Cancer immunotherapy is transforming the treatment landscape for many types of cancer, but identifying those people most likely to benefit from these new treatments is one of the biggest hurdles facing the field. Personalized targeted therapies have traditionally been prescribed based on the presence of single mutations – but immunotherapy isn’t so simple.
The biology of each person’s cancer and immune system is different, and this variability can result in different responses to this class of medicine. For example, only about 20-40% of patients respond to checkpoint inhibitor immunotherapies.
Matching the Complexity of Cancer
The task of determining who will benefit from immunotherapy is not only important, but an enormous challenge. Today, one of the most common biomarkers for immunotherapy is a protein called PD-L1. Unfortunately, while patients with high PD-L1 are more likely to respond to immunotherapy, many with high PD-L1 don’t respond, and many with low PD-L1 do. It is becoming apparent that PD-L1 expression alone is not a sufficient predictor of response.
To make a more informed treatment decision and take the next step in personalized cancer immunotherapy, we need a comprehensive view of protein biomarkers, mutational load and genomic alterations. At this year’s AACR meeting, we are excited to present new results focused on that next step.
One exciting new biomarker for predicting response to cancer immunotherapy is tumor mutational burden (TMB). TMB provides a quantitative estimate of the total number of mutations in the coding region of a tumor’s genome. Tumor cells that have higher levels of TMB may be more recognizable by the immune system, and may therefore facilitate a stronger immune response to checkpoint inhibitors.
Indeed, our new results to be presented at AACR validate the ability of TMB to predict responses to FDA-approved checkpoint inhibitor cancer immunotherapies across multiple tumor types, including lung cancer, melanoma and bladder cancer.
However, given the complexity of cancer, even with the addition of TMB to current approaches, immunotherapy response cannot be predicted perfectly. To maximize benefit for individual patients, we need to identify and integrate additional predictive markers.
Finding New Ways to Predict Responses
Using comprehensive genomic profiling (CGP) approaches in people with non-small cell lung cancer (NSCLC), we may have identified additional markers, findings which we will also discuss at the AACR Bridging Big Data and Clinical Practice Session.
Specifically, we discovered that even in patients with low TMB, many lung cancers were highly responsive to immunotherapies if they also had mutations in the BRAF or MET genes. Evaluating these results in addition to TMB could help identify additional patients who may benefit from new immunotherapies.
Additionally, we discovered that in people with high TMB, those with STK11 alterations were much less likely to respond to immunotherapy. This suggests that these patients might benefit from another treatment approach instead. While these results will require clinical validation before being routinely integrated into patient care, together they may help us more accurately identify people who will or will not respond to immunotherapy, so physicians can make more informed and precise decisions for their patients. While our work at Foundation Medicine has focused on NSCLC, these STK11, BRAF and MET alterations could also affect immunotherapy response in other tumor types. This is an exciting area for ongoing research and discovery.
Perhaps equally notable is how these findings were discovered – through marrying sophisticated genomic data with real-world clinical outcomes. These “clinical-genomic” datasets, whether from new analyses of clinical trial data, or novel methodologies like the EMR-linked dataset Foundation Medicine has developed in partnership with Flatiron Health, have the potential to accelerate precision medicine and our ability to match the right therapy to the right person.
Doing Our Part
Despite the profound benefit cancer immunotherapy provides for some, there is still a need for better predictive biomarkers to classify both responders and non-responders. Clearly identifying patients who won’t respond to specific drugs is just as important as identifying those who will, saving the entire cancer ecosystem precious time and resources. Together, we can help ensure that patients receive a treatment that is likely to work – avoiding unnecessary side effects and wasted time with one that won’t. We can help the pharmaceutical industry better discriminate the impact of their drug. And we can help researchers develop new hypotheses and potentially discover new therapies.
Foundation Medicine is dedicated to seizing opportunities like these to advance biomarker development. Together with the rest of the community, we can push toward the future of precision medicine.