Introduction The usage of transient elastography (TE) as a predictor of oesophageal varices has been widely studied. The formation of varices is the sequelae of portal hypertension (PH) which also manifest as splenomegaly leading to thrombocytopenia. We aim to establish the utility of TE in conjunction with PH in predicting varices.

Methods Patients who underwent TE with the FibroScan® between June 2010 and January 2016 in our trust were identified. Only patients with a liver stiffness measurement (LSM) of >11.5 kPa which is indicative of cirrhosis and had undergone an endoscopy were included. Thrombocytopenia was defined as a platelet count of <150 x 10⁹/L and the presence of splenomegaly was identified from radiological investigations. A TE measurement of >25 kPa has been regarded as the optimum cut-off value for predicting varices and is used in this study.

Results 2,643 patients underwent TE during the time period. 372 patients had a LSM of >11.5 kPa and of these patients, 157 patients underwent a gastroscopy. 41 patients (26.1%) had evidence of varices on gastroscopy. The sensitivity and specificity of TE in various combinations of PH in predicting varices are outlined in the table below.

In total 46 out of our 157 patients with cirrhosis had TE < 25 kPA, normal platelets and normal sized spleen on ultrasound scanning. Only 1 patient in that group (TE = 16.7 kPA) had early borderline oesophageal varices.

Conclusion In our cohort of patients, a cut-off value of >25 kPa on TE is a better predictor of varices (sensitivity 83%) than thrombocytopenia or splenomegaly. Assessment of LSM using TE in conjunction with thrombocytopenia and splenomegaly improves sensitivity further to 97%. Limiting screening endoscopy in patients with liver cirrhosis to those with any of those 3 criteria: TE > 25 kPA, thrombocytopenia or splenomegaly, may reduce demand for endoscopy in that group of patients by 30%.

Disclosure of Interest None Declared