Noninvasive predictors of esophageal varices in patients with hepatosplenic schistosomiasis mansoni.

BACKGROUND:No previous study have evaluated transient elastography for predicting esophageal varices in hepatosplenic schistosomiasis. AIM:To investigate noninvasive methods of predicting esophageal varices in patients with hepatosplenic schistosomiasis mansoni. METHODS:Cross-sectional multicentric study included 51 patients with hepatosplenic schistosomiasis. Patients underwent ultrasonography-dopplerfluxometry, upper endoscopy, complete blood cell count and transient elastography (Fibroscan®) for liver and spleen stiffness measurement (LSM and SSM). Noninvasive scores previously established for cirrhotic population were studied: platelet count to spleen diameter ratio (PSR), LSM-spleen diameter to platelet ratio score (LSPS) and varices risk score (VRS). We proposed a version of LSPS and VRS by replacing LSM with SSM and named them SSPS and modified-VRS, respectively. RESULTS:Esophageal varices were detected in 42 (82.4%) subjects. Individuals with varices presented higher SSM (73.5 vs 36.3 Kpa, p = 0.001), splenic vein diameter (10.8 vs 8.0 mm, p = 0.017), SSPS (18.7 vs 6.7, p = 0.003) and modified-VRS (4.0 vs 1.4, p = 0.013), besides lower PSR (332 vs 542, p = 0.038), than those without varices. SSPS was independently associated with varices presence (OR=1.19, 95%CI 1.03-1.37, p = 0.020) after multivariate analysis. In a model excluding noninvasive scores, SSM was independently associated with varices diagnosis (OR=1.09, 95%CI 1.03-1.16, p = 0.004). AUROC was 0.856 (95%CI 0.752-0.961, p = 0.001) for SSM and 0.816 (95%CI 0.699-0.932, p = 0.003) for SSPS (p = 0.551). CONCLUSIONS:Spleen-related variables were predictors of esophageal varices: SSM, splenic vein diameter, SSPS, modified-VRS and PSR. Multivariate models indicated that SSM and SSPS are useful tools for predicting varices in non-cirrhotic portal hypertension by hepatosplenic schistosomiasis and may be used in clinical practice.