Splenic Stiffness Is the Best Predictor of Clinically Significant Varices in Children With Portal Hypertension

Abstract

Background and Aims: 

We studied splenic stiffness measurement (SSM) by transient elastography (TE) and portal hemodynamics parameters (PHDp) on Doppler as predictors of clinically significant varices (CSV) in children.

Methods: 

All children of 6 months to 18 years of age with portal hypertension (PHT) (chronic liver disease, CLD and non-cirrhotic portal hypertension, NCPH) were enrolled. TE for spleen (SSM) and liver (liver stiffness measurement, LSM) and PHDp by Doppler ultrasonography were measured. Noninvasive indices for PHT were calculated. CSV were defined as esophageal varices ≥grade 2 and/or gastric varix. Binary logistic regression analysis (LRA) and receiver operating characteristic statistics were applied.

Results: 

A total of 150 (120 CLD and 30 NCPH) children formed the study cohort. Prevalence of CSV was higher in NCPH than CLD [73.3% vs 53.3%, Odd’s ratio (OR) 2.369, P = 0.04]. On LRA, SSM was found to be the only independent predictor of CSV in children with CLD [OR 1.19 (95% Confidence Interval (CI) 1.018–1.16), P = 0.000] as well as in NCPH [OR 1.088 (95% CI 1.018–1.16), P = 0.013]. This model improved prediction of CSV in CLD from 52.5% to 83.9% and in NCPH from 73.3% to 86.7%. In children with CLD, SSM at a cut-off ≥27.6 kPa and in NCPH, SSM at a cut-off ≥29.5 kPa predicted CSV. In children with CLD, SSM correlated with LSM (R = 0.610, P <0.001) and with noninvasive PHT indices except aspartate aminotransferase-to-platelet ratio index.

Conclusion: 

SSM is the best noninvasive predictor of CSV in childhood CLD and NCPH and can be used as screening test for endoscopy in children with PHT.