Red Cell Distribution width has Incremental Prognostic Value to B-Type Natriuretic Peptide in Acute Heart Failure

Abstract

Red blood cell distribution width (RDW) is a measure of the variation in cell volume within the circulating erythrocyte population. Red blood cell distribution width is available routinely with every complete blood count (CBC) test and is most commonly used for differentiating between different types of anaemia. In recent years, an alternative role for this haematological parameter has emerged—as a potential prognostic marker. The independent prognostic significance of RDW in acute heart failure has recently been reported.1 This novel biomarker has previously been shown to be a strong, independent predictor of prognosis in patients with chronic heart failure.2 However, neither of these studies included the B-type natriuretic peptides (BNPs) in their respective multivariable analyses. It remains unknown whether or not RDW provides incremental prognostic information to these ‘gold standard’ biomarkers in acute or chronic heart failure. Our aim was to evaluate the prognostic value of RDW in patients with acute heart failure. In particular, we studied whether or not this novel biomarker provides additional prognostic information to BNP. All admissions to the Royal and Western Infirmaries in Glasgow with acute heart failure were screened prospectively from 1 December 2006–20 July 2008. Patients were eligible if they were at least 18 years of age and able to provide written informed consent. Both acute decompensation of chronic heart failure and new onset acute heart failure were included. Patients with reduced and preserved systolic function heart failure were included. Inclusion criteria included satisfying the European Society of Cardiology (ESC) definition of acute heart failure3 and having an elevated BNP (>100 pg/mL). Exclusion criteria included acute coronary syndromes complicated by pulmonary oedema, cognitive impairment, and serious concurrent systemic disease resulting in reduced life expectancy. The study protocol was approved by the Local Research Ethics Committee, and all patients provided written informed consent. Blood samples were obtained on the first day of admission and processed immediately. Plasma BNP was measured using the Architect Assay (Abbott Laboratories, Abbott Park, IL, USA). Components of the CBC (RDW, haemoglobin, white cell count, and lymphocytes) were measured using the XE-2100 analyser (Sysmex, Japan). Anaemia was identified using the World Health Organization criteria.4 Categorical variables are expressed as number with the percentage. Continuous variables are expressed as mean with standard deviation (SD) and median with inter-quartile range (IQR) when the distribution is normal and non-normal, respectively. B-type natriuretic peptide was positively skewed and analysed as log[concentration] in a continuous fashion. A Kaplan–Meier survival curve was drawn, stratifying patients into two groups depending on whether their RDW was elevated (defined as RDW ≥14.5%) and a log-rank value was calculated to assess statistical significance. Cox proportional hazard models were used to estimate hazard ratios and 95% confidence intervals (CIs) for individual components of the CBC univariately and then together in a multivariable analysis with age, sex, and BNP. A P-value of <0.05 was considered statistically significant. We prospectively recruited 707 patients with acute heart failure.5 Baseline characteristics are shown in Table 1. Median follow-up was 421 days (range 208–792) and 212 patients died during the follow-up period. Median age was 73 years (IQR 67–80) and 52% were male. Median BNP on admission to hospital was 900 pg/mL (IQR 401–1852). Fifty-seven per cent of the cohort was anaemic. Mean RDW on admission was 15.5% (SD 2.4). Patients with an elevated RDW had a higher unadjusted risk of mortality throughout follow-up period when compared with those with an RDW level within the normal range (Figure 1). Univariate analysis demonstrated lower levels of haemoglobin and lymphocytes, as well as elevated RDW and BNP, to be predictive of adverse outcome. Each percentage increase in RDW was associated with a 9% increase in unadjusted risk of mortality (95% CI 1.04–1.14, P = 0.00013). The same increase in unadjusted risk of mortality was seen for each 1 g/dL decrease in haemoglobin concentration (95% CI 0.85–0.96, P = 0.0017). Red blood cell distribution width remained an independent predictor of all-cause mortality following multivariable analysis with little change in the adjusted hazard ratios (Table 2). B-type natriuretic peptide was a strong independent predictor of death. Intriguingly, as also demonstrated by Pascual-Figal et al.,1 haemoglobin was not an independent predictor of outcome. We recognize limitations with our study. First, some established predictors of risk in heart failure (such as renal function and left ventricular ejection fraction) were not included. On the other hand, our multivariable model included BNP that has arguably been the strongest predictor of outcome in previous heart failure multivariable models. Secondly, inflammatory markers and iron studies (both of which may be closely related to RDW) were not available in this study. We have shown, for the first time, that RDW provides incremental prognostic value to BNP in acute heart failure. The prognostic ability of RDW was independent of haemoglobin concentration, highlighting that RDW is not simply a surrogate for anaemia. The biological basis for this finding in heart failure merits further investigation. Further work is warranted to delineate the potential role of RDW as a tool to guide therapy in heart failure. Conflict of interest: none declared.