NOAC vs. Warfarin for antiphospholipid syndrome: a pattern emerging?

In the field of oral anticoagulation (OAC), great strides have been made in the last decades. Vitamin K antagonists (VKAs) such as warfarin were until recently the only option in this domain. A narrow therapeutic index and highly unpredictable pharmacokinetics make it difficult to be a patient on a VKA and difficult for health professionals to effectively administer the treatment.

Large trials have demonstrated clear benefits of non-vitamin K antagonist OAC (NOACs) over VKAs in the most common indications for OAC such as non-valvular atrial fibrillation (AF) or venous thromboembolism (VTE), leading to both reduced bleeding and reduced thrombotic events.1 This is reflected in current ESC guidelines.

However, there remain settings associated with particularly high thrombotic risk in which NOACs have not been well evaluated against VKAs, including those with mechanical heart valves, valvular AF (in the presence of at least moderate mitral stenosis) and thrombosis associated with thrombophilias such as antiphospholipid syndrome (APS). APS is an autoimmune disorder characterised by a combination of clinical thrombotic events and laboratory evidence of the presence of lupus anticoagulant, anti-cardiolipin or anti-β2 glycoprotein I antibodies. OAC with a VKA has generally been standard-of-care for patients with APS. However, use of a NOAC has been endorsed in some lower-risk cases or in those with absolute or relative contraindications to a VKA, though evidence supporting this approach has been relatively lacking.2

The recent publication by Khairani and colleagues in the Journal of the American College of Cardiology reports their meta-analysis of data from a total of 472 participants of four open-label randomised controlled trials comparing NOAC (either rivaroxaban or apixaban) with warfarin.3 They found that arterial thrombotic events occurred more commonly in those receiving NOAC (10.3 % vs. 1.3%, odds ratio [OR] 5.43 [95% CI 1.87 to 15.75], p<0.001). There was no significant difference in venous thrombotic events (1.20 [0.31 to 4.55]) nor major bleeding (1.02 [0.42 to 2.47], p=0.97).

It is worth noting the significant limitations of the analysis, which are acknowledged by the authors. Despite pooling from multiple studies, the sample size is still small and the number of events low. Nevertheless, these data are all we have at present and do not support use of NOAC over VKA for management of APS, which is perhaps a surprising finding in the context of those in non-valvular AF and VTE. 

This analysis comes after last year’s publication of the INVICTUS trial, which included 4531  patients with valvular AF secondary to rheumatic heart disease, who were randomised to either rivaroxaban or VKA.4 After a mean follow up of 3.1 years, rivaroxaban was not non-inferior to VKA in preventing the primary composite endpoint of stroke, systemic embolism, myocardial infarction, or death from vascular or unknown cause (proportional hazards ratio 1.25 [95% CI 1.10 to 1.41]) and indeed VKA demonstrated superiority (p<0.001). The authors carefully considered explanations for this but could not readily explain from their data, including detailed subgroup analysis, why VKA appeared to be superior in this setting in contrast to the opposite finding from multiple trials in patients with non-valvular AF.

Drugs such as rivaroxaban and apixaban target a single coagulation factor, Xa, whereas the chronic effect of VKA administration is broader, seeing levels of factors II, VII and IX as well as X reduced. It has been postulated that in the highest risk indications for OAC, single factor inhibition may not be adequate and this is one potential explanation for these findings. It is also worth noting that whilst the target INR in the studies of APS was generally 2.0 to 3.0, it was higher some patients e.g. with recurrent thrombosis. Conversely, the dose of rivaroxaban was 20 mg once daily (or 15 mg for renal adjustment) and apixaban 2.5 mg twice daily, which may leave some room for further intensification if stronger OAC effect is desired. 

At present, therefore, the data do not support routine use of NOACs in APS or valvular AF and VKA should remain the OAC of choice in these conditions. Further work might explore if higher clinically available doses of NOAC or combinations of oral single factor inhibitors provide better antithrombotic protection in these particularly prothrombotic settings.