Renal dysfunction in heart failure : insights on prevalence and prognosis

Abstract

Background: Kidney disease is common in heart failure (HF) and has been found to be associated with worse outcomes. The impact of different degrees of chronic kidney disease (CKD) in HF, as well as the link to, and impact of CKD in HF with reduced (HFrEF), the newly defined mid-range (HFmrEF) and preserved ejection fraction (HFpEF) have been uncertain. Studies of worsening renal function (WRF) in the various EF groups are lacking. HF treatment of HFrEF is well defined, while we lack knowledge of the effect of heart failure treatment in HFpEF, HFmrEF and in patients with HF and CKD.

Aims: 1. To examine prevalence and prognostic impact of different degrees of kidney dysfunction in unselected HF patients; 2. To perform a comprehensive comparison of CKD in HF with HFpEF, HFmrEF and HFrEF with regards to prevalence, clinical correlates and prognosis; 3. To examine the risk for and impact of WRF in HFpEF, HFmrEF and HFrEF; 4. To analyze the association between mineralocorticoid receptor antagonist (MRA) treatment and outcome in patients with myocardial infarction (MI) and HF in relation to EF groups and CKD.

Prevalence and prognostic impact of kidney disease in heart failure: We studied 47,716 patients in the Swedish Heart Failure Registry (SwedeHF) 2000-2013. Patients were divided into five renal function strata based on estimated glomerular filtration rate (eGFR). 51% of the patients had eGFR < 60 ml/min/1.73 m2 and 11% had eGFR < 30 ml/min/1.73 m2. The mortality risk increased with decreasing eGFR and persisted after adjusting for differences in baseline characteristics, severity of heart disease and medical treatment.

Associations with and prognostic impact of CKD in HFpEF, HFmrEF and HFrEF: Of 40,230 patients with measured EF in SwedeHF, 22% had HFpEF, 21% had HFmrEF, and 57% had HFrEF, with a CKD prevalence of 56%, 48%, and 45%, respectively. Associations between covariates and CKD were similar in all EF groups. There was higher mortality in all EF groups in patients with CKD. After adjustment, CKD was more strongly associated with death in HFrEF and HFmrEF than in HFpEF.

WRF in different EF categories: After merging the SwedeHF registry with the laboratory data in Stockholm Creatinine Measurement (SCREAM) database, 7,154 patients in Stockholm between 2006-2010 were studied. After discharge, the risk for WRF was higher in HFpEF than in HFmrEF and HFrEF. Variables related to more severe HF were predictive of WRF. WRF within year one after the index-HF event was strongly associated with long-term mortality, but in HFpEF only with the most severe WRF.

Outcome in MI patients with HF with or without MRA treatment: Patients with MI and HF registered in the Swedish national myocardial infarction registry, SWEDEHEART, between 2005-2014, were studied. Of 45,071 patients with MI and HF, 10% were treated with MRA. Patients with reduced EF < 40% were more often treated with MRA compared to mid-range EF 40-49% and normal EF > 50%. Of patients with CKD, 9% received MRA. After adjustment, MRA use was associated with a lower mortality in patients with EF < 40% but not with EF > 50% while the association between MRA use and outcome was similar regardless of presence or not of CKD.

Conclusions: In unselected HF patients, half of the patients have at least moderate renal dysfunction. There is a strong graded association between renal dysfunction and both short- and long-term outcome. CKD is slightly more common in patients with HFpEF but is associated with similar covariates regardless of EF. CKD is strongly associated with mortality regardless of EF group, although less strongly in HFpEF than in HFmrEF and HFrEF. The long-term risk of WRF is high in HF and especially in HFpEF. WRF within one year of discharge is a strong negative prognostic factor in all EF groups during long term follow-up, although in HFpEF only in those with the most severe WRF. In patients with MI and HF, MRA treatment is associated with better long-term survival in patients with reduced but not with preserved EF, while the association between MRA use and outcome seems to be similar regardless of presence or not of CKD.