Blog August 16, 2016

Running the Risks of High Stakes Drug Development

On August 5, Bristol-Myers Squibb (BMS) surprised many with the announcement that their PD-L1 inhibitor Opdivo® (nivolumab) had failed a Phase 3 trial assessing its use as a monotherapy in the first-line treatment of non-small cell lung cancer (NSCLC). This is a disease where the drug is already approved and used in patients who have failed prior chemotherapy, competing in that space with Merck’s Keytruda® (pembrolizumab). NSCLC accounts for approximately 80% of lung cancer cases, and analysts had projected Opdivo sales of $12 billion by 2021, with $7-8 billion of that amount coming from first-line uses of the drug. The news of the trial failure sent BMS shares down by 18%, while Merck shares gained 10% on the news. A few days later, Merck achieved an additional win for Keytruda, which received FDA approval as a treatment for head and neck cancer in patients with advanced disease who had failed prior platinum-based chemotherapy.


The trial could be seen as a high stakes gamble by BMS. In an effort to have a treatment that addressed the widest possible patient population in NSCLC without the need for a companion diagnostic, the company had enrolled 541 previously untreated patients whose tumors expressed PD-L1 at a level of 5% or greater. In contrast, Merck — who had announced positive results with Keytruda in the indication in June — had used PD-L1 levels of at least 50% as its cut-off for patient enrollment.  Their success gives them exclusive access for now to the 25% of patients who express high levels of the marker, and marks another success in favor of their personalized medicine approach.


BMS CEO Giovanni Caforio told Matt Herper of Forbes:


“Our strategy in lung cancer from the very beginning has been to answer a number of questions in order to really understand how to treat every patient with lung cancer regardless of histology or PD-1 expression or line of therapy…And the trial tells us that the role of monotherapy is likely to be limited to a very small subset of patients that express PD-L1 at very high levels, which is what we believe we know from a previous study that was reported not a long time ago.”


The failed trial results emphasize that I/O combinations are likely to be needed in all but the small group of patients who express high levels of PD-L1.  And indeed, BMS has an ongoing Phase 3 study exploring the potential of the combination of Opdivo plus Yervoy for PDL-1 positive patients, and Opdivo plus Yervoy, or Opdivo plus chemotherapy in PD-L1 negative patients.


Many more studies will clearly be needed throughout the I/O space before there is a good understanding of how these agents are best used. But BMS’ results also vividly illustrate the high cost of drug development. Risky trials like this one cost a drug developer millions of dollars with no immediate reward. But they add significantly to the overall body of knowledge that helps physicians and patients better fight disease.