Alkermes has announced the failure of two of its pivotal trials for its novel opioid-targeting antidepressant ALKS 5461 (buprenorphine/samidorphan), yet is emboldened by an efficacy signal in post-hoc analyses and is committing to expanding a third ongoing trial. While antidepressant trials are notoriously difficult to control for a widely variable placebo effect, Alkermes is unlikely to salvage any credibility for its product.
First two pivotal trials, FORWARD-3 and -4, both miss primary endpoints
The two completed trials – FORWARD-3 and FORWARD-4 – had similar designs, in which major depressive disorder patients were randomized to receive either ALKS 5461 or placebo as an adjunct to standard antidepressant therapy over a 12-week treatment period. The FORWARD-3 trial, which investigated a single dose of ALKS 5461 (2mg buprenorphine/2mg samidorphan), failed outright with no measurable treatment effect compared to placebo. FORWARD-4 showed a trend toward improvement of depression symptoms as measured with the Montgomery-Asberg Depression Rating Scale (MADRS) with the higher 2mg/2mg dose, although there was no effect at the 0.5mg/0.5mg dose. Alkermes also ambiguously claimed that post-hoc analyses reached statistical significance for the entire 2mg/2mg dose group compared to placebo on MADRS, potentially pointing toward FORWARD’s sequential parallel comparison design (SPCD), in which non-responders to placebo were re-randomized to one of the three treatment arms midway through the trial.
These data are in stark contrast to Alkermes’s Phase II trial, which showed a highly impressive seven-point reduction in MADRS scores for patients receiving ALKS 5461 compared to placebo after four weeks, with the treatment difference being significant after just one week. This trial also employed an SPCD design, which is hypothesized to reduce the placebo effect that often confounds studies involving antidepressants. No data are yet published for the larger FORWARD studies, although it is probable that a larger placebo effect was exhibited compared to the prior proof-of-concept study – indeed Alkermes noted a stronger placebo response in FORWARD-4 than in FORWARD-3.
Trials of effective antidepressants can and do fail, although Alkermes is unlikely to salvage ALKS 5461
These data are highly damaging for Alkermes and its prospects of bringing ALKS 5461 to market. Nevertheless, the company is still attempting to generate enough positive data to file for regulatory approval, expanding the size of the currently ongoing FORWARD-5 trial – its third and final pivotal efficacy study. There is a precedent for proven, effective antidepressants to fail clinical trials and still receive regulatory approval. Most recently, Allergan’s Fetzima (levomilnacipran) only showed significant efficacy in three of its four registration studies, while Lundbeck and Takeda had a range of positive and negative trials for Brintellix (vortioxetine) across a range of doses. However, these are both drugs with known antidepressant mechanisms. ALKS 5461 targets the kappa opioid receptor, which is more commonly known for its role in analgesia and addiction pathways.
As such, Datamonitor Healthcare believes that Alkermes has a very slim likelihood of being able to salvage ALKS 5461. Even if FORWARD-3 and FORWARD-4 were somehow flawed and compromised by a high placebo response, and FORWARD-5 generates antidepressant data more akin to the prior Phase II findings, it will be very difficult to argue for regulatory approval with one positive trial. It will be impossible for regulators to dismiss the data from FORWARD-3 and FORWARD-4, unless Alkermes decides to investigate higher doses. This is also unlikely, as the company had previously looked at the 8mg/8mg dose during Phase II and instead decided to progress ALKS 5461 at a much lower range.
Alkermes’s next post-ALKS 5461 steps will define its ambition and identity
Earlier in January, at the annual JP Morgan Healthcare Conference, Alkermes’s chief executive officer Richard Pops had talked optimistically about the potential to launch ALKS 5461 by the end of 2016. One week later, the same investors he was presenting to had largely wiped ALKS 5461 from their valuation models, as Alkermes’s share price fell by 44%. This shows the potential importance that the innovative antidepressant could have had on the company’s fortunes. Datamonitor Healthcare had previously forecast the drug to achieve $510m in annual US sales for the treatment of depression by 2021.
ALKS 5461 marked a break from Alkermes’s traditional business model, which was first built on its drug delivery technology. The company receives royalties from Johnson & Johnson on sales of the long-acting injectable (LAI) antipsychotics Risperdal Consta (risperidone) and Invega Sustenna (paliperidone palmitate), as well as from AstraZeneca’s Bydureon (exenatide) injection. Alkermes subsequently started developing and marketing its own LAI products, but targeting molecules and indications for which there is significant proof-of-concept, such as Vivitrol (naltrexone) for addiction and Aristada (aripiprazole lauroxil) for schizophrenia.
Alkermes must now be carefully pondering its next steps, as these will define the company’s ambition and identity. The next most advanced products in its pipeline are typical for Alkermes. ALKS 8700 is a monomethyl fumarate formulation in development for multiple sclerosis that is unashamedly targeting the multi-billion dollar market that Biogen has created for Tecfidera (dimethyl fumarate). ALKS 3831 is a co-formulation of olanzapine and samidorphan that is attempting to retain olanzapine’s best-in-market antipsychotic efficacy while reducing its debilitating metabolic side-effect profile. Beyond this, Alkermes may well seek to bolster its pipeline, or potentially its marketed portfolio, through acquisitions in the central nervous system space in which it is positioned.