MCL-079 survival prediction models (AIIMS mantle cell lymphoma overall survival [AMOS] score and AIIMS mantle cell lymphoma event-free survival [AMES] score) for patients with mantle cell lymphoma from Indian population

Abstract

Context: Mantle cell lymphoma (MCL) in Southeast Asia specifically among Indian population presents in advanced stage with bulky disease, significant extranodal involvement, and poor ECOG PS. Indigenous survival prediction models for this part of the world are lacking. Objective: Development of prediction models for overall survival (OS) and event-free survival (EFS) in Indian MCL patients. Design, Setting and Study Participants We analyzed ambispective clinical data of 90 MCL patients treated January 2013-December 2023 at Dr. B.R. Ambedkar Institute Rotary Cancer Hospital, All India Institute of Medical Sciences (AIIMS), New Delhi, India. Outcome Measures: Statistical analysis performed using STATA 13.0. Clinical and laboratory correlates of OS and EFS with P<.2 in univariate logistic regression analysis were included in multivariate analysis. Predictive models (AMOS and AMES scores) were developed using coefficients of variables from regression models. Receiver operating characteristic (ROC) curve analysis was used to set optimal cutoff for both models. Performance of cutoffs was reported using sensitivity, specificity. P<.05 defined statistical significance. Results: In multivariate analysis; age >60 years (a), presence of B symptoms (b), ECOG PS ≥2 (c), splenomegaly (d), LDH ≥ULN (e), and nonuse of rituximab maintenance (f) were significantly associated with no OS and no EFS at 1 year; male sex (g) was associated with no OS at 1 year. Each variable (a-g) was dichotomized to have a score of 0 (if variables were as mentioned above) or 1. AMOS score calculated as 18.9(a) + 20.4(b) + 10.0(c) + 11.1(d) + 19.4(e) + 33.3(f) + 14.9(g). AMES score calculated as 17.8(a) + 14.9(b) + 10.0(c) + 15.0(d) + 13.8(e) + 21.1(f). Mean (±SD) AMOS score 59.8 (±28.3); optimal cutoff at 48.8. AMOS score ≥48.8 had sensitivity, specificity of 86.0%, 81.0%, respectively, to predict OS at 1 year. Mean (±SD) AMES score 46.8 (±22.6); optimal cutoff at 42.2. AMES score ≥42.2 had sensitivity, specificity of 75.0%, 85.0%, respectively, to predict EFS at 1 year. Area under ROC (AUROC) for AMOS and AMES models were 0.86 and 0.91, respectively. Conclusions: AMOS and AMES scores are indigenous models with high predictive capacity for OS and EFS, respectively, in patients of MCL from this part of the world. This exemplifies their utility to predict survival outcome in MCL patients.