Improving Clinical Trial Design with Tumor Mutational Burden

FMI Social June Content TMB Header 4066x2033 R1

Biomarkers are playing a transformative role in cancer drug development. Not only can they help identify patients most likely to respond to a medicine, they can also make a huge difference in drug development by improving outcomes in clinical trials. Just as the right biomarker can positively influence drug response rates and clinical trial outcomes, the wrong biomarker can have profound negative impacts. The New England Journal of Medicine (NEJM) recently published data from a Phase III trial showing that an anti-PD-1 cancer immunotherapy (nivolumab) failed to meet its endpoint of improved progression-free survival (PFS) as first-line treatment for non-small cell lung cancer (NSCLC) in people that expressed the PD-L1 biomarker by immunohistochemistry (IHC).

TMB vs PD-L1

Despite its broad use in the cancer immunotherapy space, PD-L1 expression as measured by IHC has an inconsistent track record, likely due to its dynamic expression on both tumor and immune cells and the qualitative nature and lack of standardization across tests. A more reliable solution may exist in a measurement known as tumor mutational burden (TMB). We already know that TMB can help predict responses to FDA-approved checkpoint inhibitor cancer immunotherapies across multiple tumor types, including lung cancer, melanoma and bladder cancer.

It’s possible that TMB could have also had an important impact on the recent nivolumab study. Here, the authors chose to evaluate PFS in patients with at least 5% PD-L1 staining by IHC and found that those receiving nivolumab failed to show improved survival over chemotherapy (median PFS of 4.2 months vs. 5.9 months, respectively).

In a subgroup analysis performed after the trial outcome, even those with very high levels of PD-L1 expression (> 50%) were not associated with improved response to nivolumab. On the other hand, high TMB was significantly associated with both improved response rates to nivolumab over chemotherapy (47% vs. 28%, respectively) and improved PFS (median PFS of 9.7 months vs. 5.8 months, HR = 0.62).

Importantly, TMB correlated with survival across the nivolumab cohort, with TMB-high patients surviving without disease progression for an average of 9.7 months compared to 4.1 months for TMB-medium/low patients. This effect was specific to nivolumab, as TMB did not differentiate between survival within the chemotherapy cohort.

Other studies support this conclusion. For example, we recently presented data at the 2017 American Society of Clinical Oncology Annual Meeting from a real-world analysis of NSCLC patients receiving nivolumab in partnership with Flatiron Health. Results showed that higher TMB was associated with increased duration on therapy, a surrogate for PFS, as well as increased overall survival (OS). Similar to the recent study in NEJM, our findings also showed that PD-L1 expression levels were not associated with either improved PFS or OS.

The value of TMB

We believe that TMB could play a broader role in the future of cancer immunotherapy, and help influence the success of many types of clinical trials. As the oncology community continues to evaluate emerging treatments across an array of tumor types, it’s clear that improved biomarkers like TMB will play a pivotal role in advancing new and important medicines.

At the 2017 American Association of Cancer Research Annual Meeting, we presented results on TMB in over 88,000 clinical samples, where we confirmed that a relatively high proportion of advanced skin, lung, bladder and other cancers demonstrated high TMB. And in a recent study published in Genome Medicine, we analyzed over 100,000 human cancer genomes across hundreds of tumor types to more precisely map the distribution of TMB.

These large-scale studies are revealing that a percentage of some cancers for which immunotherapies are not typically considered – including brain, breast, colon and endometrial cancer – are actually TMB-high. This could mean that TMB has the ability to predict response to immunotherapy for nearly every type of cancer, and provide an important alternative to PD-L1 expression by IHC in many settings.

This inherent value is driving new partnerships with industry. In March, for example, we announced a collaboration with Bristol-Myers Squibb to evaluate TMB and MSI with FoundationOne® in patients receiving dozens of their investigational or approved immunotherapy drugs, in 35 types of cancer. Results from partnerships like these could help the pharmaceutical industry design more targeted and effective clinical trials, ultimately advancing the development of new, targeted therapies to treat cancer.

Although clinical trial setbacks can be disappointing for patients, we are committed to learning from each one, and improving the ways we bring the right medicines to the people who need them.