Over the past few years, Bristol-Myers Squibb and Merck have brought to market two rival immunotherapy drugs, so-called checkpoint inhibitors, that have begun to change the way certain cancers are treated. But immunotherapy still has a long way to go, and Bristol provided the latest evidence Friday morning, when its drug, nivolumab (Opdivo), failed a late-stage trial in patients with previously untreated lung cancer. The drug didn’t lead to a statistically significant improvement in what’s known as progression-free survival; it didn’t do a better job than chemotherapy of keeping tumors from spreading.
The news sent ripples through the public markets, as Bristol-Myers (NASDSAQ: BMY) shares fell more than 16 percent in early trading, eliminating billions in market value. Rival Merck (NYSE: MRK), on the other hand surged more than 7 percent. Merck’s drug, pembrolizumab (Keytruda), is also being developed for lung cancer patients who have yet to be treated with chemotherapy.
“This is a major surprise, possibly the biggest clinical surprise of my career,” wrote ISI Evercore analyst Mark Schoenebaum in a note to investors on Friday.
Indeed, the failure of the trial, which compared nivolumab to platinum-based chemotherapy in non-small cell lung cancer patients, is a setback for Bristol, which has rode its immunotherapy franchise—led by nivolumab and an earlier immunotherapy drug, ipilimumab (Yervoy)—to significant success. Since March 2011, when ipilimumab was approved for melanoma, Bristol shares had tripled. Nivolumab has followed and picked up FDA approvals in skin, lung (following chemotherapy), and kidney cancers, as well one form of Hodgkin lymphoma. In the most recent quarter, nivolumab and ipiliumamb generated almost $1.1 billion in sales worldwide.
In doing so, nivolumab had separated itself from Merck’s pembrolizumab, which was second to market and banked $314 million worldwide in the most recent quarter. Still, both drugs have been significant not just as revenue producers, but in helping kickstart what’s been a huge investment in cancer immunotherapy. Nivolumab and pembrolizumab—which help the immune system identify tumors—have produced encouraging results so far, but still only work in a fraction of patients. They and other checkpoint inhibitors have become the backbone of a slew of combination trials, through which companies are trying to figure out how to make the non-responders respond, how to make responses last longer, and how to broaden the use of immunotherapy overall.
Many questions as to why checkpoint drugs don’t work more frequently remain, and one of them has to do with a certain biomarker. In the trial that failed today, Bristol enrolled patients whose tumors expressed at least some levels of a protein known as PD-L1. PD-L1 levels are seen as a way to predict response to checkpoint drugs like nivolumab and pembrolizumab, but it’s unclear how much is needed to generate a response.
“It’s somewhat of a quagmire,” Cancer Research Institute CEO Jill O’Donnell-Tormey told Xconomy in an interview in June. Although there seems to be a trend that more PD-L1 expression may lead to better responses, “some patients don’t show expression of PD-L1 but still respond, and vice versa.”
At the time, O’Donnell Tormey pointed specifically to how different companies have evaluated PD-L1 expression differently. Bristol, for example, enrolled patients whose tumors express at least 5 percent PD-L1. Merck enrolled patients whose tumors have at least 50 percent PD-L1 expression. The results have now come back for both companies, and Merck has reaped the rewards for its choice. Nivolumab failed, while Merck disclosed in June that its drug helped lung cancer patients with at least 50 PD-L1 expression live longer, and hold tumors in check longer, than those on chemotherapy in a Phase 3 trial.
Lung cancer is the second most common cause of cancer (behind skin cancer) and is the leading cause of cancer related deaths, according to the American Cancer Society. The ACS predicts 224,390 new cases and 158,080 deaths from the disease this year alone. Non-small cell lung cancer is the most common form of lung cancer, accounting for 80 to 85 percent of cases, according to the ACS.
Schoenebaum estimated in a note that the market for first-line non-small cell lung cancer is more than $12 billion, and that consensus estimates for nivolumab’s peak sales in first-line lung cancer were anywhere from $7 billion to $8 billion before today.
Bristol, it’s worth noting, still has another trial ongoing in first-line lung cancer, a study called Checkmate-227 testing nivolulmab in combination with ipilimumab. As Leerink Partners’ Seamus Fernandez wrote in a research note Friday morning, today’s Phase 3 failure “narrows the opportunity for front-line monotherapy substantially, [and] places a heavy reliance” on Checkmate-227.