Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Developing a limited sampling strategy for cyclosporine area under the curve monitering in lung transplant… Dumont, Randall John 2000

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

Item Metadata

Download

Media
831-ubc_2000-0386.pdf [ 6.07MB ]
Metadata
JSON: 831-1.0099550.json
JSON-LD: 831-1.0099550-ld.json
RDF/XML (Pretty): 831-1.0099550-rdf.xml
RDF/JSON: 831-1.0099550-rdf.json
Turtle: 831-1.0099550-turtle.txt
N-Triples: 831-1.0099550-rdf-ntriples.txt
Original Record: 831-1.0099550-source.json
Full Text
831-1.0099550-fulltext.txt
Citation
831-1.0099550.ris

Full Text

DEVELOPING A LIMITED SAMPLING STRATEGY FOR CYCLOSPORINS A R E A U N D E R THE C U R V E MONITORING IN L U N G T R A N S P L A N T PATIENTS By R A N D A L L JOHN D U M O N T B.Sc.(Pharm.), The University of British Columbia, 1998 R.Ph., The College of Pharmacists of British Columbia, 1998 A THESIS SUBMITTED IN PARTIAL F U L F I L L M E N T OF THE REQUIREMENTS FOR THE DEGREE OF M A S T E R OF SCIENCE In THE F A C U L T Y OF G R A D U A T E STUDIES (Faculty of Pharmaceutical Sciences) (Division of Clinical Pharmacy) We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH C O L U M B I A May 1999 © Randall John Dumont, 2000 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. DepartirreTrt of Pkarw/ar-fU-f-i' ra \ ^ C i e*c -g $ The University of British Columbia Vancouver, Canada Date _ J ovwe DE-6 (2/88) 11 A B S T R A C T This study developed a limited sampling strategy (LSS) to provide an estimate of cyclosporine (Neoral®) area-under-the-curve (AUC) in lung transplant recipients, a population for which a cyclosporine LSS has not yet been delineated. The predictive performance of the LSS, and other published ..LSS in other transplant types, was evaluated. Finally, the pharmacokinetic parameters of the lung transplant patients were calculated. Fourteen stable lung transplant patients (n = 7 male; n = 7 female) were entered into the study. Upon administration of a steady-state morning cyclosporine dose, blood samples were collected at 0, 1, 2, 3, 4, 5, 6, 8, 9, 10, and 12 hours post-dose in 12 patients (and up to 8 hours post-dose in 2 patients on a q8h regimen). Blood samples were analyzed by monoclonal fluorescence polarization immunoassay. A U C was calculated by the linear trapezoidal method, and the LSS was calculated using multiple regression analysis. Predictive performance was evaluated using methods proposed by Sheiner and Beal. Pharmacokinetic analysis was performed using WinNonlin® computer software. Patient characteristics (mean + SD) are as follows: age: 48 ± 12 years; weight: 69 ± 17 kg; transplant type: 6 double lung, 8 single lung; total daily cyclosporine dose: 4.3 ± 1.7 mg/kg; time post-transplant: 5.1 + 3.4 years. Eight patients were used to determine the LSS. Analysis of all available concentration-time data revealed the following equation: A U C = 17.24xC6 -58.96xC8 + 23.39xC9 + 52.29xC12 - 796.07, r 2 = 0.999. In order to provide a clinically feasible LSS, the remainder of the analysis was restricted to the data collected in the first 3 hours post-dose. One 4-point, four 3-point, six 2-point, and four 1-point equations were determined. On the basis of the number of samples required, the coefficient of determination, comparison of predictive performance, and the percent prediction error in A U C estimation (%pe), we selected I l l the following equation for analysis of predictive performance: A U C = 1.46xCl + 5.36xC3 + 274.49; r 2 = 0.975; %pe (range) = -4.47 - 8.47%. For this LSS, mean prediction error (ME, bias) was 195 ngxhr/mL, and mean absolute error (MAE, precision) was 299 ngxhr/mL. There was no significant difference in predictive performance between the LSS for lung transplant patients and other published LSS in other transplant types, with 5 exceptions. This 2-concentration LSS for lung transplant patients was significantly more biased than a 3-concentration LSS developed for renal transplant patients, and was significantly less biased and significantly more precise than 2 other LSS that were developed for renal transplant patients. The best clinically feasible LSS for cyclosporine A U C estimation requires 2 concentrations drawn at 1 and 3 hours post-dose. iv TABLE OF CONTENTS A B S T R A C T i T A B L E OF CONTENTS iv LIST OF T A B L E S viii LIST OF FIGURES ix LIST OF ABBREVIATIONS x A C K N O W L E D G E M E N T S xii DEDICATION xii i CHAPTER ONE: Introduction 1 Transplantation 2 1.1 The history of transplantation 2 1.2 Brief overview of rejection 4 1.3 Primary immunosuppression 6 1.3.1 Corticosteroids 8 1.3.2 Azathioprine 10 1.3.3 Cyclophosphamide 10 1.3.4 Mizoribine 11 1.3.5 Tacrolimus 12 1.3.6 Gusperimus 13 1.3.7 Mycophenolate mofetil 13 1.3.8 Sirolimus 14 1.3.9 Brequinar 15 1.3.10 Leflunomide 16 Lung transplantation 16 1.1 The history of lung transplantation 16 1.2 Indications for lung transplantation 17 1.3 Recipient criteria 19 1.4 Donor criteria 21 1.5 Current challenges in lung transplantation 23 V 1.6 Recent advances in lung transplantation 24 1.7 The future of lung transplantation 25 Cyclosporine 25 1.1 The history of cyclosporine 25 1.2 Cyclosporine physical-chemical properties 27 1.3 Pharmacology and therapeutic use of cyclosporine 29 1.4 Considerations for cyclosporine use in lung transplantation 31 1.5 Formulations of cyclosporine used in lung transplantation 33 1.6 Pharmacokinetics of cyclosporine 35 1.6.1 Absorption 35 1.6.2 Distribution 37 1.6.3 Metabolism 38 1.6.4 Excretion 40 1.7 Cyclosporine toxicity 41 1.7.1 Nephrotoxicity 41 1.7.2 Cardiovascular toxicity 43 1.7.3 Neurotoxicity 44 1.7.4 Dermatologic toxicity 44 1.7.5 Hepatotoxicity 45 1.7.6 Gastrointestinal toxicity 45 1.7.7 Hematologic toxicity 45 1.7.8 Lymphatic and related toxicity 46 1.7.9 Hypersensitivity and other effects 47 Developing a limited sampling strategy for cyclosporine area under the curve monitoring in lung transplant patients 49 1.1 Introduction 49 1.2 Background studies 51 1.3 Significance of the research project 58 vi Research hypothesis 59 Objectives : 59 Rationale 59 CHAPTER TWO: Materials and Methods 62 2.1 Experimental design 63 2.2 Clinical research subjects 63 2.3 Clinical research subject inclusion criteria 63 2.4 Clinical research subject exclusion criteria 63 2.5 Clinical research study protocol 64 2.6 Sample analysis 66 2.6.1 Reagents and standards 66 2.6.1.1 Reagent solutions 66 2.6.1.2 Calibrator solutions 67 2.6.2 Sample preparation 67 2.6.3 Assay controls 68 2.6.4 Calibration 68 2.6.5 Sample analysis procedure ..68 2.6.6 Linearity 69 2.6.7 Precision 70 2.6.8 Standard reference intervals for cyclosporine trough levels utilized at Vancouver Hospital and Health Sciences Centre 70 2.6.9 Specificity of the fluorescence polarization immunoassay 71 2.7 Pharmacokinetic analysis 72 2.8 Sample size 74 2.9 Phase I of the clinical study 74 2.10 Phase II of the clinical study 75 CHAPTER THREE: Results 77 3.1 Lung transplant patient characteristics 78 3.2 Cyclosporine concentration-time data 78 3.3 Steady-state pharmacokinetic parameters for cyclosporine (Neoral®) 80 3.4 Phase I: multiple regression analysis 82 vii 3.5 Phase II: comparisons of predictive performance 87 CHAPTER 4: Discussion 91 4.1 Steady-state pharmacokinetic parameters for cyclosporine (Neoral®) 93 4.2 Multiple regression analysis 98 4.3 Comparisons of predictive performance 101 4.4 Conclusion 103 REFERENCES 1 r t o Vlll L I S T O F T A B L E S Table 1 Table 2 Table 3 Table 4 Table 5 Table 6 Table 7 Table 8 Table 9 Table 10 Table 11 Table 12 Table 13 Table 14 Table 15 The development of cyclosporine 27 The cyclosporine dosing protocol at Vancouver Hospital and Health Sciences Centre 29 Clinical studies evaluating area under the curve monitoring of cyclosporine A and outcome 53 Summary of clinical studies that have developed a limited sampling strategy for a transplant population 55 Expected precision of the monoclonal fluorescence polarization immunoassay used for sample analysis 68 Standard reference intervals for target trough cyclosporine concentrations at Vancouver Hospital and Health Sciences Centre 69 Cross-reactivity of the monoclonal fluorescence polarization immunoassay used at Vancouver Hospital and Health Sciences Centre 70 Patient characteristics for 14 stable, adult lung transplant patients 77 Concentration-time data for 14 stable, adult lung transplant patients 78 Mean concentration-time values for 14 stable, adult lung transplant patients 79 Individual steady-state pharmacokinetic parameters for cyclosporine (Neoral®) in 14 stable, adult lung transplant patients 80 Mean steady-state pharmacokinetic parameters for cyclosporine (Neoral®) in 14 stable, adult lung transplant patients 81 The limited sampling strategies, and their respective predictive performance, derived from concentration-time data from 8 stable, adult lung transplant patients 85 Comparisons of predictive performance between the recommended limited sampling strategy for lung transplant patients and the remaining limited sampling strategies for lung transplant patients resulting from Phase I multiple regression analysis 86 Comparisons of predictive performance between the recommended limited sampling strategy for lung transplant patients and other previously published limited sampling strategies developed for non-lung transplant patient population..88 ix LIST OF FIGURES Figure 1 The chemical structure of cyclosporine A (C62H111N11O12, molecular weight 1202.6. Cyclosporine A is a large, neutral, hydrophobic cyclic peptide composed of 11 amino acid residues 30 Figure 2 The mechanism of action of cyclosporine A 33 Figure 3 The mean concentration versus time curve for 14 stable, adult lung transplant patients 83 Figure 4 The log concentration versus time curves for 14 stable, adult lung transplant patients 84 LIST OF A B B R E V I A T I O N S A U C area under the concentration versus time curve AUCO-T area under the concentration versus time curve for one dosing interval C concentration CI confidence interval Cmax maximum concentration observed during the dosing interval CSA cyclosporine A C V coefficient of variation EMIT enzyme multiplied immunotechnique F bioavailability FPIA fluorescence polarization immunoassay HPLC high performance liquid chromatography LSS limited sampling strategy M A E mean absolute error, or precision M D A E mean difference in absolute error M D P E mean difference in prediction error M E mean prediction error, or bias M R T mean residence time M RIA monoclonal radioimmunoassay N E O Neoral® NT nephrotoxicity pe prediction error %pe percent prediction error P RIA polyclonal radioimmunoassay PO oral RIA radioimmunoassay RJ rejection SD standard deviation SIM Sandimmune® tj /2 terminal elimination half-life Tmax time of the maximum concentration observed during the dosing interval T L M trough level monitoring V d volume of distribution V H H S C Vancouver Hospital and Health Sciences Centre WB whole blood terminal elimination rate constant xii A C K N O W L E D G E M E N T S I would like to thank my supervisor, Dr. Mary Ensom, for providing me the opportunity to further my studies. I would especially like to thank her for her support and help, and for always being there when I needed her help or advice. I would also like to thank the members of my research committee for their time, support, and feedback: Dr. John Sinclair (Chair), Dr. Thomas Chang (External), Dr. Bruce Carleton, Dr. Wayne Riggs, and Dr. Kishor Wasan. Thanks also for pushing me to be a better, more confident presenter. Thanks also to Dr. Robert Levy and Dr. Nilufar Partovi for their help in my thesis project. I especially appreciate all of times that Nilu sent me answers to my many questions. Thanks to Mr. Julius Chala for his expertise at collecting samples, and the many hours of interesting conversation. He showed me that life definitely does not end at 40. An especially large amount of thanks goes to my girl, Allison. You were with me throughout my journey through grad. school, and always were there for me. Those who know me know that it is not easy putting up with all of my colors. To the various summer students and Pharm.D. students that have worked in the lab, I extend my thanks for the excellent conversation and intellectual discussion. Finally, I would like to thank U B C . My time there felt like a marathon, but it was worth it. I gave six years of my life, but in return I made good friends, received two degrees, a pharmacist's license, two loyal dogs, and a wife. Now I have the opportunity to invest another four years for a third degree. X I I I D E D I C A T I O N This work is dedicated to those who were always there throughout the years to help me get where I've gotten: my family. To my parents, John and Joanne, who gave instilled in me the drive to succeed, as well as the large amount of support (read: dollars) required to get me there. To my brothers, Aaron and Carey, with whom I've shared many battles. To my love Allison, who is the sun in my sky, and my biggest supporter. And finally to my two boys, Bishop and Merlin. I will always strive to be the kind of person that both of you think I am. CHAPTER 1 INTRODUCTION A Chinese myth, circa 300 B.C.: One day, two men, Lu and Chao, called on the surgeon, Pien Ch 'iao. He g a v e them a toxic drink, and they were unconscious for three days. Pien Ch 'iao operated and opened their stomachs and explored the heart; after removing and interchanging their organs, he g a v e a wonderful drug and the two men went home recovered. (From Bergan 1997) 2 Transplantation 1.1 The History Of Transplantation Transplantation, which, simply defined, is the transfer of tissue from one person or animal to another, or from one part of the body to another, began many centuries ago. Cave paintings are the earliest depiction of transplantation (Bergan 1997). Having surpassed many historical and ethical developments during its evolution, modern organ transplantation now offers recipients improved quality of life, often prolonged life, and for transplant physicians, a viable therapeutic option for the treatment of end-stage organ disease. The field of transplantation is many-faceted, and it is evident in the literature that while individual institutions often follow consensus guidelines, each institution has its own preferences. For example, assay methods, immunosuppressive protocols, and target immunosuppressive concentrations are not the same at all institutions. Transplantation can be described by the genetic relationship between the donor and recipient, the site of grafting, and the identity of the tissue being transplanted (Bergan 1997). There are 4 possible types of genetic relationships between donor and recipient: 1) autograft, which is where the donor and the recipient are the same individual; 2) isograft, also called syngeneic graft, which is when the donor and recipient are identically or nearly identically related (i.e., monozygotic twins or in highly inbred animals); 3) allograft, also referred to as homograft, which is when the donor and recipient are genetically unrelated individuals of the same species; and 4) xenograft, also referred to as heterograft, which is when the donor and recipient are individuals of different species (Bergan 1997). When describing transplantation according to the site of grafting, orthotopic is when the donor tissue is surrounded by native tissue of the same type or it is transplanted in the 3 anatomically correct position. A l l other sites of grafting are heterotopic, which is when the tissue has been relocated in the recipient (Bergan 1997). Currently, tissues, whole organs, or partial organs can be transplanted. Bone, heart valves, cartilage, veins, corneas, heart, liver, lung, pancreas, and sections of the gastrointestinal tract comprise the tissues and organs that can be transplanted (Bergan 1997). The oldest record of transplant-related surgery comes from skulls dated back to the Bronze Age. Trephination, an archaic surgical technique used to relieve intracranial pressure, involved the removal of a circular disc of bone from the calvarium. The disc of bone was later replaced as an orthotopic autograft (Bergan 1997). Myths and legends from history provide written accounts of other examples of transplantation. While likely fictitious, the myth of Saints Cosmas and Damian from 6 t h century Rome is often mentioned. The legend describes a man with cancer in his leg. He spends the night in the church of the two saints. During the night, the saints examine his leg. One removes the diseased limb while the other goes to the local cemetery, exhumes the body of a recently deceased Ethiopian man, amputates and gathers the replacement limb, and returns to the church. The limb is attached, and the man awakens the next morning fully cured (Bergan 1997). A painting of the legend shows that the wrong leg was grafted to the man (Muller-Ruchholtz 1999). There are other examples of historical transplantation from records originating in different countries. Nasal reconstruction, involving grafting of skin from the cheek of a patient, was performed in India around 700 B.C. (Bergan 1997). Nasal reconstruction was performed in 15 t h century Italy as well. This technique involved a skin autograft from the arm of the patient. The Italian Method, as it is now referred to, is still used today (Bergan 1997). 4 By the 19 t h century, grafts of skin, tendons, nerves, cartilage, adipose tissue, corneas, adrenal and thyroid glands, ovaries, partial digestive and urinary tracts, and muscle were documented, although the grafts were more successful in animals than in humans (Bergan 1997). By 1880, stable corneal transplants in humans were documented. Once successful surgical techniques were established, the era of modern transplantation began. The first renal transplant was performed in 1936 in Russia by Dr. Voronoy (Bergan 1997). Unfortunately, the recipient died 2 days later. In 1947, a successful, but transient renal transplant was performed in the United States. The kidney of a dying patient was transplanted into a pregnant woman with a uterine infection. The transplanted kidney helped the woman recover from her infection, and was removed after a few days before rejection could occur. In 1950, another patient received a kidney from another patient who was not a relative but was the same age and blood type. The graft was lost after 11 months (Bergan 1997). During World War II, it was discovered that skin grafts performed for burn patients were successful only when the donor was an identical twin. This information was applied to renal transplantation when the first renal transplant in which the donor and recipient were identical twins was performed in Boston in 1954 (Bergan 1997, Muller-Ruchholtz 1999). There was no rejection, but the patient died of cardiac complications 8 years after receiving the transplant. Another renal transplant of this type was performed in 1956, and the recipient lived into the 1990's (Bergan 1997). 1.2 Brief Overview Of Rejection The cascade of events collectively known as rejection is a biphasic process involving host ("self) recognition of foreign antigens ("non-self) followed by a cascade of biochemical events 5 ultimately resulting in damage to the transplanted tissue or organ. Host recognition depends on cells that are capable of recognizing antigens expressed on the transplanted organ as foreign. These antigen presenting cells, which include macrophages, dendritic reticulum cells, endothelial cells, and some parenchymal cells, are capable of triggering a specific and coordinated host immune response that, i f unmodified, results in damage to the graft, and i f severe and prolonged, graft loss (Philip and Gerson 1998). The major components of the immune system involved in the immune response are T cells and macrophages, although all components are involved to some degree (Philip and Gerson 1998). T cell sensitization involves the presentation of antigens bound to human lymphocyte antigen (commonly referred to as H L A ) cell surface molecules by antigen presenting cells (Philip and Gerson 1998). Once sensitized, T cells become activated and undergo a series of changes resulting in proliferation, activation of various effector functions, and immunologic memory (Philip and Gerson 1998). Helper T cells are stimulated first. These cells secrete lymphokines that cause the proliferation of cytotoxic T cells and B cells, which secrete antibodies. More than 70 molecules are involved in the biochemical cascade resulting in activation and proliferation of the cells of the immune system (Philip and Gerson 1998). The most important of these is interleukin-2, which has regulatory functions over the whole process. Because of this, interleukin-2 is often the target of attempts to attenuate the immune response (Philip and Gerson 1998). Graft rejection is classified as hyperacute, acute, or chronic. Hyperacute rejection is caused by pre-existing antibodies that rapidly bind to vascular endothelium in the donor organ, activate complement, and cause rapid thrombosis of vessels (Trulock 1997). When present, such antibodies are usually the result of prior exposure to alloantigens through blood transfusion, 6 pregnancy, or previous transplantation. Hyperacute rejection has virtually been eliminated by prescreening the recipient's serum for antibodies against a standard panel of cells (referred to as panel reactive antibodies), but it remains a major barrier to xenotransplantation (Trulock 1997). Acute rejection is primarily caused by T cells, but it is not known if a humoral response plays a role. In lung transplant patients, perivascular mononuclear infiltrates, with or without an accompanying lymphocytic bronchitis or bronchiolitis, are the hallmark of acute rejection (Trulock 1997). Acute rejection is rarely fatal. It is usually reversible by intravenous therapy with a corticosteroid, and does not result in significant permanent damage to the graft. Chronic rejection is an important obstacle to long-term survival of transplant patients. The defining characteristic of acute rejection is fibrous obliteration of endothelialized or epithelialized luminal structures (Trulock 1997). The exact mechanism is not known, but research has suggested both alloantigen-dependent and alloantigen-independent mechanisms are important (Trulock 1997). The lung is particularly prone to rejection, for several reasons. The lung is one of the largest transplantable organs, and it has an extensive vasculature that is perfused by the entire cardiac output and circulating immune system (Burke et al 1987). Also, the lung has a vast intrinsic immune apparatus, and the respiratory tract is constantly exposed to extrinsic infectious agents and other inhaled antigens that can cause inflammatory reactions, upregulation of alloantigen expression on bronchial epithelium, and activation of T cells (Burke et al 1987). 1.3 Primary Immunosuppression Early attempts at transplantation met with little success due to rejection of foreign tissue. Before advances in the field of immunology, it was thought that rejection was due to 7 malnutrition of the grafted tissue (Muller-Ruchholtz 1999). Peter B. Medawar, who was awarded the Nobel Prize for his pioneering work, discovered that graft rejection is immunologic in nature (Muller-Ruchholtz 1999). Following World War II, researchers in the transplantation area began to look for methods that would eliminate hostile immune reactions. Initially, it was believed that over time, a donor organ would completely integrate itself into the host. Thus, it was only necessary to control rejection for a short time following surgery (Bergan 1997). It became quickly evident that immunosuppression was required throughout the life of the recipient. The success of modern transplantation is due largely to immunosuppressive therapy. Immunosuppressive strategies can be divided into 4 classes: 1) chemical (also referred to as pharmaceutical; e.g., corticosteroids, azathiprine, mizoribine, cyclosporine A (CSA), tacrolimus, gusperimus, mycophenolate mofetil, and sirolimus, formerly known as rapamycin); 2) biological (also referred to as immunological; e.g., anti-lymphocyte globulin or A L G , anti-thymocyte globulin or A T G , monoclonal antibody against surface molecules of T cells, and blood transfusions); 3) physical (also referred to as radiological; e.g., local or total body irradiation, total lymphoid tissue irradiation); 4) surgical (e.g., thymectomy, splenectomy, lymphapheresis, plasmapheresis) (Oka and Yoshimura 1996). Tissue typing, a process which attempts to match the antigens of donor and recipient as closely as possible, helps to minimize the immune response without suppressing the recipient's immune system. Histocompatibility includes A B O blood matching and the identification of both donor and recipient antigen profiles (Bergan 1997). Total body irradiation, which was introduced in 1959, was the first attempt at inducing immunosuppression in the recipient. It was originally employed to suppress rejection in renal 8 transplant patients (Bergan 1997, Muller-Ruchholtz 1999). While effective in reducing the lymphocyte population, risk/benefit analysis revealed that the often severe side effects (most notably a tremendous susceptibility to infection and disease) were not worth the immunosuppression achieved (Bergan 1997, Muller-Ruchholtz 1999). It became clear that in order to improve postoperative prognosis in graft recipients, immunosuppressive drugs that were less toxic than radiation would need to be developed. 1.3.1 Corticosteroids Cortisone, an endogenous corticosteroid produced by the adrenal gland, was discovered in 1936. By 1964, it was documented that prednisone, a more potent synthetic derivative of cortisone, was effective in reversing rejection in renal allograft recipients (Bergan 1997). It was eventually determined that corticosteroids cross cell membranes, form complexes with specific binding proteins, and are then transported across the nuclear membrane to sites near genes involved in transcription and control of cytokines (Philip and Gerson 1998). The corticosteroid-binding protein complex inhibits gene transcription. The end result in T cells is that the interleukin-2 receptors are not expressed, thus preventing sensitization, activation, and proliferation of T cells (Philip and Gerson 1998). Many leukocytes are affected by corticosteroids. They not only reduce the number of monocytes and macrophages but also impair their functions, such as responses to lymphokines, phagocytosis, and interleukin-1 secretion (Kaplan et al, 1983). Besides the previously mentioned effects on T cell interleukin-2 receptor expression, corticosteroids can also inhibit killer T cell activity by blocking the production of interleukin-2 by helper T cells (Oka and Yoshimura 1996). The anti-inflammatory action of corticosteroids is well known. Their potent anti-inflammatory action is due to 9 inhibition of prostaglandin synthesis by inhibition of cycloxygenase activity (Oka and Yoshimura 1996). Although B cells are resistant to corticosteroids, they seem to indirectly inhibit antibody production due to their effects on macrophages and helper T cells (Oka and Yoshimura 1996). Corticosteroids have other effects, including suppression of febrile responses, release of macrophage-mediated cytokines, and inhibition of neutrophil chemoattractants and inflammatory mediators involved in cell-mediated cytotoxicity in tissues (Philip and Gerson 1998). The most desirable effect of corticosteroids as immunosuppressive agents is their immediate and selective ability to destroy cells that affect the immune system (Bergan 1997). When compared with other immunosuppressive agents, corticosteroids have a much broader spectrum of activity. However, their immunosuppressive effects are generally less potent, and their efficacy decreases or becomes non-existent with chronic usage (Oka and Yoshimura 1996). In addition, chronic use can permanently alter normal immune function and potentially cause a myriad of toxicities including: acne, hair growth, hypothalamus-pituitary-adrenal axis suppression resulting in cessation of production of endogenous corticosteroids, which could have serious deleterious consequences in times of stress (e.g., surgery or mechanical injury), psychiatric disturbances, and elevation of blood sugar due to the glucocorticoid activity of corticosteroids. More serious toxicities include iatrogenic Cushing's syndrome (which is the redistribution of body fat resulting in characteristic central obesity and a "buffalo hump"), hypertension, gastrointestinal tract toxicity, cataracts, and potential thinning and weakening of bone which can lead to hip fractures and aseptic necrosis (Bergan 1997, Philip and Gerson 1998). 10 1.3.2 Azathioprine In 1959, it was discovered that 6-mercaptopurine was able to lessen the immune response in rabbits challenged with a foreign protein antigen (Bergan 1997, Philip and Gerson 1998). Azathioprine, a purine analog of 6-mercaptopurine, was found to have greater immunosuppressive potency than 6-mercaptopurine. R N A synthesis and function are altered when azathioprine incorporates into cellular DNA, resulting in inhibition of purine synthesis and metabolism (Philip and Gerson 1998). In large doses, it inhibits B cell function and therefore, the humoral (antibody) response of the immune system (Oka and Yoshimura 1996). Azathioprine was first used clinically for immunosuppression in 1963. The primary toxicity of azathioprine is in the bone marrow. Azathioprine causes bone marrow suppression, which results in leukocytopenia, thrombocytopenia, and anemia (Philip and Gerson 1998, Oka and Yoshimura 1996). Azathioprine is also suspected to play at least a partial role in" causing lymphoma in chemically immunosuppressed transplant recipients (Philip and Gerson 1998). Azathioprine is also noted for its gastrointestinal toxicity, which manifests as nausea, vomiting, and pancreatitis. Hepatotoxicity is also encountered (Philip and Gerson 1998, Oka and Yoshimura 1996). Transient alopecia and compromised renal function are seen. Finally, skin lesions and uterine and cervical dysplasia may also manifest (Bergan 1997). 1.3.3 Cyclophosphamide The combination of corticosteroids and cyclophosphamide was initially found to be more effective than either drug alone, and the combination was used in transplantation research prior to 1964 (Philip and Gerson 1998). Because azathioprine was deemed less toxic than cyclophosphamide, azathioprine and corticosteroids became the standard immunosuppressive 11 therapy until the early 1980's when CSA was introduced into clinical practice. Cyclophosphamide, like azathioprine, is an antineoplastic drug, one of whose side effects (immunosuppression) is useful for another indication (transplantation in this case). The use of cyclophospamide in transplantation medicine is not widely recognized or known, as it was traditionally used as a second-line agent to replace azathioprine when azathioprine was withdrawn due to hepatotoxicity (Philip and Gerson 1998). 1.3.4 Mizoribine Mizoribine was developed in Japan in the late 1970's as a potential replacement for azathioprine. It has seen only limited use in Japan, and neglible use outside of Japan. Mizoribine is an imidazole nucleotide prodrug with antimicrobial activity, requiring phosphorylation to become active (Philip and Gerson 1998, Oka and Yoshimura 1996). Nucleic acid synthesis and cell proliferation are inhibited by mizoribine via disruption of the pathway from the xanthine-5'-nucleotide to the guanosine-5'-nucleotide in purine biosynthesis, resulting in guanine depletion (Philip and Gerson 1998, Oka and Yoshimura 1996). The disruption occurs because of inhibition of the enzyme inosine monophosphate dehydrogenase. Mizoribine is an effective immunosuppressant because of a key difference between immune cells and other cells in the body. Most cells use guanine generated through the salvage pathway for synthesis of DNA. Immune cells lack the enzymes active in the salvage pathway, and are dependent on the de novo synthesis of guanine. The advantage of mizoribine over azathioprine is that mizoribine is less toxic to bone marrow and liver. However, mizoribine has less immunosuppressive potency than azathioprine (Oka and Yoshimura 1996). Mizoribine toxicity is primarily gastrointestinal tract-related. This 12 toxicity is especially evident in dogs where it manifests as hemorrhagic enteritis and erosive mucosal lesions (Morris 1993). There is some controversy with this immunosuppressant: studies that have shown mizoribine to be less bone marrow- and hepatotoxic than azathioprine were not controlled. Also, the advantages of mizoribine over standard immunosuppressants have not been clearly established in clinical studies (Morris 1993). 1.3.5 Tacrolimus Tacrolimus, also commonly referred to as F K 506, was approved for clinical use in 1994 in the United States. It is a macrolide lactone metabolite produced by Streptomyces tsukubaensis (Philip and Gerson 1998, Oka and Yoshimura 1996). Although structurally distinct from CSA, the mechanism of action of tacrolimus is virtually identical to the mechanism of action of CSA, which will be discussed in a subsequent section. Briefly, tacrolimus ultimately affects helper T cells and inhibits cytotoxic T cells. The difference in the mechanism of action is that tacrolimus binds to F K binding protein, whereas CSA binds to cyclophilin. Tacrolimus is more potent than CSA in inhibiting T cell proliferation, B cell activation, and the production of other cytokines such as interleukin-3, interleukin-4, interferon-y, and granulocyte colony stimulating factor. The toxicity of tacrolimus is virtually identical to the toxicity of CSA, with differing incidences of various toxicities between the two. (The toxicity of CSA will be described in a subsequent section). As with CSA, nephrotoxicity is a frequently encountered adverse effect. Interestingly, it was originally hoped that tacrolimus would be less nephrotoxic than CSA, especially since they are structurally unrelated. However, tacrolimus is now regarded to be as nephrotoxic as CSA. In a study comparing CSA and tacrolimus as the primary immunosuppressant, recipients 13 treated with tacrolimus had fewer rejection episodes, but tacrolimus was found to be significantly more nephrotoxic (Morris 1993). 1.3.6 Gusperimus Gusperimus, also referred to as 15-deoxyspergualin, is a metabolite produced by Bacillus later osporrus, which has antineoplastic and antimicrobial activity. It is thought that gusperimus acts on macrophages, resulting in inhibition of oxidative metabolism, lysosomal enzyme synthesis, interleukin-1 production, and cell surface expression of major histocompatibility complex II antigens (Philip and Gerson 1998, Oka and Yoshimura 1996). Gusperimus also prevents antibody production by acting on B cells. Some data available show that gusperimus is effective as rescue therapy in rejection episodes (Suzuki et al 1990). Minor gastrointestinal disturbance has been reported, as well as bone marrow suppression. However, the myelotoxic effects are cytostatic in nature, and thus, bone marrow toxicity is rapidly reversible following discontinuation of therapy (Philip and Gerson 1998, Oka and Yoshimura 1996). Gusperimus is currently not available for clinical use in Canada. 1.3.7 Mycophenolate Mofetil RS-61443, or mycophenolate mofetil as it is now most commonly referred to, was originally isolated from Penicillium glaucum (Philip and Gerson 1998). Mycophenolate mofetil is a morpholinoethyl ester of mycophenolic acid. Mycophenolic acid is a potent, selective, uncompetitive, and reversible inhibitor of inosine monophosphate dehydrogenase (Hale et al 1998). Thus, it has properties in common with the much less widely used immunosuppressive agent, mizoribine. After oral administration, mycophenolate mofetil is rapidly and extensively 1 4 absorbed and is presystemically hydrolyzed to mycophenolic acid. The conversion to mycophenolic acid is so rapid that mycophenolate mofetil is not detected following oral administration. Intravenous administration is required to gather concentration-time data for the prodrug. Because of its enhanced effectiveness and decreased toxicity when compared to azathioprine, mycophenolate mofetil is now beginning to replace azathioprine as the secondary immunosuppressant in polytherapy. The lung transplant program at Vancouver Hospital and Health Sciences Centre (VHHSC; Vancouver, BC, Canada) is currently in the process of converting lung transplant recipients over to mycophenolate mofetil therapy. Patients at this center are now maintained on CSA or tacrolimus (only i f CSA is not tolerated), mycophenolate mofetil, and prednisone. Mycophenolate mofetil is usually well tolerated, and is notable in its lack of renal, hepatic, and bone marrow toxicity. The most frequently encountered adverse effects are gastrointestinal tract-related, manifesting as nausea and diarrhea. While these symptoms usually improve with dose reduction, they can be severe enough to necessitate discontinuation of therapy (Hale et al 1998). 1.3.8 Sirolimus Sirolimus, formerly known as rapamycin, is a macrolide product of Streptomyces hygroscopicus, and because of its macrolide nature, is structurally similar to tacrolimus. Like other immunosuppressive agents, it was originally discovered while screening for antimicrobial compounds. Sirolimus was reported in 1975 as having potent activity against Candida albicans (Mattel et al 1975). Sirolimus affects T and B cells directly by preventing cytokines from activating them (Morris 1993). It inhibits binding of interleukin-2 to its receptor without affecting the production of interleukin-2 (Dumont et al 1990). Another noteworthy property is 15 that tacrolimus and CSA only slightly inhibit interleukin-2 and interleukin-4 induced growth of T cells, while sirolimus is very effective (Dumont et al 1990). Like tacrolimus, sirolimus binds to F K binding protein, but unlike tacrolimus, this sirolimus-FK binding protein complex does not result in calcineurin inhibition (Liu et al 1991). Experience with sirolimus in humans is limited, but in canine animal models, sirolimus causes weight loss, testicular atrophy, lethargy, and in rat models, sirolimus is diabetogenic (Morris 1993). Sirolimus is currently not available for clinical use in Canada, although it is undergoing clinical trials in North America. 1.3.9 Brequinar Originally developed as an antineoplastic, brequinar was found to have immunosuppressive properties. It inhibits the action of dihydro-orotate dehydrogenase, an enzyme involved in the de novo biosynthetic pathway of pyrimidines involved in the synthesis of D N A and R N A (Allison et al 1993, Morris 1993). Inhibition of uridine synthesis also blocks the salvage pathway in lymphocytes (Allison et al 1993, Morris 1993). Brequinar does not, however, affect the synthesis of interleukin-2. In experimental models of transplantation, adverse effects of brequinar include thrombocytopenia, desquamative maculopapular dermatitis, mucositis, and gastrointestinal toxicity (Philip and Gerson 1998). These adverse effects appear to be dose-dependent and reversible upon discontinuation of therapy, and are consistently seen in all species that have been administered the drug. Preliminary data also show that there is a greater degree of toxicity with more frequent dosing (Makowka et al 1993). Brequinar is not currently available for clinical use in Canada. 16 1.3.10 Leflunomide All-1126, the active metabolite of leflunomide, was developed specifically as an immunosuppressant, but it was found to cause severe gastrointestinal disturbances. The prodrug, leflunomide, helps to ameliorate the gastrointestinal disturbances, as it is converted to A77-1726 following absorption (Philip and Gerson 1998). Leflunomide suppresses the immune system by inhibiting tyrosine kinase-associated growth factors (Morris 1993). Leflunomide also blocks the production and action of interleukin-2 as well as the mitogenic activity of stimulated T cells (Morris 1993). Experience with leflunomide in clinical transplantation is limited. However, data are available regarding its use in rheumatoid arthritis patients in Europe (Morris 1993). It appears to be well tolerated. More importantly, unlike other immunosuppressive drugs, it does not cause myelotoxicity or nephrotoxicity (Morris 1993). Lung Transplantation 1.1 The History Of Lung Transplantation The first human lung transplant was performed in 1963 by Dr. James Hardy at the University of Mississippi (Blumenstock et al 1993, Grover et al 1997, Hardy et al 1963). The recipient died on the 18 th postoperative day from renal failure (Grover et al 1997). Between 1963 and 1974, 36 lung transplants were performed worldwide, but only 2 recipients lived longer than 1 month (Veith et al 1974). Because of this, the area of lung transplantation was largely inactive until CSA was introduced clinically in the early 1980's (Trulock 1997). In 1981, heart-lung transplantation was utilized to treat pulmonary vascular disease (Reitz et al 1982). Pulmonary fibrosis was managed by single lung transplantation beginning in 1983 (Toronto 17 Lung Transplant Group 1986), and in 1986, double lung transplantation for obstructive lung disease was started (Cooper et al 1989). Dr. Frank Veith made many contributions to the development of lung transplantation beginning in 1969 and extending through 1983 (Grover et al 1997). During that time, he developed techniques for pulmonary artery and pulmonary venous anastomoses, as well as a telescoping anastomosis (Grover et al 1997). In 1979, Dr. Veith made the extremely important discovery regarding the effect of donor bronchial length on healing of the bronchial anastomosis (Grover et al 1997). He noted that the shorter the bronchus, the better the healing. In 1983, Dr. Veith reported improved experimental results with lung transplantation using CSA as the primary immunosuppressant (Grover et al 1997). Dr. Joel Cooper, from the University of Toronto, performed Toronto's first lung transplant in 1978. Unfortunately, the recipient survived only for 17 days, but he was considered high risk prior to the transplant (Nelems et al 1980). In 1981, Dr. Cooper and colleagues discovered that corticosteroids used in the immunosuppressive regimen decreased bronchial healing, while azathioprine and CSA did not (Lima et al 1981). In 1983, Dr. Cooper's group reported that immunosuppressive agents increased bronchial disruption and necrosis (Goldberg et al 1983). Also in 1983, he reported on studies using the omental wrap in an experimental model with improved bronchial healing and neovascularization (Morgan et al 1983). 1.2 Indications For Lung Transplantation Current indications for single lung transplantation include restrictive lung disease, emphysema, pulmonary hypertension, and other nonseptic, end-stage pulmonary disease including restrictive lung disease secondary to connective tissue disorders (Grover et al 1997, 18 Higenbottam et al 1990, Trulock 1997). Indications for bilateral sequential lung transplantation include cystic fibrosis and patients in whom there is chronic infection with end-stage pulmonary failure, including patients with bronchiectasis (Grover et al 1997, Higenbottam et al 1990, Trulock 1997). Bilateral sequential lung transplantation is preferred for patients with primary and secondary pulmonary hypertension if it is possible to do so, as that type of transplant appears to decrease reperfusion edema (Grover et al 1997). Emphysema patients with lung volumes so great that it would be very difficult to find a large enough single lung may also be selected for bilateral sequential lung transplantation (Grover et al 1997). The most common reasons for lung transplantation are chronic obstructive pulmonary disease, a i -antitrypsin deficiency emphysema, idiopathic pulmonary fibrosis, cystic fibrosis, primary pulmonary hypertension, and Eisenmenger's syndrome (Hosenpud et al 1996). Other less common reasons include bronchiectasis, sarcoidosis, lymphangioleiomyomatosis, and eosinophilic granuloma of the lung (Trulock 1997). There is some debate as to how to manage patients with primary pulmonary hypertension. Single, bilateral sequential, and heart-lung transplantation are all therapeutic options for this patient population. The advantages of single lung transplantation for these patients are that it conserves donor organs, is an easier procedure, is very effective in decreasing pulmonary vascular resistance which improves right ventricular ejection fraction, and provides good symptomatic relief (Mal et al 1989). The disadvantages are that reperfusion pulmonary edema is frequently a major problem and can be a cause of mortality, 85% of pulmonary blood flow goes to the transplanted lung which means that dysfunction in the transplanted lung is not tolerated, and rejection is more serious because of the mismatch in pulmonary blood flow (Mal et al 1989). With this procedure, the operative mortality is 26%, and the actuarial survival at 1 year is 66% 19 (Hosenpud et al 1994). There are those clinicians who believe that because primary pulmonary hypertension patients are often young, they will benefit more in the long term by receiving 2 lungs. An advantage of bilateral sequential lung transplantation is that it offers even distribution of pulmonary blood flow to both lungs, which results in less reperfusion edema and easier perioperative management. Because of the balanced flow, acute or chronic rejection is better tolerated. In addition, similar to single lung transplantation, pulmonary vascular resistance is reduced and right ventricular ejection fraction is increased. An advantage of bilateral sequential lung transplantation over heart-lung transplantation is that the heart is available for another recipient (Grover et al 1997). The disadvantages are that the procedure is more challenging than single lung transplantation, requires longer cardiopulmonary bypass time, and uses 2 lungs per patient, which is a concern given the dramatic shortage of donor lungs (Grover et al 1997). The operative mortality is 10% and the 1-year actuarial survival is 77% (Hosenpud et al 1994). 1.3 Recipient Criteria Ideally the potential recipient has disease confined to the lungs with no other major organ dysfunction or disease (Grover et al 1997). In most circumstances, the patient should be under 65 years of age for a single lung transplant and under 60 for a bilateral sequential lung transplant (Grover et al 1997, Trulock 1997). In addition, there should be no history of alcohol or recreational drug dependency, and the recipient should be psychologically stable (Grover et al 1997). Absolute contraindications for lung transplantation include multisystem disease other than lung, history of carcinoma or sarcoma with a possibility of recurrence, current infection, significant renal or hepatic dysfunction, cigarette smoking within 3 to 4 months, drug or alcohol 20 abuse, psychiatric instability, and poor medical compliance. Bronchiectasis and chronic or recurrent pulmonary infection are contraindications for single lung transplantation, and presence of these conditions necessitates bilateral sequential lung transplantation (Grover et al 1997). Relative contraindications include insulin-dependent diabetes mellitus, age above the recommended guidelines, the presence of significant coronary artery disease and/or left ventricular dysfunction, long-term ventilation support, previous thoracic surgery, use of corticosteroids greater than either 20 mg per day or 0.2 to 0.3 mg/kg/day, hemodynamic instability, extreme cachexia, morbid obesity, advanced connective tissue disease associated with other organ dysfunction, and previous lung transplantation (Grover et al 1997, Massard et al 1993, Schafers etal 1992). Lung transplantation is appropriate when other therapeutic options have been exhausted and when the patient's prognosis will be improved by the procedure. The risks of aggressive medical treatment (e.g., a vasodilator for primary pulmonary hypertension or a cytotoxic drug for idiopathic pulmonary fibrosis) are generally considered to be less than those of transplantation (Trulock 1997). Medical treatment may delay the need for transplantation or serve as a bridge through the long waiting period. Quality of life is the prime motivation for transplantation for many patients, but prognosis should be the main determinant of timing (Trulock 1997). The median waiting time for lung transplantation was 550 days for patients who entered the waiting list in 1994 (Trulock 1997). Thus, a realistic waiting time of 1 to 2 years should be incorporated into the transplantation strategy. For patients with chronic obstructive pulmonary disease and a i -antitrypsin deficiency, lung transplantation should be undertaken when post-bronchodilator forced expiratory volume in 1 second is less than 25% predicted, there is resting hypoxia (partial pressure of oxygen less than 55 to 60 mm Hg), hypercapnia, secondary 21 pulmonary hypertension, and a clinical course characterized by rapid rate of decline of forced expiratory volume in 1 second or life threatening exacerbations are present (Trulock 1997). When post-bronchodilator forced expiratory volume in 1 second is less than 30% predicted, there is resting hypoxia (partial pressure of oxygen less than 55 mm Hg), hypercapnia, and a clinical course with increasing frequency and severity of exacerbations are present, lung transplantation should be undertaken for cystic fibrosis patients (Trulock 1997). For patients with idiopathic pulmonary fibrosis, lung transplantation should be undertaken when vital capacity is less than 60 to 65% predicted, and when there is resting hypoxia, secondary pulmonary hypertension, and clinical, radiographic, or physiologic progression while on medical therapy (Trulock 1997). Finally, i f the patient has a New York Heart Association functional class III or IV rating, mean right atrial pressure greater than 10 mm Hg, mean pulmonary arterial pressure greater than 50 mm Hg, and a cardiac index less than 2.5 L/min/m 2, lung transplantation should be undertaken in patients with primary pulmonary hypertension (Trulock 1997). 1.4 Donor Criteria During the mid-1960's, surgeons were performing few transplants, and there was not much of a demand for donor organs. In the case of renal transplants, most donors were living relatives of the recipients (Bergan 1997). Because instances occurred that resulted in donor abuse (e.g., psychological abuse from relatives of a potential donor who refused to donate) and death, donation was not as complication-free as it is today, and because the demand for donated organs was slowly growing, alternatives were needed (Bergan 1997). Also, in the late 1960's, physical death was defined as the moment the heart stopped beating. Because there was no concept of brain death at that time, it was unethical for organs to 22 be harvested for transplantation from a patient who was brain dead but whose heart was still beating. This changed when a Boston physician, Dr. Moore, transplanted a liver from a brain dead police officer killed in the line of duty into a patient who desperately needed a liver transplant (Bergan 1997). The concept of brain death was formally defined by Harvard Medical School, and this increased the donor pool of organs and tissues for transplantation. It also allowed donated organs to be harvested from a healthy blood supply, allowing them to be preserved in optimum condition (Bergan 1997). Potential lung donors should have a partial pressure of oxygen greater than 300 mm Hg with ventilator settings of 100% fraction of inspired oxygen, a positive end-expiratory pressure of 5 cm H 2 O , and a tidal volume of 12 mL/kg/min (Grover et al 1997, Trulock 1997). The chest X-ray should be clear, the sputum stain should be negative for fungus and preferably negative for gram-negative rods, there should be no purulent secretions, the age of the donor should be less than 60 years, HIV and hepatitis B and C antigens should be negative, and the donor should be on a ventilator for less than 1 week (Grover et al 1997, Trulock 1997). The donor should have no history of lung disease including asthma, no history of cancer except for non-melanomatous skin cancer, and no history of "high risk" behavior. The circumference of the nipple line should be within 10 to 13 cm of the recipient's. Also, there should be reasonable matching of height, weight, and vertical and horizontal lung dimensions, a less than 30-pack year smoking history, and no gross purulence from lobar or segmental orifices on bronchoscopy (Grover et al 1997, Trulock 1997). 23 1.5 Current Challenges In Lung Transplantation Some notable problems need to be overcome before lung transplantation will offer consistent and long-term relief of symptoms and improved quality and duration of life. The most important of these problems is chronic rejection, as manifested by obliterative bronchiolitis. Although there is no way to prevent the bronchiolitis obliterans syndrome, the consensus is that aggressive prophylaxis of cytomegalovirus infection, along with switching CSA to tacrolimus in patients with repeated episodes of rejection, is important (Grover et al 1997). Early diagnosis of rejection is also very important. If either acute or chronic rejection is suspected, fiberoptic bronchoscopy and bronchoalveolar lavage with routine, fungal, and viral cultures and transbronchial biopsy should be undertaken (Grover et al 1997). Another less common but potentially fatal problem in lung transplantation is post-transplant lymphoproliferative disorder. This occurs in approximately 5% of lung transplant recipients, and is initially treated by decreasing the doses of the immunosuppressive drugs. Chemotherapy is added if this intervention is unsuccessful (Randhawa et al 1989). A controversial problem is whether or not to retransplant patients who develop either early graft failure or chronic rejection as manifested by bronchiolitis obliterans. There is a marked difference in survival between first time lung transplants and reoperative lung transplants, with a 1-year survival of approximately 40% in retransplants versus greater than 70% in first time recipients. For some as yet unknown reason, the figure for retransplants had risen to slightly greater than 50% by 1995 (Novick et al 1994, 1995). This problem offers an ethical dilemma as well: is it ethically acceptable to deplete the already limited supply of donor lungs even further by using them for retransplantation, knowing that the survival is less than that for first time recipients (Grover et al 1997)? 2 4 Another challenge that continues to be present in lung transplantation is donor shortage. In 1993, there was a 2:1 demand:supply ratio for lung transplants (Grover et al 1997). Since that time, the number of donors has remained relatively constant, but the demand for donor lungs has increased (Trulock 1997). Because of this, many patients die while on the waiting list. 1.6 Recent Advances In Lung Transplantation A study published in 1994 (Cohen et al 1994) shows that it is possible to use living related donors for cystic fibrosis patients. Seven recipients received bilateral lobes from 14 living related donors. There were no donor deaths or complications except for prolonged air leaks in 3 of the donors. These donors had also donated the middle lobe of their lung, in addition to the lower lobe. Normally, only the lower lobe of the lung is harvested for this type of procedure (Cohen et al 1994). There were no deaths in the recipients with cystic fibrosis, although there has been some mortality in infant and pediatric recipients who received lobes for other reasons (Starnes et al 1994). Living related transplantation can offer earlier transplantation for patients who are at risk for dying while waiting for a transplant. Because donors are related, there is the potential for better organ compatibility. In addition, the donor supply is enlarged, and it is a continuation of a tradition of living related organ donation that has been long established for renal transplantation and used occasionally for liver transplantation (Grover et al 1997). Current lung preservation techniques are capable of preserving harvested lungs for up to 8 hours of cold ischemia. Ongoing investigations into supplementary preservation measures, including the use of antioxidants, leukocyte depletion, nitric oxide, and cytokine manipulation, may extend this time past 8 hours (Grover et al 1997). 25 1.7 The Future Of Lung Transplantation Concurrent transplantation of bone marrow along with the primary solid organ is currently a major area of investigation. It has been shown that in recipients of cadaveric kidneys, livers, hearts, and lungs who are given donor bone marrow at the time of their transplant, and who are given tacrolimus and prednisone for immunosuppresion, there is a trend towards donor specific non-reactivity and increasing survival (Pham et al 1995). The area of xenotransplantation has been a topic of much research as well. One of the most well known examples of xenotransplantation occurred in the early 1980's. Dr. Leonard Bailey transplanted a baboon's heart into a critically i l l infant termed Baby Fae (Bergan 1997). Baby Fae survived for only 20 days, but much research was performed during this short time (Bergan 1997). Hyperacute rejection and transmission of infection are major problems with xenografts of this type. It was discovered that grafts from non-human primates stimulate a greater immune response than grafts from other mammals. This stimulated research into the use of pigs as potential donors. These animals are large, plentiful, and have a short gestation period. Finally, it is my opinion that the field of transplantation may also see the use of donor-specific organs grown in vitro i f ethical concerns are worked out. This would eliminate the need for immunosuppressive drugs and their associated toxicities, as well as the long waits for donor organs. Cyclosporine 1.1 The History Of Cyclosporine The journey from discovery to market began with a soil sample collected from the Hardanger Vidda in Norway (Borel 1983, Borel et al 1989, Stahelin 1986). The soil contained 26 fungi that were of particular interest to drug discovery scientists. The cyclosporines were originally discovered in the early 1970's by scientists at Sandoz Ltd. in Basel, Switzerland (Borel 1983, Borel et al 1989). They were performing routine screening for compounds with antifungal activity (Borel et al 1989). The crude extracts of two strains of fungi, Cylindrocarpon lucidum Booth and Tolypocladium inflation Gams, possessed antifungal activity. However, this antifungal activity was not potent enough to warrant further development (Borel 1983, Borel et al 1989). It was subsequently observed that the antifungal activity was coupled with an unusually low toxicity in mice. Because of this, the metabolite mixture was put through a limited pharmacological screening program (Borel et al 1989). In early 1972, J. F. Borel and also, H. Stahelin observed that the metabolite mixture appeared to have immunosuppressive properties in murine animal models (Borel et al 1989). CSA was purified in 1973, and it was observed that CSA exerted unique immunosuppressive effects while having no bone marrow toxicity (Borel et al 1976, Borel et al 1977). CSA was first studied clinically in 1978, and was approved for use in organ transplantation first in Switzerland and then the United States in 1983 (Borel et al 1989). Table 1 summarizes the development of CSA. 27 Table 1. The development of cyclsporine. (Adapted from Borel 1983 and Borel et al 1989.) Year Development 1970 Isolation of 2 new strains of fungi which produce antifungal metabolites by B. Thiele. Isolation of a metabolite mixture and characterization as novel neutral polypeptides by Z. L. Kis and colleagues. 1971 Isolation of the partially purified 2-component metabolite mixture (24-556) on a preparative scale for initial biological screening by Haerri and Ruegger 1972 Discovery of the immunosuppressive properties of metabolite 24-556 in mice by J F. Borel. 1973 Purification of CSA (27-400) by Ruegger. Culture and production of CSA by Dreyfus and colleagues. 1974 Animal studies of the immunosuppressive activity of CSA, in vitro and in vivo by J. F. Borel. 1975 Elucidation of the structure of CSA via X-ray studies and chemical degradation by Petcher and colleagues and Ruegger and colleagues. 1976 Toxicity studies in rats and monkeys demonstrate the selectivity of CSA for lymphocytes and a lack of effect on hematopoiesis. World-wide confirmation of specific immunosuppressive effect of CSA in experimental transplantations and other models. 1978 The first clinical trials using CSA by Calne and colleagues and Powles and colleagues. 1980 Total synthesis of CSA by Wenger. 1981 Publication of the antischistosomal and anti-malarial activity of CSA by Bueding and colleagues and Thommen. 1983 Sandimmune® first approved for clinical use in organ transplantation in Switzerland, and then the United States and some other countries later in the year 1985 The first controlled clinical trials with CSA in autoimmune diseases. 1.2 Cyclosporine Physical-Chemical Properties CSA (C62H111N11O12, Figure 1), also known as cyclosporine A , cyclosporin A, antibiotic S 748IF 1, and CsA, is normally encountered as a white powder, occasionally with a faint yellow cast (Cyclosporine AHFS monograph 1998, Sigma Product Information Sheet 2000). It has a melting point range of 148 to 151 °C, a molecular weight of 1202.6, and an optical rotation of-244° (Sigma Product Information Sheet 2000). In its natural powder form, CSA is expected to 28 be stable at 2 to 8 °C for at least 2 years i f stored sealed in the dark (Sigma Product Information Sheet 2000). CSA has a solubility of 10 mg/mL in methylene chloride, 6 mg/mL in chloroform, 10 mg/mL in ethanol, and 50 mg/mL in DMSO (Sigma Product Information Sheet 2000). Solutions are clear and colorless to faint yellow. CSA is often reported to be slightly soluble in water and saturated hydrocarbons. It is recommended that stock solutions in ethanol or DMSO should be stored at -20°C (Sigma Product Information Sheet 2000). Since CSA is poorly water soluble, but is often administered intravenously, admixtures must be thoroughly mixed to ensure that CSA is properly dissolved. Solutions in intravenous fluids must be shaken vigorously to assure proper dispersion. CSA is stable i f solutions are protected from light, but its concentration may drop due to adsorption to container walls (Sigma Product Information Sheet 2000). If diluted in glucose 5% or glucose/amino acid solutions and stored at room temperature in the dark, CSA is stable for 72 hours (Sigma Product Information Sheet 2000). However, when diluted in sodium chloride 0.9%, CSA is stable for only 8 hours under similar storage conditions (Sigma Product Information Sheet 2000). CSA is a hydrophobic cyclic peptide composed of 11 amino acid residues, all having the S-configuration except for the D-alanine in position 8, which has the R-configuration, and sarcosine in position 3 (Wenger 1983). Ten of the amino acids are known amino acids: a-aminobutyric acid in position 2, sarcosine in position 3, N-methylleucine in positions 4, 6, 9, and 10, valine in position 5, alanine in position 7, D-alanine in position 8, and N-methylvaline in position 11 (Wenger 1983). At the time of synthesis, the amino acid in position 1, a novel P-hydroxy, unsaturated, 9 carbon amino acid {(4R)-4-[(E)-2-butenyl]-4,N-dimethyl-L-threonine); MeBmt} (Kahan 1989) had not been previously isolated or known in free form (Wenger 1983). 29 1.3 Pharmacology And Therapeutic Use O f Cyclosporine CSA has antifungal activity, but it is of no use clinically because it does not have action against human pathogens except for coccidioidomycosis (Borel et al 1989). CSA also has antischistosomal and antimalarial activity (Borel 1983). In addition, CSA has been used to mitigate multidrug resistance in hematological cancer patients, due to CSA's inhibition of P-glycoprotein activity (Wu et al 1995). However, CSA is primarily used therapeutically for its potent immunosuppressive activity. While it has been used to treat autoimmune diseases (Fathman and Myers 1992) and for the prevention and treatment of graft-versus-host-disease in bone marrow transplant patients (Pai et al 1994), CSA is mainly used as the cornerstone of immunosuppressive polytherapy in solid organ transplant patients (Trulock 1997, Tsunoda and Aweeka 1996). As mentioned previously, the lung transplant program at V H H S C typically uses triple immunosuppressive therapy in lung transplant patients: CSA or tacrolimus (only i f CSA is not tolerated or contraindications are present), mycophenolate or azathioprine (although patients are being converted to mycophenolate), and prednisone. A standard protocol for CSA dosing and desired CSA trough levels is utilized (Table 2). Table 2. The cyclosporine dosing protocol at Vancouver Hospital and Health Sciences Centre Pre-operative: cyclosporine (Neoral®) 3 mg/kg Post-operative: cyclosporine (Neoral®) 9 mg/kg/day given in two divided doses Time Post-Transplant (months) Target Cyclosporine Trough Level (ug/L) < 1 350 to,400 1 to 3 300 to 350 3 to 6 250 to 300 6 to 12 200 to 250 > 12 150 to 200 30 CJ ± CJ ± CJ CJ—CJ I-TT1 O 00 2 0> 31 The immunosuppressive activity of CSA (Figure 2) appears to be mediated by intracellular, rather than cell membrane, receptors (Freeman 1991). CSA binds to several cytosolic proteins including calmodulin, cyclophilin, and peptidyl-prolyl cis-trans isomerase, an enzyme involved in the refolding of intracellular proteins, which is thought to be the same protein as cyclophilin (Freeman 1991, Halloran and Madrenas 1991, Zenke et al 1993). It is thought that the immunosuppressive activity of CSA is due to its binding to cyclophilin, which results in the formation of a complex and inhibition of cis-trans isomerase activity (Freeman 1991, Halloran and Madrenas 1991, Zenke et al 1993). This complex acts on calcineurin, a serine-threonine phosphatase that requires calcium binding for activity (Pai et al 1994). Under normal cellular conditions, calcineurin dephosphorylates the cytoplasmic subunit of the transcription factor NF-AT, allowing it to translocate to the nucleus where it is involved in transcriptional activation of lymphokine genes (Fruman et al 1992). Expression of these genes results in production of lymphokines, of which interleukin-2 is the most important in T cell activation. In vitro studies have shown that inhibition of calcineurin activity by CSA results in inhibition of the T cell signal transduction cascade and, therefore, activation of T cells (Clipstone and Crabtree 1993, Fruman et al 1992). At typical therapeutic dosages, CSA inhibits calcineurin activity by about 50% (Batiuk et al 1995a, 1995b, Cantarovich et al 1998, Quien et al 1997), which is likely one of the reasons why transplant recipients have sufficient immune system activity to fight infection and experience periods of rejection (Batiuk et al 1995b). 1.4 Considerations For Cyclosporine Use In Lung Transplantation CSA immunosuppression has helped to make lung transplantation a viable therapeutic option for the management of patients with end-stage lung disease (Higenbottam et al 1990, 32 Trull et al 1999). Compared to other types of transplant recipients, higher target CSA concentrations are used in lung transplant patients due to the severe consequences of graft failure. In spite of this, most patients still experience at least 1 acute rejection episode in the first 3 months post-transplant (Trull et al 1999). Acute rejection (both the incidence and severity) appears to be the most important risk factor for chronic graft failure (Trull et al 1999, Best et al 1996). Also, continued exposure to high concentrations of cyclosporine could result in a higher incidence of adverse effects such as nephrotoxicity, hepatotoxicity, hypertension, and severe infection (Trull et al 1999). Lung transplant patients may eventually require dialysis and subsequent kidney transplantation due to CSA nephrotoxicity. Differences in pharmacokinetics between Neoral® (NEO) and Sandimmune® (SIM), due to differences in formulation between the 2 products (which will be described in a later section), have implications in lung transplantation. Because CSA absorption with NEO is enhanced, it is possible that patients are exposed to greater amounts of CSA with NEO than with SIM when the same trough levels are maintained (Friman and Backman 1996). This could have implications in terms of incidence of adverse effects, although studies have shown that the two formulations are tolerated equally well (Kesten et al 1998a, Trull et al 1999, Zaldonis et al 1998). Increased absorption with NEO may also reduce CSA dosage requirements in lung transplant patients with cystic fibrosis, who typically have high SIM dosage requirements due to erratic absorption characteristics, poorer bioavailability, and delayed absorption (Trull et al 1999). However, one study (Trull et al 1999) has shown that there were no significant differences in doses of NEO and SIM required to achieve similar concentrations in lung transplant patients with cystic fibrosis. Reduced inter- and intra- patient variability with NEO may also help to reduce acute rejection 33 and possibly graft failure due to obliterative bronchiolitis since variability in trough levels with SIM has been identified as a significant risk factor for acute rejection (Best et al 1992). 1.5 F o r m u l a t i o n s O f C y c l o s p o r i n e U s e d I n L u n g T r a n s p l a n t a t i o n CSA is available in both parenteral and oral dosage forms. Cyclosporine concentrate for injection appears as a clear, faintly brownish-yellow solution. This formulation is a sterile solution of CSA in polyoxyl 35 castor oil (Cremophor® EL,polyethoxylated castor oil) with 32.9% alcohol (AHFS Drug Information 1999). At the time of manufacture, the air in the ampules is replaced with nitrogen (AHFS Drug Information 1999). The purpose of this is to help prevent oxidation of the formulation ingredients. Oral dosage forms of the original SIM formulation of CSA consist of an oral solution and conventional liquid-filled, soft gelatin capsules. SIM oral solution has a clear, yellow, oily appearance. The oral solution contains the drug in an olive oil and peglicol 5 oleate (Labrafil® M 1944CS) vehicle with 12.5% alcohol (AHFS Drug Information 1999). SIM capsules were available in 25, 50, and 100 mg strengths. The SIM formulation of CSA is no longer available in Canada. C S A is currently commercially available as a non-aqueous liquid formulation of the drug (NEO) that immediately forms an emulsion in aqueous fluids. This is done via mized micelles, thus forming an oil-in-water type of emulsion. The formulation is available as an oral solution for emulsion and as oral 25 and 100 mg liquid-filled soft gelatin capsules containing the oral solution for emulsion (AHFS Drug Information 1999). When exposed to an aqueous environment such as that found in the digestive tract, the oral solution for emulsion forms a homogenous transparent emulsion with a droplet size smaller than 100 nm in diameter (AHFS 35 Drug Information 1999). Because of this, NEO is often referred to as a microemulsion formulation of CSA. In this formulation, the molecular structure of CSA is unmodified, and aqueous dilution results in formation of an emulsion without reprecipitation of the drug (AHFS Drug Information 1999). In NEO, CSA is dispersed in a mixture of propylene glycol, which functions as the hydrophilic solvent, and corn oil monoglycerides, diglycerides, and triglycerides, which function as lipophilic solvents (AHFS Drug Information 1999). When dispersed, polyoxyl 40 hydrogenated castor oil functions as a surfactant, and d,l-a-tocopherol functions as an antioxidant (AHFS Drug Information 1999). The oral solution for emulsion, both as the solution and in the capsules, also contains dehydrated alcohol in a maximum concentration of 9.5% (AHFS Drug Information 1999). NEO and SIM are not bioequivalent due to formulation and pharmacokinetic differences. Thus, they cannot be used interchangeably, although this is no longer an issue in Canada where, as previously mentioned, SIM is no longer available. 1.6 Pharmacokinetics Of Cyclosporine (This section refers to the SIM formulation of CSA unless otherwise noted). 1.6.1 Absorption Following oral administration, CSA (SIM) is incompletely and erratically absorbed (Freeman 1991, Friman and Backman 1996, Lemaire et al 1990, McMillan 1989, Yee and Saloman 1992). Mean bioavailability (F) is about 30% (Freeman 1991, Friman and Backman 1996, Lemaire et al 1990, McMillan 1989, Yee and Saloman 1992) with ranges of 8 to 60% (Lemaire et al 1990) and less than 5% to 90% being reported (Yee and Saloman 1992), depending on the reference consulted. In addition, bioavailability can increase with increased 36 time post-transplant in renal transplant patients (Lemaire et al 1990, Yee and Saloman 1992). The reasons for this are unknown. It is possible that the surgery disrupts homeostasis or some normal physiological parameter responsible for CSA absorption. If CSA is absorbed across the gastrointestinal tract membrane by transporters, it is possible that the number of transporters increases post-transplant. There may be an endogenous compound that increases absorption of CSA, similar to the situation involving vitamin D and calcium absorption. Of course, this is purely speculation. Factors which can affect absorption of CSA include availability of bile, gastrointestinal status, and coadministration with food. The effect of bile on absorption of CSA was discovered from observations in liver transplant patients. In the immediate post-transplant period, absorption of CSA is poor due to gastrointestinal disturbances and disruption of the biliary system (van Mourik 1998). NEO, due to its microemulsion formulation, does not require solubilization by bile for absorption, and liver transplant patients have been successfully treated orally in the immediate post-transplant period (Tredger 1995). The effect of food on CSA absorption is controversial, with food being reported to both increase and decrease CSA absorption (Friman and Backman 1996). Administration with food or food and bile acid tablets resulted in two peaks being observed in the concentration versus time curves obtained in a study published in 1990 (Lindholm et al 1990). CSA given under fasting conditions did not produce two peaks, suggesting that enterohepatic circulation occurs under non-fasting conditions (Lindholm et al 1990). In healthy volunteers, administration of SIM with a fat-rich meal nearly doubled time to the maximum concentration encountered during the dosing interval (T m a x ) and resulted in a 37% increase in area under the concentration versus time curve (AUC) (Mueller et al 1994). This was associated with significant elevations in CSA 37 concentrations when compared to CSA concentrations obtained under fasting conditions. The influence of a fat-rich meal on the pharmacokinetics of NEO was less pronounced. The maximum concentration observed during the dosing interval (C m a x ) decreased by 26%, there was no significant change in T m a x , and A U C decreased by 15% (Mueller et al 1994). Gastrointestinal status can also alter CSA absorption. Chemoradiation-induced enteritis, acute graft-versus-host-disease involving the intestine, diarrhea, and metoclopramide therapy have all been shown to decrease CSA absorption (Freeman 1991, McMillan 1989, Yee and Saloman 1992). In renal transplant patients, conversion from SIM to NEO resulted in increases in C m a x and A U C in one study (Kovarik et al 1994) and significant increases in C m a x , A U C , and average steady-state concentrations in another (Sketris et al 1994). A n additional study in renal transplant patients (Browne et al 1994) found significantly higher C m a x values and significantly shorter T m a x values in the same patients given SIM and then NEO. One study published in 1998 involving lung transplant patients (Kesten et al 1998a) evaluated changes in pharmacokinetic parameters following conversion from SIM to NEO. A significant increase in C m a x and A U C was found after conversion to NEO. 1.6.2 Distribution CSA has a large apparent volume of distribution (Vd) of 4-6 L/kg in kidney and bone marrow transplant patients (Yee and Saloman 1992). A range of 1.8-13.8 L/kg has also been reported (Lemaire et al 1990), indicating that CSA is highly tissue bound. In addition, CSA can be measured in tissues for at least two weeks after CSA therapy is discontinued (Yee and 38 Saloman 1992). CSA is able to partition into many body tissues due to its lipophilic nature and the wide distribution of its binding protein, cyclophilin, throughout the body (Freeman 1991). In vitro, 40 to 50% of CSA distributes into the erythrocyte fraction, 10 to 20% distributes into the leukocyte fraction, and 30 to 40% distributes into the plasma fraction (Christians and Sewing 1993, Lemaire et al 1990, Yee and Saloman 1992). CSA binding to erythrocytes is saturable (Yee and Saloman 1992). In addition, erythrocyte binding is both temperature and hematocrit dependent, such that increases in both result in an increase in CSA recovered in the cellular fraction (Yee and Saloman 1992). In plasma, CSA is highly bound to lipoproteins, which account for 10 to 15% of all plasma proteins (Yee and Saloman 1992). Forty-six percent of CSA is bound to high density lipoproteins, 31% to low density lipoproteins, 16% to very low density lipoproteins, and 8% to proteins, primarily albumin (Christians and Sewing 1993). It has been further confirmed in a more recent study that the majority of CSA distributes into HDL and L D L , with lesser amounts distributing into V L D L and proteins (Wasan et al 1997). Protein binding is concentration-independent between the concentration range of 25 to 500 ng/mL in vitro and temperature-dependent, such that CSA is 90 to 95% protein bound at 20 °C and 70% bound at 4 °C (Yee and Saloman 1992). Fraction unbound in plasma of 1-1.5% (Christians and Sewing 1993) and 1-17% (depending on the method of measurement) (Yee and Saloman 1992) have been reported. 1.6.3 Metabolism CSA is extensively metabolized by the liver and in the gastrointestinal tract by cytochrome P450 3A4. Because this isoenzyme is responsible for the metabolism of many other drugs used therapeutically, there is a potential for clinically significant drug interactions 39 (Michalets 1998). It appears that following metabolism, the cyclic peptide structure remains intact, with most of the metabolites being hydroxylated, N-demethylated, or both (Lemaire et al 1990, Yee and Saloman 1992). A carboxylated metabolite and a conjugated (sulfate) metabolite have also been identified (Freeman 1991). Cyclization at amino acid 1 is another mechanism of metabolite production (Christians et al 1995, 1988). More than 30 metabolites have been identified from human bile, feces, blood, and urine (Christians et al 1995, Liu et al 1995). The primary metabolites are M l , M l 7 , and M21 (Freeman 1991). More than 50% of the concentrations of M l , M9, M l 9 , M69, M49, and M4N69 are associated with the cellular components of blood, while M4N9, M l c , M l c 9 , M14N, and M I A are found mainly in plasma (Christians and Sewing 1993). One of the reasons for using whole blood as the sampling fluid in routine clinical monitoring is because CSA metabolites detectable in blood preferentially distribute into blood cells. Thus, higher metabolite concentrations can be found in whole blood rather than plasma samples (Christians and Sewing 1993). CSA metabolites, like the parent compound, are widely distributed in tissue. Distribution into blood cells is also temperature- and hematocrit-dependent. It appears that partitioning into blood cells is facilitated by hydroxylation, while N-demethylation, carboxylation, and cyclization at amino acid 1 decrease cellular affinity (Christians and Sewing 1993). The clinical significance of CSA metabolites (i.e., their immunosuppressive and toxic activities) remains unclear (Christians et al 1995, Copeland et al 1990, Yee and Saloman 1992). CSA is a low to intermediate extraction ratio drug (Freeman 1991, Yee and Saloman 1992). Thus, upon exposure to drug metabolizing enzymes, CSA is not extensively metabolized. Since there is not a large first pass effect, more CSA is systemically available. Different apparent clearance values have been reported for different transplant types. Apparent clearances 40 of 0.34 to 0.71 L/hr/kg have been reported for kidney transplant patients, 0.33 to 0.56 L/hr/kg for liver transplant patients, 0.34 to 0.79 L/hr/kg for bone marrow transplant patients, and 0.24 to 0.31 L/hr/kg for heart transplant patients (Yee and Saloman 1992). In addition, apparent clearance is lower in elderly patients and patients with decreased levels of serum triglyceride and cholesterol, as well as in patients with hepatic impairment (Kahan 1985). It is recommended that dosing intervals should be increased in the presence of elevated serum levels of bilirubin or alanine aminotransferase, but not aspartate aminotransferase, lactate dehydrogenase, or alkaline phosphatase (Yee et al 1984). 1.6.4 Excretion Less than 1% of an administered CSA dose is excreted in the urine unchanged. In addition, very little CSA, in the form of metabolites, is excreted in the urine (Yee and Saloman 1992). The metabolites M l , M l c , and M9 are found in the urine, often in higher concentrations than the parent drug (Yee and Saloman 1992). The major route of CSA excretion is via the biliary route (Freeman 1991, Yee and Saloman 1992). However, very little parent drug is found in the bile. Nearly all of the compounds detectable in the bile are metabolites, at concentrations much higher than those found in blood (Yee and Saloman 1992). It has been found that hepatic impairment reduces metabolite elimination (Kahan 1989). The terminal elimination half-life (ti/2) of CSA is considered long, although there is considerable variability in reported values. A n elimination half-life of 6.2 to 23.9 hours has been reported (Freeman 1991). 41 1.7 Cyclosporine Toxicity The majority of studies involving CSA and its associated toxicities were performed using the SIM formulation. Studies published 1995 and later, in the majority of cases, were performed with the NEO formulation. Clinical use of CSA demonstrated the potent immunosuppressive activity of the compound. Unfortunately, it was also discovered that CSA has a narrow therapeutic index with several potential toxicities associated with its use. While nephrotoxicity is the primary adverse effect, other cardiovascular, nervous system, dermatologic, hepatic, gastrointestinal, infectious (due to bacterial, fungal, and viral pathogens), hematologic, lymphatic, and hypersensitivity effects can occur (Cyclosporine AHFS monograph 1998, Kahan 1989, Mihatsch et al 1998, Shihab 1996). 1.7.1 Nephrotoxicity One of the most well known and commonly encountered toxicities of CSA is nephrotoxicity. It is also the most widely studied toxicity. Nephrotoxicity complicates the management of transplant patients. In renal transplant patients, it is difficult to differentiate between nephrotoxicity and acute rejection, and diagnosis of one or the other often involves considerable guesswork and expense. In heart, lung, and heart-lung transplant patients, who require higher CSA levels to prevent rejection, dialysis and subsequent renal transplantation are not uncommon. In both experimental models and human clinical studies, it has been well established that CSA produces dose-dependent,, acute and reversible vasoconstriction of renal arterioles (Olyaei et al 1999, Shihab 1996). CSA-induced acute renal failure may occur as early as a few weeks or 42 months following initiation of therapy (Flechner et al 1983). Prolonged cold ischemic time, advanced donor age, hypotension, perioperative surgical complications, and donor history of acute renal failure may increase the incidence and appearance of acute renal impairment (Olyaei et al 1999). The clinical signs of renal arteriolar vasoconstriction or acute renal dysfunction include reduction in glomerular filtration rate, hypertension, hyperkalemia, tubular acidosis, increased reabsorption of sodium, and oliguria (Remuzzi 1995). The adverse effects on renal hemodynamics caused by CSA are thought to be directly related to blood concentrations (Olyaei etal 1999). Despite much research in the area, the exact mechanism and the mediators involved in the alteration of renal hemodynamics by CSA are not known. Increased production of endothelin and thromboxane A2, decreased production of renal vasodilatory prostaglandins, inhibition of nitric oxide, as well as activation of the sympathetic nervous system have all been proposed as causes of renal arteriolar vasoconstriction (Auch-Schwelk et al 1994, Bobadilla et al 1994, Bunchman and Brookshire 1991, Lanese and Conger 1993, Morgan et al 1991, Scherrer et al 1990). In solid organ transplant patients treated with CSA, chronic progressive nephrotoxicity is the major long-term toxic effect (Olyaei et al 1999). The effect of hemodynamic changes encountered in acute CSA nephrotoxicity may not be the only cause of progressive CSA nephrotoxicity. Studies of biopsies from experimental models, patients with autoimmune diseases and extrarenal solid organ transplants have shown distinct and specific pathological and morphological changes characteristic of CSA-associated chronic progressive nephropathy (Bertani et al 1987, Dische et al 1988, Nussenblatt and Palestine 1986, Palestine et al 1986). Histologically, there is destruction of arterial walls, myointimal necrosis, and progressive 43 narrowing of the arterial lumen (Olyaei et al 1999). There is also tubulointerstitial fibrosis in a striped pattern, beginning in the medulla and extending to the medullary rays of the cortex (Olyaei et al 1999). It is interesting to note that unlike CSA-induced acute renal impairment, chronic progressive nephropathy is not dose-dependent (Messana et al 1995). In most cases, chronic progressive nephropathy is associated with mild to moderate renal dysfunction. Although it is difficult to distinguish and separate chronic rejection and chronic progressive nephropathy induced by CSA in renal transplant patients, the morphological changes associated with rejection involve mostly large blood vessels, while chronic progressive nephropathy affects mainly arterioles (Mihatsch et al 1995). It is also conceivable that a given patient could have both conditions present simultaneously. Renal function, as estimated by serum creatinine, is the most widely used surrogate marker of the presence or absence of chronic CSA nephropathy. 1.7.2 Cardiovascular Toxicity A n increase in arterial blood pressure is frequently observed in CSA-treated patients (Bennett and Porter 1988). This increase leads to hypertension, which in turn can result in initiation of antihypertensive therapy. The mechanism of this increase in blood pressure is not currently known. It has been shown that angiotensin-converting enzyme inhibitors have little effect in reducing blood pressure (Curtis et al 1988), suggesting that the renin-angiotensin system is not involved. Defects in endothelium-dependent relaxation and in response of vascular smooth muscle to vasodilators have been associated with CSA-induced hypertension in experimental models (Roullet et al 1994). 44 1.7.3 Neurotoxicity Even though CSA is not believed to cross the intact blood-brain barrier, neurologic side effects occur in approximately 20 percent of kidney (Kahan et al 1987) and liver (de Groen et al 1987) transplant patients, resulting in syndromes of tremor, burning palmar and plantar paresthesias, headache, flushing, depression, confusion, and somnolence. CSA neurotoxicity is also an important cause of morbidity and mortality in bone marrow transplant patients (Shah 1999). Seizures of new onset may be triggered by hypocholesterolemia, hypertension, intravenous methylprednisolone therapy, hypomagnesemia, infection, hemorrhage, or cerebral infarction (Kahan 1989). Visual disorders, paresis, disorientation, and coma improve when CSA treatment is discontinued, but reoccur when C S A treatment is initiated again (de Groen et al 1987). As is the case with the other toxicities, the mechanism of neurotoxicity is not definitively known. On computerized axial tomography, the presence of neurotoxicity is associated with white matter hypodensity, which suggests increased water content in the brain (Kahan 1989). It has also been proposed that CSA has a direct cytotoxic effect on brain capillary endothelial cells, and the inhibition of P-glycoprotein by CSA may be partly involved in the occurrence of CSA neurotoxicity (Kochi et al 1999). 1.7.4 Dermatologic Toxicity Hypertrichosis of the face, arms, shoulders, and back develops in at least 50 percent of renal transplant patients receiving CSA as part of their immunosuppressive protocol (Kahan et al 1987). Coarsening of facial features has also been reported (Reznik et al 1987). CSA also increases the number of gingival fibroblasts, as well as collagen production by the gingival fibroblasts, producing primarily anterior gingival hyperplasia (Kahan 1989). 45 1.7.5 Hepatotoxicity CSA treatment nearly doubles the incidence of cholestasis with hyperbilirubinemia and elevation of serum levels of aminotransferases (especially alanine aminotransferase) in renal transplant patients (Pickrell et al 1988). Chronic hepatic dysfunction is associated with an increased incidence of cholelithiasis and choledocholithiasis (Kahan 1989). Experimental models of hepatotoxicity reveal centrilobular fatty changes and hepatocyte necrosis, dilated endoplasmic reticulum, and an increased number of autophagic vacuoles (Kahan 1989). In addition, it has been shown that CSA alters calcium fluxes across hepatocyte cell membranes (in vitro), elevates serum bile acids, and decreases bile flow (Kahan 1989). 1.7.6 Gastrointestinal Toxicity When the Sandimmune® formulation of CSA was still in widespread clinical use, anorexia, bloating, nausea, and/or vomiting were frequently reported following ingestion of Sandimmune® oral suspension (Kahan 1989). CSA has no direct toxic effects on the structure or function of gastrointestinal mucosa. Even in the case of oral CSA overdose, only a mild syndrome of hypertension, dysesthesias, flushing, and stomach upset is encountered (Kahan 1989). 1.7.7 Hematologic Toxicity There are reports of increased incidences of thrombosis of arterial and venous limbs of renal allografts (Kahan et al 1985, Najarian et al 1985), as well as thrombosis of systemic veins (Varenterghem et al 1985). This increased incidence of thrombosis may be due to increases in 46 platelet aggregation, thromboxane A2 release, thromboplastin generation, and factor VII activity (Carlesen and Prydz 1987, Grace et al 1987). Another common hematologic-related problem is atherogenic changes in serum lipids, with increased cholesterol, apolipoprotein B, and triglyceride concentrations. This increase in serum lipid levels often requires lipid-lowering therapy, and may also exacerbate the adverse effects of CSA on vascular endothelium (Ellis et al 1986) . 1.7.8 Lymphatic And Related Toxicity The risk of non-Hodgkin lymphoma in organ transplant patients receiving CSA immunosuppression is high (Opelz and Henderson 1993). It appears that the greatest risk is during the first year post-transplant, when target CSA concentrations are higher. In the first year post-transplant, kidney and heart transplant patients had occurrence rates of 20 and 120 times higher than the general population, respectively (Opelz and Henderson 1993). The higher incidence in heart transplant patients is thought to be due to the higher amounts of CSA that heart transplant patients are exposed to when compared to kidney transplant patients (Land 1987) . Proliferation of B cells that escape T-cell control, driven by Epstein-Barr virus, is believed to be the primary mechanism of lymphoma development (York and Qualtiere 1990). Immunosuppression is also associated with an increased incidence of certain other cancers (Penn 1986). It is important to point out that these cancers occur not only in transplant patients, but also in patients receiving immunosuppressants for autoimmune and chronic inflammatory disorders, patients who are immunosuppressed as a side effect of chemotherapy, patients with primary or genetically determined immunodeficiency disorders, patients with AIDS, and patients with chronic renal failure who are on dialysis (Penn 1986). 47 Patients receiving transplants have a 3-fold increase in various cancers in when compared with age-matched controls (Kinlen 1985). Cancers commonly encountered in the general population, including cancer of the lung, prostate, colon, rectum, female breast, and invasive cancer of the uterine cervix were not increased in incidence in transplant patients (Penn 1986). However, other cancers, such as skin, lip, Kaposi's sarcoma, other uterine, vulva, perineum, scrotum, penis, perianal skin, anus, and hepatobiliary, showed an increased incidence in transplant patients when compared to the general population (Penn 1986). While earlier data showed that the incidence of lung cancer was not increased in transplant patients, more recent data have shown that there is a high incidence of lung cancer following heart transplantation, particularly in patients who were smokers prior to transplantation (Goldstein et al 1995, 1996, Curtil etal 1997). 1.7.9 Hypersensitivity And Other Effects It has long been known that hypersensitivity to the older olive oil based and castor oil based drug vehicles manifests as a spectrum of clinical symptoms, ranging from mild flushing and hypertension to hypotension, crushing chest pain, dyspnea, and respiratory distress after oral, and especially after intravenous bolus administration of CSA (Kahan 1985). CSA occasionally produces hyperglycemia not accompanied by ketosis, ketonuria, or altered responses of endogenous insulin and C peptide to intravenous glucose administration (Gunnarsson et al 1984). Since CSA has no effect on the number or binding affinity of cellular insulin receptors, the hyperglycemia may be due to impaired hepatic synthesis of glycogen rather than hormone resistance (Betschart et al 1988). CSA increases serum prolactin but decreases testosterone levels, which leads to gynecomastia in men (Kahan 1989). This change has also 48 caused impaired spermatogenesis or sperm maturation in experimental models (Raj fer et al 1987). CSA, when used as monotherapy or as polytherapy, can potentially affect skeletal muscle (Briel and Chariot 1999, Rush 1990). Myopathy manifesting by myalgia, muscle weakness, and plasma creatine kinase elevation have all been attributed to CSA use (Briel and Chariot 1999). Because of a potential drug interaction between CSA and hydroxymethylglutaryl coenzyme A reductase inhibitors, colchicine, and pyrazinamide, the frequency of muscular complications may be increased (Briel and Chariot 1999). This is especially a concern with hydroxymethylglutaryl coenzyme A reductase inhibitors, as these agents are commonly used to treat hyperlipidemia in transplant patients. The mechanism of myotoxicity is not known, but some clinical and experimental findings have provided some clues: 1) some patients with myopathy attributed to CSA have mitochondrial abnormalities and lipid droplet accumulation in skeletal muscle (Fernandez-Sola et al 1990, Larner et al 1994); and 2) CSA decreases mitochondrial respiration in rat skeletal muscle (Hokanson et al 1995, Mercier et al 1995). Therapy is often required to prevent bone loss following institution of CSA therapy. In cardiac transplant patients, bone loss and fractures are frequent complications, particularly during the first year post-transplantation when higher CSA levels are targeted (Boncimino et al 1999). Magnesium depletion also frequently occurs with CSA therapy, and this can adversely affect many phases of skeletal metabolism. Magnesium depletion has also been implicated as a risk factor in several forms of osteoporosis (Boncimino et al 1999). However, recent evidence has shown that cardiac transplant patients with low serum magnesium levels had significantly lower rates of bone loss, lower serum parathyroid hormone concentrations, and lower bone 49 turnover (Boncimino et al 1999). The exact mechanism of this protective effect is not currently known. Developing A Limited Sampling Strategy For Cyclosporine Area Under The Curve Monitoring In Lung Transplant Patients 1.1 Introduction The development of immunosuppressive agents has revolutionized the practice of organ transplantation and brought about significant improvements in survival rates (Trulock 1997, Tsunoda and Aweeka 1996). CSA is the mainstay of immunosuppressive therapy for solid organ transplants (Trulock 1997, Tsunoda and Aweeka 1996). Early clinical studies have demonstrated the immunosuppressive efficacy of CSA (Calne et al 1978, Powles et al 1978, Starzl et al 1981). Since its introduction into clinical practice in 1983, CSA has been shown to increase survival in liver (Starzl et al 1983), heart (Macoviak et al 1985, Oyer et al 1983), heart-lung (Macoviak et al 1985), and renal (The Canadian Multicentre Transplant Group 1986) transplant patients. Despite its widespread use, however, the optimum strategy for the therapeutic drug monitoring of this drug remains undetermined. Measurement of CSA blood concentrations is essential in guiding optimal dosages to minimize its toxicity and promote its efficacy (Yee and Saloman 1992). It has been demonstrated that AUCo-T pharmacokinetic monitoring of C S A is superior to trough level monitoring (Grevel et al 1989). However, the A U C method requires numerous serial blood samples (i.e., typically 7 or more) and, thus, is expensive, cumbersome, and impractical for routine clinical use. Consequently, the usual practice consists of sampling trough levels only. Patients with identical trough levels, however, may have significantly different systemic exposure to CSA as measured by A U C (Yee and Saloman 1992). For example, two patients 50 could have identical trough levels of 200 ng/mL. However, i f concentration-time data were gathered over a dosing interval for these two patients, the resulting concentration-time curves would not be superimposable. If A U C values were calculated for the dosing interval, the values obtained would not be identical. Thus, despite similar trough levels, the patients will not have identical A U C , and therefore, exposure to CSA. Recent studies conducted primarily in kidney transplant patients (Johnston et al 1990, Grevel and Kahan 1991a, Meyer et al 1991, Meyer et al 1993, Serino et al 1994, Foradori et al 1995, Johnston et al 1996, Serafinowicz et al 1996, Amante and Kahan 1996, Keown et al 1996, Cooney et al 1996, Gaspari et al 1997, Primmett et al 1998, Meier-Kriesche et al 1998, Lemire et al 1998, Marsh 1999) and a small number of heart (Johnston et al 1990) and liver (Cooney et al 1996) transplant patients indicate that a limited number of blood samples (i.e., a limited sampling strategy) provides a reliable alternative to A U C monitoring. Some studies involved patients taking the older CSA formulation, SIM (Johnston et al 1990, Grevel and Kahan 1991a, Meyer et al 1991, Meyer et al 1993, Serino et al 1994), while others involved patients taking the new formulation, NEO (Foradori et al 1995, Johnston et al 1996, Serafinowicz et al 1996, Amante and Kahan 1996, Keown et al 1996, Cooney et al 1996, Gaspari et al 1997, Primmett et al 1998, Meier-Kriesche et al 1998, Lemire et al 1998, Marsh 1999). CSA absorption appears to be enhanced and inter- and intra-patient variability reduced with NEO, compared with SIM (Kovarik et al 1994a, 1994b, 1994c). As such, a limited sampling strategy may be an even better predictor for NEO than for SIM. However, studies involving these abbreviated monitoring strategies for NEO are nonexistent in the lung transplant population. Several studies have described CSA pharmacokinetic parameters in renal transplant patients (Ptachcinski et al 1985, Awni et al 1989, Awni et al 1990, Honcharik et al 1991, Kahan 51 et al 1992, Cooney et al 1994, Kovarik et al 1994, Sketris et al 1994, Kahan et al 1995, Kahan et al 1996, Amante et al 1997). Studies involving healthy subjects (Lindholm et al 1990, Gupta et al 1990, Drewe et al 1992, Gardier et al 1993, Kovarik et al 1994b, Mueller et al 1994), heart-lung transplant patients (Tan et al 1993, Tsang et al 1994, Eadon et al 1995), and liver transplant patients (Trull et al 1995, Tredger 1995, van Mourik et al 1999, Reynaud-Gaubert et al 1997) have also been published. However, relatively few studies describing CSA pharmacokinetics in lung transplant patients are available (Eadon et al 1995, Reynaud-Gaubert et al 1997, Kesten et al 1998a, 1998b). 1.2 Background Studies Trough level monitoring (TLM) is the traditional method of monitoring CSA therapy. Some advantages of T L M are that it is simple, practical for routine clinical use in both inpatients and outpatients, provides qualitative insight into a patient's absorption, and is the most widely studied method of drug monitoring. However, its disadvantages are that it is not a good indicator of total drug exposure, is not a good predictor of outcome since single concentration-effect relationships are weak, and provides only a rough estimation of absorption and/or elimination and no information on other pharmacokinetic parameters (Keown et al 1998, Lindholm and Sawe 1995). The therapeutic ranges of many drugs have been questioned recently, mainly because these ranges were derived from studies that used empirical observations and approximations (usually in studies with small sample sizes) rather than population studies with appropriate statistical analysis (Morris 1997). There is similar controversy with the therapeutic range of CSA, as not all studies show a strong correlation between CSA levels and efficacy or toxicity 52 (Morris 1997). Due to wide inter-patient variability in C S A pharmacokinetics and the fact that the upper and lower concentration values are, in the majority of cases, arbitrary values within which "efficacy" was achieved with a comparatively low incidence of "toxicity", there are patients who have periods of rejection despite a therapeutic trough level. Similarly, there are patients who exhibit toxicity at therapeutic levels. The therapeutic range can be considered analogous to a confidence interval in that it specifies a range of possible concentrations that are associated with efficacy (e.g., no overt rejection in the case of CSA). Using a simplified and fictional example, we could say that a given therapeutic range for CSA is an 80% confidence interval. Thus, in 80% of cases, efficacy will be achieved, but in 20% of cases, rejection or toxicity may occur (Dumont and Ensom 2000). The therapeutic range method of T D M has been criticized (Morris 1997, Ensom et al 1998, Holford 1999). Because there is a range of possible target values, there is uncertainty regarding what initial dose to prescribe (Holford 1999) and initial dosing is often a "hit and miss" process. The therapeutic range concept also implies that all concentrations within the range are equally desirable. This further implies that there is a 3-step concentration-effect relationship (Holford 1999). Thus, rejection will be encountered with sub-therapeutic levels and toxicity will occur with supra-therapeutic levels. If, for example, a level comes back a few concentration values above "therapeutic", the dose may be immediately lowered to prevent toxicity when in fact, that level is probably efficacious and the patient is experiencing no toxicity. Because of the problems with T L M of CSA therapy, researchers looked for other methods of monitoring. It is widely known that A U C gives an indication of extent of exposure to a drug. Proper calculation of A U C requires administration of a dose, followed by blood 53 collection according to an intensive sampling strategy. Concentration values obtained are used to calculate A U C , usually by the trapezoidal method. A study published in 1989 by Grevel and colleagues (Grevel et al 1989) is one of the most commonly cited studies demonstrating that A U C monitoring is better than T L M . Patients involved in the study had previously taken part in a pre-transplant pharmacokinetic study to determine individualized oral dosing. Serum concentrations of CSA were measured by polyclonal radioimmunoassay. Correlation analysis was based on 71 observations in 36 renal transplant patients 0-36 months following transplantation. Performance analysis was based on 26 observations in 14 different patients who were 0 - 4 months post-transplant. The conclusion of the study was based on two observations. First, A U C (r = 0.381, p = 0.001 for dose in mg; r = 0.538, p = 0.0001 for dose in mg/kg) but not trough level (r = 0.154, p = 0.20 for dose in mg; r = 0.136, p = 0.26 for dose in mg/kg) was significantly, albeit poorly, correlated with dose. Second, after adjusting the oral dosage to achieve the target CSA concentration, the absolute deviation of the A U C prediction was significantly smaller than the absolute deviation of the trough level prediction (14.6 ± 13.6% for A U C versus 36.0 ± 27.5% for TL, p = 0.0005; mean ± standard deviation) (Grevel et al 1989). A subsequent study (Grevel and Kahan 1991b) in renal transplant patients showed that CSA inter-patient variability could be counterbalanced by dosage individualization through the use of A U C monitoring. Some advantages of A U C monitoring are that it is the most precise indicator of drug exposure, can characterize abnormal absorption patterns, appears to be a predictor of clinical outcomes in the majority of studies (Table 3) (Kahan et al 1986, Savoldi and Kahan 1986, Kasiske et al 1988, Grevel et al 1991, Lindholm et al 1993, Schroeder et al 1994, Barone et al 1996, Bowles et al 1996), generates a concentration-time profile, allows calculation of oral 54 Table 3. Clinical studies evaluating area under the curve monitoring of cyclosporine A and outcome. Reference Patient Group Sample Matrix/ Analysis Method Indicator(s) of Outcome Findings Kahan et al 1986 Renal Serum/P RIA NT CSA A U C : dose was significantly higher in patients with NT. Savoldi and Kahan 1986 Renal Serum/RIA RJ, NT, Hepatotoxicity, Infection For IV dosing, RJ was associated with a significantly lower A U C with twice-daily but not once-daily dosing; NT was associated with a significantly higher A U C with once-daily and twice-daily dosing; Hepatotoxicity was associated with a significantly lower A U C with twice-daily dosing only; With infection, there was no significant difference in A U C . For PO dosing, infection was associated with a significantly higher A U C with once-daily dosing. There was no significant difference in A U C with the other indicators of outcome. Kasiske et al 1988 Renal WB/HPLC RJ, NT, Hypertension No correlation was found between A U C and NT. Grevel et al 1991 Renal Serum/P RIA, WB/P FPIA RJ A significant difference in A U C was found only with measurement by P RIA. Lindholm et al 1993 Renal W B / M RIA RJ, Graft Loss, Renal Function AUC values were used to calculate pharmacokinetic parameters. A significantly higher clearance was found with RJ (IV dosing). With PO dosing, those with RJ had a significantly lower average steady-state concentration and a significantly higher oral clearance. Significantly lower average steady-state concentration and bioavailability and significantly higher oral clearance were found with graft loss. AUC-monitored patients had significantly better renal function (as assessed by creatinine clearance) at 6 months post-transplant, with trends for better renal function at 1 and 2 years post-transplant 55 Schroeder etal 1994 Barone et al 1996 Renal Renal WB/HPLC, WB/P FPIA Bowles et al 1996 Mahalati etal 1999 Renal Renal WB/HPLC, WB/FPIA Length of Stay, Number of Readmissions, RJ WB/FPIA W B / M R I A Grant etal 1999 Liver Time to Death, Time to Graft Loss, Time to First RJ Episode (historical control). Retrospective. Acute RJ resulted in a significantly greater number of readmissions and length of stay for re-hospitalization the first year post-transplant. A U C was significantly lower in both acute and chronic rejection. RJ, NT No correlation was found between A U C and any indicators of outcome. RJ WB/FPIA or M RIA, or EMIT RJ No significant difference in A U C was found between nephrotoxic and non-nephrotoxic groups. There was a significant difference in A U C between patients with and without acute rejection. However, there was no significant difference in trough levels between patients with and without acute rejection. There was a significant correlation between A U C and rejection, but not between trough level and acute rejection. | ; | , Abbreviations: A U C = area under the concentration versus time curve; P RIA = polyclonal radioimmunoassay; NT = nephrotoxicity; CSA = cyclosporine A ; RIA = radioimmunoassay; WB = whole blood; HPLC = high performance liquid chromatography; RJ = rejection; M RIA = monoclonal radioimmunoassay; PO = oral; FPIA = fluorescence polarization immunoassay; EMIT = enzyme multiplied immunotechnique pharmacokinetic parameters, and reduces the problems associated with lab errors and single concentrations. Despite its appealing potential advantages, the major disadvantage of A U C monitoring is its inherent need for multiple blood samples. The increased number of samples required, when compared to T L M , makes A U C monitoring impractical for routine clinical use, expensive in the short-term due to increased sample collection, analysis and interpretation more 56 of results, and inconvenient for patients, especially those in an outpatient setting (Keown et al 1998, Lindholm and Sawe 1995). A U C monitoring showed promise, but was not possible to use in routine clinical practice. A search for methods to accurately approximate A U C without the intensive sampling led to the development of limited sampling strategies (LSS). Briefly, an LSS is developed by full A U C calculations in a sample population. Stepwise multiple regression analysis is then performed on the concentration-time points sampled. The points that do not correlate well with A U C are removed until a regression equation consisting of two or three concentration-time points is left. A number of limited sampling strategies have been developed in a variety of transplant patients (Johnston et al 1990, Grevel and Kahan 1991a, Meyer etal 1991, Meyer et al 1993, Serino et al 1994, Foradori et al 1995, Johnston et al 1996, Serafmowicz et al 1996, Amante and Kahan 1996, Keown et al 1996, Cooney et al 1996, Gaspari et al 1997, Primmett et al 1998, Meier-Kriesche et al 1998, Lemire et al 1998, Marsh 1999). (Table 4). Table 4. Summary of clinical studies that have developed a limited sampling strategy for a transplant population. Reference Johnston et al 1990 Patient Group Grevel and Kahan 1991 Meyer etal 1991 Meyer et al 1993 Renal and Heart Renal Renal Renal Formulati on SIM SIM SIM AUC Interval 12 hour 24 hour Serino etal 1994 Renal SIM SIM 24 hour Equation (C = concentration) A U C = 4 . 3 x C 3 . 5 + 5 . 5 x C 8 + 3 . 1 x C i 0 -3 3 3 A U C = 2 .91 x C 2 + 5 . 9 5 x C 6 + 1 1 . 6 8 x C i 4 + 153 24 hour 12 hour A U C = 2 . 0 x C 2 + 1 0 . 2 x C 6 + 0 .2 A U C = 8 . 6 x C 2 4 + 1 . 4 x C 2 + 6 . 2 x C 6 + 1.57 A U C = 2 . 1 1 x C 2 + 3 . 2 3 x C 4 + 5 . 6 3 x C 9 + 2 5 0 57 Foradori etal 1 9 9 5 Renal NEO 12 hour A U C = 0 . 6 8 1 x C i + 1 . 8 5 9 x C 2 5 + 3 . 4 1 1 x C 5 + 7 9 1 . 7 4 Johnston etal 1 9 9 6 Renal NEO 12 hour A U C = 1 . 9 6 x C 2 + 1 1 . 5 x C 8 + 3 5 5 . 2 Serafinowicz etal 1 9 9 6 Renal NEO 12 hour A U C = 9 . 1 3 1 x C 0 + 0 . 7 8 4 x C i + 2 . 6 1 7 x C 2 + 1 9 3 . 5 6 1 Amante et al 1 9 9 6 Renal NEO 12 hour A U C = 2 . 4 x C 2 + 7 . 7 x C 6 + 1 9 5 . 8 A U C = 1 . 5 x C i . 5 + 4 . 1 x C 4 + 5 . 6 x C 9 + 105 .5 Keown et al 1 9 9 6 Renal NEO 12 hour A U C = 1 . 8 4 x C 0 + 4 . 3 9 x C 2 + 3 1 2 . 6 6 Cooney et al 1 9 9 6 Liver (pediatric) NEO 8 hour A U C = 1 0 . 1 9 x C 0 + 4 . 4 7 x C 3 + 7 4 9 . 7 A U C = 1 5 . 1 3 x C 0 + 6 . 7 8 x C 3 - 2 . 9 4 x C 3 5 + 3 0 1 . 3 A U C = 3 . 6 9 x C 3 + 1 3 . 7 6 x C i 2 + 4 0 2 . 8 A U C = 5 . 3 2 x C 0 + 3 . 8 4 x C 3 + 1 0 . 8 3 x C i 2 - 3 3 . 6 Gaspari etal 1 9 9 7 Renal NEO 12hour A U C = 5 . 1 8 9 x C 0 1 . 2 6 7 C , + 4 . 1 5 0 x C 3 + 1 3 5 . 0 7 9 Primmett e r a / 1 9 9 8 Renal NEO 4 hour A U C = 1 2 . 3 4 x C 0 + 2 . 4 8 x C 2 + 4 4 1 . 4 2 A U C = 9 . 5 5 x C 0 + 0 . 9 6 x C i + 2 . 0 5 x C 2 + 1 1 2 . 0 7 Meier-Kriesche et al 1 9 9 8 Renal (pediatric and adult) NEO 8 hour A U C = 1 . 8 4 x C 2 + 4 . 3 9 x C 4 + 1 2 9 Lemire etal 1 9 9 8 Renal NEO 12 hour A U C = 1 . 0 2 3 x C i + 1 3 . 1 0 x C 6 + 2 4 2 Marsh 1 9 9 9 Renal NEO 12 hour A U C = 6 x ( C 0 + C 4 ) Earlier studies involved the Sandimmune® formulation, and 3 of these 5 studies used once daily dosing. Because the majority of centers dose cyclosporine twice daily, and because once daily dosing is no longer used, these LSS may no longer be applicable. LSS are generally considered superior to T L M because trough concentrations do not provide an adequate estimate of exposure to CSA. Also, equations for some LSS were chosen solely on the basis of a high coefficient of determination value, rather than looking at other 58 criteria such as prediction error (Gaspari et al 1998). LSS developed using the original Sandimmune® formulation are likely no longer applicable due to the differences between Neoral® and Sandimmune®. In addition, LSS are primarily available for renal allografts. Because of pharmacokinetic differences between transplant types, these LSS are not applicable to, and warrant further study in, other transplant types. LSS may be center-specific, as studies that have evaluated LSS developed in other centers have found that they do not perform as well in terms of percent prediction error for A U C (Gaspari et al 1997, 1998, 1993). As well, not all LSS have worked (Gaspari et al 1993). 1.3 Significance Of The Research Project To our knowledge, this is the first study to develop a limited sampling strategy for use in the therapeutic drug monitoring of C S A in lung transplant patients. A previously published report by Trull and colleagues (Trull et al 1999) stated in the abstract that they had utilized a limited sampling strategy for their population of lung transplant patients, although this limited sampling strategy was not mentioned in the body of the paper. Also, this "limited sampling strategy" was not an LSS like the ones reported in this thesis and previously published papers. They did not use an equation obtained from multiple regression analysis of concentration-time data to predict total AUCn- T. Instead, they calculated an AUCn-6 and performed repeated measures analysis of variance on the AUCo-6 obtained from pharmacokinetic studies performed at the end of weeks 1 to 4 and at the end of weeks 13, 26, 39, and 52 (Trull et al 1999). Thus, because they did not develop or even utilize an LSS developed according to previously published methods, our LSS developed specifically for lung transplant patients is, indeed, novel. The information generated from the proposed study could prove to be crucial for effective 59 management of lung transplant patients. Using an individualized limited sampling strategy for CSA developed uniquely for lung transplant patients is expected to be a convenient and accurate method to minimize toxicity and maximize efficacy of long-term immunosuppressant therapy in this patient subpopulation. Calculation of pharmacokinetic parameters in lung transplant patients taking the Neoral® formulation of CSA will add to the existing, limited data. Research Hypothesis A limited sampling strategy for C S A (Neoral®) is a precise and unbiased predictor of A U C in lung transplant patients. Objectives The objectives of this study are: 1) To define the optimal limited sampling strategy for CSA monitoring in lung transplant patients. 2) To compare the predictive performance of this optimal limited sampling strategy (with other models derived from previous studies) in lung transplant patients. 3) To determine the mean steady-state pharmacokinetic parameters of CSA (Neoral®) in lung transplant patients. Rationale Of all organs, the lung is most susceptible to rejection (Trulock 1997). Although CSA has been largely responsible for the success of lung and other organ transplantation, it is not 60 without problems. CSA has a narrow therapeutic window, thus requiring measurement of blood concentrations to maintain efficacy and minimize toxicity (Yee and Saloman 1992). The uniqueness of CSA dosing and monitoring in lung transplant patients is further illustrated by the large proportion (approximately 1/3 in the V H H S C lung transplant program) of them who have cystic fibrosis. Compared to patients without cystic fibrosis, those with cystic fibrosis are known to have greater CSA clearance, more erratic absorption, and more variable pharmacokinetics (Tan et al 1993). In the lung transplant population, the newer microemulsion formulation of CSA, NEO, demonstrates greater bioavailability and reduced variability in C S A blood concentrations compared with older CSA formulations (Kesten et al 1998b, Zaldonis et al 1998, Mikhail et al 1997, 1998, Wilczek et al 1997, Svendsen et al 1995, 1996, Girault et al 1995). In these studies, the improved pharmacokinetic characteristics of N E O were demonstrated via calculations of A U C obtained by serial blood sampling (i.e., typically 7 or more concentrations). In clinical practice, collection of multiple serial samples to monitor C S A (NEO) therapy is not possible and collection of only trough samples is the routine. It has been demonstrated that A U C monitoring of C S A is superior to trough level monitoring (Grevel et al 1989). Patients with identical trough levels, however, may have significantly different systemic exposure to CSA as measured by A U C (Yee and Saloman 1992). Suboptimal exposure to CSA can lead to acute rejection. Furthermore, multiple acute rejection episodes lead to a greater risk of chronic rejection, ultimately ending in graft failure and death (Trulock 1997). A limited sampling strategy has the advantage of minimizing the number of blood samples required by A U C monitoring and at the same time, providing relevant pharmacokinetic 61 data to guide optimal CSA dosing. Information on the utility of limited sampling strategies for CSA, either Neoral® or Sandimmune®, in lung transplant patients, however, is glaringly lacking. To our knowledge, no published reports are available in this patient subpopulation. A limited sampling strategy has served as a good predictor for C S A (older formulations and NEO) A U C in non-lung transplant patients (Johnston et al 1990, Grevel and Kahan 1991a, Meyer et al 1991, Meyer et al 1993, Serino et al 1994, Foradori et al 1995, Johnston et al 1996, Serafinowicz et al 1996, Amante and Kahan 1996, Keown et al 1996, Cooney et al 1996, Gaspari et al 1997, Primmett et al 1998, Meier-Kriesche et al 1998, Lemire et al 1998, Marsh 1999). Because of reduced inter- and intra-patient variability in C S A blood concentrations associated with NEO administration to lung transplant patients, we expect such a limited sampling strategy to be an even better predictor for NEO than for older CSA formulations. Furthermore, the inability to apply some limited sampling strategies for CSA (derived from kidney transplant patients from a particular institution to another institution) (Gaspari et al 1997, 1998, 1993) underscores the need to develop limited sampling strategies unique to the lung transplant patient subpopulation. 62 CHAPTER 2 MATERIALS AND METHODS 63 2.1 Experimental Design This was an open-label, single center, pilot clinical study. 2.2 Clinical Research Subjects Stable, adult lung transplant patients (both single and double lung graft recipients) were recruited from the outpatient solid organ transplant clinic, Vancouver Hospital and Health Sciences Centre, Vancouver, British Columbia, Canada. 2.3 Clinical Research Subject Inclusion Criteria 1) Lung transplant patients who are on a steady-state dosage of CSA (Neoral®) (Attainment of steady state was assumed when patients have taken CSA for at least 2 weeks without a dosage adjustment or change in a concurrent medication that can affect CSA metabolism). 2) Patients over 16 years of age, in accordance with a predetermined arbitrary limit between adult and pediatric age. 3) Patients able to provide informed consent 2.4 Clinical Research Subject Exclusion Criteria 1) Patients refusing or unable to provide informed consent 2) Patients younger than 16 years of age 3) Patients whose CSA (Neoral®) therapy was not at steady state, as defined previously in section 2.3, Clinical Research Subject Inclusion Criteria. 64 2.5 Clinical Research Study Protocol Subjects reported to the research clinic at the BC Transplant Society office (Vancouver, BC, Canada) prior to their morning dose of CSA (NEO). In order to have consistent start times for all patients, the start time was 0700 for all patients. Prior to arrival for the study visit, all patients were interviewed by phone regarding medication history and administration time. If a given patient did not normally take CSA at 0700, they were asked to switch to this time 2 weeks prior to the study visit. In addition, all relevant medical chart data, including date and type of transplant, serum creatinine level, demographic information, and transplant and non-transplant related medical conditions were gathered prior to the study visit. Prior to administration of their morning CSA dose, an indwelling intravenous catheter (i.e., a "butterfly") was placed in a forearm vein i f they did not have a central line or i f they preferred a "butterfly" over multiple venipunctures. Ethyl chloride spray U.S.P. (Xenex Laboratories Inc., Coquitlam, BC, Canada) was used as a local anesthetic prior to venipuncture. A l l patients, with the exception of 1, chose the intravenous catheter over multiple venipunctures. Depending on the patient, 1 of 2 types of catheter devices were used. One was an Insyte™ 22 gauge x 1 inch intravenous catheter (Becton Dickinson Infusion Therapy Systems Inc., Sandy, Utah, USA). The other was a shorter unit, which was an Argyle® intermittent infusion plug (Sherwood Medical, St. Louis, Missouri, USA). Routine clinical measurements (blood pressure, pulse, height, and weight) were also taken. A l l patients took the morning CSA dose in a fasted state. This likely does not represent the normal clinical situation, but was done for the sake of consistency, although approximately 25% of the patients who participated did not eat breakfast at all. Patients were allowed to eat normally and move about following collection of the 0-hour blood sample, routine clinical measurements, and ingestion of the dose. Because patients were allowed to eat normally immediately following 65 dosage ingestion i f desired, any benefits of fasting on absorption were negated and likely made our protocol more closely resemble the normal clinical situation. Blood samples (approximately 3 mL each) were collected at 0, 1, 2, 3, 4, 5, 6, 8, 9, 10, and 12 hours following CSA administration. A Vacutainer™ 22 gauge x 1 inch needle (Becton Dickinson and Company, Franklin Lakes, New Jersey, USA) and a Vacutainer® holder (Becton Dickinson and Company, Franklin Lakes, New Jersey, USA) were used for blood withdrawal. A l l samples were collected into Vacutainer® (Becton Dickinson and Company, Franklin Lakes, New Jersey, USA) collection tubes. These collection tubes contained 0.057 mL of 0.34 molar potassium ethylene diamine tetraacetic acid, had a draw volume of 5 mL, and were equipped with a Hemogard™ closure. In cases of a difficult draw, a Monoject® 25 guage x 5/8 inch 3 cc luer lock syringe with needle was used to withdraw the sample from either the forearm vein (in 1 case) or the collection device. The blood was then transferred from the syringe to a collection tube. Immediately after sample collection, the collection tubes were gently inverted several times, and kept refrigerated until transported to the laboratory. The collection device was then flushed with 1 mL of sodium chloride injection U.S.P. 0.9%, followed by 0.5 mL of heparin lock flush solution U.S.P. 100 U.S.P. units per mL. The heparin used in the lock flush solution was of porcine intestine mucosal origin. Prior to collection of the next sample, approximately 1 mL of blood was collected into a collection tube as described previously in order to ensure that there was no flush solution collected into the sample collection tubes. Patients ingested other medications in their medication regimen as normal, in order to simulate the normal clinical situation. In all cases, other medications were taken along with the morning CSA dose. 66 2.6 Sample Analysis Whole blood CSA concentrations were analyzed by fluorescence polarization immunoassay using a specific monoclonal antibody kit (TDx®; Abbott Inc., Abbott Park, Illinois, USA) in the central hospital laboratory located in the Department of Pathology and Laboratory Medicine at VHHSC. A typical sample run on the FPIA machine consists of a treated sample containing fluorescein tagged CSA, CSA from the patient sample, and a monoclonal antibody specific for the parent CSA molecule. When low levels of CSA are present in the patient sample, more of the monoclonal antibody is bound to the tagged CSA molecules. This binding fixes the tagged CSA molecule, preventing it from rotating freely when exposed to polarized light. Thus, with low levels of CSA in the patient sample, there are high levels of polarization, which the FPIA machine detects and uses to quantitate CSA present in the patient sample through the use of a standard curve. The opposite is true when the patient sample contains high levels of CSA. 2.6.1 Reagents And Standards 2.6.1.1 Reagent Solutions The CSA monoclonal kit consists of a 3-pot reagent pack labeled as *S, *T, or *P. *S is less than 25% CSA antiserum (mouse monoclonal) in buffer with protein stabilizer and sodium azide as preservative. *T is less than 0.01% C S A fluorescein tracer in buffer containing surfactant, protein stabilizer, and sodium azide as preservative. *P is pretreatment solution containing surfactant in buffer, with sodium azide as preservative. The reagent packs were stored at 2 to 8 °C and are stable until expiry. Prior to use, the reagent was mixed gently, the cap was removed, and all bubbles were removed with an applicator stick. The reagents did not need 67 to be warmed to room temperature before use. Accessories to the reagents include whole blood precipitation reagent/probe wash, which is zinc sulfate solution in methanol and ethylene glycol and solubilization reagent, which contains surfactants in water with sodium azide as preservative. The accessories were stored at room temperature, and no additional preparation was required prior to use. 2.6.1.2 Calibrator Solutions There are 6 CSA monoclonal whole blood calibrators, and they were stored in 6 vials labeled A through F. The concentrations of the 6 standards, with the lowest concentration stored in vial A and the highest concentration stored in vial F, were as follows: 0, 100, 250, 500, 1000, and 1500 ng per mL. These 6 concentrations comprised the 6-point calibration curve used for quantitation of patient samples. Opened and unopened calibrator packs were stored at 2 to 8 °C, and were stable until expiry. A l l calibrators were mixed gently before use. 2.6.2 Sample Preparation A pretreatment step was performed on each matrix to be measured (calibrators, controls, and patient samples) prior to testing. The purpose of the pretreatment step was to minimize interference from endogenous protein-bound fluorescent compounds. The pretreatment consisted of addition of solubilization reagent (in order to solubilize the cells) and whole blood precipitation reagent/probe wash to the sample (in order to precipitate protein), followed by centrifugation to obtain a clear supernatant. The supernatant was then used for further analysis. 68 2.6.3 Assay Controls Three different controls, stored in vials labeled L , M , and H were utilized throughout sample analysis. Control solutions were reconstituted by volumetrically pipetting 2.0 mL of Barnstead water into 1 vial. The solution was then left to stand at room temperature for 20 minutes, with occasional swirling. Prior to use, the vial was inverted several times to ensure homogeneity. Reconstituted vials were stored at 2 to 8 °C and were stable for 30 days following reconstitution. If the controls were not freshly reconstituted, they were warmed to room temperature and inverted several times prior to use. Two levels of controls were used in each run, and the levels used were alternated between runs. The control vials should read in the following ranges: vial L , 120 to 180 ng per mL; vial M , 340 to 460 ng per mL; vial H , 680 to 920 ng per mL. 2.6.4 Calibration Calibration was always required following the machine activation procedure, and i f the memory circuit is replaced. Calibration may be required when quality control is unacceptable, reagents with new lot numbers are used, buffer with a new lot number is used, when any dispenser component is replaced, and when any instrument calibration is performed. During the calibration procedure, all calibrators are run in duplicate, and all levels of control solutions are run. If possible, a fresh reagent pack and a fresh calibrator pack are used. 2.6.5 Sample Analysis Procedure A maximum of 20 samples can be run in 1 batch. A l l samples (calibrators, controls, and patient samples) were prepared as described in section 2.6.2 Sample Preparation. Following 69 preparation, the samples were mixed by gentle inversion in order to prevent foaming, which could affect the results. If excessive foaming did develop, the samples were left to stand for a few minutes until the foam dissipated. One hundred fifty uL of the sample to be analyzed was pipetted into an appropriately labeled centrifuge tube. The same pipette tip was used throughout, with the pipette tip being rinsed with normal saline several times between each transfer of sample. Fifty uL of solubilization reagent was then pipetted into the centrifuge tube, followed by 300 p.L of whole blood precipitation/probe wash solution. The centrifuge tube was then capped and vortexed for 10 seconds to ensure thorough mixing. The samples were then centrifuged for 5 minutes at 9500 g, or until a clear supernatant and a hard, compact pellet of denatured protein were obtained. After centrifugation was complete, the centrifuge tubes were uncapped and the supernatant was immediately decanted into the corresponding sample well of a sample cartridge. (A minimum of 150 uL of sample supernatant is required to perform the assay). Samples were run immediately following transfer to the sample cartridges. 2.6.6 L i n e a r i t y The standard curve is linear and useable in the range of 25 to 1500 ng/mL. Samples with a concentration of greater than 1500.00 ng/mL are diluted with an equal amount of whole blood calibrator A (0.00 ng/mL) prior to performing the solubilization step. The final sample concentration is then determined by doubling the result obtained. 70 2.6.7 Precision Reproducibility was determined by assaying the three C S A monoclonal whole blood controls, L (low, 150 ng/mL), M (medium, 400 ng/mL), and H (high, 800 ng/mL) in replicates of 5 in each of 10 independent runs on 11 instruments at 10 laboratories. Within run coefficient of variation (CV), between run C V , and between laboratory C V were calculated (Table 5). Table 5. Expected precision of the monoclonal fluorescence polarization immunoassay used for sample analysis. (Adapted from the V H H S C laboratory standard procedure manual I J J J J . Target Cyclosporine Concentration Cng/mL) 150.0 400.0 800.0 \ cy / Mean Cyclosporine Concentration Obtained (ng/mL) 146.7 390.8 791.8 C V Within Run (%) 3.2 2.3 2.3 CV Between Run (%) 3.9 2.9 3.0 CV Between Laboratory (%) 4.7 3.5 3.5 2.6.8 Standard Reference Intervals For Cyclosporine Trough Levels Utilized At Vancouver Hospital And Health Sciences Centre The standard trough concentrations targeted at V H H S C vary with the type of transplant and the time post-transplant (Table 6). Results are reported as whole numbers, and i f the result is less than the lower limit of quantitation of the assay, the result is reported as "<25 ng/mL". 7 1 Table 6. Standard reference intervals for target trough cyclosporine concentrations at Vancouver Hospital and Health Sciences Centre. Transplant Type Time Post-Transplant Target Trough Concentration (ng/mL) Kidney, Liver, and Heart Less than 1 month 400 to 450 Less than 2 months 300 to 400 2 to 6 months 200 to 300 6 to 12 months 150 to 250 Greater than 12 125 to 200 months Lung 0 to 1 month 350 to 400 1 to 3 months 300 to 350 3 to 6 months 250 to 300 6 to 12 months 200 to 250 Greater than 12 150 to 200 months 2.6.9 Specificity Of The Fluorescence Polarization Immunoassay The specificity of the high volume assay methods utilized in the routine clinical management of transplant patients receiving CSA as the primary form of immunosuppression was improved greatly when polyclonal antibodies were replaced with monoclonal antibodies that were more specific for the parent drug. However, because of the large structure of CSA and because the large cyclic structure remains intact following metabolism, there is still cross-reactivity with metabolites even with the monoclonal antibodies. The cross-reactivity of the monoclonal antibodies used in the fluorescence polarization immunoassay used at V H H S C was determined by spiking whole blood specimens containing 200 ng/mL of CSA with known amounts of the CSA metabolites M l , M8, M17, M18, and M21. The percent cross-reactivity was determined according to the following equation: Percent Cross-Reactivity = (Metabolite concentration obtained - Control concentration obtained) x 100 Metabolite concentration added 72 The results are summarized in Table 7. Occasionally, patient samples are obtained which contain unusually high levels of metabolites. This primarily occurs in patients with hepatic dysfunction or in patients who have inadequate biliary flow. Although the cross-reactivity of the assay with the metabolites tested is low (with 1 exception), the presence of these metabolites in high concentrations can result in overestimation of the trough CSA concentration. In this case, it is recommended by the manufacturer that an alternate methodology that quantitates the parent drug only, such as HPLC, be used instead. Table 7. Cross-reactivity of the monoclonal fluorescence polarization immunoassay used at Vancouver Hospital and Health Sciences Centre. Metabolite Concentration Added (ng/mL) Average Cross-Reactivity In The Presence Of Cyclosporine (%) Standard Deviation M l 250 19.4 2.6 M8 250 Less than 5 M17 500 6.7 1.7 M18 250 Less than 5 M21 250 Less than 5 2.7 Pharmacokinetic Analysis Drug concentration-time data obtained from the clinical study was modeled with WinNonlin® pharmacokinetic modeling software (Version 1.1, SCI Software, North Carolina, USA). For the WinNonlin® pharmacokinetic analysis, model 200, a noncompartmental model with extravascular input and plasma/blood data, was utilized. Initial estimates of the pharmacokinetic parameters were calculated manually using Microsoft® Excel 2000. The area under the concentration versus time curve over one dosing interval (AUCo-T) for each subject was calculated using the linear trapezoidal method. Briefly, the trapezoidal method involves the description of the concentration versus time curve by a function that depicts the curve as a series of straight lines. Hence, the area under this curve can be divided into multiple 73 trapezoids. Then, the area of each trapezoid is calculated and the sum of the areas of each trapezoid yields an estimate of the true A U C . Dose corrected A U C n - T was calculated by dividing the value obtained for A U C n - T by the cyclosporine dosage in mg/kg. The area under the first moment curve for one dosing interval ( A U M C n - x ) for each subject was calculated in a manner similar to the trapezoidal method, except a concentration x time versus time curve was described by a function that depicts the curve as a series of straight lines. Mean residence time (MRT) for each subject was calculated according to a previously published method (Pfeffer 1984): M R T = [AUMCo-x + (x x AUC 0 .co)] /AUC 0 . T The maximum concentration observed during the dosing interval (C m a x ) and the time of the maximum concentration observed during the dosing interval (Tm ax) f ° r each subject were elucidated by visual inspection of the appropriate concentration versus time curves. The terminal elimination rate constant (kz) for each subject was determined from the slope of the terminal elimination phase. This slope was determined by least-squares linear regression of the terminal elimination phase. For the WinNonlin pharmacokinetic analysis, the points used in the calculation of kz were determined by the program. For the determination of the initial estimate of kz, the 8-hour to 12-hour concentration-time points were utilized. The terminal elimination half-life (ti/2) for each subject was calculated initially according to the following equation: ti/2 = 0.693/Xz The apparent oral clearance (CL/F) for each subject was calculated as follows: CL/F = Dose/AUCo-t 74 The apparent volume of distribution (Vd/F) for each subject was calculated as follows: Vd/F = Dose/(?,zxAUCo.T) 2.8 Sample Size The total sample size used for the clinical study was a convenience sample size. A minimum of 14 patients (providing a sample size of 8 for Phase I and 6 for Phase II) was deemed reasonable given that only 23 lung transplant patients are followed at the solid organ transplant clinic based at VHHSC. One of these patients no longer lives in British Columbia, and is followed at Foothills Hospital in Calgary, Alberta, Canada. The remainder of the lung transplant patients above the 14 selected either refused to participate or were unable to participate due to medical reasons. Members of the investigative group with experience in determining pharmacokinetic parameters (in Phase I) and testing their predictive performance (in Phase II) (Radomski et al 1997, Birt and Chandler 1990, Rhoney et al 1993) suggested that 14 patients was a reasonable number to use for the clinical study. 2.9 Phase I Of The Clinical Study Data from the first 8 subjects were used to develop the limited sampling strategy (LSS) for lung transplant patients. Multiple regression analysis, with A U C as the dependent variable and the blood concentrations grouped by time as the independent variables, was performed using Statistica® statistical software (StatSoft Inc., Tulsa, Oklahoma, USA). The backward elimination method was used to calculate the initial regression equation. Briefly, this procedure involves an initial regression using all concentration-time data. Then concentration-time data are removed one collection time (e.g., all of the 6 hour post-dose samples) at a time. If this deletion made 75 little difference to the coefficient of determination, it was not included in the final equation. If the deletion was important to the regression, then it was included. Following this initial regression analysis, the remainder of the analysis was restricted to the first 3 hours post-dose. In order to generate as many potential equations as possible, the concentration-time data to be analyzed were chosen manually, and the standard regression analysis technique was performed. 2.10 Phase II Of The Clinical Study Data from the remaining 6 subjects was used to test the predictive performance of the limited sampling strategy developed in Phase I. Predictive performance of our limited sampling strategy was compared with the predictive performance of other published limited sampling strategies derived from non-lung transplant patients. Predictive performance was calculated according to the methods proposed by Sheiner and Beal (Sheiner and Beal 1981). First, concentration-time data was used to calculate A U C from the LSS for lung transplant patients, which was determined in Phase I. For each LSS evaluated (the optimal LSS developed specifically for lung transplant patients and other previously published LSS developed for other transplant populations), prediction error (pe) was determined using the following equation: pe = predicted A U C - actual A U C Percent prediction error (%pe) was calculated according to the following equation: %pe = 100 x (predicted A U C - actual AUC) (actual AUC) We arbitrarily chose an acceptable error of 10% for %pe. Thus, A U C predictions that were within ± 10%> of the actual value were deemed acceptable. These error limits are widely 76 regarded as standard in basic science and the pharmaceutical industry, and are likely considered strict for therapeutic drug monitoring of CSA. Following calculation of pe and %pe for each LSS evaluated, the mean prediction error (ME, bias) and the mean absolute error (MAE, precision) were determined according to the following equations: M E = (Zpe)/n M A E = ( £ | p e | ) / n Comparisons of predictive performance were done in a pairwise fashion. A l l comparisons of predictive performance involved the optimal LSS for lung transplant patients and 1 previously published LSS derived for other transplant patients. Using pe values for each LSS, the mean difference in prediction error (MDPE) was calculated by taking the mean of the differences between all of the respective pe. Similarly, the mean difference in absolute error (MDAE) was calculated by taking the mean of the differences between all of the respective absolute errors. Finally, the 95% confidence intervals (CI) are calculated for M D P E and M D A E . If the 95% confidence interval (CI) for the mean difference between two equations' prediction errors does not contain zero, then the respective equations differ in bias at a significance level of < 0.05. If the 95% CI for the mean difference between two equations' absolute errors does not contain zero, then those equations differ in precision at a significance level of < 0.05 (Sheiner and Beal 1981, Radomski et al 1997, Rhoney et al 1993, Welch et al 1993, Leader et al 1994, Cropp et al 1998, Davis and Chandler 1996). CHAPTER 3 RESULTS 78 3.1 Lung Transplant Patient Characteristics Fourteen stable, adult lung transplant patients participated in the clinical study. There was a 50:50 distribution in sex. A l l of the various types of lung transplant (double lung, right single lung, and left single lung) were represented. In addition, most of the common indications for lung transplantation were represented (Table 8). One patient, who participated in Phase II of the clinical study, was also the recipient of a kidney transplant necessitated by CSA nephrotoxicity. This patient had received a double lung transplant for cystic fibrosis, and had the longest time post-transplant. Another patient, who also participated in Phase II, was on dialysis due to CSA nephrotoxicity. The patient had also received a double lung transplant for cystic fibrosis, and received CSA every 8 hours. Because dialysis does not affect CSA levels due to the large size and lipophilicity of the CSA molecule, and because dialysis was not an exclusion criterium, the patient's data were included in Phase II data analysis. Table 8. Patient characteristics for 14 stable, adult lung transplant patients. Age 48 ± 12 years Sex 7 male, 7 female Weight 69 ± 17 kg Transplant Type 6 double lung, 6 right single lung, 2 left single lung Time Post-Transplant 5.1 + 3.4 years Reason For Transplant Chronic obstructive lung disease (6 cases), alpha-1 antitrypsin deficiency emphysema (3 cases), cystic fibrosis (4 cases), idiopathic pulmonary fibrosis (1 case) Immunosuppressive Regimen Triple therapy: cyclosporine (Neoral®) + mycophenolate mofetil or azathioprine + prednisone CSA Dose 4.3 ± 1.7 mg/kg/day Note: Data are presented as mean ± standard deviation 3.2 Cyclosporine Concentration-Time Data A total of 160 blood samples were gathered for analysis by FPIA at the hospital laboratory, VHHSC. Of the 160 blood samples dropped off at the laboratory, 158 were analyzed 79 (Table 9 and 10). Two samples from 1 patient in Phase I were lost prior to analysis. Fortunately, the samples were in the middle of the terminal elimination phase and thus, were not crucial for the pharmacokinetic and multiple regression analysis. In the case of 12 patients, 11 blood samples were gathered over 1 dosing interval of 12 hours. In the remaining 2 cases, 8 blood samples were gathered over 1 dosing interval of 8 hours. Both of these patients were young (with ages greater than 1 SD less than the mean) cystic fibrosis patients with double lung transplants. Both were documented as poor absorbers of CSA in their respective medical chart. One of these patient's data were included in Phase I analysis, and the other patient's data were included in Phase II analysis. Table 9. Concentration-time data for 14 stable, adult lung transplant patients. Time Pt. Pt. Pt. Pt. Pt. Pt. Pt. Pt. Pt. Pt. Pt. Pt. Pt. Pt. 001 002 003 004 005 006 007 008 009 010 011 012 013 014 0 154 193 274 169 154 213 319 177 184 169 120 302 223 123 1 935 992 691 1137 1203 1219 1323 320 1096 1588 827 3076 1818 301 2 722 749 1093 724 737 746 1482 1184 1276 1120 627 1752 972 789 3 423 457 775 472 462 518 957 751 895 651 405 949 613 314 4 310 371 637 348 397 415 633 497 705 454 266 666 485 257 5 243 308 546 286 326 338 547 395 594 319 215 528 430 194 6 198 272 420 242 295 Lost 476 341 500 268 189 454 339 160 8 148 222 238 198 218 Lost 349 245 350 200 156 376 267 109 9 142 215 — 192 200 250 300 207 327 177 138 362 240 -10 133 188 — 180 193 223 298 194 300 162 131 359 211 — 12 123 163 — 157 163 203 257 178 251 130 124 296 180 -Note: Time is in units of hours, and concentration is in units of ng/mL 80 Table 10. Mean concentration-time values for 14 stab Time Mean Concentration Standard Deviation (hours) (ng/mL) (ng/mL) 0 198 62 1 1180 689 2 998 336 3 617 215 4 460 150 5 376 133 6 320 114 8 237 82 9 229 70 10 214 70 12 185 56 e, adult lung transplant patients. 3.3 Steady-State Pharmacokinetic Parameters For Cyclosporine (Neoral ) Individual (Table 11) and mean (Table 12) steady-state pharmacokinetic parameters for CSA (Neoral® [NEO]) were calculated using concentration-time data gathered from 14 stable, adult lung transplant patients. The pharmacokinetic characteristics of the NEO formulation are more predictable than those of the SIM formulation, thus reducing inter- and intra-patient variability in pharmacokinetic parameters. Because there was only 1 study visit per patient, it was not possible to evaluate intra-patient variability in pharmacokinetic parameters. As can be seen from the ranges observed in the individual pharmacokinetic parameters (Table 13), there is noticeable inter-patient variability. The results confirm the generalizations in pharmacokinetic parameters reported in the literature for the various transplant populations. That is, CSA NEO, in stable, adult lung transplant patients, has a long Un, a large apparent Vd, and a T m a x that occurs predictably within 2 hours following dosage administration. Figure 3 depicts the mean concentration versus time curve developed from data from 14 stable lung transplant patients. Figure 4 depicts the mean log concentration versus time curve developed from the same data. Figures 5 and 6 depict all 14 concentration versus time curves on one graph, and all 14 log concentration versus time curves on one graph, respectively Table 11. Individual steady-state pharmacokinetic parameters for cyclosporine (Neoral ) in 14 stable, adult lung transplant patients. Parameter Pt. 001 Pt. 002 Pt. 003 Pt. 004 Pt. 005 Pt. 006 Pt. 007 AUCo-x 3669 4378 4688 4381 4615 5044 7301 (ngxhr/mL) Dose- 2422 2977 2297 3637 3655 4237 2434 Corrected AUCo-x (ngxhr/mL) AUMCo-x 14366 18539 15955 17677 19098 21605 31001 (ngxhr /mL) M R T (hr) 24.44 21.23 22.41 22.43 22.05 22.78 19.86 Cmax (ng/mL) 935 992 1093 1137 1203 1219 1482 Tmax (hr) 1.08 0.93 2.00 0.97 1.00 1.00 2.00 CL/F (L/hr) 15.94 20.24 27.08 14.89 12.87 12.87 15.79 CL/F (L/hr/kg) 0.24 0.24 0.37 0.18 0.15 0.15 0.32 Vd/F (L) 337 226 97 222 253 173 135 Vd/F (L/kg) 5.11 2.66 1.32 2.67 2.06 2.06 2.70 Xz (hr"1) 0.047 0.089 0.037 0.067 0.072 0.074 0.117 ti/2 (hr) 14.67 7.76 2.48 10.32 9.32 5.93 5.93 Parameter Pt. 008 Pt. 009 Pt. 010 Pt. 011 Pt. 012 Pt. 013 Pt. 014 AUCo-x 4787 6984 5450 3368 9553 6077 2272 (ngxhr/mL) Dose- 3532 3179 1526 1437 2809 2139 1263 Corrected AUCo-x (ngxhr/mL) AUMCo-x 21806 31192 19551 13324 35395 23442 7087 (ngxhr2/mL) M R T (hr) 25.61 21.29 18.31 32.20 20.55 19.18 13.15 Cmax (ng/mL) 1184 1276 1588 827 3076 1818 789 Tmax (hr) 2.00 1.98 1.00 1.00 1.00 1.00 1.97 CL/F (L/hr) 11.91 12.77 22.43 19.01 18.65 28.99 26.31 CL/F (L/hr/kg) 0.16 0.22 0.53 0.30 0.25 0.37 0.63 Vd/F (L) 242 143 213 699 243 271 140 Vd/F (L/kg) 3.27 2.51 5.08 10.93 3.30 3.42 3.35 Xz (hf 1) 0.049 0.089 0.105 0.027 0.077 0.107 0.189 ti/2 (hr) 14.07 . 7.76 6.59 25.49 9.02 6.48 3.68 82 Table 12. Mean steady-state pharmacokinetic parameters for cyclosporine (Neoral ) in 14 stable Parameter Mean Value Standard Deviation Range AUCo-x (ngxhr/mL) 5183 1835 2272 to 9553 Dose-Corrected 2682 912 1263 to 4237 AUCO-T (ngxhr/mL) AUMCo-x (ngxhr2/mL) 20712 7655 7087 to 35395 M R T (hr) 21.82 4.22 13.15 to 32.20 Cmax (ng/mL) 1330 578 789 to 3076 Tmax (hr) 1.35 0.49 0.93 to 2.00 CL/F (L/hr) 18.93 5.52 11.91 to 28.99 CL/F (L/hr/kg) 0.30 0.14 0.15 to 0.63 Vd/F (L) 242 146 97 to 699 Vd/F (L/kg) 3.64 2.33 1.32 to 10.93 Xz (hf 1) 0.082 0.041 0.027 to 0.189 Un (hr) 9.52 5.71 2.48 to 25.49 (5.93 to 25.49 if the patients receiving CSA q8h are excluded) 3.4 Phase I: Multiple Regression Analysis Multiple regression analysis using all concentration-time data, as well as concentration-time data restricted to the first 3 hours post-dose (for clinical feasibility purposes) yielded a total of 17 LSS to choose from. Of these 17 LSS, 2 required 4 blood samples. One of these LSS had the highest coefficient of determination (0.999), but required a wait time of 12 hours. This equation was derived from all of the concentration-time data. Fifteen of the other equations were derived from concentration-time data gathered during the first 3 hours post-dose. The other equation requiring 4 blood samples had a patient wait time of 3 hours. Four LSS required 3 blood samples, 6 LSS required 2 blood samples, and the remaining 4 LSS required 1 blood sample. Because these equations were derived from concentration-time data in the first 3 hours post-dose, the maximum patient wait time was 3 hours, and in 7 cases, it was less than this. Coefficients of determination were obtained for all LSS generated. In addition, in order to aid in 83 the choosing of the optimal LSS developed specifically for lung transplant patients, the bias, precision, and percent prediction error (%pe) range of each LSS was calculated using the concentration-time data gathered from the 6 patients in Phase II (Table 14). Taking into account the number of blood samples required, patient wait time, coefficient of determination, bias, precision, and %pe range, we defined the following equation as the optimal LSS for the population of lung transplant patients studied: A U C = 1.46xCi + 5.36xC3 + 274.49 In order to confirm this LSS was the optimal LSS, the predictive performance of all of the LSS developed was compared (Table 15). This comparison was performed primarily to determine i f the LSS compared favorably to equations that required more blood samples in terms of bias and precision. When compared to the 1-point LSS, there were no significant differences in bias. However, the recommended LSS was significantly more precise than 3 of the 4 1-point LSS. When the 2-point LSS were compared, it was found that there were also no significant differences in bias, while the recommended LSS was significantly more precise than 4 of 5 2-point LSS compared. There were no significant differences in bias or precision when the recommended LSS was compared to the various 3-concentration LSS, with 3 exceptions. The recommended LSS was significantly less biased than 1 equation, was significantly more precise than another, but was significantly more biased than a 3 r d 3-concentration equation. (The only unfavorable comparison, with the 3-concentration requiring blood samples at 0, 1, and 3 hours post-dose, is italicized, boldfaced, and underlined in Table 15). 85 C/3 -(—> fi CD OH fi 'EH fi fi fi CD -•—' a> <tf }-< O CD CD > o CD fi S-i o &. O &2 CD CD ctf 00 (-|ui/Bu) UO!1BJ;U90UOO Bon fi .2 ctf fi CD ctf O CD fi O M o Ofj o .fi ^ > CD 'cD , f i O = WD Finally, there were no significant differences in bias or precision between the recommended LSS and the 4-point equation derived from all of the concentration-time data gathered in the first 3 hours post-dose. Table 13. The limited sampling strategies, and their respective predictive performance, derived ^ _ from concentration-time data from 8 stable, adult lung transplant patients. A UC Equation r2 %pe (range) ME (ngxhr/mL) MAE (ngxhr/mL) a) Equation from multiple regression analysis of all concentration-time data AUC = 17.24xC6 - 58.96xC8 + 23.39xC9 + 52.29xC, 2- 796.07 0.999 -45.50 to 13.99 -876.83 1267.77 b) Equation using all concentration-time points in the first 3 hours post-dose AUC = 2.99xC0 + 1.17xC, - 2.69xC2 + 8.60xC 3- 592.31 0.936 -11.05 to 10.08 21.29 277.29 c) Three-concentration equations using concentration-time points in the first 3 hours post-dose AUC = 0.96xC0 + 1.46xC, + 3.42xC2 + 34.52 0.945 -11.58 to 44.80 440.73 710.28 AUC = 1.56xC0 + 1.34xCi + 4.96xC3 + 320.54 0.964 -6.71 to 9.01 102.86 259.02 AUC = 13.05xC0 - 8.24xC2 + 14.27xC3 + 1634.10 0.908 -51.00 to 13.23 -1419.09 1567.58 AUC = 1.42xd - 1.52xC2 + 7.61xC3 + 403.53 0.964 -2.33 to 10.08 44.94 253.63 d) Two-concentration equations using concentration-time points in the first 3 hours post-dose AUC = 19.13xCo-0.03xQ + 1167.63 0.846 -33.35 to 54.53 -918.91 1355.62 AUC = 14.13xC0 + 1.13xC2 + 1053.43 0.886 -27.04 to 62.08 -692.80 1193.02 AUC = 13.32xC0 + 1.95xC3 + 1119.87 0.899 -26.79 to 48.33 -764.87 1177.76 AUC = 1.54xd + 3.59xC2 - 8.61 0.963 -10.36 to 44.67 519.50 760.78 AUC = 1.46xd + 5.36xC3 + 274.49 0.975 -4.47 to 8.47 194.60 298.62 AUC = -9.10xC2 + 18.74xC3 + 2342.01 0.741 -53.95 to 25.49 -1235.06 1694.17 e) One-concentration equations using concentration-time points in the first 3 hours post-dose AUC = 19.06xC0+ 1150.60 0.884 -33.31 to 53.80 -905.49 1336.46 AUC= 1.28xd + 3597.49 0.130 -28.40 to 75.27 -162.37 1222.13 AUC = 3.45xC2 + 1638.14 0.737 -19.58 to 91.88 -220.80 1078.78 AUC = 5.22xC3 + 1789.53 0.770 -29.41 to 50.88 -498.12 1062.24 87 Table 14. Comparisons of predictive performance between the recommended limited sampling strategy for lung transplant patients and the remaining limited sampling strategies for LSS Equation MDPE 95% CI MDAE 95% CI A U C = 19.06xC 0+ 1150.60 1100 -49 to 2249 -1038 -1825 to -250 A U C = 1.28xCi + 3597.49 357 -841 to 1555 -924 -1670 to -178 A U C = 3.45xC 2+ 1638.14 415 -703 to 1534 -780 -1524 to -36 A U C = 5.22xC3 + 1789.53 693 -453 to 1839 -764 -1565 to 37 A U C = 19.13xC 0 -0.03xCi + 1167.63 1114 -52 to 2279 -1057 -1853 to -261 A U C = 14.13xC0 + 1.13xC2 + 1053.43 887 -167 to 1942 -894 -1574 to -215 A U C = 13.32xC 0+ 1.95xC3 + 1119.87 959 -78 to 1997 -879 -1571 to -188 A U C = 1.54xCi + 3.59xC 2-8.61 -325 -812 to 162 -462 -855 to -69 A U C = -9.10xC 2 + 18.74xC3 + 1430 -438 to 3297 -1396 -2859 to 68 2342.01 A U C = 0.96xC 0 + 1.46xCi + 3.42xC2 -246 -711 to 219 -412 -775 to -48 + 34.52 A U C = 1.56xC0 + 1.34xCi + 4.96xC 3 92 5 to 179 40 -74 to 153 + 320.54 A U C = 13.05xC0 - 8.24xC2 + 1614 83 to 3145 -1269 -2586 to 48 14.27xC3 + 1634.10 A U C = 1.42xC, - 1.52xC2 + 7.61xC3 150 -59 to 358 45 -49 to 139 + 403.53 A U C = 2.99xC 0+ 1.17xCi-2.69xC 2 173 -12 to 359 21 -62 to 104 + 8.60xC 3-592.31 3.5 Phase II: Comparisons Of Predictive Performance Of the 21 LSS reported from the 16 clinical studies identified, 11 were deemed eligible for comparison in Phase II of the clinical study (Table 16). Five of the LSS deemed not eligible for comparison were developed using the SIM formulation, and 3 of these studies with SIM used once daily dosing. These LSS were excluded from analysis because of the pharmacokinetic differences between the SIM and NEO formulations of CSA. Also, once daily dosing is no longer used in clinical practice. The LSS proposed by Foradori and colleagues (Foradori et al 1995) was appropriate for comparison. However, the 3-concentration equation required a blood sample at 2.5 hours post-88 dose, for which we did not have concentration-time data. For this reason, the LSS was excluded from analysis. The 3-concentration LSS proposed by Amante and colleagues (Amante et al 1996) required a blood sample for which we did not have concentration-time data (at 1.5 hours post-dose). For the 4 LSS reported by Cooney and colleagues (Cooney et al 1996), only 1 was eligible for further evaluation. One LSS required a blood sample at 3.5 hours post-dose, for which we did not have concentration-time data. The other 2 LSS, one a 2-concentration LSS, and the other, a 3-concentration LSS, were deemed not eligible because they both required a 12 hour post-dose trough concentration. This is not clinically feasible, except perhaps in an inpatient setting. However, because the stable, adult lung transplant patients studied were followed exclusively on an outpatient basis, this was not appropriate. A l l of the LSS compared with our derived, optimal LSS for lung transplant patients involved either 2-concentration or 3-concentration LSS. There were no significant differences in bias or precision with 5 exceptions. Of these 5 exceptions, only 1 was unfavorable. Our optimal LSS was significantly less biased than the LSS proposed by Meier-Kriesche and colleagues (Meier-Kriesche et al 1998) and Marsh (Marsh 1999), and significantly more precise than the LSS proposed by the same investigators. The derived LSS was significantly more biased than the 3-concentration LSS proposed by Gaspari and colleagues (Gaspari et al 1997). 89 Table 15 . Comparisons of predictive performance between the recommended limited sampling strategy for lung transplant patients and other previously published limited sampling strategies developed for non-lung transplant patient population. Study Population Equation Johnston et al 1 9 9 6 Renal A U C = 1 . 9 6 x C 2 h + 1 1 . 5 x C 8 h + 3 5 5 . 2 Serafinowicz et al 1 9 9 6 Renal A U C = 9 . 1 3 1 x C 0 h + 0 . 7 8 4 x C i h + 2 . 6 1 7 x C 2 h + 1 9 3 . 5 6 1 Amante et al 1 9 9 6 Renal A U C = 2 . 4 x C 2 h + 7 . 7 x C 6 h + 1 9 5 . 8 Keown et al 1 9 9 6 Renal A U C = 1 . 8 4 x C 0 h + 4 . 3 9 x C 2 h + 3 1 2 . 6 6 Cooney et al 1 9 9 6 Pediatric Liver A U C = 1 0 . 1 9 x C 0 h + 4 . 4 7 x C 3 h + 7 4 9 . 7 Gaspari et al 1 9 9 7 Renal A U C = 5 . 1 8 9 x C 0 h 1 . 2 6 7 C i h + 4 . 1 5 0 x C 3 h + 1 3 5 . 0 7 9 Primmett etal 1 9 9 8 Renal A U C = 1 2 . 3 4 x C 0 h + 2 . 4 8 x C 2 h + 4 4 1 . 4 2 Primmett etal 1 9 9 8 Renal A U C = 9 . 5 5 x C 0 r i + 0 . 9 6 x C i h + 2 . 0 5 x C 2 h + 1 1 2 . 0 7 Meier-Kriesche etal 1 9 9 8 Renal A U C = 1 . 8 4 x C 2 h + 4 . 3 9 x C 4 h + 1 2 9 Lemire et al 1 9 9 8 Renal A U C = 1 . 0 2 3 x C i h + 1 3 . 1 0 x C 6 h + 2 4 2 Marsh 1 9 9 9 Renal A U C = 6 x ( C 0 h + C 4 h ) 90 Table 15. (Continued) Study Population %pe (range) ME (ngxhr/mL) MAE (ngxhr/mL) 9 5%CI (ngxhr/mL; bias/precision) Johnston et al Renal -15.07 to -332.34 630.14 -219.43 to 1273.31/ 1996 38.85 -798.75 to 135.72 Serafmowicz Renal -13.05 to 270.77 574.67 -563.93 to 411.59/ etal 1996 59.19 -704.79 to 152.70 Amante et al Renal -17.34 to -355.77 746.87 -307.24 to 1407.98/ 1996 46.16 -977.45 to 80.95 Keown et al Renal -17.88 to -178.53 785.51 -467.31 to 1213.57/ 1996 76.15 -1111.59 to 137.82 Cooney et al Pediatric -17.88 to -112.49 629.03 -438.99 to 1053.17/ 1996 Liver 49.92 -853.29 to 192.47 Gaspari et al Renal -11.33 to 8.16 -27.15 241.50 62.74 to 380.77/ 1997 -184.32 to 298.56 Primmett Renal -20.71 to -94.71 433.87 -211.36 to 789.99/ etal1998 40.52 -671.77 to 401.28 Primmett Renal -15.88 to -168.65 753.07 -432.89 to 1159.39/ etal1998 72.33 -1034.64 to 125.75 Meier- Renal -34.30 to -1410.95 1556.49 569.73 to 2641.37/ Kriesche et al 19.21 -2062.55 to -453.18 1998 Lemire et al Renal -2.27 to 16.44 279.40 375.70 -643.61 to 474.00/ 1998 -405.73 to 251.57 Marsh 1999 Renal -39.20 to 0.34 -1663.14 1665.68 808.31 to 2907.17/ -2318.53 to -415.58 CHAPTER 4 DISCUSSION 92 The objectives of our study were threefold. First, we sought to define the optimal limited sampling strategy for CSA monitoring in lung transplant patients. Second, we wished to compare the predictive performance of this optimal limited sampling strategy with the predictive performance of previously published limited sampling strategies in other (i.e., non-lung) transplant populations. Finally, we wished to determine the mean steady-state pharmacokinetic parameters of C S A (Neoral®) in our population of lung transplant patients. To meet the first objective, we defined the optimal limited sampling strategy as a 2-concentration equation for A U C with blood samples required at 1 and 3 hours post-dose. This recommendation was based on the number of blood samples required, clinical feasibility, coefficient of determination, %pe range, bias, precision, as well as comparisons of predictive performance with the other 1-, 2-, 3-, and 4-concentration LSS developed specifically for our population of lung transplant patients. The equation is as follows: A U C = 1.46xCi + 5.36xC3 + 274.49 For the second objective, it was found that the predictive performance of the optimal LSS for lung transplant patients compared very favorably with the predictive performance of previously published LSS developed for non-lung transplant populations. In all comparisons except for 5, there were no significant differences in bias or precision. Of these 5 comparisons in which there were significant differences found in either bias or precision, only 1 was unfavorable. The optimal LSS was significantly less biased than the LSS proposed by Meier-Kriesche and colleagues (Meier-Kriesche et al 1998) and Marsh (Marsh 1999), and significantly more precise than the LSS proposed by the same investigators. The optimal LSS was significantly more biased than the 3-concentration LSS proposed by Gaspari and colleagues (Gaspari etal 1997). 93 Of the 14 lung transplant patients who participated in the study, 12 received CSA every 12 hours, and 2 received CSA every 8 hours. This brings to light possible concerns regarding sample homogeneity. We chose to include the 2 lung transplant patients (1 patient in Phase I, and 1 patient in Phase II) who received CSA every 8 hours in all analyses performed for several reasons. The most important reason was that we wanted to develop a monitoring method that was useable for all patients encountered in routine clinical practice. Also, our aim was to develop an LSS that gave a precise and unbiased estimate of A U C during the dosing interval. Because we were interested specifically in the AUCo-T, the 2 lung transplant patients could be included, because we sampled sufficiently to calculate their AUCo-T- Thus, our experiment was designed such that the length of the dosing interval did not matter, as long as the drug regimen was at steady-state. 4.1 Steady-State Pharmacokinetic Parameters For Cyclosporine (Neoral®) The mean steady-state pharmacokinetic parameters for CSA (NEO) in lung transplant patients were determined using concentration-time data gathered from 14 stable, adult lung transplant patients. As mentioned previously, 12 of these patients received CSA every 12 hours, and the other 2 patients, both of whom were documented poor absorbers of CSA, received CSA every 8 hours. The pharmacokinetic parameters of interest clinically (i.e. CL/F, Vd/F, and Ua) were determined. Mean CL/F was 0.30 L/hr/kg, mean Vd/F was 3.64 L/kg, and mean ti/2 was 9.52 hours. Mean bioavailability of CSA is generally regarded in the literature as 30%. However, this value is for the SIM formulation of CSA, which is erratically and incompletely absorbed when compared to NEO. To my knowledge, no mean apparent bioavailability figure is available for the NEO formulation. One study has shown that the relative bioavailability of CSA 94 in the NEO formulation was 1.84 and 2.09 times greater than SIM at dosages of 200 and 800 mg, respectively in heart-transplant candidates (Tan et al 1995). However, relative bioavailability cannot be used to convert CL/F to C L and Vd/F to Vd. Thus, I did not believe it was appropriate to calculate apparent clearance (CL) and apparent volume of distribution (Vd) using an F of 30%. The mean apparent bioavailability of CSA in the NEO formulation is certainly higher than 30%, but it is not possible to speculate by how much at this time. CL/F is an important piece of information, as that parameter has been associated with outcome in renal transplant patients in 2 large clinical studies (Lindholm et al 1993, Lindholm and Kahan 1993). Both of these studies compared patients with and without rejection, and patients who had suffered graft loss, and patients who had not. Both studies found that patients with rejection and patients who suffered graft loss had a significantly higher oral clearance (Lindholm et al 1993, Lindholm and Kahan 1993). Lindholm and colleagues found that an oral clearance of >85 L/hr was associated with an increased incidence of rejection, while an oral clearance of >101 L/hr was associated with an increased incidence of graft loss. An oral clearance of <59 L/hr was associated with a decreased incidence of acute rejection, while an oral clearance of <63 L/hr was associated with a decreased incidence of graft loss. Lindholm and Kahan found that an oral clearance of >90 L/hr was associated with decreased graft survival and an increased incidence of acute rejection. The mean oral clearance obtained in our current clinical study was approximately 19 L/hr. One possible explanation for the large difference in oral clearances between our study and the one by Lindholm and colleagues is that their study was conducted in the early post-transplant period. 95 The apparent volume of distribution observed in our clinical study confirms that CSA, likely due to its lipophilic nature, has a large apparent volume of distribution. However, because the parameter calculated is actually Vd/F and not Vd, this parameter also depends on bioavailability. To my knowledge, Vd/F has not been associated with outcome in a clinical trial. Thus, this parameter has limited use both clinically and academically until a bioavailability value for NEO becomes available. CSA is regarded as a drug with a long terminal elimination half-life, and our value of 9.52 hours is consistent with that generality. A mean terminal elimination value of 5.5 hours has been reported in lung transplant patients (Kesten et al 1998a), and values of 6.3 and 8.1 hours have been reported in heart-lung transplant patients (Tsang et al 1994). Our value was higher than these. Because of the low numbers of patients involved in these studies (14 lung transplant patients for ours, and an additional 20 lung and 10 heart-lung transplant patients in the other 2), it is just possible that all of the values were within 1 standard deviation of the true mean of the total worldwide population of lung transplant patients. Another possibility may be that the differences are due simply to inter-patient variability. As mentioned previously, 2 patients, both of whom have cystic fibrosis and were documented poor-absorbers of CSA, received CSA every 8 hours. It is possible that the sampling time was not over at least one half-life as is required, resulting in half-life values that may be falsely low. This may be the case, as half-life values of 2.5 and 3.7 hours were obtained for these patients. A mean half-life of 10.6 hours, with a range of 5.9 to 25.5 hours was obtained i f the data from these two patients were excluded from analysis. Terminal elimination half-lives of 10.7, 7.4, 15.1, 12.8, 7.4, and 8.7 hours have been reported in renal transplant patients, and a terminal elimination half-life of 6.4 hours has been reported for cardiac transplant patients (Yee and Saloman 1992). The half-life of 9.52 hours obtained in the 96 current clinical study is consistent with these values. Possible differences could be due to inter-individual variablility, assay methodology, sampling fluid, and differences in pharmacokinetics among different transplant types. In general, the lung transplant patients who participated in this study were stable, long-term allograft recipients who were free of major adverse effects from CSA. One patient had received a kidney transplant necessitated by C S A nephrotoxicity, but is now stable and free of adverse effects from CSA. Another patient was on dialysis awaiting kidney transplantation, but otherwise was stable and free of major C S A adverse effects. Otherwise, there were no notable instances of major C S A adverse effects that negatively impacted quality of life. One patient had a case of CSA-associated tremor, and several others had well controlled hypertension and hyperlipidemia. Others had noticeable cases of hypertrichosis on the arms and torso. The mean AUCo-x obtained was 5183 ngxhr/mL (range 2272 to 9553 ngxhr/mL), and the mean dose-corrected AUCo-x was 2682 ngxhr/mL/mg/kg (range 1263 to 4237 ngxhr/mL/mg/kg). Values of 5944 (Kesten et al 1998a), 5318 (Kesten et al 1998b), and 7013 (Reynaid-Gaubert et al 1997) ngxhr/mL have been reported in lung transplant recipients. Our value was close to 2 of these reported values, but was approximately 2000 units less than the third. The difference may be due to differences in dose, or inter-patient variability in absorption or gastrointestinal cytochrome P-450 3A4 activity. A U C values of 4543 ngxhr/mL (van Mourik et al 1999) have been reported in stable liver transplant patients. A U C values of 5951 (Awni et al 1989), 2085 (Honcharik et al 1991), 3485 (Kovarik et al 1994a), and 5020 (Sketris et al 1994) ngxhr/mL have been reported in stable renal transplant patients. Inter-patient variability in absorption or metabolism cannot be discounted as reasons for the differences in A U C between the various transplant populations and the population of lung transplant patients studied in the current 97 clinical study. Also, there are well-documented differences in pharmacokinetic parameters among the various transplant types. The most likely reasons, however, are that larger doses are given to lung transplant patients in order to achieve the higher target CSA concentrations for this type of transplant, and also due to pharmacokinetic differences between SIM and NEO in the older studies. A U C values of 10405 (Gardier et al 1993), 2234 (Drewe et al 1992), 4471 (Gupta et al 1990), 7453 (Lindholm et al 1990), 3441 (Kovarik et al 1994), and 2981 (Mueller et al 1994) ngxhr/mL have been reported in healthy subjects. Despite several studies published involving A U C monitoring or LSS, the target A U C values for stable patients are uncertain even in the renal transplant population. A tentative "therapeutic A U C range" has been recently proposed for renal transplant patients (Belitsky et al 1999). They defined an A U C of 9500 to 11500 ngxhr/mL as the range in which the incidence of acute rejection was the lowest without an increased incidence of nephrotoxicity (Belitsky et al 1999). This range is not potentially applicable to our population of lung transplant patients, or even other types of stable transplant patients because the study was done in the immediate post-transplant period. In the study by Belitsky and colleagues (Belitsky et al 1999), concentration-time data were gathered after only 3 days on NEO therapy following 1 to 3 days of intravenous CSA. Only 5 blood samples were collected at 0, 1, 2, 3, and 4 hours post-dose, and A U C was extrapolated out to 12 hours. Another study, by the same investigative group (Mahalati et al 1999), proposed an AUCo-4 range of 4400 to 5500 ngxhr/mL, which corresponded to the same A U C 0 -12 range of 9500 to 11500 ngxhr/mL reported previously. As the study was conducted in the early post-transplant period, these ranges are of little interest to the stable, long-term transplant population. Because the lung transplant patients who participated in our clinical study were stable and free of major adverse effects, it would be reasonable to suggest that the target 98 A U C value may be in the 5000 to 6000 ngxhr/mL range for lung transplant patients. This suggestion is based on our results, as well as previously published values. However, much more study in this area is needed. It is important to note that maximum absorption of CSA occurred predictably within 2 hours post-dose. Even in the cases of the 2 documented poor absorbers of CSA, T m a x occurred within 2 hours post-dose. The parameter, AUMCo-t, simply defined the mean of a probability distribution, is of little use clinically, and was included mainly to calculate M RT. It is interesting to note that the mean M R T value obtained (21.82 hours) was much greater than the terminal elimination half-life value obtained. It is not unusual for the values to be different, as they have different meanings. By definition, terminal elimination half-life is the time it takes for half of the amount/concentration of drug to be eliminated from the body. Mean residence time, on the other hand, is the arithmetic average of the time that each molecule remains in the body. It can also be more simply defined as the time it takes for 63% of the drug to be eliminated from the body. Because it takes longer to remove 63% than 50% of the body, the M R T value should theoretically be longer than tm- However, the mean M R T value was approximately double the value for the mean terminal elimination half-life. The reasons for the result are uncertain. It is possible that the method used to calculate M R T somehow takes into account the time it takes for a given CSA molecule to exit the peripheral compartment. 4.2 Multiple Regression Analysis Multiple regression analysis is an excellent way of generating an equation that allows the calculation of values of a dependent variable using values from one or more independent variables. Following development of a regression equation, a coefficient of determination is 99 generated, which gives an indication of the degree of association of the independent and dependent variables. It can also give an indication of variance. For example, an r value of 0.999 explains 99.9% of the variance. Because it is not appropriate to use an r 2 value to recommend an LSS, no further related statistical calculations were required. Determination of a Pearson's correlation coefficient, for example, was inappropriate for the current analysis for 2 reasons. First, we wanted to be consistent with the methods of previous investigators, who did not perform this test. Second, correlation analysis does not assume a relationship between the concentration and A U C data, and because our goal was to determine a relationship between these data (via our recommended LSS). Multiple regression analysis, with A U C as the dependent variable, and whole blood CSA concentrations grouped by time as the independent variable, yielded 17 potential LSS capable of providing unbiased and precise estimates of A U C . To our knowledge, this is the first study that has developed an LSS from concentration-time data obtained specifically from lung transplant recipients. Even with restricting the multiple regression analysis to concentration-time data in the first 3 hours post-dose, 15 potential LSS for lung transplant recipients were delineated. Despite having a very high coefficient of determination, the 4-concentration LSS developed from all of the concentration-time data performed quite poorly in terms of the measures of predictive performance. This is an excellent example of why predictive performance should be evaluated before making a recommendation. Some of the earlier published LSS may not have been optimal, and it has been proposed that equations for some LSS were chosen solely on the basis of a high coefficient of determination value, rather than looking at other criteria such as prediction error (Gaspari et al 1998). The 4-concentration LSS developed from concentration-time data developed in the first 3 hours post-dose performed well in all calculated measures. However, the requirement for 4 100 blood samples was deemed not clinically feasible, and later comparison with the recommended LSS confirmed that there was no additional advantage in using that LSS over the recommended LSS. Two of the 3-concentration equations also performed well. There were no unfavorable significant differences in predictive performance with one exception. Specifically, our recommended LSS was significantly more biased than the LSS requiring Co, C i , and C 3 . We believed that the advantage of one less blood sample outweighed the disadvantage of a small, albeit significant, over-prediction of A U C . The recommended LSS outperformed the various 1- and 2-concentration LSS in all assessments of predictive performance. A n additional, practical benefit of using C i instead of Co is that it eliminates the problem of patients inadvertently taking their CSA dose before the trough sample is drawn, thus making it essentially useless in helping to guide therapy. Although the reason for inclusion of the 2 lung transplant patients who received CSA every 8 hours was adequately justified in a previous section, we wanted to be thorough. Multiple regression analysis was performed on concentration-time data obtained in the first 3 hours post-dose from the 7 lung transplant patients in Phase I who received CSA every 12 hours. The following equation was the result: A U C = 1.52xC, + 5.31xC3 + 267.89 An r 2 value of 0.971 was obtained for this equation. The concentration-time data from the 6 patients in Phase II was used to assess the predictive performance of this LSS. Percent pe ranged from -4.26 to 12.64%, which exceeded our previously defined limit of acceptable error in A U C prediction. Bias was 243 ngxhr/mL, and precision was 342 ngxhr/mL. When the predictive performance of this LSS was compared to the predictive performance of our recommended LSS, 101 there was no difference in bias. However, the recommended LSS for lung transplant patients was significantly more precise. These results further support our reasons for inclusion of the lung transplant patients receiving C S A every 8 hours. Exclusion of 1 of these patients in the development of the LSS made it less than optimal for predicting the A U C of all of the different patients encountered in routine clinical practice. This is not surprising as A U C values were similar to patients receiving CSA every 12 hours. Also, because the data analysis was restricted to the first 3 hours post-dose, when all patients were sampled the same regardless of dosing interval, the distinction between patients receiving C S A every 12 hours and patients receiving CSA every 8 hours is not discernable. 4.3 Comparisons Of Predictive Performance We chose to compare the predictive performance of our LSS with the predictive performance of other published LSS developed for other transplant populations, in part, to see how the A U C estimation in lung transplant patients was compared to A U C estimation by the other LSS. There is some evidence that LSS may be patient-specific, as studies that have evaluated LSS developed in other centers have found that they do not perform as well in terms of percent prediction error for A U C estimation (Gaspari et al 1993, 1997, and 1999). Because these studies did not assess predictive performance as suggested by Sheiner and Beal (Sheiner and Beal 1981), and because of the pharmacokinetic differences between transplant types, we believed that comparisons of predictive performance between the recommended LSS for lung transplant recipients and LSS derived from primarily renal transplant recipients would provide some additional insight. 102 No published LSS developed for other transplant populations had %pe within 10% of the actual A U C value. Because one study was close, and another study was within 20%, it is possible that these LSS would be acceptable for use in lung transplant recipients. However, as might be expected, the recommended LSS, developed specifically using concentration-time data from the population of interest, performed better in terms of %pe. In the majority of comparisons with other LSS, however, there were no differences in predictive performance. In two comparisons, the recommended LSS had more favorable predictive performance. In the only unfavorable comparison, the recommended LSS was significantly more biased than the LSS of Gaspari and colleagues (Gaspari et al 1997). However, this LSS required 3 blood samples, and we believe that the advantage of 1 less blood sample outweighed the disadvantage of a small (relative to total A U C ) but significant over-prediction of A U C . Thus, our recommended LSS is the most suitable for future use in our population of lung transplant recipients. However, before this LSS can replace routine trough level monitoring, prospective studies directly comparing the two monitoring strategies, looking at well-defined outcomes such as nephrotoxicity, episodes of rejection, and/or incidence of graft failure, are warranted. Currently, there are some prospective data evaluating outcomes in liver (Grant et al 1999), heart (Cantarovich et al 1999), and renal (Kelles et al 1999) transplant recipients. However, there are no published studies that compare the 2 monitoring methods in lung transplant patients. In addition, these 3 studies (Grant et al 1999, Cantarovich et al 1999, Kelles et al 1999) do not compare the two monitoring methods head-to-head. That is, 1 method (e.g., trough level monitoring) is utilized as a result of data gathering from another method (e.g., AUC0-4). There are not 2 separate groups (i.e., 1 group of transplant patients followed with trough level monitoring, and 1 group of patients followed using an LSS). 103 Our full evaluation of predictive performance fully tested the appropriateness of our recommended LSS for further prospective use. Earlier studies of this nature in other transplant populations recommended LSS based solely on high r values, which is completely inappropriate. Other studies went further and calculated %pe, which provided some additional, useful information. However, very few studies have determined these values and also carried out the full analysis of predictive performance according to the methods of Sheiner and Beal. 4.4 Conclusion A limited sampling strategy developed specifically for lung transplant patients is a precise and unbiased method of estimating actual A U C to allow the A U C monitoring of lung transplant patients. The LSS developed in this clinical study clearly outperformed other previously published LSS developed for non-lung transplant populations in terms of percent prediction error range. When predictive performance was compared, in general, there were no significant differences with 3 exceptions. Two of these were favorable, and the unfavorable one indicated an over-prediction in A U C that was small relative to total A U C . In addition, the LSS that was significantly less biased than ours required 3 blood samples, and thus would be more expensive and less clinically feasible than the LSS for lung transplant patients. The NEO formulation of CSA has a long terminal elimination half-life in lung transplant patients. The value of 9.52 hours was longer than previously published values for lung transplant patients. CSA has a large apparent volume of distribution (Vd/F = 3.64 L/kg), and an apparent oral clearance of 0.30 L/hr/kg. In addition, T m a x occurred predictably within 2 hours of dosage administration, with a mean T m a x of 1.35 hours. 104 In conclusion, based on the number of blood samples required, the patient wait time, coefficient of determination, %pe, and comparisons of predictive performance, the best clinically feasible LSS for cyclosporine A U C estimation for lung transplant patients requires 2 concentrations drawn at 1 and 3 hours post-dose: A U C = 1.46xCl + 5.36xC3 + 274.49 Few studies that have developed LSS have included appropriate comparisons of predictive performance, thus allowing statistical comparisons of the various LSS. Because we have done so with our LSS, our recommendation of the appropriateness of our LSS for lung transplant patients is more definitive than the recommendations of previous studies. The most obvious limitation to the clinical study is the relatively small sample size. A convenience sample size was the only option available to the investigative group. Given the limited number of potential subjects available, and based on previously published studies that I have read, 14 is actually a fairly large sample size. Several useful insights were discovered despite this limitation. However, it would have been interesting to compare the pharmacokinetic parameters between lung transplant patients with and without cystic fibrosis. Such statistical comparisons could have been made as is, but given that only 6 patients had cystic fibrosis, there would not have been sufficient power to detect a difference. Another limitation is the lack of blood samples, and therefore concentration-time data in the absorption phase. There was only 1 blood sample to characterize the upswing of the curve, and thus, it was not possible to calculate the absorption rate constant. This resulted in the loss of reporting of an oral dosing pharmacokinetic parameter. Knowledge of the absorption rate constant also would have allowed the calculation of the volume of distribution at steady state, thus giving an estimate of volume of distribution that did not depend on mean absolute 105 bioavailability, unlike the Vd/F parameter calculated in this study. Additional blood samples (e.g., at 15 minutes, 30 minutes, and 45 minutes post-dose) would have been useful. However, the primary concern was in obtaining enough samples to give an accurate calculation of A U C for multiple regression analysis, and not the calculation of pharmacokinetic parameters. Because this study enrolled stable, long-term lung transplant patients, the results provide no insight into another common area of transplant medicine: the early post-transplant period. This period is critical for lung transplant patients given the common occurrence of acute rejection, and the prognostic significance of those episodes in terms of chronic rejection. Pharmacokinetic data for lung transplant patients are relatively sparse in the literature, especially when compared with the pharmacokinetic data available for renal transplant patients. Some of the estimates of the mean pharmacokinetic parameters generated by this clinical study, such as apparent oral clearance, and especially terminal elimination half-life, can potentially aid in dosage adjustments, for example. Specifically, by using a combination of available concentration-time data, and literature values of pharmacokinetic parameters such as oral clearance, patient-specific parameters such as terminal elimination half-life and apparent volume of distribution could be calculated. These parameters could be used to calculate patient-specific dosages and dosing intervals for patients found to be not in the "therapeutic range". We have developed a feasible LSS available for use in lung transplant patients followed at V H H S C solid organ transplant clinic. Following publication in a transplantation journal, the LSS will be available to other centers. Further randomized, prospective studies comparing the LSS with trough level monitoring are required before the LSS can replace routine trough level monitoring as the standard therapeutic drug monitoring method. We have provided a potentially 106 useful tool for future changes in the practice of therapeutic drug monitoring of CSA in lung transplant patients. As stated in section 4.5, our LSS cannot be applied to lung transplant patients in the early post-transplant period, when prevention of acute rejection is of the utmost importance. In addition, the LSS was developed using lung transplant patients who were on a steady-state dosage of CSA. In the early post-transplant period, there are usually several dosage adjustments, and steady-states can often take 1 or more months to achieve. Currently, it is not known what A U C values to target. There has been an AUC0-12 and an AUC0-4 range proposed for renal transplant patients in the very early post-transplant period. However, this is of little use for our target transplant population. A possible future clinical study is one that attempts to define a "therapeutic A U C range" for lung transplant patients. Because the mean A U C obtained from the current clinical study was approximately 5200 ngxhr/mL, it would be reasonable to include a patient group of randomly assigned lung transplant patients who are followed by targeting an A U C range of 5000 to 6000 ngxhr/mL. There could be an additional group followed by targeting a higher A U C range, and one followed by targeting a lower A U C range. The most appropriate A U C range could then be determined by looking at surrogate markers of outcome, such as incidence of acute rejection, or incidence of nephrotoxicity. Once an appropriate "therapeutic A U C range" has been established, another clinical study that would allow the prospective comparison of the LSS for lung transplant patients with the current standard method of therapeutic drug monitoring (i.e., trough level monitoring) could be undertaken. Again, well defined surrogate markers of outcome could be assessed. Because the study needs to be of sufficient duration to assess the superiority of one method over another, 107 markers such as incidence of graft loss, number of hospital readmissions, length of stay, etc. could be assessed in addition to the usual markers. These 2 studies would then define the role of LSS in the routine monitoring of lung transplant patients. Additional clinical studies using the pharmacokinetic data obtained from our clinical study could also be performed. A prospective clinical study, with a blood sampling schedule similar to the current clinical study, could be performed to see i f any pharmacokinetic parameters could be associated with outcome, as has been done in renal transplant patients. We have recommended a LSS developed specifically for lung transplant patients, and to our knowledge, we are the first to do so. Our recommendation was based on evaluation of all possible criteria. Our investigative group now has a monitoring strategy that is useable in further clinical studies involving lung transplant patients to see i f A U C monitoring through the use of LSS has a place in the therapeutic drug monitoring of CSA in lung transplant patients. Because the results of our clinical study will be published in a transplantation journal, the monitoring strategy will be available for further clinical studies in lung transplant patients at other centers. Thus, the results of our study have contributed to, and filled a gap in, the existing transplantation and pharmacokinetic literature. 108 REFERENCES 1 0 9 Allison A C , Kowalski W Y , Muller CJ, et al. (1993). Mycophenolic acid and brequinar, inhibitors of purine and pyrimidine synthesis block the glycosylation of adhesion molecules. Transplant. Proc. 25, 67-70. Amante A J , Kahan BD. (1996) Abbreviated area-under-the-curve strategy for monitoring cyclosporine microemulsion therapy in the immediate posttransplant period. Clin. Chem. 42, 1294-1296. Amante A J , Meier-Kriesche H U , Schoenberg L , et al. (1997). A pharmacokinetic comparison of the corn oil versus microemulsion gelcap formulation of cyclosporin used de novo after renal transplantation. Transpl. Int. 10,217-222. Auch-Schwelk W, Duske E, Hink U , et al. (1994). Vasomotor responses in cyclosporin A-treated rats after chronic angiotensin blockade. Hypertension. 23, 832-837. Awni W M , Kasiske B L , Heim-Duthoy K , et al. (1989). Long-term cyclosporine pharmacokinetic changes in renal transplant recipients: effects of binding and metabolism. Clin. Pharmacol. Ther. 45,41-48. Awni W M , Heim-Duthoy K , Kasiske B L . (1990). Monitoring of cyclosporin by serial posttransplant pharmacokinetic studies in renal transplant patients. Transplant. Proc. 22, 1343-1344. Barone G, Chang CT, Choc M G , et al. (1996). The pharmacokinetics of a microemulsion formulation of cyclosporine in primary renal allograft recipients. Transplantation. 61, 875-880. Batiuk TD, Pazderka F, Halloran PF. (1995a). Cyclosporine-treated renal transplant patients have only partial inhibition of calcineurin phosphatase activity. Transplant. Proc. 27, 840-841. Batiuk TD, Pazderka F, Halloran PF. (1995b). Calcineurin activity is only partially inhibited in leukocytes of cyclosporine-treated patients. Transplantation. 59, 1400-1404. Belitsky P, Mahalati K , West K, et al. (1999). Influence of drug formulation on utilization and outcomes: Neoral and monitoring by sparse sample area under the curve. Transplant. Proc. 31, 1667-1668. Bennett W M , Porter GA. (1988). Cyclosporine-associated hypertension. Am. J. Med. 85, 131-133. Bergan A. (1997). Ancient myth, modern reality: a brief history of transplantation. J. Biocommun. 24, 2-9. Bertani T, Perico N , Abbate M , et al. (1987). Renal injury induced by long term administration of cyclosporine A to rats. Am. J. Pathol. 127, 569-579. 110 Best N G , Tan K K C , Trull A K , et al. (1996). Pharmacodynamics of cyclosporine in heart and heart-lung transplant patients. Transplantation. 62,1436-1441. Best N G , Trull A K , Tan K K C , et al. (1992). Blood cyclosporin concentrations and the short-term risk of lung rejection following heart-lung transplantation. Br. J. Clin. Pharmacol. 34, 513-520. Betschart J M , Virji M A , Shinozuka H . (1988). Cyclosporine A-induced alterations in rat hepatic glycogen metabolism. Transplant. Proc. 20, 880-884. Birt JK, Chandler M H H . (1990). Using clinical data to determine vancomycin dosing parameters. Ther. Drug Monit. 12, 206-209. Blumenstock DA, Lewis C. (1993). The first transplantation of the lung in a human revisited. Ann. Thorac. Surg. 56, 1423-1425. Bobadilla N A , Tapia E, Franco M , et al. (1994). Role of nitric oxide in renal hemodynamic abnormalities of cyclosporin nephrotoxicity. Kidney Int. 46, 773-779. Boncimino K , McMahon DJ, Addesso V , et al. (1999). Magnesium deficiency and bone loss after cardiac transplantation. J. Bone. Miner. Res. 14,295-303. Borel JF. (1983). Cyclosporine: historical perspectives. Transplant. Proc. 15, 2219-2229. i Borel JF, Di Padova F, Mason J, et al. (1989). Pharmacology of cyclosporine (Sandimmune). I. Introduction. Pharmacol. Rev. 41, 239-242. Borel JF, Feurer C, Gubler H U , et al. (1976). Biological effect of cyclosporin A : a new antilymphocytic agent. Agent. Act. 6,468-475. Borel JF, Feurer C, Magnee C, et al. (1977). Effects of the new anti-lymphocytic peptide cyclosporine A in animals. Immunol. 32,1017-1025. Bowles M H , Waters JB, Lechler RI, et al. (1996). Do cyclosporin profiles provide useful information in the management of renal transplant recipients. Nephrol. Dial. Transplant. 11, 1597-1602. Briel M , Chariot P. (1999). Muscle disorders associated with cyclosporine treatment. Muscle Nerve. 22,1631-1636. Browne BJ, Jordan MS, van Buren C, et al. (1994). Diet and cyclosporin A - pharmacokinetic comparison between Neoral and Sandimmune gelatin capsules. Transplant. Proc. 26, 2959-2960. Bunchman TE, Brookshire C A . Cyclosporine-induced synthesis of endothelin by cultured human endothelial cells. J. Clin. Invest. 88,310-314. I l l Burke C M , Glanville AR, Theodore J, et al. (1987). Lung immunogenicity, rejection, and obliterative bronchiolitis. Chest. 92, 547-549. Calne R Y , White DJG, Thiru S, et al. (1978). Cyclosporin A in patients receiving renal allografts from cadaver donors. Lancet. 2,1323-1327. Cantarovich M , Barkun JS, Tchervenkov JI, et al. (1998). Comparison of Neoral dose monitoring with cyclosporine trough levels versus 2-hr postdose levels in stable liver transplant patients. Transplantation. 66, 1621-1627. Carlesen E, Prydz H. (1987). Enhancement of procoagulant activity in stimulated mononuclear blood cells and monocytes by cyclosporine. Transplantation. 45, 543-548. Christians U , Sewing K F . (1993). Cyclosporin metabolism in transplant patients. Pharmacol. Ther. 57,291-345. Christians U , Kohlhaw K , Surig T, et al. (1995). Parallel blood concentrations of second-generation cyclosporine metabolites and bilirubin in liver graft recipients. Ther. Drug Monit. 17, 487-498. Christians U , Schlitt HJ, Bleck JS, et al. (1988). Measurement of cyclosporine and 18 metabolites in blood, bile and urine by high performance liquid chromatography (HPLC). Transplant Proc. 20,614-622. Clipstone N A , Crabtree GR. (1993). Calcineurin is a key signaling enzyme in T lymphocyte activation and the target of the immunosuppressive drugs cyclosporin A and FK506. Ann. N. Y. Acad. Sci. 696, 20-30. Cohen R G , Barr M L , Schenkel F A , et al. (1994). Living-related donor lobectomy for bilateral lobar transplantation in patients with cystic fibrosis. Ann. Thorac. Surg. 57,1423-1427. Cooney GF, Lum B L , Meligeni JA, et al. (1996). Pharmacokinetics of a microemulsion formulation of cyclosporine in pediatric liver transplant patients. Transplant. Proc. 28, 2270-2272. Cooney GF, Heifets M , Bell A , et al. (1994). Utility of pretransplantation cyclosporine pharmacokinetic studies. Ther. Drug Monit. 16,151-154. Cooper JD, Patterson G A , Grossman R, et al. (1989). Double-lung transplant for advanced chronic obstructive lung disease. Am. Rev. Respir. Dis. 139, 303-307. Copeland K R , Yatscoff RW, McKenna R M . (1990). Immunosuppressive activity of cyclosporine metabolites compared and characterized by mass spectroscopy and nuclear magnetic resonance. Clin. Chem. 36, 225-229. 112 Cropp CD, Davis G A , Ensom M H H . (1998). Evaluation of aminoglycoside pharmacokinetics in postpartum patients using Bayesian forecasting. Ther. Drug Monit. 20, 68-72. Curtil A , Robin J, Tronc F, et al. (1997). Malignant neoplasms following cardiac transplantation. Eur. J. Cardio. Thorac. Surg. 12, 101-106. Curtis JJ, Luke RG, Jones D, et al. (1988). Hypertension in cyclosporine-treated renal transplant recipients is sodium dependent. Am. J. Med. 85, 134-138. Cyclosporine monograph. (1998). In: AHFS Drug Information. Bethesda, M D : ASHP, Inc. Davis G A , Chandler M H H . (1996). Predictive performance and utility of a modified Cockcroft-Gault equation to estimate creatinine clearance in trauma patients. Am. J. Health-Syst. Pharm. 53, 1028-1032. De Groen PC, Aksamit A J , Rakela J, et al. (1987). Central nervous system toxicity after liver transplantation: the role of cyclosporine and cholesterol. N. Engl. J. Med. 317, 861-866. Dische FE, Neuberger J, Keating J, et al. (1988). Kidney pathology in liver allograft recipients after long-term treatment with cyclosporin A . Lab. Invest. 58, 395-402. Drewe J, Meier R, Vonderscher J, et al. (1992). Enhancement of the oral absorption of cyclosporin in man. Br. J. Clin. Pharmacol. 34, 60-64. Dumont FJ, Staruch M J , Koprak SL, et al. (1990). Distinct mechanism of suppression of murine T cell activation by the related macrolides FK-506 and rapamycin. J. Immunol. 144, 251-258. Dumont RJ, Ensom M H H . (2000). Methods for clinical monitoring of cyclosporine in transplant patients. Clin. Pharmacokinet. 38, 427-447. Eadon H, Rose M , O'Neill R, et al. (1995). A pharmacokinetic comparison of cyclosporin oral solution and cyclosporin capsules in heart and lung transplant recipients. Transpl. Int. 8, 35-40. Ellis C N , Gorsulowsky DC, Hamilton TA, et al. (1986). Cyclosporine improves psoriasis in a double-blind study. J.A.M.A. 256, 3110-3116. Ensom M H H , Davis GA, Cropp CD, et al. (1998). Clinical pharmacokinetics in the 21 s t century: does the evidence support definitive outcomes? Clin. Pharmacokinet. 34, 265-279. Fathman CG, Myers BD. (1992). Cyclosporine therapy for autoimmune disease. N. Engl. J. Med. 326, 1693-1695. Fernandez-Sola J, Campistol JM, Casademont J, et al. (1990). Reversible cyclosporin myopathy. Lancet. 335, 362-363. 113 Flechner S M , Payne WD, Van Buren C, et al. (1983). The effect of cyclosporine on early graft function in human renal transplantation. Transplantation. 36, 268-272. Foradori A C , Martinez L , Elberg A , et al. (1995). Preliminary pharmacokinetic evaluation of a new galenical formulation of oral cyclosporine A : Neoral T M . Transplant. Proc. 27, 1813-1814. Freeman DJ. (1991). Pharmacology and pharmacokinetics of cyclosporine. Clin. Biochem. 24, 9-14. Friman S, Backman L. (1996). A new microemulsion formulation of cyclosporin: pharmacokinetic and clinical features. Clin. Pharmacokinet. 30, 181-193. Fruman D A , Klee C B , Bierer BE , et al. (1992). Calcineurin phosphatase activity in T lymphocytes is inhibited by F K 506 and cyclosporin A . Proc. Natl. Acad. Sci. USA. 89, 3686-3690. Gardier A M , Mathe D, Guedeney X , et al. (1993). Effects of plasma lipid levels on blood distribution and pharmacokinetics of cyclosporin A . Ther. Drug Monit. 15, 274-280. Gaspari F, Anedda M F , Signorini O, et al. (1997). Prediction of cyclosporine area under the curve using a three-point sampling strategy after Neoral administration. J. Am. Soc. Nephrol. 8, 647-652. Gaspari F, Perico N , Signorini O, et al. (1998). Abbreviated kinetic profile in area-under-the-curve monitoring of cyclosporine therapy. Kidney Int. 54,2146-2150. Gaspari F, Ruggenenti P, Torre L, et al. (1993). Failure to predict cyclosporine area under the curve using a limited sampling strategy. Kidney Int. 44, 436-439. Girault D, Ffaloun A , Viard L . (1995). Sandimmun Neoral improves the bioavailability of cyclosporin A and decreases inter-individual variations in patients affected with cystic fibrosis. Transplant. Proc. 27, 2488-2490. Goldberg M , Cooper JD, Lima O, et al. (1983). A comparison between cyclosporin A and methylprednisolone plus azathioprine on bronchial healing following canine autotransplantation. J. Thorac. Cardiovasc. Surg. 85, 821-826. Goldstein DJ, Williams DL, Oz M C , et al. (1995). De novo solid malignancies after cardiac transplantation. Ann. Thorac. Surg. 60, 1783-1789. Goldstein DJ, Austin JHM, Zuech N , et al. (1996). Carcinoma of the lung after heart transplantation. Transplantation. 62, 772-775. Grace A A , Barradas M A , Mikhailidis DP, et al. (1987). Cyclosporine A enhances platelet aggregation. Kidney Int. 32,889-895. 114 Grevel J, Welsh M S , Kahan BD. (1989). Cyclosporine monitoring in renal transplantation: area under the curve monitoring is superior to trough-level monitoring. Ther. Drug Monit. 11, 246-248. Grevel J, Kahan BD. (1991). Abbreviated kinetic profiles in area-under-the-curve monitoring of cyclosporine therapy. Clin. Chem. 37,1905-1908. Grevel J, Napoli K L , Welsh MS, et al. (1991). Prediction of acute graft rejection in renal transplantation: the utility of cyclosporine blood concentrations. Pharm. Res. 8, 278-281. Grover FL , Fullerton D A , Zamora MR, et al. (1997). The past, present, and future of lung transplantation. Am. J. Surg. 173, 523-533. Gunnarsson R, Klintman G, Lundgren G, et al. (1984). Deterioration in glucose metabolism in pancreatic transplant recipients after conversion from azathioprine to cyclosporine. Transplant. Proc. 16,709-712. Gupta SK, Manfro RC, Tomlanovich SJ, et al. (1990). Effect of food on the pharmacokinetics of cyclosporine in healthy subjects following oral and intravenous administration. J. Clin. Pharmacol. 30, 643-653. Halloran PF, Madrenas J. (1991). The mechanism of action of cyclosporine: a perspective for the 90's. Clin. Biochem. 24, 3-7. Hardy JD, Webb WR, Dalton M L , et al. (1963). Lung transplantation in man. JAMA. 186, 1065-1074. Higenbottam T, Otulana B A , Wall work J. (1990). Transplantation of the lung. Eur. Respir. J. 3, 594-605. Hokanson JF, Mercier JG, Brooks GA. (1995). Cyclosporine A decreases rat skeletal muscle mitochondrial respiration in vitro. Am. J. Respir. Crit. Care Med. 151, 1848-1851. Honcharik N , Yatscoff RW, Jeffery JR, et al. (1991). The effect of meal composition on cyclosporine absorption. Transplantation. 52,1087-1089. Hosenpud JD, Novick RJ, Bennett L E , et al. (1996). The registry of the Internation Society for Heart and Lung Transplantation: thirteenth official report-1996. J. Heart Lung Transplant. 15, 655-674. Hosenpud JD, Novick RJ, Brren TJ, et al. (1994). The registry of the International Society for Heart and Lung Transplantation: eleventh official report- 1994. J. Heart Lung Transplant. 13, 561-570. 115 Johnston A , Sketris I, Marsden JT. (1990). A limited sampling strategy for the measurement of cyclosporine A U C . Transplant. Proc. 22,1345-1346. Johnston A , Kovarik JM, Mueller EA, et al. (1996). Predicting patients' exposure to cyclosporin. Transpl. Int. 9, S305-S307. Kahan BD. (1989). Drug therapy: cyclosporine. N. Engl. J. Med. 321, 1725-1738. Kahan BD. (1985). Individualization of cyclosporine therapy using pharmacokinetic and pharmacodynamic parameters. Transplantation. 40, 457-476. Kahan BD, Flechner S M , Lorber MI , et al. (1987). Complications of cyclsporine-prednisone immunosuppression in 402 renal allograft recipients exclusively followed at a single center from one to five years. Transplantation. 43,197-204. Kahan BD, van Buren CT, Flechner S M , et al. (1985). Clinical and experimental studies with cyclosporine in renal transplantation. Surgery. 97,125-140. Kahan BD, Welsh M , Rutzky L , et al. (1992). The ability of pretransplant test-dose pharmacokinetic profiles to reduce early adverse events after renal transplantation. Transplantation. 53,345-351. Kahan BD, Dunn J, Van Buren D, et al. (1995). Reduced inter- and intrasubject variability in cyclosporine pharmacokinetics in renal transplant recipients treated with a microemulsion formulation in conjunction with fasting, low-fat meals, or high-fat meals. Transplantation. 59, 505-511. Kahan BD, Welsh M , Schoenberg L , et al. (1996). Variable oral absorption of cyclosporine: a biopharmaceutical risk factor for chronic renal allograft rejection. Transplantation. 62, 599-606. Kahan BD, Kramer WG, Wideman CA, et al. (1986). Analysis of pharmacokinetic profiles in 232 renal and 87 cardiac allograft recipients treated with cyclosporine. Transplant. Proc. 18, 115-119. Kaplan MP, Lysz K , Rosenberg SA, et al. (1983). Supression of interleukin 2 production by methylprednisolone. Transplant. Proc. 15,407-412. Kasiske B L , Heim-Duthoy K , Rao K V , et al. (1988). The relationship between cyclosporine pharmacokinetic parameters and subsequent acute rejection in renal transplant recipients. Transplantation. 46,716-722. Keown P, Landsberg D, Halloran P, et al. (1996). A randomized, prospective multicenteer pharmacoepidemiologic study of cyclosporine microemulsion in stable renal graft recipients. Transplantation. 42, 1744-1752. 116 Keown P, Kahan BD, Johnston A , et al. (1998). Optimization of cyclosporine therapy with new therapeutic drug monitoring strategies: report from the international Neoral T D M advisory consensus meeting (Vancouver, November 1997). Transplant. Proc. 30, 1645-1649. Kesten S, Scavuzzo M , Laurin L , et al. (1998a). Conversion from standard cyclosporine to Neoral in lung transplant recipients. Transplant. Proc. 30,1895-1897. Kesten S, Scavuzzo M , Chaparro C, et al. (1998b). Pharmacokinetic profile and variability of cyclosporine versus Neoral in patients with cystic fibrosis after lung transplantation. Pharmacotherapy. 18, 847-850. Kinlen LJ . (1985). Incidence of cancer in rheumatoid arthritis and other disorders after immunosuppressive treatment. Am. J. Med. 78, 44-49. Kochi S, Takanaga H , Matsuo H , et al. (1999). Effect of cyclosporin A or tacrolimus on the function of blood-brain barrier cells. Eur. J. Pharmacol. 372, 287-295. Kovarik J M , Mueller E A , van Bree JB, et al. (1994a) Cyclosporine pharmacokinetics and variability from a microemulsion formulation- a multicenter investigation in kidney transplant patients. Transplantation. 58, 658-663. Kovarik J M , Mueller E A , Van Bree JB, et al. (1994b). Reduced inter- and intra-individual variability in cyclosporine pharmacokinetics from a microemulsion formulation. J. Pharm. Sci. 83, 444-446. Kovarik J M , Mueller E A , Van Bree JB, et al. (1994c). Within-day consistency in cyclosporine pharmacokinetics from a micremulsion formulation in renal transplant patients. Ther. Drug Monit. 16,232-237. Land W. (1987). Optimal use of cyclosporine in clinical organ transplantation. Transplant. Proc. 19, 130-135. Lanese D M , Conger JD. (1993). Effects of endothelin receptort antagonist on cyclosporine-induced vasoconstriction in isolated rat renal arterioles. J. Clin. Invest. 91, 2144-2149. Larner A J , Sturman SG, Hawkins JB, et al. (1994). Myopathy with ragged red fibres following renal transplantation: possible role of cyclosporin-induced hypomagnesemia. Acta. Neuropathol. 88, 189-192. Leader WG, Tsubaki T, Chandler M H H . (1994). Creatinine-clearance estimates for predicting gentamicin pharmacokinetic values in obese patients. Am. J. Hosp. Pharm. 51,2125-2130. Lemaire M , Fahr A , Maurer G. (1990). Pharmacokinetics of cyclosporine: inter- and intra-individual variations and metabolic pathways. Transplant. Proc. 22, 1110-1112. 117 Lemire J, Capparelli D, MacDonald D, et al. (1998). Estimated area-under-the-curve monitoring of Neoral in a stable pediatric renal transplant population: one-year experience. Transplant. Proc. 30, 1983-1984. Lima O, Cooper JD, Peters WJ, et al. (1981). Effects of methylprednisolone and azathioprine on bronchial healing following lung autotransplantation. J. Thorac. Cardiovasc. Surg. 82, 211-215. Lindholm A , Henricsson S, Dahlqvist R. (1990). The effect of food and bile acid admnistration on the relative bioavailability of cyclosporin. Br. J. Clin. Pharmacol. 29, 541-548. Lindholm A , Welsh M , Rutzky L , et al. (1993). The adverse impact of high cyclosporine clearance rates on the incidences of acute rejection and graft loss. Transplantation. 55, 985-993. Lindholm A, Kahan BD. (1993). Influence of cyclosporine pharmacokinetics, trough concentrations, and A U C monitoring on outcome after kidney transplantation. Clin. Pharmacol. Ther. 54,205-218. Lindholm A , Sawe J. (1995). Pharmacokinetics and therapeutic drug monitoring of immunosuppressants. Ther. DrugMonit. 17,570-573. Liu J, Farmer JD, Lane WS, et al. (1991). Calcineurin is a common target of cyclophilin-cyclosporin and FKBP-FK506 complex. Cell. 86, 807-815. Liu WT, Levy G A , Wong PY. (1995). Measurement of AM19 and other cyclosporine metabolites in the blood of liver transplant patients with stable liver function. Ther. Drug Monit. 17, 479-486. Macoviak JA, Oyer PE, Stinson EB, et al. (1985). Four-year experience with cyclosporine for heart and heart-lung transplantation. Transplant. Proc. 17,97-101. Mahalati K , Belitsky P, Sketris I, et al. (1999). Neoral monitorin by simplified sparse sampling area under the concentration-time curve. Its relationship to acute rejection and cyclosporine nephrotoxicity early after kidney transplantation. Makowka L , Chapman F, Cramer DV. (1993). Historical development of brequinar sodium as a new immunosuppressive drug for transplantation. Transplant. Proc. 25, 2-7. Mai H , Andreassian B , Pamela F, et al. (1989). Unilateral lung transplantation in end-stage pulmonary emphysema. Am. Rev. Respir. Dis. 140, 797-802. Marsh CL. (1999). Abbreviated pharmacokinetic profiles in area-under-the-curve monitoring of cyclosporine therapy in de novo renal transplant patients treated with Sandimmune or Neoral. Ther. Drug Monit. 21, 27-34. Martel RR, Klicius J, Galet S. (1975). Inhibition of the immune response by rapamycin, a new antifungal antibiotic. Can. J. Physiol. Pharmacol. 55, 48-51. 118 Massard G, Shennib H , Metras D, et al. (1993). Double lung transplantation for mechanically ventilated patients with cystic fibrosis. Ann. Thorac. Surg. 55, 1087-1092. McMillan M A . (1989). Clinical pharmacokinetics of cyclosporin. Phamracol. Ther. 42, 135-156. . Meier-Kriesche H U , Kaplan B, Brannan P, et al. (1998). A limited sampling strategy for the estimation of eight-hour Neoral areas under the curve in renal transplantation. Ther. Drug Monit. 20,401-407. Mercier JG, Hokanson JF, Brooks G A . (1995). Effects of cyclosporine A on skeletal muscle mitochondrial respiration and endurance time in rats. Am. J. Respir. Crit. Care Med. 151, 1532-1536. Messana J M , Johnson K J , Mihatsch MJ . Renal structure and function effects after low dose cyclosporine in psoriasis patients: a preliminary report. Clin. Nephrol. 43, 150-153. Meyer M , Bennet W, Udeajah J, et al. (1991). Evaluation of the efficacy of area under the curve (AUC) cyclosporine (CSA) monitoring, (abstract) J. Am. Soc. Nephrol. 2, 808. Meyer M M , Munar M , Udeaja J, et al. (1993). Efficacy of area under the curve cyclosporine monitoring in renal transplantation. J. Am. Soc. Nephrol. 4, 1306-1315. Michalets E L . (1998). Update: clinically significant cytochrome P-450 drug interactions. Pharmacotherapy. 18,84-112. Mihatsch M J , Ryffel B, Gudat F. (1995). The differential diagnosis between rejection and cyclosporine toxicity. Kidney Int. 52, S63-S69. Mihatsch M J , Kyo M , Morozumi K , et al. (1998). The side effects of ciclosporine-A and tacrolimus. Clin. Nephrol. 4 9 , 356-363. Mikhail G, Eadon H , Leaver N . (1997). A n investigation of the pharmacokinetics, toxicity, and clinical efficacy of Neoral cyclosporin in cystic fibrosis patients. Transplant. Proc. 29, 599-601. Mikhail G, Eadon H , Leaver N , et al. (1998). Comparison of Neoral and Sandimmun cyclosporines for de novo lung transplantation in cystic fibrosis patients. Transplant. Proc. 30, 1510-1511. Morgan WE, lima O, Goldberg M , et al. (1983). Improved bronchial healing in canine left lung reimplantation using omental pedicle wrap. J. Thorac. Cardiovasc. Surg. 85, 134-139. Morgan BJ , Lyson T, Scherrer U , et al. (1991). Cyclosporine causes sympathetically mediated elevations in arterial pressure in rats. Hypertension. 18,458-466. 119 Morris RE. (1993). New small molecule immunosuppressants for transplantation: review of essential concepts. J. Heart Lung Transplant. 12, S275-S286. Morris R G . (1997). Target concentration strategy for cyclosporin monitoring. Clin. Pharmacokinet. 32, 175-179. Mueller E A , Kovarik J M , van Bree JB, et al. (1994). Influence of a fat-rich meal on the pharmacokinetics of a new oral formulation of cyclosporine in a crossover comparison with the market formulation. Pharm. Res. 11,151-155. Muller-Ruchholtz W. (1999). Glances at the history of transplantation immunology. Transplant Proc. 31, 1443-1451. Najarian JS, Fryd DS, Strand M , et al. (1985). A single institution, randomized, prospective trial of cyclosporin versus azathioprine antilymphocyte globulin for immunosuppression in renal allograft recipients. Ann. Surg. 201,142-157. Nelems J M , Rebuck AS , Cooper JD, et al. (1980). Human lung transplantation. Chest. 78, 569-573. Novick RJ, Andreassian B, Schafers HG, et al. (1994). Pulmonary retransplantation for obliterative bronchiolitis: intermediate results for a North American-European series. J. Thorac. Cardiovasc. Surg. 107, 755-763. Novick RJ, Schafers HJ, Stitt L, et al. Seventy-two pulmonary retransplantations for obliterative bronchiolitis: predictors of survival. Ann. Thorac. Surg. 60, 111-116. Nussenblatt RB, Palestine A G . (1986). Cyclosporine: immunology, pharmacology and therapeutic uses. Surv. Ophthalmol. 31,159-169. Oka T, Yoshimura N . (1996). Immunsuppression in organ transplantation. Jpn. J. Pharmacol. 71, 89-100. Olyaei A J , de Mattos A M , Bennett W M . (1999). Immunosuppressant-induced nephropathy: pathophysiology, incidence and management. Drug Safety. 21, 471-488. Opelz G, Henderson R. (1993). Incidence of non-Hodgkin lymphoma in kidney and heart transplant recipients. Lancet. 342,1514-1516. Oyer PE, Stinson E B , Jamieson SW, et al. (1983). Cyclosporine in cardiac transplantation: a 2 V2 year follow-up. Transplant. Proc. 4, 2546-2452. Pai SY, Fruman D A , Leong T, et al. (1994). Inhibition of calcineurin phosphatase activity in adult bone marrow transplant recipients treated with cyclosporine A . Blood. 84, 3974-3979. 120 Palestine A G , Austin H A , Balow JE, et al. (1986). Renal histopathologic alterations in patients treated with cyclosporine for uveitis. N. Engl. J. Med. 314, 1293-1298. Penn I. (1986). Cancer is a complication of severe immunosuppression. Surg. Gyn. Obstet. 162, 603-610. Pfeffer M . (1984). Estimation of mean residence time from data obtained when multiple-dosing steady state has been reached. J. Pharm. Sci. 73, 854-856. Pham S M , Keenan RJ, Rao AS, et al. (1995). Perioperative donor bone marrow infusion augments chimerism in heart and lung transplant recipients. Ann. Thorac. Surg. 60, 1015-1020. Philip AT, Gerson B . (1998). Toxicology and adverse effects of drugs used for immunosuppression in organ transplantation. Clinic. Lab. Med. 18,755-765. Pickrell M D , Sawers R, Michael J. (1988). Pregnancy after renal transplantation: severe intrauterine growth retardation during treatment with cyclosporin A . Br. Med. J. 296, 825. Powles RL , Barret A J , Clink H , et al. (1978). Cyclosporin A for the treatment of graft-versus-host disease in man. Lancet. 2,1327-1331. Primmett D R N , Levine M , Kovarik J M , et al. (1998). Cyclosporine monitoring in patients with renal transplants: two- or three-point methods that estimate area under the curve are superior to trough levels in predicting drug exposure. Ther. Drug Monit. 20, 276-283. Ptachcinski RJ, Venkataramanan R, Rosenthal JT, et al. (1985). Cyclosporine kinetics in renal transplantation. Clin. Pharmacol. Ther. 38, 296-300. Quien R M , Kaiser B A , Dunn SP, et al. (1997). Calcineurin activity in children with renal transplants receiving cyclosporine. Transplantation. 64, 1486-1489. Radomski K M , Davis GA, Chandler M H H . (1997). General versus subpopulation values in Bayesian prediction of aminoglycoside pharmacokinetics in hematology/oncology. Am. J. Health-Syst. Pharm. 54, 541-544. Rajfer J, Sikka SC, Lemmi C, et al. (1987). Cyclosporine inhibits testosterone biosynthesis in the rat testis. Endocrinology. 121, 586-589. Randhawa PS, Yousem SA, Paradis IL, et al. (1989). The clinical spectrum, pathology, and clonal analysis of Epstein-Barr virus associated lymphoproliferative disorders after lung transplantation. Am. J. Clin. Pathol. 92, 177-185. Reitz B A , Wallwork JL, Hunt SA, et al. (1982). Heart-lung transplantation. Successful therapy for patients with pulmonary vascular disease. N. Engl. J. Med. 306, 557-564. 121 Remuzzi G, Perico N . (1995). Cyclosporine-induced renal dysfunction in experimental animals and humans. Kidney Int. 52, S70-S74. Reynaud-Gaubert M , Viard L , Girault D, et al. (1997). Improved absorption and bioavailability of cyclosporine A from a microemulsion formulation in lung transplant recipients affected with cystic fibrosis. Transplant. Proc. 29,2450-2453. Reznik V M , Jones K L , Durham B L , et al. (1987). Changes in facial appearance during cyclosporin treatment. Lancet. 1,1405-1407. Rhoney DR, Leader W G , Chandler M H H . (1993). Modified Michaelis-Menten equation for estimating unbound-phenytoin concentrations. Clin. Pharm. 12, 913-917. Roullet JB, Xue H , McCarron DA, et al. (1994). Vascular mechanisms of cyclosporine-induced hypertension in the rat. J. Clin. Invest. 93, 2244-2250. Rush D N . (1990). Cyclosporine toxicity to organs other than the kidney. Clin. Biochem. 24, 101-105. Savoldi S, Kahan BD. (1986). Relationship of cyclosporine pharmacokinetic parameters to clinical events in human renal transplantation. Transplant. Proc. 18, 120-128. Schafers HJ, Wagner TOF, Demertzis S, et al. (1992). Preoperative cortical steroids: a contradindication to lung transplantation? Chest. 102,1522-1525. Scherrer U , Vissing SF, Morgan BJ, et al. (1990). Cyclosporine-induced sympathetic activation and hypertension after heart transplantation. N. Engl. J. Med. 323, 693-699. Schroeder TJ, Hariharan S, First MR. (1994). Relationship between cyclosporine bioavailability and clinical outcome in renal transplant recipients. Transplant. Proc. 26, 2787-2790. Serafinowicz A , Gaciong Z, Backowska T, et al. (1996). Limited sampling strategy to estimate exposure to cyclosporine A in renal allograft recipients treated with Sandimmun-Neoral. Transplant. Proc. 28, 3138-3139. Serino F, Citterio F, Pozzetto U , et al. (1994). Abbreviated three-point kinetic profile in the 12-hour area under the curve for pharmacokinetic monitoring of cyclosporine. Transplant. Proc. 26, 2807-2808. Shah A K . (1999). Cyclosporine A neurotoxicity among bone marrow transplant recipients. Clin. Neuropharmacol. 22, 67-73. Sheiner L B , Beal SL. (1981). Some suggestions for measuring predictive performance. J. Pharmacokinet. Biopharm. 9, 503-512. 122 Shihab FS. (1996). Cyclosporine nephropathy: pathophysiology and clinical impact. Seminar. Nephrol. 16,536-547. Sigma Product Information Sheet for C3662: Cyclosporin A . (2000). http//www. sigma. sial .com/sigma/proddata/c3662.htm. Sketris IS, Lawen JG, Beauregard-Zollinger L , et al. (1994). Comparison of the pharmacokinetics of cyclosporine Sandimmune with Sandimmune Neoral in stable renal transplant patients. Transplant. Proc. 26,2961-2963. Stahelin H . (1986). Historical background. Prog. Allergy. 38, 19-27. Starnes V A , Barr M L , Cohen RG. (1994). Lobar transplantation- indications, technique and outcome. J. Thorac. Cardiovasc. Surg. 108,403-410. Starzl TE, Klintmalm G B G , Porter K A , et al. (1981). Liver transplantation with the use of cyclosporin A and prednisone. N. Engl. J. Med. 305, 266-269. Starzl TE, Iwatsuki S, Van Thiel D H , et al. (1983). Report of Colorado-Pittsburgh liver transplantation studies. Transplant. Proc. 4, 2582-2585. Suzuki S, Amemiya H . (1990). 15-deoxyspergualin- a novel immunosuppressant, experimental studies and clinical trial. Transplant. Immunol. Lett. 7, 17. Svendsen U , Larsen K , Allermand H . (1995). Neoral conversion study: shift from Sandimmune classic formulation to Sandimmune Neoral in heart and lung transplant patients. Transplant. Proc. 27,3477. Svendsen U G , Larsen KR, Allermand H . (1996). Neoral conversion study: shift from Sandimmune classic formulation to Neoral in heart and lung transplant patients. Transplant. Proc. 28,2281. Tan K K C , Trull A K , Hue K L , et al. (1993). Pharmacokinetics of cyclosporine in heart and lung transplant candidates and recipients with cystic fibrosis and Eisenmengers's syndrome. Clin. Pharmacol. Ther. 53, 544-554. Tan K K , Trull A K , Uttridge JA, et al. (1995). Relative bioavailability of cyclosporin from conventional and microemulsion formulations in heart-lung transplant candidates with cystic fibrosis. Eur. J. Clin. Pharmacol. 48, 285-289. The Canadian Multicentre Transplant Study Group. (1986). A randomized clinical trial of cyclosporine in cadaveric renal transplantation. Analysis at three years. A f . Engl. J. Med. 314, 1219-1225. Toronto Lung Transplant Group. (1986). Unilateral lung transplantation for pulmonary fibrosis. N. Engl. J. Med. 314,1140-1145. 123 Tredger J M . (1995). Using cyclosporine Neoral immediately after liver transplantation. Ther. Drug Monit. 17, 63 8-641. Trull A , Steel L , Sharpies L , et al. (1999). Randomized, trough blood cyclosporine concentration-controlled trial to compare the pharmacodynamics of Sandimmune and Neoral in de novo lung transplant patients. Ther. Drug Monit. 62,1436-1441. Trull A K , Tan K K C , Tan L , et al. (1995). Absorption of cyclosporin from conventional and new microemulsion oral formulations in liver transplant recipients with external biliary diversion. Br. J. Clin. Pharmacol 39, 627-631. Trulock EP. (1997). Lung transplantation. Am. J. Respir. Crit. Care Med. 155, 789-818. Tsang VT, Johnston A , Heritier F, et al. (1994). Cyclosporin pharmacokinetics in heart-lung transplant recipients with cystic fibrosis: effects of pancreatic enzymes and ranitidine. Eur. J. Clin. Pharmacol. 46, 261-265. Tsunoda S M , Aweeka FT. (1996). The use of therapeutic drug monitoring to optimize immunosuppressive therapy. Clin. Pharmacokinet. 30,107-140. Vancouver Hospital Hospital and Health Sciences Centre. (1999). Laboratory standard procedure manual. Deparment of Pathology and Laboratory Medicine. van Mourik IDM, Melendez H V , Thomsom M , et al. Efficacy of Neoral in the immediate postoperative period in children post-liver transplantation. Liver Transplant. Surg. 4,491-498. van Mourik IDM, Thomson M , Kelly DA. (1999). Comparison of pharmacokinetics of Neoral and Sandimmune in stable pediatric liver transplant recipients. Liver Transplant. Surg. 5, 107-111. Veith FJ, Koerner SK. (1974). Problems in the management of human lung transplant patients. Vase. Surg. 8,273-282. Wasan K M , Pritchard PH, Ramaswamy M , et al. (1997). Differences in lipoprotein lipid concentration and composition modify the plasma distribution of cyclosporine. Pharm. Res. 14, 1613-1620. Welch LP, Leader WG, Chandler M H H . (1993). Predicting vancomycin pharmacokinetics by using aminoglycoside pharmacokinetics. Clin. Pharm. 12,909-913. Wenger R. (1983). Synthesis of cyclosporine and analogues: structure, activity, relationships of new cyclosporine derivatives. Transplant. Proc. 15,2230-2241. 124 Wilczek W H , Berglin E, Blohme I, et al. (1997). The Swedish nationwide experience of converting transplant patients treated with Sandimmun to the new galenic formulation Sandimmun-Neoral. Transplant. Proc. 29, 280-283. Wu G, Baraldo M , Pea F, et al. (1995). Effects of different sampling strategies on predictions of blood cyclosporine concentrations in haematological patients with multidrug resistance by Bayesian and non-linear least squares methods. Pharmacol. Res. 32, 355-362. Yee GC, Saloman DR. (1992). Cyclosporine. In: Applied pharmacokinetics: principles of therapeutic drug monitoring. Evans WE, Schentag JJ, Jusko WJ (eds.), 3 r d ed. Vancouver, W A : Applied Therapeutics, Inc., 28-1-28-40. Yee GC, Kennedy M S , Storb R, et al. (1984). Effect of hepatic dysfunction on oral cyclosporin pharmacokinetics in marrow transplant patients. Blood. 64,1277-1279. York LJ , Qualtiere LF . (1990). Cyclosporine abrogates virus-specific T cell control of E B V -induced B-cell lympho-proliferation. Viral. Immunol. 3,127-136. Zaldonis DB, Keenan RJ, Pham S M , et al. (1998). Neoral conversion in stable thoracic transplant patients leads to dose reduction. Transplant. Proc. 30, 1158-1159. Zenke G, Baumann G, Wenger R, et al. (1993). Molecular mechanisms of immuosuppression by cyclosporins. Ann. N. Y. Acad. Sci. 685, 330-335. 

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:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0099550/manifest

Comment

Related Items