Exploratory IND Application: Evaluation of [11C](-)-RWAY
PET IMAGING OF BRAIN 5-HT1A RECEPTORS USING [11C](-)-RWAY
I.Introductory Statement of Purpose and General Plan
II.Study Protocol: PET Imaging of Brain 5-HT1A Receptors Using [11C](-)-RWAY
A.Precis
III.Chemistry, Manufacture and Control
IV.Pharmacology and Toxicology
A.Pharmacology.
B.In Vitro Receptor Binding.
C.Toxicology.
V.Animal Experimentation.
A.Metabolism and Clearance of (-)-RWAY.
B.PET Imaging in Nonhuman Primates.
C.Pharmacological Effects in Nonhuman Primates.
D.Estimated Receptor Occupancy.
E.Radiation Dosimetry Estimation from Non-human Primates
V.Human Experience
VI.Environmental Assessment
VII.Case Report Form
VIII.References
IX.APPENDICES
A.Investigator data: CV of the investigator
B.Study Protocol.
C.IRB Protocol Approval.
D.Dr. Mann's letter to FDA authorizing reference of his IND
E.Dr. Herscovitch’s letter to FDA authorizing reference of his IND
F.Case Report Form
G.Dosimetry Report from Nonhuman Primates
NOTE: Pages in the Appendices are numbered A1, A2, A3 etc; then B1, B2, B3, etc.
Chemistry, Manufacturing, and Controls
Document No. / Section1 / CMC-RWAY
2 / Master Batch Record
3 / QC Form
4 / Radiopharmacy Form
5 / SOP: Standard Operating Procedures
6 / Annual Testing for Radionuclidic Identity
7 / Certificate of Analysis for starting Materials
8 / Data Form for Generating Calibration Curve
9 / Precursor Acceptance Testing and Form
10 / Validation Runs
PET IMAGING OF BRAIN 5-HT1A RECEPTORS USING [11C](-)-RWAY
I.Introductory Statement of Purpose and General Plan
Overview. This application is being submitted under the exploratory IND Guidance. We request permission to study no more than 15 healthy subjects to determine whether [11C](-)-RWAY is a good PET probe for the serotonin 5-HT1A receptor. Our imaging and metabolism studies in rats and monkeys suggest it will be an excellent tracer. The results of kinetic brain imaging in humans with [11C](-)-RWAY will be compared with published studies using [11C]WAY-100635 and [18F]FCWAY. If the results with [11C](-)-RWAY are adequately promising, we understand that we must have approval from the FDA to study any additional subjects.
We request permission for a single injection in no more than 15 healthy subjects. The injected activity will be ≤ 20 mCi, with an associated mass ≤ 10 µg. For a 70 kg subject, this corresponds to a maximum mass dose of 0.14 µg/kg. For comparison purposes, the no effect level of FCWAY in 14-day study was 300 µg/kg in rabbits (corresponding to HED = 97 µg/kg). Thus, the requested single mass dose is 690 fold lower (= 97 / 0.14) than the no effect level from a 14-day study in rabbits.
Serotonin Receptor Subtypes. Serotonin (5-HT) is a major brain neurotransmitter that elicits a multitude of physiological function by interactions with seven classes of receptors (5-HT1-5-HT7) and their 14 subtypes which have been identified to date. Among all these receptors, 5-HT1A received the most attention, and therefore is structurally, pharmacologically, and functionally the best characterized (Lanfumey and Hamon 2004). 5-HT1A receptors show an abundant and comparable expression in the brain of mammals, including humans. Brain 5-HT1A receptors are located both pre- and postsynaptically. Presynaptic 5-HT1A receptors (5-HT1A autoreceptors) provide a mechanism for the feedback inhibition of the 5-HT system. Activation of postsynaptic 5- HT1A receptors is generally believed to decrease the firing rate of postsynaptic cells (Toth, 2003).
There is substantial evidence from studies of postmortem tissue to link abnormal levels of brain 5-HT1A receptors to neuropsychiatric diseases, including depression (Matsubara et al., 1991) and schizophrenia (Burnett et al., 1997). Furthermore, the 5-HT1A receptor has been in the past and remains today an important target for drug discovery. Selective agonists (e.g., buspirone) are used widely as anxiolytics and antidepressants (Goldberg and Finnerty, 1979). Moreover, there is now important evidence that certain 5-HT1A antagonists (e.g., pindolol) accelerate the antidepressant efficacy of widely prescribed selective serotonin reuptake inhibitors (SSRIs) (Blier and Bergeron, 1995). Therefore, it is highly desirable to be able to image 5-HT1A receptors in human brain in vivo to gain insight into the involvement of these receptors in neuropsychiatric diseases and their current therapies.
PET Imaging of 5-HT1A Receptors. An important breakthrough for imaging 5-HT1A receptors was the radiolabeling of the potent and selective silent antagonist, WAY-100635 (Fig. 1, middle panel), with carbon-11 (11C, T1/2 = 20.3 min) to provide the first radioligand for imaging 5-HT1A receptors in human brain with positron-emission tomography (PET) (Pike et al., 1995a). Initially, the radiolabel was placed in the O-methoxy position (Fig. 2, Pike et al., 1995b). However, in primates this radioligand is primarily metabolised by amide hydrolysis to the labelled amine ([11C]WAY-100634) (Osman et al., 1996) which has high affinity for 5-HT1A receptors and α1-adrenoceptors. This metabolite was labelled with carbon-11 in the laboratory and was shown to penetrate monkey brain to bind nonspecifically in cerebellum and other regions, which detracts from the accuracy of applying a simplified reference tissue model (Osman et al., 1996). [carbonyl-11C]WAY-100635 (Fig. 2) is effective for imaging brain 5-HT1A receptors in vivo, but suffers from rapid metabolism by amide hydrolysis, which acts against definition of an arterial-input function for compartmental modeling. This factor requires that this radioligand be labeled at its carbonyl position through a technically demanding process, and also detracts from the ease with which ligand-binding parameters may be calculated from PET data. Thus, [carbonyl-11C]WAY-100635 has two significant limitations: very rapid hydrolysis of the amide bond and significant difficulty to synthesize. That is, “internal” labeling in the carbonyl position is much more difficult than “external” labeling of the methoxy position in WAY-100635.
[18F]FCWAY (Fig. 1, lower panel) was developed at NIH as a longer-lived alternative to 11C-labeled WAY-100635 (Lang et al, 1999). Based upon imaging in nonhuman primates, [18F]FCWAY was an excellent candidate (Carson et al., 2000). Unfortunately, humans (unlike monkeys) showed significant metabolic defluorination with subsequent uptake of 18F fluoride in bone (Toczek et al., 2003). In fact, the skull uptake of fluoride from [18F]FCWAY is so significant that spillover of activity precludes accurate quantitation of underlying neocortex in humans.
In summary, each of the three commonly used PET tracers for the 5-HT1A receptor has a significant deficiency. [O-methoxy-11C]WAY-100635 is metabolized at the amide position and generates a radiometabolite that enters brain; [carbonyl-11C]WAY-100635 is difficult to synthesize; and [18F]FCWAY generates significant uptake of skull activity. We sought to develop an improved PET ligand that was resistant to amide hydrolysis; labeled by 11C (and thus avoid skull uptake); and also easily labeled in an “external” methoxy position.
Based on animal studies, [11C](-)-RWAY (Fig. 1. upper panel) fulfills these three criteria. First, [11C](-)-RWAY is easily labeled in the methoxy position. Second, Even though similar to WAY-100635, [11C](-)-RWAY has a ‘reversed’ amide linkage (Fig. 1, upper panel). Mass spectrometry (MS) studies of (-)-RWAY have shown that the indicated amide bond in (-)-RWAY is metabolically resistant (Shetty 2005). Third, PET imaging in non-human primates with [11C](-)-RWAY shows that it readily enters brain but has a slower washout than [carbonyl-11C]WAY-100635 (McCarron et al., 2004).
Research on 5-HT1A receptor ligands has demonstrated the difficulties of developing an optimal PET tracer and the limitation of using even a closely related species like rhesus monkey to predict results in human subjects. For example, [18F]FCWAY showed excellent results in monkeys but later demonstrated significant defluorination in humans. We understand that one goal of the exploratory IND is to facilitate the development of PET biomarkers by allowing a new ligand to be studied in a limited number of subjects to more quickly “rule in” or “rule out” a new candidate. Thus, the purpose of this protocol is to test [11C](-)-RWAY in a limited number of healthy human subjects to see if it is a good ligand compared to the currently available tracers [18F]FCWAY (Lang et al, 1999) and [carbonyl-11C]WAY-100635 (Pike et al., 1995a). [11C](-)-RWAY will be assessed based upon the amount of brain uptake, the ratio of uptake in target (e.g., cortex) and background regions (e.g., cerebellum), and the identifiability of distribution volume measured with both compartmental and reference tissue models (Ichise et al., 2003).
II.Study Protocol: PET Imaging of Brain 5-HT1A Receptors Using [11C](-)-RWAY
A.Precis
The complete protocol is located in Appendix B. A brief summary is included below.
The 5-hydroxytryptamine (serotonin, 5-HT) 5-HT1A receptors subtype is a target for drug therapy in the treatment of anxiety and depress (Lanfumey and Hamon, 2004). Radioligands currently in use with PET for studying human brain 5-HT1A receptors in clinical research or drug development are based on WAY-100635 (Fig. 1, middle panel). Though variously effective, they each suffer from one or more drawbacks with respect to measuring relative regional receptor densities. These drawbacks include: rapid metabolism and very low non-specific binding for [carbonyl-11C]WAY-100635; defluoridation for [18F]FCWAY. Therefore, we have recently developed [11C](-)-RWAY (Fig. 1, upper panel) as an alternative radioligand for brain 5HT1A receptors. Even though similar to WAY-100635, [11C](-)-RWAY has a ‘reversed’ amide linkage. Reversal of the direction of the amide bond may confer resistance to amide hydrolysis and make acceptable easy labeling at its O-methoxy group. In the present protocol, we plan to perform a kinetic brain imaging study in healthy human subjects to measure 5HT1A receptors in brain regions with [11C](-)-RWAY (active enantiomer of RWAY).
III.Chemistry, Manufacture and Control
The CMC Section is a separate portion of this application.
IV.Pharmacology and Toxicology
A.Pharmacology.
The pharmacological profile of (-)-RWAY as a selective 5-HT1A antagonist was well established by Cliffe and Dourish et al in their patent applications (Eur. Patent Appl. 0481744 and UK Patent Appl.GB 2303 303). For example, in Cliffe’s patent application, (-)-RWAY was tested for 5-HT1A receptor binding activity in rat hippocampal membrane homogenate and was found to have an IC50 of 3 nM. (-)-RWAY was also tested for 5-HT1A receptor antagonism activity in a test involving the antagonism of 8-OH DPAT syndrome in the rats with a minimum effective dose (MED) of 0.03 mg/kg, s.c. In functional assay, (-)-RWAY was tested for potential anxiolytic activity by a test procedure measuring mouse exploratory activity in two-compartment light/dark box with the MED of 0.03 mg/kg, s.c. (Cliffe, Eur. Patent Appl. 0481744). In Dourish et al’s patent application (UK Patent Appl.GB 2303 303), functional tests were performed to verify that (-)-RWAY acts as an antagonist but not agonist in pharmacological models of 5-HT1A receptor by either electrophysiologically monitoring the activity of the neurons to measure their firing rate or by studying the effect of (-)-RWAY on 5-HT release in the hippocampus using in vivo microdialysis. Both studies strongly indicated that (-)-RWAY, like its closely related chemical analog WAY-100635, is a selective 5-HT1A antagonist.
B.In Vitro Receptor Binding.
Radioligand-binding assays were performed by using the resources of the NIMH-PDSP (Pharmacology Drug Screening Program). Detailed on-line protocols are available at the NIMH-PDSP web site ( Ki values were calculated using LIGAND program. The results demonstrated that (-)-RWAY has high affinity for cloned human 5-HT1A receptors with a Ki of 0.6 nM, which is at least 10-fold higher than its affinity for 5-HT1D, 5-HT1B, alpha1a, D3, D4 receptors and 100-fold higher than its affinity for the rest of the receptors on the list of Table 1.
Table 1. Binding properties of (-)-RWAY and WAY-100635 at receptors (mean ± SD)(-)-RWAY / WAY100635
Receptor Subtype / Ki (nM) / Ratio to 5-HT1A / Ki (nM) / Ratio to 5-HT1A
5HT1a / 0.6 / 1 / 2.2 / 1
5HT1b / 444.7 / 741
5HT1d / 9.487 / 16 / >10,000 / >10,000
5HT2a / 147.5 / 246 / 6260 / 2845
5HT2b / 7.2 / 12 / 24.2 / 11
5HT2c / 1682 / 2803 / >10,000 / >10,000
5HT5a / 306 / 510 / >10,000 / >10,000
5HT7 / 65.4 / 109
alpha1a / 10.35 / 17 / 19.9 / 9
alpha1b / 222 / 370 / 322.8 / 147
alpha2a / 1311 / 2185
alpha2b / 1476 / 2460
alpha2c / 92 / 153 / 2573 / 1170
Beta1 / 770 / 1283
Beta2 / >10,000 / >10,000
D1 / 732 / 1220
D2 / 34.5 / 58 / 936 / 425
D3 / 5.1 / 9 / 365.5 / 166
D4 / 15.6 / 26 / 16.4 / 7
H1 / 335.4 / 559 / >10,000 / >10,000
H2 / 3296 / 5493
Kappa Opiate / 1609 / 2682 / 411 / 187
SERT / 3269 / 5448
The values for (-)-RWAY were compared to WAY-100635, which had previously been studied by PDSP. Please note that WAY-100635 does not have stereoisomers. [11C]WAY-100635 is the most commonly used PET tracer for the 5-HT1A receptor (Pike et al., 2000). To compare (-)-RWAY and WAY-100635, the above Table lists the ratio of the Ki value for each receptor to that for the 5-HT1A receptor. The relative affinities of (-)-RWAY and WAY-100635 are similar except for the 5-HT1D and the D3 receptors. The actual binding to these alternate sites will depend upon affinity and the density of sites. (-)-RWAY is 9 fold less potent at D3 than 5-HT1A receptors, and the density is about 10 times lower (see below). Thus, the binding of this radioligand to D3 would be more than 900 fold lower than to 5-HT1A, which itself is occupied at tracer amounts. Binding to the 5-HT1D receptor is even less problematic, since the affinity is 16 fold lower, and the density of these sites is very low in human brain (Bonaventure et al., 1997).
The density of of 5-HT1A in human brain target ranges is about 100 – 200 fmol/mg wet tissue (Burnet et al., 1997). In contrast the density of D3 receptors is about 1-3 fmol/mg wet weight tissue (Quik et al., 2000). [Please note that some papers report receptor density per mg protein, which can be converted to mg wet weight assuming 1 mg protein in brain corresponds to ~10 mg tissue.]
We do not have full receptor screening data on FCWAY, but its Ki for 5-HT1A receptors is reported to be 1.0 nM (Lang et al., 1999)
C.Toxicology.
The FDA description of the exploratory IND states: “This guidance clarifies what preclinical and clinical approach (including chemistry, manufacturing, and controls) should be considered when planning exploratory IND studies in humans, including studies of closely related drugs or therapeutic biologics products , under an investigational new drug (IND) application (21 CFR 312)”. We have interpreted this statement to mean that the exploratory IND can be used to study several chemically related compounds for a single target. For example, the investigator may have 5 similar chemical structures as ligands for the dopamine D2 receptor. Animal toxicity study for one of the agent can presumably be used to study a small number of subjects with each compound. The current exploratory IND application is based on this principle. (-)-RWAY is a closely related chemical analog of both FCWAY and WAY-100635. We have obtained animal toxicity data on these two related compounds: FCWAY via Peter Herscovitch, MD (NIH PET Department) and WAY-100635 via J. John Mann, MD (Columbia University). The data from these two toxicity studies provide strong evidence of the safety of the proposed mass dose of (-)-RWAY. Please note that Dr. Mann’s IND or other studies have safely used 700 µg/kg daily dose of WAY-100635 in human. In the 14-Day Toxicity Study of FCWAY in Rabbits (IND #59,163), intravenous administration of FCWAY at doses of 0.15 and 0.3 mg/kg/day for 14 consecutive days did not result in treatment-related histologic lesions. The no-effect level for the study was 0.3 mg of FCWAY/kg body weight /day (corresponding to HED = 97 µg/kg). For the current study, the injected mass dose of (-)-RWAY will be less than 10 µg, which corresponds to less than 0.14 µg/kg. Thus, the requested single mass dose is 690 fold lower (= 97 / 0.14) than the no effect level from a 14-day study in rabbits.
IV.Hidden Section IV
V.Animal Experimentation.
A. Metabolism and Clearance of (-)-RWAY.
The in vivo stability of the [11C](-)-RWAY was investigated in both rat and rhesus monkey (Macacca mulatta). One Sprague Dawley rat (300 g) was injected with [11C](-)-RWAY (1.0 mCi) and another with (-)-RWAY (500 µg). The radioactive and non-radioactive animals were killed at 30 min and 2.4 h after injection, respectively. Urine and plasma samples were collected from both animals. The radioactive samples were analyzed by RP-radio-HPLC. The LC-MS analysis of urinary metabolites (non-radioactive) involved HPLC separation on a C18 column, electrospray ionization and MS and MS-MS detections using ThermoFinnigan’s LCQ Deca instrument.
In rat administered with [11C](-)-RWAY, the urine and plasma radiometabolite profiles were found to be comparable. LC-MS analysis of the urine sample (from the non-radioactive experiment) generated mass spectral data that account for several of the metabolites predicted for (-)-RWAY. The metabolite screening by MS indicated that the amide bond in [11C](-)-RWAY or ()-RWAY is metabolically stable, unlike that in its structural analog, WAY-100635.
In our monkey study, the 11.9 kg monkey received 4.25 mCi (1.3 µg) of [11C](-)-RWAY while under 1.6% isoflurane and 98.4% O2 anesthesia. Non-radioactive [12C](-)-RWAY was injected IV 40 min after the radioactivity dose and urine was collected, thereafter, from the catheterized monkey for 2 h. Since we have found out that urine radiometabolites of [11C](-)-RWAY mirror those of the plasma (Shetty et al 2005), we therefore subjected the monkey urine for LC-MS analysis. Urine analysis detected parent (-)-RWAY along with two hydroxy-, a keto- and an N-desphenyl (-)RWAY metabolites, but at lower concentrations than that in the rat. Similar to the rat study findings, there were no metabolic fragments arising from the hydrolysis of the amide bond of (-)-RWAY thus confirming the in vivo stability of the amide bond in [11C](-)-RWAY in a nonhuman primate.
B.PET Imaging in Nonhuman Primates.
Animals: The PET study of uptake of [11C](-)-RWAY comprised four rhesus monkeys (weight: 13.2±1.8 kg), three of which underwent arterial blood sampling. One monkey without arterial blood sampling was examined with the active and inactive enantiomers. One of monkeys with arterial blood sampling participated also in the pre-blocking study, in which cold compound of WAY-100635 (0.5 mg/kg i.v.) was injected prior to [11C](-)-RWAY.
Positron emission tomography studies:All of PET scans were performed with a GE Advance scanner (General Electric Medical Systems, Waukesha, WI), with reconstructed resolution of 6 mm full-width half-maximum in all directions in 3D mode. Coronal slices covering the whole brain were obtained. PET scans were acquired for 90 min (33 frames with scan duration ranging from 30 s to 5 min). After transmission scan of 6 min with a 68Ge-68Ga source, a bolus of 201-215 MBq of [11C](-)-RWAY, with specific activity of 23-65 GBq/μmol at time of injection. In the study of the uptake of [11C](+)-RWAY, a bolus of 195.7MBq of ligand with specific radioactivity of 35 GBq/μmol was injected.
The PET images were coregistered to magnetic resonance images using SPM2 (Wellcome Department of Cognitive Neurology, London, U.K.), and regions of interest were defined over the cerebellum, raphe nucleus, thalamus, prefrontal cortex and temporal cortex.
After iv administration of [11C](-)-RWAY, arterial blood was collected at 30 and 45 s, and 1, 3, 5, 20, 40, 60, and 90 min after tracer injection. Each blood sample was separated into plasma and blood cell fraction by centrifugation. The plasma time-activity curve was corrected by the amount of unchanged ligand in plasma.
Estimation of distribution volume with arterial input function:The brain and plasma data were analyzed with one-tissue (1T) and two tissue (2T) compartment models (1-TCM and 2-TCM) using rate constants (K1, k2, k2', k3 and k4) that were described previously (Laruelle M, 1994). In the one-compartment model, the total distribution volume Vt is calculated as: