DOACs vs Vitamin-K Antagonists for Thrombotic Antiphospholipid Syndrome

Can patients with bona fide antiphospholipid syndrome (APS) who have an indication for anticoagulation be effectively and safely treated with a direct oral anticoagulant (DOAC) rather than with warfarin or another vitamin K antagonist (VKA)?

This meta-analysis of clinical outcomes — arterial and venous thrombosis and major bleeding — in patients with APS evaluated the data from four randomized controlled trials that compared treatment outcomes of DOACs with those of VKAs. Notably, three of the four studies investigated rivaroxaban (a total of 211 patients in the rivaroxaban-treatment arms) and one study looked at apixaban (23 patients in the apixaban arm). While altogether 474 patients were enrolled in these four trials, only 472 were eligible for formal evaluation in this meta-analysis. Unfortunate wording in the manuscript may have led to the impression that miscalculations occurred in this meta-analysis publication (wording in the Abstract: "472 patients were included"; wording in the Results section: “474 patients were included in the analysis”). However, miscalculation did not occur. Personal communication with the meta-analysis’ senior author, Dr. Bikdeli, led to the clarification that 2 patients (from the RAPS trial) had to be excluded from the final analyses.

Overall, 56% of patients were triple antiphospholipid antibody (APLA)–positive (lupus anticoagulant, anticardiolipin IgG or IgM, and anti–beta-2-glycoprotein-I IgG– or IgM–positive), and the remaining 44% were either single or double APLA–positive. The follow-up periods during the study were 7, 12, 20, and 36 months in the four studies. It appears that 36% (n = 130) of 358 patients with APS, in whom the pre-enrollment thrombotic event had been specified, had had an arterial event. Furthermore, 16.4% of 358 patients in the three trials, in whom aspirin status was reported, were on an anticoagulant plus aspirin.

Finding #1: Arterial thrombosis occurs more commonly in patients with APS treated with a DOAC than in those treated with a VKA. The arterial events that occur on DOACs are almost all strokes, not myocardial infarction or ischemic limb events. There was a significantly higher risk for developing arterial thrombosis in the overall group of patients treated with DOACs (10.3% vs 1.3 %; OR, 5.43; 95% CI, 1.87–15.75; P < .001), and these arterial events were strokes, not myocardial infarction or major ischemic limb events (Figure 2 in the publication). Venous thromboembolism (VTE) events were uncommon and not significantly different between the two treatment groups (1.7% vs 1.3% in the DOAC and VKA groups, respectively) and neither were major nor clinically relevant non-major bleeding events (10.3% vs 7.1% in the DOAC and VKA groups, respectively; Figures 3 and 4).

Finding #2: Triple APLA–positive patients clearly have a higher rate of DOAC failure, with the development of arterial thromboembolic events (Figure 5). In addition, single and double APLA–positive patients may have a higher rate of arterial events if treated with a DOAC rather than with a VKA. However, this trend towards a higher DOAC failure in this latter group was not statistically significant, possibly due to the relatively small number of patients in this group (Figure 5). Noteworthy is, though, that the one study (RAPS) that had the lowest percentage of triple-positive patients in its cohort reported that no thrombotic events occurred in any of their 57 patients treated with a DOAC.

Finding #3: Even the patients with APS who had no known history of arterial thrombotic events prior to study enrollment had a higher rate of arterial events on a DOAC (Figure 5).

Consequences of the data from this meta-analysis

1. Triple-positive patients with APS: As it had already become clear when the results of the individual TRAPS and Ordi-Ros trials were published, triple APLA–positive patients with APS who have an indication for anticoagulation are best treated with warfarin or another VKA, not a DOAC. This is not only true for patients who had a previous arterial thrombotic event but also for those on anticoagulation with a history of VTE without a previous clinical arterial thrombotic event.
2. Single- or double-positive patients with APS: Even the patients with APS who are single- or double-positive for APLA may have a higher DOAC failure and arterial thrombosis risk. However, at this point, this is not a solid conclusion, given the limited amount of data. This needs to be discussed with a patient when making a decision on which anticoagulant to use. However, caution in using a DOAC in these patients seems appropriate. One cannot be dogmatic that warfarin or another VKA should or must be used in these patients.
3. Is a DOAC plus aspirin the preferred treatment choice in patients with APS over anticoagulation alone? Unfortunately, no sub-analysis was performed in this meta-analysis to determine what impact on the occurrence of arterial events and DOAC failure the simultaneous use of aspirin and an anticoagulant had. This is an unanswered — yet clinically important — question when it comes to the best management of patients with APS and arterial events, particularly those who are triple-positive: is a combination of antiplatelet therapy with an anticoagulant, particularly when one uses a DOAC as the anticoagulant, superior in preventing future arterial events compared with the use of an anticoagulant alone? Given that 59 of 358 patients (16.4%), in whom the aspirin information was available for this meta-analysis, were on aspirin plus anticoagulant therapy, it would have been nice if these data had been analyzed and reported. This seems to be a missed opportunity. It would be very useful if a future trial addressed the combination of a DOAC plus aspirin in patients with APS — maybe this would be a more effective and better choice than testing a higher dose of a DOAC, as is being tested in the RISAPS trial (NCT03684564).

Two additional issues to point out

1. VKAs may have looked better in the four trials included in this meta-analysis than they actually are in the day-to-day management of patients with APS. The authors of this meta-analysis did not discuss in their “study limitation” section the issue of "effect modification by initial treatment," which may have influenced the results of the four individual trials, leading to a lower failure rate of VKAs compared with the DOACs studied; the four randomized studies may well have enrolled patients who, because of their history of having tolerated a VKA well prior to enrollment, were expected to continue to respond well to a VKA, as has been previously discussed more in detail in a previous study.¹
2. Collaboration: Finally, I read with interest that only three of the four trialist groups of the randomized trials provided additional study-level information. I wonder why that was. APS is challenging to study owing to its rarity. Additional data from the fourth group would have been nice to have to make this superb and very-welcomed meta-analysis stronger.

Reference

1. Moll S. Rivaroxaban Versus Vitamin K Antagonist in Antiphospholipid Syndrome. Ann Intern Med. 2020;172(7):510. https://www.acpjournals.org/doi/10.7326/L20-0053?url_ver=Z39.88-2003&rfr_id=ori:rid:crossref.org&rfr_dat=cr_pub%20%200pubmed