Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

An automated chemistry module for the (18F)FDG production Wu, Jason Shao-Chun 2002

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

Item Metadata


831-ubc_2002-0289.pdf [ 5.14MB ]
JSON: 831-1.0085130.json
JSON-LD: 831-1.0085130-ld.json
RDF/XML (Pretty): 831-1.0085130-rdf.xml
RDF/JSON: 831-1.0085130-rdf.json
Turtle: 831-1.0085130-turtle.txt
N-Triples: 831-1.0085130-rdf-ntriples.txt
Original Record: 831-1.0085130-source.json
Full Text

Full Text

A N AUTOMATED CHEMISTRY M O D U L E FOR T H E [ F]FDG PRODUCTION 18  By JASON S H A O - C H U N W U B.Sc, The University of British Columbia, 2000 A THESIS SUBMITTED IN P A R T I A L F U L F I L M E N T OF T H E REQUIREMENTS FOR T H E D E G R E E O F M A S T E R OF SCIENCE In T H E F A C U L T Y OF G R A D U A T E STUDIES (Department of Physics and Astronomy) We accept this theSiZaT^mformrn^ to d j e ^ q m ^ e d ^ s ^ —  T H E UNIVERSITY OF BRITISH C O L U M B I A February 2002 © Jason Shao-Chun Wu, 2002  In presenting this  thesis in partial  degree at the University of  fulfilment of  the  requirements  for  an advanced  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 department  or  by  his  or  her  representatives.  It  is  understood  that  copying  my or  publication of this thesis for financial gain shall not be allowed without my written permission.  Department The University of British Columbia Vancouver, Canada Date  DE-6 (2/88)  Abstract This thesis presents the design, chemical procedures, operation, optimization, and performance of a [ F ] F D G synthesis module. Together with the E B C O / T R I U M F TR 14 ,8  cyclotron in C.H.U.S., this module forms an integrated [ F]FDG production line, 18  employing aqueous [ F]fluoride as precursors and nucleophilic reaction as the 18  fluorination step. 18  Among many short-lived positron emitting radionuclides,  F is the most  important and widely used nuclide due to its convenient half-life (ti/2 = 109 minutes). The emergence of [ F]FDG as the most popular PET radiopharmaceutical has 18  necessitated the development of routine methods of its production in the PET radiochemistry clinic. A synthesis module is required to produce large quantities of [ F]FDG reliably, efficiently, and economically. In order to achieve this goal, the fluid 18  transfer system, the heating/cooling mechanism, and the automated control of the module are the key design components. A modified chemistry module based on an E B C O / T R I U M F design is constructed. After extensive cold tests, 40 hot tests are conducted in C.H.U.S. to optimize the module. Proper cleaning and operation procedures are developed and the optimum reaction parameters are established. The results are satisfactory, while the consistency in the yields displays a great reliability.  ii  Table of Contents Abstract  ii  Table of Contents  iii  List of Tables  v  List of Figures  vi  Chapter 1  Introduction  1  Chapter 2  Theory  3  2.1  2.2  Chapter 3  Nuclear and Target Physics  3  2.1.1  Q-Value and Threshold Energy  3  2.1.2  Reaction Cross-Sections  4  2.1.3  Stopping Power and Range  5  2.1.4  Production of Radionuclides  5  Chemical Synthesis of [ F]FDG  10  2.2.1  Electrophilic Fluorination  11  2.2.2  Nucleophilic Fluorination  13  2.2.3  Clean-up Columns  16  l8  Experimental & Design  18  3.1  TR-14 Cyclotron, Targets, and Operating Interface  18  3.1.1  TR-14 Cyclotron  18  3.1.2  Targets  19  3.1.3  Computer Control S ystem  22  3.2  Chemistry Module & Peripherals  22  3.2.1  22  Overview  iii  3.2.2  Construction  29  3.2.3  Control Software  39  3.2.4  Fluid Transfer  43  3.2.5  Heat Control  47  3.2.6  Accessories  50  3.2.7  Reagent Preparation  53  3.2.8  Cleaning  54  Chapter 4  Results & Product Analysis  56  Chapter 5  Conclusion  62  Bibliography  64  Appendix  66  iv  List o f Tables Table 2. 1: Range of proton in water and production rate of F from a 100% 18  , 8  0-  enriched water target as a function of proton energy  9  Table 2. 2: Effect of Fluorinating Agent on the Percentage of 2-FDG/ 2 - F D M  12  Table 3.1: Summary of the Hammacher method  23  Table 3. 2: The summary of [ F]FDG synthesis sequencing steps and their optimum l8  parameters Table 3. 3: Aseptic cleaning sequencing steps  27 54  List of Figures Figure 2.1: The excitation function of the 0(p,n) F reaction  4  Figure 2.2: Saturated yield of the 0(p,n) F reaction as a function of energy  8  ,8  18  18  18  Figure 2.3: The production rate curve of the 0(p,n) F reaction as a function of proton 18  18  energy  10  Figure 2.4: Reaction of 3,4,6-tri-O-acetyl-D-glucal with fluorinating agents  11  Figure 2.5: Nucleophilic fluorination in the synthesis  13  Figure 2.6: Reaction scheme for the synthesis of [ F]FDG  15  Figure 3.1: The base plate of the reaction vessel assembly  32  Figure 3.2: The jacket of the reaction vessel assembly  33  Figure 3.3: The P E E K header of the reaction vessel assembly  36  Figure 3.4: The [ 0]H20 recovery vial holder  37  Figure 3.5: The electrical schematics of the [ F]FDG synthesis system  38  Figure 3.6: The control screen showing the first water wash  41  Figure 3.7: The production report  42  Figure 3.8: R C circuit used to prevent valves from overheating  44  Figure 3.9: The wiring diagram of a PCM-SSR system  48  Figure 3.10: The heating and cooling curves  49  Figure 3.11: The radiation detector assembly  51  Figure 3.12: The radiation amplifier circuit  52  Figure 4.1: Summary of [ F]FDG yield  56  18  18  18  18  Figure 4.2: Dependence of forming [ F]FDG on sodium hydroxide concentration and 18  reaction time during the alkaline hydrolysis  vi  59  Figure A . l : Front-side view of the chemistry module  66  Figure A.2: Back-side view of the chemistry module  67  Figure A.3: The P E E K header and the reaction vessel  67  Figure A.4: The reaction vessel assembly  68  Figure A.5: The control box  ..68  vii  Chapter 1. Introduction  Chapter 1 Introduction Positron emission tomography (PET) has become the most popular imaging technique for many diagnostic applications in nuclear medicine in the last decade. It is the only non-invasive method that permits quantitative dynamical measurements of radioactivity, making it an extremely powerful tool for studying physiological, pharmacological, and biochemical phenomena in oncology and neurology. PET imaging is based on the simultaneous detection of two 0.511 M e V photons, at almost 180° to each other, released from the annihilation of a positron and an electron. A suitable amount of a tracer substance labelled with a positron-emitting nuclide such as " C , 0 , or F is 1 5  1 8  administered to the patient. Since these radionuclides do not alter the physiological properties of the tracer substance, depending on the nature of the tracer, images can be constructed around the targeted organs or cells. For instance, [ F]2-deoxy-2-fluoro-Dl8  glucose, or simply [ F]FDG, has one of its hydroxyl groups replaced by F and still l8  1 8  pertains some of its original biological functions, such that it localizes in the same way as a normal glucose. Typically, clinical PET studies require scanning time in the order of one to several hours. Consequently, the half-life of the radiotracer should be comparable to the time of measurements.  18  F , currently the most widely used PET nuclide, has a convenient half-  life of 109.6 minutes and decays nearly exclusively via positron emission. In addition, it can be routinely produced in high yield by proton bombardment of 0 enriched water or 1 8  oxygen gas. The threshold energy of the nuclear reaction Q(p,n) F is about 2.5 MeV, a 18  1  18  Chapter 1.  Introduction  relatively low value that can be easily achieved by a small medical cyclotron.  , 8  F , with  activity ranging from 1 to 2 C i , produced at the end of each bombardment, will then be extracted from the target as fluorine [ F]F2, hydrogen fluoride [ F]HF, or fluoride 18  18  [ F]F, depending on the subsequent needs of the radiopharmaceutical synthesis. 18  Among many F-labelled tracers, [ F]FDG is the most important PET l8  18  radiopharmaceutical due to its use in measuring glucose metabolism. Because of the high glucose uptake by many types of aggressive tumor cells, [ F]FDG provides a means to 18  precisely localize tumors (Adler et al. 1991). In recent years, the increased efforts in cancer treatment have greatly broadened the demand for [ F]FDG, but the production of 18  this tracer is somewhat unreliable. The chemical yield in Centre Hospitalier Universitaire de Sherbrooke (C.H.U.S.) typically varies from 70% to 30% decay-corrected, giving activities ranging from 250 mCi to 600 mCi. Ideally, the amount of [ F]FDG needed for 18  a full working day is about 400mCi to 500 mCi; therefore, occasionally, some scheduled scans will have to be cancelled due to a low yield, creating inconvenience to both the patient and the clinic. The goal of this study is to develop a reliable and fast [ F]FDG production, using 18  the existing medical cyclotron and targets in C.H.U.S. and a new chemistry module. This new module should be remotely monitored and controlled; it should also have the capability of giving a high, consistent uncorrected yield of more than 40% (400mCi of [ F]FDG from 1 C i of F). With a careful evaluation and modification of published 18  18  procedures, chemical steps and reaction parameters optimal for the module will be established. Other features such as preparing and cleaning the module quickly and keeping the construction costs low are also the objectives of this study.  2  Chapter 2. Theory  Chapter 2 Theory In this chapter, some physics and chemistry concepts will be briefly discussed. These theories are important in the understanding of the design and choices made for the targets and the chemistry module.  2.1  Nuclear and Target Physics  2.1.1 Q-Value and Threshold Energy By the conservation of total mass-energy, the difference in the initial mass and final mass of a nuclear reaction is called Q-value (Krane, p.381). More explicitly: Q = [(Sum of atomic masses of reactants) - (Sum of atomic masses of Products)]- c  This Q-value is the energy released in a nuclear reaction. Clearly, for a negative Q-value, energy must be put into the system to initiate the reaction. In the scope of this study, this energy input is given by the incident particle beams generated from a small medical cyclotron. Furthermore, a certain amount of additional kinetic energy is needed since there's a transfer of momentum from the incident particle to the target nucleus. For an incident particle of mass m and a target nucleus of mass Af, the minimum kinetic energy or the threshold energy required for the reaction to take place is defined as (Krane, p.382): T n A a  =-  Q  ! H l K  (2  .1)  M  Via this equation, the Q-value of 0(p,n) F reaction is calculated to be -2.4 MeV. 18  18  3  Chapter  2.  Theory  2.1.2 Reaction Cross-Sections Reaction cross section o is a measure of the relative probability for a particular nuclear reaction to occur in a collision. It is different from the total cross section a  t >  which represents the probability of all possible nuclear reactions occurring in a collision. The unit of cross section is the barn. 1 barn = 10" cm 24  2  Every nucleus exhibits a specific cross-section for each different nuclear reaction and the values of a depend on the kinetic energy of the incident beam: A useful plot having reaction cross-sections as a function of incident energies serves a good guide for choosing an optimal incident energy. This plot is called the excitation function of the reaction.  18  0(p,n) F 18  Excitation Function  800 -,  Proton Energy (MeV)  Figure 2. 1:  The excitation function of the 0(p,n) F reaction 18  18  (Ruth and Wolf 1979). This figure is recreated from the data set given in the original paper.  4  Chapter 2. Theory  2.1.3 Stopping Power and Range The theoretical relationship between the range of an energetic particle in matter and the energy of the projectile particle can be obtained from a quantum mechanical calculation of the collision process. The calculation, also called Bethe-Bloch formula, is a differential equation involving many parameters such as the density of the stopping material, the velocity of the projectile, and the ionization potential. The outcome of this equation is the magnitude of the energy loss per unit length (dE/dx) or the stopping power of the target. The total amount of energy loss in a given amount of target thickness can then be evaluated via integration. In addition, the mean range of the incident particles in the target can also be calculated by integrating (dE/dx)" over the energies of the incident particle. 1  (2.2) where T is the kinetic energy of the incident particle (Krane, p. 195). If the entire incident beam is stopped in a target, the target is said to be a thick target. Conversely, if a portion of the beam goes through the target, it is then a thin target. Since the incident beam penetrates a thin target, other materials outside the target may become radioactive. In the confined and crowded space around the target changer in a cyclotron, it is highly desirable to minimize the number of incident particles passing through the target. Consequently, thick targets are chosen in this study.  2.1.4 Production of Radionuclides  5  Chapter 2. Theory Production Rate When a particle beam strikes a thin target, the number of nuclei produced per unit volume per unit time R can be expressed as p  R = NooIAx  (2.3)  P  where / is the beam current, tris the reaction cross section, and Ax is the thickness of the thin target, and No is the number of target nuclei per unit volume, which can be calculated as No = -?-Ao M  (2.4)  where p and M are the density and the molecular weight of target material respectively, and Ao is Avogadro's number. To convert to the number per gram per unit time, simply divide Equation 2.3 by p: r - nooIAx  (2.5)  If the product nuclei decay to a stable nuclide with a decay constant A, the build-up of product nuclei over an irradiation time t is given by (Pavan 1997) irr  n{tirr) = HOOlAx[l  - eXp(-  (2.6)  Atirr)\  A The number of radionuclides produced at a certain time t after the end of bombardment (EOB) can be expressed as n(tirr, t) = nooIAx  & ~ exp(- Atirr)[-exp(-  At)  A  Radioactive Decay Radioactive decay at a certain time t after E O B can be calculated by differentiating equation 2.7:  6  (2.7)  Chapter 2. Theory - exp(- / U , - r r ) ] - exp(- At)  - — = noo/Ajc[l  dt  (2.8)  Equation 2.8 is only a statistical result; one cannot predict precisely the time at which the decay of a specific atom will occur. However, the decay process is exponential and, given the huge number of nuclei involved in the reaction, the total radioactivity can be assumed to be exact. The unit used to measure radioactivity is the becquerel (Bq): 1 Bq = 1 decay per second Another convenient and commonly used unit is the curie (Ci), which is equivalent to the activity of one gram of radium: 1 C i = 3.7x 1 0 B q 1 0  Since the activities used in this study are relatively large, the curie will be the unit of choice for the remainder of this report. Thick Target Activity In a thick target, the incident beam energy is degraded to the threshold energy Ethr-  At this energy, the incident particles will not activate nuclei in the target and the  total activity produced at EOB can be calculated by integrating equation 2.8 over the energy range. Setting Ax = A E / (dE / dx), the activity produced can be written as:  AEOB =  /no[l - exp(- A  o{E) ' dE "" dE/dx  t i r M  "  K  (2.9)  iE  With an irradiation time much longer than the half-life, Tm, the exponential term in equation 2.9 vanishes, making the activity produced independent of the irradiation time. In this case, the target reaches its maximum activity or the saturation activity: o(E)  A~ = ln*r--22±<iE **  dE/dx  J  7  (2.10)  Chapter 2. Theory  Saturation Yield The saturation yield,  Y , sat  is the maximum activity produced per unit current at a  certain incident energy. Using the equation of Y much larger than T1/2, Y  s a t  in (Solin 1988) and taking tin-to be  can be expressed as  s a t  Asal Ysat — •  (2.11)  where / is the irradiating beam current The saturation yield of the 0(p,n) F reaction is shown in Figure 2.2 (Bishop et al. 18  18  1996a). s 18/  O(p,n) F 18  Saturated Yield in a Thick-Target  Energy (MeV)  Figure 2.2: Saturated yield of the  l8  0(p,n) F reaction as a function of l8  energy. Yields of two other common reactions producing F are also 18  shown. Evidently, at a beam energy greater than 5 MeV, the yield curve of the O(p,n) F reaction is substantially higher than the other two, making it ls  18  the most effective and most widely used reaction for the F production. l 8  8  Chapter 2. Theory  Production Rate Production Rate, R d, is the radioactivity at EOB divided by the integrated beam pro  current and the irradiation time:  I-tirr  Table 2.1 shows the values of the theoretical production rate for the O(p,n) F reaction ls  18  on a 100% enriched O water target and the corresponding proton beam energies, E , s  p  (Guillaume 1991). Using the data set in Table 2.1, plotting R d as a function of Ep as p r 0  shown in Figure 2.3, it becomes clear that irradiating at the beam energy of -14 M e V is the most economical, since the value of R d increases very slowly passing the 15 M e V pr0  mark. Ep  (MeV)  Range in Water (mg/cm ) 23.3 34.5 47.6 62.7 79.6 98.3 119 140 165 190 216 246 277 312 345  Rprod  (mCi/u.Ah)  2  4 5 6 7 8 9 10 11 12 13 14 15 16 17 18  2.2 6.4 16.2 23.2 28.9 33.8 39.1 43.8 48.4 52.2 54.5 56.4 58 59.5 60.5  Table 2. 1: Range of proton in water and production rate of as a function of proton energy. Note that the values of R  9  prod  l 8  F from a 100% O-enriched water target ls  are obtained with t^ < 1 hour.  Chapter 2. Theory  Ep(MeV) Figure 2. 3: The production rate curve of the 0(p,n)' F reaction as a function of 18  8  proton energy. The curve starts to plateau at energies above 15 MeV.  2.2  Chemical Synthesis of [ F]FDG 18  To develop a new chemistry module, it is important to investigate the chemistry of each [ F]FDG production method. The method chosen should be simple, reliable, and 18  fast with minimum purification and quality control at the end of the synthesis. In general, after irradiation, F ions are formed in the target. These ions are made 18  into a desirable fluorination reagent via some chemical processes. Depending on the type of fluorination processes chosen, a precursor of glucose is added into the reaction vessel and F is labeled onto the sugar. After the completion of labeling, acid or base 1 8  hydrolysis is conducted to make the labeled sugar into [ F]FDG. The most crucial 18  chemical step in the synthesis is the fluorination of the sugar since it determines the type of fluorination reagents and the purity and yield of the final product.  10  Chapter 2. Theory  2.2.1 Electrophilic Fluorination The electrophilic reaction involves the formation of F electrophilic fluorination. 18  agent and then the addition across the double bond of a glucal as shown in Figure 2.4 (Bishop et al. 1996b).  (TAG)  Figure 2. 4:  (2-FDQ)  (2-FDM)  Reaction of 3,4,6-tri-O-acetyl-D-glucal with fluorinating agents. Process (1) indicates  the actual fluorination and (2) is the subsequent acid hydrolysis.  The major concern in this synthetic method is that the product is a mixture of FDG and FDM. FDM, 2-fluoro-deoxy-D-mannose, is structurally very similar to FDG, making the subsequent separation step difficult. The formation of FDM is inevitable in the electrophilic reaction since the fluorination agent can randomly add to either side of the double bond. The original synthesis of [ F]FDG was performed by reacting 18  triacetylglucal (TAG, 1) with [ F]F in Freon, which gives a mixture of [ F]FDG (2) 18  18  2  and [ F]FDM (3) in the ratio of 6:4 (Tewson 1989). Although this ratio is reported to be 18  slightly higher in the later study as shown in Table 2.2 (Bishop et al. 1996b), more than a quarter of the activity is lost as [ F]FDM. l8  Also shown in Table 2.2, the ratios of [ F]FDG and [ F]FDM resulting from the 18  18  reaction with TAG are sensitive to the nature of the fluorination agents. Among them, F-acetylhypofluorite, AcOF, appears to have very high stereo-selectivity in the  18  11  Chapter 2. Theory  eletrophilic addition forming [ F]FDG. In addition, compared to fluorine (F2), AcOF is 18  also considerably less violent in its reaction and much more soluble in a wider range of solvents. Unfortunately, the conversion of F2 and OF into AcOF, a relatively simple chemical process, is not properly documented and the success rate varies from 85% to as low as 50% according to different literary sources.  Percent of 2-FDM 2-FDG 94 6  Fluorinating agent AcOF" AcOF*  95  5  F  73  27  70  30  67  33  2  OF  d 2  OF through NaOAc-3H 0 2  2  Table 2. 2: Effect of Fluorinating Agent on the Percentage of 2-FDG/  2-FDM  8  * In these experiments, the fluorinating agent was bubbled into T A G in Freon and the product after acid hydrolysis analyzed by F NMR to determine the relative percentages of 2-FDG and 2-FDM. 19  b  Prepared by passing 1% F in O2 through NaOAc-3H 0. 2  2  Prepared by passing 1% F in argon through NaOAc-3H20.  c  2  d  1% in oxygen  Besides the problematic mixture of [ F]FDG and [ F]FDM, the production of 18  18  [ F]fluorine via 0(p,n) F reaction is slightly more complex. According to (Nickles et 18  18  l8  al. 1984), enriched [ 0]02 gas is bombarded in a fluorine-passivated nickel target and 18  [ 0]0  is withdrawn and trapped for re-use at the end of the bombardment. All F ions 18  18  2  stick to the target walls. The target is then filled with 2% F / neon and bombarded briefly 2  to facilitate the exchange of F with F , forming [ F]F . The radioactive fluorine can 18  18  2  2  12  Chapter 2. Theory then be extracted from the target and used as a fluorination agent or made into AcOF. Compared to making [ F]fluoride in water, the fluorination agent used in a nucleophilic 18  reaction, the production of [ F]F2 is a more elaborate process. Furthermore, since 18  [ F]F2 is in the form of F- F and both fluorides can add to the double bond of a glucal, 18  18  19  the maximum labeling yield is limited to 50%. Similarly, only half of the AcOF made from [ F]F2 has radioactive fluorides; the labeling yield using this modified fluorination 18  agent will also be less than 50%.  2.2.2 Nucleophilic Fluorination Conceptually a nucleophilic synthesis of [ F]FDG should be much simpler, more 18  efficient and more flexible than an electrophilic reaction. The reaction requires a suitably protected mannose derivative with a good leaving group on the 2 position, which will react with fluoride ion by nucleophilic displacement and inversion of configuration at 2 to give the protected [ F]FDG as shown in Figure 2.5 (Tewson 1989). 18  Figure 2. 5: Nucleophilicfluorinationin the synthesis of [ F]FDG I8  X = a leaving group. R = any molecule containing carbon.  13  Chapter 2. Theory  The R groups replace the hydroxyl groups (OH) in a normal mannose, preventing interactions between OH and the leaving group. With these R groups attached, the sugar is said to be in a protected form. The labeled sugar can then be de-protected via acid or base hydrolysis. 18  There are a number of successful nucleophilic syntheses of [ F]FDG published. Among them, the method developed by K. Hamacher gives the highest yield. The 1 8  1 8  greatest challenge in a nucleophilic approach to making [ F1FDG is that [ F]fluoride is an extremely poor nucleophile, making the substitution at the 2 position of hexoses 18  extremely difficult. To increase nucleophilicity of the substituting group, F must be first presented as F associated with a cation that will allow it to dissolve in an 1 8  anhydrous, non-hydrogen bonding solvent such that the formation of [ F]HF is 18  minimized. In the study done by Hamacher, a solution of 20 mg 1,3,4,6-tetra-O-acetyl2-O-trifluoimethanesulfonyl-P-D-mannopyranose (or mannose triflate for short) in 1 ml anhydrous acetonitrile ( C H 3 C N ) is added to the dry residue of [ F]fluoride containing 18  aminopolyether-complex potassium carbonate, which is also commonly abbreviated as Kryptofix 2.2.2 or K-222. This mixture is heated for about 5 minutes under refux and later concentrated to -0.4 ml. At this stage, the displacement of the triflate group using [K-222]K F is completed. Since K-222 acts only as a catalyst, it remains in the solution +I8  and must be removed because of its toxicity. Consequently, the solution and an addition of a small amount of water is passed through a C-18 SEP-PAK cartridge which traps acetylated [ F]FDG while allowing hydrophilic K-222 to go through. The acetylated l8  sugars are desorbed with THF. 1 M hydrochloric acid is then added and the solution is heated under reflux for 15 minutes at temperature of 130°C. This hydrolysis step 14  Chapter 2.  Theory  eliminates the acetyl groups. The summary of these reactions is shown in Figure 2.6 (Hamacher et al. 1986). Finally, the product is decolorized, neutralized, and sterilized via passing through C-18 SEP-PAK cartridge, a column packed with ion retardation resin, and a Millipore filter respectively.  Figure 2. 6: Reaction scheme for the synthesis of [ F]FDG. l8  Tf = -S0 CF , triflate group 2  3  The uncorrected yield is reported to be about 44 ± 4% and the total time varies from 45 to 50 minutes. Besides the use of K-222, tetraalkylammonium salts are suggested as a more reactive alternative (Korguth et al. 1988). However, tetraalkylammonium-mediated syntheses appear to be slightly less successful than the K-222-mediated ones. The problem is thought to be the metal ions coming off from the target walls in the water containing [ F]fluoride. These metal ions compete with the added tetraalkylammonium 18  for fluoride anion, forming insoluble fluoride salts in the organic solvent used and, therefore, reducing the availability of F needed in the labeling step. The trace amount 1 8  of metal ions in water is inevitable since the corrosion and the accompanying dissolution of the target body is part of the target-aging phenomenon due to activation of the target materials by bombardment. K-222, on the other hand, is apparently more forgiving with the metal ion contamination. TMA, tetramethylammonium, is also used to replace K-222  15  Chapter 2. Theory (Mock et al. 1996). The results provide some evidence showing that TMA is equally forgiving as K-222 with respect to the metal contamination. Unfortunately, TMA appears to be more toxic, and the final yield of [ F]FDG does not have any significant 18  increase. To conclude, the chemistry of the [ F]FDG synthesis will be based on the work 18  of Hamacher, using K-222 as the cation catalyst and mannose triflate as the sugar precursor. Some small alterations to the actual steps described earlier will be implemented if deemed advantageous.  2.2.3 Clean-up Columns After hydrolysis, the reaction mixture is transferred to the ion exchange column, alumina cartridge, and C-18 cartridge for purification. Knowing the type of ions/molecules trapped in each column, one can quickly determine the cause of a poor yield by measuring the activity in each clean-up column. •  Ion Exchange Column  This column is packed with BioRad AG50X8 resin (on the top, near the inlet) and AG11A8 resin (on the bottom). AG50X8 removes K-222 from the product mixture and AG11A8 traps excess H* and Cl" ions, neutralizing and desalinating the solution. A small portion of F ions that fail to replace the triflate group during the fluorination step will 18  also stay in this column. •  Alumina Sep-Pak  The majority of F ions are trapped in the alumina column; as a result, a high 18  reading of activity from this column indicates poor labeling yield.  16  Chapter 2. Theory •  C-18 Sep-Pak  C-18 Sep-Pak separates the non-polar, unhydrolyzed mannose from the polar product solution. If a large activity is found in this column, an incomplete hydrolysis reaction is likely to be the reason for a low yield. The order of these clean-up columns is essential to the success of product purification. For instance, the product solution needs to be neutralized by the ion exchange column before reaching the alumina column, which will not trap fluoride ion if the pH is lower than 5. Columns should be arranged in such manner so product solution will first enter the ion exchange column, through the alumina column, and exit via the C18 column.  17  Chapter 3. Experimental & Design  Chapter 3 Experimental & Design 1R  In this chapter, the setup of each component used in the [ F]FDG synthesis will be briefly mentioned. Any special design feature and its benefits will also be pointed out.  3.1 TR-14 Cyclotron, Targets, and Operating Interface Although during the course of this study, no significant change was done to the existing setup of the cyclotron and the targets in C.H.U.S., it is useful to outline them, since, to a certain extent, the procedures and the chemistry module developed for the [ F]FDG production is best suited under these conditions. 18  3.1.1 T R - 1 4 Cyclotron The medical cyclotron installed in C.H.U.S. is a model developed in T R I U M F capable of emitting two 14 M e V proton beams simultaneously. The beam energy generated in TR-14 is selected to place a yield at the optimum power use in [ F]FDG 18  production as discussed earlier in section 2.1.4. The TR-14 cyclotron is equipped with two beam extractors, 180° apart, so that two beams can be extracted simultaneously. Extraction is achieved by stripping the electrons off the negative hydrogen ions using a thin-film carbon foil. The extracted beams exit the TR14 cyclotron along two extraction ports, which are designated as Beamlines 1 and 2. The ability to irradiate either Side 1 or Side 2 greatly enhances the ability to work in research and development. In C.H.U.S., Side 1 of the cyclotron is routinely used for the production of radioisotopes for clinical  18  Chapter 3. Experimental  & Design  use. After bombardment, Side 1 is too radioactive to work with and a cool-down period of a few days is needed. With the option of irradiating the other side, no such wait is necessary and new tests can be made quickly. In addition, there is one target holder on each side, and each target holder can house four different targets. These target holders further minimize disturbance to the existing production line by allowing the operator to bombard one of the eight targets, making experiments involving new targets more accessible. A key feature of the TR-14 cyclotron is the vacuum system design. The vacuum is achieved using cryopumps, eliminating the contamination of oil diffusion pumps. The design also separates the vacuum in the main tank from the subsystems such as ion source, target holders, and extractors. This minimizes the need to vent the vacuum tank during maintenance and avoids the delay due to system pump down. The local shields around the target changers are able to attenuate fast neutrons and gamma rays up to 100 fold, ensuring minimum radiation fields and allowing sensitive equipment to be placed inside the main vault.  3.1.2 Targets The target chosen for the production of [ F]FDG is a niobium water target with 18  Havar target foils. Niobium is very inert and it does not bind to F ions irreversibly. 1 8  Compared to titanium, another popular material for the construction of target bodies, niobium has higher heat conductivity. The chance of neutron activation in niobium target is also relatively low (thermal neutron cross section =1.1 barn). Consequently, after bombardment, a niobium target is less radioactive and the aging process is slower.  19  Chapter 3. Experimental  & Design  Havar foils, composed primarily of cobalt, chromium, and nickel, when activated, have a 1  long half-life and they become very radioactive after repetitive irradiation. However, these foils have extremely high tensile strength of 250 kpsi, capable of holding a high target pressure. In addition, the attenuation in passing a proton beam through Havar foils is very minute, meaning that only a small amount of heat is deposited onto the foils and cooling can be done easily via a helium coolant. On the contrary, a much greater amount of energy is constantly put into the target material and the target body. Heat generated in these parts must be taken out more efficiently to prevent both evaporation and a dramatic pressure build-up. Cooling is done via a running water system at temperature of ~10°C. Since the heat conductivity of niobium body is only satisfactory, some local hot spots are expected. It is then important to frequently monitor the irradiation process to ensure that the internal pressure does not exceed the limit of 500 psi. According to the range values tabulated in Table 2.1, the required thickness of a thick target for a 14 M e V proton beam is about 2.2 mm. The niobium target has 7 mm effective target thickness and holds 0.9 ml of target material. Clearly, the target thickness used is significantly greater than required; this larger thickness prevents the accelerated protons from penetrating target material even when some vapor bubbles are formed. The arrival of the [ O ] H 0 bolus at the target is detected using a bolus detector that measures ls  2  the change in the index of refraction in a transparent tube during the passing of the fluid (Zeisler et al. 1993). [ 0 ] H 0 in the niobium target is bombarded with 14 M e V protons 18  2  at a beam current of -20 uA. The typical yield of the water target is about 1.2 mCi in an hour.  The composition of Havar is Co 42.5%, Cr 20%, Ni 13%, W 2.8 %, Mo 2%, Mn 1.6%, C 2000 ppm, Be 400 ppm, Fe balance. 1  20  Chapter 3. Experimental & Design Gas target has been reported to have some advantages over water targets in the synthesis of [ F]fluoride. Blessing et al. 1986 and Ruth et al. 2001 analyzed the yields 18  of a neon target and an [ 0]C»2 target respectively. In both experiments, [ F]fluoride 18  l8  adhering to the target walls is eluted by 1-2 water washes and trapped in a QMA SEPPAK column. The advantages of this production method are the superior recycling 1  Q  efficiency and the higher beam current permissible. Close to 99% of the non-reacted O2 is withdrawn and trapped in a container for future use. Due to a higher operating current, the total irradiation time of a gas target can be significantly shortened. The major problems associated with gas targets are the time needed to complete the washes and the strict temperature dependence for a successful wash. Blessing has reported that water washes give good yields but the extra steps are too cumbersome for a routine production. In our experience, to accomplish two manual water washes requires at least 30 minutes. Efforts are also needed to clean and dry the target before reuse. In Ruth's target system, time lost to washing a target seems to be less problematic; however, water temperature is shown to be crucial in releasing [ F]fluoride from the target walls. 18  At the optimal temperature of ~ 80°C, two water washes are able to recover more than 70 % of the activity produced in a nickel-plated target. Higher recovery can be achieved by using stainless steel or glassy carbon targets. The geometry of the target also has to be carefully designed to allow water to wash the entire inner surface. A nickel-plated copper gas target was tested in C.H.U.S. Unfortunately, preliminary tests gave a very poor total yield of less than 200 mCi in two water washes. Consequently, the liquid niobium target is used as the target of choice.  21  Chapter 3. Experimental  & Design  3.1.2 Computer Control System The entire production of F is automated by the Computer Control System 1 8  developed by E B C O Technologies. This system is very user friendly and provides the operator with a Graphical User Interface (GUI) with the cyclotron systems. The Computer Control System allows real-time monitoring and control of all aspects of the cyclotron operation and records all parameter information related to the cyclotron. Another key feature of the control system is the safety interlock system. This system monitors any building services required for safe operation of the cyclotron, including the electrical power, the operation of the control system, the chilled water flow and temperature, and room ventilation. A fault in any of these services will trigger a safety interlock to shut down or prevent the start up of the cyclotron.  3.2 Chemistry Module & Peripherals 3.2.1 Overview The goal of the chemistry module is to produce high purity, pyrogen-free [ F]FDG suitable for clinical use. The purpose of the synthesis procedure is to react the 18  [ F]fluoride produced in the target with chemical reagents for a preset of period of time 18  at a controlled temperature. As discussed earlier, the chosen synthesis procedure is based on the Hamacher method, the no carrier-added nucleophilic fluorination of mannose triflate. Culbert et al. 1995 and Mock et al. 1996 both employed the Hamacher method for their production of [ F]FDG and their procedures are summarized in Table 3.1. 18  22  Chapter 3. Experimental  & Design  Step  Step  1. Target contents return 2. Fluoride trapping 3. Fluoride elution 4. First evaporation 5. C H C N addition 6. Azeotropic evaporation 7. Triflate addition 8. Labeling of triflate  9. Third evaporation 10. Addition of 1 . 5 N H C L 11. Hydrolysis 12. First transfer 13. Addition of first wash 14. Second cleanup 15. Addition of second wash 16. Third cleanup  3  Table 3. 1: Summary of the Hammacher method.  The set temperature and reaction time for each step  varies between research groups. Consequently, one of the goals of this study is to determine the optimum parameters for the new chemistry module.  These steps are used as a guideline in this study and changes are made along the way.  Target Contents Return The irradiated [ 0 ] H 0 is routed out of the niobium target through a long Teflon 18  2  tube of 1/8" inner diameter into a 10 ml V-vial inside a shielded dose calibrator. With helium bleed of 130 psi, almost all of the target contents are collected into this bolus catch vial, where the activity of [ F]fluoride is accurately counted. Using another 18  Teflon tube of 1/16" i.d. extending to the bottom of the vial, [ F]fluoride is drawn out l8  into the chemistry module in the initial step of the [ F]FDG synthesis. 18  [ F]Fluoride trapping 18  The QMA-light ion exchange Sep-Pak (130mg, Waters) is preconditioned with saturated sodium bicarbonate (10 ml), and then washed with distilled water (20 ml). To minimize the reduction in the enrichment of the [ 0]H20 due to dilution, air is drawn 18  through the column to remove the residual water. The target contents are passed over the column using negative pressure of -10 to -14 psi, generated by a diaphragm pump. More  23  Chapter 3. Experimental  & Design  than 98% of [ F]fluoride is trapped in the column and the excess [ 0 ] H 0 is recycled 18  18  2  into a recovery vial for purification and future use.  [ F]Fluoride elution 18  The trapped fluoride is eluted with 2.4 ml of K-222/ water/ C H C N solution and 3  only about 1 to 2% of the activity remains. This solution is drawn slowly over the Q M A ion exchange Sep-Pak using controlled negative pressure. The mixture containing the [ F]fluoride is drawn into the reaction vessel. 18  First Evaporation 1R  Any water in the mixture will compete with K-222 to react with [ F]fluoride, forming [ F]HF, a hopeless nucleophile unable to replace the triflate group in the 18  fluorination step. Consequently, following the elution, the mixture is evaporated to dryness using vacuum, controlled helium flow, and heat. For this module, heat is provided from an air heater capable of raising air temperature up to above 150°C in less than 30 seconds. Since a sudden rise of vessel temperature will produce intense steam that splashes the solution in the reaction vessel, causing a high loss of activity to vacuum, it is important to heat the vessel in steps of gradually increasing temperature. In the first phase, 1 psi of helium is applied and the temperature of the hot air is set to 88°C. When the vessel temperature is above 83°C, helium is turned off and the 2  nd  phase temperature  advances to 93°C with a new cut-off vessel temperature at 90°C. In the last phase, the hot air temperature is further increased to 100°C and the cut-off remains unchanged.  Addition of CH CN & Azeotropic Evaporation 3  In many published journals, addition of anhydrous  CH3CN  and azeotropic  evaporation are steps taken to remove the traces of water inside the reaction vessel before  24  Chapter 3. Experimental & Design the labeling/fluorinationstep. However, in the course of this study, it is found that these two steps do not change the labeling yield substantially. Factoring the time required to complete these two steps (~8 minutes), the practice of azeotropic evaporation actually decreases thefinal[ F]FDG yield. Therefore, the addition of C H 3 C N is abandoned and l8  the azeotropic evaporation is replaced by a secondary evaporation following the first one. Second Evaporation 1R  The second evaporation is used to ensure the dryness of [ F]fluoride. Similar to the first evaporation, there are three phases in this step. The set temperatures of the first two phases are at 100°C and 110°C with cut-off temperature of the vessel unchanged. To prepare for the addition of mannose triflate, the set temperature of the last phase is 0°C, which shuts off the power of the air heater and allows cold air to cool the reaction vessel to less than 50°C. Helium of 2 psi is also applied into the vessel to speed up the cooling process. Cooling prevents the formation ofrigorousvapor and bubbles that may carry 1R  [ F]fluoride into the vacuum line upon addition of mannose triflate. Addition of Mannose Triflate The mannose triflate dissolved in  CH3CN  is added slowly into the vessel using  controlled negative pressure. Labeling/ Fluorination The labeling of [ F]fluoride onto mannose triflate is optimized at the vessel 18  temperature of 80°C for 9 minutes. The labeling yield is about 75 to 85%. Efforts have been taken to increase the labeling yield either by lengthening the reaction time or increasing the vessel temperature. However, these two measures result in an unknown dark brown substance adhering to the bottom of the vessel at the end of the synthesis. 25  Chapter 3. Experimental & Design  This brown substance is thought to be burnt sugars and it can cause a loss of up tolO% of the initial activity. Third Evaporation Upon the completion of labeling step, the toxic C H 3 C N is evaporated off under vacuum. Indicated by a large peak in the monitor recording the activity released to the ventilation system, this step is the source of a major loss of activity. At least 10% of the decay-corrected initial activity cannot be accounted for by adding the activities in the product and the components of the module. The way to minimize this loss is to cool the vessel before the evaporation. By blowing cold air for one minute, the peak on the monitor screen appearing at the beginning of the evaporation is halved both in height and width. Like previous evaporation steps, the heating is done in steps via air temperature of 85°C and then 93°C, with the cut-off temperatures set at 80°C and 90°C respectively. Helium pressure is applied to reduce the vacuum power. The reaction vessel is later cooled to <50°C before proceeding to the next step. Addition of HC1 Instead of using 1.5 N HC1, 3 ml of 1 M HC1 is added. Hydrolysis The temperature of the reaction vessel is raised to 115°C to facilitate the hydrolysis of the protecting groups (the acetyl groups) from the intermediate product to form crude [ F]FDG. This step takes 12 minutes to complete. 18  Transfers and Water Washes Using a controlled flow of helium gas, the crude [ F]FDG in hydrochloric acid 18  solution is purified, neutralized, and sterilized by forcing it out of the reaction vessel  26  Chapter 3. Experimental  & Design  through the following set of columns: ion exchange column packed with 1.5 grams of AG50WX8 (OH, Bio-Rad) and 3.0 grams of AG11-A8 (Bio-Rad) ion retardation resins, alumina Sep-Pak (Waters), C-18 Sep-Pak (Waters), and a sterile filter (Millipore). Successive rinses of the reaction vessel and purification line are performed using the sterile USP water of volumes 10 ml and 5 ml, both at 60°C. AG11-A8 ion retardation resin is prepared in saline (10 ml) and water (10 ml) before packing it into the ion exchange column. C-18 Sep-Pak is preconditioned with absolute ethanol (10 ml) and washed with sterile USP water (10 ml). Both alumina Sep-Pak and AG50WX8 are conditioned and prepared with 20 ml of USP water. At this point, the production is considered complete. The product vial is removed from the lead pig into a dose calibrator, where the activity of the product is measured. Finally, a series of quality checks are performed on samples of the product. A l l procedures discussed above are tabulated in Table 3.2.  Step 1  Step Description  Action  [ 0 ] H 0 and [ F]fluoride transferred to the dose calibrator Load [ F]fluoride onto Q M A Sep-Pak  Transfer is done by a controlled flow of helium gas  ,8  2  ,8  2  18  3  Elute [ F]fluoride from Q M A Sep-Pak  4  Evaporation l a  5  Evaporation l b  18  The fluid flow is achieved via vacuum Draw K-222 solution over Q M A via vacuum and 1 psi of helium* Apply vacuum and 1 psi of helium. The set temperature is at 88°C Apply vacuum. The set temperature is at 93°C  27  Comment  [ 0 ] H 0 is collected in the recovery vial [ F]fluoride is eluted to the reaction vessel , 8  2  18  The cut-off temperature is 83°C  The cut-off temperature is 90°C  Chapter 3. Experimental  & Design  6  Evaporation l c  7  Evaporation 2a  8  Evaporation 2b  9  Evaporation 2c (cooling)  10 11  Addition of mannose triflate Labeling  12  Cooling  13  Evaporation 3a  14  Evaporation 3b  15  Evaporation 3c (cooling)  16  Addition of H C L  17  Hydrolysis  18  Dispense 1  19  Wash 1  20  Dispense 2  Apply vacuum. The set temperature is at 100°C Apply vacuum with the set temperature at 100°C Apply vacuum and 1 psi of helium. The set temperature is at 110°C Apply vacuum and 1 psi of helium. The set temperature is at 0°C Apply vacuum and 2 psi of helium Heat at 80°C The set temperature is at 0°C Apply vacuum and 3 psi of helium. The set temperature is at 85°C Apply vacuum at 93°C Apply vacuum and 2 psi of helium. The set temperature is at 0°C Apply vacuum and 4 psi of helium Heat at 115°C Transfer production through clean-up columns into the product vial via 10 psi of helium Draw water from the corresponding reagent vial into the reaction vessel. The set temperature is at 60°C Repeat Dispense 1  The cut-off temperature is 90°C The cut-off temperature is 90°C The cut-off temperature is 90°C  The vessel is cooled for 90 seconds to < 50°C  9 minute labeling time 1 minute cooling time The cut-off temperature is 80°C  The cut-off temperature is 90°C 80 second cooling time  12 minute hydrolysis  Chapter 3. Experimental  & Design  21 22  Wash 2 Dispense 3  23  End of synthesis  24  Product analysis**  Repeat Wash 1 Repeat Dispense 2 with a higher helium pressure (20 psi). Remove the product vial Check the purity and sterility of the product  Table 3. 2: The summary of [ F]FDG synthesis sequencing steps and their optimum parameters. l8  •Helium is applied to adjust the vacuum efficiency in order to control the flow of the solution. Further discussion on this topic will be given in section 3.2.4. ••Product analysis will be discussed in detail in the next chapter.  3.2.2 Construction The [ F]FDG production is established with an integrated system of a main 18  chemistry module, a control box, and a helium control box. Chemistry Module The chemistry module executes all of the chemical reactions described in the previous section. •  Stainless Steel Frame The synthesis module is a modified T R I U M F design, consisting of primarily  commercially available parts mounted on a stainless steel frame. This frame has six components bolted together to give physical support and protection for the sensitive electronics inside the module. These components are: a front panel, a back panel, a base plate, a valve tower, a bracket, and a heater enclosure, all made of stainless steel plates. The front panel houses eight valves, five in the front and three in the back. It also holds a Q M A Sep-Pak, a radiation detector beside the Q M A , a series of clean-up columns, and a [ 0]H20 collector. The back panel has many electrical ports for the connection between 18  29  Chapter 3. Experimental  & Design  the module and the control system. Helium, compressed air, and vacuum supplies as well as the target contents are all connected to the module through this panel. The reaction vessel assembly, the valve tower, the bracket, and the reagent vial holder are mounted on the base plate. The valve tower houses an addition of three valves and a mechanical pressure sensor that monitors the compressed airflow. The bracket is used to secure a pressure transducer measuring the pressure inside the reaction vessel. Since the temperature of the air heater can rise above 100°C during the synthesis, a heater enclosure is necessary not only to fix the heater in place but also to insulate heat transfer. The low thermal conductivity of this stainless steel enclosure prevents potential electrical malfunctions and injuries. The enclosure is bolted to the back of the front panel. A l l fasteners used in the module are also made of stainless steel and the thread sizes are 4-40 U N C , 6-32 U N C , M 3 , or M4. The frame does not form a completely sealed box; the top and the two sizes are empty, providing an easy access to the internal parts. The overall dimensions are 27-cm x 31-cm x 26-cm (depth x width x height). •  The Reaction Vessel Assembly The reaction vessel assembly is consisted of a Teflon base, a Teflon jacket, a 10 ml  reaction vessel, and a P E E K header with nine tapped holes of 14-28 thread and a central cavity tapped for the thread of the reaction vessel. The base plate, as shown in Figure 3.1, has a <])l/8"-NPT fitting, connecting to the air heater. The drilled interior duct forms a 90° turn that directs compressed air to the center of the hollow jacket mounted on top of this base plate. The jacket (Figure 3.2) is designed to fit tightly between the base and the header and to surround the reaction vessel (9591-23, Ace). Its hollow interior is concentric with the center of the vessel and has a diameter larger than that of the vessel  30  Chapter 3. Experimental & Design  by an additional 2-cm. This extra space allows compressed air circulate and exit via a <]>l/8"-NPT tapped hole, 3 cm below the top of the jacket. A 3-mm hole is drilled all the way to the central air duct near the jacket-base junction; a thermocouple is inserted through this opening to detect the air temperature. At the other end of the jacket, another 3-mm hole tapped for a M3 fastener is made. When fitting the header with the jacket, this opening coincides with a groove 3 mm deep on the PEEK header. With a M3 fastener in place, the header is locked firmly to the jacket, preventing the compressed air from lifting the header. Due to its low thermal conductivity, Teflon is the material of choice for the base and the jacket in that heat transfer from the hot air to the surrounding is minimized.  31  Chapter 3. Experimental  & Design  LlC'O] 9»  Figure 3. 1:  The base plate of the reaction vessel assembly  32  Chapter 3. Experimental  & Design  Chapter 3. Experimental & Design Mounted on top of the jacket is a P E E K header, which serves as a manifold connecting the reaction vessel, a thermocouple, and seven Teflon tubes (Figure 3.3). There are eight tapped holes evenly spaced around a 3.8-cm circle on the header; seven of them are linked to 1/16" i.d. Teflon tubes through A-2S SuperFlangless fittings (PEEK, l  Upchurch) and the spare hole is plugged. These SuperFlangless fittings are very convenient. They can be tightened quickly on to a 1/16" tubing and rotate freely around the tube, a great feature that makes wiring tubes an easier task. However, these fittings are designed to be tightened "finger tight" only. Over tightening can collapse the tubing which may result in blockage. The spare hole is reserved for any additional reagent one may want to experiment with; for example, the azeotropic evaporation step can be easily brought back by plumbing a 1/16" tubing between the spare hole and a CH3CN vial. The last tapped hole, located in the center of the header, is used to insert a thermocouple to the bottom of the reaction vessel. A l l of these holes open to a central cavity that can be sealed tightly to the reaction vessel with a #16-Polyurethane O-ring. This type of O-ring is chosen for its high resistance to radiation damage. P E E K , having superb chemical and radiation resistance and low thermal conductivity, is a great material to use in places such as the header and the tube fittings, where frequent exposures to chemicals are expected. •  f 0lH Q 18  7  collector  In the step of trapping [ F]fluoride, the expensive [ 0 ] H 0 is recycled through a 18  18  2  Teflon holder into a glass vial located in the upper left corner of the front panel. A n Oring (#114 B U N A - N ) is used to fit the recovery vial (3 ml, Supelco) tightly against the holder in the same manner as the seal formed between the reaction vessel and the header.  34  Chapter 3. Experimental  & Design  There are two small channels bored into the holder. [ 0]H20 is drawn through a %-28 18  fitting into one of the channels and is collected in the recovery vial. The suction is provided through the other channel that connects to a vacuum pump via Teflon tubing (See Figure 3.4).  Control Box The control box contains many types of electrical components including digital DC output modules (ODC5, Dutec), analog voltage output modules (OV5, duTec), analog voltage input modules (IV10, duTec), analog current output module (OI420, duTec), analog temperature input module (ITCT, duTec), and signal amplifiers. All of them are connected to a master I/O board (IOP-AD/3+, duTec) or an expansion board (IOP-DE, duTec). A power supply (level 3 TUV, Power-One) is used to provide currents of 5, 12, or 24 volts to each electrical component. The overall dimensions of the control box is 19" x 18" x 6.5" (depth x width x height). The control box functions as a processing interface between the computer and the chemistry module. It converts analog signals from devices such as the radiation detectors, the thermocouples, and the pressure transducers into digital engineering units comprehensible for the computer. In addition, signals from the computer are amplified inside the control box to activate electronics such as the valves and the heater in the module. The complete electrical schematics are shown in Figure 3.5.  35  Chapter 3. Experimental  & Design  Figure 3.3: The PEEK header of the reaction vessel assembly  36  Chapter 3. Experimental  & Design  -[.SZ'O] 81  <  i  1  OJ o  L g \0  Figure 3. 4: The [  OJ  18  0]H 0 recovery vial holder 2  37  Chapter 3. Experimental & Design Chemistry module is labeled as  _!'EDG_box!! As shown, every component is wired to the control box  Figure_3. 5: JThe_electrieal schematics ofihe [ F]FDG synthesis system 18  38  Chapter 3. Experimental & Design Helium Control Box The function of the helium control box is to precisely control the pressure of helium permitted into the reaction vessel. Helium stored in a gas tank enters the box through a 1/8" Swagelock fitting and comes out through a A-28 fitting into the chemistry l  module. This unit contains a pressure regulator (Wilkerson) and an electropneumatic converter (Omega) that adjusts helium flow according to the signal voltage sent from the control box. The depth, width, and height of the helium box are 4", 4", and 6" respectively.  3.2.3  Control Software The control software used to control, automate, and monitor the entire [ F]FDG  synthesis is called Lookout, a measurement and automation software package from National Instruments. Lookout employs an object-based architecture, enabling process control and monitor through connections between each object. This software serves as a container for objects that may be programmed by other programming language such as Visual Basic, Java, or C++. To achieve automation, all objects required in the production need to be created and properly connect to each other. An excel file, defining the states of controllable objects and the sequence of reaction processes, is written and adopted by Lookout, which will operate objects accordingly. There are two basic screens created in Lookout: the control screen and the production report screen. In a control screen, objects such as valves, heater, and helium regulator can be actuated and monitored. Connection between the reaction vessel and reagent vials following an opening or closing of a valve will be indicated by a color change in the tubing. Small animations such as air blowing on the reaction vessel and  39  Chapter 3. Experimental & Design formation of bubbles in the reaction vessel are made to assist visual monitoring. A builtin alarm system warns the operator unusual events before or during the production. Readings of temperature, pressure, radiation, and process time are given in real time on the screen. The object-base architecture also allows quick addition of new components into the production. Figure 3.6 shows a typical control screen. At the end of a synthesis, excel parameters, statistics such as the maximum, minimum, and average of all process measurements, trend lines, and the amount of time elapsed in each step are organized and neatly presented in a production report as shown in Figure 3.7. This report will be automatically printed out, while a state file containing important measurements recorded for the run will be produced. B y comparing production reports, causes for a sudden drop or increase of yield and any other inconsistency can be discovered and studied quickly.  40  Chapter 3. Experimental & Design  Figure 3. 6: The control  screen showing the first water wash. The network of tubing is also shown.  41  Chapter  3. Experimental  &  Design  Figure 3. 7: The production report.  Rx TCI = temperature inside the reaction vessel (red line)  (*Lines are in different colors on screen)  Block TC2 = air temperature inside the Teflon base (blue line) Rxn Vessel PT = pressure inside the reaction vessel (black line) 42  Chapter 3. Experimental & Design  3.2.4 Fluid Transfer One of the most essential tasks of the [ F]FDG synthesis module is to transfer the target 18  contents and the reagents to and from the reaction vessel. This task is achieved by an integrated fluid transfer system composed of a network of Teflon tubing, a variety of different types of fittings, a vacuum pump, a helium supply box, an one-way check valve, and 12 solenoid valves. The transfer system is illustrated in Figure 3.6. •  Solenoid Valves Solenoid valves control or, in the case of three-way valves, redirect the passage of  the fluid. A two-way valve (HP225T012, Nresearch) is normally closed when there's no electrical input; it is actuated upon the arrival of a 24-V signal from the control box, allowing an open path for the fluid being transferred. On the contrary, an idle three-way valve (HP225T032, Nresearch) has a flow path between the common port, or simply called "the common", and one of the other two ports. In the actuated state, the valve opens the normally closed port and blocks the normally open one, redirecting the flow into or out of the common. For instance, as shown in the flow schematics, SV10-13 are two-way valves used to block the tubing between the reagent vials and the reaction vessel until the steps where reagents are needed.  SV8 and SV2 are three-way valves used to  bring target contents and K-222 mixtures, through a Q M A Sep-Pak, into the [ 0 ] H 0 18  2  recovery vial and the reaction vessel respectively. The location of the valves is crucial for the success of the synthesis. Many coldtests (tests done without radioactive materials) demonstrate that, during labeling and hydrolysis steps, solution in the reaction vessel is driven out by the rising internal pressure, flowing out of the header along the dispense line which extends to the bottom of  43  Chapter 3. Experimental & Design the vessel. Thefluidthat comes out will not be heated properly; bothfluorinationand hydrolysis in this portion of the solution will not be as efficient as the rest inside the vessel. Consequently, SV14 is mounted directly on top of the PEEK header to reduce the dead volume. SV6 is a three-way valve normally allowing passage between SV14 and the clean-up columns. With SV14 and SV6 actuated, helium of controlled pressure can be directly applied to the bottom of the vessel, stirring the solution. SV4 and SV5 govern the application of vacuum to either the recovery vial or the reaction vessel. To avoid valve overheating as a result of the excess energy deposited onto the solenoid coils, a simple RC circuit consisting of a 150 Q resistor and a 1000 u>F capacitor in parallel is wired to a solid-state relay inside the control box (Figure 3.8).  v+  Figure 3. 8: RC circuit used to prevent valves from overheating. *Valve R = 90 ohms  When a signal isfiredfrom the computer, the relay amplifies the signal voltage to 24V, which actuates a corresponding valve. The current voltage then drops quickly to  44  Chapter 3. Experimental & Design ~9V as the capacitor is fully charged, forcing current to go through the resistor. The lower voltage is sufficient to hold the valves at the energized state while keeping the valve temperature only slightly higher than the ambient temperature. With this implementation, a valve can be energized for more than 12 hours without damage. Since the valves are constantly in contact with chemicals, Teflon valves are chosen for their high chemical resistance. Over a period of two months with 40 hot tests conducted, no valve failure has yet been observed. •  Controlled Vacuum In an ideal transfer, target contents and reagents should be drawn into the reaction  vessel rapidly without any splash. Fluid coming out of the channels inside the P E E K header should also fall onto the bottom of the reaction vessel drop by drop without any loss to the vacuum line. Unfortunately, during a rapid transfer, the solution in motion usually splashes at the end of the transfer line instead of forming a stream of droplets. The portion splashed onto the interior wall of the header is quickly drawn into the tubing of vacuum supply, causing a significant loss of valuable target contents or reagents. Consequently, vacuum power has to be carefully controlled by applying helium gas simultaneously. Application of helium gas of a few psi can effectively reduce the suction on the fluid, lowering the rate at which the fluid is being drawn. Based on the observation made in cold-tests, desirable transfer rates are decided and they are achieved by using helium gas at different pressures. Eluting [ F]fluoride trapped on the Q M A 18  Sep-Pak, for instance, requires the use of 1 psi helium coupled with the vacuum to achieve a smooth transfer of K-222 mixtures from the reagent vial into the reaction vessel (for the optimal use of helium gas at each reaction stage, see Table 3.2).  45  Chapter 3. Experimental •  & Design  Tubing Setup Near PEEK Header  Although a controlled vacuum minimizes the occurrence of a splash, some residual droplets can still adhere to the cavity wall of the header and drift into one of the many reagent lines converged in the header. Besides losing a portion of the transferred solution to vacuum, cross-contamination of reagents is likely to happen. As a small amount of the reagent flows into other reagent lines, the chemical conditions inside the reaction vessel will be impaired in the later steps, during which, the previously trapped reagent is drawn into the vessel along with the intended chemical. Inefficient labeling and reduced chemical purity are the common outcomes of this contamination. To avoid mixing of the reagents, Vi-28 PEEK nuts are locked to Teflon tubes a few inches away from the tip. As a result, when these nuts are tightened into the header, the tubes are pushed through the inner channels, extending out of the interior wall of the header cavity. In this setup, droplets adhering to the opening of a tube will have to travel a much longer distance to enter other tubes reserved for other chemicals. Since SV14 is located immediately on top of the header, a male-to-male fitting, made with two 1/4-28 nuts placed tightly against each other is used to connect the valve and the header. However, SV14 receives no support from the stainless steel frame and the soft Teflon tube between the two nuts cannot sustain the weight of a solenoid valve. To prevent blockage, a brass column is fabricated to clamp the fittings together, forming a one-piece union able to uphold SV14. In addition, since the tubing reserved for water washes are the longest with the largest potential space for storing chemicals, a one-way check valve (CV3301, Upchurch) is installed to prevent fluid from backing up the water line.  46  Chapter 3. Experimental  & Design  3.1.5 Heat Control Temperature of the reaction vessel is arguably the most crucial factor determining the final yield of [ F]FDG. Heating at a proper temperature accelerates each chemical 18  process and strongly favors the equilibrium to the product side. In Table 3.2, the optimal set temperature at each stage of the synthesis is listed and a heat control system using an air heater, a solid-state relay, a pulse control module, and two thermocouples is developed. The cylindrically shaped air heater (S-50, Farnam Custom), with one end connected to the base of the reaction vessel assembly and the other end, a compressed air supply, is able to generate thermal energy from 120V current at the maximum rate of 400 Watts. Its 6.4" long metal body heats up air stream from room temperature to >100°C in less than 15 seconds. When the heater is switched off, air temperature quickly falls to ~30°C; the rapid flow of cool air now becomes a great coolant, bringing down the vessel temperature quickly. The previous chemistry module in C.H.U.S. utilizes a thin copper pad installed inside the jacket for heating. This method is good in that heating is done gently and precisely since temperature of the heating pad rises gradually. The obvious drawback is the time required to complete three evaporation steps being extremely long (-35 minutes). On the contrary, the present heating device consumes only 10 minutes for the three evaporations and the practice of step heating is actually implemented to slow down the heating processes. The typical total evaporation time now stands at 20 to 25 minutes. The power output of the air heater is regulated closely via the use of a pulse control module ( P C M module) and a solid state relay (SSR). With the P C M module  47  Chapter 3. Experimental & Design mounted directly to the input terminals of the solid state relay, a simple conversion of an ON/OFF solid state relay to a proportional power regulator is made (see Figure 3.9).  Figure 3. 9: The wiring diagram of a PCM-SSR system *The heater is represented as "LOAD" in this diagram  Based on the inputs from a 4-20 mA analog current output module inside the control box to the P M C module, the proportional power regulator provides a rectangular shape, time proportional control signal to the heater. The average power to the load will be proportional to the 4-20mA input, which is in turn controlled by the feedback of the thermocouple measuring air temperature. Consequently, when the reading of air temperature is above a certain set value, the temperature input and current output modules reduce the current going to the P M C module, resulting a decline in heater power. Together, with a reverse mechanism applied when air temperature is below a set point, the temperature of the air stream flowing through the reaction vessel can be controlled and maintained readily. The thermocouple measuring the reaction vessel temperature is not directly involved in the heating control. Instead, it serves as a trigger of the transition from one step to another. For instance, Evaporation l a will not be terminated if the reading of the vessel temperature is not above the 83°C mark. This trigger mechanism ensures the  48  Chapter 3. Experimental & Design dryness of the vessel. Figure 3.10 shows the heating and cooling curves based on data collected via the two thermocouples. Heating and Cooling Curves  —  —Air temp. Vessel temp.  Time (second) Figure 3. 10: The heating and cooling curves. The set temperatures are 100°C at the first 5 minutes (0-300 seconds) and 0°C for the next 2 minutes (300-420 seconds).  Besides the advantages in the use of this new heater, accidental overheating of the metal body often occurs when an operator forgets to turn on the compressed air supply. Without adequate airflow, the thermocouple inside the jacket will remain at room temperature despite the fact that thermal energy is being generated and deposited onto the heater body at the maximum rate. In other words, feedback from the control box to the P C M module will always be positive since the reading of air temperature stays below the set point. In less than one minute, the metal body will reach >200°C and become permanently damaged. To prevent this hazard, the 4-20mA current is routed to the P M C  49  Chapter 3. Experimental & Design module through a mechanical pressure switch (SMC) measuring airflow at the entrance of the heater. As compressed air flowing out of a l/8"i.d. Teflon tube connected to the exit hole of the jacket, it generates loud noise. This problem is treated with the addition of a stainless steel muffler (Swagelok) at the end of the tube.  3.1.1 Accessories •  Radiation Detectors Two radiation detectors are used in the chemistry module and their readings are  plotted as trend lines on the control screen in real time. One of the detectors is fixed to the frame next to Q M A Sep-Pak, measuring the trapping and releasing of  F ; the other,  strapped on the jacket, monitors the activity inside the reaction vessel. Both detectors are assembled by sealing a photodiode and a CdW04 crystal inside a small box filled with non-conducting black epoxy (see Fig 3.11). The signals generated from the photodiodes are amplified by the circuit shown in Fig 3.12. With an uncertainty of about 0.05 C i , these radiation detectors can track the flow of radioactive fluoride ions. Moreover, for a synthesis involving more than 1.0 C i , they are a powerful tool in establishing the problematic step where activity drops faster than the natural decay line. •  Pressure Transducer The pressure transducer is a diaphragm meter made by Data Instrument. It is  chosen for its high resistance against chemicals such as hydrochloric acid, one the reagent used in the synthesis. In C.H.U.S., the same pressure transducer is used for over the span of one year; no malfunction has yet been encountered.  50  Chapter 3. Experimental & Design  Figure 3.11: Theradiationdetector assembly  51  Figure 3. 12: The radiation amplifier circuit  52  Chapter 3. Experimental  & Design  3.1.2 Reagent Preparation The personnel in charge of operating [ F]FDG synthesis module should arrive 90 18  minutes prior to the scheduled production since approximately one hour is needed to complete the reagent preparations and the cleaning procedures. Brief descriptions on preparing K-222 and mannose triflate are given; other reagents are very common and their preparation will not be discussed. •  Preparing K-222 mixture Place 10 mg of Kryptofix 2.2.2 (Aldrich) and 2 mg of Potassium Carbonate  (Aldrich) into a sterile container. These chemicals are dissolved by 200 U.L of sterile USP water. Use an "all-plastic" syringe to add 2.0 mL of anhydrous acentonitrile (Aldrich) and transfer the mixture into a sterile, nitrogen-filled vial. From our experience, the quality of K-222 solution does not show signs of deterioration over a period of one week. Therefore, a large batch of K-222 mixture equivalent to the amount needed for the week is usually prepared and stored in a properly sealed vial. Before each synthesis, 2.0 mL of K-222 solution will be drawn out from this batch into a reagent vial. •  Preparing Mannose Triflate Mannose triflate is very reactive with water in the atmosphere; therefore, it is  prepared immediately before the FDG synthesis. 15 mg of Mannose Triflate is weighted into a small sterile, dry beaker, dissolved in 2.5 mL of anhydrous CH3CN, and transferred to a sterile, nitrogen-filled vial.  53  Chapter 3. Experimental & Design  3.1.3 Cleaning 18  One of the principal factors leading to a successful synthesis of [ F]FDG is the sterile and chemically uncontaminated initial environment inside the tubing and the reaction vessel. To achieve this, a series of aseptic cleaning procedures is developed for this chemistry module and it is presented in Table 3.3. In this study, a jump of more than 10% is observed in the final uncorrected yield when proper cleaning steps are performed prior to a synthesis. The cleaning procedures also include inspections on the seals of the valves and the reaction vessel; necessary maintenance can then be done before the start of the synthesis.  Step 1  Description  Action*  Vessel Cleaning  Take out the reaction vessel from the PEEK header and washed it with chromium acid and plenty of water Open Sv9 and apply 20 psi helium. Close Sv9 and the helium supply after the pressure transducer reads >12 psi  Leak Check  2  Addition of CH CN 3  Clean K-222 and mannose triflate lines with 5 mL of CH CN. Open valves corresponding to these reagent lines and manually inject the cleaning solution. Manually inject 4 mL of ethanol into HCL line and 3 mL into each water line. Open every reagent valve and the dispense valve. Apply 20 psi helium and increase heater temperature to 120°C. 3  3  Addition of Ethanol  4  Drying 1  54  Comments Screw the vessel back on after it is cleaned.  Pressure of the vessel should not drop more than 0.3 psi in one minute. If pressure drops too rapidly, one needs to find the leak and fix it. After cleaning the two lines, there should be 10 mL of CH CN in the reaction vessel. Extract the solution via the dispense line (Svl4 is opened). Drain ethanol through the dispense line. 3  This step should take 15 to 20 minutes. Meanwhile, start preparing reagents and the columns.  Chapter 3. Experimental & Design 5  Drying 2  6  Cooling  7  Final Leak Check  Close valves and open Sv5. Apply vacuum and keep the heater at 120°C. Close Sv5 and open SvlO. Turn the heater temperature to 0°C and apply 20 psi helium. Allow the vessel temperature to drop below 30°C. Repeat Step 1  1 to 2 minutes are given for this step. Helium flow accelerates the rate of cooling; SvlO serves as the exit for helium. This step takes about 3 to 5 minutes. At the end of this step, open one valve to release the pressure.  Table 3. 3: Aseptic cleaning sequencing steps. * See Figure 3.8 in section 3.2.3 for the location of each valve  Time required to complete every cleaning step should be within 35 minutes if the module is free of leakage.  55  Chapter 4. Results and Product Analysis  Chapter 4 Results & Product Analysis In this chapter, results of each production are organized in figures or tables. The methods and outcomes of three quality checks will also be documented.  •  f FlFDG !8  Yield  Yields of [ F]FDG are plotted in Figure 4.1. The data of two productions 18  ignored due to operating errors.  Figure 4. 1: Summary of [  l8  F]FDG yield.  56  are  Chapter 4. Results and Product  Analysis  Percent yield in Figure 4.1 is defined as the activity of the product, decay-corrected or uncorrected, divided by the initial activity of F and multiplied by 100%. Between the 1 8  two different ways of denoting yield, decay-corrected yield (also known as the radiochemical yield) indicates the efficiency of chemical reactions in the synthesis; 1 8  1 8  whereas, the uncorrected yield shows the true conversion rate of F ions to [ F]FDG. Since the uncorrected yield is the figure of real physical importance, for the rest of this paper, the percent yield is equivalent to the uncorrected yield unless specified otherwise. The total activity accounted is the sum of activities remaining in Q M A , ion exchange column, alumina column, C-18 column, and reaction vessel. Accordingly, subtracting this value from 100%, the amount of activity lost to the ventilation system through vacuum is easily calculated. Before the 5 run, cleaning procedures are not fully developed; consequently, th  yields are lessened run after run as the reaction vessel and the tubing becomes more and more contaminated. This trend is shown by sharp negative slopes in the yield curves between the 1 run and the 4 run. When properly cleaned, the uncorrected yield in the st  th  5 run is increased by ~15%. th  The value of percent yield fluctuates substantially in the first 15 runs. It is because the parameters such as the reaction time, the set temperature, and the helium pressure for each stage have not been determined, and changes to reaction conditions are being made constantly in an attempt to optimize the yield. At the end of the 16 run, th  both yield and the total activity accounted are satisfactory; it is then assumed that the optimum production will occur in conditions close to that of this run. For the subsequent productions, changes to the parameters are kept very moderate and only a single  57  Chapter 4. Results and Product  Analysis  parameter per run is allowed to differ. Based on this method, parameters given in Table 3.2 are the most updated conditions and they are deemed as the optimum. After Run 16, the percent yield is around the 40% mark, except for a depression from Run 28 to Run 30 and a sharp drop at Run 39. During the 28 to the 30 runs, more th  th  drastic changes to the parameters are experimented in order to shorten the time allocated to evaporation, labeling, and dispense. In the 28 run, the first two evaporations fail to th  eliminate water completely before the labeling step. The decay-corrected percent activity of alumina column increases from the usual 12% to 24%, indicating a poor labeling reaction. In the 29 and the 30 runs, the activity in C-18 column is measured to be th  th  above 10%, characterizing an incomplete hydrolysis. This depression in the yield line is quickly fixed by setting the parameters back to that of the 27 run. In Run 38, the th  thermocouple inside the reaction vessel does not function properly and sugar products are overheated as a result. Activities remained on the reaction vessel and C-18 jump from the expected 2% and 4% to 12% and 11% respectively. The activity in the vessel is in the form of burnt sugars stuck to the glass wall. Yield returns to >40% after the thermocouple is adjusted. In addition, the trend line for the total activity accounted shows a significant increase from -70% to -90% in Run 14 and remains at 80% to 90% for the subsequent tests. This improvement is accomplished by the implementation of step heating as discussed in section 3.2.5. Overall, the uncorrected yield is 41 % (± 5 % s.d., n = 25), ignoring the first fifteen runs when optimization of the module is still in progress. If the depression and the drop in Run 38 are taken out, the yield further increases to 42.1 (± 3 % s.d., n = 21).  58  Chapter 4. Results and Product Analysis •  Base Hydrolysis Compared to the acid hydrolysis, it has been suggested that alkaline hydrolysis  offers special advantages such as the shorter reaction time and mild thermal requirements. This method is not extensively used in this study due to the fact that base hydrolysis is reported to produce both [ F]FDG and [ F ] F D M as the final products. The 18  18  ratio of [ F]FDG and [ F ] F D M seems to be highly dependant on the concentration of 18  18  OH" in solution and the hydrolysis time. The work done by Fuchtner et al. 1996 shows that under 0.3 M N a O H at room temperature, the reaction will be completed within 1 minute and the optimum yield will be achieved within 2 minute as shown in Figure 4.2.  ; yield (%FDG]  20 10 • 01 0 1 2 3 4 5 6 7 8 9 10 Figure 4. 2: •  .  i  i  I  I  Dependence of forming [ F]FDG on sodium hydroxide 18  time [min]  concentration and reaction time during the alkaline hydrolysis. Note that the yield here is not the same as the percent yield defined previously. It only shows the relative success rate in producing [ F]FDG under different conditions. 18  59  Chapter 4. Results and Product Analysis These optimum hydrolysis conditions are later updated to 0.33 M NaOH at a temperature below 40°C for less than 5 minutes (Meyer et al. 1999). The epimerization of [ F]FDG and [ F ] F D M is limited to 0.5%. In the light of these encouraging results, a 18  18  test run is conducted, replacing the original acid hydrolysis with 0.3 M NaOH at room temperature for 3 minutes. The chemistry module is quickly adjusted for the alkaline hydrolysis step, and the acid form of AG11A8 resin is packed into the ion exchange column. The initial outcome is a promising 33% uncorrected yield and a careful investigation similar to the procedures developed by Van Rijn et al. 1985 indicates no significant trace of [ F]FDM. It is likely that the yield will at least reach the 40% mark 18  if efforts are made to optimize the process. •  Product Analysis To measure the radiochemical purity, thin layer chromatography (TLC) is performed.  In a T L C test, an aliquot of the product is placed on a silica plate and allowed to developed in solution containing 10% water and 90% C H C N . When the plate is 3  developed, it is observed under a device called Instant Imager, capable of identifying 18  spots of high activity and accessing the radiochemical purity accordingly. [ F]FDG produced in this module typically has a radiochemical purity of 99.0 ± 0.5%. The concentration of C H 3 C N is evaluated using a gas chromatography (GC) equipped with a flame ionization detector (FID). The outcome is < 250 ppm for the productions after the 14 run, below the 400 ppm permissible level set by F D A (Channing et al. th  2001). Finally, the product is checked for the concentration of Kryptofix 2.2.2 before it is administered to patients. In a number of reports, the amount of K-222 that constitutes a  60  Chapter 4. Results and Product Analysis 50% lethal dose (LD50) in rats is 35 mg/kg; the toxicity of K-222 is clearly a concern in Hamacher's method. In C.H.U.S., the silica plate used in TLC test is placed in an iodoplatinate reagent consisting of hexachloroplatinic acid (FhPtCle) and potassium iodide (KI). Free K-222 binds to KI, forming a [K-222] T complex similar to [K+  222] F. Iodide in the complex will be oxidized to I2 when in contact with H^PtCk, and +18  a colored platinum complex is formed. Although some authors have reported a detection limit of less than 0.003 mg/mL (Mock et al. 1997), the limit found here is about 0.01 mg/mL. During the course of this study, no visible color spot is found on the plate, proving that K-222 concentration in the product is below 0.01 mg/mL.  61  Chapter 5. Conclusion  Chapter 5 Conclusion iR  A modified chemistry module producing epimerically pure [ F]FDG with full automation is built with commercially available parts. The construction costs of this module is slightly less due to fewer electronic components inside the control box. The design of this module and its control software allows alterations in the chemical procedures to be done quickly. A set of cleaning procedures is developed and the optimum reaction parameters are found. From a total of 25 hot tests done to optimize the module, the uncorrected yield stands at 41 ± 5% with great reliability. The synthesis time is about 55 minutes, approximately 15 minutes less than the previous module in C.H.U.S. The product passes all of the required quality checks and it is now part of the [ F]FDG production line in the 18  radiochemistry lab in the hospital. A similar model is now being tested in Vancouver General Hospital (V.G.H). Within 3 trials, the uncorrected yield reaches 41%, showing a promising start. Besides [ F]FDG, with some modifications to the module and the automation 18  program, it is possible to produce other F-labeled tracers. For instance, by switching to 18  18  the appropriate reagents and reconfiguring several transfer lines, the original [ F]FDG module can be quickly adapted to the production of [ F]FESDS, a PET tracer for the 18  estrogen receptor (Romer et al. 2001). In addition, according to the reactions described in (Studenov and Berridge 2001) and (Shiue et al. 2001), a modified module should be capable of producing a variety of potential myocardial blood flow tracers and tracers like 62  Chapters.  Conclusion  [ F]FHPG and [ F]FHBG that are useful for assessing the efficiency of gene therapy. 18  18  Trials involving the production of a different F-labeled pharmaceutical should be 18  conducted in the future.  63  Bibliography  Bibliography Adler, L., Blair, H., and Makley, J. (1991). "Noninvasive grading of musculoskeletal tumors using PET." J. Nucl Med, 32, 1508-1512. Bishop, A., Satyamurth, N., Bida, G., Phelps, M., and Barrio, J. (1996a). "Production of [18p]F2  U s i n  S  t n e  16  0 ( H e , p ) F Reaction." Nuclear Medicine & Biology, 23, 3  18  385-389. Bishop, A., Satyamurthy, N., Bida, G., and Barrio, J. (1996b). "Chemical Reactivity of the  1 8  F Electrophilic Reagents from the 0(p,n) F Gas Target Systems." 18  18  Nuclear Medicine & Biology, 23, 559-565.  Blessing, G., Coenen, H., Franken, K., and Qaim, S. (1986). "Production of [ F]F2, 18  H F , and F 1 8  1 8  using the Ne(d,cc) F process." Appl. Radiat. hot., 37, 113520  a q  18  1139. Channing, M., Huang, B., and Eckelman, W. (2001). "Analysis of residual solvents in 2[18p]FDG by GC." Nuclear Medicine & Biology, 28, 469-471.  Culbert, P., Adam, M., Hurtado, E., Huser, J., Jivan, S., Lu, J., Ruth, T., and Zeisler, S. (1995). "Automated Synthesis of [ F]FDG using Tetrabutylammonium Bicarbonate." Appl. Radiat. hot., 46(9), 887-891. Fuchtner, F., Steinbach, J., Mading, P., and Johannsen, B. (1996). "Basic Hydrolysis of 18  2-[18p]Fluoro-l,3,4,6-tetra-0-acetyl-D-glucose in the Preparation of 2[ F]Fluoro-2-deoxy-D-glucose." Appl. Radiat. hot., 47(1), 61-66. Guillaume, M. (1991). "Recommendation for Fluourine-18 Production." Appl. Radiat. hot., 42(8), 749. Hamacher, K., Coenen, H., and Stocklin, G. (1986). "Efficient Stereospecific Synthesis 18  of No-Carrier-Added 2-[ F]-Fluoro-2-Deoxy-D-Glucose Using Aminopolyether Supported Nucleophilic Substitution." J Nucl Med, 27, 235-238. Korguth, M., DeGrado, T., and Holden, J. (1988). "Effects of reaction conditions on rates of incorporation of no-carrier added F-18 fluoride into several organic 18  compounds." J. Labeled Compd. Radiopharm., 25, 4.  Krane, K. S. (1988). Introductory Nuclear Physics, John Wiley and Sons, Toronto. Meyer, G., Matzke, K., Hamacher, K., Fuchtner, F., Steinback, J., Notohamiprodjo, G., and Zijlstra, S. (1999). "The stability of 2-[ F]fluoro-deoxy-D-glucose towards epimerisation under alkaline conditions." Appl. Radiat. hot., 51, 37-41. 18  Mock, B., Varek, T., and Mulholland, G. (1996). "Back to Back "One-Pot" [ F]FDG Syntheses in a Single Siemens-CTI Chemistry Process Control Unit." Nuclear 18  Medicine & Biology, 23, 497-501.  Mock, B., Winkle, W., and Vavrek, M. (1997). "A Color Spot Test for the Detection of Kryptofix 2.2.2 in [ F]FDG Preparations." Nuclear Medicine & Biology, 24, 193-195. 18  Nickles, R., Doube, M., and Ruth, T. (1984). "An 0 2 target for the production of 1 8  [ F]F ." Int. J. Appl. Radiat. hot., 35,117-122. 18  2  64  Bibliography Pavan, R. (1997). "An H 2 0 Water Target for the Production of F-fluoride in the TR13 Cyclotron," engineering physics, University of British Columbia, Vancouver. Romer, J., Fuchtner, F., Steinbach, J., and Kasch, H. (2001). "Automated synthesis of 16a-[18F]fluoroestradiol-3,17 P-disulphamate." Appl. Radiat. hot., 55,631-639. Ruth, T., Buckley, K., Chun, K., Hurtado, E., Jivan, S., and Zeisler, S. (2001). "A proof 1 8  18  of principle of targetry to produce ultra high quantities of l F-fluoride." Appl. Radiat. hot., 55,457-461. Shiue, G., Shiue, C , Lee, R., MacDonald, D., Hustinx, R., Eck, S., and Alavi, A. (2001). "A simplifed one-pot synthesis of 9-[(3-[ F]Fluoro-l-hydroxy-2propoxy)methyl]guanine ([ F]FHPG) and 9-(4-[ F]Fluoro-3hydroxymethylbutyl)guanine ([ F]FHBG) for gene therapy." Nuclear Medicine & Biology, 28, 875-883. Solin, O. (1988). "Production of F from Water Targets." Appl. Radiat. hot., 39(10), 1056. Studenov, A., and Berridge, M. (2001). "Synthesis and properties of l F-labeled potential myocardial blood flow tracers." Nuclear Medicine & Biology, 28,683693. Tewson, T. (1989). "Procedures, Pitfalls and Solutions in the Production of [ F]2Deoxy-2-fluoro-D-glucose: a Paradigm in the Routine Synthesis of Fluorine-18 Radiopharmaceuticals." Nuclear Medicine & Biology, 16(6), 533-551. Van Rijn, C , Herscheid, J., Visser, G., and Hoekstra, A. (1985). "On the Stereoselectivity of the Reaction of [l F]Acetylhypofluorite With Glucals." Appl. Radiat. hot., 36(2), 111-115. Zeisler, S., Ruth, T., Rector, M., and Gschwandtner, G. "Detectors and transducers for target operation and automated PET chemistry." Proc. Fifth Int. Workshop on 8  l8  18  18  i8  1 8  8  18  8  Targetry and Target Chem., Brookhaven.  65  Appendix  Appendix In this section, several pictures of the chemistry module and its components will be shown.  Figure A. 1: Front-side view of the chemistry module  66  67  Appendix  Figure A. 5: The control box.  68  


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



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"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items