Editorial on Poli et al. Thromb Haemost 2012; 107: 1100-1106. The prevalence of chronic kidney disease (CKD) has doubled in the past 10 years related to the rise in obesity, diabetes and hypertension, becoming recognised as a silent epidemic (1). CKD is associated with an increased risk for cardiovascular morbidity and mortality, with the frequency of atrial fibrillation (AF) being especially high. Indeed, AF is common in patients with CKD at different stages of severity (1–4). Both conditions share several pathophysiological and clinical risk factors, such as hypertension, diabetes mellitus and coronary heart disease (5). Thus, the prevalence of AF increases with decreasing glomerular filtration rate (6), and conversely, CKD increases the risk of thromboembolism in AF independently of other risk factors (7). AF is the most common arrhythmia in clinical practice and associates a high morbidity and mortality due mainly to its fivefold increase in ischaemic stroke. The prevalence of AF increases with age, and age is also a risk factor for stroke, so the combination of age and AF means that stroke prevention in elderly AF patients is a huge challenge (8), added to the fact that CKD also increases with age. Oral anticoagulation reduces substantially the risk of stroke; however, elderly people are less likely to receive antithrombotic therapy, in spite of the net clinical benefit of oral anticoagulation in such patients (9, 10). Renal impairment is an established bleeding risk factor in patients taking oral anticoagulation. Several studies have reported the association between CKD and a significantly increased risk for bleeding in anticoagulated patients with AF or venous thromboembolism (11–14). Thus, CKD has been included in the HAS-BLED bleeding score (15) which has been proposed in current AF clinical guidelines (16, 17), or other risk scores like the HEMORRHA2GES (18), or more recently, the ATRIA based bleeding risk score (19). In a similar way, the RIETE investigators have derived and validated a score to predict the risk of major bleeding for venous thromboembolism which also includes renal impairment (20). In the June 2012 issue of Thrombosis and Haemostasis, the investigators of the EPICA study, reported that CKD is an independent risk factor for bleeding and that a wide proportion of aged patients taking oral anticoagulation had severe or moderate CKD independently of the method used for measuring glomerular filtration rate (21). Whilst the abbreviated MDRD (Modification of Diet in Renal Disease) formula is more accurate in estimating the glomerular filtration rate in elderly patients (22), it has recently been proposed that the CockroftGault formula is the better method to evaluate renal function in heart failure patients (where it was predictive of mortality), although the latter patients were about 10 years younger (23). In the study by Poli et al., the Cockroft-Gault formula seemed to overestimate the incidence of severe CKD, as previously reported – also, independent of the formula used to calculate, severe CKD patients (creatinine clearance less than 30 ml/ minute) was associated with an independent increased risk of bleeding (21). There is much renewed interest regarding to the evaluation of renal function in AF patients, as the new oral anticoagulants are contraindicated in patients with severe renal impairment (24), which could have implications for thromboprophylaxis. Indeed, the oral direct thrombin inhibitor dabigatran is highly excreted by the kidneys so there could be an increased risk of bleeding secondary to the accumulation of the drug (25). Indeed, patients with severe renal dysfunction were excluded from the recent clinical trials with new oral anticoagulants (26). Also, the management of warfarin therapy in severe CKD patients is difficult as they require lower warfarin dosages, spend less time on therapeutic range and are at risk of overanticoagulation (27). Nonetheless, stroke risk is closely related to bleeding risk in AF patients (28), as many risk factors for stroke are also associated with a higher risk for anticoagulation associated haemorrhage (17). CKD does not only coexist with advancing age, but also with hypertension, diabetes, heart failure and vascular disease, which are included within current stroke risk stratification schemes such as the CHA2DS2-VASc score (29). Severe renal dysfunction is not included in neither of the two stroke risk stratification scores, but it has already been informally proposed in 2010 that it could be included in CHA2DS2-VASc score, with the little “c” letter indicating “chronic severe renal impairment” pending validation studies (3). Renal impairment is at least included in the HAS-BLED score, which also does have some modest predictive value for cardiovascular events and mortality, but less so compared to its predictiveness for major bleeding (where as a bleeding score, HAS-BLED performs as good as a multivariate analysis) (12). Several studies agree that stroke risk stratification in severe CKD patients should be individualised which is essential to guide the selection of the most appropriate thromboprophylactic measures (30, 3 © Schattauer 2012
Read full abstract