The participants were treated at the Department of Hepatobiliary and Pancreatic Surgery at the First Affiliated Hospital of Ningbo University between January 2021 and December 2024. Investigators classified nutritional status using the GLIM criteria and used logistic regression to examine the relationships between PNI, SI, SII and the presence of malnutrition. Diagnostic performance was evaluated through the area under the receiver operating characteristic curve (AUC), with a DeLong test used to compare the combined PNI-SI model against PNI by itself. Of the 210 patients included, 68, representing 32.38% of the cohort, were diagnosed with malnutrition under the GLIM criteria. Multivariable analysis indicated that low PNI, low SI and high SII were independent risk factors for malnutrition, with each association reported at P below 0.05.

Among the individual indices, PNI had an AUC of 0.772, with a 95% confidence interval of 0.708 to 0.836 and P below 0.001. Its sensitivity was 0.683 and its specificity was 0.765. SI achieved an AUC of 0.688 (95% confidence interval 0.613 to 0.762; P below 0.05), with sensitivity of 0.577 and specificity of 0.750. SII had the highest sensitivity of the individual measures, at 0.882, but its specificity was 0.606 and its AUC was 0.647 (95% confidence interval 0.569 to 0.725; P below 0.05). These findings distinguish PNI’s stronger overall diagnostic performance from SII’s greater sensitivity and lower specificity in this surgical cohort. The analysis also shows that all three indices retained independent associations with GLIM-defined malnutrition after adjustment in the multivariable model.

Combining PNI and SI produced an AUC of 0.779, with a 95% confidence interval of 0.716 to 0.843 and P below 0.001. The combined model had sensitivity of 0.912 and specificity of 0.577, giving it a higher sensitivity and lower specificity than PNI alone. However, the DeLong test did not show a statistically significant difference in AUC between the combined model and PNI: the reported 95% confidence interval for the difference was −0.030 to 0.016, with Z = −0.629 and P = 0.529. The findings therefore support PNI, SI and SII as auxiliary predictors that can complement clinical nutritional assessment under GLIM criteria, with the PNI-SI combination providing a high-sensitivity option. Because the study was retrospective, limited to surgically treated HCC patients and conducted at one institution, the results do not establish causation or demonstrate that the indices will perform similarly in non-surgical populations or other clinical settings.

Subject of Research: Diagnostic performance of nutritional and immune indices for GLIM-defined malnutrition in surgically treated hepatocellular carcinoma.

Article Title: The value of the prognostic nutritional index, sarcopenia index, and systemic immune-inflammatory index as auxiliary tools in the diagnosis of malnutrition in liver cancer patients according to the GLIM criteria

Article References: Qiu, J., Wang, J., Lu, L., Xu, Q., Ma, L.-L., Pei, H., Jiang, X., & Yang, X. (2026). The value of the prognostic nutritional index, sarcopenia index, and systemic immune-inflammatory index as auxiliary tools in the diagnosis of malnutrition in liver cancer patients according to the GLIM criteria. BMC Cancer. https://doi.org/10.1186/s12885-026-16877-4

Image Credits: AI Generated

DOI: 10.1186/s12885-026-16877-4

Keywords: hepatocellular carcinoma, malnutrition, GLIM criteria, Prognostic Nutritional Index, sarcopenia index, systemic immune-inflammatory index, nutritional assessment, retrospective study, diagnostic accuracy, value, prognostic, nutritional