Cardiovascular Outcomes in Dialysis Patients: One Size Does Not Fit All

Brendan Bowman; Emaad M. Abdel-Rahman


Eur Heart J. 2019;40(11):899-901. 

In This Article


Cardiovascular disease (CVD) is the largest contributor to mortality in patients with the most severe form of kidney disease, end-stage kidney disease (ESKD).[1] While this fact is widely understood, the mortality risk of the nearly half million US dialysis patients is truly hard to fathom: 166 per 1000 patient-years, a rate surpassing most metastatic cancers.[1] Nearly half of this stunning mortality rate is attributed to CVD (Figure 1).[1] Despite significant advances in our understanding of the pathology and treatment of CVD in the general population, little has been shown to improve the cardiovascular risk profile of the dialysis patient, aside from transplantation.[2] Complicating matters, in ESKD patients, the leading cause of CVD-related death is lethal arrhythmias/sudden cardiac death, not coronary artery disease/myocardial infarction.[1] Recent studies of implantable loop recorders in ESKD have shown that most fatal arrhythmias appear to be bradycardic arrhythmias or pulseless electrical activity in nature, explaining the difficulty in demonstrating a mortality benefit of implantable cardiac defibrillators.[3]

Figure 1.

The staggering toll of cardiovascular disease in ESKD.

Multiple factors account for the poor state of knowledge regarding cardiac complications in ESKD. Patients requiring dialysis are some of the most medically complex patients cared for in any health system. As such, they are typically excluded from the large prospective trials from which the cardiology field has rightfully earned acclaim. In the rare instances where a sizeable randomized controlled trial has been undertaken, results have generally been mixed. One example, the Deutsche Diabetes Dialyse (4 D) trial, randomized 1255 patients with < 2 years of dialysis, diabetes, and abnormal lipid profiles to atorvastatin vs. placebo and followed the patients for 3 years of therapy. Despite marked lipid profile improvement, no significant difference was seen in the composite cardiovascular outcome.[4]

If traditional risk factors are, at minimum, less predictive of cardiovascular outcomes in ESKD, are there non-traditional risk factors that have been identified? Over the years, many have been suggested. Chronic anaemia, hormonal derangements such as elevated levels of fibroblast growth factor 23, and, more recently, recurrent myocardial stunning during dialysis, to name just a few, have all been posited as non-traditional CVD risk factors in dialysis patients. However, there are no large randomized controlled trials of a successful intervention to reduce CVD in dialysis to date. This leaves the nephrologist and cardiologist with an evidence base of retrospective analyses subject to confounding, observational trial data, and small underpowered prospective trials.[5]

The limited interventional trials in CVD and ESKD proceed from the assumption that patients share a homogenous disease where interventions will be generalizable to the cohort as a whole. This belief persists despite a growing body of literature suggesting that ESKD is a heterogeneous mix with differing implications for health. Komatsu et al.,[6] showed that patients with ESKD due to immunoglobulin A (IgA) nephropathy had favourable outcomes.[6] In contrast, Chen et al. compared clinical data of patients with diabetic nephropathy (DN) on haemodialysis (HD) vs. non-DN on HD.[7] They showed that the incidence of hypertension, coronary heart disease, and cerebral thrombus in DN patients was higher than those in non-DN patients (P < 0.001). More recently, Al-Thani et al. confirmed that patients with diabetes mellitus and lupus nephritis (LN) undergoing HD had a significantly higher vascular disease burden and mortality compared with non-diabetic patients on HD.[8] Levy et al. demonstrated higher mortality of patients with LN on HD when compared to controls with polycystic kidney disease, further suggesting that outcomes of dialysis patients may vary based on the initial cause of the kidney disease.[9]

Narrowing focus to glomerular disease, a recent New Zealand study examined patients with biopsy-proven primary glomerulonephritis (GN) with the goal of estimating the long-term risk of ESKD and death based on GN subtype.[10] The authors demonstrated that patients with rapidly progressive GN, anti glomerular basement membrane (anti-GBM) disease, and membranous nephropathy had the highest mortality risk, while patients with minimal change disease had the lowest. Though the study did not delineate cause of death in these patients, it highlighted that the aetiology of GN is a major risk factor for the development of ESKD/death.

In an earlier paper, the authors of the current study in this issue examined a smaller data set extracted from the US Renal Data System (USRDS) of 84 301 adult patients with incident (1996–2011) ESKD secondary to different subtypes of GN.[11] They aimed to examine outcome differences among GN subtypes with respect to demographic and clinical attributes at declaration of ESKD and prognosis once ESKD therapy was initiated. The authors found a significant survival discrepancy in outcomes varying by aetiology of ESKD and GN, with patients with DN and LN having the highest mortality, and patients with IgA nephropathy having the lowest.