"As research on pharmacological treatments for CUD continues, a few key findings are of note. First, cannabinoid agonists (nabilone, dronabinol in combination with lofexidine, and lofexidine alone), were the only drugs that decreased drug-taking in a human laboratory model of relapse, supporting the notion that agonist replacement and attenuation of noradrenergic activity show promise for relapse prevention. Although dronabinol alone failed to clinically reduce cannabis use, a higher dose might have been more effective. Further, that study was designed to evaluate the initiation of abstinence; dronabinol or the more bioavailable agonist nabilone, might have greater utility in the prevention of relapse [25•]. These studies also support further testing of lofexidine in combination with other drugs, and generally illustrate the utility that can be gained from combining medications.
"Second, gabapentin and NAC were the only drugs tested in placebo-controlled clinical trials that decreased cannabis use (abstinence induction). Third, the ability of a drug to reduce cannabis withdrawal symptoms is not predictive of its ability to alter drug-taking behaviors (reduce use or prevent relapse). However, all the studies that reported positive changes in drug use also reported a reduction in withdrawal during early abstinence, suggesting that this feature is an important component of an efficacious medication."

Source

Rebecca E. Balter, Ziva D. Cooper, and Margaret Haney, "Novel Pharmacologic Approaches to Treating Cannabis Use Disorder," Current Addiction Reports, March 1, 2014, DOI 10.1007/s40429-014-0011-1.
http://link.springer.com...