Biological correlates of self-reported new and continued abstinence in cannabis cessation treatment clinical trials

Highlights

• Traditional 11-nor-9-carboxy-Δ9-tetrahydrocannabinol (THCCOOH) abstinence cut-offs are not sensitive to recent abstinence initiation.

• Agreement between cannabis use and urine THCCOOH measures are not well defined.

• Combining current THCCOOH with recent changes in CN-THCCOOH may be more sensitive.

• Inclusion of a contingency management may decrease the agreement.

Abstract

Background

The agreement between self-reported cannabis abstinence with urine cannabinoid concentrations in a clinical trials setting is not well characterized. We assessed the agreement between various cannabinoid cutoffs and self-reported abstinence across three clinical trials, one including contingency management for abstinence.

Methods

Three cannabis cessation clinical trials where participants reported use and provided weekly urine samples for cannabis and creatinine concentration measurements were included. Bootstrapped data were assessed for agreement between self-reported 7+ day abstinence and urine cannabinoid tests using generalized linear mixed effects models for clustered binary outcomes. One study implemented contingency management for cannabis abstinence. Four hundred and seventy-three participants with 3787 valid urine specimens were included. Urine was analyzed for 11-nor-9-carboxy-Δ9-tetrahydrocannabinol and creatinine using immunoassay methods Biological cutoffs of 50, 100, and 200 ng/ml, as well as changes in CN normalized THCCOOH (25%/50% decrease), were assessed for agreement with self-reported abstinence during the three clinical trials.

Results

Agreement between measured THCCOOH and self-reported abstinence increases with increasing cutoff concentrations, while the agreement with self-reported non-abstinence decreases with increasing cutoff concentrations. Combining THCCOOH cutoffs with recent changes in CN-THCCOOH provides a better agreement in those self-reporting abstinence. Participants in the studies that received CM for abstinence had a lower agreement between self-reported abstinence and returned to use than those in studies that did not have a contingency management component.

Conclusion

Using combinations of biological measurements and self-reported abstinence, confirmation of study related abstinence may be verifiable earlier and with greater accuracy than relying on a single measurement.

Introduction

Cannabis is the most widely used illicit substance in many western countries and around the world (Vega et al., 2002). Cannabis use prevalence in the United States more than doubled from 4.1% to 9.5% between 2001/2 and 2012/13 (Hasin et al., 2015). In addition, approximately 12% of individuals who have used cannabis in the past year meet criteria for cannabis use disorder (CUD; Richter et al., 2017); however, measures of changes in cannabis use patterns over time are difficult to validate. Detection of positive cannabinoid concentrations in urine is subject to both physiological and pharmacological influences, including individual metabolism and hydration, as well as recency, frequency, concentration, and quantity of cannabis use. The primary psychoactive ingredient in cannabis, Δ⁹-tetrahydrocannabinol (THC), is highly lipophilic and stored in the adipose tissues in the body, with increasing concentrations at increasing use frequencies. THC then metabolizes to the non-psychoactive 11-nor-9-carboxy- Δ⁹-tetrahydrocannabinol (THCCOOH) and is excreted in the urine (Huestis and Smith, 2005).

Urine collection measures have a well-defined methodology and a well-characterized historic record of use in clinical trials. However, frequent cannabis users may submit positive urine samples for extended periods of time beyond initial abstinence. A pressing issue in cannabis cessation clinical trials is biological confirmation of new and continued abstinence during treatment. Pre-determined urinary THCCOOH cut-offs for biological confirmation of self-reported abstinence ignore the potentially lengthy detection times of urinary metabolites in frequent cannabis users (Dackis et al., 1982; Kelly and Jones, 1992). Upon abstinence, participants may have urine testing results that alternate between positive and negative over an extended period of time-based on a pre-determined cut-off. To accommodate this, some researchers have used higher THCCOOH cutoffs to determine biological evidence of abstinence (Levin et al., 2013), and Schwilke et al. (2011) developed a model to differentiate new cannabis use from residual THCCOOH excretion in a daily cannabis-using population using a cut-off derived from the previous urine measurement. Creatinine-normalized THCCOOH (CN-THCCOOH) values have also been used, as they have the advantage that two measures taken at separate visits are adjusted for differences that may occur due to hydration variability. The collection of cannabis use data through self-report during a clinical trial provides the added benefit of detailed use patterns as well as the ability to examine changes in frequency and intensity of use. However, the validity and accuracy of self-reported substance use amounts vary within a population and may be underreported in cases in which there is a reward for reduced use or abstinence (Carey, 1997; Del Boca and Noll, 2000; Williams and Nowatzki, 2005). In some cases, abstinence may be reported when the individual is in fact, not abstinent.

The primary goal of this analysis was to assess the biological correlates and their agreement with self-reported 7+ day abstinence in a cannabis cessation clinical trial setting with a population of participants with CUD. Specifically, we aimed to characterize the agreement between self-reported 7+ day abstinence with concurrent THCCOOH concentration cutoffs [widely used cut-off values (50 ng/ml (Substance Abuse and Mental Health Services Administration, 2004) and 100 ng/ml (Levin et al., 2013)) as well as higher values (200 ng/ml)] and recent changes in CN-THCCOOH. Additionally, we intend to examine the impact of contingency management implementation on the agreement between self-reported abstinence and biological measures.