A prognostic model for resectable acral melanoma patients on the basis of preoperative inflammatory markers.

Acral melanoma is a rare disease, but is common in Asia. Knowledge of its prognostic indicators is limited. Growing evidence indicates that inflammation plays a critical role in the development and progression of acral melanoma. We developed a novel prognostic model on the basis of preoperative inflammatory markers and examined its prognostic value in a cohort of patients. This retrospective study included 232 acral melanoma patients who underwent radical surgical resection between 2000 and 2010 at the Sun Yat-sen University Cancer Center. Significant predictive factors were identified by multivariate Cox regression analyses, and a prognostic model on the basis of these variables was constructed to predict survival. Kaplan-Meier curves were plotted to estimate overall survival. Multivariate analyses showed that C-reactive protein, albumin/globulin ratio, age, lactic dehydrogenase, and lymph node positivity were related independently to survival. After analyzing these variables, we classified patients into three risk groups. The new prognostic model identified three categories of patients with different prognoses (P < 0.001) and significantly stratified patient prognosis into different tumor stages. The area under the curve of the new prognostic model was 0.684 (95% confidence interval: 0.620-0.743), which was significantly higher than that of the other variables (P < 0.001). C-reactive protein and albumin/globulin ratio were independently related to survival in our study population and the prognostic model developed using inflammatory-based scores was useful in stratifying patients into different risk groups. Thus, this model will be a valuable complement to the 2009 American Joint Committee on Cancer staging for Asian patients with acral melanoma. Copyright (C) 2017 Wolters Kluwer Health, Inc. All rights reserved.