Approval Withdrawn

Two of my recent post topics (on antibody-drug conjugates and on accelerated approval) have intersected today. In the former, I mentioned in passing the ADCs that are conjugated to a very cytotoxic anticancer agent called monomethyl auristatin E, and there are several. A related compound (monomethyl auristatin F) has also been used in a recent ADC from GlaxoSmithKline, Blenrep. In that one, the antibody is directed against BCMA (B-cell maturation antigen), which is a cell-surface receptor protein that’s primarily found on B-cell lymphocytes. As you might guess from that lineage, you might well want such an ADC for various types of leukemia, and indeed, Blenrep seemed to show efficacy a couple of years ago in a tough-to-treat population, patients with relapsed or refractory multiple melanoma.

That was a good sign, because getting single-agent efficacy in patients who have already exhausted their chemotherapy options is not at all easy. Blenrep was given a conditional approval by the FDA a few months later as a fifth-line therapy for MM, and (see that second post linked above) this was tied to the results holding up in a confirmatory trial. In this case, the trial was already going, as the FDA wants to see from now on in general, but unfortunately those results came out a couple of weeks ago and they were not good. The drug missed its primary endpoint in this larger trial, progression-free survival. If you read over that press release and don’t think too hard about it, you could tell yourself that Blenrep was an improvement, because things like duration-of-response were better with it than with the standard of care (pomalidomide and dexamethasone). Median progression-free survival was actually better with Blenrep, but the main way to square those numbers would be if patients deteriorated more quickly after that in the Blenrep treatment group. Overall survival was identical, and the overall response rate (tumor load/shrinkage) looks basically identical as well. 

And today the FDA did rescind that conditional approval. GSK is not contesting the decision, but is continuing with their existing trials of Blenrep (in comparison to other chemotherapy options, for example) hoping to see a niche benefit that will stand out somewhere. This is exactly what should happen: conditional approval should be conditional, and if everyone has agreed on the way to test the drug going forward, then everyone should abide by that plan. It’s also worth noting that since this conditional approval the FDA has approved three other therapies for refractory multiple myeloma, so honestly, it’s not like this latest decision is going to leave a huge gap in the treatment landscape.

But this should all serve as yet another round of proof for some fundamental truths in drug discovery. One, solid biochemical rationales don’t always lead to the results you’d want in human patients. Two, promising data in Phase II is only that – promising, until you confirm it in another (larger) trial. Remember always the story of Jonas Salk, the morning that it looked like a polio vaccine candidate had finally shown a protective effect in an ex vivo blood assay and everyone in the lab was celebrating: “OK people, hold on, let’s see if we can do it again“. All too often, the rule in clinical trials is “Einmal ist keinmal“, literally “one time is no time”. I think that accelerated approval is still a good idea in such cases, especially for deadly diseases, but you have to be prepared for the chance – the real chance – that you’re going to be treating people with something that’s not as effective as you thought it was.