Prediction of esophageal variceal bleeding in B-viral liver cirrhosis using the P2/MS noninvasive index based on complete blood counts.

Abstract

BACKGROUND/AIM: Periodic endoscopy for esophageal varices (EVs) and prophylactic treatment of high-risk EVs, i.e., medium/large EVs, small EVs with the red-color sign or decompensation, are recommended in cirrhotic patients. We assessed the cumulative risks for future EV bleeding using the following simple P2/MS index: (platelet count)2/[monocyte fraction (%) × segmented neutrophil fraction (%)].

METHODS: We enrolled 475 consecutive B-viral cirrhosis patients for 4 years, none of whom experienced EV bleeding. All underwent laboratory work-ups, endoscopy and ultrasonography. Those with EV bleeding took a nonselective β-blocker as prophylaxis. The major endpoint was the first occurrence of EV bleeding, analyzed using the Kaplan-Meier and Cox regression methods.

RESULTS: Among patients with EV bleeding (n = 131), 25 experienced their first EV bleeding during follow-up. To differentiate the risk for EV bleeding, we divided them into two subgroups according to their P2/MS value (subgroup 1: P2/MS ≥9 and subgroup 2: P2/MS <9). The risk was significantly higher in subgroup 2 (p = 0.029). From multivariate analysis, a lower P2/MS (p = 0.040) remained a significant predictor for EV bleeding along with large varix size (p = 0.015), red-color sign (p = 0.041) and Child-Pugh classification B/C (p = 0.001). In subgroup 1, the risk for EV bleeding was similar to that of patients with low-risk EVs (p = 0.164).

CONCLUSIONS: The P2/MS is a reliable predictor for the risk of EV bleeding among patients with EV bleeding. According to risk stratification, different prophylactic treatments should be considered for the subgroup with a P2/MS <9.