Journal club

Abstract

Metastatic malignant pleural effusion (MMPE) is associated with highly variable survival, highlighting the need for accurate prognostic tools to guide clinical decision-making. The LENT score has traditionally been used to estimate median survival, whereas the PROMISE score was developed to predict 3 month mortality. A recent external validation of both models (ERJ Open Res 2025, DOI: 10.1183/23120541.01019–2024) in 773 patients revealed limitations. Although both scores demonstrated moderate discriminative ability for overall survival (Harrell’s C-statistics 0.72 for LENT, 0.73 for PROMISE), their median survival estimates lacked precision. Performance improved at fixed timepoints (3-, 6-, and 12 months mortality), with C-indices approaching 0.8. However, both scores underestimated survival in low-risk patients and lacked integration of tumor-specific features or oncogenic drivers such as EGFR mutations. Additionally, LENT includes mesothelioma and unconfirmed effusions, which may inflate survival estimates, while PROMISE requires biomarkers not always available in routine practice. To overcome these limitations, the GASENT model was recently developed prospectively in Galicia (Arch Bronconuemol 2025, DOI: 10.1016/j.arbres.2025.04.001) using a derivation cohort (n=475) and validated in an independent cohort (n=205). It incorporates six widely available variables: Age, Sex, ECOG-Performance Status, Neutrophil-to-Lymphocyte Ratio, and Tumour type. GASENT …