Immune-nutritional indexes for predicting the response to nivolumab immunotherapy in gastric cancer

Abstract Background: Immune checkpoint inhibitors reportedly improve survival outcomes in patients with gastric cancer (GC). However, most patients develop disease progression, for which treatment options are limited. Although research has shown some benefits from nivolumab, accurate prediction of the response to nivolumab has not been achieved. We examined biomarkers for identification of patients with good response for promising treatment like immune checkpoint inhibitors. Methods: We analyzed 40 patients with clinical unresectable or recurrent GC who received nivolumab immunotherapy and divided them into a responder group (n=7) and non-responder group (n=33) based on the RECIST criteria. We prioritized the following biomarkers: nutritional markers (prognostic nutritional index [PNI], lymphocyte-to-plasma C-reactive protein ratio [LCR], and neutrophil–lymphocyte ratio [NLR]), sarcopenia, and osteopenia. Results: Our model predicted the response to nivolumab with excellent accuracy in immune-nutritional markers (LCR: area under the curve [AUC]=0.74, NLR: AUC=0.75, PNI: AUC=0.60), osteopenia (AUC=0.83), and sarcopenia (AUC=0.73). Moreover, the predictive accuracy of the risk-assessment model was superior to that of each parameter (AUC=0.93). The high-risk group based on the risk-assessment model was associated with poor overall survival (P<0.01) and progression-free survival (P=0.08). Moreover, univariate Cox proportional hazard regression analysis revealed that our model was an independent predictor of survival outcomes (hazard ratio=3.22, 95% confidence interval=1.62–6.42, P<0.01). Conclusions: Measurement of nutritional markers, osteopenia, and sarcopenia allowed prediction of a response to nivolumab immunotherapy in patients with GC. Our novel risk-assessment model for predicting the response to nivolumab has potential for clinical translation.