SPLENO-SYSTEMIC SHUNTS AND COVERT HEPATIC ENCEPHALOPATY

Background and Aims: Portal-systemic shunts can involve blood from the intestine (patent paraumbilical vein, inverted left gastric vein) or from the spleen (spleno-systemic shunts [SSS]). A peculiar condition is that of total shunt in patients with inverted portal flow. The relationship between portal-systemic shunt and the occurrence of overt hepatic encephalopathy has long been known. However, that between SSS and the subtle neuropsychiatric alterations termed covert hepatic encephalopathy (CHE) is less clear. The aims of this study were to: i) assess the likelihood of CHE screening in relation to the presence of SSS; ii) evaluate the relationship between SSS and quantitative CHE indices. Methods: Three-hundred-and-thirty-one patients with cirrhosis, independently referred for hepatic Doppler-US between Jan-2009 and August-2012 were enrolled. They were qualified as having SSS if convoluted, anechoic spleno-renal and spleno-retroperitoneal channels were detected, and venous flow confirmed by colour- Doppler. Flow direction within the portal vein was also established. Information was obtained on independent referral of the same patients for CHE screening, including electroencephalography, within 6 months of Doppler-US. Results: Eighty-eight/331 (27%) patients were qualified as having SSS; this was spleno-renal in 17 (19%) and spleno-retroperitoneal in 71 (81%). Eight/331 (2%) patients, all with SSS, had inverted portal flow. Forty-three/331 (13%) patients underwent CHE screening, the prevalence of which was higher in those with SSS (34 vs. 5%; c2 = 47.2, p < 0.0001). Significant differences in spectral EEG featureswere observed between patients with/without inversion of the portal flow in the entire population (EEG frequency: 8.3±2.5 vs. 10.6±1.7 Hz, p < 0.05; slow delta activity: 18±17 vs. 6±5%, p < 0.01) and in the SSS group (EEG frequency: 8.3±2.5 vs. 10.3±1.7 Hz, p < 0.05; slow delta activity: 18±17 vs. 7±5%, p < 0.05). In patients without portal flow inversion, no differences in EEG parameters were observed in relation to SSS. Conclusion: A significant association was observed between SSS presence and the likelihood of CHE screening. However, EEG parameters were not different in patients with/without SSS. In contrast, the EEG was slower in patients with inverted portal flow compared to those with SSS only, suggesting that flow inversion is a risk factor for CHE.