What happened

After a surprising decision by Roche Holding (RHHBY 1.07%) to exit its co-development deal on a new cancer drug, NewLink Genetics Corporation (NLNK) shares were crashing 32% as of 1 p.m. EDT.

So what

Hope that success for NewLink Genetics' IDO-inhibiting cancer drugs could get its share price back on track, following a disappointing trial failure last year, were dashed today when Roche announced it was returning rights to co-develop GDC-0919.

A person hides their head underneath an open laptop that's flipped upside down.

IMAGE SOURCE: GETTY IMAGES.

The news is a bit surprising given that Roche just presented data from a trial of GDC-0919 at the high-profile American Society of Clinical Oncology (ASCO) annual meeting this week.

However, it becomes a little less surprising when we dig into the data and discover that the benefit of adding GDC-0919 to Roche's PD-L1 checkpoint inhibitor Tecentriq didn't move the needle much. The combination did improve response rates, but the improvement appears too slight for Roche to justify committing any more research to it. Specifically, the overall response rate in patients diagnosed with various solid tumor cancers was only 9%, and all responses were partial responses.

Further sealing GDC-0919's fate was that the response rate appears to trail that of Incyte's (INCY -0.52%) epacadostat, a drug that has the same IDO-inhibiting target as GDC-0919. At ASCO, Incyte reported that teaming epacadostat up with Merck & Co.'s (MRK -0.05%) PD-1 checkpoint inhibitor Keytruda delivered overall response rates of 30% to 35% in various cancers, including kidney cancer and non-small cell lung cancer.

Now what

Roche's decision caps off a week of disappointing news at NewLink Genetics. On Jun. 2, it reported that another of its IDO-inhibitors, indoximod, failed a mid-stage breast cancer trial, casting doubt on its use in indications outside of melanoma, where results have been solid.

In April, management said that combining indoximod with Keytruda resulted in an objective response rate of 52%, and a disease control rate of 73%. Those are good numbers, but they don't appear to be much different from epacadostat, which delivered a 58% overall response rate in advanced melanoma patients when used alongside Keytruda, and a 74% disease-control rate.

Overall, IDO-inhibitors appear to work, but their efficacy might not be as strong across all cancers as hoped; undeniably, that puts peak sales forecasts for this class of drugs at risk. Only time will tell what NewLink Genetics does now that it has 100% rights to both indoximod and GDC-0919, but it may be a bit premature to assume that the company's IDO-inhibitors won't eventually make it to market.