Respir Med. 2021 Oct 13;189:106650. doi: 10.1016/j.rmed.2021.106650. Online ahead of print.
ABSTRACT
BACKGROUND AND OBJECTIVE: Prognostication of malignant pleural effusion (MPE) guides treatment strategies but existing prognostication scores are yet to be validated in Asians. We aimed to evaluate the performance of these scores in an Asian population. A refined score was also proposed based on the impact of EGFR mutation on survival.
METHODS: Survival and clinical data of histocytologically-confirmed MPE patients from a Hong Kong hospital were analyzed with the LENT, modified-LENT, PROMISE and SELECT (converted from its original model) scores. A refinement of the LENT score for Asians was proposed by inclusion of EGFR status (EGFR-LENT), which was compared with the LENT score and validated in an independent patient cohort.
RESULTS: All prognostication scores performed well on risk stratification by Kaplan-Meier curve (log rank p < 0.0001) in 368 MPE patients except for LENT in low-risk group. C-statistics for LENT, modified LENT, PROMISE and SELECT in predicting 3-month mortality were 0.77, 0.80, 0.80 and 0.82, respectively. The proposed LENT score refinement (EGFR-LENT) improved stratification among low-risk patients; with a higher C-statistic (0.83) in 3-month mortality prediction than LENT (0.77, p = 0.0121), PROMISE (0.80, p = 0.3713), and SELECT (0.82, p = 0.7908) scores. Validation of EGFR-LENT in an independent cohort (124 patients) confirmed good performance in predicting 3-month mortality (C-statistic 0.87, vs 0.79 in LENT, p = 0.0444).
CONCLUSION: All existing scores had reasonable performance in prognosticating MPE, and LENT score refinement by inclusion of EGFR mutation status improved its performance among Asian MPE patients.
PMID:34688123 | DOI:10.1016/j.rmed.2021.106650
