Clinical trials of combination immunotherapy treatments are showing impressive overall survival rates.
Today, the only FDA-approved regimen for treating people who have pleural malignant mesothelioma is chemotherapy. But that might change soon. Impressive improvements in patients’ overall survival rates when treated with combination immunotherapy in clinical trials indicate that more effective treatment is on the way.
MAPS2, an academic Intergroupe Francophone de Cancérologie Thoracique-sponsored, non-comparative randomized phase 2 trial, evaluated the PD-1 inhibitor Opdivo (nivolumab) alone and in combination with the CTLA-4 inhibitor Yervoy (ipilimumab) in patients with mesothelioma who relapsed on prior treatment with Alimta (pemetrexed) and platinum chemotherapy.
Lead MAPS2 study author Gerard Zalcman, M.D., presented the results of the study during the 2017 European Society for Medical Oncology (ESMO) Congress. These results showed that the 1-year overall survival (OS) rate was 51 percent with Opdivo and 58 percent with the combination of Opdivo and Yervoy.
In an interview with CURE®, Zalcman, head of Thoracic Oncology Department, Hôpital Bichat-Claude Bernard, Université Paris-Diderot, reviewed the MAPS2 trial and discussed the promise of immunotherapy in patients with mesothelioma.
Zalcman: This was a trial assessing Opdivo, or Opdivo plus Yervoy, in patients with mesothelioma in the second- or third-line setting. It was a noncomparative randomized trial, so we are not allowed to compare the 2 arms directly. The median follow-up was 15 months.
The most remarkable data of this presentation are the OS data. This showed that, with Opdivo, the median OS is 13.6 months, which is amazing in this setting and these patients. The median OS of the Opdivo plus Yervoy arm has not yet been reached after 15 months of follow-up, suggesting that it could be over 15 months.
This trial ran very fast because we accrued 125 patients within 5 months, and the last patient was accrued in September 2016. There were no specific signals to worry about in this toxicity profile. There were slightly more immunological side effects in the combination arm, but nothing that was statistically significant. There were 3 toxic deaths in the combination arm, which occurred very early in the study course, and with no more grade 5 events after. This suggests that the investigators have learned how to manage such immunologic toxicities.
We also presented data about PD-L1 immunohistochemistry (IHC) expression, showing that PD-L1 expression as defined by more than 1 percent of tumors cell stain with the anti—PD-L1 antibody was associated with overall response. When we selected the stronger expressers — meaning more than 25 percent of tumor cells — this was associated with overall response and disease control rates.
For OS, it is more complex. In the Opdivo arm, PD-L1—positive patients did better than the PD-L1–negative group. However, in the combination arm, the PD-L1 IHC had no influence on OS. Therefore, PD-L1 as a biomarker could be a good marker for Opdivo, but is not a good biomarker for the combination.
Currently, first-line mesothelioma treatment is based on the backbone of pemetrexed/cisplatin chemotherapy plus or minus bevacizumab. The median OS ranges from 15 to 19 months and there is currently no recognized second-line treatment for this disease. Therefore, the positivity of this MAPS2 trial supports the use of immunotherapy as second- and third-line therapy. This has actually been written as an option by the National Comprehensive Care Network (NCCN) panel. The company developing these drugs is going to seek a breakthrough application with the FDA.
There is a large phase 3 trial running in the first-line setting comparing the combination of immunotherapy with standard chemotherapy. In the near future, we will use chemotherapy in the second- and third-line settings. Moreover, if the phase 3 trial is positive, it could go to the first-line setting with double immunotherapy of Opdivo plus Yervoy.
Currently we don't know what could be a good biomarker for mesothelioma. One reason is that this is a tumor where the mutational tumor burden is very low. In the famous schematic representation that has been published in Nature, mesothelioma stands at the left of the scheme, meaning that there are few mutations. Therefore, something like mutational burden will not be a good biomarker. We will try to test, but I don't think we are expecting it to be a good biomarker.
We see that PD-L1, at least with the antibody we tested, might be difficult to go to clinical practice with since it is no use for the combination therapy. We will have to study quantitatively and qualitatively the density of lymphocyte infiltration in the tumor. We have such data, we have done the study, but it is not statistically analyzed yet. We will have to see if the density of lymphocytic infiltration could add something as a biomarker to predict response. The good biomarker will ultimately be the one that predicts survival, not just response — because it is not always correlated.
These trials are clearly an advance for the treatment of patients, provided that these drugs are accepted by regulatory agencies. This is a rare disease, so the development of drugs in rare diseases is different. The dose that will be submitted to the FDA will have these data for supporting registration. I believe that authorities are ready to accept such a registration with such data, and I am pretty sure these drugs will soon be available — at least in the United States. The situation could be more difficult in Europe where we may be more reluctant to register these drugs. The patients and advocacy groups will have an important role to play to support this registration.