Proposal of improved prognostic index for patients with extranodal natural killer/T cell lymphoma treated with non-anthracycline based treatment

Although serum beta-2 microglobulin (B2M) has been suggested as a potential prognostic predictor for patients with extranodal natural killer/T cell lymphoma (ENKTL), there are no prognostic models using B2M. We aimed to investigate the prognostic role of B2M by incorporating B2M into the most recently prognostic models, Prognostic Index of Natural Killer Lymphoma (PINK) and PINK-E (Epstein- Barr virus). Between January 2005 to December 2014, 141 patients with ENKTL were identified in the database of the Asan Medical Center, Lymphoma Registry. Among them, 108 patients were treated with non-anthracycline based treatment. Median B2M value was 2.45 mg/L (range, 1.0-22.0) and baseline B2M was elevated in patients (45.4%). With median follow-up duration of 32.1 months (range, 0.3-131.0), and median overall survival (OS) was not reached. In univariate analysis, elevated B2M level was significantly associated with poorer OS (HR = 3.66; 95% CI: 1.96-6.82; p < 0.0001). We performed multivariate analysis with risk groups by PINK and elevated B2M was demonstrated as a significant prognostic factor for OS (HR = 2.12; 95% CI: 1.07-4.19; p = 0.032). Considering the multicollinearity between elevated B2M and EBV infection, multivariate analysis of B2M with risk groups by PINK-E was not conducted. We tried to develop a new prognostic model with 4 elements of PINK and B2M (PINK-B). Three risk groups were composed as followings: low risk (0-1 points), intermediate risk (2-3 points), and high risk (4 or more points). The current model, PINK-B showed better discriminative power compared with PINK and similar with PINK-E for predicting 3-year OS of low-, intermediate-, and high-risk group; 79%, 74%, and 27% for PINK, 80%, 46%, and 23% for PINK-B, 83%, 47% and 26% for PINK-E, respectively. For ENKTL patients treated with non-anthracycline based therapy, we suggest a new prognostic index, consisting of age, stage, distant node involvement, non-nasal type disease and serum B2M, which could be good for discriminating poor risk groups and convenient to apply real practice.