In this edition of “How I will treat my next patient,” I highlight three studies presented at the World Conference on Lung Cancer (WCLC 2019) regarding the use of tissue biomarkers to predict benefit when immune checkpoint inhibitors (ICIs) are combined with chemotherapy in the treatment of stage IV non–small cell lung cancer (NSCLC) patients.
TMB in nonsquamous NSCLC
Marina C. Garassino, MD, and colleagues examined tumor mutation burden (TMB) to predict benefit from pembrolizumab in the KEYNOTE-189 study. KEYNOTE-189 was a double-blind comparison of first-line chemotherapy plus either pembrolizumab or placebo in 616 patients with stage IV nonsquamous NSCLC who were randomized 2:1 to the treatment arms.
Overall, adding pembrolizumab to pemetrexed and platinum significantly improved overall survival (hazard ratio, 0.49), progression-free survival (HR, 0.52), and overall response rate (47.6% vs. 18.9%). Benefit was observed in all analyzed subgroups, including patients with programmed death-ligand 1 (PD-L1) of less than 1%, 1%-49%, and 50% or greater.
In 293 patients – less than 50% of the total participants in the trial – with evaluable TMB data, TMB as a continuous variable showed no significant association with overall survival, progression-free survival, or overall response rate. There was no cut point at which TMB predicted outcome from treatment.
Similarly, Corey Langer, MD, and colleagues presented an exploratory analysis of the randomized, phase 2 KEYNOTE-021 trial (open label, pembrolizumab plus chemotherapy in 70 stage IV nonsquamous NSCLC patients). There was no association between tissue TMB and overall survival, progression-free survival, or overall response rate. In patients with tissue TMB greater than 175 mutations/exome and less than 175 mutations/exome, the overall response rate was 71% and 61%, respectively.
Both presenters recommended that tissue TMB not yet be used in therapeutic decision making.
How these results influence clinical practice
TMB has been associated with response to ICIs, but there is little information regarding whether TMB predicts for response to chemotherapy, either given alone or with ICIs. Logistical issues have limited the clinical utility of TMB. There are a variety of methodologies to measure TMB and no consensus on the ideal cut point for defining benefit from ICI therapy.
While TMB remains a marker of interest, the two presentations at WCLC 2019 demonstrate that additional research is needed to define whether TMB needs to be combined with other markers in an algorithm or matrix to guide decision making or whether we should focus entirely on identifying better biomarkers of immunogenicity.
PD-L1 expression and overall survival
Federico Cappuzzo, MD, and colleagues reported a subset analysis of IMpower131, a randomized, phase 3 trial of chemotherapy plus or minus atezolizumab as first-line therapy in 1,021 patients with stage IV squamous NSCLC. Patients were randomized to arm A (atezolizumab plus carboplatin plus paclitaxel), arm B (atezolizumab plus carboplatin plus nab-paclitaxel) or arm C (carboplatin plus nab-paclitaxel). Investigator-assessed progression-free survival, reported at the 2018 annual meeting of the American Society of Clinical Oncology, showed a small (about 21 days), but statistically significant, improvement in median progression-free survival in arm B versus arm C. The progression-free survival benefit was seen in all PD-L1-positive subgroups. At WCLC 2019, he reported the final overall survival results of arms B versus C.