While the anticoagulant didn’t raise the odds of major bleeding, nonmajor clinically relevant bleeds did trend higher.
MADRID, Spain—For patients who experience venous thromboembolism (VTE) provoked by surgery, immobility, or other factors and who face ongoing risk factors, 12 months of low-intensity apixaban (Eliquis; Bristol Myers Squibb) therapy reduces the odds of recurrent VTE by nearly 90%, according to the HI-PRO trial.
Moreover, researchers report, the direct oral anticoagulant does so while offering a low risk of major bleeding.
Gregory Piazza, MD (Brigham and Women’s Hospital, Boston, MA), when presenting the late-breaking results this weekend at the European Society of Cardiology (ESC) Congress 2025, explained that the distinction between provoked VTE—the focus of HI-PRO—and nonprovoked VTE is relevant to the duration of anticoagulant therapy that follows.
“Unprovoked venous thromboembolism is typically treated without a predetermined stop date and provoked venous thromboembolism given time-limited anticoagulation of 3 months,” he told attendees. But observational studies have suggested that patients with provoked VTE may be at higher risk of recurrence if they have other provoking factors that persist.
Both the ESC and the International Society on Thrombosis and Haemostasis (ISTH) have stressed the need to understand this particular profile, Piazza continued. In 2019, the ESC guidelines for acute pulmonary embolism added clarity by classifying various risk levels for long-term VTE recurrence, he pointed out. “Unfortunately, no set of guidelines clearly provides recommendations on what to do about duration of anticoagulation in [this subset].”
With these data, which were simultaneously published in the New England Journal of Medicine, “we see that one more piece of the extended-duration anticoagulation puzzle has fallen into place,” he said. HI-PRO highlights “a population with provoked venous thromboembolism and enduring risk in which the risk of recurrence off anticoagulation is high enough to warrant consideration of continued [therapy].”
As the trial’s discussant in the Hot Line session, Christopher B. Granger, MD (Duke University Medical Center, Durham, NC), characterized the results as “clear and important.” He acknowledged HI-PRO does have a few limitations, most notably that it was performed at a single center and involved a “modest” number of events.
Still, said Granger, “in the context of other trials of extended treatment for provoked VTE, this trial will change my practice.”
Both Primary Endpoints Met
For the double-blind trial, investigators enrolled 600 adults (mean age 59.5 years; 57.0% women; 19.2% nonwhite) who had experienced VTE due to a transient provoking factor, had at least one enduring risk factor, and had completed at least 3 months of anticoagulation. They randomized them to receive either oral apixaban (2.5 mg twice daily) or placebo for another 12 months.
The most frequent provoking factors were surgery (33.5%), immobility (31.3%), trauma (19.2%), and acute medical illness (18.3%), while the most common enduring risk factors were chronic inflammatory or autoimmune disorder (52.2%), body mass index ≥ 30 kg/m2 (48.2%), atherosclerotic cardiovascular disease (29.3%), and chronic lung disease (22.3%).
Median number of days adhering to the assigned therapy was 350 in the apixaban group and 338 in the placebo group. Six patients in each group did not take any of their assigned regimen.
First symptomatic recurrent VTE, the primary efficacy outcome, occurred in four (1.3%) of the apixaban-treated patients and 30 (10.0%) of those given placebo (HR 0.13; 0.04-0.36). One instance of ISTH major bleeding, the primary safety outcome, developed in the apixaban group (0.3%)—it was a 3-mm parafalcine subdural hematoma that occurred when a patient fell from a horse and did not result in either hospitalization or drug discontinuation. None of the placebo group had major bleeds.
However, 14 of 294 apixaban patients (4.8%) did have clinically relevant nonmajor bleeding: most often vaginal (1.7%), hematuria (1.4%), and rectal (1.0%). Five of 294 patients (1.7%) in the placebo group had such an event, most commonly vaginal (1.0%). This difference was nonsignificant, but risk trended higher with apixaban (HR 2.68; 95% CI 0.96-7.43; P = 0.06).
One patient assigned to apixaban died, as did three assigned to placebo—no deaths were attributed to CV or hemorrhagic causes. The rate of nonfatal adverse events unrelated to bleeding was 2.0% in each group.
Considerations for Practice
Granger, in his discussion, drew attention to the clinically relevant nonmajor bleeds, which he said may be more of an issue for patients “in general practice, who may be higher risk than this population.” He also asked whether the findings apply to all patients with provoked VTE and enduring risk, or if they could be more relevant to particular subgroups within that category.
“That’s a terrific question,” said Piazza. “There are a couple of ways to think about it. First, there were patients in this study who had multiple enduring risk factors, and I think that’s an easy group of patients that we can conceive of being the right fit for this therapy.
“Future studies,” he continued, “should probably go beyond just clinical markers and look at things like polygenic risk, [artificial intelligence]-informed risk scores, and even biomarkers, if we ever get some high-quality ones, to identify and more precisely pick patients for extended-duration thromboprophylaxis.”
Neil A. Zakai, MD (University of Vermont, Burlington), who wrote an editorial for the paper, notes that the field has begun to consider the best course of action after provoked VTE.
“For decades, the guidance was simple: treat a provoked venous thromboembolism . . . for 3 to 6 months, stop, and move on. Safer anticoagulants and emerging evidence now challenge this long-held approach,” he writes. “Recurrence risk does not reset to zero after a provoked event, and bleeding risk during the receipt of anticoagulation is far from negligible.”
The HI-PRO results, says Zakai, “invite us to reconsider whether the ‘provoked’ label is a license to discontinue anticoagulation or the beginning of a more nuanced conversation with our patients.”
The binary categories of VTE—provoked versus unprovoked—“are too blunt to guide care because they group together patients with hugely different recurrence risks,” he comments. “Thus, we need to develop and validate models that integrate baseline risk factors, features of the acute event, bleeding risk, patient values, and clinician experience to determine who may benefit from extended VTE prevention.”
These factors must be better defined to ensure consistent treatment but also leave room for conversations about an individual patient’s benefits and risks with extended anticoagulation, says Zakai.