New data find reduced bleeding risk for DOACs used in extended VTE prophylaxis

FOR YEARS, DATA have shown that oral anticoagulants can help prevent VTE in medically ill patients both in the hospital and in the post-discharge setting. But there’s been a catch: Those same data have shown that giving medically ill patients oral anticoagulants leads to an elevated risk of major bleeding.

That was the case in an earlier large study—the MAGELLAN trial—looking at the use of oral anticoagulants. Researchers found that the DOAC rivaroxaban was noninferior to enoxaparin in short-term use and superior to enoxaparin in long-term use compared to short-term use of enoxaparin followed by placebo. But there was also an elevated risk of major and fatal bleeding.

Read more about the risks of VTE in post-discharge patients here.

When researchers looked at the data more closely, however, they noticed that patients with five bleed risk factors accounted for most of the bleeding issues. By excluding those patients from the trial, researchers discovered they could slash bleeding events associated with VTE prophylaxis while still preventing VTE events.

“Modeling studies very consistently find that about half of hospitalized patients are at risk for VTE, and about one-quarter of those are at high VTE risk.”

~Alex C. Spyropoulos, MD
Northwell Health

In terms of preventing VTE, the study found that by extending thromboprophylaxis to 35 days, rivaroxaban could prevent fatal or irreversible thrombotic events in 12,000 patients a year in the U.S. It could also prevent symptomatic VTE and VTE-related death, as well as nonfatal and fatal PE, in about 24,000 patients globally a year.

Those data were strong enough that the FDA approved rivaroxaban for VTE prophylaxis for acutely ill medical patients both during hospitalization and post-discharge. That gives hospitalists a unique opportunity to prevent VTE in patients after they leave the hospital.

Today’s Hospitalist talked to Alex C. Spyropoulos, MD, lead author of the study that examined the subpopulation data from the MAGELLAN trial and a member of the executive committee of the MAGELLAN trial. Dr. Spyropoulos, who serves as system director of anticoagulation and clinical thrombosis services at Northwell Health at Lenox Hill Hospital in New York, has been working on VTE prophylaxis for more than 20 years.

Which patients were excluded from the therapy to reduce bleeding risks?
We focused on patients with five characteristics: any active gastroduodenal ulcer within three months of randomization or currently symptomatic, bleeding within three months prior to randomization or during index hospitalization prior to randomization, active cancer at randomization, medical history of severe bronchiectasis or pulmonary cavitation, and dual antiplatelet therapy at baseline. What’s interesting is that these patients accounted for only about 20% of the total trial population, but they were responsible for most major bleeds.

What happened to bleeding rates when these patients were excluded?
Major bleeding was drastically lowered. The major bleed rates in the MAGELLAN subpopulation were cut in half in both the hospital and the post-discharge populations. In terms of critical site and fatal bleeding, the incidence was < 0.1% or lower in the control group and the rivaroxaban group. And we found that the efficacy of rivaroxaban was maintained.

We had tested the hypothesis that excluding these patients would reduce bleeding events in another trial, MARINER, which was the case, so we were confident that we could identify a post-discharge population in which we could safely use rivaroxaban for extended prophylaxis.

What impact do you think these data will have on hospitalists?
Modeling studies very consistently find that about half of hospitalized patients are at risk for VTE, and about one-quarter of those are at high VTE risk. Let’s make sure that we get these people on appropriate pharmacologic prophylaxis. The rates of post-discharge prophylaxis currently are as low as 4%, so very few patients being discharged in the U.S. in 2019 are getting appropriate VTE prophylaxis.

We should formalize VTE risk assessment at discharge as well as at admission. That’s not being done right now in the U.S. The original trials of VTE prophylaxis were in the six- to 14-day range, so we’re finding that most patients are receiving inappropriate durations of prophylaxis. If patients are only in the hospital for three or four days, they’re not there long enough for the therapy to be effective.

How long should medically ill patients receive thromboprophylaxis?
Previous trials say that prophylaxis works if given for six to 14 days, but most data indicate that prophylaxis needs to be at least 10 days. I would say the minimum duration of prophylaxis should be 10 to 14 days. That’s what the studies have shown us.

What obstacles exist to extending VTE prophlyaxis in the medically ill?
To date, the largest roadblock has been the lack of an oral option. A lot of patients who need VTE prophylaxis are elderly individuals with limited cardiopulmonary reserves who have comorbid conditions. Three years ago, if you tried to convince an 80-year-old patient with rheumatoid arthritis to self-inject with heparin for a week, it wouldn’t happen. Now we have a real-world oral-only option.

What’s the next step for hospitalists?
Hospitalists on the front-line will want to identify what patients who will benefit from extended prophylaxis look like. To identify that group, pick a low-bleed risk population by excluding the five key criteria we used in the study, then target a population at high risk of VTE by using a scoring tool like IMPROVE VTE. That’s your patient population. If you do the math, about one-quarter of all your hospitalized medical patients will fit into that category.

How can doctors implement the lessons of the study?
At my institution, we have designed and hopefully implementing a randomized control trial to see if we can automate this decision-making process for hospitalists using a decision-support tool embedded in the EHR. We want to see if electronic alerts at admission and discharge will decrease VTE events in medically ill patients over usual medical care.