1. Introduction to Application We thank the reviewers for their enthusiasm for the initial proposal and for their detailed reviews. The average scores for Significance, Investigators and Innovation were each 1.67 and that for the research environment (primarily the Yale PET Center and the facilities for performing the Alcohol Drinking Paradigm (ADP) ) scored a perfect 1.0. The primary criticisms of the Approach fell into two main types: (1) lack of in vivo data on selectivity of new kappa (K) opioid receptor (KOR) tracer, [11C]LY-2975050 (aka [11C]PKAB), and (2) lack of preliminary K imaging data in heavy drinkers (HD). These main concerns – and minor criticisms as well – have been addressed thanks to new preliminary data in the present resubmission. Finally, we have examined PET data in healthy control subjects (HC) at baseline and we propose to expand our original proposal to include a complete cohort of HC to address questions regarding the effects of prolonged drinking on baseline KOR expression. We have also included exploratory opioid genetic analyses. Summary of new preliminary data: 1. PET imaging has been performed in rhesus monkeys to investigate the selectivity of our novel tracer, [11C]PKAB, in vivo for KOR over MOR. In two sets of experiments, the animals were pre-dosed with varying concentrations of unlabeled PKAB. Subsequently, PET imaging was performed with either [11C]PKAB or the mu-specific tracer, [11C]carfentanil (CFN). At comparable doses of unlabeled PKAB, there was much greater blocking of [11C]PKAB than of [11C]CFN. These results – taken together with earlier in vitro work – indicate that the affinity of [11C]PKAB is more than 9 times greater for KOR than MOR, in vivo. (See Fig. A; note: new figs were given letters; Figs 2,4,5,8 have been superseded by new results in heavy drinkers using an arterial input function. Figs 3 ,6 and 9 were updated. All original and updated figures retain original numbers. 2. Preliminary PET images with our K-selective tracer have been acquired in 3 male heavy drinkers (two positive for family history of alcoholism (FHP), one negative (FHN)). The PET, like the ADP, was performed on these subjects before and 1 week after treatment with 100 mg/day of NTX. These exciting new data suggest a dramatic difference between FHP and FHN drinkers – both in (percent) reduction of drinking after NTX and in occupancy of KOR by NTX. As we hypothesized in our first submission, the occupancy of KOR by NTX appears to be much higher in FHP (96%) than in FHN (20%). In addition, NTX treatment reduced drinking in the two FHP subjects (average of 76%) more than in the FHN subject (only 50%). (Figs. D & E.) In the course of our preliminary work, we addressed two supplementary aims of the original proposal. First, we have determined that there is no reliable reference region in the brain for the K tracer (Figure B). (As we explain in the text, this does not prevent us from accomplishing our goals, but it does establish the ongoing need for arterial cannulation in all subjects.) Discussion of choice of reference regions and models has been removed to make way for newer results. Second, we have established that a 90 minute-long scan is adequate for reliable estimates of tracer uptake (Figure C). Shorter scans should be more easily tolerated by subjects. 3. A comparison of the baseline KOR uptake suggests that FHP heavy drinkers have lower KOR availability than heavy drinkers who are FHN, supporting Specific Aim
Transcript
1. Introduction to Application We thank the reviewers for their enthusiasm for the initial proposal and for their detailed reviews. The average scores for Significance, Investigators and Innovation were each 1.67 and that for the research environment (primarily the Yale PET Center and the facilities for performing the Alcohol Drinking Paradigm (ADP) ) scored a perfect 1.0. The primary criticisms of the Approach fell into two main types: (1) lack of in vivo data on selectivity of new kappa (K) opioid receptor (KOR) tracer, [11C]LY-2975050 (aka [11C]PKAB), and (2) lack of preliminary K imaging data in heavy drinkers (HD). These main concerns – and minor criticisms as well – have been addressed thanks to new preliminary data in the present resubmission. Finally, we have examined PET data in healthy control subjects (HC) at baseline and we propose to expand our original proposal to include a complete cohort of HC to address questions regarding the effects of prolonged drinking on baseline KOR expression. We have also included exploratory opioid genetic analyses. Summary of new preliminary data:1. PET imaging has been performed in rhesus monkeys to investigate the selectivity of our novel tracer, [11C]PKAB, in vivo for KOR over MOR. In two sets of experiments, the animals were pre-dosed with varying concentrations of unlabeled PKAB. Subsequently, PET imaging was performed with either [11C]PKAB or the mu-specific tracer, [11C]carfentanil (CFN). At comparable doses of unlabeled PKAB, there was much greater blocking of [11C]PKAB than of [11C]CFN. These results – taken together with earlier in vitro work – indicate that the affinity of [11C]PKAB is more than 9 times greater for KOR than MOR, in vivo. (See Fig. A; note: new figs were given letters; Figs 2,4,5,8 have been superseded by new results in heavy drinkers using an arterial input function. Figs 3 ,6 and 9 were updated. All original and updated figures retain original numbers. 2. Preliminary PET images with our K-selective tracer have been acquired in 3 male heavy drinkers (two positive for family history of alcoholism (FHP), one negative (FHN)). The PET, like the ADP, was performed on these subjects before and 1 week after treatment with 100 mg/day of NTX. These exciting new data suggest a dramatic difference between FHP and FHN drinkers – both in (percent) reduction of drinking after NTX and in occupancy of KOR by NTX. As we hypothesized in our first submission, the occupancy of KOR by NTX appears to be much higher in FHP (96%) than in FHN (20%). In addition, NTX treatment reduced drinking in the two FHP subjects (average of 76%) more than in the FHN subject (only 50%). (Figs. D & E.)
In the course of our preliminary work, we addressed two supplementary aims of the original proposal. First, we have determined that there is no reliable reference region in the brain for the K tracer (Figure B). (As we explain in the text, this does not prevent us from accomplishing our goals, but it does establish the ongoing need for arterial cannulation in all subjects.) Discussion of choice of reference regions and models has been removed to make way for newer results. Second, we have established that a 90 minute-long scan is adequate for reliable estimates of tracer uptake (Figure C). Shorter scans should be more easily tolerated by subjects. 3. A comparison of the baseline KOR uptake suggests that FHP heavy drinkers have lower KOR availability than heavy drinkers who are FHN, supporting Specific Aim P3 in the current proposal. Furthermore, we analyzed baseline [11C]PKAB scans obtained from 12 healthy controls participating in another project. In most brain regions examined, HC subjects had baseline KOR volumes of distribution (VT = specific plus nonspecific uptake) that was intermediate between FHN and FHP. (See figure F.) Given properly matched cohorts of HC with/without family history of alcoholism, observation of differences between HC and HD could help to inform the effects of long-term drinking on KOR. Thus, we propose to add two additional cohorts of HC (properly matched for FH and other demographics to each cohort of HDs) to our project to begin to investigate the effects of heavy drinking on the K system. Additional criticisms and responses:Reviewer 1: Backup plan to examine lower doses of NTX than 100 mg is inconsistent with previous findings by investigators showing differential responses (FHP vs FHN) to drug at that dose. Response: Preliminary findings suggest large group differences in occupancy and behavior at 100 mg. No other doses will be used. Reviewer 2: No genetic data to be collected. Response: We will collect blood samples from all subjects to compare the frequency of gene variants (such as the OPRK1 and OPRM1 alleles) in FHP vs FHN populations and explore relationships with NTX responsivity. Reviewer 3: Concern about MOR effects of repeated dosing of “kappa antagonists”. Response: The KOR tracer being used in this study is delivered in such small mass amounts that it has no pharmacological effects. We would like to point out that NTX achieves complete mu blockade at a dose of 50 mg/day; the purpose of this project is to examine if variability in NTX responsivity in FHP and FHN drinkers is related to differences in KOR levels. In response to another suggestion, we have added an expert kappa opioid pharmacologist, Prof. Bryan Roth of UNC, to our team.
2. Specific Aims Naltrexone (NTX), a non-specific opioid antagonist that binds dose-dependently to mu, delta, and kappa opioid receptors, has been shown to be efficacious in the treatment of alcohol dependence [1]. Problem: Understanding how NTX reduces drinking is needed to optimize NTX treatment and identify targets for the development of new pharmacotherapeutic agents to treat alcohol dependence. Recent evidence suggests that kappa opioid receptors (KOR’s) play an important but complex role in mediating alcohol drinking behavior. This proposal convenes a translational team of scientists to examine the in vivo biochemistry of NTX in heavy drinkers (HD), via high resolution PET imaging of the KOR and relate this to clinical outcomes of NTX efficacy. Important finding by our behavioral science team: Using a laboratory-based human Alcohol Drinking Paradigm (ADP; [2]) that we pioneered, we have shown that family history (FH) of alcoholism may moderate (encode the direction of) response to NTX [3]. Specifically, pretreatment with a 100 mg/day dose of NTX was associated with lower drinking in the ADP in heavy drinkers with a positive FH (FHP), and higher drinking in those with a negative FH (FHN). What could be the underlying biochemistry? Existing evidence suggests that, at a NTX dose of 50 mg/day, mu-opioid receptors are saturated and therefore unlikely to determine variability in NTX responsivity [4]. At 100 mg/day, NTX should antagonize KOR, raising the possibility that the differences observed in the two cohorts of drinkers may be related to differences in KOR. KOR levels can be examined using PET imaging. However, until recently, no tracer was available to facilitate such an examination.Recent development of KOR PET tracer by our PET Center. [11C]PKAB is a selective, high affinity, KOR antagonist tracer with favorable kinetics for imaging KOR and drug occupancy in humans [5]. As we demonstrate in Preliminary PET Studies, the new tracer has excellent test/retest reproducibility and is well-suited to quantitating KOR levels in striatal and extra-striatal regions. Preliminary results indicate a >9 times higher affinity of [11C]PKAB for KOR over MOR in non-human primates, in vivo. Initial collaborative work conducted by our group with the KOR tracer suggests that heavy drinkers who are FHP have much higher levels of KOR occupancy by NTX after a week of treatment than HD with FHN. In turn, high occupancy of KOR by NTX correlates with a greater percent reduction in drinking. Preliminary results also suggest that healthy controls (HC) have baseline uptake of the KOR tracer that is intermediate between FHP and FHN.Collaboration. This proposal unites experts in developing new tracers and in conducting human PET studies with experts in testing treatments for alcoholism. Together, w e will conduct the first in vivo PET study of KOR and NTX occupancy in two cohorts of HD (18 FHP and 18 FHN) and relate our imaging findings to FH and alterations of drinking behavior by NTX in the ADP. This work will advance our understanding of (a) the degree of involvement of the kappa system in reducing drinking in humans, (b) the contribution of FH to this relationship, (c) refine the use of a new PET tracer, and (d) investigate the effects of heavy drinking on KOR in HD and FH-matched HC. Our measurements of NTX occupancy will inform the development of future targeted pharmacotherapies for alcoholism and our studies of baseline KOR will probe effects of long-term drinking on the KOR/dynorphin system in vivo . Primary (P), Supplementary (S),and Exploratory (E) Specific Aims follow:P1. Occupancy of KOR by NTX and drinking: To determine the degree to which occupancy of KORs by a 100 mg/day dose of NTX mediates (influences the strength of) responsivity to NTX treatment in all heavy drinkers. It is expected that KOR occupancy will significantly mediate the relationship between NTX treatment and responsivity. ‘Responsivity’ is the percent change in # drinks consumed during ADP from baseline to NTX.P2. Family History of Alcoholism as a moderator: To determine whether the relationship between NTX
responsivity and occupancy of KOR is different in FHP vs. FHN heavy drinkers. P3. Baseline KOR differences. (a) To determine if baseline levels of KOR differ between FHP and FHN
heavy drinkers. (b) To determine if baseline KOR level is related to either baseline drinking or responsivity to NTX. (c) To determine the effect of long-term drinking on KOR by comparing baseline levels of KOR in family-history-matched HC vs HD.
S1. Behavioral measures: To determine if drug occupancy of KOR affects alcohol craving or stimulation.E1. Genetics. To compare the frequency of the OPRK1 and OPRM1 alleles in FHP and FHN populations and
possible correlations between the alleles and baseline KOR levels and/or NTX occupancy.We hypothesize that (1) degree of NTX occupancy of KOR during treatment will be correlated with the degree of responsivity to NTX in all heavy drinkers, (2) FHP HD, who respond positively to NTX, will have lower
baseline KOR levels when compared with FHN HD, (3) FHP HDs have lower baseline KOR availability than family-history-matched HCs.
3. Research Strategy A. Significance Alcohol abuse is one of the leading causes of disability in the United States. At present, three medications
are approved by the FDA for the treatment of alcohol dependence: disulfiram, naltrexone, and acamprosate. However, the efficacy of these agents ranges from modest at best, and they are under-utilized in clinical community settings [6, 7]. There is great need for medications that are more effective in reducing alcohol use.
Development of new pharmacotherapeutic approaches to treating alcohol dependence should be based on our understanding of the behavioral and neurochemical mechanisms mediating alcohol drinking, as well the efficacy of agents known to reduce alcohol drinking. Naltrexone (NTX) is a non-specific opioid antagonist that has been shown to have minimal-modest efficacy in treating alcohol dependence. Clinical trials indicate that treatment-seeking heavy drinkers who receive NTX at a dose of 50 mg/day [1, 8] or 100 mg/day [9], have lower levels of relapse to drinking during the treatment period than do those receiving placebo. Subpopulations of alcohol dependent patients may respond better to NTX [10] and family history (FH) of alcoholism is emerging as an important predictor of NTX response [3, 11].
Our earlier work, using a laboratory-based alcohol drinking paradigm (ADP) revealed that NTX responsivity was altered by the presence or absence of a FH of alcoholism; specifically NTX treatment was associated with lower drinking in those with a positive FH (FHP) and actually with higher drinking in those with a negative FH (FHN) [3]. Developing a better pharmacological understanding of this differential response could help enhance the efficacy of NTX treatment. Naltrexone’s ability to reduce alcohol drinking is believed to be mediated, in part, through its effects on the endogenous opioid system, specifically through dose dependent antagonism of mu, delta and kappa opioid receptors. Weerts and colleagues [4] have shown that the most commonly used dose of naltrexone, 50 mg/day, results in almost complete occupancy (saturation) of mu-opioid sites in the brain with variable, partial occupancy of delta opioid receptor sites. This evidence suggests that the differential efficacy of NTX is unlikely to be mediated by mu-opioid binding but could be related to variability in binding to other (delta or kappa) opioid receptor sites. While some investigations have focused on the delta system [4], to date, no one has examined the relationship of kappa opioid receptors (KOR) to alcohol drinking or NTX responsivity in human alcoholics. Dynorphin is the endogenous opioid ligand that binds to both mu and KORs [12]. In general, activation of KORs is aversive. Animals will not self-administer dynorphin, administration of dynorphin results in conditioned place aversion in animals, and dynorphin activation of KOR results in decreased dopamine release in brain reward areas [13, 14]. The potential role of KORs in mediating alcohol drinking behavior has been the focus of several studies in animals and humans. Mice that lack the KOR drink more alcohol and have greater release of dopamine in response to alcohol [15]. Rodents with an increased propensity to consume alcohol have lower dynorphin levels in several brain regions that are involved in the control of alcohol drinking [16-18]. In humans, the presence of the OPRK1 allele (which decreases expression of KOR) is associated with increased risk of alcoholism [19]. The above evidence suggests that differences in the status of the KOR system may mediate differences in propensity to drink alcohol. Preliminary evidence (shown below) is consistent with KOR differences between FHP and FHN drinkers. The current proposal will be the first to compare KOR occupancy in FHP and FHN drinkers and correlate KOR levels with drinking in the ADP.
The effect of stimulating or inhibiting the KOR opioid system on alcohol drinking appears to depend, in part, on the prior history of alcohol exposure. In alcohol naïve animals, stimulation of KOR by agonists decreases alcohol drinking [20], while inhibition of KOR by antagonists increases alcohol drinking [21]. Chronic alcohol exposure dysregulates the dynorphin/KOR system; specifically, dynorphin levels are enhanced and number of KOR are reduced [22, 23]. The aversive effect of dynorphin/KOR activation is decreased in individuals whose dynorphin/KOR system has been chronically stimulated, such as those with chronic pain or those with a long history of polysubstance abuse, which may reflect a reduced number or binding of KOR’s [24, 25]. In contrast to the above evidence, in alcohol dependent animals, KOR antagonists reduce alcohol drinking [26], suggesting that chronic alcohol drinking also alters the KOR system. The above evidence suggests that differences in the KOR system may mediate differential responsivity to NTX.
The proposed project would be the first to assess the potential contribution of the KOR system to individual differences in naltrexone-mediated changes in alcohol drinking using an established ADP in alcohol dependent, heavy drinkers (HD). Additionally, the current proposal will be the first to examine KOR levels and NTX responsivity in FHP and FHN heavy drinkers; this will help us understand if the differential responsivity to
NTX observed in FHP and FHN drinkers is related to KOR occupancy by NTX . This revised proposal also includes two cohorts of healthy, social drinking, controls (FHP and FHN) to allow for investigation of the effects of long-term drinking on KOR, and testing of OPRM1 and OPRK1 genotypes to explore genetic associations with KOR levels and naltrexone response.
Knowledge GapOur project will investigate a provocative finding in alcohol treatment research [3]. Namely, that FH of
alcoholism may influence (or be a marker of an underlying trait that influences) the response of heavy drinkers (HD) to NTX. Simply put, not every HD responds favorably to NTX but lower levels of drinking can be achieved and risk of relapse is diminished in some. It is essential for clinical treatment and future drug development to understand how NTX works. Preclinical results have implicated KOR in changes in alcohol drinking, but the directionality of these changes are complex. This project will bring together a stellar group of investigators to examine the role of KOR in mediating drinking behavior and treatment efficacy of naltrexone, as well as the effects of long-term heavy drinking on KOR expression. This translational project could not only help maximize treatment effectiveness of naltrexone and result in identification of new medication treatment targets for more optimal reduction of alcohol drinking, but also answer the fundamental question of what effects long term alcohol drinking might exert on KOR expression, and its relation with FH status.
B. InnovationThe project that we propose is inherently innovative in both its conception and application because it
identifies a timely and important gap in clinical knowledge –how does naltrexone work to reduce drinking - and then brings to bear two disparate but perfectly tailored experimental tools (one biochemical, one behavioral) to investigate it. The combination of a novel imaging tool and a behavioral science tool to examine the very relevant clinical question of NTX’s mode of action is at the heart of our innovation.
A novel aspect of our project is the use of the PET tracer for KOR’s. PET imaging is the most direct, sensitive and straightforward means of probing functional neurochemistry of human subjects, in vivo. It has unparalleled molecular specificity – behavior of a particular receptor subtype can be probed – provided the proper tracer is available. However, this approach has not previously been applied to the KOR system due to the lack of a specific KOR ligand. Thanks to one of our collaborators (HH) and his Chemistry team at the Yale PET center, a tracer to probe the KOR subtype’s role in NTX action is now available. Our tracer, [11C]PKAB has already been tested in humans [5] and nonhuman primates and shown to have key qualities associated with a good [11C]-labeled PET tracer. Namely, it has good test/retest reproducibility and fast kinetics, can be blocked by a drug specific for intended target.
In addition to a novel PET tracer, our center uses the world’s highest sensitivity and resolution human brain scanner available, the Siemens High Resolution Research Tomograph (HRRT). High resolution, enhanced by our colleagues’ considerable theoretical work on iterative reconstruction [27] allows us to precisely image small brain regions like the amygdala which appear to have high KOR specific binding (see Preliminary Studies).
The behavioral component of our project is the ADP. The ADP - pioneered by our group (Dr’s O’Malley and Krishnan-Sarin) - has been used as a time-efficient, controlled laboratory-based method for testing medication effects on drinking, and has been shown to be sensitive to NTX-induced changes in drinking [2, 3]. It has also been adopted by other leading researchers in academia and industry. Thanks to the expertise of our team (SK-S, SSO) we will be able to recruit adequate numbers of HDs, determine their FH of alcoholism, administer NTX treatment in a reliable fashion, and precisely measure the effects of NTX on drinking behavior using the ADP. Combining the ADP with an imaging study is a logical and innovative extension of its use.
Finally, this proposal also includes exploratory examinations of OPRM1 and OPRK1 genes, which encode mu and kappa opioid receptors, respectively. A number of studies have reported positive associations between alcohol dependence and an OPRM1 functional variant A118G ([28]; [29]; [30]; [31]) as well as with OPRK1 haplotypes ([19]; [32]). Similarly polymorphisms of the OPRM1 gene have also been shown to be associated with response to naltrexone treatment ([33]; [34]; [35]). Our proposed examinations of the relationship between OPRM1 and OPRK1 genotypes and KOR levels as well as response to naltrexone will significantly add to our understanding of how naltrexone works to reduce drinking behavior.
Figure 1. Number of drinks consumed during 2-hr choice period. Groups of subjects received either 0, 50 or 100mg NTX. FHP subjects (solid circles) receiving 100mg dose drank less than FHP who received 0 mg. FHN (triangles) at 100 mg drank more during ADP than FHN at 0mg.
Thus, the present proposal combines the unique tools we have available to us as a group: (1) a laboratory-based means of assessing changes in drinking brought on by NTX treatment in populations known to respond differently to the drug with (2) high resolution PET imaging of human subjects’ KOR subtype before and after drug treatment, 3) comparisons with social drinking controls and 4) exploratory examinations of opioid genetic associations. We believe this synthesis of techniques is uniquely armed to shed new light on the actions of NTX and the dynorphin/KOR system in alcohol drinking behavior.
C. Approach1. Preliminary Studies
i. Alcohol Drinking Results – Effect of NTX on Drinking in FH groups
Members of this research group have extensive experience with conducting treatment and laboratory-based research in alcohol-dependent subjects, and with examining the influence of opioid antagonists. We have also developed and validated an excellent alcohol drinking paradigm (ADP) [2] that has been used to evaluate the effects of many medications on drinking behaviors by our [3, 36] and other [37-39] research groups (including a pharmaceutical company trial of an investigational agent where we were the lead site).
The ADP consists of exposure to a priming drink (PD) of alcohol, followed by a 40 minute observation period, and then free-choice self-administration (SA) periods during which participants are asked to choose between alcoholic drinks and money. In our original work with naltrexone (described below) we used two, one hour SA periods (total of 8 choices) but in our ongoing work with memantine (P50AA12870) we have extended this to three, one hour SA periods (total of 12 choices) in order to increase variability in drinking. Initial evidence from our memantine work [40] indicates that this manipulation was successful in increasing the number of drinks consumed by participants. Thus, we will use the three-hour format in the current proposal.
In our completed project (P50AA12870; Program Project PI: Krystal; Project PI: Krishnan-Sarin; [3]) we used the ADP to evaluate the influence of six days of pretreatment with different doses of naltrexone (0, 50, 100 mg/day), on drinking behavior in alcohol-dependent heavy drinkers who were FHP or FHN for alcoholism. Participants (n=92) were mostly male (70 males, 22 females) with a racial distribution consistent with that of New Haven County (70 Caucasians, 14 African Americans, 4 Hispanics, 1 Native American, 2 Asians, and 1 other). 58.7% were FHN and 41.3% were FHP. There were no FH X MED interactions on any of the baseline drinking variables, and within each FH group these variables were comparable for the three medication conditions. The drinking results were presented in Krishnan-Sarin et al. [3]. The FH groups and the medication (MED) conditions (0, 50, 100 mg naltrexone) within each FH group were compared on demographic and drinking characteristics using ANOVA or logistic regressions (for binary outcomes). Total numbers of drinks consumed during the SA period were compared using an ANOVA with FH and MED groups as between-subjects factors.
During the two-hour SA period, there was a significant FH by MED interaction [F(2,86) = 3.2, p<0.05]. Post-hoc analyses indicate that drinking was lower in FHP participants receiving 100 mg of naltrexone than in those receiving a placebo (p<0.05; Figure 1), while drinking was not significantly different for FHN participants at either dose of naltrexone. Secondary analyses in male drinkers found a significant FH by MED interaction on the total number of drinks consumed [F(2,64) = 4.58, p<0.01], with post-hoc analyses indicating lower drinking in FHP males following 100 mg of naltrexone (p<0.05) but higher drinking in FHN males following 50 (p=0.09) and 100 (p<0.05) mg naltrexone (data not shown) when compared with 0 mg dose. In contrast, in female participants, there were no significant main or interactive effects of FH or MED status on drinks consumed (data not shown). But due to the small sample of women in this study this latter finding was hard to interpret. Moreover, other adequately powered clinical trials have observed efficacy of naltrexone in women drinkers (e.g. COMBINE [9]). Therefore, in the current proposal we will recruit male and female heavy drinkers.
The current proposal will use the ADP to measure drinking at baseline and following seven days of NTX treatment. “Responsivity to NTX” will be defined as “percentage change in number of drinks taken
Figure 6. Percent change in BPP from test to retest scans using 2T model to estimate k3/k4. Each bar is based on the average of 7 HC subjects. Regions: amygdala, caudate, posterior cingulate, frontal cortex, hypothalamus, insula, occipital, pallidum, putamen, temporal, thalamus.
Figure A. Occupancy plots for cold PKAB at the kappa (top) and mu (bottom) opioid receptor sites as determined by blocking of [11C]PKAB and [11C]CFN, respectively. Top curve is based on multiple doses in two monkeys (solid symbols). Bottom curve is based same two monkeys (open symbols). The cold dose of 135ug/kg that would occupy 90% of KOR (based on top curve) occupied an average of 50% of MOR.
Figure 3. Selected regional time activity curves for a single FHP HD subject at baseline (dashed) and after a week of NTX treatment (solid). Baseline curves represent the regions of high uptake: anterior cingulate, insula, putamen, as well as cerebellum. Note: all regions including cerebellum show reduced uptake of tracer in presence of NTX, suggesting absence of a valid reference region.
during the ADP from baseline to NTX condition”.
ii. PET Imaging Results –Imaging of KOR Sites
a. IntroductionThe preliminary data in this section represent our initial
experience with the new PET tracer [11C]LY2795050 (aka [11C]PKAB) in humans (both healthy controls and heavy drinkers) and monkeys. A number of opioid receptor radiotracers are currently available for PET imaging in humans: [11C]carfentanil (aka [11C]CFN; mu-specific), [11C]/[18F]diprenorphine (nonspecific), [11C]buprenorphine (nonspecific), [18F]cyclofoxy (mu, kappa), and [11C]methylnaltrindole (delta) [41]. None of these existing tracers is selective for KOR. So far, the results with [11C]PKAB suggest that it is taken up consistent with the known distribution of KOR [42-44]; It also appears to have good contrast to background properties and good test/retest reproducibility. Our data suggest that there is no reference region (i.e., brain region devoid of KOR).
b. Definition of experimental endpoints
How will we measure KOR level at baseline? At best, PET can measure Bavail, the available receptors (i.e., not bound to an endogenous ligand). Often, PET data are used to estimate binding potential (BP), a measure of the steady state ratio of bound to free tracer in the target tissue. There are a few variants of BP [45], the most common of which is BPND, defined as BPND = fNDBavail/KD. fND is the free fraction of tracer in the non-displaceable tissue compartment. KD is the equilibrium dissociation constant of the tracer at the KOR site.
Total Volume of Distribution - VT, is a related concept to BP. VT is the equilibrium ratio of the concentration of parent tracer in tissue to parent tracer in plasma. It is a combined measure of the specific and the non-displaceable uptake of tracer by the tissue. Blocking of specific binding by a drug is reflected in a diminution of VT (Innis et al,. 2007).
Occupancy of KOR by NTX – or simply “occupancy” – can be measured as a percentage change in BPND (BP) from the baseline condition to the drug condition (in our case, a week of NTX treatment). If we assume that neither the KD nor the fND of the tracer change in an individual between scans, then occupancy can be seen to be the fractional change in available receptors due to the drug.
0
10
20
30
40
50
60
70
80
90
100
0 20 40 60 80 100 120 140 160
Occ
upan
cy (%
)
cold pkab dose (ug/kg)
Figure B. Displacement can be seen as reduction in VT from baseline scan (blue) to post-NTX scan (red hashed) in all regions, including cerebellum. Average VT (n=2) in FHP subjects is given here for the same regions shown in (updated) Figure 3, above. VT is reduced by NTX in all regions. Thus, there does not appear to be a valid reference region for [11C]PKAB.
Figure C. Comparison of VT for different brain regions based on 90 (x axis) or 120 (y axis) min of data. 90 min of data yields identical results for VT estimated from 2T model with metabolite-corrected arterial plasma input. This allows us to shorten all future scans to 90 min. Dashed line is identity.
Occupancy of KOR by NTX without a Reference Region It is possible to estimate occupancy of a drug even in the absence of a reference region for the tracer. One way is estimate k3/k4 (=BPND) from a 2T model fit to the time activity curves and then to use BP, as explained above. However, this method is prone to produce variable estimates of k3/k4
so an alternative method is used that depends on VT and change in VT from baseline to drug condition. This latter method, based on the Lassen Plot ([46]) has become the de facto standard approach to measuring occupancy in drug studies when there is no reference region. It relies on the additional assumption that the occupancy is uniform across brain regions. These methods require arterial sampling to determine a plasma input function.
c. Selectivity of [11C]PKAB,i. In vitro results
Radioligand competition assays were conducted to assess the in vitro binding affinities of PKAB to opioid receptors. PKAB is a KOR antagonist and displays high binding affinity and great selectivity to KOR in vitro (Ki = 1.37, 87. 3, and 475 nM, respectively, for kappa, mu, and delta receptors) as determined in radioligand competition assays using cloned human KOR, MOR and DOR [47].
ii. In vivo results in monkeysWe performed two types of in vivo studies on selectivity of the tracer in vivo. First, to establish the EC50 of the tracer at the KOR, we measured the increased blocking of [11C]PKAB with increasing doses of unlabeled PKAB (See top curve in Figure A). Based on 3 doses each in two monkeys, this curve yields an estimate of IC50
KOR = 12.84 g/kg. From the top curve in Figure A, we can project to a dose of 135 g/kg to achieve >90% occupancy of KOR. In our second study, we scanned two monkeys with [11C]carfentanil ([11C]CFN), a selective MOR tracer [48] The occupancy at MOR for at 135 ug/kg was shown to be approximately 50%. Using an Emax model with maximum occupancy set to 100% to fit the [11C]CFN data (lower curve in Fig. A), we find the EC50MOR of our KOR-ligand is ~119.2 g/kg. That is, the affinity of [11C]PKAB is 9.3 times greater for KOR than MOR. One useful way to understand this ratio is to recognize that the ratio of the specific binding at KOR to MOR is BKOR/BMOR = (Bmax
KOR/BmaxMOR )*( KD
MOR/KDKOR)
where Bmaxi and 1/KD
i are the receptor density and affinity of the tracer, respectively, for site i. If the density of KOR and MOR are approximately equal in a given brain region, then the estimated specific binding (ie., BP) of [11C]PKAB will be 9.3/(9.3 + 1) or 90% attributable to binding at the KOR.
d. Time-activity curves in humans
Estimates of kinetic parameters require regional tracer concentration values over time. Automatic regions-of-interest (ROI) were applied (see PET methods) to the dynamic emission images to generate time-activity curves (TACs) in 13 ROI’s. Examples of TACs in three select regions before and after NTX are shown in Figure 3. The data from an FHP HD in Figure 3 are normalized by injected dose per bodyweight (called SUV) to allow display of both scans on one axis. The lowering of each respective curve after NTX is an indicator of reduced specific binding. The fast kinetics of the tracer are advantageous for human imaging and presage robust estimation of VT with scan durations shorter than 120 min. (see section h).
e. Percent change between ‘test’ and ‘retest’ scans: How small an effect can we detect?
Small test/retest variability is essential for us to be able to detect small differences between conditions. For each subject, the test-retest variability was measured by computing the ratio of the difference between binding potentials to their means: 100*(BP(TEST) - BP(RETEST)) / [ ½ ( BP(TEST) + BP(RETEST) ) ].
Figure D. Occupancy of NTX at KOR in FHP and FHN heavy drinkers as illustrated by VT images at baseline and after a week of treatment. Degree of reduction in VT after a week of NTX treatment indicates the degree of NTX occupancy at the receptor. Dramatic decrease in VT in an FHP drinker (compare top two rows) indicates high (>90%) NTX occupancy, whereas little change in BP (compare bottom two rows) indicates low NTX occupancy in FHN subject. VT images were created using the “RE-GP” graphical analysis method of Zhou et al, (2010).
Figure E. Two FHP drinkers (solid symbols) both had high (>95%) occupancy of NTX and they reduced their drinking by 75%. One FHN drinker (open symbol) had only 21% occupancy of NTX after the treatment week and reduced his drinking by only 50%.
The mean test-retest variability was 8.2% (2T) indicating that BPND values increased during the retest scan (conducted in the afternoon of the same day) compared to the test scan, across all regions. This level or reliability is comparable to other popular PET tracers. Figure 6 shows the test/retest comparison as an average value over subjects for all regions.
f. Test/retest imaging in humans
Seven healthy volunteers, age = 26 ± 6 (5 males and 2 females), participated in a test-retest study of [11C]PKAB, an antagonist PET radiotracer for the KOR. Each subject received two bolus injections of up to 20 mCi of radioactivity and mass dose of < 10 µg for each injection. The injected activity dose was 13.0 ± 4.6 mCi, specific activity was 0.59 ± 0.22 mCi/nmol at time of injection, and injection mass was 8.8 ± 1.2 µg across all subjects. Arterial (plasma) input function (AIF) curves were prepared from arterial samples taken during the study. Metabolite fraction was applied to yield an input curve for the native tracer only. BP throughout the brain (calculated by estimating k3/k4 from the 2T model and Logan plot (data not shown) with arterial input (Logan, 1991)) was used as the study endpoint.
f. Pre-blocking of [11C]PKAB with a known kappa-antagonist
We can verify that uptake and retention of [11C]PKAB is due to specific binding to KOR by looking at pre-blocking studies with well-characterized (unlabeled) drugs. Pre-blocking studies were performed in 3 rhesus monkeys. The experiments were similar to what is being proposed herein to measure occupancy of KOR by NTX. In the preliminary studies, rhesus monkeys were given iv, 1 mg/kg naloxone, a non-selective KOR antagonist, or varying doses of LY2456302, a selective kappa antagonist developed by an industry partner (Ki = 0.60, 16.4 and 122 nM for KOR, MOR and DOR, unpublished data). Dose-dependent reduction in binding of [11C]PKAB was observed across brain regions in response to increasing doses of LY2456302 (Fig. 5 removed for new data.)
g. Displacement indicates lack of reference region Reduction in time activity curves following NTX (Figure 3) were the first evidence that we were imaging the occupancy of NTX at KOR. The reduction of VT in all brain regions is a quantitative indication that specific binding (ie., receptor availability) is decreased by NTX. Figure B shows the reduction in VT (in the same
regions as Fig 3) due to NTX. Note: VT is reduced in cerebellum along with other regions. We found no region without displacement.
h. 90 min vs 120 min: Reducing scan time
A 90 min PET scan is slightly less burdensome for the subject than a 2 hr scan and given the fairly rapid kinetics of the tracer. In work we performed since our original proposal, we have determined that estimates of VT using 90 min of data are indistinguishable from estimates using 120 min. of data. Thus, all scans will be 90 min. (Fig C).
i. Occupancy of NTX in FHP and FHN drinkers
Figure D shows VT images made from [11C]PKAB scans on FHP and FHN heavy drinkers. The VT images were produced with a bi-graphical method
0.0
1.0
2.0
3.0
4.0
5.0
[undefined]
0.0
1.0
2.0
3.0
4.0
5.0
[undefined]
FHN
0.0
1.0
2.0
3.0
4.0
5.0
[undefined]
0.0
1.0
2.0
3.0
4.0
5.0
[undefined]
0.0
1.0
2.0
3.0
4.0
5.0
[undefined]
0.0
1.0
2.0
3.0
4.0
5.0
[undefined]
baseline
0.0
1.0
2.0
3.0
4.0
5.0
[undefined]
0.0
1.0
2.0
3.0
4.0
5.0
[undefined]
0.0
1.0
2.0
3.0
4.0
5.0
[undefined]
0.0
1.0
2.0
3.0
4.0
5.0
[undefined]
0.0
1.0
2.0
3.0
4.0
5.0
[undefined]
FHP
NTX 100mg
0.0
1.0
2.0
3.0
4.0
5.0
[undefined]
baseline
NTX 100mg
0.0
1.0
2.0
3.0
4.0
5.0[undefined]
0.0
1.0
2.0
3.0
4.0
5.0
[undefined]
0.0
1.0
2.0
3.0
4.0
5.0
[undefined]
0.0
1.0
2.0
3.0
4.0
5.0
[undefined]
FHN
0.0
1.0
2.0
3.0
4.0
5.0
[undefined]
0.0
1.0
2.0
3.0
4.0
5.0
[undefined]
0.0
1.0
2.0
3.0
4.0
5.0
[undefined]
0.0
1.0
2.0
3.0
4.0
5.0
[undefined]
baseline
0.0
1.0
2.0
3.0
4.0
5.0
[undefined]
0.0
1.0
2.0
3.0
4.0
5.0
[undefined]
0.0
1.0
2.0
3.0
4.0
5.0
[undefined]
0.0
1.0
2.0
3.0
4.0
5.0
[undefined]
0.0
1.0
2.0
3.0
4.0
5.0
[undefined]
FHP
NTX 100mg
0.0
1.0
2.0
3.0
4.0
5.0
[undefined]
baseline
NTX 100mg
0.0
1.0
2.0
3.0
4.0
5.0[undefined]
0.0
1.0
2.0
3.0
4.0
5.0
[undefined]
0%10%
20%30%
40%50%
60%70%
80%90%
100%
0% 50% 100%
Occupancy of NTX at KOR
Redu
ction
in D
rinki
ng
Figure F. In most brain regions (10/14) examined, the mean VT value of 12 healthy controls (red) was intermediate between that of FHP (blue) and FHN (green) drinkers. Two FHP had mean values mostly higher than mean VT for HC; the 1st
FHN subject had lower VT in most regions than the means of the HC cohort.
Figure 9. Timeline for a week of lab testing, NTX treatment, and imaging. All subjects participate in Phase I; only HD participate in Phase II. Sometimes ADP in a given condition precedes PET to accommodate scheduling.
introduced by Zhou et al ([49, 50]) for the characterization of reversible tracers. The Zhou method appears to avoid the negative bias with noise that afflicts the Logan plot. A metabolite corrected plasma curve was used as the input function. The images reveal a dramatic difference in the blocking of [11C]PKAB uptake by NTX and hence a dramatic difference in the occupancy of NTX at the KOR after a week of NTX treatment (described below).
j. Percent reduction in drinking follows occupancy of NTX
In our preliminary cohort of 2 FHP and one FHN drinkers, we also examined their drinking behavior (see description of Alcohol Drinking Paradigm in Methods sect iv). As we hypothesize in Specific Aims, the FHP drinkers had very high NTX occupancy at KOR and large percentage reduction in drinks as measured by the ADP. In contrast, the first FHN drinker scanned had low NTX occupancy and correspondingly lower reduction in drinking after the same week of NTX treatment. See Figure E.
k. Baseline KOR in HD vs HC
We have examined baseline KOR values in 12 healthy control subjects being scanned as part of a separate project. The comparison between baseline uptake of [11C]PKAB by drinkers and HC suggests (assuming no group differences in nondisplaceable uptake) that KOR levels in HC are intermediate between FHP and FHN drinkers (see Figure F). These observations prompted us to extend our proposal to 4 years) in order to perform comparisons of baseline KOR levels in heavy drinkers and HC subjects with properly-matched family histories of drinking(details, in Methods. We note that the regional distribution of [11C]PKAB is similar across groups. By expanding our plan to a comparison of drinkers and HCs, we can begin to address the effects of long-term drinking on the kappa system.
Summary: As we have shown above, we have developed and validated a specific KOR ligand that has high specificity and sensitivity and that can be combined with our established ADP to evaluate the hypotheses that are central to the current proposal.
2. Methodsi. Four year plan
A total of 72 subjects (36 heavy drinkers, 36 social drinking controls; 50% FHP and 50% FHN) will be tested over the 5-year long project (18 completers per yr).
ii. Study designThe basic study design is in Figure 9 above. Controls will complete Phase I (through PET I and no ADP I). In heavy drinkers, because we are interested in investigating the implications of baseline KOR, as well as Occupancy of KOR by NTX, we must perform 2 PET scans on each subject; PET I at baseline and PET 2
following 5-11 days of NTX pretreatment. Similarly, because we are interested in relating Responsivity to Occupancy of KOR by NTX we must make two assessments of drinking behavior using the ADP; ADP 1 at baseline and ADP 2 after 6-11 days of treatment with NTX. All eligible subjects will complete one mock PET scan and one mock ADP at the Hospital Research Unit (HRU) of Yale New Haven Hospital (YNHH) and an MRI scan and then complete PET 1. In heavy drinkers who will also complete the baseline ADP the order of the PET I and ADP I can be be switched to allow for flexibility in scheduling at the HRU where the ADP is conducted and the PET center. Heavy drinkers will then start NTX dosage will be initiated at 25 mg and tapered up over the first three days to 100 mg/day (dose will be given po every morning at 10am). Subjects will be maintained outpatient on the 100 mg/day dose from Day 4 through Day 11 (depending on when the second ADP and second PET scan can be scheduled); The second ADP and PET scan will be held on Days 6-11 in order to 1) allow for NTX levels to stabilize following three half-lives, 2) to ]closely mirror the timing used in our original paper [3] and 3) to allow for flexibility in scheduling at the PET center and the HRU. During this outpatient treatment period, participants will told be to not alter their drinking behavior. The second PET scan and ADP order will also be interchangeable in order to allow for any scheduling concerns. The second PET scan will be conducted at peak NTX levels (roughly 1-4 hours after oral dosing).
In designing the study, we had to balance various competing interests and we provide justifications below:a. Lack of Placebo and use of within subjects design: We considered including a placebo control. But then in
order to control for “order” effects, we would have had to randomize placebo and NTX conditions AND allow a significant washout period between medication conditions, especially for those receiving NTX first. This would have led to problems with retention and increased the sample size. We also considered a between subjects design but this would have also increased sample size and costs. Therefore, in order to contain costs and increase feasibility, we decided to use a within subjects design examining a change from baseline, and configured the overall duration of participation to be as short as possible, in order to enhance subject retention. While our proposed design could still have “order” effects we do not believe that it will influence our ability to detect an initial signal.
b. Use of Mock ADP and PET sessions: To minimize the possible stress and novelty associated with a first PET scan and the ADP sessions, we will conduct a mock PET scan and a mock ADP session prior to the baseline PET and ADP. During the mock PET session the subject will lie in a replica of the PET scanner, be told when the scanner is being “turned on” and when blood samples are “being drawn”. During the mock ADP session, participants will be exposed to the HRU rooms in YNHH where the ADP’s are conducted and will be walked through the ADP procedures and given the opportunity to taste and drink their three chosen beverages (of which they will chose one for the baseline and NTX ADP; see below). These procedures will make them more comfortable drinking in the HRU room and will also help reduce order effects in the current design.
c. Timing of PET scan: The preliminary PET test/restest data suggest that VT and BP for our new KOR tracer may be higher in the afternoon (Figure 6). (Note: This cannot explained by a carryover effect of a high-mass injection of tracer in the morning (test) scan, since that would cause BPND to be lower in the second scan.) We speculate that this effect could be due to diurnal fluctuations in dynorphin level. Regardless of cause, we will control for it by (a) scheduling a subject’s two scans at the same time of day and (b) guarding against possible biases in scheduling between FHP and FHN cohorts.
d. Dose, Timing of NTX and PK samples: We will taper subjects up to a NTX dose of 100 mg/day. 100 mg/day is the dose of NTX that has shown efficacy in many recent clinical trials [9], and our own preliminary evidence suggests that this dose may produce the greatest reduction in drinking in FHP drinkers but increased drinking in FHN drinkers. We considered testing more doses of NTX but we felt that this would not have been justified and would have been too costly. NTX dose will be administered at 10 am each day. All participants will start with 25 mg/day and be tapered up to 100 mg by day 3; following this participants will be maintained at the 100 mg dose for 6 days (to mirror the administration period used in our earlier work [3] to 11 days (to allow for flexibility in scheduling at the PET center and HRU). We will obtain two blood samples on the day of the second PET scan, and one on the day of the NTX ADP, to confirm that steady state levels of NTX were in effect during these procedures. These PK samples will be tested for NTX and its major active metabolite, 6(beta)-naltrexol.
e. Tobacco Use and Smoking breaks : In our experience, 50% of alcoholics who participate in our projects smoke cigarettes or use other tobacco products. In order to avoid nicotine withdrawal, on the ADP days,
participants will be given smoke breaks until about an hour prior to the start of the ADP. We will not allow them to smoke during the ADP since this could not only alter their craving for alcohol but could also pose a potential risk (since YNHH is a nonsmoking facility and the participant would have to be walked out for a smoke break while intoxicated). On the PET days, we will allow regular smoke breaks up until about two hours prior to PET scans. It is very unlikely that the smoking on the morning of the scans would cause kappa receptors to up-regulate. Since these individuals have been smoking and drinking regularly, we expect that any receptor regulation from smoking probably happened a long time ago. Further, if there is any acute dynorphin release associated with a smoking break, it will occur sufficiently early relative to the start of the PET scan to minimize any residual effect of transiently elevated dynorphin on our measurements of KOR level (VT).
iii. Recruitment and screening We anticipate that we will need to recruit 44 male and female non-treatment-seeking, heavy drinking, alcohol-dependent volunteers, and 40 age- and gender- matched light social drinkers who meet all eligibility criteria. Based on our current experience, we anticipate a 10% drop out during initial recruitment (all) and a subsequent 10% dropout (of HD) following the baseline ADP leaving us with a final completed sample of 36 heavy drinkers and 36 social drinkers. All heavy drinkers will meet current DSM-IV criteria for alcohol dependence, will be drinking at hazardous drinking levels as defined by NIAAA (25-50 drinks/week for men and 20-45 drinks per week for women) and will report not being abstinent more than 3 days per week. The age- and gender-matched social drinkers will be light drinkers consuming less than or equal to 5 drinks per week in the past 90 days and reporting less than 5 binge drinking episodes in the past year (similar to those used in studies of social drinkers conducted by [51]; [52]; [53]) and not meet past or current criteria DSM-IV abuse or dependence criteria. All other criteria, designed to protect subjects from risks of participation are described in the human subjects section and are similar to those used by our group in earlier projects [2, 3, 36]. Participants will be recruited through advertisements in local newspapers, community TV channels and postings in community locations (bars, alcohol/coffee shops, grocery stores). We have also had success with recruiting participants from postings on Craigslist and social networking sites like Facebook as well as our study URL, www.paidalcoholstudy.com, which links subjects to confidential surveys that assess some preliminary eligibility criteria and provides preliminary information about the project. Each potential participant’s eligibility will be assessed over the phone by the research assistant who will then set them up for the initial intake appointment. We do not anticipate any problems with recruiting these subjects since their profile is similar to that of the participants we have been successfully recruiting for our ongoing projects. a. Women : Based on our previous experience with testing alcoholics in laboratory studies of NTX, and on the
basis of the ratio of alcohol dependence in men and women in the community, we anticipate that women will make up 25-30% of our sample. We will exclude women who are pregnant, nursing or not using a reliable form of birth control. We will also obtain a detailed 3 month timeline of menstrual cycle phase.
b. Minorities : Based on our previous experience, we anticipate that the demographics for the proposed project will be 76% Caucasian, 21% African-American, and 3% Asian, as well as 11% Hispanic. Participants will be scheduled for an appointment at the Substance Abuse Center in the Connecticut Mental Health Center, New Haven, CT, where informed consent will be obtained and eligibility will be assessed. At this DSM-IV criteria (SCID, [54], Family history of alcoholism (FHAM; [55]) and drinking over the past 90 days (TLFB; [56, 57]) will be assessed. The research assistant will schedule a physical exam (including EKG) and routine laboratory work, and hepatic, kidney, and thyroid function tests and pregnancy tests for all females. Subject characteristics and medical history will be reviewed by the PI’s and the study physician (Dr. Julia Shi, medical director of CMU) to ensure that the subject meets all the eligibility criteria. Eligible participants will be scheduled for the study procedures (as described in figure 9) and will be reminded to not use any illicit drugs or other medications throughout the duration of the study.
iv. Laboratory-based Alcohol Drinking Paradigm(ADP): responsivity to NTXAll the procedures described below are similar to those used by our group in multiple projects [ 2 , 3 , 36 ] including ongoing grants (P50AA12870) that use the ADP. The alcohol drinking procedures used follow the NIAAA (2005) council guidelines. a. Mock Lab session: All eligible participants will be scheduled to participate in the mock ADP, during which they
will be shown the HRU room where the ADP’s will be conducted, walked through all the procedures of the ADP,
and will have the opportunity to taste three different drinks (see section c. below) and consume alcohol in the HRU room. Blood samples for genetic analyses will be collected at this appointment.
b. Medication Administration: Eligible participants will receive NTX for a minimum of seven days. The dose of NTX will be tapered up over the first four days as shown in Figure 9 above. Participants will come into our clinic on a daily basis to take their medications and provide reports on drinking behavior and adverse events.
c. Alcohol Drinking Paradigm: The procedures for all the alcohol-drinking sessions will be similar. Participants will be told not to consume alcohol starting at 11 pm the earlier night and arrive at the Hospital Research Unit (HRU) of Yale New Haven Hospital, New Haven, at 9 am. We will first assess breath alcohol levels and conduct urine drug tests. If the urine drug tests are positive and/or breath alcohol levels are 0.05 or greater, then the session will be rescheduled. The ADP will be conducted in a private room in the HRU. It will start at 3 pm with the priming dose period, which will be followed by three one-hour drinking periods (4-5 pm, 5-6 pm, and 6-7 pm), and will conclude at 7 pm.
i. Priming dose (PD) periodThe PD of alcohol will be provided at 3 pm to model a “lapse” situation and subjects will have 5 minutes to drink it. A 40 minute absorption period will follow during which the alcohol craving (AUQ; [58]), stimulation/sedation (BAES; [59]) and physiological effects (heart rate, blood pressure, breath alcohol levels) of this priming dose of alcohol will be monitored every 10 minutes.ii. Alcohol self-administration (SA) periods Following the PD, participants will be exposed to three one-hour SA periods designed to model a “relapse” situation. During each SA period they will be permitted to drink up to four alcoholic drinks designed to raise Blood Alcohol Level (BAL) by 0.015 mg% of alcohol, or to receive cash (equivalent to the price of each drink that is not consumed). The first SA period will begin at 4 pm, when the research assistant will take 4 prepared drinks into the room along with a "tab" sheet worth $12. The participant will be informed that these 4 drinks will be available to him/her for the next 60 minutes (i.e., until 5 pm). S/he can choose either to drink or to keep the money; each drink will cost $3. For example, if the participant chooses to drink only one drink in the next one hour, s/he will earn $9. The money will be given to them the next morning before they leave the hospital. The second and third SA periods will begin at 5 pm and 6 pm, respectively, and will be similar to the SA period. Thus, participants can choose to consume up to 12 additional drinks over this 3 hour period or to receive up to $36 to take home the next morning. Number of drinks that each participant consumes during this period will serve as the primary outcome for this portion of the experiment. iii. Beverage content and mixersThe YNHH Investigational Pharmacy will calculate the alcohol dose for each participant. The PD dose will be designed to raise blood alcohol levels to 0.03 mg% and will be based on the formula specified by Watson [60] which takes into account gender, weight, and age of the subject. The subsequent drinks provided in the SD blocks will raise BAL by 0.015 mg% each, using the same formula. The alcohol doses will be delivered to the HRU unit and any unused doses will be returned to the pharmacy. Alcoholic beverages administered during this study will consist of 1 part 80 proof liquor of the subject's choosing to 3 parts mixer chosen from a selection of equi-caloric, noncaffeinated, non-carbonated drinks. The research assistant will prepare the drinks using the alcohol doses prepared by the YNHH pharmacy. Participants would have already chosen their favorite alcohol and mixers on an earlier day. Specifically, they will be asked to choose their favorite alcohol and three mixers to go with it from a list that will be provided to them at the intake appointment. The mixers on this list have been chosen to be equicaloric and to not contain any carbonation (which could alter absorption of alcohol). During their mock ADP session subjects will be asked to taste all three mixed drinks and pick the one that they would like to use in the baseline and NTX ADP’s.
d. End of alcohol self-administration period and overnight stay in HRU: The alcohol administration portion of the study will end at 7:05 pm. Following this, breath alcohol levels and craving will be assessed every 30 minutes until the subject's breath alcohol level falls below 0.02. Participants will get dinner and will stay in the hospital overnight and will be discharged between 6 and 7 am the next day.
e. Assessments during and after the three choice periods : During the second, third and fourth hour of the laboratory session alcohol craving (AUQ; [58]) and stimulation/sedation (BAES; [59])and breath alcohol levels (assessed using Alco-Sensor 3; Intoximeter, St. Louis, MO following rinsing of mouth) will be
assessed every thirty minutes. The range of assessments, however, is limited to avoid interfering with the evaluation of drinking behavior. Changes in craving and stimulation/sedation can then be correlated to KOR occupancy. Breath alcohol levels will be used to confirm that there are no differences in alcohol absorption between subjects.
f. Follow up appointments: Subjects will participate in a one week follow-up appointment. Drinking over the past week will be determined using TLFB techniques and any remaining adverse events will be monitored. A brief motivational intervention will be provided to encourage the subject to address their alcohol problem and an immediate referral to treatment will be made if subjects are interested. Even though the subjects participating in this study are not seeking treatment for their drinking, we feel that their participation in this project provides us with a “teaching moment” to address their drinking behavior. We have found that similar brief advice resulted in decreases in alcohol-drinking behavior and increased motivation to quit drinking [61]. As previously done, this intervention will be based on the principles of Miller’s Motivational Enhancement Therapy (MET) [62]. We will also provide them with the NIAAA brochure, “Rethinking Drinking.”
v. PET Methodsa. Data Acquisition : PET scans will be conducted on the Siemens High Resolution Research Tomograph
(HRRT), the world’s highest sensitivity and resolution human brain PET scanner, with a practical resolution of 2.5-3 mm [63]. The Yale PET center has conducted more than 400 human scans on the HRRT with 20 different tracers. As shown in Figure 8, the excellent spatial resolution of the HRRT will be critical in discerning small structures [64]. The HRRT is a list-mode machine so we can bin up the dynamic data anyway we choose. We will create a graduated series of time-frames stating with 30 second frames and lengthening to 5 minutes. Hardware motion correction using the Vicra system (NDI Systems, Waterloo, Ontario) will be performed on an event-by-event basis. Arterial blood samples will be acquired via a continuous withdrawal device (PBS-101, Veenstra Instruments, Joure, The Netherlands) for the first 5-10 minutes of the scan and manually thereafter, and radioactive metabolites identified by HPLC. Plasma concentration of free parent tracer over time will be used for subsequent kinetic modeling analysis. Emission images will be reconstructed iteratively with built-in corrections for attenuation, normalization, scatter, randoms, deadtime and subject motion [27].
b. PK samples : PK samples will be collected during the NTX PET session. Days 7-11 were chosen for the NTX PET scan to build some flexibility into scheduling but maximize the likelihood that NTX levels would be at equilibrium, which will be confirmed by PK analyses. We will also look (on a population basis) for a relationship between occupancy of KOR and NTX (and 6(beta)-naltrexol) levels and interaction with FH of alcoholism. Levels of naltrexone and 6-b-naltrexol will be assayed National Medical Services (Willow Grove).
c. ROI analysis and parametric image generation: After data reconstruction with all corrections, a new summed image (0-10 min) will be registered to the subject's T1-weighted MR image, which will be registered to an MR template so that PET and MR images are in the same MNI space. ROI’s, including the 13 examined in the Preliminary data section in Figure 6 will be taken from a template (Anatomical Automatic Labeling (AAL) for SPM2). Using the arterial input curve, images of VT can be generated pixel-by-pixel in various ways, e.g., the Logan graphical method ([65]) or the more recent refinement by Zhou et al [49, 50]) as shown in Figure D.
d. Possible to dispense with arterial sampling, 2T model fitting? : Since the initial submission of our proposal, we have determined that there is a reliable reference region does not exist for the tracer, [11C]PKAB (see Figure B), so arterial lines will be required for all subjects. 2T model fitting of regional time-activity data and estimation of VT will be the primary outcome of modeling. Measurement of receptor occupancy will still be possible ([46]).
e. Possible to shorten scan duration? Since the initial submission of our proposal, we have determined that it will be possible to shorten the scan duration from 2 hrs to 1 ½ hrs without any loss of accuracy or precision (see Figure C).
f. Analysis of Initial Cohort: Thanks to the preliminary data acquired since the initial submission of our proposal, we have now implemented an optimal scanning protocol and will not need to analyze an initial cohort for the purpose of setting the final protocol.
g. Modeling of PET data to yield BP in regions of interest and parametric images: Standard kinetic modeling approaches
With plasma input function measurements in hand, we can compare multiple standard methods of estimating our desired endpoints to determine the best model or technique [66] that yields the lowest variance in the estimates. Because there is no reliable reference region for our KOR tracer, VT will be our primary endpoint for KOR at baseline. 2T-model fitting will be used to produce estimates of a secondary endpoint, BPND (= k3/k4), but these may prove highly variable. Our primary endpoint for effect of NTX at KOR, OccupKappa, will be calculated via Lassen plot using regional estimates of VT pre- and post-NTX ([46]) Measuring occupancy does not depend on displacement, time of day It is important to note that the half-life of NTX is very long (10.5 hrs in the blood at 100 mg oral dose) and subjects will have been on 100 mg for at least 2 days before the NTX scan. Therefore, NTX in the brain will be at a nearly constant level prior to and during the NTX-scan. Thus, we are not measuring displacement of [11C]PKAB but rather, the effect of a pre-blocking (i.e., a measurement at a second steady state). This is an important distinction. The latter does not depend on the displaceability of the tracer (a kinetic issue), and thus, we have not addressed it in our preliminary data. Since the initial submission of our proposal we scanned two FHP heavy drinkers before and after the prescribed NTX period. One subject’s post-NTX PET scan occurred 1.5 hours after the final NTX dose whereas the other subject’s post-NTX PET scan occurred 8 hours after NTX. Both FHP drinkers were shown to have OccupKappa > 95%. This gives further support to our assertion that after 7-11 days of NTX, drug is at a steady state level and time of PET scanning (as long as pre- and post-NTX times are matched) will not affect results.
h. Problems and anticipated solutions: We do not anticipate any problems with recruiting social and heavy drinkers. (1) Due to the longer duration of participation required in this project we have anticipated a dropout rate of 10%; we propose to examine these rates at the end of the first year and if need be alter and enhance our recruitment rates. (2) We have also allowed for problems with scheduling concerns by using a flexible scheduling design. (3) While we anticipate that steady state levels of naltrexone will be achieved within 3-5 days of dosing, we will obtain PK samples to confirm NTX levels prior to KOR PET imaging.
vi. Statistical considerations a. All Data will be entered and cleaned prior to analyses. Statistical tests will be conducted using SAS v9.1 and SPSS v16 (or later) and will be performed at 2-tailed alpha level of 0.05Primary Aim 1. Occupancy of KOR by NTX and drinking: To determine the degree to which NTX occupancy of kappa mediates responsivity to NTX in heavy drinkers. Responsivity to NTX is defined as the change in the total number of drinks consumed in the NTX ADP versus the baseline ADP. Tests for mediation will be conducted using the regression strategies described initially Baron and Kenny [67], and the more recent elaboration by Kraemer et al. [68]. The proposed mediator (i.e., the potential mechanism of change), occupancy of KORs, is measured as the fractional change in BPND before and after treatment with NTX. The analyses will determine whether occupancy is associated with a) NTX treatment, and b) NTX responsivity. Primary Aim 2. Family History as a moderator: To determine if FH status alters the relationship between NTX responsivity and occupancy of KOR. FHP and FHN will be compared on NTX responsivity and KOR occupancy. Tests for moderation will be conducted using the strategies of Baron and Kenny [67], and Kraemer et al. [68] to look for significant FH by dose interactions on NTX response and occupancy.Primary Aim 3. Baseline KOR differences: To determine if baseline KOR levels differ between FHP and FHN (both HC and HD). For HD, determine if KOR levels are related to either drinking in baseline or NTX ADP. FHP and FHN groups will be compared using chi-square tests for categorical variables, and using Mann-Whitney or t-tests for continuous variables. All continuous variables will be examined for adherence to the normal distribution using normal probability plots and Kolmogorov-Smirnov tests. If normality is not satisfied and transformations do not help with achieving normality, alternative analytic strategies will be considered such as generalized estimating equations or nonparametric methods. The relationship between baseline KOR and drinking will be examined via Pearson product-moment correlations. Secondary Aim 1 Kappa and behavioral indicators relevant for alcoholism: To examine the correlations between occupancy of KOR (also baseline VT) and alcohol craving, stimulation and sedation. A series of Spearman rank-order and Pearson correlation analyses will be conducted. Exploratory Aim 1: To determine if polymorphisms of the OPRM1 and OPRK1 genes are related to KOR receptor levels or naltrexone response. A series of correlational analyses will be used.b. Power analysis: The sample size was based on power calculations for the hypotheses of (a) reduced drinking in the NTX condition, and (b) power to detect an interaction by FH. Power calculations were generated
using G*Power software Version 3.0 [69]. Effect sizes were estimated based on the preliminary FH and NTX effects on drinking behavior in the ADP described in [6] and in our preliminary studies. The effect size for the effect of FH is .91. To detect this effect size at alpha=.05 with 80% power, a total of 16 participants would be required for each arm. Therefore the proposed sample size of n=18 per arm would be sufficient to detect differences in drinking between FHP and FHN at the 100 mg dose of NTX.
References
1. O'Malley, S.S., et al., Naltrexone and coping skills therapy for alcohol dependence. A controlled study. Arch Gen Psychiatry, 1992. 49(11): p. 881-7.
2. O'Malley, S.S., et al., Naltrexone decreases craving and alcohol self-administration in alcohol-dependent subjects and activates the hypothalamo-pituitary-adrenocortical axis. Psychopharmacology (Berl), 2002. 160(1): p. 19-29.
3. Krishnan-Sarin, S., et al., Family history of alcoholism influences naltrexone-induced reduction in alcohol drinking. Biol Psychiatry, 2007. 62(6): p. 694-7.
4. Weerts, E.M., et al., Differences in delta- and mu-opioid receptor blockade measured by positron emission tomography in naltrexone-treated recently abstinent alcohol-dependent subjects. Neuropsychopharmacology, 2008. 33(3): p. 653-65.
5. Huang, Y., et al., Imaging Kappa Opioid Receptor in Humans with a Novel Antagonist Tracer, [11C]LY2795050. Annual Meeting of Society of Nuclear Medicine, 2010. Abstract(Submitted).
6. Garbutt, J.C., et al., Pharmacological treatment of alcohol dependence: a review of the evidence. Jama, 1999. 281(14): p. 1318-25.
7. Krishnan-Sarin, S., S. O'Malley, and J.H. Krystal, Treatment Implications: Using Neuroscience to Guide the Development of New Pharmacotherapies for Alcoholism. Alcohol Research & Health, 2008. 31(4): p. 400-407.
8. Volpicelli, J.R., et al., Naltrexone in the treatment of alcohol dependence. Arch Gen Psychiatry, 1992. 49(11): p. 876-80.
9. Anton, R.F., et al., Combined pharmacotherapies and behavioral interventions for alcohol dependence: the COMBINE study: a randomized controlled trial. Jama, 2006. 295(17): p. 2003-17.
10. Gueorguieva, R., et al., New insights into the efficacy of naltrexone based on trajectory-based reanalyses of two negative clinical trials. Biol Psychiatry, 2007. 61(11): p. 1290-5.
11. Monterosso, J.R., et al., Predicting treatment response to naltrexone: the influence of craving and family history. Am J Addict, 2001. 10(3): p. 258-68.
12. Kosterlitz, H.W., A.D. Corbett, and S.J. Paterson, Opioid receptors and ligands. NIDA Res Monogr, 1989. 95: p. 159-66.
13. Mucha, R.F. and A. Herz, Motivational properties of kappa and mu opioid receptor agonists studied with place and taste preference conditioning. Psychopharmacology (Berl), 1985. 86(3): p. 274-80.
14. Shippenberg, T.S., A. Zapata, and V.I. Chefer, Dynorphin and the pathophysiology of drug addiction. Pharmacol Ther, 2007. 116(2): p. 306-21.
15. Kovacs, K.M., et al., Decreased oral self-administration of alcohol in kappa-opioid receptor knock-out mice. Alcohol Clin Exp Res, 2005. 29(5): p. 730-8.
16. Fadda, P., et al., Differences in the opioid system in selected brain regions of alcohol-preferring and alcohol-nonpreferring rats. Alcohol Clin Exp Res, 1999. 23(8): p. 1296-305.
17. Marinelli, P.W., K. Kiianmaa, and C. Gianoulakis, Opioid propeptide mRNA content and receptor density in the brains of AA and ANA rats. Life Sci, 2000. 66(20): p. 1915-27.
18. Winkler, A. and R. Spanagel, Differences in the kappa opioid receptor mRNA content in distinct brain regions of two inbred mice strains. Neuroreport, 1998. 9(7): p. 1459-64.
19. Edenberg, H.J., et al., A regulatory variation in OPRK1, the gene encoding the kappa-opioid receptor, is associated with alcohol dependence. Hum Mol Genet, 2008. 17(12): p. 1783-9.
20. Lindholm, S., et al., The selective kappa-opioid receptor agonist U50,488H attenuates voluntary ethanol intake in the rat. Behav Brain Res, 2001. 120(2): p. 137-46.
21. Mitchell, J.M., M.T. Liang, and H.L. Fields, A single injection of the kappa opioid antagonist norbinaltorphimine increases ethanol consumption in rats. Psychopharmacology (Berl), 2005. 182(3): p. 384-92.
22. Lindholm, S., et al., Repeated ethanol administration induces short- and long-term changes in enkephalin and dynorphin tissue concentrations in rat brain. Alcohol, 2000. 22(3): p. 165-71.
23. Rosin, A., et al., Downregulation of kappa opioid receptor mRNA levels by chronic ethanol and repetitive cocaine in rat ventral tegmentum and nucleus accumbens. Neurosci Lett, 1999. 275(1): p. 1-4.
24. Pande, A.C., et al., Analgesic efficacy of the kappa-receptor agonist, enadoline, in dental surgery pain. Clin Neuropharmacol, 1996. 19(1): p. 92-7.
25. Walsh, S.L., et al., Enadoline and butorphanol: evaluation of kappa-agonists on cocaine pharmacodynamics and cocaine self-administration in humans. J Pharmacol Exp Ther, 2001. 299(1): p. 147-58.
26. Walker, B.M. and G.F. Koob, Pharmacological evidence for a motivational role of kappa-opioid systems in ethanol dependence. Neuropsychopharmacology, 2008. 33(3): p. 643-52.
27. Carson, R.E., et al., Design of a motion-compensation OSEM list-mode algorithm for resolution-recovery reconstruction for the HRRT, in 2003 Ieee Nuclear Science Symposium, Conference Record, Vols 1-5, S.D. Metzler, Editor 2004. p. 3281-3285.
28. Schinka, J.A., et al., A functional polymorphism gene and risk for abuse of within the mu-opioid receptor alcohol and other substances. Molecular Psychiatry, 2002. 7(2): p. 224-228.
29. Town, T., et al., Association of a functional μ-opioid receptor allele (+118A) with alcohol dependency. American journal of medical genetics, 1999. 88(5): p. 458-461.
30. Kim, S.-G., et al., Association of Functional Opioid Receptor Genotypes With Alcohol Dependence in Koreans. Alcoholism: clinical and experimental research, 2004. 28(7): p. 986-990.
31. Bart, G., et al., Increased Attributable Risk Related to a Functional [mu]-Opioid Receptor Gene Polymorphism in Association with Alcohol Dependence in Central Sweden. Neuropsychopharmacology, 2004. 30(2): p. 417-422.
32. Zhang, H., et al., The OPRD1 and OPRK1 loci in alcohol or drug dependence: OPRD1 variation modulates substance dependence risk. Mol Psychiatry, 2007. 13(5): p. 531-543.
33. Oroszi, G., et al., OPRM1 Asn40Asp Predicts Response to Naltrexone Treatment: A Haplotype-Based Approach. Alcoholism: clinical and experimental research, 2009. 33(3): p. 383-393.
34. Kim, S.-G., et al., A mu opioid receptor gene polymorphism (A118G) and naltrexone treatment response in adherent Korean alcohol-dependent patients. Psychopharmacology, 2009. 201(4): p. 611-618.
35. Anton, R.F., et al., An Evaluation of {micro}-Opioid Receptor (OPRM1) as a Predictor of Naltrexone Response in the Treatment of Alcohol Dependence: Results From the Combined Pharmacotherapies and Behavioral Interventions for Alcohol Dependence (COMBINE) Study. Arch Gen Psychiatry, 2008. 65(2): p. 135-144.
36. McKee, S.A., et al., Varenicline reduces alcohol self-administration in heavy-drinking smokers. Biol Psychiatry, 2009. 66(2): p. 185-90.
37. Anton, R.F., et al., Naltrexone effects on alcohol consumption in a clinical laboratory paradigm: temporal effects of drinking. Psychopharmacology (Berl), 2004. 173(1-2): p. 32-40.
38. Drobes, D.J., et al., A clinical laboratory paradigm for evaluating medication effects on alcohol consumption: naltrexone and nalmefene. Neuropsychopharmacology, 2003. 28(4): p. 755-64.
39. Voronin, K., et al., Aripiprazole effects on alcohol consumption and subjective reports in a clinical laboratory paradigm--possible influence of self-control. Alcohol Clin Exp Res, 2008. 32(11): p. 1954-61.
40. Krishnan-Sarin, S., et al., Memantine effects on human ethanol consumption: initial impressions. Alcoholism: Clinical and Experimental Research, 2009. 33(s1): p. 290A.
41. Henriksen, G. and F. Willoch, Imaging of opioid receptors in the central nervous system. Brain, 2008. 131(Pt 5): p. 1171-96.
42. Quirion, R., C. Pilapil, and J. Magnan, Localization of kappa opioid receptor binding sites in human forebrain using [3H]U69,593: comparison with [3H]bremazocine. Cell Mol Neurobiol, 1987. 7(3): p. 303-7.
43. Ko, M.C., et al., Studies of micro-, kappa-, and delta-opioid receptor density and G protein activation in the cortex and thalamus of monkeys. J Pharmacol Exp Ther, 2003. 306(1): p. 179-86.
44. Slater, P. and A.J. Cross, Autoradiographic distribution of dynorphin1-9 binding sites in primate brain. Neuropeptides, 1986. 8(1): p. 71-6.
45. Innis, R.B., et al., Consensus nomenclature for in vivo imaging of reversibly binding radioligands. J Cereb Blood Flow Metab, 2007. 27(9): p. 1533-9.
46. Cunningham, V., et al., Measuring drug occupancy in the absence of a reference region: the Lassen plot re-visited. Journal of cerebral blood flow and metabolism, 2010. 30(1): p. 46-50.
47. Zheng, M.-Q., et al., [C-11]LY2795050, An Antagonist Tracer for PET Imaging of the Kappa Opioid. Annual Meeting of Society of Nuclear Medicine 2010. Abstract(Submitted).
48. Henriksen, G., et al., Recent development and potential use of u- and k-opioid receptor ligands in positron emission tomography studies. Drug Dev Res, 2006. 67: p. 890-904.
49. Zhou, Y., et al., A consistent and efficient graphical analysis method to improve the quantification of reversible tracer binding in radioligand receptor dynamic PET studies. NeuroImage, 2009. 44(3): p. 661-670.
50. Zhou, Y., et al., Multi-graphical analysis of dynamic PET. NeuroImage, 2010. 49(4): p. 2947-2957.51. King, A.C. and J.A. Byars, Alcohol-induced performance impairment in heavy episodic and light social
drinkers. Journal of Studies on Alcohol, 2004. 65(1): p. 27.52. Holdstock, L. and H. de Wit, Ethanol Impairs Saccadic and Smooth Pursuit Eye Movements Without
Producing Self-Reports of Sedation. Alcoholism: clinical and experimental research, 1999. 23(4): p. 664-672.
53. Holdstock, L., A.C. King, and H. de Wit, Subjective and Objective Responses to Ethanol in Moderate/Heavy and Light Social Drinkers. Alcoholism: clinical and experimental research, 2000. 24(6): p. 789-794.
54. First, M.B., et al., Structured Clinical Interview for DSM-IV Axis I Disorders, Research Version, Patient Edition1996, New York: Biometrics Research, New York State Psychiatric Institute.
55. Rice, J.P., et al., Comparison of direct interview and family history diagnoses of alcohol dependence. Alcohol Clin Exp Res, 1995. 19(4): p. 1018-23.
56. Cooper, A.M., et al., Validity of alcoholic's self-reports: duration data. Int J Addict, 1981. 16(3): p. 401-6.57. Maisto, S.A., et al., Comparison of two techniques to obtain retrospective reports of drinking behavior
from alcohol abusers. Addict Behav, 1982. 7(1): p. 33-8.58. Bohn, M.J., D.D. Krahn, and B.A. Staehler, Development and initial validation of a measure of drinking
urges in abstinent alcoholics. Alcohol Clin Exp Res, 1995. 19(3): p. 600-6.59. Martin, C.S., et al., Development and validation of the Biphasic Alcohol Effects Scale. Alcohol Clin Exp
Res, 1993. 17(1): p. 140-6.60. Watson, T.E., Total body water and alcohol levels: updating the fundamentals, in Human Metabolism of
Alcohol, K.E. Krow and R.D. Batt, Editors. 1989, CRC Press: Florida. p. 41-66.61. Sinha, R., et al., Naturalistic follow-up of drinking behavior following participation in an alcohol
administration study. J Subst Abuse Treat, 1999. 17(1-2): p. 159-62.62. Miller, W.R., et al., Motivational Enhancement Therapy manual: A clinical research guide for therapists
treating individuals with alcohol abuse and dependence1992, Rockville, MD: National Institute on Alcohol Abuse and Alcoholism.
63. de Jong, H.W., et al., Performance evaluation of the ECAT HRRT: an LSO-LYSO double layer high resolution, high sensitivity scanner. Phys Med Biol, 2007. 52(5): p. 1505-26.
64. Heiss, W.D., et al., Metabolic rates in small brain nuclei determined by high-resolution PET. J Nucl Med, 2004. 45(11): p. 1811-5.
65. Logan, J., et al., Graphical Analysis of Reversible Radioligand Binding from Time-Activity Measurements Applied to [N-11C-methyl]-(-)-Cocaine PET Studies in Human Subjects. J Cereb Blood Flow Metab, 1990. 10(5): p. 740-747.
66. Gallezot, J.D., et al., Kinetic modeling of the serotonin 5-HT(1B) receptor radioligand [(11)C]P943 in humans. J Cereb Blood Flow Metab, 2010. 30(1): p. 196-210.
67. Baron, R.M. and D.A. Kenny, The moderator-mediator variable distinction in social psychological research: conceptual, strategic, and statistical considerations. J Pers Soc Psychol, 1986. 51(6): p. 1173-82.
68. Kraemer, H., et al., Mediators and moderators of treatment effects in randomized clinical trials. Archives of General Psychiatry, 2002. 59: p. 877-883.
69. Faul, F., et al., G*Power 3: A flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behavior Research Methods, 2007. 39: p. 175-191.
