RESUMO
BACKGROUND: Management options for unresected malignant pleural mesothelioma (MPM) are largely limited to palliative chemotherapy and best supportive care. This study sought to delineate subgroups most likely to benefit from chemotherapy. PATIENTS AND METHODS: The National Cancer Database was queried for newly-diagnosed unresected sarcomatoid, biphasic, and/or metastatic (M1) MPM. Statistics included Kaplan-Meier overall survival (OS) analysis with and without propensity matching, landmark Kaplan-Meier analysis to address immortal time bias, and multivariable Cox proportional hazards modeling in all patients as well as within histologic/M-classification-based subcohorts. RESULTS: Of 4655 patients (48% chemotherapy, 52% best supportive care), 15%, 27%, and 40% had epithelioid, biphasic, and sarcomatoid disease, respectively; 41% had M1 disease. The median OS in the chemotherapy and BSC cohorts was 10.4 versus 4.8 months (P < .001). OS differences persisted following landmark analysis (P = .038) and propensity matching (P < .001). Chemotherapy was associated with higher OS in M1 cases with unknown histology and M1 epithelioid patients (P < .001 for both). For non-epithelioid cases, chemotherapy was associated with higher OS for M0 (P < .001 for sarcomatoid and biphasic) but not M1 (P > .05 for both) disease. CONCLUSIONS: Chemotherapy may benefit metastatic epithelioid and non-metastatic non-epithelioid MPM to a greater degree than metastatic non-epithelioid disease. Causation, however, is not implied, and careful patient selection in this population cannot be understated.