Rivaroxaban May Yet Protect After a Hospital Stay: MARINER

A secondary analysis suggests both arterial and venous events are reduced at a 10-mg dose, with potential lessons for COVID-19.

Rivaroxaban May Yet Protect After a Hospital Stay: MARINER


(UPDATED) A standard dose of rivaroxaban (Xarelto; Bayer/Janssen) reduces the combined risk of venous and arterial thromboembolic events after discharge among patients recently hospitalized for acute illness, according to a fresh look at data from the randomized MARINER trial. There appears to be no penalty paid in major bleeding.

As reported in 2018, the study had failed to meet its primary endpoint, showing no decrease in the composite of symptomatic venous thromboembolism (VTE) or VTE-related death among patients randomized to placebo or rivaroxaban (either 10 mg daily or, for those with creatinine clearance < 50 mL/min, a lower dose of 7.5 mg daily) for 45 days. The lower dose was deemed to have been ineffective for the subset of patients in whom it was used. As such, the new analysis focused on patients whose baseline creatinine clearance of ≥ 50 mL/min qualified them for randomization to the 10-mg dose.

“We now, after 15 years’ worth of research, can identify a high-VTE-risk, low-bleeding-risk population that can benefit from extended [thromboembolic prophylaxis], and I think we’re confident to know this constitutes about a quarter of all hospitalized patients,” lead author Alex C. Spyropoulos, MD (Northwell Health at Lenox Hill Hospital, New York, NY), told TCTMD. “Changes should be done really at an institutional level [to ensure] there’s the option to do this, and then of course clinicians can use their . . . criteria to make this happen at the patient level.”

For Spyropoulos, who gave the caveat that this prespecified secondary endpoint can only be hypothesis-generating, the hint that the direct oral anticoagulant (DOAC) might have wide-ranging effects is good news. Specifically, he said, there may be room for rivaroxaban-based treatment in COVID-19 patients.

“We think these results, if anything, are probably more beneficial in the COVID-19 population because of their baseline risk of arterial and venous events,” he predicted. This concept is set to be tested in numerous trials.

MARINER Revisited

Of 9,822 patients (mean age 67.8 years; 55.5% men) in the current analysis, published online this week in the Journal of the American College of Cardiology, 4,909 were randomized to rivaroxaban 10 mg and 4,913 to placebo. Mean baseline creatinine clearance was 87.8 mL/min, and the mean length of hospitalization was 6.7 days.

The composite efficacy endpoint included symptomatic lower-extremity deep vein thrombosis (DVT) and nonfatal pulmonary embolism (PE), myocardial infarction, nonhemorrhagic stroke, and cardiovascular death. On the whole, risk of this endpoint was reduced by a relative 28%, with an absolute difference of less than 0.5%; for each individual component, however, the difference didn’t reach statistical significance.

Outcomes at 45 Days After Discharge




HR (95% CI)

Composite Efficacy



0.72 (0.52-1.00)




0.20 (0.04-0.91)




0.36 (0.12-1.14)




1.62 (0.67-3.92)




0.54 (0.28-1.06)

CV Death



0.93 (0.60-1.44)

Major Bleeding



1.44 (0.62-3.37)

What stands out about these results, said Spyropoulos, is the “concept that arterial and venous thrombotic events are intimately linked. We used to have in many ways an artificial distinction . . . thinking that because of the platelet-derived aspects versus the fibrin/thrombin-derived aspects of the other that an antiplatelet versus an anticoagulant strategy is the way to go.”

Behnood Bikdeli, MD, MS (NewYork-Presbyterian/Columbia University Irving Medical Center, New York, NY), who was not involved in the research, described this analysis and the larger MARINER trial as rigorously done.

However, despite the fact that the MARINER researchers used the IMPROVE score to target patients vulnerable to VTE and tried to exclude those vulnerable to bleeding, the “absolute risk reductions are fairly small and the risk of bleeding isn’t miniscule,” Bikdeli observed to TCTMD. “They’re only reporting major bleeds. I wouldn’t be surprised if you looked at clinically relevant bleeds [that that rate] would be a little increased.”

Moreover, the numbers of arterial events in this analysis are so small “that we shouldn’t make conclusive comparisons,” said Bikdeli. That said, rivaroxaban has already been studied for acute coronary syndromes in the ATLAS ACS-2 TIMI-51 trial and for stable atherosclerotic vascular disease in the COMPASS trial, he added.

Samuel Z. Goldhaber, MD (Brigham and Women’s Hospital, Boston, MA), in an editorial, agrees that the study by Spyropoulos et al “informs us that VTE outcomes are interwoven with cardiovascular death and stroke outcomes. This “theme of reduction of overall cardiovascular events, including VTE,” has also been seen with the anti-Xa agent betrixaban (Bevyxxa; Portola Pharmaceuticals), he points out.

Both betrixaban, on the basis of the APEX trial, and rivaroxaban, on the basis of MAGELLAN, are US Food and Drug Administration-approved for extended-duration VTE prophylaxis. Yet Goldhaber sees barriers to uptake. Specific to rivaroxaban, he mentions the FDA’s exclusion criteria related to major bleeding and predicts that the “1% absolute VTE reduction is not enough to spawn ‘local champions’ at US hospitals,” where budgets are likely to be strained during and after COVID-19.

As Goldhaber notes, the process of adoption may be slow—it took 11 years for the idea of VTE prophylaxis in hospitalized surgical patients to gain credibility. As data came in showing further benefit with anticoagulation in the month after discharge, the surgical community was quick to adopt “extended-duration VTE prophylaxis as the norm, not the exception,” he writes.

Peter K. Henke, MD (University of Michigan, Ann Arbor), lead author on recent American Heart Association guidance for hospital-setting VTE prevention, told TCTMD that the study’s most interesting aspect is that the main driver of the composite endpoint was symptomatic VTE—exactly what the therapy is aiming to prevent. Knowing that the benefit is there, what’s needed next is a more precise way to balance the need for anticoagulation and its risks.

Bleeding is a real problem for some, Henke noted. In the MARINER analysis, two patients in the rivaroxaban died as a result of their major bleeds. “So that’s not something you can totally ignore. Again, balancing that risk-benefit is really key,” he emphasized.

Around 35% of patients in the data set had an IMPROVE score of 4 or above, placing them in the highest-risk category for VTE; it would be interesting, Henke said, to know rivaroxaban’s effects specifically in this group.

Beyond drug therapy, though, educating patients to walk and get physical activity after discharge is in itself important, he added. And there may be a role for aspirin.

A Role in COVID-19?

Now, in part due to COVID-19, the adoption of VTE prophylaxis in discharged patients may speed up, predicted Spyropoulos. This pandemic has drawn attention to the “fact that thromboembolic disease is real in hospitalized patients—it’s not a joke, it’s real, we see it happen all the time,” he said. And given that hospital stays have grown shorter in recent years than in the past, Spyropoulos said, protection in the early period after hospitalization is important.

Bikdeli, too, sees a potential for rivaroxaban in COVID-19. He, along with Mahesh Madhavan, MD (NewYork-Presbyterian/Columbia University Irving Medical Center), and colleagues, included the drug in their recent review of pharmacologic agents aimed at “thromboinflammation” in COVID-19 patients.

It will be important to understand which patients are most at risk of catastrophic thromboembolic events, he said. “What we know is anticoagulants are going to suppress the coagulation cascade, so they’re going to give us some [protection], but that’s not going to be a free lunch—it’s going to come at the cost of bleeding.”

Various DOACs are being studied—and even used—for COVID-19, Bikdeli said, but rivaroxaban holds an edge thanks to its prepandemic evidence base in hospitalized, medically ill patients.

Henke also said he’s seen high-risk COVID-19 patients given DOACs, most commonly rivaroxaban and apixaban, in the hospital and after discharge. Though there are trade-offs between the two drugs in terms of side effects and number of daily doses, he said, “in terms of efficacy, I would imagine they are both very similar.”

  • The MARINER study is sponsored by Janssen Research & Development LLC.
  • Spyropoulos reports having served as a consultant for Janssen Research & Development LLC, Bayer, Portola, Boehringer Ingelheim, and Bristol-Myers Squibb; having received research support from Boehringer Ingelheim and Janssen; having served on an advisory board for Daiichi Sankyo; and having received a stipend from the ATLAS group.
  • Goldhaber reports no relevant conflicts of interest.
  • Bikdeli reports that he is a consulting expert, on behalf of the plaintiff, for litigation related to a specific type of inferior vena cava filters.
  • Henke reports no relevant conflicts of interest.