Direct oral anticoagulants for treatment of thromboembolic disease: Have we reached a plateau?

In recent years, a number of direct oral anticoagulants (DOACs, also known as novel anticoagulants or NOACs) have been developed for the treatment and prevention of venous thromboembolism (VTE), and for stroke prevention in non-valvular atrial fibrillation (NVAF). Commercially available DOACs include the direct thrombin inhibitor dabigatran and the Xa inhibitors, rivaroxaban, apixaban and edoxaban. The advantages of these agents are minimal drug and food interactions and predictable dose responses, thus eliminating the need for regular blood test monitoring and dose adjustments, as well as minimal need for low-molecular weight heparin (LMWH) bridging, and reducing the risk of breakthrough thrombosis and bleeding due to the unstable international normalized ratios often seen on vitamin K antagonists (VKAs) such as warfarin. Standard-dose DOACs are associated with a lower hazard for stroke/systemic embolism death and intracranial bleeding, and are non-inferior for major bleeding.1 Clinical trials have also shown DOACs to be non-inferior to VKAs for VTE treatment, with a significant reduction in the risk of intracranial and fatal bleeding.2 In comparison to VKAs, the gastrointestinal bleeding risk is higher with dabigatran, rivaroxaban and high-dose edoxaban; lower with low-dose edoxaban; and unchanged with apixaban.1 Consequently, international guidelines now recommend DOACs as first-line therapy for stroke prevention in NVAF and for treatment of VTE. Warfarin use is therefore restricted to individuals with severe renal impairment (creatinine clearance <15–30 ml/min), mechanical heart valves, breastfeeding women and individuals with antiphospholipid syndrome (APLS). The direct Xa inhibitors, edoxaban, rivaroxaban and apixaban are also non-inferior to LMWHs for the treatment of cancer-associated VTE. However, there was an increased risk of major bleeding in the edoxaban and rivaroxaban studies, and increased clinically relevant non-major bleeding in the apixaban study,3 when the DOAC arms were compared with the LMWH arms. Most bleeding events in these studies occurred in patients with gastrointestinal cancer; therefore, in patients with gastrointestinal cancer, DOAC use should be considered on an individual basis, after consideration of the risk–benefit ratio. Nevertheless, DOACs are a safe and effective alternative for many patients with cancer-associated thrombosis. While there are no large randomized clinical trials for DOAC use in unusual site VTE, data from numerous studies, including smaller randomized controlled trials, observation cohorts and case series, also attest to the efficacy and safety of DOACs in such patients, including those with cerebral venous and splanchnic vein thrombosis.4, 5 So, have we reached a plateau in our research? Potentially, however, there are still several patient populations where DOACs are used only cautiously or not at all, and there is still work to do in these areas. First, there are limited data regarding the safety of DOACs for breastfeeding women and at present, their use in this setting is not recommended. Similarly, transplacental transfer of apixaban, dabigatran and rivaroxaban has been demonstrated in animal models and DOACs are contraindicated in pregnancy. Reports on DOAC exposure during pregnancy do not suggest a high risk of embryopathy6; however, it is unlikely that DOACs will ever be used in pregnancy. Conversely, further research seems quite feasible in breastfeeding women, and may well show DOACs to be safe in this patient group. Second, although patients with moderate to severe renal impairment were excluded from most pivotal DOAC trials, these agents are increasingly being used in this group. The prescribing information for rivaroxaban and apixaban allows for reduced dosing in those with creatinine clearance 15–29 ml/min and NVAF, with no dose adjustments suggested for those with end-stage renal disease (ESRD) and VTE. However, these doses are based on pharmacodynamic studies rather than clinical outcome data. Warfarin use is also problematic in patients with ESRD because of its potential contribution to vascular calcification and calciphylaxis.7 In one large retrospective study of 25,523 patients with ESRD on dialysis and AF, patients starting warfarin were matched with those starting apixaban. The study found there was no difference in the rate of ischaemic stroke but that the bleeding risk was lower in those on apixaban (hazard ratio 0.72; 95% CI 0.59–0.87).8 The same study showed that standard-dose apixaban (5 mg twice a day) was associated with significantly lower risks of embolism and death when compared with either reduced-dose apixaban (2.5 mg twice a day) or warfarin. Similarly, small studies which have included patients with ESRD and VTE have shown no difference in VTE recurrence.9 Third, although studies have shown the thrombotic recurrence rate to be higher in the DOAC arms of studies on patients with triple-positive APLS, largely due to an excess of arterial events, there may be a group of lower-risk patients with single- and double-positive APLS, or APLS patients who have previously presented with VTE without arterial events, who could benefit from DOACs.10 Finally, the risk of bleeding on anticoagulation cannot be completely obviated, with a risk of major bleeding of approximately 1%–3% on DOACs overall. There remains a group of high-risk patients, especially in the elderly NVAF cohort, where DOACs are withheld due to bleeding concerns. Accordingly, research into factor XIa inhibitors has been a focus of interest for the past decade, since it is well established that congenital FXI deficiency is rarely associated with spontaneous bleeding but affords protection from ischaemic stroke and VTE.11 A recent phase 2 clinical trial showed equivalent efficacy of an oral FXIa inhibitor to LMWH in the setting of VTE prevention post orthopaedic surgery, with a low risk of bleeding, and further studies are needed to determine whether FXIa inhibitors can dissociate thrombosis from haemostasis.11 FXII inhibitors are also currently under development, although their use is likely to be restricted to inhibiting contact activation in those being treated with extracorporeal circuits rather than for VTE and NVAF.12 Are we there yet? Almost, but not quite. Because until we have answered this question for every patient group—why use VKAs or LMWH when a DOAC will do—then there is work yet to do. None declared.