CLINICAL PHARMACOKINETIC CONCEPTS
Clin. Pharmacokinet. 22 (4): 274-283, 1992 0312-5963/92/0004-0274/$05.00/0 © Adis International Limited. All rights reserved. CPKl
Uses and Limitations of Positron Emission Tomography in Clinical Pharmacokinetics/Dynamics (Part II) Laura L. Boles Ponto and James A, Ponto PET Imaging Center, Department of Radiology, University of Iowa, Iowa City, Iowa, USA
Contents 274 274 275 275 277 277 278
Summary
Summary 3. The Use of PET in Pharmacokinetics 3.1 Route of Administration: Intra-Arterial vs Intravenous 3.2 Receptor Binding and Occupancy 3.3 Toxicodynamics: Alcohol 3.4 Influence of Enantiomeric Forms 4. Conclusion
Positron emission tomography (PET) involves imaging the biodistribution and tissue localisation of small amounts of radiolabelled biomolecules or drugs. In Part I of this article, which appeared in the previous issue of the Journal, the applications of pharmacokinetics in PET were discussed in order to derive quantitative measures of physiological function. Part II examines the use of PET imaging as a tool to study the pharmacokinetics and pharmacodynamics of specific drugs.
3. The Use of PET in Pharmacokinetics The use of PET imaging as a pharmacokinetic tool involves the radiolabelling of the drug of interest and the subsequent imaging and analysis of the time-course of the disposition of the agent. This approach has been employed primarily for receptor-binding ligands, especially those binding to eNS receptors, and to cancer chemotherapeutic agents. The use of PET imaging as a pharmacodynamic tool involves the imaging and analysis of the degree and time-course of pharmacologically-induced modifications in fundamental biological processes such as blood flow and/or metabolism. This ap-
proach is applicable whenever pharmacological or toxicological effects involve the induction of change in 1 of these fundamental processes (e.g. investigations into the effect of a2-adrenoceptor agonists on myocardial perfusion using successive 82Rb dynamic PET studies (Budinger et al. 1990) and the effect of adenosine on cerebral blood flow using [ 150]water (Sollevi et al. 1987)). PET studies of pharmacokinetics were reviewed previously by Yamamoto and Diksic (1985). They examined the subject from a pharmaceutical transport perspective. In contrast, this review will present representative examples of the types of phar-
275
Positron Emission Tomography
macokinetic/pharmacodynamic issues that can only be addressed in vivo by PET imaging. 3.1 Route of Administration: Intra-Arterial vs Intravenous A critical objective of any treatment regimen, but especially for those regimens involving highly toxic drugs such as cancer chemotherapeutic agents, is to maximise the desired therapeutic responses while minimising undesirable responses (adverse or toxicological effects). One of the choices which potentially influences this therapeutic balancing act is that of the route of administration. The pharmacological advantage of intra-arterial vs intravenous administration of 2 cancer chemotherapeutic agents, cisplatin and carmustine, has been evaluated for CNS tumours using the 13N_ and IIC-labelled derivatives, respectively (Ginos et al. 1987; Tyler et al. 1986). Intracarotid administration of cisplatin produced 1.1 to 2.5 times greater tumour/brain concentration ratios than did intravenous administration (Ginos et al. 1987). Superselective arterial administration of carmustine into a major tumour-feeding vessel (e.g. middle cerebral artery) provided tumour/healthy brain tissue concentration ratios of 50: I, whereas intravenous administration provided ratios of only 1.2: I (Tyler et al. 1986). Furthermore, ratios of tumour llC radioactivity concentrations between the intraarterial and intravenous routes ranged from 2.5 to 99 (n = 10). The 2 highest ratios were associated with patients demonstrating the most dramatic responses and the lowest ratios were associated with individuals experiencing essentially no therapeutic response. Similar investigations have been undertaken to evaluate the regional administration of fluorouracil in the treatment of metastatic liver disease (Strauss & Conti 1991). Examination of the short and lon~ term uptake of [18F]fluorouracil into metastatic lesions after intravenous and intra-arterial administration demonstrated that enhanced drug exposure through regional administration was only I factor involved in determining potential responsiveness to therapy. For fluorouracil, preferential
perfusion must also be accompanied by high tumour cell transport and metabolism. Therefore, delivery of the therapeutic agent, in this case fluorouracil, to the tumour is a necessary but not a sufficient predictor of response (Strauss & Conti 1991). 3.2 Receptor Binding and Occupancy Pharmacokinetic/pharmacodynamic investigations rely heavily on drug concentrations measured in accessible body fluids such as serum, plasma, whole blood, saliva, semen and/or urine. The use of these measures is predicated on the assumption that the concentrations in these fluids reflect the concentration at the site of action. Whenever the response pattern mirrors the concentration pattern in the measured fluid, this assumption is reinforced. When an inconsistent relationship is exhibited, the assumption is in question and a more direct measure is necessary. 3.2.1 Dopamine D2-Receptors The antipsychotic effect of neuroleptic drugs is believed to be due to the ability of these agents to block central dopamine receptors (Creese 1987; Losonczy et al. 1987). Unfortunately, plasma concentrations of neuroleptic agents are known to be imperfect predictors of response (Cohen et al. 1988; Kirch et al. 1988; Midha et al. 1989). Therefore, PET imaging studies have taken a more direct approach to the examination of the issues surrounding neuroleptic treatment. Specifically, 3 issues have been examined with respect to dopamine D2receptors and neuroleptic therapy (Farde et al. 1988a; Smith et al. 1988; Wolkin et al. 1989). First, what degree of receptor blockade is induced by clinically-administered dosages of various neuroleptic agents? Secondly, what is the time-course of receptor occupancy and plasma concentrations on neuroleptic withdrawal? Thirdly, do responders and nonresponders differ in neuroleptic-induced receptor blockade? Using [11C]raclopride, Farde et al. (1988a) studied central dopamine D2-receptor occupancy in 14 patients with schizophrenia who had received a
276
Clin. Pharmacokinet. 22 (4) 1992
single antipsychotic drug, at conventional dosages, for at least 1 month. Receptor occupancy ranged from 65 to 84% for all agents. One patient receiving nortriptyline exhibited no receptor occupancy. The authors concluded that, although all 11 neuro1eptics were chemically distinct entities, the degree of occupancy of central dopamine D2-receptors supported the proposed mechanism of antipsychotic drug action. The single patient in the study who experienced extrapyramidal side effects (akathisia) had the highest receptor occupancy at 86% on a dosage of haloperidol 6mg twice daily. After reduction of the dosage to 4mg twice daily, the side effects disappeared, but the receptor occupancy remained at 84%. Because of the relatively low incidence of extrapyramidal effects in this sample and the change in symptoms associated with a minor change in receptor occupancy, the investigators hypothesised that the antipsychotic action may require a lower receptor occupancy than for the development of extrapyramidal effects or that the 2 effects may be mediated by different receptor mechanisms. Smith et al. (1988) took an indirect approach to examining the relationship between [lsF]Nmethylspiroperidol ([lSF]NMS) uptake in the striatum and plasma concentration in healthy volunteers and patients with schizophrenia treated with either haloperidol or chlorpromazine. They found a decrease in the affinity of the [lSF]NMS for striatal tissues in the drug-treated patients with schizophrenia compared with the unmedicated volunteers, with the magnitude of this decrease directly related to the antipsychotic drug dosage (relative to the clinical potencies of the agents) and plasma concentration. When the antipsychotic drug was withdrawn from patients, the ratio index, a measure of receptor occupancy, approached that observed in healthy volunteers. A log-linear relationship was observed between the ratio index (y) and plasma haloperidol concentrations (x): y = 1.974 e-O.082x
r2 =
0.7482
(Eq. 35)
Using a more direct approach, Farde et al. (l988a) examined the serial receptor occupancy vs serum drug concentration after withdrawal of therapy in
3 of the patients described above. In patients treated with sulpiride 600mg twice daily and haloperidol 6mg twice daily, receptor occupancy remained essentially unchanged at approximately 65% and 85%, respectively, 27 and 54h after withdrawal, even though the serum drug concentrations had declined to approximately 25% of initial values. The third patient underwent a stepwise dosage reduction from sulpiride 800mg twice daily to 0 mg/day. The receptor occupancy exhibited a curvilinear relationship to the total daily dose whereas the serum concentration exhibited a linear relationship. The investigators concluded that the hyperbolic nature of the occupancy curve explained why major changes in drug concentration resulted in only minor changes in the receptor occupancy. Furthermore, they stated that PET imaging may be a method by which minimum individual dosages may be defined which provide adequate receptor occupancy for antipsychotic effects but minimise the potential for adverse effects. Although the above studies support the concept of a receptor occupancy-response relationship, additional factors may be involved in the response of schizophrenic patients to antipsychotic therapy. Wo1kin et al. (1989) examined dopamine-D2 blockade using [lsF]NMS in 2 biological subgroups of patients with schizophrenia, responders and nonresponders, treated with a maximum of haloperidol 100mg daily (minimum plasma concentration = 10 IAog/L) for 4 to 6 weeks. Comparable degrees of haloperidol-induced receptor blockade were found between responders and nonresponders. The authors concluded, therefore, that nonresponse to therapy may not necessarily be caused by inadequate concentrations of drug in the central nervous system. Instead, responders and nonresponders may differ in the pathophysiology of the schizophrenic symptoms and not in the pharmacokinetics of antipsychotic drug disposition. 3.2.2 p-Opiate Receptors
The lAo-opiate receptor system has been investigated with the opiate agonist [llC]carfentanil (Fisher & Frost 1991; Frost et al. 1989). Just as with the neuroleptic agents, similar questions re-
277
Positron Emission Tomography
garding the degree of receptor blockade and the time-course of occupancy vs plasma concentrations can be examined with PET. Naltrexone, an orally administered opiate receptor antagonist, is used to treat narcotic dependency states. On the basis of the [II C]carfentanil time-activity curves, the mean percentage blockade of the ~-opiate receptors at 48, 72, 120 and 168h after administration of naltrexone 50mg was 91, 80, 46 and 30%, respectively (Lee et al. 1988). Although naltrexone and its major active metabolite ~-naltrexol have apparent plasma t'/2 values of approximately 4 and 12h, respectively, the effective t'/2 of receptor occupancy was estimated to be 72 to 108h. Thus, the t'/2 for return to baseline opiate receptor occupancy corresponded more closely with the estimated t'/2 of the tertiary, terminal phase of the plasma-time activity curve (i.e. 96h) and the time-course for inhibition of the physiological and subjective effects of heroin (i.e. 72h) [Lee et al. 1988; Verebey et al. 1976] than it did with apparent plasma t'/2 values.
3.3 Toxicodynamics: Alcohol Alcohol is a drug with a significant potential for toxic effects. Using PET imaging, Volkow et al. (1988, 1990) have studied the toxicokinetic/dynamic effects of alcohol in healthy volunteers (social drinkers) and in alcoholics. The effects of low (0.5 gfkg) and moderate (1 g/kg) alcohol dosages on cerebral blood flow in healthy volunteers were investigated with [ 15 0]water (Volkow et al. 1988). Scans were performed at baseline and at 40 and 60 min after ingestion, and the results were compared with blood alcohol concentrations, subjective measures of intoxication and scores on a fine motor coordination examination. Although there were no changes in heart rate or blood pressure, subjective measures of intoxication ('relaxing effect', increased talkativeness, euphoria) started after approximately 20 min and peaked 50 min after administration. All study participants experienced impairment in fine motor coordination at the moderate dosage. After alcohol consumption, a significant relative reduction in blood flow to the cerebellum occurred with a relative increase in blood
flow to the prefrontal and right temporal cortices. The largest effects were recorded at the 60 min scan when the blood alcohol concentrations were the highest, a mean of 0.044 mg% at the low dosage and 0.087 mg% at the moderate dosage. The reductions in blood flow to the cerebellum were consistent with the observed disturbances in fine motor coordination. The increased blood flow to the prefrontal and temporal cortices was believed to be related to the observed general mood activation. The investigators concluded that the sensitivity of cerebellar blood flow to the effects of alcohol may explain the high incidence of ataxia during alcohol intoxication. Furthermore, the cerebral vasoactive response to alcohol could pose a possible explanation for the higher incidence of stroke in 'binge' drinkers. The effect of moderate dosages (1 gfkg) of alcohol on regional brain glucose transport and metabolism was studied in healthy volunteers and in alcoholics using [I8F]FDG (Volkow et al. 1990). Although there were no significant effects on the partial pressures of carbon dioxide and oxygen or on plasma glucose concentrations, alcohol administration resulted in a significant reduction in brain regional metabolism in cortical areas and cerebellum with relative sparing of the basal ganglia and corpus callosum. The pattern of decreased metabolism reflected the known distribution of benzodiazepine receptors. Examination of pharmacokinetic parameters revealed that the decreased metabolism stemmed from an alteration in k3 (phosphorylation rate) and not from alterations in either kl or k2 (transport into and out ofthe brain, respectively). The alcoholic patients experienced a greater decrease in the brain glucose metabolic rate than that experienced by the healthy volunteers. The authors hypothesised that this finding may be the result of an increased sensitivity of the benzodiazepine/'y-aminobutyric acid receptor complex in patients with chronic alcoholism. 3.4 Influence of Enantiomeric Forms
Drugs that possess an asymmetric carbon atom exist in enantiomeric forms. If the asymmetric carbon is located in a critical area or functional site
278
of the molecule, the enantiomeric form will influence the corresponding function (e.g. transport, metabolism, receptor binding) [Levy & Boddy 1991]. All processes that depend solely on the chemical nature of the molecule, such as passive diffusion, are believed to remain unchanged. While natural products generally exist in the active enantiomeric form only, synthetic products generally exist as racemic mixtures, unless specifically purified. The enantiomeric form of radiopharmaceuticals is important in PET imaging. Inactive enantiomers have played a role in the characterisation of nonspecific binding in PET receptor-binding studies (Farde et al. 1988b, 1989) where it is assumed that all dispositional factors between the enantiomeric forms are identical, with the exception of the specific binding component. In some cases, however, various dispositional processes may be influenced by enantiomeric form. The examples of PET imaging in the study of the disposition of nicotine and cocaine, both natural products and potent CNS stimulants, are presented below. 3.4.1 Nicotine The synthetic (+)-(R-) and the natural (-)-(L-) forms of nicotine, labelled with IIC, have been used to visualise nicotinic receptors in the brain (Halldin et al. 1991; Nyback et al. 1989a,b). After intravenous administration to humans, the IIC activity peaked in the arterial plasma after 2 min and in the brain after 4 to 6 min. Plasma concentrations did not differ between the (+) and (-) forms within individuals, but the steady-state plasma concentrations were slightly higher for smokers than nonsmokers. By the end of the imaging period (54 min), 25% of the (+) form had been metabolised to [11C]cotinine, whereas only 15% of the (-) form appeared as the metabolite. Recent studies indicate that this metabolite does not cross the blood-brain barrier (Halldin et al. 1991). In the brain, the regional distributions of activity resulting from the administration of the (+) and (-) enantiomers were similar, with the highest accumulations recorded in the cortical and subcortical regions and the lowest in the pons, cerebellum, occipital cortex and white
Clin. Pharmacokinet. 22 (4) 1992
matter of centrum semiovale. Distributional differences were observed between smokers and nonsmokers. The (+) form exhibited a lower peak and a slower decline in activity during the time-course of the study than did the (-) form. The investigators hypothesised that these differences were due to different binding profiles of the enantiomers to the nicotinic receptors and not a complete lack of affinity to the receptor by the synthetic enantiomer. 3.4.2 Cocaine [ II C]Cocaine has been used to map cocaine binding sites in the human brain (Fowler et al. 1989). After intravenous administration, IIC activity peaked in the brain (specifically the corpus striatum) at between 4 and 10 min and then declined with a t'l2 of 25 min. This pattern paralleled the time-course of cocaine-induced euphoria. In an effort to characterise the nonspecific binding, the inactive (+) enantiomer was labelled with IIC; when administered it resulted in no visible brain uptake (Gatley et al. 1990). Since the transport of cocaine into the brain is not believed to be stereoselective, alternative explanations were sought. It was found that by 30 sec after administration, plasma concentrations of the (+ )-cocaine were undetectable. In vitro work found that the (+)-cocaine was hydrolysed primarily to (+ )-ecgonine methyl ester by butyry1cholinesterase. This enzyme metabolises (+ )-cocaine 2000 times faster than the (-) form. Therefore, the difference in metabolism appears to be the determining factor for the distribution differences observed between the enantiomeric forms of cocaine.
4. Conclusion The purpose of this review is to acquaint the reader with the uniquely productive relationship between PET imaging and pharmacokinetics/pharmacodynamics. The latter provide the necessary tools for image quantification in PET imaging, and PET imaging provides noninvasive, in vivo distributional information necessary to address some of the difficult questions confronting pharmacokin-
Positron Emission Tomography
etic/pharmacodynamic investigation. PET imaging has developed into both a tool for and an application of clinical pharmacokinetics/dynamics in the 1990s and beyond.
Acknowledgements The authors would like to acknowledge the support and help of Dr G. Leonard Watkins and Dr Richard D. Hichwa in the preoparation of this manuscript.
References Andreasen NC, Carson R, Diksic M, Evans A, Farde L, et al. Workshop on schizophrenia, PET, and dopamine 02 receptors in the human neostriatum. Schizophrenia Bulletin 14: 471-484, 1988 Armbrecht JJ, Buxton DB, Brunken RC, Phelps ME, Schelbert HR. Regional myocardial oxygen consumption determined noninvasively in humans with (l_IIC)acetate and dynamic positron tomography. Circulation 80: 863-872, 1989 Arnett CD, Fowler JS, MacGregor RR, Schlyer OJ, Wolf AP, et al. Turnover of brain monoamine oxidase measured in vivo by positron emission tomography using L-[IIC)deprenyl. Journal of Neurochemistry 49: 522-527, 1987 Arnett CD, Fowler JS, Wolf AP, Logan J, MacGregor RR. Mapping brain neuroleptic receptors in the live baboon. Biological Psychiatry 19: 1365-1375, 1984 Arnett CD, Fowler JS, Wolf AP, Shiue CY, McPherson OW. [ISF)_ N-methylspiroperidol: the radioligand of choice for PETT studies of the dopamine receptor in human brain. Life Science 36: 1359-1366, 1985 Arnett CD, Shiue CY, Wolf AP, Fowler JS, Logan J, et al. Comparison of three IsF-labeled butyrophenone neuroleptic drugs in the baboon using positron emission tomography. Journal of Neurochemistry 44: 835-844, 1985 Arnett CD, Wolf AP, Shiue CY, Fowler JS, MacGregor RR, et al. Improved delineation of human dopamine receptors using [ISF)-N-methylspiroperidol and PET. Journal of Nuclear Medicine 27: 1878-1882, 1986 Bahn MM, Huang SC, Hawkins RA, Satyamurthy N, HotTman JM, et al. Models for in vivo kinetic interactions of dopamine D2-neuroreceptors and 3-(2'-[ ISF)fluoroethyl)spiperone examined with positron emission tomography. Journal of Cerebral Blood Flow and Metabolism 9: 840-849, 1989 Baron JC, Rougemont 0, Soussaline F, Bustany P, CrouzeI C, et al. Local interrelationships of cerebral oxygen consumption and glucose utilization in normal subjects and in ischemic stroke patients: a positron tomography study. Journal of Cerebral Blood Flow and Metabolism 4: 140-149, 1984 Barrio JR, Satyamurthy N, Huang SC, Keen RE, Nissenson CHI
Clin. Pharmacokinet. 22 (4): 274-283, 1992 0312-5963/92/0004-0274/$05.00/0 © Adis International Limited. All rights reserved. CPKl
Uses and Limitations of Positron Emission Tomography in Clinical Pharmacokinetics/Dynamics (Part II) Laura L. Boles Ponto and James A, Ponto PET Imaging Center, Department of Radiology, University of Iowa, Iowa City, Iowa, USA
Contents 274 274 275 275 277 277 278
Summary
Summary 3. The Use of PET in Pharmacokinetics 3.1 Route of Administration: Intra-Arterial vs Intravenous 3.2 Receptor Binding and Occupancy 3.3 Toxicodynamics: Alcohol 3.4 Influence of Enantiomeric Forms 4. Conclusion
Positron emission tomography (PET) involves imaging the biodistribution and tissue localisation of small amounts of radiolabelled biomolecules or drugs. In Part I of this article, which appeared in the previous issue of the Journal, the applications of pharmacokinetics in PET were discussed in order to derive quantitative measures of physiological function. Part II examines the use of PET imaging as a tool to study the pharmacokinetics and pharmacodynamics of specific drugs.
3. The Use of PET in Pharmacokinetics The use of PET imaging as a pharmacokinetic tool involves the radiolabelling of the drug of interest and the subsequent imaging and analysis of the time-course of the disposition of the agent. This approach has been employed primarily for receptor-binding ligands, especially those binding to eNS receptors, and to cancer chemotherapeutic agents. The use of PET imaging as a pharmacodynamic tool involves the imaging and analysis of the degree and time-course of pharmacologically-induced modifications in fundamental biological processes such as blood flow and/or metabolism. This ap-
proach is applicable whenever pharmacological or toxicological effects involve the induction of change in 1 of these fundamental processes (e.g. investigations into the effect of a2-adrenoceptor agonists on myocardial perfusion using successive 82Rb dynamic PET studies (Budinger et al. 1990) and the effect of adenosine on cerebral blood flow using [ 150]water (Sollevi et al. 1987)). PET studies of pharmacokinetics were reviewed previously by Yamamoto and Diksic (1985). They examined the subject from a pharmaceutical transport perspective. In contrast, this review will present representative examples of the types of phar-
275
Positron Emission Tomography
macokinetic/pharmacodynamic issues that can only be addressed in vivo by PET imaging. 3.1 Route of Administration: Intra-Arterial vs Intravenous A critical objective of any treatment regimen, but especially for those regimens involving highly toxic drugs such as cancer chemotherapeutic agents, is to maximise the desired therapeutic responses while minimising undesirable responses (adverse or toxicological effects). One of the choices which potentially influences this therapeutic balancing act is that of the route of administration. The pharmacological advantage of intra-arterial vs intravenous administration of 2 cancer chemotherapeutic agents, cisplatin and carmustine, has been evaluated for CNS tumours using the 13N_ and IIC-labelled derivatives, respectively (Ginos et al. 1987; Tyler et al. 1986). Intracarotid administration of cisplatin produced 1.1 to 2.5 times greater tumour/brain concentration ratios than did intravenous administration (Ginos et al. 1987). Superselective arterial administration of carmustine into a major tumour-feeding vessel (e.g. middle cerebral artery) provided tumour/healthy brain tissue concentration ratios of 50: I, whereas intravenous administration provided ratios of only 1.2: I (Tyler et al. 1986). Furthermore, ratios of tumour llC radioactivity concentrations between the intraarterial and intravenous routes ranged from 2.5 to 99 (n = 10). The 2 highest ratios were associated with patients demonstrating the most dramatic responses and the lowest ratios were associated with individuals experiencing essentially no therapeutic response. Similar investigations have been undertaken to evaluate the regional administration of fluorouracil in the treatment of metastatic liver disease (Strauss & Conti 1991). Examination of the short and lon~ term uptake of [18F]fluorouracil into metastatic lesions after intravenous and intra-arterial administration demonstrated that enhanced drug exposure through regional administration was only I factor involved in determining potential responsiveness to therapy. For fluorouracil, preferential
perfusion must also be accompanied by high tumour cell transport and metabolism. Therefore, delivery of the therapeutic agent, in this case fluorouracil, to the tumour is a necessary but not a sufficient predictor of response (Strauss & Conti 1991). 3.2 Receptor Binding and Occupancy Pharmacokinetic/pharmacodynamic investigations rely heavily on drug concentrations measured in accessible body fluids such as serum, plasma, whole blood, saliva, semen and/or urine. The use of these measures is predicated on the assumption that the concentrations in these fluids reflect the concentration at the site of action. Whenever the response pattern mirrors the concentration pattern in the measured fluid, this assumption is reinforced. When an inconsistent relationship is exhibited, the assumption is in question and a more direct measure is necessary. 3.2.1 Dopamine D2-Receptors The antipsychotic effect of neuroleptic drugs is believed to be due to the ability of these agents to block central dopamine receptors (Creese 1987; Losonczy et al. 1987). Unfortunately, plasma concentrations of neuroleptic agents are known to be imperfect predictors of response (Cohen et al. 1988; Kirch et al. 1988; Midha et al. 1989). Therefore, PET imaging studies have taken a more direct approach to the examination of the issues surrounding neuroleptic treatment. Specifically, 3 issues have been examined with respect to dopamine D2receptors and neuroleptic therapy (Farde et al. 1988a; Smith et al. 1988; Wolkin et al. 1989). First, what degree of receptor blockade is induced by clinically-administered dosages of various neuroleptic agents? Secondly, what is the time-course of receptor occupancy and plasma concentrations on neuroleptic withdrawal? Thirdly, do responders and nonresponders differ in neuroleptic-induced receptor blockade? Using [11C]raclopride, Farde et al. (1988a) studied central dopamine D2-receptor occupancy in 14 patients with schizophrenia who had received a
276
Clin. Pharmacokinet. 22 (4) 1992
single antipsychotic drug, at conventional dosages, for at least 1 month. Receptor occupancy ranged from 65 to 84% for all agents. One patient receiving nortriptyline exhibited no receptor occupancy. The authors concluded that, although all 11 neuro1eptics were chemically distinct entities, the degree of occupancy of central dopamine D2-receptors supported the proposed mechanism of antipsychotic drug action. The single patient in the study who experienced extrapyramidal side effects (akathisia) had the highest receptor occupancy at 86% on a dosage of haloperidol 6mg twice daily. After reduction of the dosage to 4mg twice daily, the side effects disappeared, but the receptor occupancy remained at 84%. Because of the relatively low incidence of extrapyramidal effects in this sample and the change in symptoms associated with a minor change in receptor occupancy, the investigators hypothesised that the antipsychotic action may require a lower receptor occupancy than for the development of extrapyramidal effects or that the 2 effects may be mediated by different receptor mechanisms. Smith et al. (1988) took an indirect approach to examining the relationship between [lsF]Nmethylspiroperidol ([lSF]NMS) uptake in the striatum and plasma concentration in healthy volunteers and patients with schizophrenia treated with either haloperidol or chlorpromazine. They found a decrease in the affinity of the [lSF]NMS for striatal tissues in the drug-treated patients with schizophrenia compared with the unmedicated volunteers, with the magnitude of this decrease directly related to the antipsychotic drug dosage (relative to the clinical potencies of the agents) and plasma concentration. When the antipsychotic drug was withdrawn from patients, the ratio index, a measure of receptor occupancy, approached that observed in healthy volunteers. A log-linear relationship was observed between the ratio index (y) and plasma haloperidol concentrations (x): y = 1.974 e-O.082x
r2 =
0.7482
(Eq. 35)
Using a more direct approach, Farde et al. (l988a) examined the serial receptor occupancy vs serum drug concentration after withdrawal of therapy in
3 of the patients described above. In patients treated with sulpiride 600mg twice daily and haloperidol 6mg twice daily, receptor occupancy remained essentially unchanged at approximately 65% and 85%, respectively, 27 and 54h after withdrawal, even though the serum drug concentrations had declined to approximately 25% of initial values. The third patient underwent a stepwise dosage reduction from sulpiride 800mg twice daily to 0 mg/day. The receptor occupancy exhibited a curvilinear relationship to the total daily dose whereas the serum concentration exhibited a linear relationship. The investigators concluded that the hyperbolic nature of the occupancy curve explained why major changes in drug concentration resulted in only minor changes in the receptor occupancy. Furthermore, they stated that PET imaging may be a method by which minimum individual dosages may be defined which provide adequate receptor occupancy for antipsychotic effects but minimise the potential for adverse effects. Although the above studies support the concept of a receptor occupancy-response relationship, additional factors may be involved in the response of schizophrenic patients to antipsychotic therapy. Wo1kin et al. (1989) examined dopamine-D2 blockade using [lsF]NMS in 2 biological subgroups of patients with schizophrenia, responders and nonresponders, treated with a maximum of haloperidol 100mg daily (minimum plasma concentration = 10 IAog/L) for 4 to 6 weeks. Comparable degrees of haloperidol-induced receptor blockade were found between responders and nonresponders. The authors concluded, therefore, that nonresponse to therapy may not necessarily be caused by inadequate concentrations of drug in the central nervous system. Instead, responders and nonresponders may differ in the pathophysiology of the schizophrenic symptoms and not in the pharmacokinetics of antipsychotic drug disposition. 3.2.2 p-Opiate Receptors
The lAo-opiate receptor system has been investigated with the opiate agonist [llC]carfentanil (Fisher & Frost 1991; Frost et al. 1989). Just as with the neuroleptic agents, similar questions re-
277
Positron Emission Tomography
garding the degree of receptor blockade and the time-course of occupancy vs plasma concentrations can be examined with PET. Naltrexone, an orally administered opiate receptor antagonist, is used to treat narcotic dependency states. On the basis of the [II C]carfentanil time-activity curves, the mean percentage blockade of the ~-opiate receptors at 48, 72, 120 and 168h after administration of naltrexone 50mg was 91, 80, 46 and 30%, respectively (Lee et al. 1988). Although naltrexone and its major active metabolite ~-naltrexol have apparent plasma t'/2 values of approximately 4 and 12h, respectively, the effective t'/2 of receptor occupancy was estimated to be 72 to 108h. Thus, the t'/2 for return to baseline opiate receptor occupancy corresponded more closely with the estimated t'/2 of the tertiary, terminal phase of the plasma-time activity curve (i.e. 96h) and the time-course for inhibition of the physiological and subjective effects of heroin (i.e. 72h) [Lee et al. 1988; Verebey et al. 1976] than it did with apparent plasma t'/2 values.
3.3 Toxicodynamics: Alcohol Alcohol is a drug with a significant potential for toxic effects. Using PET imaging, Volkow et al. (1988, 1990) have studied the toxicokinetic/dynamic effects of alcohol in healthy volunteers (social drinkers) and in alcoholics. The effects of low (0.5 gfkg) and moderate (1 g/kg) alcohol dosages on cerebral blood flow in healthy volunteers were investigated with [ 15 0]water (Volkow et al. 1988). Scans were performed at baseline and at 40 and 60 min after ingestion, and the results were compared with blood alcohol concentrations, subjective measures of intoxication and scores on a fine motor coordination examination. Although there were no changes in heart rate or blood pressure, subjective measures of intoxication ('relaxing effect', increased talkativeness, euphoria) started after approximately 20 min and peaked 50 min after administration. All study participants experienced impairment in fine motor coordination at the moderate dosage. After alcohol consumption, a significant relative reduction in blood flow to the cerebellum occurred with a relative increase in blood
flow to the prefrontal and right temporal cortices. The largest effects were recorded at the 60 min scan when the blood alcohol concentrations were the highest, a mean of 0.044 mg% at the low dosage and 0.087 mg% at the moderate dosage. The reductions in blood flow to the cerebellum were consistent with the observed disturbances in fine motor coordination. The increased blood flow to the prefrontal and temporal cortices was believed to be related to the observed general mood activation. The investigators concluded that the sensitivity of cerebellar blood flow to the effects of alcohol may explain the high incidence of ataxia during alcohol intoxication. Furthermore, the cerebral vasoactive response to alcohol could pose a possible explanation for the higher incidence of stroke in 'binge' drinkers. The effect of moderate dosages (1 gfkg) of alcohol on regional brain glucose transport and metabolism was studied in healthy volunteers and in alcoholics using [I8F]FDG (Volkow et al. 1990). Although there were no significant effects on the partial pressures of carbon dioxide and oxygen or on plasma glucose concentrations, alcohol administration resulted in a significant reduction in brain regional metabolism in cortical areas and cerebellum with relative sparing of the basal ganglia and corpus callosum. The pattern of decreased metabolism reflected the known distribution of benzodiazepine receptors. Examination of pharmacokinetic parameters revealed that the decreased metabolism stemmed from an alteration in k3 (phosphorylation rate) and not from alterations in either kl or k2 (transport into and out ofthe brain, respectively). The alcoholic patients experienced a greater decrease in the brain glucose metabolic rate than that experienced by the healthy volunteers. The authors hypothesised that this finding may be the result of an increased sensitivity of the benzodiazepine/'y-aminobutyric acid receptor complex in patients with chronic alcoholism. 3.4 Influence of Enantiomeric Forms
Drugs that possess an asymmetric carbon atom exist in enantiomeric forms. If the asymmetric carbon is located in a critical area or functional site
278
of the molecule, the enantiomeric form will influence the corresponding function (e.g. transport, metabolism, receptor binding) [Levy & Boddy 1991]. All processes that depend solely on the chemical nature of the molecule, such as passive diffusion, are believed to remain unchanged. While natural products generally exist in the active enantiomeric form only, synthetic products generally exist as racemic mixtures, unless specifically purified. The enantiomeric form of radiopharmaceuticals is important in PET imaging. Inactive enantiomers have played a role in the characterisation of nonspecific binding in PET receptor-binding studies (Farde et al. 1988b, 1989) where it is assumed that all dispositional factors between the enantiomeric forms are identical, with the exception of the specific binding component. In some cases, however, various dispositional processes may be influenced by enantiomeric form. The examples of PET imaging in the study of the disposition of nicotine and cocaine, both natural products and potent CNS stimulants, are presented below. 3.4.1 Nicotine The synthetic (+)-(R-) and the natural (-)-(L-) forms of nicotine, labelled with IIC, have been used to visualise nicotinic receptors in the brain (Halldin et al. 1991; Nyback et al. 1989a,b). After intravenous administration to humans, the IIC activity peaked in the arterial plasma after 2 min and in the brain after 4 to 6 min. Plasma concentrations did not differ between the (+) and (-) forms within individuals, but the steady-state plasma concentrations were slightly higher for smokers than nonsmokers. By the end of the imaging period (54 min), 25% of the (+) form had been metabolised to [11C]cotinine, whereas only 15% of the (-) form appeared as the metabolite. Recent studies indicate that this metabolite does not cross the blood-brain barrier (Halldin et al. 1991). In the brain, the regional distributions of activity resulting from the administration of the (+) and (-) enantiomers were similar, with the highest accumulations recorded in the cortical and subcortical regions and the lowest in the pons, cerebellum, occipital cortex and white
Clin. Pharmacokinet. 22 (4) 1992
matter of centrum semiovale. Distributional differences were observed between smokers and nonsmokers. The (+) form exhibited a lower peak and a slower decline in activity during the time-course of the study than did the (-) form. The investigators hypothesised that these differences were due to different binding profiles of the enantiomers to the nicotinic receptors and not a complete lack of affinity to the receptor by the synthetic enantiomer. 3.4.2 Cocaine [ II C]Cocaine has been used to map cocaine binding sites in the human brain (Fowler et al. 1989). After intravenous administration, IIC activity peaked in the brain (specifically the corpus striatum) at between 4 and 10 min and then declined with a t'l2 of 25 min. This pattern paralleled the time-course of cocaine-induced euphoria. In an effort to characterise the nonspecific binding, the inactive (+) enantiomer was labelled with IIC; when administered it resulted in no visible brain uptake (Gatley et al. 1990). Since the transport of cocaine into the brain is not believed to be stereoselective, alternative explanations were sought. It was found that by 30 sec after administration, plasma concentrations of the (+ )-cocaine were undetectable. In vitro work found that the (+)-cocaine was hydrolysed primarily to (+ )-ecgonine methyl ester by butyry1cholinesterase. This enzyme metabolises (+ )-cocaine 2000 times faster than the (-) form. Therefore, the difference in metabolism appears to be the determining factor for the distribution differences observed between the enantiomeric forms of cocaine.
4. Conclusion The purpose of this review is to acquaint the reader with the uniquely productive relationship between PET imaging and pharmacokinetics/pharmacodynamics. The latter provide the necessary tools for image quantification in PET imaging, and PET imaging provides noninvasive, in vivo distributional information necessary to address some of the difficult questions confronting pharmacokin-
Positron Emission Tomography
etic/pharmacodynamic investigation. PET imaging has developed into both a tool for and an application of clinical pharmacokinetics/dynamics in the 1990s and beyond.
Acknowledgements The authors would like to acknowledge the support and help of Dr G. Leonard Watkins and Dr Richard D. Hichwa in the preoparation of this manuscript.
References Andreasen NC, Carson R, Diksic M, Evans A, Farde L, et al. Workshop on schizophrenia, PET, and dopamine 02 receptors in the human neostriatum. Schizophrenia Bulletin 14: 471-484, 1988 Armbrecht JJ, Buxton DB, Brunken RC, Phelps ME, Schelbert HR. Regional myocardial oxygen consumption determined noninvasively in humans with (l_IIC)acetate and dynamic positron tomography. Circulation 80: 863-872, 1989 Arnett CD, Fowler JS, MacGregor RR, Schlyer OJ, Wolf AP, et al. Turnover of brain monoamine oxidase measured in vivo by positron emission tomography using L-[IIC)deprenyl. Journal of Neurochemistry 49: 522-527, 1987 Arnett CD, Fowler JS, Wolf AP, Logan J, MacGregor RR. Mapping brain neuroleptic receptors in the live baboon. Biological Psychiatry 19: 1365-1375, 1984 Arnett CD, Fowler JS, Wolf AP, Shiue CY, McPherson OW. [ISF)_ N-methylspiroperidol: the radioligand of choice for PETT studies of the dopamine receptor in human brain. Life Science 36: 1359-1366, 1985 Arnett CD, Shiue CY, Wolf AP, Fowler JS, Logan J, et al. Comparison of three IsF-labeled butyrophenone neuroleptic drugs in the baboon using positron emission tomography. Journal of Neurochemistry 44: 835-844, 1985 Arnett CD, Wolf AP, Shiue CY, Fowler JS, MacGregor RR, et al. Improved delineation of human dopamine receptors using [ISF)-N-methylspiroperidol and PET. Journal of Nuclear Medicine 27: 1878-1882, 1986 Bahn MM, Huang SC, Hawkins RA, Satyamurthy N, HotTman JM, et al. Models for in vivo kinetic interactions of dopamine D2-neuroreceptors and 3-(2'-[ ISF)fluoroethyl)spiperone examined with positron emission tomography. Journal of Cerebral Blood Flow and Metabolism 9: 840-849, 1989 Baron JC, Rougemont 0, Soussaline F, Bustany P, CrouzeI C, et al. Local interrelationships of cerebral oxygen consumption and glucose utilization in normal subjects and in ischemic stroke patients: a positron tomography study. Journal of Cerebral Blood Flow and Metabolism 4: 140-149, 1984 Barrio JR, Satyamurthy N, Huang SC, Keen RE, Nissenson CHI
