One of the most pressing issues in immuno-oncology is predicting which treatments will be most effective for which patients. While some individuals respond in dramatic ways to current immunotherapies, others show little, if any, improvement after receiving the treatment. Many companies have tried to identify genes and biomarkers that could help them figure out how effective treatments might be. While progress has been slow, a few notable achievements exist. For example, researchers several years ago noticed that there is a correlation between response to the PD-1 checkpoint inhibitor pembrolizumab and cancer-related mutations in the genome.
More recently, Roche and Foundation Medicine introduced a blood test that examines tumor mutational burden (TMB) and could help predict patient responses to immuno-oncological treatments. The TMB is the total number of mutations within the cancer genome. This test is significantly less invasive than the traditional tissue biopsies presently used to figure out mutation rates. In theory, researchers could also use the test to look at particular biomarkers that help formulate a prognosis. Roche is particularly interested in biomarker research as a way of matching therapies to people most likely to benefit from them and sees the new blood test as a step toward personalized cancer care.
High levels of TMB have been linked to a better response from immune checkpoint inhibitors across several different types of cancer, including bladder and lung cancer, as well as melanoma. In fact, researchers have found that TMB is a better biomarker for predicting response to anti-PD-L1 treatments than testing for levels of PD-L1.
Research in Support of Roche’s New Blood Test
When Roche unveiled its new blood test, the company presented data from the Phase II POPLAR and Phase III OAK studies, both of which were looking into the efficacy and safety of atezolizumab, a PD-L1 blocker. The investigators looked at plasma samples from 794 patients in these trials as a retrospective analysis of association between TMB and atezolizumab response. In these two trials, the drug manufacturer found that patients with non-small cell lung cancer and high TMB had longer progression-free survival times when they received atezolizumab than patients with lower TMB.
Because retrospective studies are prone to bias, Roche is currently undertaking two new studies: one that is semi-prospective, and another that is wholly prospective with patients who are receiving atezolizumab and/or alectinib for non-small cell lung cancer. These prospective studies will provide better proof of concept for the TMB blood assay. In the B-F1RST study, blood samples of patients collected prospectively will be tested retrospectively to establish a clearer link between TMB and drug response. The BFAST study will test patients for a high TMB score and oncogenic somatic mutations and assign patients to receive atezolizumab or alectinib, respectively.
The Need for Better Predictive Measures
While Roche is taking a step in the right direction, the larger immuno-oncology industry has known for years that it needs more information on biomarkers in cancer treatment in general, but especially with cancer immunotherapy. In addition to assessing the likelihood of tumors responding to a particular treatment, biomarkers also have the potential to predict adverse effects in certain patients, which would direct clinicians to use different modalities or create a management plan before the effects arise.
Biomarkers could also be useful in tracking actual response to the treatment. Conventional methods for judging tumor reactions include the Response Evaluation Criteria in Solid Tumors (RECIST) and World Health Organization (WHO) imaging standards, but these methods largely rely on tumor size. Immuno-oncology treatments have the ability to prolong life without actually reducing the size of a tumor. Furthermore, these treatments can actually cause tumors to get bigger due to lymphocyte infiltration. Post-treatment biomarker strategies could provide a much more helpful means of judging treatment response.
The Struggle to Identify Meaningful Biomarkers
The main difficulty in identifying good biomarkers is the fact that immunotherapy involves a complex set of biomarkers, not a single, easily identifiable one. To move forward, researchers need to look at protein biomarkers, immune cell signatures, mutational changes, and more. All of these factors contribute to the uniqueness of each individual’s immune system. The complexity of the immune system provides some explanation as to why so few patients respond to immunotherapies and why some of those who do experience an incredible effect.
Measuring protein biomarkers like PD-L1 has proven largely inconsistent because of dynamic expression on tumors and immune cells alike. For this reason, researchers have been more interested in more complex, quantitative approaches, such as TMB and microsatellite instability (MSI), both of which look at genomic changes in cancer. While TMB looks at total mutations, MSI focuses on short, repeated DNA sequences that are attributed to defective DNA repair mechanisms. Patients with colorectal cancer who have a high MSI score were found to be more likely to respond to certain checkpoint inhibitors.
The potential of immuno-oncology to radically change the face of cancer treatment is clear, but better and more biomarkers are necessary to make the transition to personalized medicine.