20 years managing malignant pleural mesothelioma: a review of our experience and the quest for the optimal prognostic marker

• Main Author: Nunes Caldeira, João
• Publication Date: 2024
• Other Authors: Rodrigues Sousa, Sofia, Figueiredo, Ana Maria, Pêgo, Alice, Barata, Fernando
• Format: Article
• Language: eng
• Source: Repositórios Científicos de Acesso Aberto de Portugal (RCAAP)
• Download full: https://doi.org/10.32932/gecp.2022.02.018
• topic: Mesothelioma; malignant; survival analysis; prognosis; Mesotelioma pleural maligno; sobrevida; prognóstico

Introduction: Malignant pleural mesothelioma (MPM) is a rare, but aggressive incurable cancer. A variety of prognostic tools have been developed to guide clinicians and patients in the management of MPM regarding the most appropriate therapy and survival time. Objectives: To evaluate the usefulness of the Brims’ decision tree analysis, the EORTC prognostic score, the CALGB prognostic groups, the modified Glasgow Prognostic Score (mGPS), and the LENT prognostic score in our MPM patients. Methods: Local retrospective database analysis of patients with MPM diagnosed between 2000 and 2020. The different prognostic scores were applied and MPM group risks’ evolution was analyzed. Statistical analysis was performed with IBM® SPSS® Statistics 25. Results: A total of 67 patients were evaluated (55 males, 82.1%). Epithelioid MPM was the commonest histological type (51 patients, 76.1%). Median overall survival (OS) was 11 months (IQR 23) and median progression free-survival (PFS) was 7 months (IQR 12.5). Survival distributions were statistically significantly different for the Brims’ and the mGPS’ groups (respectively, χ2 = 7.188, p = .027, and χ2 = 6.46, p = .04). For EORTC, CALGB and LENT score, there were not statistically significant differences in survival distributions (respectively, χ2 =.57, p = .811; χ2 = 7.978, p = .157; and χ2 = 1.23, p = .267). Brims’ model and mGPS statistically significantly predicted OS (respectively, F(1,57) = 11.1, p < 0.01, and F(1,32) = 6.846, p = .01). EORTC, CALGB and LENT failed to statistically significantly predict OS (respectively, F(1,44) = .003, p = .955; F(1.43) = .722, p = 0.4; and F(1,14) = 1.193, p = .293). Conclusion: The small number of patients included and missing data for some parameters are limitations of our analysis. However, to our knowledge, this was the first study and the largest cohort to address MPM prognostic scores in our country. In our cohort, the Brims’ model and mGPS were useful in predicting survival. Both are simple tools with easily accessible parameters and should be considered for clinical practice.