Background: Direct oral anticoagulants (DOACs) are recommended for the management of thrombosis prophylaxis, especially in patients with atrial fibrillation. As substrates of cytochrome P450 (CYP) 3A4 and/or P-glycoprotein, they are implicated in potential drug–drug interactions. NS5A/NS5B inhibitors are direct-acting agents (DAAs) against the Hepatitis C Virus (HCV) infection that exert a mild inhibition of P-glycoprotein without effects on CYP3A4. A DOAC and NS5A/NS5B inhibitor co-administration may lead to an increased risk of bleeding. Real-world data on the concomitant use of DOACs and DAAs are scarce. On this purpose, we perform a retrospective analysis on the risk of vascular adverse events (bleeding and thrombosis) among HCV patients under DOAC/DAA therapy, even in advanced liver disease. Methods: Between May 2015 and April 2023, patients treated with sofosbuvir-based DAA regimens and DOACs were consecutively included in this study from 12 Italian medical centers. Baseline characteristics, especially concerning bleeding risk and liver function, were collected. The occurrence of bleeding events, classified as major and minor, was the primary endpoint. Secondary endpoints were the rate of any thrombotic events and/or the need for discontinuation of one or both treatments. Moreover, a cohort of patients, matched by demographic characteristics (age and sex), that switched to vitamin K antagonists (VKAs) during the antiviral treatment was compared with the DOAC/DAA group. Results: A total of 104 patients were included. Thirty-eight of them (36.5%) were cirrhotic. Atrial fibrillation was an indication for anticoagulation in almost all cases (76%). Rivaroxaban (35.6%) was the most used DOAC, followed by apixaban (26.9%), dabigatran (19.2%) and edoxaban (18.3%). Sofosbuvir/velpatasvir (78.8%) was the most prescribed DAA, and all patients were already on anticoagulant therapy before the start of DAAs. During concomitant DOAC/DAA treatment, no major bleeding events were recorded, while four minor bleeding events occurred, but none resulted in DAA or DOAC discontinuation. At univariate analysis, the only additional risk factor statistically related to bleeding events was the anticoagulant therapy (hazard ratio [HR]: 13.2, 95% confidence interval 1,6-109). Performing an evaluation by a LOGIT binomial analysis with demographic characteristics, the antiplatelet therapy remained statistically associated to bleeding events. No significant differences were found in the rate of clinically relevant bleeding when the main population was compared with the VKA-switched cohort. A single major bleeding event leading to anticoagulation and DAA discontinuation was reported in VKA-switched matched cohort. Conclusions: In our study, the concomitant use of NS5A/NS5B inhibitors with DOAC showed good safety, and the only risk factor associated with bleeding events was the concomitant antiplatelet therapy. These findings support the use of DOACs during sofosbuvir-based HCV treatment, even in advanced liver disease. Replacing DOACs with VKAs does not appear to be of clinical benefit.

Background and aims: Management of severe thrombocytopenia poses significant challenges in patients with chronic liver disease. Here, we aimed to evaluate the first real-world European post-marketing cohort of cirrhotic patients treated with lusutrombopag, a thrombopoietin receptor agonist, verifying the efficacy and safety of the drug. Methods: In the REAl-world Lusutrombopag treatment in ITalY (REALITY) study, we collected data from consecutive cirrhotic patients treated with lusutrombopag in 19 Italian hepatology centers, mostly joined to the “Club Epatologi Ospedalieri” (CLEO). Primary and secondary efficacy endpoints were the ability of lusutrombopag to avoid platelet transfusions and to raise the platelet count to ≥50,000/μL, respectively. Treatment-associated adverse events were also collected. Results: A total of 66 patients and 73 cycles of treatment were included in the study, since 5 patients received multiple doses of lusutrombopag over time for different invasive procedures. Fourteen patients (19%) had a history of portal vein thrombosis (PVT). Lusutrombopag determined a significant increase in platelet count [from 37,000 (33,000–44,000/μL) to 58,000 (49,000–82,000), p < 0.001]. The primary endpoint was met in 84% of patients and the secondary endpoint in 74% of patients. Baseline platelet count was the only independent factor associated with response in multivariate logistic regression analysis (OR for any 1000 uL of 1.13, CI95% 1.04–1.26, p 0.01), with a good discrimination power (AUROC: 0.78). Notably, a baseline platelet count ≤ 29,000/μL was identified as the threshold for identifying patients unlikely to respond to the drug (sensitivity of 91%). Finally, de novo PVT was observed in four patients (5%), none of whom had undergone repeated treatment, and no other safety or hemorrhagic events were recorded in the entire population analyzed. Conclusions: In this first European real-world series, lusutrombopag demonstrated efficacy and safety consistent with the results of registrational studies. According to our results, patients with baseline platelet counts ≤29,000/μL are unlikely to respond to the drug.