PE & DVT prevention for cancer patients (API-CAT trial)
Low-dose apixaban worked well, but was it "noninferior"?
Patients with cancer who develop pulmonary embolism or deep venous thrombosis have the highest rates of recurrence. As many as 20-25% of people with active cancer will experience a second PE or DVT in the first 12 months after their first event.
Historically, they were required to self-inject with low molecular weight heparin such as enoxaparin once or twice a day, essentially for the rest of their lives or until they declined the therapy.
A pragmatic randomized trial published in 2023 suggested that direct oral anticoagulants could be an effective and much more convenient alternative to enoxaparin in cancer-associated venous thromboembolism.
Among 671 patients with invasive solid cancers or blood cancers with a new diagnosis of VTE, those randomized to DOAC had nominally fewer VTE recurrences (6%) compared to those receiving any low molecular weight heparin (8.8% VTE recurrence), meeting the trial’s criteria for noninferiority.
Major bleeding occurred equally (~5%) in both groups, higher than that in DOACs in patients without cancer. About 11% more patients were adherent with DOACs (i.e., more patients stopped injecting LMWH than stopped taking a DOAC).
In prior trials, apixaban resulted in less bleeding (relative to LMWH) than rivaroxaban; most of those randomized to DOACs were taking apixaban in this trial. Many oncologists had already begun using DOACs as secondary prevention in VTE associated with cancer, and this study likely widened the practice to become first-line.
The AMPLIFY-EXT trial, funded by Bristol-Myers Squibb and Pfizer and published in 2013, suggested that lower-dose apixaban might be as effective as the full dose at preventing VTE in patients without cancer.
In a new trial, they sought to show that apixaban can achieve adequate VTE prevention at a lower dose in the highest-risk cancer patients.
Did they succeed?
The API-CAT Trial
In the API-CAT trial, 1,766 patients with active cancer (of whom two-thirds were metastatic) and proximal DVT or pulmonary embolism at 121 centers in 11 countries were randomized to take either full dose (5 mg bid) or low-dose (2 mg bid) apixaban for one year, after 6 months of initial treatment for DVT/PE.
After 12 months of follow-up,
Recurrent VTE occurred nominally less often in the reduced-dose group (2.1% vs. 2.8% in the full-dose group) with an adjusted subhazard ratio of 0.76 and a 95% confidence interval of 0.41 to 1.41.
Significant bleeding occurred less often in the reduced-dose group (12% vs. 15.6% in the full-dose group, an adjusted subhazard ratio of 0.75 with a 95% CI of 0.58 to 0.97 and P=0.03).
Mortality was nominally lower in the reduced-dose arm (17.7% vs 19.6%, adjusted hazard ratio, 0.96; 95% CI, 0.86 to 1.06).
Outcomes (PE, DVT and bleeding) were centrally adjudicated with blinded adjudication, reducing the risk of bias.
But Was The Reduced Dose “Noninferior”?
As is typical for noninferiority trials, the odds were rigged in the house’s favor.
Achieving “noninferiority” required only staying within a margin of 2.0 for the upper limit of the 95% confidence interval of the subhazard ratio. This meant low-dose apixaban could have resulted in up to twice the number of PE/DVTs and been “noninferior”.
According to the authors, this choice was not influenced by Bristol-Myers Squibb or Pfizer’s funding:
The funder had no role in the trial design or conduct of the trial; in the collection, analysis, or interpretation of the data; or in the review or editing of the manuscript. The funder provided apixaban in bulk, free of charge; packaging of the drug and placebo tablets was the responsibility of [the French team overseeing the trial].
Any Differences In the Rates of Really Bad PEs?
Pulmonary embolism represents an enormous spectrum, from incidentally discovered and asymptomatic to rapidly fatal.
In this trial, there were no adjudicated fatal PEs in either group. There were three unexplained sudden deaths in the lower-dose group (0.4% incidence) and two in the full-dose group (0.3%).
Into the Weeds: What’s a Subhazard Ratio?
Hazard ratios are similar to risk ratios (i.e., relative risk), but are used in slightly different contexts.
Relative risk (= risk ratio) is used in situations in which the cumulative risk is being sought. The rate of events in one arm is compared against the rate in the other arm. Simple.
Hazard ratios are used when time influences the outcome, and the instantaneous or time-dependent event rates are sought. This is more applicable in survival analysis, such as with this population of patients, two-thirds of whom had metastatic cancer.
A subhazard ratio is a hazard ratio in the “subdistribution”, meaning the event rate (PE/DVT) distributed over time after accounting for competing risks (here, mortality).
The subhazard ratio in this case accounts for the expected and observed high rates of death throughout the trial (18% died over one year), which censored patients (prevented them from ever having a PE or DVT that could be counted).
If more patients were to die in one arm by chance, this could bias the results by making it incorrectly appear that there was a lower risk of PE in that arm, whereas if more patients had happened to survive, they would have lived to develop PEs in this hypothetical example.
Using a widely accepted statistical modeling technique by statisticians Jason Fine and Robert Gray, the subhazard ratio accounts for these competing risks (in this case, mortality) between trial arms.
Who Are Fine and Gray? Tell Me More About These Guys
No. We’re not going that far into the weeds.
Conclusion
Low-dose apixaban (2.5 mg bid) appeared to be at least as effective and safe as the full dose (5 mg bid) in preventing recurrent pulmonary embolism and DVT in people with active cancer, after 6 months of initial treatment.
The lower dose of apixaban has already been established for use in patients with atrial fibrillation who are elderly, frail, or have significant renal impairment, and is used off-label for patients considered to be higher risk for bleeding.
The API-CAT trial may extend apixaban’s enormous market advantage over its competitors. Dabigatran and rivaroxaban also have flexible dosing, but have a smaller evidence base supporting their use at lower doses.