Should "provoked" PEs get prolonged anticoagulation?

HI-PRO trial for long-term apixaban succeeds on both medical and business levels

Increasing evidence shows that most people who experience pulmonary embolism or deep venous thrombosis (collectively called venous thromboembolism) that is not clearly caused by a transient risk factor are at elevated risk for recurrence after treatment with anticoagulants.

Many of them found out the hard way, and through no fault of their own, after experiencing VTE recurrence despite treatment according to the historical standard of care (through ~2008 or so):

• For decades, experts advised treatment with anticoagulation for as few as 3 months for so-called “unprovoked” PE or DVT—those that come seemingly out of the blue without a clear risk factor or precipitant.

• PEs or DVTs occurring in the context of recent surgery, trauma, prolonged bed rest, pregnancy, or estrogen use were considered to be “provoked” by these factors and prescribed only 3 months of anticoagulation in most cases.

Historically, a large proportion (~30%) of the “unprovoked” patients had eventual VTE recurrence, but longitudinal studies were slow to confirm their persistently elevated risk after shorter courses of anticoagulation. A smaller but significant percentage of patients with “provoked” PE/DVT also experienced recurrence after cessation of anticoagulation, due to some unrecognized cofactor(s).

Today, patients with unprovoked PE or DVT are advised to consider indefinite (lifelong) anticoagulation. Those with provoked DVT or PE should undergo a careful analysis of potential cofactors that may warrant extending anticoagulation longer than 3 months.

These recommendations are individualized, based (among many other things) on how dangerous the clot was (large central PEs tend to recur as large PEs; DVTs tend to recur as DVTs; subsegmental PEs appear to be lower-risk) and the patient’s bleeding risk, along with patient preferences.

Pressure to Misclassify VTE as “Provoked”

Patients and physicians both desire the shortest possible treatment/prophylaxis course after a venous thromboembolism, both for lifestyle reasons and bleeding risk.

Anecdotally, this has commonly led to clinicians’ stretching the definition of “provoked” PE/DVT. Common examples of “pseudo-provocation” have included:

• Long car rides (long-haul flight >8 hours is the classic minor provoking factor);

• Acute illness without immobilization or bedrest;

• Any preceding illness generally.

As another example, although estrogen use is considered a minor transient risk factor, estrogen-containing contraception is used by tens of millions of women. Although it’s natural to blame the estrogen, an unknown percentage of them also have DVT-PE cofactors that do not reset to zero after 3 months of treatment for their “provoked” VTE.

When patients and physicians are both motivated to shorten the duration of anticoagulation, it’s tempting to identify “provoking” factors that weren’t, really—or were just one of multiple contributors.

In a new study in NEJM, this common clinical scenario was explored, without explicitly saying so.

A Provocative Trial Design

At Brigham and Women’s (Harvard) in Boston, investigators enrolled 600 patients with VTE who had both an (ostensibly) transient risk factor:

• Major surgery

• “Acute medical illness” (not usually hospitalized; see below)

• Trauma

• Immobility

• Estrogen therapy

• “Long-haul travel” (a surprising 17% of the total; see below)

• “Other” (9% of the total; see below)

and also what authors called an “enduring” risk factor, which could include

• Obesity

• Heart failure of any type

• COPD, asthma, or interstitial lung disease

• Chronic kidney disease of any severity

• Chronic liver disease

• Chronic inflammatory conditions (autoimmune or infectious)

• Atherosclerotic cardiovascular disease (CAD, stroke, PAD)

• Persistent immobility (e.g., wheelchair- or bedbound)

After at least 3 months of treatment (and usually >6 months) with regular-dose apixaban, they were randomized to receive low-dose apixaban (2.5 mg twice daily) or placebo for one year.

Those taking apixaban had far fewer recurrent symptomatic VTE episodes (1.3% vs. 10%), with a hazard ratio of 0.13, 95% CI, 0.04 to 0.36; P<0.001.

Only one patient taking apixaban experienced major bleeding (vs. zero in the placebo group).

About 5% of apixaban-treated patients had “clinically relevant nonmajor bleeding” (i.e., that required medical attention) vs. about 2% taking placebo.

No one died from PE in either group.

The Bristol Myers Squibb-Pfizer Alliance funded the study, but “had no direct role in the design, execution, statistical analysis, or reporting of the trial.”

Always Be Positive

Like most clinical trials, and virtually all industry-funded trials, the HI-PRO design was cleverly and strongly tilted toward producing a positive result.