Prognostic associations of PD-L1 expression and BAP1 loss in pleural mesothelioma: a retrospective single-center cohort from Colombia

Abstract

Background Evidence on the prognostic associations of programmed death-ligand 1 (PD-L1) and BRCA1-associated protein 1 (BAP1) in pleural mesothelioma (PM) remains inconsistent, and joint data from Latin American cohorts are limited. We evaluated these biomarkers in a Colombian cohort. Methods This retrospective single-center cohort included adults with histologically confirmed PM treated at the Instituto Nacional de Cancerología, Bogotá, from January 2011 through December 2023. PD-L1 was quantified using tumor proportion score (TPS), with ≥1% as the primary threshold; BAP1 was classified as retained or lost by nuclear immunoreactivity. Overall survival (OS) was measured from pathologic diagnosis to death. Event-free survival (EFS) was measured from diagnosis to first documented progression or death; patients without an event were censored at last contact. Kaplan–Meier methods and separate multivariable Cox models were used. Results Sixty-seven patients were included; 60/67 (89.6%) had locally advanced or metastatic disease. PD-L1 TPS ≥1% was present in 26/67 (38.8%), and BAP1 loss in 47/61 evaluable tumors (77.0%). All 67 patients died. Median OS was 9.20 months (95% CI, 6.11–14.13), and median EFS was 6.51 months (95% CI, 5.49–9.13). PD-L1 TPS ≥1% was not clearly associated with OS (adjusted HR, 1.03; 95% CI, 0.57–1.86) or EFS (adjusted HR, 1.11; 95% CI, 0.62–2.00). BAP1 loss also showed no clear adjusted association with OS (HR, 1.57; 95% CI, 0.78–3.16) or EFS (HR, 1.65; 95% CI, 0.82–3.33). Conclusions Neither PD-L1 TPS ≥1% nor BAP1 loss showed a clear adjusted association with survival. These jointly measured estimates extend the evidence from an underrepresented Latin American population but require external validation; the exploratory ≥5% and ≥10% analyses were based on only 14 and 7 patients, respectively.