Prognostic factors for mortality in malignant pleural effusion: Is it possible to predict survival?

Abstract

e21638 Background: The aim of this study was to identify predictors of overall survival (OS) following palliative pleural procedures in patients with malignant pleural effusion (MPE). Methods: Data was collected from our database between January 2012 and December 2016 including all patients with MPE. All patients were followed-up for at least 30 days after the pleural procedure. We evaluated radiological aspects, biochemical and hematologic parameters as well as clinical features. For OS analysis, patients were divided into two groups. Group I included OS < 30 days and Group II included OS > 30 days. Prognostic factors for OS were identified by univariate analysis, using Fisher's exact for the categorical variables and Student's t-test for the quantitative ones. The cutoff points for any significant continuous variables were determined by the ROC curve. Subsequently, significant variables were entered into a multivariate logistic regression analysis ( p < 0.05). Results: A total of 238 patients were included in the analysis (37,6% males). Median age was 62 year-old (range 24-965) and median follow-up time for surviving patients was 127 days. Median OS was 83 days for therapeutic pleural aspiration (TPA), 353 days for pleurodesis, 135 days for VAT drainages without pleurodesis due to trapped lung and 39 days for indwelling pleural catheter. Regarding primary site, median OS was 170 days for lung (n = 91), 253 days for breast (n = 70), 90 days for gynecological (n = 18), 33 for gastrointestinal (n = 34) and 305 days for others (n = 26). Factors affecting OS in univariate analysis with p < 0.0001 were ECOG, procedure, protein, albumin, neutrophil to lymphocyte ratio(NLR) and hemoglobin, pulmonary infiltrate (p < 0.0438), hematocrit (p < 0.0002), and chemotherapy line (p < 0.0229). In multivariate analysis the following variables were identified as independent prognostic factor: ECOG 3/4 (p < 0.0095), pulmonary infiltrate (p < 0.0011), albumin < 2.5 (p < 0.0325), and hematocrit < 35 (p < 0.0055). Conclusions: Patients with MPE who presented with ECOG 3/4, pulmonary infiltrate, albumin < 2.5 and hematocrit < 35 had shorter OS. The identification of these prognostic factors may assist physicians in choosing the optimal palliative technique.