Sometimes it feels like biotech’s successes and failures come in waves—even when it involves entirely unrelated clinical programs. This past month has had many in industry feeling like they’re underwater.
Sometimes it feels like biotech’s successes and failures come in waves—even when it involves entirely unrelated clinical programs. This past month has had many in the industry feeling like they’re underwater.
But perhaps the biggest and broadest disappointment of the year is Incyte’s (Nasdaq: INCY) epacadostat. This IDO inhibitor was the toast of ASCO (the American Society of Clinical Oncology) last year, as its ability to boost the activity of checkpoint inhibitors seemed borne out by early clinical work.
Incyte boldly launched this drug into nine full-scale pivotal trials in combination with checkpoint drugs from Merck, Bristol-Myers Squibb, and AstraZeneca. But melanoma was supposed to be the low-hanging fruit. This, after all, was a disease in which checkpoint inhibitors have a strong early effect but tend to fade, and in which epacadostat showed strong response rates in early studies.
That the drug failed so definitively in melanoma makes the other eight ongoing trials look like serious longshots. It makes competing IDO inhibitors from NewLink Genetics and Bristol look like Hail Mary prayers.
But it goes even further than that.
Clinical trials fail for lots of reasons. They fail because they were designed to reflect a theory that was incorrect, or because they don’t actually do what the theory says they should. Sometimes they fail for reasons no one understands. Sometimes good drugs simply get tested they wrong way.
Sadly, it appears that epacadostat falls in the first category: It dutifully reduced kynurenine levels by about 90%, which is exactly what it was supposed to do. The theory was that tumors, by boosting endogenous IDO activity, turn tryptophan into kynurenine as a means of evading the immune system, and that blocking this process would make tumors more susceptible to immunotherapy.
Well…maybe not.
That’s unfortunate for patients and everyone trying to develop IDO inhibitors. But it is also a potential problem for the multitude—literally over a thousand—of ongoing combo therapy trials involving checkpoint inhibitors.
Work on IDO inhibitors has been going on for over a decade, based on what we thought was a fairly clear understanding of some of the pathways involved. So why do these drugs seem not to work? How were we wrong about the theory? It brings to mind the experience of combining PD-1 drugs with CTLA-4 drugs (e.g., Yervoy). Theory suggests this combination, working through largely independent mechanisms, should be complementary. And indeed, a combo of Keytruda and Yervoy is approved in melanoma. But combos of these drugs have been disappointing in other indications like non-small cell lung cancer and renal cell carcinoma. It suggests that we may not know as much as we think we do about the cancer immunity cycle.
Certainly that’s reason to be at least a little nervous about combos of checkpoint inhibitors and PARP inhibitors, checkpoint inhibitors and BRAF/MEK inhibitors, checkpoint inhibitors and CAR-T, checkpoint inhibitors and CD27 antibodies, etc., etc.
Of course, it’s far too early to say that checkpoint combinations of any sort won’t work, and I remain hopeful that some will. Many there’s even a glimmer of hope left for IDO inhibitors.
Of course, there’s another possibility. A 2017 paper in Nature Reviews Clinical Oncology that suggests that drugs that have little or no effect as monotherapies (like IDO inhibitors) are far less likely to succeed in combinations. That at least offers a little more hope for many of the other combo therapy trials underway.
On the other hand, it might send a chill down the spines of Halozyme investors.
