Safety Study to Assess Atomoxetine With MA Abusers and Healthy Controls

Methamphetamine (MA) abuse is a national public health concern. People who are dependent on MA have problems with mental functions (e.g., learning, remembering, focusing attention, solving problems). Such problems can interfere with their treatment for MA abuse, and thereby may promote continued drug use. While the effects of MA have been studied in rodents and non-human primates, its effects on the human brain have not been well characterized. This is a study of nontreatment seeking individuals who use MA compared to individuals who do not use MA(control participants). The study has three goals: 1. it aims to identify the brain regions and pathways that may contribute to the problems of MA abusers in performing mental tasks; 2. it will serve as a double-blind, placebocontrolled,within-subjects study to determine the safety and tolerability, and positive effects of MA in MA-abusing volunteers treated with atomoxetine or placebo; 3. It aims to compare the brain activity as measured by structuraland functional magnetic resonance imaging (fMRI). These are noninvasive brain imaging procedures, that will be used to study brain function while control and MA using participants take atomoxetine or placebo and perform tests of memory and concentration. MA abusing participants will undergo a 1-day outpatient screening and if it is safe for the participants to proceed with the study they will participate in two inpatient phases of the study that will occur in the UCLA research setting, the General Clinical Research Center. The first inpatient stay will be 15 days, and the second will be a 9 days stay that includes drug administration and assessments. There will be at least a two week interval between inpatient phases. During the inpatient phases participants will receive alternating study drugs; atomoxetine or placebo and four sessions of IV MA administration or placebo. The study schedule for control participants will include a 1-day outpatient screening and two phases of outpatient administration of atomoxotime or placebo with a two week study drug free interval between the phases. Four to five of the outpatient study visits will involve cognitive tests and brain imaging studies.

As a preliminary step, the investigators will screen the experimental data for normality and for potential associations between outcome variables with demographic variables at baseline. In cases of strong deviations from normality, log or power transformations will be used as appropriate. For the subjective effects questionnaires, the peak effect will be identified for each subject and used as the dependent variable. The primary analytical technique will be linear mixed effects regression. The general linear mixed model (GLMM) is equivalent to repeated measures ANOVA when all subjects have complete data but automatically handles missing observations, producing unbiased estimates as long as the values are missing at random. This approach will allow use of the data from subjects who complete only part of the protocol or who have invalid measurements on some tests, greatly enhancing our power. The sample size is as large or greater than those in the previous human laboratory studies assessing safety and initial efficacy of agonist-like pharmacotherapies for stimulant dependence. For example, these previous studies used samples of 7 (Herin et al., 2010; Stoops et al., 2008; Dackis et al., 2003), and 10 (Sofuoglu et al., 2009) subjects to obtain significant reductions in subjective effects of d-amphetamine by atomoxetine and those of cocaine by modafinil. To be conservative, the investigators base power estimates on completers but will use available data from all 30 subjects in GLMM analyses. This should substantially increase power (power for GLMM and repeated measures ANOVA are equivalent except that ANOVA cannot utilize incomplete data, while a GLMM can).