A new laboratory-based algorithm to predict microvascular invasion and survival in patients with hepatocellular carcinoma

Abstract

BackgroundPreoperative serum inflammatory markers have been correlated with survival outcomes after resection of hepatocellular carcinoma (HCC). Whether they can predict microvascular invasion (MVI) in HCC is still unknown. This study aimed to evaluate the association of inflammatory markers with MVI, and develop a simple and inexpensive preoperative prediction model for MVI. MethodsWe developed a novel index using routine laboratory tests to predict MVI. The index was developed based on a study on patients with HCC, and validated in an internal cohort and another external cohort. The infiltration of CD8+ T cells in tumors was measured using immunohistochemistry. The prediction accuracy was evaluated with the area under the receiver operating characteristic curve (AUC). ResultsThere were 165 patients in the training cohort, 107 patients in the internal validation cohort and 80 patients in the external validation cohort. On multivariable analysis in the training cohort, alkaline phosphatase (ALP) and lymphocyte count were independent predictors of MVI. Thus, the ALP-to-lymphocyte ratio (ALR) was developed. The AUCs of the ALR for MVI were higher than the other conventional clinical indices. An optimal cutoff point for the ALR of 69.9 stratified HCC patients into the high (≥69.9) and low (<69.9) groups. An ALR ≥69.9 was significantly associated with worse overall and disease-free survival outcomes. The performance of ALR was validated in the internal and in external cohorts. The CD8+ T cell counts were significantly higher in HCC in the ALR<69.9 groups. ConclusionALR was a simple, accurate and inexpensive alternative to predict MVI and an independent risk factor of prognosis for HCC patients. The dismal survival outcomes in patients with high ALR scores were related to decreased infiltrations of CD8+ T cells in tumors.