Prognostic Nutritional Index for Oral Cancer: A predictive parameter of surgical site infection and prognosis in oral cancer patients undergoing extensive resection with reconstruction-a single-centre retrospective study

Abstract

Objective

The incidence of surgical site infection (SSI) after extensive oral cancer surgery has now fallen to about 20 % but remains a challenge perioperative management. In this study, we investigated potential factors predictive of SSI, with a particular focus on nutrition, in patients undergoing extensive oral cancer surgery, and created an assessment tool specifically designed for predicting the onset of SSI in patients with oral cancer.

Methods

117 Patients treated at Gunma University Hospital between April 2009 and March 2018 were assessed for inclusion in the study. We investigated the relationship between tumour characteristics, nutritional factors, and SSI onset during post-operative hospitalisation. Based on the SSI predictors extracted by logistic regression analysis, the Prognostic Nutrition Index for Oral Cancer (PNI-OC) score was devised.

Results

A total of 29 patients (24.8 %) developed SSIs. In the logistic analysis, body mass index (BMI) (p < 0.05), total lymphocyte count (TLC) (p < 0.05), and total cholesterol level (T-cho) (p < 0.05) were identified as independent factors for SSI onset. The PNI-OC score as a predictive tool for SSI had an AUC was 0.84, a sensivitivity of 82.3 %, and a negative predictive value of 92.3 %. The PNI-OC (high risk score 4) was significantly associated with overall survival (OS) (log-rank test: p = 0.032).

Conclusion

In the setting of low SSI incidence, BMI, TLC, and T-cho were independent predictors. The PNI-OC score calculated using these three factors represented a useful nutrutional assessment tool for ruling out the perioperative SSI risk. Furthermore, the PNI-OC was determined to be an oncological predictive factor of OS.