Blog October 17, 2017

ESMO 2017 — Part 2: Building on Biomarkers

The news from ESMO in September once again has highlighted the utility and need for biomarkers to help determine not only who will likely benefit from a particular immuno-oncology (IO) treatment, but also who may require a combination of such therapies right from the start.


While higher levels of PD-L1 have generally correlated with improved response rates, there have been exceptions. Some patients with high PD-L1 remain resistant to treatment with PD-1/PD-L1 therapies while some patients with low PD-L1 levels have responded well. Moreover, there is no standard assay for PD-L1 (although standardization studies are being run), making it difficult to compare results across clinical trials. So clearly, better assays — and new validated biomarkers — are needed. Two such new biomarkers caught the ESMO meeting spotlight: LAG3 and tumor mutational burden.


LAG-3 is a protein in the tumor microenvironment that boosts T cell activity against tumors and inhibits other T cells that can suppress such anticancer activity. Bristol-Myers Squibb presented data showing that for LAG-3+ tumor cells, the response rate to its PD-1 therapy, Opdivo, triples. The company is developing an anti-LAG-3 agent, relatimab, which it is testing in combination with Opdivo. In a Phase 1/2a study in melanoma patients who had not responded or became resistant to immunotherapies, 18% of LAG-3+ patients responded to the combination of Opdivo and relatimab as compared to 5% of patients where LAG3 expression was less than 1%.


At ESMO, considerable attention was paid to a similar broad-based biomarker of tumor genetic instability, tumor mutational burden (TMB). TMB is a measure of the number of mutations within a tumor genome, which investigators hope will predict which patients benefit most from IO therapy. Roche and Bristol-Myers Squibb, both in partnership with Foundation Medicine, have made TMB a focus of efforts to better understand which patients might benefit from which therapy or therapeutic combination.


Foundation Medicine has previously shown that measuring TMB in tissue biopsy samples could reliably predict patient response to treatment.  At ESMO, Foundation Medicine and Roche AG presented data on the measurement of TMB in blood, which if validated, could be especially valuable for assessing patients where obtaining a standard tissue biopsy is not feasible, due to the patient’s age or state of disease progression.


The companies looked at the POPLAR and OAK studies of Tecentriq (atezolizumab) in second line lung cancer. In the POPLAR trial, they filtered the trial population by different TMB scores to establish threshold associations of those scores with improved progression-free survival (PFS) and overall survival (OS). Foundation Medicine then applied the same endpoints in a blinded manner to the OAK data to see whether stratifying patients by TMB score could predict improved responses. The data showed a high correlation between high TMB scores and longer PFS, while preserving the OS survival benefit. The two companies are now prospectively validating blood-based TMB testing in two clinical studies, using the measurement to select either Tecentriq or Alecensa (Alectinib) as a first-line cancer treatment.


While improvements in blood-based assays are needed, especially with regard to sensitivity, new biomarkers such as LAG3 and TMB could greatly expand the use of IO to some cancers, like breast, which have historically seen low response to such therapies. Moreover, they might help indicate which patients could be treated with a single IO therapeutic (e.g., based on both high TMB and high PD-L1) and which might best benefit from a combination approach from the start.