What are the implications of the spontaneous spleno-renal shunts in liver cirrhosis?

Background: Although significant advances are expected to be made in the assessment of the portal hypertension-related complications, the prognostic role of spleno-renal shunts has not been fully explored so far. Clarifying this aspect could help tackle the life-treating events occurring in patients suffering from liver cirrhosis. The aim of the study was to analyze the relationships between the spleno-renal shunts presence at doppler ultrasound and the liver cirrhosis complications. Methods: Design: eighty one patients out of 129 formed the study population (35 females). Chronic liver damage in these patients was caused by HCV (66), HBV (2), alcohol abuse (2) or unknown etiology, likely non-alcoholic steatohepatitis (11). Setting: two Liver Units of university/primary hospitals in Southern Italy. Main outcome measures: grading of esofageal varices; detection of ascites: assessment of hepatic encephalopathy; evaluation of liver cirrhosis severity; tracking hepatocellular carcinoma; doppler features of spleno-renal shunts and splenic flow velocity; spleen longitudinal diameter at sonography. Results: The prevalence of spleno-renal shunts was 18.5%, without no difference concerning the etiology (HCV versus non-HCV, p = 0.870); the prevalence of hepatocellular carcinoma in patients with spleno-renal shunts was superior to that of patients without them (Pearson Chi-square, p = 0.006, power of sample size 74%), also after adjustment for liver decompensation (p = 0.024). The median score of hepatic encephalopathy in patients with and without spleno-renal shunts was similar, i.e., 0 (range, 0-2) versus 0 (0-3), p = 0.67. The median splenic vein flow velocity in patients with spleno-renal shunts was significantly inferior to that of patients without them, i.e., 13 cm/sec (95% confidence intervals, 6-18) versus 21 cm/sec (17-24), p < 0.0001. By far the largest percentage of large esophageal varices was in patients without spleno-renal shunts (p = 0.005). In contrast, the frequency of ascites and hepatic encephalopathy severity was overlapping in the two groups. BMI values but not Child-Pugh's classification predicted spleno-renal shunts (Ors = 1.84, 95% confidence intervals = 1.28-2.64, p = 0.001 and 1.145, 95% confidence intervals = 0.77-1.51, p = 0.66). Conclusion: Taking into consideration the relatively small sample size, patients with spleno-renal shunts are burdened by an increased incidence of hepatocellular carcinoma. BMI predicted the spleno-renal shunts presence.