Value of CRP, albumin, and lymphocyte index in predicting survival of patients with gastrointestinal malignancies: a systematic review and meta-analysis

Background We conducted this systematic review to present high-quality evidence on the prognostic ability of CRP, albumin, and lymphocyte (CALLY) index for gastrointestinal (GI) malignancies. Methods PubMed, Embase, Scopus, Web of Science, and Wanfang databases were searched till 15 th January 2025 for studies reporting the prognostic ability of CALLY for all GI malignancies. Hazard ratios (HR) were pooled in a random-effect model for overall survival (OS) and progression-free survival (PFS). Results 18 studies were included. CALLY index was found to be a significant predictor of poor OS (HR: 1.89 95% CI: 1.720, 2.077 I 2 = 12%) and PFS (HR: 1.617 95% CI: 1.444, 1.809 I 2 = 1%) in GI malignancies. Low CALLY was a significant predictor of OS in pancreatic cancer (HR: 1.772 95% CI: 1.279, 2.456), cholangiocarcinoma (HR: 2.07 95% CI: 1.106, 3.875), colorectal liver metastasis (HR: 1.67 95% CI: 1.032, 2.702), gastric cancer (HR: 1.884 95% CI: 1.606, 2.210 I 2 = 15%), colorectal cancer (HR: 2.284 95% CI: 1.737, 3.004 I 2 = 0%), hepatocellular cancer (HR: 1.649 95% CI: 1.308, 2.079 I 2 = 0%), and esophageal cancer (HR: 2.133 95% CI: 1.607, 2.831 I 2 = 62%). Likewise, low CALLY was associated with worse PFS in pancreatic cancer (HR: 1.289 95% CI: 1.006, 1.652), esophageal cancer (HR: 2.171 95% CI: 1.543, 3.056 I 2 = 0%), hepatocellular cancer (HR: 1.468 95% CI: 1.195, 1.801 I 2 = 0%), gastric cancer (HR: 1.904 95% CI: 1.539, 2.356 I 2 = 0%) and cholangiocarcinoma (HR: 2.13 95% CI: 1.163, 3.902). Random-effect meta-regression using sample size, age, male gender, TNM stage III/IV, lymph node metastasis, CALLY cut-off, low CALLY percentage, and follow-up as moderators were non-significant. Conclusions CALLY can be a simple and easy-to-use prognostic marker for GI malignancies. Further research is needed to decipher its role in specific GI malignancies and improve the quality of evidence. Systematic review registration https://www.crd.york.ac.uk/prospero/ , identifier CRD42025636999.