Provision of Community-Based Health Services to Injection Drug Users

Frederick L. Altice, M.D.
Yale University AIDS Program, New Haven, CT

Much of what is known about the health status and health service utilization of injection drug users (IDUs) is derived from drug treatment-derived cohorts and hospital/clinic settings. Little is known about the relationship between out-of-treatment IDUs and the health care delivery system. To address the needs of this population, the first mobile needle exchange (NEP)-based health services program was deployed in New Haven, Connecticut, in 1993 and expanded in June 1996 to provide a comprehensive array of clinical, case management, and drug treatment services. To evaluate this community intervention, an independent cohort of 373 out-of-treatment IDUs was recruited using respondent-driven sampling methodology.

This presentation will describe the health status and co-morbid conditions (chronic diseases, infectious diseases complications, and mental illness) of this population. The presentation will also describe the characteristics of the individuals who used NEP-based health services and pertinent medical, drug treatment, and case management issues encountered. It will also examine the correlates of emergency room, inpatient hospitalization, and primary care services by this population. Lastly, to evaluate our interventions effect on one marker of fragmented health care, we will describe the impact of NEP-based mobile health services on the use of emergency room services by those out-of-treatment IDUs who used the services versus those who did not.

A Controlled Comparison of the Buprenorphine-Naloxone Tablet and Methadone for Opioid Maintenance Treatment: Interim Results

Leslie Amass
University of Colorado School of Medicine, Denver, CO

A combination buprenorphine-naloxone tablet (BNX) containing 8 mg of buprenorphine and 2 mg of naloxone is pending approval by the FDA for opioid dependence treatment. This ongoing, controlled study of 300 patients is comparing the efficacy of BNX to that of methadone (METH) for opioid maintenance treatment. An interim analysis was conducted of 162 opioid-dependent outpatients assigned randomly to either BNX8, BNX16, METH45, or METH90 mg dose groups for 118 days. Daily dosing was double-blind and double-dummy. Urine samples were collected thrice weekly, manualized behavioral counseling was provided twice per month, and Addiction Severity Index (ASI) ratings were collected at intake and 8 and 16 weeks. There were no significant differences across groups in the number of patients completing the study, medication compliance, rates of opioid use, or self-reported drug use. All groups improved significantly over time on ASI composite scores of Drug, Opiates, Employment, and Legal, with a trend toward improvement on the Family measure. Fifty-five of the 162 patients (34 percent) completed the entire study. For treatment completers, patients receiving BNX tended to have higher rates of opioid abstinence (64 percent for BNX8 and BNX16) than those receiving METH (36 percent for METH45 and 52 percent for METH90). Patients receiving BNX were also significantly better at achieving at least 2 weeks of continuous opioid abstinence (84 percent for BNX8 and 70 percent for BNX16) than patients on METH (41 percent for METH45 and 67 percent for METH90). Finally, Treatment Effectiveness Scores (TES) for opioids were significantly higher for patients receiving BNX (32.2 \u177\'b1 3.6) as compared with patients receiving METH (21.4 \u177\'b1 2.7). These interim results suggest that BNX is as efficacious as METH and that treatment improves several aspects of psychosocial functioning. Further, for a subset of patients receiving at least 4 months of treatment, 8 to 16 mg of BNX is at least as effective as 90 mg METH in reducing illicit opioid use. (Supported by NIDA Grant DA11160)

Adherence With Antiretroviral Therapy in HIV-Infected Drug Users

Julia H. Arnsten, M.D., M.P.H.
Montefiore Medical Center, Bronx, NY

Objective

To describe adherence with antiretroviral therapy (ART) in HIV-infected drug users and determine its impact on HIV viral load.

Methods

Adherence is assessed for 6 months using self-report and medication event monitoring devices (MEMS), which record the time and date of each pill bottle opening. HIV viral load is quantified monthly.

Results

Data are available for 48 patients who have completed 6 months of followup or withdrew (mean length of followup, 5 months). The mean age is 44; patients are 60 percent male, 62 percent hispanic, and 25 percent black; 94 percent are on methadone and 100 percent are insured by medicaid. These preliminary data suggest that adherence is markedly different when measured by MEMS than when measured by self-report. By MEMS, subjects took 54 percent of all prescribed medication doses during the monitored period (among those monitored by MEMS, median length of monitoring is 165 days). By self-report, subjects took 83 percent of all prescribed medication doses. There is a significant negative linear relationship between MEMS adherence and HIV viral load (r=-.33; p=.03) that is not observed between self-reported adherence and viral load (r=-.016; p=.9), suggesting that self-report may exaggerate medication use to an extent that obviates its use. There is significant heterogeneity of adherence rates in this cohort of HIV-infected drug users: 35 percent of subjects achieved greater than 80 percent adherence (by MEMS), while 23 percent achieved less than 20 percent adherence. Two sociodemographic factors were the strongest predictors of poor MEMS adherence: female gender and receipt of social security disability benefits (SSI). Active drug use, more medication side effects, fewer HIV-related symptoms, and depression were associated with poor adherence, but not significantly. It is expected that these factors will become significant predictors when the projected sample size is achieved. Of note, features of the ART regimen were not associated with adherence.

Conclusions

Among HIV-infected drug users, adherence rates by MEMS and self-report are widely divergent, and only MEMS adherence is associated with viral load.

Thomas H. Brandon, Ph.D.
University of South Florida and the H. Lee Moffitt Cancer Center and Research Institute, Tampa, FL

Substance abuse treatments tend to be plagued by poor initial cessation rates and high rates of relapse. Contemporary learning theory may offer avenues toward improving the outcomes of such treatments. Two recent theories are being tested for their ability to explain substance abuse behavior and treatment outcome, with the long-term goal of applying the theories to the development of improved interventions.

The first is a behavioral theory of motivation that may explain the variance among individuals in exhibiting the persistence necessary to cease substance use. The second model illustrates the importance of learning context and may explain why treatments may fail to generalize beyond the clinic.

Michael Gary Byas-Smith, M.D.
Emory University School of Medicine, Atlanta, GA

The overall aim of this presentation is to highlight the potential of imaging technology to identify neurophysiological changes that occur in chronic pain patients on opioid medications. Our hypothesis is that chronic opioid usage in the setting of persistent painful inputs results in abnormal responses to both opioids and acute painful stimulation. To achieve adequate relief, these patients require higher doses of medication and may have lower thresholds to painful stimuli. We ultimately will compare these differences with opioid abusers. Two abstracts will be presented that show images acquired using positron emission tomography (PET) in normal individuals (no prior exposure to opioid medications) and patients who ingest opioids daily. Both regional and global changes were evaluated in these studies. These preliminary results indicate that patients indeed respond differently both to opioid administration and to painful stimulation. The functional significance of these differences has yet to be explained thoroughly. However, we believe that these differences are at least the signature of tolerance development. Through pharmacological manipulation, we are hopeful that we can prevent and reverse the clinical downside of chronic opioid exposure and use imaging studies to characterize and monitor these changes.

Linda Chang, M.D.
University of California, Los Angeles, Torrance, CA

Methamphetamine (METH) is a popular drug of abuse. However, little biological data are available on the potential toxic effects of METH on the human brain. Using advanced neuroimaging techniques, including proton magnetic resonance spectroscopy (1H MRS) and perfusion MRI, and neuropsychological (NP) tests, we evaluated 20 previously METH-dependent, abstinent subjects (10 males and 10 females; lifetime exposure, 9234±3568 grams; last METH use, 7.8±2.1 months) and 20 control subjects without a history of drug abuse. Compared with the control subjects, METH users demonstrate evidence of neuronal loss (with decreased N-acetyl aspartate, a neuronal marker, on 1H MRS) and persistent blood flow and cognitive abnormalities (especially on tasks for psychomotor speed and attention). These findings reflect the underlying neuronal damage associated with prior METH exposure. Ongoing studies are being performed to determine whether some of these abnormalities on neuroimaging studies and the NP tests are reversible. (Supported by grants from NIDA [K-20 DA00280-04; R01 DA12734-01])

Howard D. Chilcoat, Sc.D.
Henry Ford Health Science Center, Detroit, MI

Early initiation of drug use in childhood is an important predictor of later, more problematic drug involvement. In this program of research, we test suspected causal pathways leading toward early drug initiation, using a well-defined epidemiologic sample of low and normal birth-weight children from an ongoing study of the neuropsychiatric consequences of low birth weight. This project combines epidemiologic research with behavioral laboratory research on individual differences in drug use. A total of 473 low birth-weight (<2500 g) and 350 normal birth-weight children were selected from lists of newborns from an urban and a suburban hospital in southeast Michigan. These children and their mothers were assessed when the children were 6 years old and again at age 11 years. Initial findings indicate that many of the sequelae of low birth weight are markers of individual susceptibility for early drug use. Specifically, children with ADHD and externalizing problems measured at age 6 were at increased risk for drug use by age 11, compared with children without these problems. Further, high levels of IQ and school achievement appeared to protect against drug use. Independent of these sequelae, low birth weight signaled increased risk of drug use in boys. Regardless of the presence of these vulnerability markers, children with high levels of parent monitoring had reduced risk of drug use. This finding suggests that protective factors in the social environment can reduce individual risk. Research currently under way tests for mechanisms through which ADHD and behavior problems increase childrenÕs risk for drug use. Specifically, we test whether behavioral measures of risk-taking and impulsivity can be used to guide our understanding of the links between ADHD, behavior problems, and drug use.

Monique Ernst, M.D., Ph.D.
Brain Imaging Center, National Institute on Drug Abuse, Baltimore, MD

Nicotine enhances vigilance and improves cognitive performance. The anatomical substrates of nicotine and its effects on working memory were studied in 11 smokers and 11 nonsmokers using PET and H2 15O water. During the placebo condition, significant activation was present mainly in the prefrontal cortex in both groups, but with L>R lateralization in nonsmokers and R>L lateralization in smokers. This laterality effect suggests that different neural circuits mediate cognitive activation as a function of previous exposure to smoking. During nicotine administration, cortical activation had a similar pattern in both groups, but was enhanced in nonsmokers and unchanged or even reduced in smokers. The lack of significant enhancement of activation in smokers, in contrast to that in nonsmokers, suggests a tolerance effect.

William Brooks Gentry, M.D.
University of Arkansas for Medical Sciences, Little Rock, AR

The goal of this research is to develop therapeutic strategies for stimulant abuse which alter the pharmacokinetic properties of these drugs. We are characterizing the pharmacokinetic and pharmacodynamic properties of two prototypic drugs of abuse (phencyclidine and methamphetamine) during the onset and offset of effects in animal models designed to mimic human drug use. The knowledge gained from these studies is being utilized in the development and testing of antibody-based medications, which alter the concentration vs. time profiles of drugs of abuse. These studies are part of a multidisciplinary approach to the rational development of therapeutic strategies for the treatment of drug abuse.

Randy L. Gollub, M.D., Ph.D.
Massachusetts General Hospital, Charlestown, MA

Introduction

We have previously demonstrated the feasibility of using fMRI to investigate the acute effects of cocaine infusion. We reported networks of brain regions activated by acute cocaine administration that were associated with subjective reports of euphoria and craving in drug-dependent subjects. We have just completed a new study using a higher field strength magnet to determine whether activation in the regions associated with euphoria and drug craving are linked in a dose-dependent manner to cocaine intake.

Methods

Medically cleared cocaine-dependent subjects (n=11) were given three infusions (saline, 0.1 mg/kg, and 0.6 mg/kg; max dose, 40 mg) using a magnet-compatible, automatic injection system (MedRad) separated by 45-120 minutes under double-blind conditions. Physiological parameters (ECG, HR, ETCO2, BP; InVivo OmniTrac 3100, Orlando, FL) and subjective ratings (RUSH, HIGH, LOW, and CRAVING) were measured during the scans. Functional scans were obtained using a T2*-weighted gradient echo sequence (TR=4000 ms, TE=30 ms, Flip=60°, in-plane resolution=3.1 x 3.1 mm, through-plane resolution=4 mm). The whole brain was imaged by placing 29 contiguous slices oriented parallel to the AC-PC line; 272 images per slice were collected (total scan duration, 18:08 minutes). Functional data were motion corrected, normalized, and transformed into Talairach space. Scans with uncorrectable motion were discarded. Group average scans for saline and 0.1 and 0.6 mg/kg cocaine were constructed by vertical averaging. Statistical maps (Kolmogorov-Smirnoff [KS] with drift correction and smoothing) were used to identify infusion-related activation in the individual and group-averaged data sets.

Results

• Physiology and Subjective Ratings: The physiologic responses to saline and 0.1 mg/kg cocaine were indistinguishable. Predictably, HR rose rapidly from 56±9 to 80±6 beats/min, MBP rose from 91±9 to 109±18 mmHg, and ETCO2 dropped from 41±2 to 37±3 torr in response to 0.6 mg/kg cocaine. Subjective ratings of euphoria (RUSH and HIGH) dose dependently increased.

• fMRI Data: KS statistical images constructed from both group-averaged and individual data comparing pre-infusion and post-infusion epochs demonstrate activation in the basal forebrain (BF), amygdala, and ventral tegmentum (VT), anterior cingulate (ant. cing.), insula, caudate, putamen, and hypothalamus in response to 0.6 mg/kg cocaine but not saline, consistent with results from two previous studies performed at 1.5T. Qualitatively, inspection of the KS maps from individual subjects suggests a pattern of dose-dependent recruitment for the anterior and posterior parahippocampus, BF, VT, insula, and ant. cing. For a more quantitative estimate of magnitude of activation, both individual and group-averaged data were used to calculate percent signal change during the time epoch associated with peak euphoria (minutes 7-11 of the scan) in the BF, amygdala, VT, ant. cing., and insula. Dose-dependent activation, defined as those regions showing an intermediate level of activation in response to the 0.1 cocaine infusion, was noted in the bilateral BF, amygdala, and VT, and some but not all regions of the insula and ant. cing. The data suggest that the BF and VT show incremental signal change while other regions are much less responsive to the low dose of cocaine despite robust activation at the higher dose.

Conclusions

In this pilot study, data from individual subjects demonstrate the same pattern of regional brain activation in response to acute cocaine administration using 3T fMRI as do previously published group-averaged data at 1.5T. Activation in a subset of these regions appears to scale in magnitude with dose of cocaine. Specifically, the data suggest an association between the BF, amygdala, VT, and portions of the ant. cing. and insula, and the euphoric response to cocaine. (Supported by NIDA 00275, 00265, 09467 and NIH M01 RR01066 to MGH-GCRC)

Denise Hallfors, Ph.D.
University of North Carolina, Chapel Hill, NC

A major focus of community alcohol, tobacco, and other drug (ATOD) abuse prevention efforts has been to increase the age of first use and reduce the amount of use by adolescents. Evaluating the effects of community initiatives on young people has been difficult, however. While national and State-level surveys have tracked annual trends, these data have not been accessible to evaluate community programs. Local surveys have become more common, yet they too have been underutilized.

In this study, all available community-level school survey data are being collected from archives spanning 15 years in 58 study communities. Twenty of these communities received 5 or more years of funding from either the Robert Wood Johnson Foundation (Fighting Back demonstration) or the Center for Substance Abuse Prevention (Community Partnership Project) to develop locally driven coalitions that would reduce demand for substances. These treatment communities were matched with the remaining 38 communities in a quasi-experimental design for the national evaluation of each of these programs. Survey data are now being meta-analyzed to assess the effects of these two programs on youth.

This presentation provides preliminary findings on the effects on age of first use. Data are limited to communities that surveyed either 7th or 8th graders over more than 1 year, and to questionnaires that included an item on truancy and/or grade point average (proxies for risk). Dependent variables are recent (30-day) alcohol use, binge-drinking, tobacco use, marijuana use, and other drug use. Multilevel modeling is used to examine the predictive value of the following explanatory variables: presence of a coalition program, whether it was youth focused, level of community poverty, quality of the survey sample, risk level of individual students, and a measure of school-based prevention.

Carl R. Lupica, Ph.D
National Institute on Drug Abuse, Baltimore, MD

The ability of the psychoactive components of marijuana, known as cannabinoids (CBs), to disrupt learning, short-term memory, and attention is well established. Because of the prominent role the hippocampus plays in these processes, and because it contains high densities of CB receptors, it has been implicated in mediating the cognitive effects of marijuana. One possible site of CB action is the axon terminals of GABAergic neurons. In an effort to determine whether CB receptors alter GABAergic neurotransmission, and to determine the mechanism(s) through which this may occur, we utilized electrophysiological techniques in rodent brain slices. Whole-cell voltage clamp recordings were obtained from hippocampal CA1 pyramidal neurons, and GABA-mediated inhibitory postsynaptic currents (IPSCs) were measured. IPSCs were reduced by the CB1 receptor agonist WIN 55-212,2 (EC50=138 nM), and this was reversed by the CB1 antagonist SR141716A. We next attempted to identify the ion channels involved using selective blocking agents. Thus, blockade of voltage-dependent Na+ or Ca2+ channels with tetrodotoxin or cadmium, respectively, reversed the inhibition of GABA release by WIN 55-212,2. However, the inhibition of IPSCs by WIN 55-212,2 was unaltered by the presynaptic K+ channel blockers barium or 4-aminopyridine. We conclude that CB1 receptors reduce GABA release exclusively through the inhibition of Ca2+ channels and that K+ channels are not involved in this process. We hypothesize that the inhibition of GABA release represents one mechanism by which marijuana disrupts cognitive processing in the hippocampus.

Catherine A. Martin, M.D.
University of Kentucky College of Medicine, Lexington, KY

Testosterone and carbon monoxide (CO) levels were obtained from 30 female smokers when they participated in a laboratory study at age 21. These smokers had completed questionnaires about their nicotine use in 7th through 10th grades and again at age 21 as part of a NIDA-supported longitudinal cohort study of drug use. History of pubertal onset was also obtained at age 21. In the 30 female smokers, testosterone levels were correlated with cigarette use in the previous 30 days (p<.01), CO levels (p<.05), cigarette use reported in the 7th and 10th grades (p<.05), and history of early pubertal onset (p<.001). Further investigation of the relationships of testosterone, nicotine, and pubertal onset in young female smokers is needed.

Michael A. Mathier, M.D.
Allegheny General Hospital, Pittsburgh, PA

It is well recognized that cocaine use is associated with adverse cardiovascular sequelae, including the development of cardiomyopathy and heart failure. The nature of the interaction between cocaine and the myopathic heart is not well understood, however. To explore this interaction, we studied the cardiovascular effects of intravenous cocaine (1 mg/kg) in nine chronically instrumented dogs prior to and following the induction of moderate heart failure by rapid ventricular pacing (240/min) for 2 weeks. Dogs with heart failure demonstrated exaggerated chronotropic and inotropic responses and blunted coronary blood flow responses to cocaine. This resulted in increased myocardial oxygen extraction and impaired global LV function. Responses to cocaine were not only exaggerated in magnitude following the induction of heart failure, but were also much more rapid, suggesting a role for altered autonomic control. To explore this, three dogs were similarly studied following surgical aortic baroreceptor denervation. These dogs exhibited qualitatively and quantitatively similar responses to the dogs with heart failure. We conclude that in the setting of moderate heart failure, cocaine produces rapid and exaggerated cardiovascular effects largely due to a loss of reflex buffering. These exaggerated effects may account for the presentation of patients with decompensated heart failure following cocaine use.

Lisa Metsch, Ph.D.
University of Miami School of Medicine, Miami, FL

Epidemiologic research data collected on HIV seropositive drug users in Miami, Florida, suggest the need to develop intervention strategies to address primary and secondary prevention needs. Data were collected from 490 seropositive injection drug users (IDUs) and non-injection drug users who initially tested positive at the University of Miami Outreach and Assessment Center. Comparing respondentsÕ risk behaviors at baseline to their 6-month followup visit, significant decreases (approximately 50 percent) in risky sexual practices were reported by both IDUs and non-IDUs, but approximately one-third of HIV-positive injectors and one-half of HIV-positive non-injectors continued to engage in risky injection practices. In another study of access to health services among 245 HIV-infected IDUs and non-IDUs, findings indicated that less than half received any type of HIV treatment and only 22 percent received standard highly active antiretroviral therapies (HAART). Predictors of use of HIV treatment included interference of pain with daily activities, using crack cocaine, having a usual source of care other than the emergency room, having sought care after HIV diagnosis, and having been diagnosed with HIV more than 5 years ago.

Eric Moolchan, M.D.
National Institute on Drug Abuse, Baltimore, MD

Adolescent smokers suffer the same withdrawal symptoms as adults, try to quit and fail, and are frequently interested in treatment. Yet there are few studies of pharmacologic interventions for adolescent smokers, and those studies have typically lacked control groups and biochemical validation. They have also suffered from low participation, high attrition, and low quit rates. To help meet the need for high-quality research on adolescent smoking cessation, we have established an outpatient program called the Teen Tobacco Addiction Treatment Research Clinic. Studies will span the basic sciences (physiology, chemistry/drug metabolism) and clinical interests (pharmacology, neuroimaging, genetic markers of nicotine dependence, novel treatment approaches). Already under way is a 6-month outpatient study to investigate the safety, tolerability, compliance, and efficacy in adolescents of two different forms of nicotine-replacement therapy, in combination with counseling/group support. Upcoming studies include a safety and efficacy trial of buproprion in adolescent smokers and an in vivo nicotine receptor brain imaging study, to be done initially in adults.

Yavin Shaham, Ph.D.
National Institute on Drug Abuse, Baltimore, MD

Using a reinstatement procedure (an animal model of relapse to drugs), we have shown that intermittent footshock stress reliably reinstates drug seeking in rats after prolonged drug-free periods. Here we will present data on the role of corticotropin-releasing factor (CRF) and corticosterone, known to be involved in the coordination of responses to stress, in relapse to drug seeking. CRF infusions into the brain were found to reinstate drug seeking, and CRF receptor antagonists attenuated footshock stress-induced reinstatement of heroin, cocaine, and alcohol seeking. In contrast, blockade of stress-induced corticosterone secretion had no effect on reinstatement of drug seeking induced by the stressor. These data suggest that CRF contributes to stress-induced relapse to drug seeking via its actions on extra-hypothalamic brain sites. Inasmuch as relapse to drug use in humans can be instigated by stress, CRF receptor antagonists might be useful in the treatment of relapse to drugs. (Supported by NIDA and the MRC of Canada)

Mark E. von Zastrow, M.D., Ph.D.
University of California, San Francisco, San Francisco, CA

Our laboratory is interested in membrane trafficking mechanisms that determine the subcellular localization of signal-transducing receptors, as these mechanisms mediate diverse functions in the regulation of signal transduction and differentially regulate closely homologous subtypes of receptor co-expressed in the same cells. We have examined mechanisms that mediate and regulate the rapid endocytosis of several classes of heptahelical G protein-coupled receptor. Studies of the beta-2 adrenergic receptor have identified a multistep mechanism that regulates the ligand-dependent redistribution of receptors between distinct microdomains of the plasma membrane, thus controlling the rate of receptor endocytosis via clathrin-coated pits. Studies of D1 and D2 dopamine receptors indicate that individual subtypes of GPCR are selectively endocytosed by distinguishable molecular mechanisms and membrane pathways, and they demonstrate that this machinery is capable of physically segregating individual receptors within distinct microdomains of the plasma membrane before endocytosis occurs. Studies of opioid receptors reveal a remarkable level of agonist-selectivity in the regulation of the endocytic machinery, which we propose may have important implications for understanding the physiological effects of addictive opiate drugs. Studies in progress are examining protein interactions that control receptor sorting after endocytosis.

Shaomeng Wang, M.D.
Georgetown University Medical Center, Washington, DC

A potent lead compound, representing a novel class of DAT inhibitors, was discovered through 3D-database pharmacophore searching. Our biological evaluation showed that this lead compound has a similar potency to cocaine in binding and uptake but has a significant effect in vitro in antagonizing the ability of cocaine to block the reuptake of dopamine. Furthermore, the lead compound has a different selectivity profile from cocaine at the three transporter sites. Unlike cocaine, the lead compound did not exhibit significant stimulant activity at concentrations from 1 to 30 mg/kg in locomotor activity tests in mice. Chemical modifications led to the identification of a much more potent compound with Ki values of 11 nM and 51 nM in binding and uptake, respectively. Our molecular modeling studies showed that in addition to the pharmacophore, hydrophobicity and conformational preference are two other important factors for the activities of this class of compounds. Taken together, our data suggest that this class of DAT inhibitors may represent promising leads for further development and evaluation as potential therapies for the treatment of cocaine abuse.