Prophylactic rivaroxaban improves clinical outcomes in discharged COVID-19 patients

In individuals who were discharged following hospitalization due to COVID-19, thromboprophylaxis with the novel oral anticoagulant rivaroxaban for 35 days led to improved clinical outcomes without increased bleeding compared with no anticoagulation post-discharge, according to findings from the MICHELLE* trial.

“There is a clear indication for in-hospital pharmacologic thromboprophylaxis for every patient with COVID-19 after bleeding risk assessment. However, there is no consensus on the role of extended thromboprophylaxis,” said Dr Eduardo Ramacciotti from the Science Valley Research Institute, Santo Andre, Sao Paulo, Brazil, who presented the findings during one of the late-breaking trial sessions at ESC 2021. [J Thromb Haemost 2020;18:1859-1865; Chest 2020;158:1143-1163]

Ramacciotti and colleagues therefore sought to evaluate whether post-hospital discharge thromboprophylaxis is warranted or not in this patient setting. [Thromb Haemost 2020;120:1597-1628]

The study comprised 320 patients (mean age 57 years, 60 percent male) who have been hospitalized for a minimum of 3 days due to COVID-19, received a standard dose of thromboprophylactic agents**, and had a high D-dimer (≥4). At discharge, participants were randomized 1:1 to receive either rivaroxaban 10 mg QD or no anticoagulation for 35 days. Sixty-two percent of participants had an IMPROVE*** score of 2–3, while the rest had a score of ≥4. [ESC 2021, Late Breaking Trials – COVID 19]

At day 35, the incidence of the primary efficacy^# outcome was lower among those who received rivaroxaban compared with those who did not receive any anticoagulation therapy (3.14 percent vs 9.43 percent), leading to a relative risk reduction of 67 percent (relative risk [RR], 0.33; p_superiority=0.03).

The primary outcome in the rivaroxaban arm was primarily driven by symptomatic pulmonary embolism (PE; 0.63 percent), asymptomatic PE (0.63 percent), and asymptomatic deep vein thrombosis (1.89 percent).

Secondary efficacy outcomes also favoured rivaroxaban and moved towards the direction of the primary endpoint, added Ramacciotti, as reflected by the lower incidences of symptomatic and fatal venous thromboembolism (VTE; 0.63 percent vs 5.03 percent; RR, 0.13), symptomatic VTE and all-cause death (2.52 percent vs 5.66 percent; RR, 0.44), and the composite of symptomatic VTE, MI, stroke, and cardiovascular (CV) deaths (0.63 percent vs 5.66 percent; RR, 0.11).

None had major bleeds. There were also no significant differences between the rivaroxaban and no-anticoagulation arms in terms of clinically relevant non-major bleeding (1.26 percent for both), another bleeding (1.26 percent vs 0.63 percent), and a combination of major, clinically relevant non-major, and other bleeds (2.51 percent vs 1.89 percent).

Subgroup^## analyses also generated results that aligned with the primary outcome, all in favour of rivaroxaban, with no signs of heterogeneity with point estimates moving towards the correct direction, said Ramacciotti.

“[We evaluated] patients discharged after hospitalization due to COVID-19 with increased IMPROVE score, and half of them spent some time at the intensive care unit, so [this is a cohort of] really sick people,” said Ramacciotti. “[Our findings suggest that] in this setting, thromboprophylaxis with rivaroxaban 10 mg QD for 35 days improved clinical outcomes without increasing bleeding compared with no out-of-hospital anticoagulation.”