Anesth Prog 37:113-120 1990
Pain and
Opiate Receptors: Considerations for tie
Design of Positron Emission Tomography Studies B. Sadzot, MD, and J.J. Frost, MD, PhD The Johns Hopkins Medical Institutions, Baltimore, Maryland
nucleotides, etc. These factors are known to modulate many receptor-ligand interactions. For example, the binding of agonists to opiate receptors in vivo should be sensitive to the local concentration of NaCl and GTP which have been shown to modulate the receptor state in vitro. 14 Nonetheless, most binding assays are conducted in the absence of these ions and nucleotides. Second, the neuronal networks in which the receptors are included are disrupted during the process of tissue preparation. Finally, only one measurement can be performed in a given individual, after his death, whereas repeated measurements could be better to assess the natural history of a clinical condition or the response to a drug administration. In vivo studies avoid some of these problems. Attempts to label receptors after systemic administration of a tracer to a living animal have rapidly followed in vitro receptor identification.23 This approach still requires the death of the animal followed by regional tissue dissection, solubilization, and counting or tissue sectioning for autoradiography, and thus were not applicable to the living human. Parallel to the development of these binding techniques, positron emission tomography (PET) became available as a new method to externally, and thus noninvasively, detect with high sensitivity the distribution of trace amounts of radioactive tracer in a cross section of an organ, such as the brain. With appropriate calibration, it is also possible to accurately quantify the regional radioactivity distributed in the organ and therefore extract from the data parameters that are familiar to neuroscientists. PET was first devoted to the study of cerebral blood flow, and glucose or oxygen metabolism, and has provided new insights into human brain function.4 Rapid progress in radiochemistry has made it possible to label, with positron emitters, several of these specific ligands that were used to characterize receptors in vitro, without modifying their pharmacologic properties. It therefore appeared natural to extend these neuroreceptor binding assays to the living man using radiotracer techniques and PET. In vivo labeling of dopamine,5 benzodiazepine,68 serotonin,9 and opiate receptorsl' in the human brain has been reported. This research has the potential to enhance our understanding of the biochemical organization of the brain and
Opiate receptors in the brain are the target of endogenous opioids and of exogenous synthetic opiates. These receptors play a major role in the modulation of pain perception. Using the appropriate ligands, positron emission tomography now allows investigators to monitor neuroreceptors in vivo. We have used "IC-diprenorphine and the extremely potent mu opiate receptor agonist, "C-carfentanil, to image the distribution of opiate receptors in the brain and to quantify their density, their affinity, and their occupancy. Several important aspects of the in vivo opiate receptor labeling with positron emission tomography in relation to the study of pain are considered in this paper. Monitoring receptor occupancy by opiate drugs as a function of pain relief has the potential to reveal better ways to treat pain.
he development of relatively simple in vitro binding techniques has resulted in the identification, localization, and characterization of a large number of neuroreceptors. These specific macroproteins, targets of endogenous neurotransmitters, and exogenous psychoactive drugs, have been extensively studied in animal and in postmortem human brains using in vitro binding techniques such as membrane binding assays or autoradiography. These techniques have been very useful in rapidly characterizing drug-receptor interactions and have been decisive to advances in neuropharmacology. In vitro methods, however, suffer several limitations. First, tissue homogenization disrupts the biochemical environment of the receptors, e.g., ions, pH, temperature,
Address correspondence to J.J. Frost, MD, PhD, Department of Radiology, Division of Nuclear Medicine, the Johns Hopkins Hospital, Nelson Tower Basement, B1-119, Baltimore, MD, 21205. X 1990 by the American Dental Society of Anesthesiology
ISSN 0003-3006/90/$3.50
113
Anesth Prog 37:113-120 1990
114 Opiate Receptors and PET in Pain
the mechanisms of drug action. Though the prospect of having a direct access to biochemical processes in the living brain is very exciting, our enthusiasm must be somewhat tempered by the realization that a number of technical, scientific, and economic issues limit the investigations that can be carried out with this technique. In the last few years, our group has acquired substantial experience in labeling opiate receptors in vivo with PET. 10-13 In this paper, we will briefly review our experience with this receptor and we will choose the clinical condition, pain, to illustrate some general principles and limitations of receptor research with PET, as well as problems that must be considered when designing clinical studies with this technique. SELECTION OF RADIOLIGANDS FOR IMAGING RECEPTOR IN VIVO Imaging receptors with PET depend critically on the selection of appropriate radiotracers. Ligands have to meet some general criteria, valid for any type of binding assay, but also a number of criteria that are specific to the in vivo situation. It is a general rule that natural neurotransmitters cannot be used to label receptors in vivo. They are either of low affinity, rapidly metabolized, or do not cross the blood brain barrier (BBB). Fortunately, more and more synthetic drugs are available that mimic endogenous neurotransmitters. Drugs must have some of the following characteristics before any in vivo binding assay can be considered. The drug must be of very high affinity for its receptor: all the radioligands available so far for in vivo studies have a KD equal or less than 1 nM when the affinity is measured in vitro. Because the binding assay cannot be systematically conducted with unlabeled competitors as under in vitro conditions, preference sh6uld be given to drugs known to be selective for one receptor type or receptor subtype
whenever possible. The specificity of the ligand can be inferred from limited competition studies as well as from the distribution of the radioactivity in the appropriate neuroanatomical system. It then goes without saying that the drug must cross the blood brain barrier. Indications of a BBB passage can be obtained from pharmacologic or behavioral studies after systemic administration of a reasonable dose of the unlabeled ligand. Then, the radiochemistry must be considered. The drugs must be labeled with a positron emitter ("lC, 18F, 75Br), the choice of which will depend on the molecular structure of the drug, the ease and speed of the synthesis route, the type of cyclotron available, and radiodosimetry considerations (the longer the half-life, the highest the absorbed dose for a same injected dose, which can be a
limiting factor for repeated studies). Receptor concentration in the brain is low (nM), and receptor measurement should be performed at receptor occupanices not greater than 10%, in most cases. For these reasons, high specific activities (SA) must be achieved to limit the mass of ligand administered and avoid partial saturation of the receptors, which would lower the signal. The high sensitivity of PET allows detection of subnanomolar concentration of ligand. The background, or nonspecific binding, which can be estimated either by a given saturating dose of cold competitor, or in a region devoid of receptors, must be as low as possible to maximize the sensitivity of the method. It is determined by several factors including drug lipophilicity, metabolism, and ionization. It is generally difficult to predict from in vitro methods if a radioligand will have a low level of nonspecific binding in vivo; preliminary animal studies are useful in this respect. Finally, the radioactivity detected in the brain should correspond only to the tracer itself. Metabolites released from peripheral organs into the circulation might cross the BBB, or metabolism of the drug could occur locally. The presence of radiolabeled metabolites would complicate the interpretation of the PET measurement.
RADIOLIGANDS FOR OPIATE RECEPTORS Our group has selected and labeled with "lC two ligands that proved to be very useful to label opiate receptors, "C-carfentanil and "IC-diprenorphine (Table 1). Carfentanil is an 4-anilinopiperidine derived from fentanyl (Figure 1). It is an extremely potent and selective mu opiate agonist. Its potency in pharmacologic tests of analgesia is 8-10 thousand times that of morphine,14 and its affinity in vitro is 0.05 nM.10 Carfentanil is labeled with 1"C. It is noteworthy that very high SA is required and to avoid the side-effects the administered mass of carfentanil must be kept under 0.1 ,ug/kg. Carfentanil is approximately 90and 250-fold less potent at delta and kappa receptors than
Table 1. Comparison of Some Characteristics of "C-Carfentanil and "C-Diprenorphine
11C-Carfentanil In vitro affinity Receptor selectivity (in vitro) BBB permeability Pharmacology Typical SA
0.05 nM mu receptors
'1C-Diprenorphine 0.2 nM mu, delta and kappa receptors
++
+++
Agonist 1900
Weak partial agonist 2300
(mCi/mmol)
BBB: blood brain bamier.
Sadzot and Frost 1 1 5
Anesth Prog 37:113-120 1990 o 1 C-0- ' CH3
G& CH2-CH2-N?
NCH2-
Pain and
Opiate Receptors: Considerations for tie
Design of Positron Emission Tomography Studies B. Sadzot, MD, and J.J. Frost, MD, PhD The Johns Hopkins Medical Institutions, Baltimore, Maryland
nucleotides, etc. These factors are known to modulate many receptor-ligand interactions. For example, the binding of agonists to opiate receptors in vivo should be sensitive to the local concentration of NaCl and GTP which have been shown to modulate the receptor state in vitro. 14 Nonetheless, most binding assays are conducted in the absence of these ions and nucleotides. Second, the neuronal networks in which the receptors are included are disrupted during the process of tissue preparation. Finally, only one measurement can be performed in a given individual, after his death, whereas repeated measurements could be better to assess the natural history of a clinical condition or the response to a drug administration. In vivo studies avoid some of these problems. Attempts to label receptors after systemic administration of a tracer to a living animal have rapidly followed in vitro receptor identification.23 This approach still requires the death of the animal followed by regional tissue dissection, solubilization, and counting or tissue sectioning for autoradiography, and thus were not applicable to the living human. Parallel to the development of these binding techniques, positron emission tomography (PET) became available as a new method to externally, and thus noninvasively, detect with high sensitivity the distribution of trace amounts of radioactive tracer in a cross section of an organ, such as the brain. With appropriate calibration, it is also possible to accurately quantify the regional radioactivity distributed in the organ and therefore extract from the data parameters that are familiar to neuroscientists. PET was first devoted to the study of cerebral blood flow, and glucose or oxygen metabolism, and has provided new insights into human brain function.4 Rapid progress in radiochemistry has made it possible to label, with positron emitters, several of these specific ligands that were used to characterize receptors in vitro, without modifying their pharmacologic properties. It therefore appeared natural to extend these neuroreceptor binding assays to the living man using radiotracer techniques and PET. In vivo labeling of dopamine,5 benzodiazepine,68 serotonin,9 and opiate receptorsl' in the human brain has been reported. This research has the potential to enhance our understanding of the biochemical organization of the brain and
Opiate receptors in the brain are the target of endogenous opioids and of exogenous synthetic opiates. These receptors play a major role in the modulation of pain perception. Using the appropriate ligands, positron emission tomography now allows investigators to monitor neuroreceptors in vivo. We have used "IC-diprenorphine and the extremely potent mu opiate receptor agonist, "C-carfentanil, to image the distribution of opiate receptors in the brain and to quantify their density, their affinity, and their occupancy. Several important aspects of the in vivo opiate receptor labeling with positron emission tomography in relation to the study of pain are considered in this paper. Monitoring receptor occupancy by opiate drugs as a function of pain relief has the potential to reveal better ways to treat pain.
he development of relatively simple in vitro binding techniques has resulted in the identification, localization, and characterization of a large number of neuroreceptors. These specific macroproteins, targets of endogenous neurotransmitters, and exogenous psychoactive drugs, have been extensively studied in animal and in postmortem human brains using in vitro binding techniques such as membrane binding assays or autoradiography. These techniques have been very useful in rapidly characterizing drug-receptor interactions and have been decisive to advances in neuropharmacology. In vitro methods, however, suffer several limitations. First, tissue homogenization disrupts the biochemical environment of the receptors, e.g., ions, pH, temperature,
Address correspondence to J.J. Frost, MD, PhD, Department of Radiology, Division of Nuclear Medicine, the Johns Hopkins Hospital, Nelson Tower Basement, B1-119, Baltimore, MD, 21205. X 1990 by the American Dental Society of Anesthesiology
ISSN 0003-3006/90/$3.50
113
Anesth Prog 37:113-120 1990
114 Opiate Receptors and PET in Pain
the mechanisms of drug action. Though the prospect of having a direct access to biochemical processes in the living brain is very exciting, our enthusiasm must be somewhat tempered by the realization that a number of technical, scientific, and economic issues limit the investigations that can be carried out with this technique. In the last few years, our group has acquired substantial experience in labeling opiate receptors in vivo with PET. 10-13 In this paper, we will briefly review our experience with this receptor and we will choose the clinical condition, pain, to illustrate some general principles and limitations of receptor research with PET, as well as problems that must be considered when designing clinical studies with this technique. SELECTION OF RADIOLIGANDS FOR IMAGING RECEPTOR IN VIVO Imaging receptors with PET depend critically on the selection of appropriate radiotracers. Ligands have to meet some general criteria, valid for any type of binding assay, but also a number of criteria that are specific to the in vivo situation. It is a general rule that natural neurotransmitters cannot be used to label receptors in vivo. They are either of low affinity, rapidly metabolized, or do not cross the blood brain barrier (BBB). Fortunately, more and more synthetic drugs are available that mimic endogenous neurotransmitters. Drugs must have some of the following characteristics before any in vivo binding assay can be considered. The drug must be of very high affinity for its receptor: all the radioligands available so far for in vivo studies have a KD equal or less than 1 nM when the affinity is measured in vitro. Because the binding assay cannot be systematically conducted with unlabeled competitors as under in vitro conditions, preference sh6uld be given to drugs known to be selective for one receptor type or receptor subtype
whenever possible. The specificity of the ligand can be inferred from limited competition studies as well as from the distribution of the radioactivity in the appropriate neuroanatomical system. It then goes without saying that the drug must cross the blood brain barrier. Indications of a BBB passage can be obtained from pharmacologic or behavioral studies after systemic administration of a reasonable dose of the unlabeled ligand. Then, the radiochemistry must be considered. The drugs must be labeled with a positron emitter ("lC, 18F, 75Br), the choice of which will depend on the molecular structure of the drug, the ease and speed of the synthesis route, the type of cyclotron available, and radiodosimetry considerations (the longer the half-life, the highest the absorbed dose for a same injected dose, which can be a
limiting factor for repeated studies). Receptor concentration in the brain is low (nM), and receptor measurement should be performed at receptor occupanices not greater than 10%, in most cases. For these reasons, high specific activities (SA) must be achieved to limit the mass of ligand administered and avoid partial saturation of the receptors, which would lower the signal. The high sensitivity of PET allows detection of subnanomolar concentration of ligand. The background, or nonspecific binding, which can be estimated either by a given saturating dose of cold competitor, or in a region devoid of receptors, must be as low as possible to maximize the sensitivity of the method. It is determined by several factors including drug lipophilicity, metabolism, and ionization. It is generally difficult to predict from in vitro methods if a radioligand will have a low level of nonspecific binding in vivo; preliminary animal studies are useful in this respect. Finally, the radioactivity detected in the brain should correspond only to the tracer itself. Metabolites released from peripheral organs into the circulation might cross the BBB, or metabolism of the drug could occur locally. The presence of radiolabeled metabolites would complicate the interpretation of the PET measurement.
RADIOLIGANDS FOR OPIATE RECEPTORS Our group has selected and labeled with "lC two ligands that proved to be very useful to label opiate receptors, "C-carfentanil and "IC-diprenorphine (Table 1). Carfentanil is an 4-anilinopiperidine derived from fentanyl (Figure 1). It is an extremely potent and selective mu opiate agonist. Its potency in pharmacologic tests of analgesia is 8-10 thousand times that of morphine,14 and its affinity in vitro is 0.05 nM.10 Carfentanil is labeled with 1"C. It is noteworthy that very high SA is required and to avoid the side-effects the administered mass of carfentanil must be kept under 0.1 ,ug/kg. Carfentanil is approximately 90and 250-fold less potent at delta and kappa receptors than
Table 1. Comparison of Some Characteristics of "C-Carfentanil and "C-Diprenorphine
11C-Carfentanil In vitro affinity Receptor selectivity (in vitro) BBB permeability Pharmacology Typical SA
0.05 nM mu receptors
'1C-Diprenorphine 0.2 nM mu, delta and kappa receptors
++
+++
Agonist 1900
Weak partial agonist 2300
(mCi/mmol)
BBB: blood brain bamier.
Sadzot and Frost 1 1 5
Anesth Prog 37:113-120 1990 o 1 C-0- ' CH3
G& CH2-CH2-N?
NCH2-
