Modified geriatric nutrition risk index as a prognostic predictor of esophageal cancer.

Abstract

This study aimed to establish a simple and useful prognostic indicator for elderly esophageal cancer patients. We designed the modified geriatric nutrition risk index (mGNRI) using the inverse of C-reactive protein (CRP) instead of albumin and compared its prognostic value with those of the GNRI and other indices. We included 128 patients aged > 65years who underwent esophagectomy for esophageal cancer. We defined mGNRI as (1.489/CRP in mg/dL) + (41.7 × present/ideal body weight) and divided patients into two groups: the low-mGNRI (mGNRI < 70, n = 50) and high-mGNRI (mGNRI ≥ 70, n = 78) groups. We retrospectively examined the relationship between mGNRI and long-term prognosis. The low-mGNRI group had more advanced cancer by stage, higher rates of recurrence, and earlier recurrence than the high-mGNRI group. Univariate analysis identified the following factors as significantly associated with poor overall survival (OS): a lower American society of anesthesiologist performance status (ASA-PS), male gender, CRP-albumin ratio ≥ 0.1, CRP ≥ 1.0, low-mGNRI, tumor depth ≥ T3, Charlson comorbidity index ≥ 2, tumor size ≥ 40mm, and age > 75years. A low-mGNRI, ASA-PS 3, age > 75years, and tumor depth ≥ T3 were independent unfavorable prognostic factors for OS. A low-mGNRI was an independent poor prognostic factor for relapse-free survival. We performed model selection analysis to identify the most clinically useful indices; mGNRI was the best predictive model. mGNRI in patients with esophageal cancer correlated with early recurrence and was a useful independent prognostic factor.