Splenic stiffness and platelet count to predict varices needing treatment in pediatric extrahepaticportal vein obstruction.

Abstract

Surveillance endoscopy to detect varices needing treatment (VNT) is important to prevent bleeding and morbidity in portal hypertension. In adult and pediatric cirrhosis, platelet count and liver stiffness measurement (LSM) are useful in selecting patients for endoscopy. Such recommendations do not exist for extrahepatic portal vein obstruction (EHPVO). Splenic stiffness measurement (SSM) has been studied in adult and pediatric EHPVO with conflicting results and methodological errors. This study evaluates the role of platelet counts and SSM to predict VNT and bleeding in pediatric EHPVO while comparing LSM and SSM between pediatric EHPVO and controls. One hundred and seven children (55 with EHPVO and 52 controls) were recruited. Clinical, biochemical, hematological, and radiographic parameters of all children were noted. All children with EHPVO underwent endoscopy. Of the 55 children with EHPVO, 48 (87.3%) had VNT. There was no difference in the platelet counts (85,000/mm3 vs. 120,000/mm3, p = 0.58) and SSM (3.62 vs. 3.19, p = 0.05) between EHPVO children with VNT and those without. They had poor sensitivity and specificity to predict VNT. EHPVO children with bleeding had higher SSM that those without. LSM was higher among EHPVO than among controls (1.19 vs. 1.10, p = 0.003). Those with LSM higher than controls had normal liver histology. SSM is higher in EHPVO bleeders but SSM and platelet counts are unreliable to predict VNT in pediatric EHPVO. Surveillance endoscopies may be needed in all pediatric EHPVO until better screening strategies are available. Not applicable.