Spleen stiffness measurement rules out high-risk varices in patients with chronic portal vein thrombosis without cirrhosis

Background and aims Screening endoscopy can be spared in patients with compensated cirrhosis when spleen stiffness measurement (SSM) by vibration-controlled transient elastography (VCTE) is ≤40 kPa, as they have a low probability of high-risk varices (HRV). Conversely, endoscopy is required in all patients with chronic portal vein thrombosis (PVT) without cirrhosis. The objective was to evaluate the performance of SSM-VCTE to exclude HRV in patients with chronic PVT. Methods We retrospectively included patients with chronic PVT without cirrhosis, who underwent an upper endoscopy within 2 years before or after SSM-VCTE in 16 VALDIG centers, divided into a derivation and a validation cohort. Results 159 patients were included in the derivation cohort; 43% had HRV. 187 patients were included in the validation cohort; 32% had HRV. By univariable analysis, myeloproliferative neoplasm, ascites, hemoglobin, bilirubin, albumin, splenomegaly, portosystemic collaterals, liver stiffness – spleen diameter to platelet ratio score, liver stiffness measurement and SSM-VCTE were associated with HRV in both cohorts. By multivariable binary logistic regression analysis, only SSM-VCTE (p <0.005) remained associated with HRV in both cohorts. In the derivation cohort, SSM-VCTE ≤ 40 kPa had a sensitivity of 97% to rule out HRV, and could spare 41% of endoscopies, with 3% of HRV missed, and a 97% negative predictive value (NPV). In the validation cohort, SSM-VCTE ≤ 40 kPa could spare 43% of endoscopies, with 5% of HRV missed, and a 96% NPV. Conclusions This study gathering a total of 346 patients with chronic PVT without cirrhosis showed that SSM-VCTE ≤ 40 kPa can be used to identify patients with a probability of HRV ≤5%, in whom endoscopy can be spared. Impact and Implications Patients with chronic portal vein thrombosis who do not have cirrhosis usually have low liver stiffness measurement values; the liver stiffness cut-offs used to rule out high-risk varices in patients with cirrhosis cannot therefore be used in this population. We show here that spleen stiffness measurement by vibration-controlled transient elastography ≤40 kPa is able to identify patients with chronic portal vein thrombosis with a very low probability of high-risk varices, in whom screening endoscopy can be spared. Annual surveillance of spleen stiffness measurement could reduce the need for repeated screening endoscopies throughout a person's lifetime. This approach could enhance quality of life while also reducing risks associated with endoscopic procedures, particularly those related to anesthesia.