Abstract
Background: Patients with kidney disease are more likely to develop atrial fibrillation (AF) than individuals with normal renal function, and more likely to suffer ischemic stroke (IS)/thromboembolism (TE). We investigated the relationship of kidney function evolution to IS/TE, mortality and bleeding in AF patients.
Methods: In a cohort of 8962 AF patients, 2653 had serum creatinine data, with 10894 patient-years of follow-up. Patients were stratified into quartiles of estimated glomerular filtration rate (eGFR) evolution (in mL/min per 1.73 m2/year).
Results: Rates of events (IS/TE, bleeding, mortality) increased with worsening eGFR by quartiles. The risk of events was particularly increased when patients in the 4th quartile were compared to others. Renal impairment per se was not an independent predictor of IS/TE but was an independent predictor of bleeding, whilst eGFR worsening was an independent predictor both for IS/TE (Hazard Ratio [HR] 1.573, 95%CI 1.160-2.134 for patients in the last quartile) and for bleeding events (HR 1.543, 95%CI 1.157-2.004). Worsening eGFR did not improve the predictive ability of the CHA2DS2VASc and HAS-BLED scores for identifying a higher risk of IS/TE or bleeding events, respectively. When the benefit of IS reduction was balanced against the increased risk of bleeding events, the net clinical benefit was positive in favor of OAC use (vs non-use) in patients with worsening eGFR.
Conclusions: Rates of IS/TE, mortality and bleeding increased with worsening eGFR >4.81 mL/min per 1.73 m2. Worsening eGFR was an independent predictor of IS/TE and of bleeding, and a better predictor of IS/TE than renal impairment in AF.
Methods: In a cohort of 8962 AF patients, 2653 had serum creatinine data, with 10894 patient-years of follow-up. Patients were stratified into quartiles of estimated glomerular filtration rate (eGFR) evolution (in mL/min per 1.73 m2/year).
Results: Rates of events (IS/TE, bleeding, mortality) increased with worsening eGFR by quartiles. The risk of events was particularly increased when patients in the 4th quartile were compared to others. Renal impairment per se was not an independent predictor of IS/TE but was an independent predictor of bleeding, whilst eGFR worsening was an independent predictor both for IS/TE (Hazard Ratio [HR] 1.573, 95%CI 1.160-2.134 for patients in the last quartile) and for bleeding events (HR 1.543, 95%CI 1.157-2.004). Worsening eGFR did not improve the predictive ability of the CHA2DS2VASc and HAS-BLED scores for identifying a higher risk of IS/TE or bleeding events, respectively. When the benefit of IS reduction was balanced against the increased risk of bleeding events, the net clinical benefit was positive in favor of OAC use (vs non-use) in patients with worsening eGFR.
Conclusions: Rates of IS/TE, mortality and bleeding increased with worsening eGFR >4.81 mL/min per 1.73 m2. Worsening eGFR was an independent predictor of IS/TE and of bleeding, and a better predictor of IS/TE than renal impairment in AF.
| Original language | English |
|---|---|
| Pages (from-to) | 39-45 |
| Journal | American Heart Journal |
| Volume | 198 |
| Early online date | 30 Dec 2017 |
| DOIs | |
| Publication status | Published - 1 Apr 2018 |