Daniel V.T. Catenacci, MD, of the University of Chicago Medical Center and Biological Sciences, emphasized the value of the patient having microsatellite instability–high (MSI-H) status as a biomarker for immunotherapy. The analysis by Chao et al “highlights how well patients with MSI-H tumors do, compared to microsatellite-stable patients, and how much better they do in a randomized setting, being exposed to immunotherapy as compared to standard-of-care chemotherapy…. The data also show that this is a poor-prognosis subset of the disease, with the disease progressing relatively more rapidly with standard-of-care treatment compared to [microsatellite-stable (MSS)] tumors.”

Daniel V.T. Catenacci, MD

Combined Positive Score

The combined positive score (CPS) uses the proprietary 22C3 antibody and is the only such assay approved to quantify and score [programmed cell death ligand 1 (PD-L1)] staining in the United States. It stains not only for tumor cells but also for infiltrating immune cells over the total number of viable cells, he explained.

For many studies, a CPS ≥ 1 has been considered “positive” for expression, which occurs in approximately 50% to 60% of patients. Based on the available evidence, CPS < 1 tumors are not likely to benefit from monotherapy immunotherapy checkpoint blockade compared to standard chemotherapy. “This is most readily seen in this subgroup of patients who had worse survival compared to the paclitaxel control in the randomized phase III KEYNOTE-061 study. A CPS < 1, therefore, is a good negative predictive biomarker, and these patients should not be exposed to monotherapy with a checkpoint inhibitor,” he said.

Moving Forward

“By increasing the level of CPS defining positivity to ≥ 10, approximately 15% to 20% of patients would be considered positive and qualify for immunotherapy at that cutoff,” he noted. “By this enrichment, [the data] are going in the right direction, and this looks to be predictive at least in this study but not others, including ­KEYNOTE-062 and KEYNOTE-181, where there is a subgroup that seems to have a detriment,” he concluded. “What would it look like if you made the cutoff higher? Would the curve look even more exceptional and avoid harm to this subset of patients?”

He added, “We have to continue to strive to find the patients who will not benefit from monotherapy and continue to strive to find the patients who will benefit based on subgroup analyses. We also have to note that more than 90% of MSI-H tumors are PD-L1–positive at the CPS 1 cutoff, whereas two-thirds are positive at the CPS 10 cutoff,” he said. “Therefore, since we recognize that patients with MSI-H status do very well and that there is this overlap between the MSI-H and PD-L1 CPS biomarkers, when searching for other predictive biomarkers, we should exclude the MSI-H subgroup from these analyses to isolate any added benefits that may or may not exist—that is, assess PD-L1 CPS > 10 (or another higher cutoff) in the MSS subgroup to limit confounding.”

Dr. Catenacci also had advice for clinicians tempted to “give every patient at least a chance to get immunotherapy” because they “don’t want to miss the chance of having an exceptional responder.” For the vast majority of patients, outcomes are disappointing for a number of reasons, he pointed out, and the chance of serendipitously finding that rare patient, after all is said and done, is small. “For example, if you include the countless numbers of patients who are not fit enough to be able to wait to enroll on these studies, the  ones with heavy burden of disease and/or rapidly progressing disease—all factors unlikely to derive benefit from immunotherapy—the rate of a good long-term outcome in this disease as a whole is very, very low, particularly after excluding those with MSI-H tumors,” he said. 

DISCLOSURE: Dr. Catenacci has received honoraria from, consulted or advised for, or served on the speakers bureaus of Amgen, Astellas Pharma, Bristol-Myers Squibb, Five Prime Therapeutics, Foundation Medicine, Genentech/Roche, Genmab, Gritstone Oncology, Guardant Health, Lilly, Merck, NantOmics, OncoPlex Diagnostics, Taiho Pharmaceutical, and Merck.