Anticoagulation in sub‐Saharan Africa: Are direct oral anticoagulants the answer? A review of lessons learnt from warfarin

Warfarin has existed for >7 decades and has been the anticoagulant of choice for many thromboembolic disorders. The recent introduction of direct‐acting oral anticoagulants (DOACs) has, however, caused a shift in preference by healthcare professionals all over the world. DOACs have been found to be at least as effective as warfarin in prevention of stroke in patients with atrial fibrillation and in treatment of venous thromboembolism. In sub‐Saharan Africa, however, the widespread use of DOACs has been hampered mainly by their higher acquisition costs. As the drugs come off patent, their use in sub‐Saharan Africa is likely to increase. However, very few trials have been conducted in African settings, and safety concerns will need to be addressed with further study before widespread adoption into clinical practice.


| INTRODUCTION
Medical conditions that require anticoagulation are important causes of morbidity and mortality worldwide, but are often unrecognized or under-treated in low-and middle-income countries (LMIC) including sub-Saharan Africa (SSA). 1 Venous thromboembolism (VTE), presenting clinically as deep venous thrombosis (DVT) and/or pulmonary embolism is a key example. 2 The prevalence of DVT varies between 2.4 and 9.6% in postoperative patients, and between 380 and 448 per 100 000 births per year in pregnant and postpartum women in SSA. 1,2 Many hospitalized patients are at risk of VTE, but a multinational cross-sectional study of over 1500 hospitalized patients in SSA found that while 50.4% were at risk for VTE, only 51.5% of these received recommended forms of prophylaxis. 3 The mortality of patients diagnosed with pulmonary embolism is alarmingly high ranging from 40 to 69.5%. 1 Life expectancy is increasing across SSA 4 due to reduced childhood mortality 5 and improved control of infectious diseases, especially HIV. 6 This increasing life expectancy is associated with increased prevalence of cardiovascular diseases such as hypertension and atrial fibrillation (AF). In addition, rheumatic heart disease (RHD), a leading cause of heart failure in SSA, 7 is present in 12-66% of hospitalized patients presenting with AF in this region. 8 The rise in such cardiovascular diseases has created an increased requirement for the use of anticoagulant therapy.
Warfarin, a vitamin K antagonist (VKA) has existed for >70 years and is the mainstay of anticoagulation therapy in SSA. The directacting oral anticoagulants (DOACs) have been shown to be at least as effective as warfarin for treatment of VTE 9 and for stroke prevention in atrial fibrillation 10 and are already widely used in high-income countries. DOAC use is increasing in LMIC and is likely to increase further as the drugs come off patent. However, there remain unanswered questions regarding their use and situations where warfarin remains useful. This review aims to examine the role of DOACs in SSA by applying lessons learnt with warfarin use.

| Bleeding and reversal
The clinical pharmacology of warfarin and DOACs (including factor Xa inhibitors apixaban, edoxaban and rivaroxaban, and the direct thrombin inhibitor dabigatran) has been extensively described elsewhere. [11][12][13] We focus on specific challenges of anticoagulants resulting from their pharmacological properties, with reference to the SSA context. Firstly, any clinical benefit of preventing thrombosis must be measured against the risk of bleeding, which can be challenging in LMIC, where laboratory and clinical services are inadequate.
The lack of validated bleeding risk prediction tools leaves patients starting anticoagulation at risk of bleeding. A cross-sectional study in South Africa found warfarin among the top 4 causes of adverse drug reactions and the leading cause of preventable drug-related haemorrhage. 14 Unfortunately, there is limited knowledge on the rates of bleeding or risk profiles among patients in SSA taking warfarin or DOACs. In large international phase III RCTs, major bleeding rates with DOACs have been generally lower compared to warfarin 10 but patients on dabigatran and rivaroxaban had higher rates of major gastrointestinal bleeding compared to those receiving warfarin. 15,16 For warfarin reversal, vitamin K is effective, 17  In addition, their expense is prohibitive for routine use even in highincome countries. 18,19 Concerns about the lack of reversal agents may limit DOAC prescribing, 20 especially in SSA where plasma derived medicinal products that are useful in clinical management of bleeding such as prothrombin complex concentrate and fresh frozen plasma, are expensive and scarce. 21

| Drug monitoring
Warfarin dose can be adjusted using standardized dosing schedules to achieve the desired International Normalized Ratio (INR) target, 22 with the success of these adjustments expressed by the time in therapeutic range (TTR). Higher TTR is generally associated with fewer bleeding events and lower mortality. 23 Studies conducted in the operational context in SSA show that the median TTR is low, ranging between 29 and 47%. [24][25][26][27] This contrasts with clinical trial participants on the warfarin arm of studies comparing warfarin with DOACS who achieved TTRs between 58-69%, 15,16,28,29 which is closer to that seen in high income countries 30 where specialist clinics, more frequent monitoring including patient self-testing, guidelines and the use of clinical dosing algorithms may account for better control. The inadequate patient knowledge and adherence counselling 31,32 leading to low TTR may also lead to suboptimal treatment with DOACs.
INR monitoring can aid assessment of warfarin adherence and may actually improve this by increasing frequency of clinical interactions. Assessing DOAC adherence relies solely on adherence assessment tools that are prone to error. Few studies explore factors associated with adherence to oral anticoagulants in SSA. However, research among heart failure patients indicates that medication adherence is a major concern, often resulting from poor patient treatment knowledge. 33 Studies from high-income countries comparing adherence to warfarin and DOACs show varying results. 34,35 Factors influencing anticoagulant adherence in SSA need to be explored especially for DOACs where drug monitoring may not be feasible.
It may be necessary to perform DOAC monitoring to assist in clinical decision-making for patients with extremes of body weight,

What is already known about this subject
• Warfarin is the mainstay of anticoagulation therapy in sub-Saharan Africa (SSA).
• Direct-acting oral anticoagulants (DOACs) have been shown to be at least as effective as warfarin for treatment of venous thromboembolism and for stroke prevention in atrial fibrillation in phase III international clinical trials.
• DOAC use is widespread in high income countries and its use is increasing in low/middle-income countries.

What this study adds
• We highlight specific gaps in anticoagulation management in SSA such as management of bleeding and drug monitoring that may hinder widespread use of DOACs.
• We also highlight important patient groups such as patients with valvular heart disease and pregnant/lactating women where further data are required on the safety of DOACs.
• Lastly, we highlight several areas for anticoagulant research in SSA that may inform safe and cost-effective use of DOACs. acute renal injury, recurrent thrombosis and drug-drug interactions. 36 Other patient groups that are not routinely included in clinical trials of DOACs and yet are frequently encountered in SSA, such as those on antiretroviral therapy for HIV or those on rifampicin for tuberculosis may also be at risk of either bleeding or thrombosis due to exaggerated or inadequate effects resulting from drug-drug interactions with DOACs. 37,38 In addition, Asian race has been shown to have a significant impact on clearance of dabigatran 39 and modest nonclinically significant effect on apixaban clearance. 40 We do not know the impact of the African race on the pharmacokinetics of DOACs.
Therefore, having a measure of pharmacodynamic response (INR in the case of warfarin) is important. Whereas DOACs may prolong prothrombin time, INR and activated partial thromboplastin time, changes observed in these at therapeutic doses are highly variable and correlate poorly with DOAC effect. 41 Therefore specific drug anti-factor Xa assays using reagents and reference ranges that are specific to the SSA population should be established for DOAC monitoring 42 to inform dosage adjustment.

| ANTICOAGULATION IN AF AND VALVULAR HEART DISEASE
Few studies have assessed the prevalence of AF in SSA, reporting prevalence ranging from 0.3 to 0.7% in rural communities 43 to >4% in urban communities. 44,45 It is highest in patients with known cardiac disease 43 , more frequently due to valvular heart disease (VHD), often resulting from RHD, in SSA than in Europe or North America. 46 In the Randomized Evaluation of Long-Term Anticoagulation Therapy (RE-LY) AF registry, which included 1137 patients from 9 countries in SSA, 21.5% of these patients had RHD compared to 2.2% of patients from North America and 1.5% among those enrolled from western Europe. 46 Hospital based studies have reported coexisting VHD in 10-53% of patients with AF in SSA. [47][48][49][50] The mortality rate of patients with AF, particularly valvular, ranges from 10 to 40% 48-50 after 1 year. This has been attributed to late presentation of patients with advanced disease and suboptimal use of oral anticoagulant therapy 47 compounded by suboptimal anticoagulation control in patients who are taking warfarin. 48,50 Fortunately, subgroup analyses of the trial participants of the 4 major Phase III trials comparing warfarin to DOACs found that DOACs can be safely used in patients with forms of VHD such as aortic stenosis, aortic regurgitation and mitral regurgitation. 51,52 DOACs may therefore prove to be as effective as warfarin in patients with VHD, which will positively impact RHD care in SSA. Consequently, the efficacy of rivaroxaban is being compared to warfarin in patients with RHD and AF at high risk of stroke in the INVICTUS trial, an international, multicentre, randomized open-label trial that has enrolled >25% of its participants (1150) from 14 SSA countries 53 (Table 1).
Although still scarce in many countries in this region, heart valve replacement surgery accounts for some need for long term anticoagulation in SSA. 25,26,54 In these patients, warfarin is superior to DOACs since the mechanism by which valve components induce thrombin generation would require much higher DOAC doses than those currently recommended. 55 64 Conversely, warfarin was undetectable in the breast milk and infant plasma in small studies where mothers received a range of doses. [65][66][67] It is yet to be established if presence of DOACs in breast milk has implications for the nursing infant. This warrants further study in SSA where up to 96% of infants are breastfed. 68

| COST OF WARFARIN VS. DOACS
Treatment costs must consider direct costs such as the cost of medication, expenses associated with monitoring, admissions due to complications and additional treatments to achieve effective INR. 69,70 Additionally, indirect costs such as lost employment time and productivity associated with anticoagulation should be estimated. The biggest hindrance to using DOACs especially in LMIC has been inaccessibility and high cost of drug compared to vitamin K antagonists. In Uganda, a 5 mg warfarin tablet costs approximately 0.2 USD, significantly lower than the 2-4 USD that would be required to purchase a 15 mg tablet of rivaroxaban. 27   Catriona Waitt https://orcid.org/0000-0003-0134-5855