Heavy menstrual bleeding on rivaroxaban

Direct oral anticoagulants (DOACs) offer many advantages over vitamin K antagonists (VKA) whilst maintaining an equivalent efficacy and safety profile. Major bleeding events with DOACs are reported as equivalent or reduced compared to VKA (Ruff et al, 2014). While the female gender has been associated with a higher risk of major and clinically relevant non-major bleeding when prescribed anticoagulant therapy, including with DOACs (Alotaibi et al, 2013), the clinical implications of gender-specific bleeding with DOACs are unknown. Further, real-world experience might reveal different patterns of bleeding that require specific management in clinical practice. One such manifestation is heavy menstrual bleeding (HMB), affecting 10–30% women during some point of their child bearing years in the absence of anticoagulation (Lethaby & Farquhar, 2003; Liu et al, 2007), which is a well-recognized adverse event with VKA treatment though not frequently reported (Sjalander et al, 2007; Huq et al, 2011). HMB is often overlooked as an adverse effect of anticoagulation. Indeed, a recent meta-analysis of published clinical trials evaluating bleeding observed with rivaroxaban therapy has no specific mention of HMB (Wasserlauf et al, 2013). Perhaps this should come as no surprise, as the International Society on Thrombosis and Haemostasis (ISTH) definition of clinically relevant non-major bleeding (Schulman & Kearon, 2005) does not specify HMB as a sub-category of bleeding to report. This lack of formal recognition perpetuates the absence of specific information in the literature on this important area and, although rarely associated with serious events, HMB has an overall negative impact on women′s quality of life, incurring significant direct and indirect costs (Liu et al, 2007). Patients anticoagulated with DOACs for venous thromboembolism (VTE) include a younger demographic compared with those receiving anticoagulation for stroke prevention. These patients have much to gain from the advantages that DOACs have to offer on one hand, and much to lose with reduced quality of life associated with HMB, on the other. It is therefore important to explore any potential associations between increased bleeding and specific patient characteristics, in order to better counsel and manage potential complications during anticoagulation with DOACs in order to optimise adherence and patient satisfaction. Early experience from our clinic suggests an increase in HMB in females taking rivaroxaban, currently the main DOAC used for VTE treatment in the UK. The clinical course of women aged between 16 and 55 years, treated with rivaroxaban between September 2012 and September 2014 was reviewed. From a total of 128 women, HMB was reported in 26 (20%) with a mean age of 38·7 years. Fifteen were of African-Caribbean origin, 9 Caucasian and 2 of unspecified ethnicity (Table 1). The predominant indication for anticoagulation in the whole cohort was deep vein thrombosis in 67% (n = 89) followed by pulmonary embolism in 15% (n = 19), as compared to 84% (n = 22) and 12% (n = 3) respectively in the group of women with documented HMB (Table 1). Long-term anticoagulation was intended for 50% of the whole cohort (n = 64) and 62% (n = 16) of those with HMB. (Table 1). Forty-seven women (37%) were anticoagulated with an alternative agent prior to rivaroxaban, of whom 5 (11%) had previous reports of HMB. When prescribed rivaroxaban, 10 of these 47 women experienced HMB, of whom only two had reported HMB previously (Table 1). Anaemia was documented in 23% (n = 6) of those with HMB, with a mean haemoglobin (Hb) drop of 18 g/l and mean Hb of 114 g/l (range 72–140 g/l). There was one episode of major bleeding that necessitated hospitalization, blood transfusion and anti-fibrinolytic tranexamic acid, due to severe HMB. Uterine fibroid disease was found in 12 of the 15 women with documented HMB who had a pelvic scan arranged, with an average age of 43·2 years. Surgical intervention was planned in three patients, of which one awaits hysterectomy and two await myomectomy, both of whom had to discontinue anticoagulation with rivaroxaban. Only seven women without HMB had documented fibroid disease. Based on clinical experience, strategies implemented to manage HMB included omission (7/26) or halving of the dose (6/26) for the first 2 d of menses. These strategies were successful in reducing HMB, allowing continued anticoagulation with rivaroxaban; only 6% (8/1286) discontinued rivaroxaban due to HMB. No thrombotic events have occurred to date. We recommend that menstruating women should be counselled about potential HMB prior to commencing anticoagulation. Further investigation of HMB and anticoagulation is warranted including the impact on quality of life and management of HMB during treatment. The Einstein Choice trial (ClinicalTrials.gov NCT02064439) will further elucidate the efficacy and safety of decreasing the dose of rivaroxaban as an acceptable strategy in the prevention of secondary venous thromboembolism. In our clinic, we have insufficient real-world experience of the use of dabigatran and apixaban in the female population aged 16–55 years to enable us to comment whether our observation is specific to rivaroxaban or common to DOACs or anti-Xa inhibitors. With wider use of other DOACs in the younger female population, this should become clearer. Finally, in order to better understand the scale of the problem, we propose that the ISTH consider altering their definition of clinically relevant non-major bleeding to include HMB specifically, so that researchers conducting work reporting bleeding outcomes with anticoagulant therapy specify this, as part of their sub-group analysis. Melanie Ferreira and Sarah Barsam performed the research. Sarah Barsam, Jignesh Patel and Melanie Ferreira designed the research study. All authors analysed the data and contributed in the writing of the paper.