Study of relationship between systemic immune inflammatory index, prognostic nutritional index and prognosis of patients with laryngeal cancer

Objective To analyze the establishment and predictive effectiveness of prognostic graph prediction model of laryngeal cancer patients based on systemic immune inflammation index (SII) and prognostic nutrition index (PNI).

Methods A total of 168 patients with laryngeal cancer who underwent surgical treatment in our hospital from January 2016 to January 2020 were selected as the study subjects. Receiver operating characteristic (ROC) curve was used to analyze the optimal truncation values of preoperative SII and PNI in predicting postoperative death of laryngeal cancer patients, and the optimal truncation values were used for grouping. The survival curve was drawn to analyze the survival of the patients and determine the risk factors affecting the prognosis of the patients, and the prognosis histogram was constructed according to the results. Bootstrap method was used for internal verification of the line graph model. The prediction efficiency and differentiation of the graph model were evaluated by area test under receiver operating curve (ROC) and consistency index (C index), respectively.

Results The median overall survival (OS) of the laryngeal cancer patients was 29 months ［95% confidence interval (CI) : 23–36］, and the 1-year, 2-year, and 3-year survival rates were 96.20%, 80.38%, and 71.52%, respectively. The survival curve showed that compared with high SII group, the laryngeal cancer patients in low SII group had a higher survival rate (χ²=30.231, P<0.001) and compared with low PNI group, the laryngeal cancer patients in high PNI group had a higher survival rate (χ²=28.347, P<0.001). Cox regression analysis showed that T stage, T3 stage, tumor differentiation degree, lymph node metastasis, SII and PNI were risk factors for postoperative death in the patients with laryngeal cancer. According to the results of Cox regression analysis, the prognosis prediction model of the laryngeal cancer patients was constructed. The results of Bootstrap method and C-index showed that the model had good calibration and differentiation efficacies. ROC curve results showed that the area under ROC curve (AUC) of the model was 0.852 (95%CI: 0.682–0.983) (P<0.001).

Conclusion Preoperative SII, PNI, T stage, tumor differentiation degree and lymph node metastasis are closely related to postoperative prognosis of laryngeal cancer patients. The column graph model constructed based on SII, PNI, T stage, tumor differentiation degree and lymph node metastasis has high accuracy, and good differentiation and clinical predictive efficacies.