Author + information
- †Baylor University Medical Center, Dallas, Texas
- ‡Baylor Heart and Vascular Institute, Dallas, Texas
- §Baylor Jack and Jane Hamilton Heart and Vascular Hospital, Dallas, Texas
- ‖The Heart Hospital, Plano, Texas
- ↵∗Reprint requests and correspondence:
Dr. Peter A. McCullough, Baylor Heart and Vascular Institute, 621 North Hall Street, H030, Dallas, Texas 75226.
Bleeding in patients with chronic kidney disease (CKD) or “uremic bleeding” is a well-recognized complication. It has been described as far back as 1907 by Reisman (1), classically as a complication of Bright disease. The continued debate over the mechanism of this disease only highlights the complicated biochemical milieu that is present in the blood of patients with CKD. In this issue of the Journal, Bonde et al. (2) use the Danish national registries to explore which direction benefit risk is tipped in patients with CKD receiving chronic anticoagulation. There are at least 92 known uremic retention solutes (3), all of which likely play some role in the body’s essential biochemical functions, and many of which may play a role in the coagulation cascade and platelet function. A number of factors have been identified in CKD patients that promote as well as inhibit coagulation. The untreated patient with CKD is generally more prone to thrombosis; however, when treated with antiplatelet agents, heparin, warfarin, and novel anticoagulants, the same patient is more prone to bleed than is a patient with normal renal function. Figure 1 is a detailed, but not comprehensive list of the commonly identified processes involved (4–7). This complex biologic balance and the lack of adequately powered, randomized controlled trials help to explain the equipoise in recommendations on anticoagulation in patients with atrial fibrillation (AF) and CKD. In 2011, the Kidney Disease Outcomes Quality Initiative committee changed from recommending anticoagulation for hemodialysis patients with AF to recommending anticoagulation only as secondary prevention and with close monitoring (8).
The investigators use an individual-level linkage of data from several Danish national patient registries to examine the outcomes of all patients discharged with nonvalvular AF from 1997 to 2011. A total of 154,259 patients were identified; 4,419 with pre-dialysis CKD, and 1,142 receiving renal replacement therapy. They examined 4 endpoints: composite of death or hospitalization from stroke or bleeding; composite of fatal stroke or fatal bleeding; cardiovascular death; and all-cause mortality. The study found anticoagulation in patients with CHA2DS2-VASc (Congestive heart failure; Hypertension; Age ≥75 years; Diabetes mellitus; previous Stroke, transient ischemic attack, or thromboembolism; Vascular disease; Age 65 to 74 years; Sex category) score >2 to be associated with better outcomes at all levels of renal function. In pre-dialysis patients, there was a reduction in composite endpoints of fatal stroke or fatal bleeding, cardiovascular death, and all-cause mortality. In renal replacement therapy patients, however, there was only a reduction in all-cause mortality, suggesting possible confounding by indication and not a therapeutic effect of stroke prevention.
Whereas this work has significant limitations inherent to cohort studies, it is among the largest and most complete datasets analyzing this complex management question. The investigators acknowledge that even with rigorous review of the population to ensure the cohorts were matched, it is impossible to eliminate some confounding factors. Interestingly, the overall rates of warfarin use were surprisingly low. Among those with no CKD, the rate of warfarin use was 23.6% despite 79.2% having a CHA2DS2-VASc score of 2 or higher. For those with CKD and end-stage renal disease, warfarin was prescribed in fewer than 20% of patients. It is not known how many patients were taking dabigatran, rivaroxaban, or apixaban, and thus a treatment effect from these agents cannot be excluded. Certain assumptions were made due to incomplete data. Elements of the CHA2DS2-VASc score, including hypertension, diabetes, and heart failure, were compiled based on available filled prescription data. Because as little as a 1-point difference in the CHA2DS2-VASc score can change treatment recommendations, these assumptions carry great weight in the analysis. It is likely that this method underestimates patients’ risk and lowers the CHA2DS2-VASc score in the database. With this limitation in mind, doctors must be very cautious in their interpretation of the reported mortality benefit found in pre-dialysis patients with a CHA2DS2-VASc of >1.
The results of this study suggest that at any CHA2DS2-VASc score, the risk of stroke or embolism is roughly double that of non-CKD patients. In all groups there was a large (4-fold or more) increase in the rate of bleeding in those in CHA2DS2-VASc ≥2 with a smaller step up for CKD and end-stage renal disease. Thus, one would conclude on balance that there is a greater opportunity for benefit than harm when the stroke risk is higher in AF in patients with background renal disease. This study provides powerful data about the role of anticoagulation in CKD; however, additional data are needed in CKD patients with AF to better understand the role of antiplatelet agents, warfarin, and the novel anticoagulants. In addition, the optimal approach to AF and stroke prevention with end-stage renal disease is still yet to be defined. For now, these data are helpful to physicians in choosing warfarin in both advanced CKD patients and those on dialysis when the CHA2DS2-VASc score is calculated.
↵∗ Editorials published in the Journal of the American College of Cardiology reflect the views of the authors and do not necessarily represent the views of JACC or the American College of Cardiology.
Dr. Wheelan owns stock in Medtronic and Johnson & Johnson; receives lecturing fees from Medtronic, Pfizer, and Boehringer Ingelheim; and receives research support from Medtronic and Boston Scientific. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- American College of Cardiology Foundation
- Bonde A.N.,
- Lip G.Y.H.,
- Kamper A.-L.,
- et al.
- Muntner P.,
- He J.,
- Astor B.C.,
- Folsom A.R.,
- Coresh J.