The Food and Drug Administration’s Psychopharmacologic Drugs Advisory Committee voted 12-7 on Jan. 12 to approve a subdermal depot formulation of buprenorphine, despite some reservations about who should be using – and implanting – the drug. To be marketed as Probuphine by Braeburn Pharmaceuticals, the long-lasting version of the partial mu opioid receptor agonist is meant for people with opioid addiction who have been stable on 8 mg/day or fewer of transmucosal buprenorphine therapy.
The pivotal phase III clinical study upon which the submission was based, PRO-814, enrolled subjects stable on 8 mg or less of transmucosal buprenorphine for at least 90 days. The double-blind, double-dummy trial randomized 87 patients to receive Probuphine and placebo transmucosal buprenorphine, and 90 patients to receive sham implants and active transmucosal buprenorphine. Participants received scheduled and random urine drug testing, and the study kept track of requirements for additional transmucosal buprenorphine.
Participants could have 2 months during the study period with a positive urine drug screen and/or self-reported illicit opioid use within the study period and still be considered responders; though data were tracked, there was no cap on additional buprenorphine requirements. Braeburn’s analysis of a modified intention-to-treat population showed that the Probuphine group met the primary endpoint of 20% noninferiority over the standard of care arm by a statistically significant margin (proportion difference, 0.088; 95% confidence interval, 0.009-0.167; P = .034).
The approval vote came after a long day of discussion that circled around three main areas of concern for the committee’s panelists. First, committee members struggled with the study’s definition of the eligible population, which consisted of people who had been on buprenorphine for at least 24 lifetime-weeks, with 3 months of negative urine drug testing before enrollment. Panelists expressed concern that a patient who was not fully stabilized on buprenorphine might be at grave risk of relapse during the week or so after implantation, when the drug is reaching steady-state.
The identification of eligible subjects informed another aspect of the study, the allowance of additional “rescue” transmucosal buprenorphine in both study arms. One public health aim of developing depot formulations of substances with abuse and diversion potential is to avoid having pills in the home, since these can be ingested by young children or be diverted. If people on Probuphine might need occasional rescue medication, this public health aim might be largely unmet. Committee members generally felt that some rescue medication soon after implantation would be acceptable but would consider people with ongoing rescue needs nonresponders to Probuphine.
Another set of concerns had to do with data analysis. Three subjects were very quickly lost to follow-up, and those patients were not included in the study sponsor’s analysis of their data. This methodology was roundly criticized by the study panel. However, when the FDA’s sensitivity analyses included these three missing patients as treatment failures, the study still met its predetermined 20% noninferiority endpoint.
Another point of contention within the data analysis was the sponsor’s imputation of a mix of positive and negative results to missing or incomplete urine toxicology screenings. Several clinicians in the room noted that they always assume a missing test is a positive test when working in addiction medicine. However, when the FDA imputed positive results to all the missing or urine toxicology data, Probuphine again met its primary noninferiority endpoint.
Some presenters felt that the sponsor’s inclusion of individuals with up to two positive urine drug tests during the 6-month study period was too lenient. Dr. Rachel Skeete, the FDA’s clinical reviewer, said, “We think that the analysis in which any positive months represent a treatment failure comes closer.”
However, Dr. Dawn F. Ionescu of the Depression Clinical and Research Program, Massachusetts General Hospital, Boston, argued against an expectation of perfection in this difficult population, pointing out that “If you think about other chronic conditions, our patients don’t come in with perfect blood pressure or perfect glucose.”
Finally, there were concerns about the implantation and removal of the subdermal implants. Many buprenorphine prescribers are not experienced proceduralists, several panelists noted. The recommended removal procedure requires suturing, a skill not kept current by all prescribers. Only four sites, two on each arm, are currently identified by the study sponsor. Though Braeburn plans to do pharmacokinetic studies to assess whether implantation sites can be reused, that’s not currently known. For some panelists, this was a big deal. Dr. Margaret Kotz, professor of psychiatry and anesthesiology at Case Western Reserve University, Cleveland, explaining her “no” vote, said, “What do you do after 2 years? That still is a huge question for me.”