Despite all the great progress made harnessing the immune system to fight cancer, there is still much work to be done to maximize its potential. The latest evidence: disappointing results today from a roughly 1,100-patient study called “Mystic,” a highly anticipated trial from AstraZeneca testing a combination of immunotherapies in newly diagnosed lung cancer patients.
AstraZeneca disclosed this morning that a combination of two of its immunotherapy drugs, durvalumab (Imfinzi) and tremelimumab, failed to best chemotherapy at keeping patients’ lung cancers from spreading, one of the main goals of the Mystic trial.
It’s important to note that AstraZeneca (NYSE: AZN) doesn’t yet know whether the drug combination will ultimately help these patients live longer. Those results are expected next year, and offer at least a chance that the study could still succeed in the long run. But the data are nonetheless a big blow to AstraZeneca, a worrisome sign for one of the immunotherapy leaders, Bristol-Myers Squibb (NYSE: BMY), a boost for rival Merck (NYSE: MRK), and fresh evidence of the many unknowns that remain in combining immunotherapy drugs with one another.
For AstraZeneca, the Mystic results were so devastating that shares fell more than 15 percent in pre-market trading. AstraZeneca is a latecomer in cancer immunotherapy, which has begun to change the way multiple cancers are treated. It just won approval of durvalumab in bladder cancer in May, well after rival immunotherapy drugs from Bristol-Myers (nivolumab (Opdivo)), Merck (pembrolizumab (Keytruda)), and others hit the market. But a successful durvalumab/tremelimumab combination would’ve given AstraZeneca a chance at billions in revenue and an opportunity to go toe to toe with Merck. Lung cancer is the second most common cancer (behind skin cancer) and is the leading cause of cancer related deaths, according to the American Cancer Society. Only Merck’s pembrolizumab is approved for newly diagnosed lung cancer patients. Nivolumab is approved for those who have failed prior treatment.
Upon the study’s failure, AstraZeneca announced a deal to hand partial rights to two other cancer drugs, olaparib (Lynparza) and the experimental treatment selumetinib, to Merck, and then test the drugs in combination with other therapies including pembrolizumab in multiple cancers. In a research note, Leerink Partners analyst Seamus Fernandez called the agreement an “admission of defeat” in the cancer immunotherapy race. Merck is paying AstraZeneca $1.6 billion up front, and possibly $8.5 billion overall, in the deal.
The Mystic results also spell potential trouble for Bristol, which is testing a similar combination of cancer immunotherapy drugs in a trial of lung cancer patients called Checkmate-227. The results, expected next year, are immensely important for Bristol. A previous study of nivolumab alone in newly diagnosed lung cancer patients failed, and Merck capitalized when pembrolizumab succeeded in a subsequent study. Checkmate-227 is Bristol’s chance to bounce back. While there are plenty of caveats in comparing one trial to another, expectations have just been ratcheted down. The Mystic data “sharply reduces” the combination’s chances of success, Fernandez wrote, for example. No surprise, then, that Bristol shares fell more than 6 percent Wednesday morning. (Merck, meanwhile, was today’s big winner, with shares climbing 5 percent.)
More broadly, however, the study is the latest reality check for immunotherapy. Immunotherapy drugs are now approved for several cancers, and in some cases produce stunning, long-lasting results for patients who might otherwise be at death’s door. But the majority of patients don’t respond to treatment, and researchers still don’t understand why. Merck’s drug succeeded in lung cancer, for example, while Bristol’s failed. Earlier this week, pembrolizumab failed in head and neck cancers, while nivolumab previously succeeded. Yet it’s unclear if there is much difference between the two drugs. In the meantime, drug companies are testing a slew of combinations hoping to boost results.
The bottom line is more biomarkers—molecular and genetic signatures that may predict a drug response—are needed to help improve the odds of success. Expression of a protein called PD-L1, for example, is seen as one of the few known potential predictors of response to immunotherapy. But the relationship between PD-L1 expression and response is unclear. And given the Mystic study enrolled patients whose tumors express PD-L1 on 25 percent or more of their cancer cells, the results are sure to raise more questions about that relationship going forward.