Frontiers in Oncology (Jun 2020)

Prognostic Nutritional Index, Another Prognostic Factor for Extranodal Natural Killer/T Cell Lymphoma, Nasal Type

  • Ningning Yao,
  • Qing Hou,
  • Shuangping Zhang,
  • Huan Xiao,
  • Yu Liang,
  • Xiaokai Xu,
  • Ruyuan Guo,
  • Hongwei Li,
  • Shengmin Lan,
  • Hongwei Si,
  • Jianzhong Cao,
  • Jianzhong Cao

DOI
https://doi.org/10.3389/fonc.2020.00877
Journal volume & issue
Vol. 10

Abstract

Read online

Objective: The prognostic nutritional index (PNI) is a significant prognostic factor in diffuse large B cell lymphoma, follicular lymphoma, and other malignancies. The current study aimed to explore its prognostic role in extranodal natural killer/T cell lymphoma (ENKTL).Methods: Patients diagnosed with ENKTL and treated during 2002 and 2018 (n = 184) were retrospectively recruited. PNI was calculated from albumin concentration (g/L) and total lymphocyte count (*109/L). The association of PNI and overall survival (OS) or progression-free survival (PFS) was assessed in univariate analysis and multivariate Cox regression validated by the 10-fold cross-validation method.Results: Survival analyses showed that both OS and PFS differed significantly between PNI groups stratified by a cutoff value of 49.0. The 3- and 5-year OS were 42.5 and 36.3% in the low-PNI (PNI < 49) subgroup and 70.6% and 63.9% (P < 0.001) in the high-PNI (PNI ≥ 49) subgroup, respectively. The corresponding PFS showed a similar pattern (38.4, 32.4 vs. 64.8, 54.0%, P < 0.001). Multivariate analysis indicated that PNI was significantly independent for both OS (HR = 0.517, 95% CI = 0.322–0.831, P = 0.006) and PFS (HR = 0.579, 95% CI = 0.373–0.899, P = 0.015). Furthermore, integrating PNI into the models of IPI (International Prognostic Index), KPI (Korean Prognostic Index), and PINK (prognostic index of natural killer lymphoma) could improve the area under the curve (AUC) and reduce the integrated Brier score (IBS) and Akaike Information Criterion (AIC) value of each model.Conclusion: PNI was a significant prognostic indicator for ENKTL.

Keywords