Relationship Between Short-Term Outcomes and PD-L1 Expression Based on Combined Positive Score and Tumor Proportion Score in Recurrent or Metastatic Head and Neck Cancers Treated With Anti-PD-1 Antibody Monotherapy

Background

PD-L1 expression in tumors and immune cells is a biomarker for the efficacy of anti-PD-1 antibody (APA) therapy across diverse cancers. Based on the results from the KEYNOTE-048 trial, pembrolizumab monotherapy is indicated for platinum-sensitive recurrent/metastatic head and neck squamous cell carcinoma (R/M-HNSCC) with a positive combined positive score (CPS). Conversely, nivolumab is utilized for platinum-pretreated R/M-HNSCC regardless of the positive tumor proportion score (TPS) following the results of the CheckMate-141; however, its subgroup analysis indicated that TPS-positive population tended to have a relatively high overall response rate and progression-free survival (PFS). Although, the superior PD-L1 evaluation method for predicting APA therapy efficacy in R/M-HNSCC and the appropriate cut-off value remain undetermined. This study aims to elucidate the relationship between short-term outcomes and PD-L1 expression based on CPS and TPS in R/M-HNSCC patients undergoing APA monotherapy.

Methods

R/M-HNSCC patients receiving APA monotherapy from 2018 to 2021 with available samples were enrolled. An experienced pathologist evaluated CPS and TPS utilizing the PD-L1 IHC 22C3 pharmDx assay. Short-term outcomes were assessed by clinical benefit rate (CBR), objective response rate (ORR), and PFS.

Results

Fifty-three R/M-HNSCC patients received APA monotherapy. Forty-seven had CPS ≥ 1, and 44 had TPS ≥ 1%. By receiver-operating characteristic curve analysis, the CPS cut-off value for predicting better CBR was determined to be 50. The ORR/CBR tended to be higher when CPS was positive. Although differences in PFS were not observed for a cut-off value of 1 or 20, they were observed for 50 (3.2 vs. 8.4 months; hazard ratio 0.44, p = 0.02). ORR and CBR were respectively 12.5% and 12.5% in the TPS < 1% group and 33.3% and 48.9% in the ≥ 1% group. The TPS < 1% group showed significantly poorer PFS (1.9 vs. 4.5 months, hazard ratio 0.40, p = 0.01).

Conclusion

The short-term efficacy of APA monotherapy in R/M-HNSCC patients tended to be better when CPS was positive. TPS helps predict the population that does not benefit from APA monotherapy.