AbstractBackground & aims Patients with non-cirrhotic non-tumoral portal venous system thrombosis (NCNT-PVT) require long-term anticoagulation to prevent rethrombosis (either splanchnic or at any site, referred to as overall rethrombosis - oRT) when there is an underlying high-risk thrombophilic disorder or history of previous thrombotic episodes. However, up to 25% of patients without such indications still develop oRT. Elevated factor VIII (≥ 150%), D-dimer (≥ 500 ng/mL) or the presence of high molecular risk (HMR) variants have been proposed as risk predictors for oRT in these patients. This study aimed to compare the prognostic value of these three biomarkers in a cohort of NCNT-PVT patients. Methods Multicenter, retrospective, observational study including 123 patients with NCNT-SVT without indication for long-term anticoagulation. Results During a median follow-up of 89.3 months (IQR 36–129), 33 patients (27%) developed venous oRT, with a cumulative incidence of 7%, 17% and 26% at 1, 5 and 10 years. Factor VIII ≥ 150% and HMR were associated with oRT risk (HR 5.92, 95% CI 2.67–13.15, p Conclusions Patients with unprovoked or local factor–associated PVT have a clinically relevant long-term risk of recurrence, with a 25% cumulative incidence of overall rethrombosis at 10 years. Elevated FVIII ≥150% identifies a subgroup at particularly increased risk, and its combination with high molecular risk variants may further refine risk stratification. Impact and Implications This study addresses an important clinical gap by evaluating whether prothrombotic and genetic biomarkers can help identify patients with non-cirrhotic, non-tumoral portal vein thrombosis who remain at risk of recurrent thrombosis despite the absence of classical thrombophilia. Our results show that persistently elevated factor VIII levels, alone or combined with high molecular risk variants, are strong predictors of rethrombosis in this population. These findings may help refine recurrence-risk stratification in patients traditionally considered low risk and could support more individualized decisions regarding long-term anticoagulation and follow-up intensity.
Baiges et al. (Sun,) studied this question.