Aims Distinguishing epithelioid malignant mesothelioma (EMM) from poorly differentiated lung adenocarcinoma (PD-LUAD) remains challenging, particularly when 21.7% of PD-LUADs lack lineage-specific markers (thyroid transcription factor-1 (TTF-1)/Napsin A), creating a diagnostic blind spot. While GATA-binding protein 3 (GATA3) is established in sarcomatoid mesothelioma, its complementary diagnostic value and prognostic relevance in EMM are not well defined. Methods This retrospective study analysed 115 tissue specimens (55 EMMs; 60 PD-LUADs). Immunohistochemistry for GATA3, calretinin, Wilms’ tumour gene 1 (WT-1), TTF-1, Napsin A and pan-cytokeratin was performed. Results were correlated with clinicopathological parameters and overall survival (OS) using Kaplan–Meier and multivariate Cox regression analyses. Results GATA3 was expressed in 78.2% of EMM but only 6.7% of PD-LUAD cases (p<0.001). Although not specific enough for standalone diagnosis, GATA3 provided meaningful complementary value: in TTF-1/Napsin A–negative PD-LUAD, GATA3 remained negative in 92.3%, helping to exclude EMM when used within a broader panel. Incorporating GATA3 with calretinin and WT-1 improved panel sensitivity to 96.4% while maintaining 100% specificity. High GATA3 expression in EMM correlated significantly with advanced T stage, higher International Mesothelioma Interest Group stage and poor functional status (Karnofsky performance status/Eastern Cooperative Oncology Group). Multivariate analysis identified GATA3 expression (p=0.037), smoking (p=0.041) and clinical T stage (p<0.001) as independent predictors of shorter OS. A qualitative inverse relationship between tumorous GATA3 and GATA3-positive tumour-infiltrating lymphocytes was also noted. Conclusions GATA3 serves as a useful adjunct within established immunohistochemical panels, particularly in resolving ambiguity in double-negative PD-LUAD. Beyond its supportive diagnostic role, GATA3 demonstrates independent prognostic significance and may reflect underlying immune-microenvironmental features, meriting further exploration in biomarker-guided therapeutic stratification.
Thabit et al. (Thu,) studied this question.