Two recent studies from Yale Pathology highlight the challenges pathologists face when interpreting PD-L1 expression using the combined positive score (CPS) method of interpretation of immunohistochemistry (IHC) results for patients with gastric/gastroesophageal junction/esophageal cancers.
The studies, from Marie Robert, MD, professor of pathology and of medicine (digestive diseases), and David Rimm, MD, PhD, Anthony N. Brady Professor of Pathology and professor of medicine (medical oncology), appeared the May 2023 issue of Modern Pathology.
“Reliable, reproducible methods to interpret PD-L1 expression on tumor cells and immune cells are needed for pathologists to inform decisions associated with checkpoint inhibitor therapies,” Dr. Robert noted in the publication.
“Although the CPS method is prescribed for more indications than any other PD-L1 scoring system, its reproducibility has never been rigorously assessed,” Rimm shared in his publication.
The study led by Rimm collected a series of 108 gastric or gastroesophageal junction cancer cases, stained and scanned them, and then circulated them to 14 pathologists at 13 institutions for the assessment of interpretative concordance for the CPS system. While the pathologists showed better agreement at higher cutoffs (CPS >20), overall agreement for the < or ≥ 1 cutoff was only 30% independent of the level of experience of the pathologist. This study also showed no relationship between objectively measured mRNA and the CPS score.
“In summary, we showed that CPS shows high subjective variability among pathologist readers and is likely to perform poorly in the real-world setting,” the study noted. “This system may be the root cause of the poor specificity and relatively low predictive value of IHC companion diagnostic tests for PD-1 axis therapies that use the CPS system.”
The study by Robert and colleagues assessed agreement of PD-L1 expression using CPS among 12 pathologists under standardized conditions on samples from 100 patients with gastric/gastroesophageal junction adenocarcinoma both before and after online interactive CPS training. Interobserver agreement was also assessed for the three components that comprise CPS (total number of tumor cells, number of PD-L1 positive tumor cells and number of PD-L1 positive immune cells) in a subgroup of samples. Interobserver agreement was low to moderate both for CPS and for the individual elements comprising CPS. Importantly, agreement did not improve after training.
“The fact that two independent studies found similar results emphasizes the need for objective techniques to standardize the interpretation of PD-L1 expression using the CPS methodology to accurately identify patients most likely to benefit from immune checkpoint inhibitor therapy,” Dr. Robert said.