The Tumor Area Positivity score demonstrated substantial agreement with the Combined Positive Score in assessing programmed death-ligand 1 expression and predicting survival outcomes in patients with gastric and esophageal cancers treated with tislelizumab, according to a recent study.
A pooled analysis of three randomized, phase 3 trials – RATIONALE-305, RATIONALE-302, and RATIONALE-306 – examined the concordance between two immunohistochemical methods used to assess programmed death-ligand 1 (PD-L1) expression: the Tumor Area Positivity (TAP) score and the Combined Positive Score (CPS). The study, led by Markus Moehler of the Department of Internal Medicine, Gastrointestinal Oncology, Johannes Gutenberg University Clinic, Mainz, Germany, included patients with advanced or metastatic gastric, gastroesophageal junction adenocarcinoma (GC/GEJC), or esophageal squamous cell carcinoma (ESCC) who were treated with tislelizumab, a programmed cell death protein-1 (PD-1) inhibitor, with or without chemotherapy.
Published in Modern Pathology, PD-L1 expression was evaluated using an investigational version of the VENTANA PD-L1 (SP263) assay. TAP score and CPS were both independently applied by pathologists trained in each scoring method. TAP score measures PD-L1–positive staining across tumor and stromal areas, while CPS calculates the ratio of PD-L1–positive tumor and immune cells to viable tumor cells.
At PD-L1 thresholds of 1 percent or more, 5 percent or more, and 10 percent or more, agreement between TAP score and CPS ranged from substantial to almost perfect, with Cohen’s κ values between 0.64 and 0.85 and intraclass correlation coefficients 0.73 or more. Across all three trials, PD-L1–positive subgroups identified by either TAP score or CPS showed similar hazard ratios for overall survival. For example, in RATIONALE-305, the OS hazard ratios for TAP score 5 percent or more and CPS 5 or more were 0.72 and 0.73, respectively. In RATIONALE-302, hazard ratios for TAP score 10 percent or more and CPS 10 or more were 0.52 and 0.54, respectively. Outcomes for progression-free survival, objective response rate, and duration of response were also comparable between the two scoring methods.
Safety profiles were assessed across PD-L1 subgroups defined by both TAP score and CPS. The incidence of any-grade, grade 3 or more, serious, and fatal treatment-related adverse events was similar across TAP and CPS subgroups in all trials.
The researchers found that TAP score and CPS demonstrated comparable performance in identifying patients with GC/GEJC or ESCC receiving tislelizumab. Both scoring methods showed analytical agreement and were associated with similar clinical and safety outcomes across multiple PD-L1 expression thresholds.
Newsletters
Receive the latest pathologist news, personalities, education, and career development – weekly to your inbox.
