnant ewes, pregnant ewes and fetal sheep as well as in fetal tracheal and amniotic fluids following i.v. bolus dosing, constant rate infusions to steady-state and intra-amniotic injection. Changes to the method involved: i . a switch from the split to splitless mode of sample injection i i . an increase in final sample reconstitution volume from 0.2 to 0.8 mL. These modifications have provided the following advantages: i . automatic sample injection which was not possible with the previous split mode of sample introduction. i i . improved sensitivity with a decrease in the lower limit of the assay from the previous low of 4 to 2 ng/mL. The pharmacokinetics of MCP were studied in chronically catheterized pregnant (n = 5) and nonpregnant ewes (n = 5). MCP was administered to the ewe by i.v. bolus injections, on a crossover basis, of 10, 20 and 40 mg, with an additional 80 mg dose in the nonpregnant animals. An additional 4 pregnant ewes received a single 40 mg bolus dose only. Serial arterial blood samples were collected from the ewe and/or fetus over six hours. Amniotic and fetal tracheal fluid samples were also collected from the ewes receiving the single 40 mg bolus dose. The findings from these studies may be summarized as follows: i . transfer of MCP from the ewe to the fetal lamb was rapid with peak concentrations in the initial 1 min sample in most cases i i . there was significant fetal exposure to the drug, with an average fetal to maternal AUC ratio of 0.74. 213 i i i . MCP concentrations followed a biexponential process in the majority of pregnant ewes and their fetuses, with a delay to peak concentration in 4 experiments. Plasma concentration-time profiles were biexponential in all cases for the nonpregnant animals. iv. terminal elimination half-lives in maternal and fetal plasma averaged 71.3 and 86.8 min, respectively, with fetal half- l i fe being significantly longer. An elimination half- l i fe of 67.5 min was calculated in the nonpregnant ewes. v. pharmacokinetic parameters were compared in the pregnant and nonpregnant ewes at the 10, 20 and 40 mg doses, with no significant differences except for total body (systemic) clearance which was 24% higher in the nonpregnant animals. This may be accounted for by lower hepatic blood flow in pregnant compared to nonpregnant sheep, since MCP's clearance is thought to be limited by liver blood flow. v i . the area under the plasma concentration versus time curves for the ewes showed a good correlation with increasing dose, indicating that the drug followed dose-independent kinetics over the respective 4- and 8-fold dose ranges studied in the pregnant and nonpregnant animals. v i i . volume of distribution at steady-state was determined by compartmental and noncompartmental methods for the pregnant and nonpregnant ewes and showed good agreement with an average difference of =6%. These results lend support to the premise that the data were adequately described by the proposed two compartment open model (Scheme 2) with elimination occurring from the central compartment. v i i i . maternal MCP administration resulted in minimal fetal effects with no change in arterial pressure, heart rate, arterial pH or Pco2 and only a small (=1.8 mm Hg), transient decline in Po 2. Amniotic pressure was also not affected. Fetal breathing-like and electrocortical activities were 214 also assessed in the single maternal 40 mg i.v. bolus studies, with no consistent or significant changes from control. Although not reported in this thesis, MCP had no effect on maternal heart rate or arterial pressure. ix. MCP accumulates in amniotic and fetal tracheal fluids, with tracheal fluid concentrations exceeding those in fetal plasma by an average of 15-fold. x. tracheal fluid could represent a significant route of drug elimination in the fetus. x i . since fetal lung fluid enters the amniotic fluid compartment or is swallowed, the high concentrations in tracheal fluid may contribute significantly to amniotic drug levels. x i i . recirculation of MCP in the fetus as a result of swallowed tracheal and/or amniotic fluids and diffusion across the pulmonary epithelium and chorioallantoic membranes may be, at least in part, responsible for the larger half- l i fe of MCP in fetal plasma compared with that in the ewe. x i i i . it would appear that at least some of the cellular elements within the lung could be exposed to very high drug concentrations; this would not be predicted by estimates of relative fetal drug exposure based on fetal/maternal plasma concentration ratios. The pharmacokinetics of MCP were also examined in 9 chronically instrumented pregnant ewes following paired maternal and fetal infusions to steady-state. The infusions were separated by 48-72 hours. MCP was administered to the ewe as a 15 mg i.v. bolus loading dose followed by a constant rate infusion of 0.21 mg/min for 90 min. The fetal loading dose and infusion rate were one-third those used in the maternal experiments. Serial samples of maternal and fetal arterial blood, amniotic and fetal tracheal fluid were collected for 6 hours following termination of the 215 infusion in most cases. In some preparations sampling was continued for 28 hours. The following summarized results were obtained: i . like the i.v. bolus studies transfer of MCP across the placenta from the ewe to the fetus was rapid with peak concentrations in the initial 5 min sample. Transfer of MCP from the fetus to the ewe was also rapid with peak concentrations in maternal plasma in the initial 5 min sample, in most cases, following fetal drug administration. i i . steady-state plasma MCP concentrations were attained in the ewe and fetus between 30 and 45 min after the initiation of maternal or fetal dosing. An average fetal/maternal steady-state concentration ratio of 0.57 was obtained during the maternal infusion experiments. i i i . maternal concentration versus time curves were adequately described by a two compartment open model (Scheme 2) following a biexponential decline after stopping the maternal infusion. This was also true for MCP in fetal arterial plasma following termination of the fetal infusions. iv. terminal elimination half-lives in the ewe and fetus averaged 64.7 and 122.9 min in the ewe and fetus, respectively, following the maternal infusions. Like the i.v. bolus studies, fetal half- l i fe was significantly longer. Respective values of 87.7 and 115.9 min were calculated in the ewe and fetus following the fetal infusions and were not significantly different. As previously mentioned, recirculation of drug within the fetus due to swallowing of amniotic and/or tracheal fluids as well as diffusion of drug across fetal membranes may be partly responsible for the longer persistence of drug in the fetus compared to the ewe. v. there was very good agreement between the estimates obtained for maternal terminal elimination half- l i fe, volume of distribution and total 216 body clearance following the i.v. bolus and steady-state infusion experiments. v i . protein binding was determined in maternal and fetal plasma at steady-state using ultrafiltration, with values of 49.5% and 39.5%, respectively, similar to that reported in humans (=40%). The extent of binding was significantly lower in the fetus. v i i . as observed in the i.v. bolus studies, MCP accumulated in amniotic and fetal tracheal f luid, with tracheal drug concentrations exceeding those in fetal arterial plasma by 13-15-fold following either maternal or fetal dosing. MCP persisted in both of these fluids for more than 28 hours after the infusion was stopped following an apparent triexponential process in tracheal fluid and a monoexponential decline in amniotic f luid. The reported binding of MCP to mitochondria as well as possible binding to surfactant within lung cells and in tracheal fluid itself may result in reservoirs ("pools") of drug within and on the surface of the lung. The diffusion and release of MCP from pools of varying "depths" into tracheal f luid, then, may be responsible for the triexponential decline observed in this f luid. Release of drug from these sites, and reentry into the fetal circulation may also account for the longer fetal than maternal elimination half - l i fe . v i i i . placental (transplacental) and nonplacental clearances of MCP were determined using a two compartment open model (Scheme 3) with elimination occurring from both compartments (Szeto et al., 1982b). Total clearance of MCP from the fetal compartment averaged 7.9 L/h/kg and was significantly greater than that in the ewe (2.9 L/h/kg) on a weight corrected basis. The transplacental clearance of MCP from the fetus to the ewe averaged 6.2 L/h/kg and was significantly larger than that from the maternal to the fetal compartment (4.3 L/h/kg). This may be due to the lower extent of 217 plasma protein binding in the fetus resulting in higher fetal concentrations of free drug being available for diffusion and/or because of the greater proportion of cardiac output delivered to the placenta by the fetus (=42%) compared to the ewe (=16%). Nonplacental elimination of MCP by the fetus accounts for =21% of the total fetal drug clearance. Specifically which fetal nonplacental routes are involved in MCP's elimination remains to be more fully evaluated. As previously mentioned, the excretion of MCP in fetal lung fluid appears to represent a significant route of elimination. The accumulation of MCP in amniotic fluid may also indicate renal elimination, since both fetal lung fluid and urine are responsible for amniotic fluid volume. This, however, remains to be established as does possible biotransformation by the fetal l iver. Maternal nonplacental clearance accounts for =96% of the total maternal clearance. As in the fetus, the nonplacental routes of MCP elimination in the ewe remain to be established although qualitative studies of maternal urine (Appendix) suggest considerable biotranformation and renal excretion of metabolites and intact drug. ix. the placental clearance of MCP (6.7 L/h/kg) is similar to that reported for antipyrine (=4.5-7.2 L/h/kg) suggesting that its placental transfer is flow-limited and that its rate of transfer is largely determined by the rate of delivery of drug to the exchange surface. x. as observed in the maternal i.v. bolus experiments, there were no marked or consistent effects of MCP on the fetus during either the maternal or fetal drug infusions, with no significant differences from control values for arterial blood gases, heart rate, arterial pressure, amniotic pressure, electrocortical or breathing-like activities. The pharmacokinetics of MCP were also studied in 4 pregnant ewes and their fetuses following intra-amniotic drug injection (10 mg bolus). 218 Serial maternal arterial, fetal arterial and umbilical blood, amniotic and fetal tracheal fluid were collected for 28 hours after drug administration. The results of these experiments can be summarized as follows: i . MCP appeared rapidly in the fetal circulation with measurable concentrations in umbilical venous and arterial plasma in the initial 5 min sample. i i . Measurable concentrations were not obtained in maternal plasma until between 10-20 min following intra-amniotic drug injection and concentrations remained lower than those in fetal plasma throughout the period of study, with an average fetal/maternal concentration ratio of 0.13. i i i . Drug accumulation in tracheal fluid also occurred following this route of administration, again exceeding fetal plasma concentrations by 15-fold. iv. umbilical venous MCP concentrations were consistently higher than those in the fetal artery throughout the f irst 1.5 hours following drug injection suggesting uptake of MCP from amniotic fluid by the chorioallantoic membranes with delivery to the fetus via the umbilical vein. There were brief intervals of arterial crossover following this init ial interval, however, umbilical venous concentrations were significantly higher than those in arterial plasma, with an average umbilical venous/fetal arterial concentration ratio of 1.3. v. elimination half- l i fe in fetal arterial plasma averaged 5.1 hours and is =2.5-3.5 times larger than the estimates calculated following the maternal i.v. bolus and paired maternal and fetal infusion experiments. Maternal elimination half- l i fe (mean 4.3 hours) was also larger than (=2.5-4-fold) the values calculated following intravascular drug administration. 219 v i . MCP was not detected in maternal plasma 10 hours after injection, but was s t i l l present in fetal arterial and umbilical venous plasma at 28 hours in two fetuses. Measurable concentrations were also present in tracheal and amniotic fluids at 28 hours. v i i . MCP appears to enter the maternal circulation as a result of transfer across the placenta from the fetus as well as by direct diffusion across the chorioallantoic membranes and uterine wall. v i i i . the findings of these experiments suggest that amniotic fluid serves as a reservoir for the fetus with drug entering the fetal circulation via fetal swallowing and uptake by the chorioallantois. These observations are relevant to drug administration and disposition in human pregnancy and suggest a method of preferential drug administration to the fetus by injection into amniotic fluid. 220 6. REFERENCES Albert KS, Sedman AJ and Wagner JG. Pharmacokinetics of orally administered acetaminophen in man. J Pharmacokinet Biopharm 2:381-393, 1974. Albibi R and McCallum RW. Metoclopramide: Pharmacology and clinical application. Ann Int Med 98:86-95, 1983. Altman PL and Dittner DS (Eds.). Biology Data Book (2nd ed.), Vol. III. Federation of American Societies for Experimental Biology. Bethesda MA, pp 1703-1705, 1972. Amiel-Tison C, Barrier G, Schnider SM, Levinson G, Hughes SC and Stefani SJ. A new neurologic and adaptive capacity scoring system for evaluating obstetric mediciation in full-term newborns. Anesthesiology 56:340-350, 1982. Anderson MW, Orton TC, Pickett RD and Eling TE. Accumulation of amines in the isolated perfused rabbit lung. J Pharmacol Exp Ther 189:456-466, 1974. Anderson DF, Phernetton TM and Rankin JHG. The measurement of placental drug clearance in near-term sheep: Indomethacin. J Pharmacol Exp Ther 213:100-104, 1980. Anderson DF, Phernetton TM and Rankin JHG. Prediction of fetal drug concentrations. Am J Obstet Gynecol 137:735-738, 1980a. Anderson DF, Phernetton TM and Rankin JHG. The placental transfer of acetyl salicylic acid in near-term ewes. Am J Obstet Gynecol 136:814-818, 1980b. Anderson DF, Bissonnette JM, Faber JJ and Thornburg KL. Central shunt flows and pressures in the mature fetal lamb. Am J Physiol 241:H60-H66, 1981. Arita T, Hori R, Ito K, Ichikawa K and Uesugi T. Transformation and excretion of drugs in biological systems. II. Transformation of metoclopramide in rabbits. Chem Pharm Bull 18:1663-1669, 1970. Arvela P, Jouppila R, Kauppila A, Pakarinen A, Pelkonen 0 and Tuimala R. Placental transfer and hormonal effects of metoclopramide. Eur J Clin Pharmacol 24:345-348, 1983. Bakke 0M and Segura J . The absorption and elimination of metoclopramide in three animal species. J Pharm Pharmac 28:32-39, 1976. Bateman DH, Kahn C, Mashiter K and Davies DS. Pharmacokinetics and concentration-effect relationships with intravenous metoclopramide. Br J Clin Pharmacol 6:401-407, 1978. Bateman DN and Gokal R. Metoclopramide in renal failure. Lancet 1:982, 1980. Bateman DN, Kahn C and Davies DS. The pharmacokinetics of metoclopramide in man with observations in the dog. Br J Clin Pharmac 9:371-377, 1980. 221 Bateman DN, Gokal R, Dodd TRP and Blain PG. The pharmacokinetics of single doses of metoclopramide in renal failure. Eur J Clin Pharmacol 19:437-441, 1981. Bateman DN. Clinical pharmacokinetics of metoclopramide. Clin Pharmacokin 8:523-529, 1983. Bateman DN, Craft AW, Nicholson E and Pearson ADJ. Dystonic reactions and the pharmacokinetics of metoclopramide in children. Br J Clin Pharmac 15:557-559, 1983. Beckett AH, Behrendt WA and Hadzija BW. Bioavailability of controlled-release metoclopramide. Arzneim-Forsch/Drug Res 37:221-224, 1987. Beeley L. Adverse effects of drugs in later pregnancy. Clin Obstet Gynaecol 13:197-214, 1986. Bend JR, Serabjit-Singh CJ and Philpot RM. The pulmonary uptake, accumulation, and metabolism of xenobioticsi Ann Rev Pharmacol Toxicol 25:97-125, 1985. Benford DJ and Bridges JW. Xenobiotic metabolism in the lung. In: Progress in Drug Metabolism, Vol. 9. Bridges JW and Chasseaud LF (Eds.). Taylor and Francis Ltd, pp 53-94, 1986. Bloedow DC, Ralston DH and Hargrove JC. Lidocaine pharmacokinetics in pregnant and nonpregnant sheep. J Pharm Sci 69:32-37, 1980. Bogaert MC and Thiery M. Pharmacokinetics and pregnancy. Europ J Obstet Gynec Reprod Biol 16:229-235, 1983. Boulos BM, Davis LE, Larks SD, Larks GG, Sirtori CR and Almond CH. Placental transfer of drugs. IV. Placental transfer of chlorpromazine, pentobarbital and phenylbutazone with fetal electrocardiographic changes as determined by direct lead tracing. Arch Int Pharmacodyn 194:403-414, 1971. Boxenbaum HG, Riegelman S and Elashoff RM. Statistical estimations in pharmacokinetics. J Pharmacokinet Biopharm 2:123-148, 1974. Boyd RDH, Haworth C, Stacey TE and Ward RHT. Permeability of the sheep placenta to unmetabolized polar non-electrolytes. J Physiol 256:617-634, 1976. Brace RA. Amniotic fluid volume and its relationship to fetal fluid balance: Review of experimental data. Sem Perinatol 10:103-112, 1986. Brazier JL, Ribon B, Desage M, Comet F, Lievre M, Berland M and Salle B. Placental tranfer model in the pregnant ewe: Use of stable isotopes. Dev Pharmacol Ther 7:52-59, 1984. Brien JF, Loomis CW, Tranmer J and McGrath M. Disposition of ethanol in human maternal venous blood and amniotic f luid. Am J Obstet Gynecol 146:181-186, 1983. 222 Brien JF, Clarke DW, Richardson B and Patrick J . Disposition of ethanol in maternal blood, fetal blood, and amniotic fluid of third-trimester pregnant ewes. Am J Obstet Gynecol 152:583-590, 1985. Brock-Utne JG, Dow TGB, Welman S, Dimopoulos, BE and Moshal MG. The effect of metoclopramide on the lower oesophageal sphincter in late pregnancy. Anaesth Inten Care 6:26-29, 1978. Brock-Utne JG, Downing JW and Mankowitz E. Drugs and the fetomaternal unit. S Afr Med J 58:366-369, 1980. Bylsma-Howell M, Riggs KW, McMorland GH, Rurak DW, Ongley R, McErlane B, Price JDE and Axelson JE. Placental transport of metoclopramide: Assessment of maternal and neonatal effects. Can Anaesth Soc J 30:487-492, 1983. Cassin S and Perks AM. Studies of factors which stimulate lung fluid secretion in fetal goats. J Dev Physiol 4:311-325, 1982. Casteels-van Daele M, Jacken J , Van der Schueren P, Zimmerman A and Van Den Bon P. Dystonic reactions in children caused by metoclopramide. Arch Dis Child 45:130-133, 1970. Charlton V. Fetal nutritional supplementation. Sem Perinatol 8:25-30, 1984. Chestnut DH, Vandewalker GE, Owen CL, Bates JN and Choi WW. Administration of metoclopramide for prevention of nausea and vomiting during epidural anesthesia for elective cesarean section. Anesthesiology 66:563-566, 1987. Chiba K, Ishizaki T, Tabuchi T, Wagatsuma T and Nakazawa Y. Antipyrine disposition in relation to lowered anticonvulsant plasma level during pregnancy. Obstet Gynecol 60:620-626, 1982. Ching MS, Jones DB, Morgan DJ, Mihaly GW, Hardy KJ and Smallwood RA. Fetal exposure to cimetidine following chronic administration to pregnant sheep. Res Comm Chem Path Pharmacol 50:139-142, 1985. Chou R and Levy G. Effect of pregnancy on the pharmacokinetics of phenytoin in rats. J Pharmac Exp Ther 229:351-358, 1984,. Clarke DW, Patrick J , Wlodek ME, Smith GN, Richardson B and Brien JF. The role of fetal urinary excretion in the transfer of ethanol into amniotic fluid after maternal administration of ethanol to the near-term pregnant ewe. Can J Physiol Pharmacol 65:1120-1124, 1987. Cohen SE, Jasson J , Talafre M, Chauvelot-Moachon L and Barrier G. Does metoclopramide decrease the volume of gastric contents in patients undergoing cesarean section? Anesthesiology 61:604-607, 1984. Collins JM and Dedrick RL. Contribution of lungs to total body clearance: Linear and nonlinear effects. J Pharm Sci 71:66-70, 1982. Comline RS and Silver M. Recent observations on the undisturbed fetus in utero and its delivery. In: Recent Advances in Physiology. Linden RJ (Ed). Churchill Livingstone, Edinburgh, p 405, 1974. 223 Cowan DA, Huizing G and Beckett AH. Identification of four new metabolic products of metoclopramide using mass spectrometry. Xenobiotica 6:605-616, 1976. D'Arcy PF and McElnay JC. Drug interactions involving the displacement of drugs from plasma protein and tissue binding sites. Pharmac Ther 17:211-220, 1982. Dawes GS, Fox HE, Leduc BM, Liggens GC and Richards RT. Respiratory movements and rapid eye movement sleep in the foetal lamb. J Physiol 220:119-143, 1972. De Vil l iers M, Parkin D, Van Jaarsveld P and Van der Walt B. A radioimmunoassay for metoclopramide. J Immunol Methods 103:33-39, 1987. de Vries JIP, Visser GHA and Prechtl HFR. The emergence of fetal behaviour. I. Qualitative aspects. Early Hum Dev 7:301-322, 1982. de Vries JIP, Visser GHA and Prechtl HFR. The emergence of fetal behaviour. II. Quantitative aspects. Early Hum Dev 12:99-120, 1985. de Zeeuw RA, Bogusz M, Franke J and Wijsbeek J . Drug screening by capillary column gas chromatography. Advances in Analytical Toxicology, Volume 1. Baselt RC (ed.) Biomedical Publications, Foster City CA, 1984. Denisoff 0 and Molle L. Binding study of benzamides to serum albumins by fluorescence probe technique. Drug Res 28:2156-2157, 1978. Desmond PV and Watson KJR. Metoclopramide - a review. Med J Australia 144:366-368, 1986. Dilts PV Jr. Placental transfer. Clin Obset Gynecol 24:555-559, 1981. Dobbing J . Vulnerable periods in brain growth and somatic growth. In: The Biology of Fetal Human Growth. Roberts DF and Thomson EM (eds.) Taylor and Francis Ltd., London,.pp 137-147, 1976. Dvorchik BH, Woodward G, Sitar DS and Tweed WA. Hydroxylation and glucuronidation of various xenobiotics by hepatic microsomes from the fetal lamb, pregnant ewe and human fetus. Dev Pharmacol Ther 9:282-289, 1986. Edelstone DI. Regulation of blood flow through the ductus venosus. J Dev Physiol 2:219-238, 1980. Ervin MG, Ross MG, Leake RD and Fisher DA. Fetal recirculation of arginine vasopressin. Am J Physiol 250:E253-E258, 1986. Faraggiana T, Vil lari D, Jagirdar J and Patil J . Expression of sialic acid on the alveolar surface of adult and fetal human lungs. J Histochem Cytochem 34:811-816, 1986. Fitzgerald BP and Cunningham FJ. Effects of metoclopramide and bromocriptine on prolactin secretion in the pregnant ewe. J Endocr 93:41-46, 1982. 224 Freeman RR. High Resolution Gas Chromatography (2nd ed.). Hewlett-Packard Co., Avondale PA, 1981. Friesen C, Yarnell R, Bachman C, Meatheral R and Biehl D. The effect of lidocaine on regional blood flows and cardiac output in the non-stressed and stressed foetal lamb. Can Anaesth Soc J 33:130-137, 1986. Gardner MJ, Schatz M, Cousins L, Zeiger R, Middleton E and Jusko WJ. Longitudinal effects of pregnancy on the pharmacokinetics of theophylline. Eur J Clin Pharmacol 31:289-295, 1987. Gibaldi M. Estimation of the pharmacokinetic parameters of the two-compartment open model from post-infusion plasma concentration data. J Pharm Sci 58:1133-1135, 1969. Gibaldi M. Effect of mode of administration on drug distribution in a two-compartment open system. J Pharm Sci 58:327-331, 1969a. Gibaldi M and Perrier D. Pharmacokinetics. Swarbrick J (Ed.). Marcel Dekker, New York NY, 1975. Gibaldi M and Weintraub H. Some considerations as to the determination and significance of biologic half - l i fe. J Pharm Sci 60:624-626, 1971. Gibaldi M and Perrier D. Pharmacokinetics. 2nd Edition. Swarbrick J (Ed.). Marcel Dekker, New York NY, 1982. Gipson SL, Stovall TG, Elkins TE and Crumrine RS. Pharmacologic reduction of the risk of aspiration. South Med J 79:1356-1358, 1986. Gluckman PD, Marti-Henneberg C, Thomsett MJ, Kaplan SL, Rudolph AM and Grumbach MM. Hormone ontogeny in the ovine fetus. VI. Dopaminergic regulation of prolactin secretion. Endocrinology 105:1173-1177, 1979. Gourley GR, Mogilevshky W and Odell GB. Protein binding of metoclopramide. J Pediatrics 101:631-632, 1982. Graffner C, Lagerstrom P and Lundborg P. Pharmacokinetics of metoclopramide intravenously and orally determined by liquid chromatography. Br J Clin Pharmac 8:469-474, 1979. Gresham EL, Rankin JHG, Makowski EL, Meschia G and Battaglia FC. An evaluation of fetal renal function in a chronic sheep preparation. J Clin Invest 51:149-156, 1972. Grimes JD, Hassan MN and Preston DN. Adverse neuroleptic effects of metoclpramide. CMA J 126:23-25, 1982. Gugler R, Brand M and Somogyi A. Impaired cimetidine absorption due to antacids and metoclopramide. Eur J Clin Pharmacol 20:225-228, 1981. Haagsman HP and van Golde LMG. Lung surfactant and pulmonary toxicology. Lung 163:275-303, 1985. Harding R. Perinatal development of laryngeal function. J Dev Physiol 6:249-258, 1984. 225 Harding R, Booking AD, Sigger JN and Wickham PJD. Composition and volume of fluid swallowed by fetal sheep. Quart J Exp Physiol 59:487-495, 1984. Harding R, Sigger JN, Poore ER and Johnson P. Ingestion in fetal sheep and its relation to sleep states and breathing movements. Quart J Exp Physiol 69:477-486, 1984a. Harding R, Sigger JN, Wickham PJD and Booking AD. The regulation of flow of pulmonary fluid in fetal sheep. Respir Phsyiol 57:47-59, 1984b. Harrington RA, Hamilton CW, Brogden RN, Linkewich JA, Romankiewicz JA and Heel RC. Metoclopramide: An updated review of its pharmacological properties and clinical use. Drugs 25:451-494, 1983. Havsteen H, Nielsen H and Kjaer M. Antiemetic effect and pharmacokinetics of high dose metoclopramide in cancer patients treated with cisplatin-containing chemotherapy regimens. Eur J Clin Pharmacol 31:33-40, 1986. Hellstern A, Hellenbrecht D, Sailer R, Gatzen M, Manus B, Achtert G, Brockmann P, Hausleiter HJ and Leuschner U. Absolute bioavailability of metoclopramide given orally or by enema in patients with normal liver function or with cirrhosis of the liver. Arzneim-Forsch/Drug Res 37:733-736, 1987. Hill MD, Craft JB and Abramson FP. Elevated plasma glycoproteins in fetal sheep do not bind basic lipophilic drugs. J Pharm Sci 75:881-882, 1986. Hill MD and Abramson FP. The significance of plasma protein binding on the fetal/maternal distribution of drugs at steady-state. Clin Pharmacokinet 14:156-170, 1988. Holt WF and Perks AM. The effect of prolactin on water movement through the isolated amniotic membrane of the guinea pig. Gen Com Endocrinol 26:153-164, 1975. Howard FA and Sharp DS. Effect of metoclopramide on gastric emptying during labour. Br Med J 1:446-448, 1973. Huizing G, Beckett AH and Segura J . Rapid thin-layer chromatographic photodensitometric method for the determination of metoclopramide and cleobopride in the presence of some of their metabolic products. J Chromatogr 172:227-237, 1979. Hutchings DE. In: Studies on the Development of Behaviour and the Nervous System. Early Influences, Vol. 4, Gottleib G (ed.). Academic Press, New York, NY, pp 7-34, 1978. Israel R, O'Mara V, Austin B, Bellucci A, and Meyer BR. Metoclopramide decreases renal plasma flow. Clin Pharmacol Ther 39:261-264, 1986. Jansen AH and Chernick V. Development of respiratory control. Physiol Rev 63:437-483, 1983. Jansen AH, Ioffe S and Chernick V. Drug-induced changes in fetal breathing activity and sleep state. Can J Physiol Pharmacol 61:315-324, 1983. 226 Jennings W. Gas Chromatography with Glass Capillary Columns (2nd ed.). Academic Press, New York NY, 1980. Joss RA, Galeazzi RL, Bishcoff AK, Pirovino M, Ryssel HJ and Brunner KW. The antiemetic activity of high-dose alizapride and high-dose metoclopramide in patients receiving cancer chemotherapy: A prospective, randomized, double-blind t r ia l . Clin Pharmacol Ther 39:619-24, 1986. Juchau MR. Drug biotransformation reactions in the placenta. In: Perinatal Pharmacology and Therapeutics. Mirkin BL (Ed). Academic Press, New York, NY, pp 71-118, 1976. Juchau MR, Chao ST and Omiecinski CJ. Drug metabolism by the human fetus. Clin Pharmacokinet 5:320-339, 1980. Kapil RP, Axelson JE, Ongley R and Price JDE. Nonlinear bioavailability of metoclopramide in the rat: Evidence for saturable first-pass metabolism. J Pharm Sci 73:215-218, 1984. Kauppila A, Arvela P, Koivisto M, Kivinen S, Ylikorkala 0 and Pelkonen 0. Metoclopramide and breast feeding: Transfer into milk and the newborn. Eur J Clin Pharmacol 25:819-823, 1983. Kennedy RL, Miller RP, Bell JU, Doshi D, deSousa MD, Kennedy MJ, Heald DL and David Y. Uptake and distribution of bupivacaine in fetal lambs. Anesthesiology 65:247-253, 1986. Kimmel CA. A profile of developmental toxicology. In: Developmental Toxicology. Kimmel CA and Buelke-Sam J (eds.). Raven Press, New York, NY, pp 321-331, 1981. Kirch W, Janisch HD, Santos SR, Duhrsen U, Dylewicz P and Ohnaus EE. Effect of cisapride and metoclopramide on digoxin bioavailability. Eur J Drug Metab Pharmacokinet 11:249-250, 1986. Krauer B, Krauer F and Hytten FE. Drug disposition and pharmacokinetics in the maternal-placental-fetal unit. Pharmac Ther. 10:301-328, 1980. Krauer B, Nau H, Dayer P, Bischof P and Anner R. Serum protein binding of diazepam and propranolol in the feto-maternal unit from early to late pregnancy. Br J Obstet Gynaecol 93:322-328, 1986. Kwong TC. Free drug measurements: Methodology and clinical significance. Clin Chim Acta 151:193-216, 1985. Lehmann CR, Heironimus JD, Collins CB, 0'Nei1 TJ, Pierson WP, Crowe JT, Melikian AP and Wright GJ. Metoclopramide kinetics in patients with impaired renal function and clearance by hemodialysis. Clin Pharmacol Ther 37:284-289, 1985. Leontic EA, Schruefer JJ , Andreassen B, Pinto H, and Tyson JE. Further evidence for the role of prolactin on human fetoplacental osmoregulation. Am J Obstet Gynecol 133:435-438, 1979. 227 Levy G. Pharmacokinetics of fetal and neonatal exposure to drugs. Obstet Gynecol 58:9S-16S, 1981. Levy G, Gibaldi M and Jusko WJ. Multicompartment pharmacokinetic models and pharmacologic effects. J Pharm Sci 58:422-424, 1969. Levy G, and Hayton WL. Pharmacokinetic aspects of placental drug transfer. In Fetal Pharmacology, Boreus L. (Ed.), Raven Press, New York NY, pp 29-39, 1973. Lindeberg S, Lundborg P, Regardh CG and Sandstrom B. Disposition of the adrenergic blocker metoprolol and its metabolite OH-metoprolol in maternal plasma, amniotic fluid and capillary blood of the neonate. Eur J Clin Pharmacol 33:363-368, 1987. Low LCK and Goel KM. Metoclopramide poisoning in children. Arch Dis Child 55:310-312, 1980. Lumbers ER. A brief review of fetal renal function. J Dev Physiol 6:1-10, 1983. Makowski EL, Meschia G, Droegemueller W and Battaglia FC. Measurement of umbilical arterial blood flow to the sheep placenta and fetus in utero. Circ Res 23:623-631, 1968. McGarry JM. A double-blind comparison of the anti-emetic effect during labour of metoclopramide and perphenzaine. Brit J Anaesth 43:613-615, 1971. McLeod W, Brien J , Loomis C, Carmichael L, Probert C and Patrick J . Effect of maternal ethanol ingestion on fetal breathing movements, gross body movements, and heart rate at 37 to 40 weeks' gestational age. Am J Obstet Gynecol 145:251-257, 1983. Mellor DJ and Slater JS. Daily changes in amniotic and allantoic fluid during the last three months of pregnancy in conscious, unstressed ewes, with catheters in their foetal fluid sacs. J Physiol 217:573-604, 1971. Mellor DJ and Slater JS. Some aspects of the physiology of sheep foetal fluids. Br Vet J 130:238-247, 1974. Mescher EJ, Platzker ACG, Ballard PL, Kitterman JA, Clements JA and Tooley WH. Ontogeny of tracheal f luid, pulmonary surfactant, and plasma corticoids in the fetal lamb. J Appl Physiol 39:1017-1021, 1975. Meschia G, Battaglia FC and Bruns PD. Theoretical and experimental study of transplacental diffusion. J Appl Physiol 22:1171-1178, 1967. Mihaly GW, Morgan DJ, Smallwood R and Hardy KJ. The developing liver: The steady-state disposition of propranolol in pregnant sheep. Hepatology 2:344-349, 1982. Mihaly GW, Jones DB, Morgan DJ, Ching MS, Webster LK, Smallwood RA and Hardy KJ. Placental transfer and renal elimination of cimetidine in maternal and fetal sheep. J Pharmacol Exp Ther 227:441-445, 1983. 228 Mihaly GW and Morgan DJ. Placental drug transfer: Effects of gestational age and species. Pharmac Ther 23:253-266, 1984. Mitani GM, Steinberg I, Lien EJ, Harrison EC and Elkayam U. The pharmacokinetics of antiarrhythmic agents in pregnancy and lactation. Clin Pharmacokinet 12:253-291, 1987. Mirkin BL and Singh H. Placental transfer of pharmacologically active molecules. In: Perinatal Pharmacology and Therapeutics. Mirkin BL (Ed). Academic Press, New York, NY, ppl-69, 1976. Morishima HO, Finster M, Pedersen H, Fukunaga A, Ronfeld RA, Vassallo HG and Covino BG. Pharmacokinetics of lidocaine in fetal and neonatal lambs and adult sheep. Anesthesiology 50:431-436, 1979. Morselli.PL. Clinical pharmacokinetics in neonates. Clin Pharmacokinet 1:81-98, 1976. Mucklow JC. The fate of drugs in pregnancy. Clin Obstet Gynaecol 13:161-175, 1986. Murad SHN, Tabsh KMA, Conklin KA, Shilyanski G, Ziadlourad F, Kapur PA and Flacke WE. Verapamil: Placental transfer and effects on maternal and fetal hemodyamics and atrioventricular conduction in the pregnant ewe. Anesthesiology 62:49-53, 1985. Natale R, Cleewlow F and Dawes GS. Measurement of fetal forelimb movements in the lamb in utero. Am J Obstet Gynecol 140:545-551, 1981. Nathanielsz PW. Monographs in Fetal Physiology 2. Animal Models in Fetal Medicine I. Elsevier/North-Hol1 and Biomedical Press, Amsterdam, 1980. Nathanielsz PW, Bailey A, Poore ER, Thorburn GD and Harding R. The relationship bewtween myometrial activity and sleep state and breathing in fetal sheep throughout the last third of gestation. Am J Obstet Gynecol 138:653-659, 1980. Nation RL. Meperidine binding in maternal and fetal plasma. Clin Pharmacol Ther 29:472-479, 1981. Nijhuis JG, Prechtl HFR, Martin CB Jr and Bots RSGM. Are there behavioural states in the human fetus? Early Hum Dev 6:177-195, 1982. Norbiato G, Bevilacqua M, Chebat E, Bertora P, Cavaiani P, Baruto C, Fumagalli S and Raggi U. Metoclopramide increases vasopressin secretion. J Clin Endocrinol Metab 63:747-750, 1986. Okumura K, Yoshida H and Hori R. Tissue distribution and metabolism of drugs. III. Accumulation of drugs by the isolated perfused rat lung. J Pharm Dyn 1:230-237, 1978. Olver RE and Strang LB. Ion fluxes across the pulmonary epithelium and the secretion of lung liquid in the foetal lamb. J Physiol 241:327-357, 1974. 229 Owens JA, Falconer J and Robinson JS. Effect of restriction of placental growth on umbilical and uterine blood flows. Am J Physiol 250:R427-R434, 1986. Pagnini G and Di Carlo R. Investigations on the interaction of 2-methoxy-5-chloro-procainamide (metoclopramide) with serum proteins. Drug Res 22:780-783, 1972. Patrick J , Richardson B, Hasen G, Clarke D, Wlodek M, Bousquet J and Brien J . Effects of maternal ethanol infusion on fetal cardiovascular and brain activity in lambs. Am J Obstet Gynecol 151:859-867, 1985. Paxton JW. Alphas-acid glycoprotein and binding of basic drugs. Methods Find Exp Clin Pharmacol 5:635-648, 1983. Peck CC, Beal SL, Sheiner LB and Nichols Al . The problem of choosing weights in nonlinear regression analysis of pharmacokinetic data. Drug Metab Rev 15:133-148, 1984. Peck CC, Beal SL, Sheiner LB and Nichols Al . Extended least squares nonlinear regression: A possible solution to the "Choice of Weights" problem in analysis of individual pharmacokinetic data. J Pharmacokinet Biopharm 12:545-558, 1984a. Pedersen PV. Curve fitting and modeling in pharmacokinetics and some practical experiences with NONLIN and an new program FUNFIT. J Pharmacokinet Biopharm 5:513-531, 1977. Pedersen H, Morishima HO, Finster M, Arthur GR and Covino BG. Pharmacokinetics of etidocaine in fetal and neonatal lambs and adult sheep. Anesth Analg 61:104-108, 1982. Pelkonen 0. Biotransformation of xenobiotics in the fetus. Pharmac Ther 10:261-281, 1980. Perrier D and Mayersohn M. Noncompartmental determination of the steady-state volume of distribution for any mode of administration. J Pharm Sci 71:372-373, 1982. ERRATA, ibid. . p 1427. Philipot RM, Anderson MW and Eling TE. Uptake, accumulation and metabolism of chemicals by the lung. In: Metabolic Functions of the Lung. Bakhle YS and Vane JR (Eds.). Marcel Dekker, New York NY, ppl23-171, 1977. Piafsky KM and Borga 0. Plasma protein binding of basic drugs, II: Importance of a i - ac id glycoprotein for interindividual variation. Clin Pharmacol Ther 22:545-548, 1977. Piafsky KM. Disease-induced changes in the plasma binding of basic drugs. Clin Pharmacokinet 5:246-262, 1980. Piercy WN, Day MA, Niems AH and Williams RL. Alteration of ovine fetal respiratory-like activity by diazepam, caffeine, and doxapram. Am J Obstet Gynecol 127:43-49, 1977. 230 Pinder RM, Brogden RN, Sawyer PR, Speight TM and Avery GS. Metoclopramide: A review of its pharmacological properties and clinical use. Drugs 12:81-131, 1976. Ponte CD and Nappi JM. Review of a new gastrointestinal drug -metoclopramide. Am J Hosp Pharm 38:829-833, 1981. Popovic J . High-pressure liquid chromatographic method for determination of metoclopramide in serum, urine, and saliva, with a pharmacokinetic study in patiens. Ther Drug Monit 6:77-82, 1984. Reynolds F. Transfer of drugs. In: Placental Transfer. Chamberlain GVP and Wilkinson AW (Eds.). University Park Press, Baltimore MA, pp 166-181, 1979. Reynolds GJ. Metoclopramide in young children. Br Med J 2:1713, 1978. Reynolds F. Distribution of drugs in amniotic f luid. In: Amniotic Fluid and Its Clinical Significance. Sandler M (Ed.). Marcel Dekker, New York NY, pp 261-275, 1981. Rigatto H, Blanco CE and Walker DW. The response to stimulation of hindlimb nerves in fetal sheep, in utero, during the different phases of electrocortical activity. J Dev Physiol 4:175-185, 1982. Riggs DS. The mathematical approach to physiological problems. Williams and Wilkins, Baltimore MD, p 210, 1963. Riggs KW. Metoclopramide: Fused si l ica capillary GLC-ECD; Placental transfer in humans and sheep. M. Sc. Thesis, The University of British Columbia, Vancouver, B.C., 1982. Riggs KW, Axelson JE, Rurak DW, Hasman DA, McErlane BA, Bylsma-Howel1 M, McMorland GH, Ongley R and Price JDE. Electron-capture determination of metoclopramide in biological fluids using fused si l ica capillary columns: Application to placental transport studies in sheep and humans. J Chromtogr 276:319-328, 1983. Riggs KW, Rurak DW, Yoo SD, Taylor SM, McErlane BA, McMorland GH and Axelson JE. Drug accumulation in fetal lamb tracheal f luid. Abstracts, 13th Annual Conference of The Society for The Study of Fetal Physiology and the 1st International Christie Conference on Perinatal Development of the Heart and Lung. Banff, Alta, Canada, July 1986. Riggs KW, Rurak DW, Yoo SD, McErlane BA, Taylor SM, McMorland GH and Axelson JE. Drug accumulation in lung fluid of the fetal lamb following maternal or fetal administration. Am J Obst Gynecol 157:1286-1291, 1987. Riggs KW, Axelson JE, Gruber NC, McErlane BA, McMorland GH and Rurak DW. Metoclopramide pharmacokinetics in pregnant and nonpregnant sheep. J Pharm Sci 77:373-379, 1988. Riggs KW, Yoo SD, Axelson JE, Taylor SM, McErlane BA and Rurak DW. Recirculation'of drugs from amniotic fluid to the fetus in pregnant sheep. Abstracts. 12th Annual Ontario-Quebec Perinatal Investigators Meeting. Kingston, Ontario, November, 1988a. 231 Robinson OPW. Metoclopramide - side effects and safety. Postgrad Med J 49:77-80, 1973. Roerig DL, Kotrly KJ, Vucins EJ, Ahlf SB, Dawson CA and Kampine JP. First pass uptake of fentanyl, meperidine and morphine in the human lung. Anesthesiology 67:466-472, 1987. Rosenfeld CR. Distribution of cardiac output in ovine pregnancy. Am J Physiol 232:H231-H235, 1977. Ross MG, Ervin MG, Leake RD, Oakes G, Hobel C and Fisher DA. Bulk flow of amniotic fluid water in response to maternal osmotic challenge. Am J Obstet Gynecol 147:697-701, 1983. Ross-Lee LM, Eadie MJ, Bochner F, Hooper WD and Tyrer JH. Electron-capture gas chromatographic assay for metoclopramide in plasma. J Chromatogr 183:175-184, 1980. Ross-Lee LM, Eadie MJ, Hooper WD and Bochner F. Single-dose pharmacokinetics of metoclopramide. Eur J Clin Pharmacol 20:465-471, 1981. Roth RA and Wiersma DA. Role of the lung in total body clearance of circulating drugs. Clin Pharmacokinet 4:355-367, 1979. Roth RA. Biochemistry, physiology and drug metabolism - Implications regarding the role of the lungs in drug disposition. Clin Physiol Biochem 3:66-79, 1985. Ruckebusch Y, Gaujoux M and Eghbali B. Placental transfer of central nervous system depressants in sheep. Eur J Pharmacol 37:193-196, 1976. Rudolph AM. Distribution and regulation of blood flow in the fetal and neonatal lamb. Cir Res 57:811-821, 1985. Rudolph AM and Heymann MA. Circulatory changes during growth in the fetal lamb. Cir Res 26:289-299, 1970. Runciman WB, Ilsley AH, Mather LE, Carapetis R and Rao MM. A sheep preparation for studying interactions between blood flow and drug disposition. I: Physiological profile. Br J Anaesth 56:1015-1027, 1984. Runciman WB, Mather LE, Ilsley AH, Carapetis R and McLean CF. A sheep preparation for studying the interactions between blood flow and drug disposition. II: Experimental applications. Br J Anaesth 56:1117-1129, 1984a. Ruppert F, Sulyok E, Sarkany I, Zambo K and Csaba IF. Effect of metoclopramide treatment on thyrotropin and prolactin levels in sick neonates. Biol Neonate 50:141-146, 1986. Sailer R, Hellenbrecht D, Hellstern A and Hess H. Improved benefit/risk ratio of higher-dose metoclopramide therapy during cisplatin-induced emesis. Eur J Clin Pharmacol 29:311-312, 1985. 232 Sailer R, Hellenbrecht D, Briemann L, Hellstern A, Hess H, Mitrou P, Hodgson M, Achtert G, Brockmann P and Hausleiter HJ. Metoclopramide kinetics at high-dose infusion rates for prevention of cisplatin-induced emesis. Clin Pharmacol Ther 37:43-47, 1985. Schellenberg, JC, Liggins GC, Manzai M, Kitterman JA and Lee CCH. Synergistic hormonal effects on lung maturation in fetal sheep. J Appl Physiol 65:94-100, 1988. Schulze-Delrieu K. Metoclopramide. New Eng J Med 305:28-33, 1981. Seeds AE. Current concepts of amniotic fluid dynamics. Am J Obstet Gynecol 138:575-586, 1980. Shaughnessy AF. Potential uses for metoclopramide. Drug Intel 1 Clin Pharm 19:723-728, 1985. Sheiner LB and Beal SL. Pharmacokinetic parameter estimates from several least squares procedures: Superiority of extended least squares. J Pharmacokinet Biopharm 13:185-201, 1985. Singh MS and Lean TH. The use of metoclopramide in hyperemesis gravidarum. Proceedings of the Obstetrical and Gynecological Society, Singapore, 1:43-49, 1970. Slordal L, Prytz PS, Aasebo U and Aarbakke J . A simple HPLC method for measuring metoclopramide in serum. Acta Pharmacol Et Toxicol 58:240-242, 1986. Smith MA and Salter FJ. Metoclopramide. Drug Intel 1 Clin Pharm 14:169-176, 1980. Solanki DR, Suresh M and Ethridge HC. The effects of intravenous cimetidine and metoclopramide on gastric volume and pH. Anesth Analg 63:599-602, 1984. Sommers DK, Meyer EC, van Wyk M and de Vi11iers LS. Aldosterone response to metoclopramide is mediated through the autonomic nervous system in man. Eur J Clin Pharmacol 33:609-612, 1988. Stacey TE, Weedon AP, Haworth C, Ward RHT and Boyd RDH. Fetomaternal transfer of glucose analogues by sheep placenta. Am J Physiol 234:E32-E37, 1978. Stock BH. Hepatic drug metabolism in pregnancy. Rev Drug Metab Drug Inter 5:53-81, 1984. Strang LB. Growth and development of the lung: Fetal and postnatal. Ann Rev Physiol 39:253-276, 1977. Szeto HH. Effects of narcotic drugs on fetal behavioral activity: Acute methadone exposure. Am J Obstet Gynecol 146:211-216, 1983. Szeto HH, Mann LI, Bhakthavathsalan A, Lui M and Inturrisi CE Meperidine pharmacokinetics in the maternal-fetal unit. J Pharmacol Exp Ther 206:448-459, 1978. 233 Szeto HH, Zervoudakis IA, Cederqvist LL and Inturrisi CE. Amniotic fluid transfer of meperidine from maternal plasma in early pregnancy. Obstet Gynecol 52:59-62, 1978a. Szeto HH, Kaiko RF, Clapp JF, Larrow RW, Mann LI and Inturrisi CE. Urinary excretion of meperidine by the fetal lamb. J Pharmacol Exp Ther 209:244-248, 1979. Szeto HH, Clapp JF, Larrow RW, Inturrisi CE and Mann LI. Renal tubular secretion of meperidine by the fetal lamb. J Pharmacol Exp Ther 213:346-349, 1980. Szeto HH, Umans JG and McFarland J . A comparison of morphine and methadone disposition in the maternal-fetal unit. Am J Obstet Gynecol 143:700-706, 1982. Szeto HH, Umans JG, Umans HR and McFarland JW. The relationship between maternal and fetal plasma protein binding of methadone in the ewe during the third trimester. Life Sci 30:1271-1279, 1982a. Szeto HH, Umans JG and Rubinow SI. The contribution of transplacental clearances and fetal clearance to drug disposition in the ovine maternal-fetal unit. Drug Metab Dispos 10:382-386, 1982b. Tarn YK. The pharmacokinetics of metoclopramide in normal rats and in rats with experimental renal and hepatic dysfunction. Ph. D. Thesis, The University of British Columbia, Vancouver, B.C., 1980. Tarn YK and Axelson JE. Sensitive electron-capture GLC determination of metoclopramide in biological fluids. J Pharm Sci 67:1073-1077, 1978. Tarn YK, Axelson JE and Ongley R. Modification of metoclopramide GLC assay: Application to human biological specimens. J Pharm Sci 68:1254-1256, 1979. Tarn YK, Axelson JE, McErlane B, Ongley R and Price JDE. Dose-dependent pharmacokinetics of metoclopramide in rat: An effect of hemoperfusion. J Pharmacol Exp Ther 217:764-769, 1981. Tarn YK, Axelson JE, McErlane B, Kapil RP, Riggs KW, Ongley R and Price JDE. The pharmacokinetics of metoclopramide in rats with experimental renal and hepatic dysfunction. J Pharmacol Exp Ther 219:141-146, 1981a. Taylor WB and Bateman DN. High dose metoclopramide - preliminary pharmacokinetic studies. Br J Clin Pharmac 16:341-342, 1983. Taylor WB, Proctor SJ and Bateman DN. Pharmacokinetics and efficacy of high-dose metoclopramide given by continuous infusion for the control of cytotoxic drug-induced vomiting. Br J Clin Pharmac 18:679-684, 1984. Taylor WB, Simpson JM and Bateman DN. High-dose metoclopramide by infusion: A double-blind study of plasma concentration-effect relationships in patients receiving cancer chemotherapy. Eur J Clin Pharmacol 33:161-165, 1987. 234 Teng L, Bruce RB and Dunning LK. Metoclopramide metabolism and determination by high-pressure liquid chromatography. J Pharm Sci 66:1615-1618, 1977. Tomoda S, Brace RA and Longo LD. Amniotic fluid volume and fetal swallowing rate in sheep. Am J Physiol 249:R133-R138, 1985. Tomoda S, Brace RA and Longo LD. Amniotic fluid volume regulation: Basal volumes and responses to fluid infusion or withdrawal in sheep. Am J Physiol 252:R380-R387, 1987. Tronick E. Editorial. A critique of neonatal neurologic and adaptive capacity score (NACS). Anesthesiology 56:338-339, 1982. Van Petten GR, Mathison HJ, Harris WH and Mears GJ. Chronic preparation of the pregnant ewe and fetus for pharmacological research: The placental transfer and fetal effects of bunitrolol. J Pharmacol Meth 1:45-60, 1978. Vella L, Francis D, Houlton P and Reynolds F. Comparison of the antiemetics metoclopramide and promethazine in labour. Brit Med J 290:1173-1175, 1985. Vree TB, Reekers-Ketting JJ , Fragen RJ and Arts THM. Placental transfer of midazolam and its metabolite 1-hydroxymethylmidazolam in the pregnant ewe. Anesth Analg 63:31-34, 1984. Waddell WJ and Marlowe GC. Disposition of drugs in the fetus. In: Perinatal Pharmacology and Therapeutics. Mirkin BL (Ed.) Academic Press, New York NY, pp 119-268, 1976. Waddell WJ and Marlowe C. Transfer of drugs across the placenta. Pharmac Ther 14:375-390, 1981. Wadhwa NK, Schroeder TJ, O'Flaherty E, Pesce AJ, Myre SA and First MR. The effect of oral metoclopramide on the absorption of cyclosporin. Transplantation 43:211-213, 1987. Wagner JG. A modern view of p h a r m a c o k i n e t i c s . J P h a r m a c o k i n e t Biopharm 1:363-401, 1973. Walker DW. Placental clearance of antipyrine and urea during hypoxia and adrenaline infusions in sheep. J Physiol 27_1:65P-66P, 1977. Wang LH, Rudolph AM and Benet LZ. Distribution and fate of acetaminophen conjugates in fetal lambs in utero. J Pharmacol Exp Ther 235:302-307, 1985. Wang LH, Rudolph AM and Benet LZ. Pharmacokinetic studies of the disposition of acetaminophen in the sheep maternal-placental-fetal unit. J Pharmacol Exp Ther 238:198-205, 1986. Ward RM, Singh S and Mirkin BL. Fetal clinical pharmacology. In: Drug Treatment. Avery GS (Ed). Adis Press, New York, NY, pp77-80, 1980. 235 Webb D, Buss DC, Fifield R, Bateman DN and Routledge PA. The plasma protein binding of metoclopramide in health and renal disease. Br J Clin Pharmac 21:334-336, 1986. Wilkening RB, Anderson S and Meschia G. Non-steady state placental transfer of highly diffusible molecules. J Dev Physiol 6:121-129, 1984. Wilkinson GR and Shand DG. A physiological approach to hepatic drug clearance. Clin Pharmac Ther 18:377-390, 1975. Wilson AGE, Law FCP, Eling TE and Anderson MW. Uptake, metabolism and efflux of methadone in "single pass" isolated perfused rabbit lung. J Pharmacol Exp Ther 199:360-367, 1976. Wlodek ME and Patrick JE. The effects of hypoxemia on fetal renal and cardiovascular function. Abstracts, 13th Annual Conference of The Society for The Study of Fetal Physiology and the 1st International Christie Conference on Perinatal Development of the Heart and Lung. Banff, Alta, Canada, July 1986. Wood M and Wood AJJ. Changes in plasma drug binding and a i - ac id glycoprotein in mother and newborn infant. Clin Pharmac Trier 29:522-526, 1981. Wright GP, Melikian AP and Crowe JT. Dose proportionality of metoclopramide. Clin Pharmacol Ther 35:284, 1984). Wright MR. Metoclopramide disposition in normal and uremic humans. M. Sc. Thesis, The University of British Columbia, Vancouver, B.C., 1987. Wright MR, Axelson JE, Rurak DW, McErlane B, McMorland GH, Ongley RC, Tarn YK and Price JDE. Linearity of metoclopramide kinetics at doses of 5-20 mg. Br J Clin Pharmac 26:469-473, 1988. Wright MR, Axelson JE, Rurak DW, McErlane B, McMorland GH, Ongley RC, Tarn YK and Price JDE. Effect of haemodialysis on metoclopramide kinetics in patients with severe renal failure. Br J Clin Pharmac 26:474-477, 1988a. Yamaguchi R, Hatakeyama Y and Yoshida T. A study of fetal utilization of substances injected in amniotic f luid. Tohoku J Exp Med 111:25-32, 1973. Yoo SD, Axelson JE, Taylor SM and Rurak DW. Placental transfer of diphenhydramine in chronically instrumented pregnant sheep. J Pharm Sci 75:685-687, 1986. Yoo SD, Rurak DW, Kwan E, Taylor SM and Axelson JE. Placental and nonplacental clearances of diphenhydramine in the chronically instrumented pregnant sheep. Abstract. The Japan-United States Congress of Pharmaceutical Sciences, Honolulu, Hawaii, December 1987. Yoo SD, Rurak DW, Riggs KW, Kwan E, Taylor SM and Axelson JE. Abstracts, Drug recirculation from amniotic fluid to the ewe and fetus in pregnant sheep. The 44th Annual Meeting of the Society of Obstetricians and Gynaecologists of Canada, Vancouver, B.C., June 1988. Yoshida H, Okumura K and Hori R. Subcellular distribution of basic drug accumulated in the isolated perfused lung. Pharm Res 4:50-53, 1987. Yuen GJ, Hansten PD and Collins J . Effect of metoclopramide on the absorption of an oral sustained-release quinidine product. Clin Pharm 6:722-725, 1987. Yurth DA. Placental transfer of local anesthetics. Clin Perinatol 9:13 29, 1982. 237 7. APPENDIX Gas Chromatography-Mass Spectrometry Gas chromatographic-mass spectrometric (GC-MS) analysis was performed using a Hewlett-Packard (HP) Model 5987A GC-MS system. An HP Model 2623A terminal and a Series 1000E HP computer were used for data processing. Electron impact (EI), positive-ion chemical ionization (CI) and negative-ion chemical ionization GC were used to study the mass spectra of standard samples of the heptafluorobutyryl (HFB) derivatives of MCP (HFB-MCP) and its di-deethylated metabolite (HFB-DE-MCP), and to subsequently identify these compounds and other possible metabolites in plasma and urine obtained from pregnant sheep. The column and GC operating conditions were essentially the same as those used for routine sample analysis by GC-ECD (Section 2.6.1) briefly: injection port temperature 260 °C; initial column temperature 205 °C increased at a rate of 4 °C/min to a temperature of 240 °C for 3.5 min; followed by a second increase at 15 °C/min to 255 °C; splitless injection mode with an inlet purge time of 30 sec following sample injection. For GC-MS, 3 /zL of sample were injected and helium was used as the carrier gas. The MS operating conditions were: interface temperature 250 °C; ion source temperature 240 °C; electron ionization energy 70 eV for EI, and 200 eV for CI; emission current 0.3 mA. Methane was used as the reagent gas for chemical ionization. Sample preparation - Except for urine samples, all standards and plasma samples were prepared using the extraction and derivative formation procedure described in Section 2.4 for routine GC-ECD analysis. The nitrogen dried extracts were reconstituted to 200 /zL instead of the usual 800 /zL for derivative formation with HFBA. Following removal of excess reagent this volume was further reduced to =20 /iL under a gentle stream of nitrogen in order to obtain sufficient concentrations for GC-MS analysis. Metoclopramide standard: 1.0 mL of a stock solution containing 3 /zg/mL of MCP equivalent to base (m.w. 299). Di-deethylated MCP (m.w. 243) standard: 1.5 mL of a stock solution containing 2.4 /zg/mL. Plasma (maternal): 1.0 mL (combined 0.5 mL aliquots) obtained at 15 and 20 min after a maternal 40 mg i.v. bolus dose of MCP. Urine (maternal) collected in a clean stainless steel tray: 1.5 mL of a sample collected =1^  hours following a maternal 40 mg i.v. bolus dose of MCP. The procedure used for the preparation of urine samples was the same as that described in Section 2.4 except that only a single extraction with IN NaOH was performed; urine samples were not back extracted with 5N HCL. RESULTS AND DISCUSSION Standard Samples Electron Impact GC-MS - The total ion chromatogram and mass spectrum obtained for the HFB derivative of the MCP standard (retention time 11.79 min) are presented in Fig. 1. The proposed fragmentation pattern is illustrated in Fig. 2 and is very similar to that previously reported by Tarn, 1980. The fragmentation pattern is also quite comparable to those obtained by Cowan et a7., 1976 and Teng et ah, 1977 using direct-probe GC-MS. The base peak is at m/z 86 and the other most abundant fragmentation ions, in order of decreasing intensity, at 99, 380, 184 and 423. The derivative of MCP cleaved at the amide bond (m/z 423) as well as the carbonyl-amide bond (m/z 99 and 380). The base peak (m/z 86) is the result of cleavage at the carbon-carbon bond beta to the amide nitrogen (Fig. 2). 238 The chromatogram and electron impact mass spectrum obtained for the HFB derivative of the di-deethylated MCP standard are shown in Fig. 3. The postulated fragmentation pattern is presented in Fig. 4 and is comparable to that reported by Cowan et a7 . , 1976. An ion with a m/z of 635 was observed suggesting the formation of a disubstituted HFB molecular ion, (M+-). The pattern of fragmentation is similar to that of MCP with rupture of the carbonyl-amide and amide bonds. Unlike MCP, however, the base peak occurred at m/z 380 with the other most abundant ions at 184, 169, 226, 409, 423 and 396, in order of decreasing intensity. Chemical Ionization - Both positive and negative chemical ionization were employed to detect the molecular ion. The spectra obtained for the HFB derivative of the MCP standard are illustrated in Fig. 5. A very intense m/z 496 ion corresponding to the molecular ion, (M+H)+, was obtained following positive chemical ionization (Fig. 5a). A molecular ion, (M-H)" m/z 494, of relatively low intensity was also obtained with negative ion chemical ionization (Fig. 5b). With negative CI, however, the most intense ion fragment occurred at m/z 475 and appears to represent the molecular ion less a F atom. The mass spectra obtained for the HFB derivative of the di-deethylated MCP standard are presented in Fig. 6. An intense molecular ion, (M+H)+ m/z 636, corresponding to a disubstituted HFB derivative was detected following positive CI (Fig. 4). A very low intensity (M-H)" ion, m/z 634, was obtained with negative CI, again confirming the formation of a disubstituted HFB derivative (Fig. 4). As previously seen with MCP, there was loss of a F atom during the negative CI process resulting in an intense fragment ion at m/z 615 (Fig. 6b). Plasma Analysis The total ion current chromatogram and electron impact mass spectra of a plasma extract obtained from a pregnant ewe receiving a 40 mg i.v. bolus dose of MCP are shown in Fig. 7. The most abundant ions for the peak with a retention time of 11.69 min were obtained at m/z 86 and 99 and correspond to the base peak and the next most intense ion fragment of the MCP standard (Figs. 1 and 2). Further analysis with positive CI revealed an intense molecular ion, (M+H)+, at m/z 496 representing the HFB derivative of MCP (Fig. 2). With negative CI, a molecular ion was not detected, but as observed with the MCP standard an intense ion fragment at m/z 475, corresponding to the molecular ion (M-H)" minus a fluorine atom, was present. Analysis of the peak with a retention time of 11.45 min resulted in the following ion fragments at m/z 380, 226 and 256. These three fragments have been previously reported by Tarn, 1980 for the HFB derivative of mono-deethylated MCP observed in rat urine following MCP dosing. The ions at m/z 226 and 256 have also been reported by Cowan et a7., 1976, following the analysis of authentic samples of mono-deethylated MCP by solid inlet mass spectrometry. The proposed fragmentation pattern for this compound, following EI mass spectrometry, is presented in Fig. 8. (Note: the fragmentation pattern presented in this figure was obtained following analysis of the urine peak with the same retention time as in plasma, and contains ion fragments additional to those seen in plasma). The sample was also subjected to positive and negative chemical ionization for further structural confirmation. As previously reported by Tarn, 1980, an intense molecular ion (M+H)+ was obtained at m/z 664 with positive CI suggesting the formation of a disubstituted derivative of the mono-deethylated MCP metabolite. A molecular ion (M-H)" was not observed with negative CI. An intense peak was obtained at m/z 643, however, which would appear to 239 correspond to a disubstituted HFB molecular ion (i.e. m/z 662) minus a F atom. This metabolite appears as a relatively small peak eluting just prior to MCP (baseline resolution is obtained) in our routine GC-ECD analysis of plasma. It is generally apparent by =15-20 min after maternal dosing but varies in size and occurrence from animal to animal. It is also occasionally observed, although to a much smaller extent, in fetal plasma. While it is possible that this metabolite may be formed as a result of fetal hepatic metabolism, it is much more likely that its presence in fetal plasma is the result of placental transfer from the ewe. During negative CI, a relatively intense ion at m/z 615 was obtained for the peak with a retention time of 10.35 min (Fig. 7). This corresponds to the molecular ion ((M-H)" m/z 634) for the disubstituted HFB derivative of the di-deethylated MCP standard and suggests the presence of this metabolite in maternal sheep plasma. Extraction of larger volumes of plasma are required, however, in order to obtain the higher concentrations needed to verify this observation. Urine Analysis Urine was collected from two ewes following maternal 40 mg i.v. bolus injections of MCP. In one animal the sample was collected at =7 h following dosing and in the other at »1J; h. Aliquots (100 ill extraction) of both were analyzed by GC-ECD. Both chromatograms contained the same major peaks, however, those in the sample collected at =1J; were much larger. On this basis only the latter sample was used for GC-MS analysis. The total ion chromatogram obtained from EI mass spectrometry for the extract of maternal urine is presented in Fig. 9. Analysis of the peak with a retention time of 11.77 min resulted in ion fragments at m/z 86, 380, 99, 423 and 184, in order of decreasing intensity, and correspond to the base peak (m/z 86) and other most abundant ions obtained for the HFB derivative of the MCP standard (Figs. 1 and 2). An intense molecular ion, (M+H)+, was obtained at m/z 496 with positive CI, confirming the compound to be the HFB derivative of metoclopramide (Fig. 2). A molecular ion was not detected following negative CI, but an intense ion was present at m/z 475 which would correspond to the (M-H)- ion minus a fluorine atom. The electron impact mass spectra and proposed fragmentation pattern for the peak with a retention time of 11.53 min are presented in Figs. 9b and 8, respectively. The base peak was observed at m/z 380 and the other most abundant ions occurred at m/z 226, 169, 256, 184, 396 and 423 in order of decreasing intensity. This pattern of fragmentation is very similar to that reported by Tarn, 1980, for the mono-deethylated metabolite of MCP in urine obtained from MCP dosed rats. A molecular ion (M+-) was obtained at m/z 663 suggesting the formation of a disubstituted derivative following reaction with HFBA (Fig. 8). An intense molecular ion (M+H)+ was detected at m/z 664 following positive CI, and confirms the formation of a disubstituted HFB derivative of mono-deethylated metoclopramide. With negative CI, an intense ion fragment at m/z 643 was detected which would correspond to the molecular ion (M-H)", m/z 662, minus a F atom. The most prominent fragment ions obtained for the peak with a retention time of 10.09 min (Fig. 9a), following EI were at m/z 380, 169, 184, 226, 423 and 409, in order of decreasing intensity. This pattern is the same as that obtained for the disubstituted HFB derivative of the di-deethylated MCP standard (Fig. 4). Confirmation of the disubstituted derivative was obtained using chemical ionization. With positive CI, an intense peak corresponding to the (M+H)+ molecular ion (Fig. 4) was detected at m/z 636 . A low intensity molecular ion ((M-H)-, m/z 634) was obtained following negative ion CI and also confirms the formation of a 240 disubstituted HFB derivative of di-deethylated MCP. The most intense ion fragment for negative CI occurred at m/z 615 as the result of loss of a F atom fragment from the molecular ion. The electron impact spectra for the peak with a retention time of 6.84 min (Fig. 9a) showed ion fragments similar to those obtained for the HFB derivatives of the di-deethylated MCP standard (Fig. 4) and the mono-deethyl ated metabolite observed in plasma and urine extracts (Fig. 8). The base peak occurred at m/z 380 with the next most intense ion fragments at m/z 254, 226 and 169. With chemical ionization, intense peaks were obtained at m/z 582 and 581 following positive and negative CI, respectively. We were not, however, able to assign a structure to this compound on the basis of the EI spectra and the "apparent" molecular ions obtained following chemical ionization. While this may represent another metabolite in sheep, further analysis is required to determine its structural identity. In summary, the mono-deethylated metabolite of MCP has been identified in plasma and urine obtained from a pregnant ewe and on the basis of qualitative chromatography appears to represent a major metabolite in this species. This compound has been reported to be the major metabolite in urine obtained from rats (Bakke and Segura, 1976; Teng et al., 1977; Tarn, 1980) and dogs (Bakke and Segura, 1976; Teng et al., 1977) and a minor metabolite in rabbits (Arita et al., 1970). The di-deethylated metabolite of MCP was not detected in these species or in man (Teng et al., 1977), however, it appears to be an important metabolic product in sheep. 241 • f i l e ' U R f c 18 .0-788.& - »co 3 U Q dried down TIC ie.eeee-1 11.79 14808H 12000&H i 8 e « « 8 H ;.0006-l 40086H •'0880H 9 .92 Jul 1 " 1 ' 1 4.0 5.0 6.8 7.0 8.0 9.8 10.0 11.0 12.8 13 .814 .815 .816 .017 .8 b. F i l e > U R G Bpk fib 184488 11080&J 188888: 90880-88888-78888^ : £.0000^ 50868: 49880: 38888-28088^ 10088-n»cp> 3ug dried down 38 36 99 / 148 1852ie 302 \ T 58 100 158 200 250 388 • inj 3ul ow<« 423 493 337 V 483 y v " 1 • " T Scan 164 11.79 » i 1 r l 6 f 188 98 j-80 78 68 :48 -38 •29 •10 358 480 458 Fig. 1. a) Total ion current mass chromatogram and b) electron impact mass spectrum of the HFB-derivative of MCP. 242 m/z 423 m/z 380 ci H 2N _<fjW OCH, CH--CH-N "1+ / 2 V -C 2 H 5 C H " 1 + C 2 H 5 m/z 184 m/z 99 m/z 86 Fig. 2. The postulated fragmentation pattern of the HFB-derivative of MCP based on the most prominent fragment ions obtained with electron impact GC-MS. 243 F i l e >UR17 80008-1 7 0 0 0 H 6 e e e e - 4 586884 48808H 38808H 20888H 18088H . 8-8 .8 A I » U . detlCP 1.5 d r i e d i n j 3ul TIC 18.07 5 .85 K 4 . 6 2 .86 8.73 " 3 J\ 7 .<4** / \ 4 A 6 . 0 ' I ' ' 1 ' I ' 7.0 S . 0 -i—j—i— i l U * 7 12.22 =---^j!r-11.8 12.0 13.0 F i l e JUR12 Epk fib 2S312 23888H 300no'ul deHCP 3ul 24888H 28888H 16868-1 j 12888-1 3088H 4080H 63 / 148 135 218 i s y 255 3 1 f c 3B8 i 489 4 4 6 ^ 188 ? » « " ~ * +88 588 Scan 129 18.12 • i r l 6 188 H98 88 •78 rr68 J-58 -38 •28 -18 635 t~rT-,-f 600 Fig. 3. a) Total ion current mass chromatogram and b) electron impact mass spectrum of the HFB-derivative of di-deethylated MCP. 244 CI C 3 r 7 - C O - N H - ^ ^ - C O - N H - C H 2 - C H 2 - N H - C O - C 3 F 7 OCH, M + l m/z 635 Cl 1 + ci C3F7-C0-NH ^-C0-NH-CH 2»CH 2 CjFy-CO-NH-^^^-CO-NH-CHj 0CH3 m/z 423 m/z 409 Fig. 4. The postulated fragmentation pattern of the HFB-derivative of di-deethylated MCP based on the most prominent fragment ions obtained with electron impact GC-MS. a. F i l e >UR21 Bpk Hb 42408 44888^ 48880: -23800-; 24800: 28088^ 16088-^  1208* 3006-4800-MCP 3 dried 143 214 253 179 \ . J J / _ \ X N 0 'i i M p-T-r-rf-^-i-f-i-r-T-rr'-ry-T ( I r I • I I i11 I I ' • ^ I ' 158 268 250 308 29? in.: 3ul Scan 179 11.77 496 345 379 381 - • ' T r r r 358 4^88 423 / 44; 524 I ' ' • • I 1 ' • >-p - m T* ' 458 508 1,1 188 (-98 88 •76 68 f tee r48 38 i-28 18 b. Ti le >UR34 Bpk fib 463555 486888-448080-4 88008-368880-320888-2S0000-240888--200808-168868-1 28888-80808-48080-1 178 9-f 1 286^ fICP o Li y dried inj 3cl 385 333 SS ' 8 273 - \ 388 439 482 "1—•—I" 1 1 .79 mi 47! 460 I 240 288 328 368 r r 1—r -1—•• 408 440 L L 2 496 -VIB I1! 80 98 88 70 P 6 8 68 j-4& 30 20 18 -,—r— r 480 S2e Fig. 5. Chemical ionization mass spectra of the HFB-derivative of MCP. a) positive-ion. b) negative-ion a. F i l e >UR22 Bpk Fib 5483 6088-5588-5800-4588-4888-3588-3888-2588-2888-1586-1888-580- 354 y r*r'1 _r' 368 388 dertCP 1.5 d r i e d 458 \ 400 448 488 i - i - p in i 3ul Scan 148 10.86 miri& 636 ,16 -j-^T-i-r-r-r 568 r (-168 90 -88 :70 60 t -48 -38 664 f28 18 688 648 b. F i l e >WR3S Bpk fib 347458 368800-1 320000-288880-240080-288808-166688^ 120800-88888^ 48880-178 219 deMCP 1.Sua d r i e d in j 3ul >88 258 31b 358 361 482 466 497 529 561 Scan 128 18.12 mry{b 680j 388 400 450 L 108 j^ 98 t-ee 78 -60 ^8 -48 38 28 10 588 558 608 Fig. 6. Chemical ionization mass spectra of the HFB-derivative of di -deethylated MCP. a) positive-ion. b) negative-ion f i l e >i»IRl4 18.8-? 4488-4800-3688-3288 : 2S88-2488^ 2886-08-8 a«u. p l a s a a 15+20 d r i e d (TIP H .69 ir.j 3ul 1690-1200^ 808^ 408-0 1.4!: 9.30 J 8.643-97 A 18.35 9.92 i . ,!\ 92 | )I2.88 j * - * - 1 6 ' - i - c , e • • i • • • • i • • • • i •' • • i 9.8 10.0 1' * .« ' ' I ' ' • • I 13.8 b. F i l e >WR14 Bpk fib 2696 2888-Dlasisa 15+28 d r i e d i n j 3u] 2408-2880-1688-1288-880-488-86 28 53 aJ. I il M I ill " —i—i—i—f—i—i—i 58 1 3 5 168 ; i 28 J Sc an 11.69 n 162 -hit [-188 :90 • :S0 L70 • L50 -48 L38 446 \ L20 -18 108 > • II -r-IL0 150 208 258 388 358 480 Fig. 7. a) Total ion current mass chromatogram for a plasma extract obtained from a pregnant ewe receiving a 40 mg i.v. bolus of MCP. b) electron impact mass spectrum of the peak with a retention time of 11 69 min corresponding to the HFB-derivative of MCP. 248 \ — / C0-C3F7 OCH3 M+* m/z 663 m/z 466 m / z 4 2 3 m /z 396 m/z 380 m/z 256 Fig. 8. The postulated fragmentation pattern of the HFB-derivative of the mono-deethylated metabolite of MCP based on the most prominent fragment ions obtained with electron impact GC-MS. 249 a. r i i e >WR1J -i ! i 51068-: I H 0 0 0 6 ~ 98886-; i i 70000-^ -60868-5 fc.0008-4 0080-38800-1 10. a- /0«, 0 =.»u. ; . - i 1.25 J . b mi IC i n : 3L ;1 » . 36 28888-1 0-J-,-gg 18 .691 i . ti4f 16.46 . 8 8.8 9.0 16.0 11.8 12.8 13.8 14.8 15.8 16.8 17.6 b. T i l e :WR11 Bpk fib 36192 \ 36088-1 1 1 22088-j » 23886-24088-26888-16980-1 -t _ i j 1 2880-8868-— 4000^ r • "' in.i 3u3 Scan 157 11.53 mi 1-98 rb0 t t?8 169 1 4 0 I I :16 ^^-i-r-T-J 1 1 1 T-y-|-i----vT"»-r--r-i-pr-i----r-TT-'--r-r ; ' • » ' T1"^ r~1' " | " " 466 1 8 8 I'60 3 8 0 488 1 1 - 1 " I — T"T" 5 8 8 r " r T r ' ' ' I 6 8 8 i-48 t: f-38 U r L Fig. 9. a) Total ion current mass chromatogram for a urine extract obtained from a pregnant ewe receiving a 40 mg i.v. bolus of MCP. b) electron impact mass spectrum of the peak with a retention time of 11.53 min corresponding to the HFB-derivative of the mono-deethylated metabolite of MCP. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0098178/manifest

Comment

Related Items