Red cell distribution width in chronic heart failure: a new independent marker for prognosis?

Abstract

Chronic heart failure (HF) is often accompanied by high co-morbidity and mortality and remains one of the leading causes of death in the western world. The syndrome differs greatly between younger and elderly individuals in many respects, such as prevalence, aetiology, symptomatology, pathophysiology, and comorbidity. Owing to this heterogeneity, there is a need for multiple objective parameters to facilitate diagnosis, prognosis, and therapy. Over the last decade, the inclusion of the natriuretic peptides, brain natriuretic peptide (BNP) and N-terminal pro-brain natriuretic peptide (NT-proBNP), as biomarkers in the management of HF has greatly improved our ability to diagnose and make prognostic prediction in HF patients.1 However, their added benefits in guiding therapy remain controversial, especially among the elderly. In addition, elevation of these peptides occurs in many clinical conditions. Therefore, natriuretic peptides are not always sufficient, in particular when they are increased to only a moderate level. Other biomarkers have been identified, such as inflammatory cytokines, high-sensitivity C-reactive protein, natriuretic peptides, neurohormones as well as markers of remodelling and oxidative stress, which could be beneficial for diagnosis and prognosis. Recently, elevated red cell distribution width (RDW) has been identified and proposed to be of importance.2 Red cell distribution width is a measure of the variability in size of circulating erythrocytes and is expressed as the coefficient of variation of the erythrocyte volume. As several routine haematology instruments can analyse erythrocyte volume, RDW is available in most clinical settings. It has been known for some time that elevated RDW predicts poor outcome in both diseased and normal populations.3,4 It is therefore not surprising that RDW can also predict mortality in patients with HF. Three studies published recently in this Journal have shown that elevated RDW predicts mortality in HF independently of BNP or NT-proBNP, thus adding RDW to our repertoire of prognostic markers in patients with HF. However, this remains to be confirmed in a larger population study. Moreover, to serve as a prognostic marker is one aspect and to function as a biomarker for improved therapy is another. In combination with other outcome predictors, the clinical use of RDW will be greatly enhanced if it can be used to guide effective therapy. In addition, there is a low relationship between high RDW and mortality, limiting its predictive value for the individual patient. Finally, the value of RDW is instrument-dependent, forcing each laboratory to establish its own reference values. It is therefore important to identify the underlying mechanism that links high RDW to poor prognosis in HF and to examine whether the connection is upstream or downstream of the failing heart. A common underlying cause of high RDW is iron or B12/folate deficiency, where normal erythrocytes are mixed with smaller or larger ones produced during the deficiency. A similar increase in RDW occurs during iron and B12/folate replacement therapy when the reticulocyte count increases. Red cell distribution width is also increased following blood transfusions, as well as in haemolytic anaemia5 and thrombotic conditions6,7 where erythrocytes are fragmented in the circulation. Elevated RDW is also associated with liver disease,8 alcohol abuse,9,10 inflammatory conditions,11 and renal disease,12 but here the mechanism behind the variation in erythrocyte size is more complex. In the February 2010 issue of the European Journal of Heart Failure, Januzzi and co-workers13 presented findings of RDW in acute HF. They concluded that RDW seems to carry added prognostic information above NT-proBNP. Together with two studies published in the December issue,14,15 these studies in acute HF and RDW clearly demonstrate that RDW is independent of anaemia. The study from Januzzi and co-workers indicates, perhaps unexpectedly, that RDW is not linked to inflammation, iron, or B12/folate deficiency or to blood transfusion, although the number of patients actually analysed for these parameters was low. In contrast, another recently published study indicates that markers of iron deficiency, inflammation, and renal impairment all correlated well with high RDW in a population with HF,16 indicating that RDW could in fact be a non-specific marker of many different conditions, all of which can result in a worsening of the HF prognosis. The CHARM study revealed a strong correlation between elevated bilirubin and poor outcome in HF, suggesting that haemolytic processes could be one link between elevated RDW and poor prognosis.17 Unfortunately, none of these studies analysed the reticulocyte count, which in combination with fragmented erythrocytes increases RDW during haemolysis. However, the fact that elevated haptoglobin is linked to poor prognosis in HF18 makes this explanation unlikely, since even minute levels of haemolysis result in a drastic lowering of haptoglobin. The correlation with bilirubin could also be due to liver damage and excessive alcohol intake, resulting in macrocytosis and increased RDW.9 High alcohol intake is linked to poor prognosis in most conditions. However, no correlation between liver damage markers such as alanine aminotransferase, aspartate aminotransferase, and alkaline phosphatase was found in this study or the study from Januzzi and co-workers, making this explanation unlikely. It would, however, be interesting to re-examine these patient data using more specific markers such as carbohydrate deficient transferrin, which increases in relation to alcohol intake irrespective of liver damage. Clearly, great uncertainty remains regarding the underlying mechanism of RDW in the context of HF. In this respect, there are several fundamental issues that need to be addressed: how is RDW affected in HF when compared with other diseases? Is RDW equally affected among different forms of HF and between acute HF and chronic HF? Can RDW be affected by HF medications? Owing to different aetiologies of RDW in various diseases, much work is required to clarify its physiological role at different stages of HF. In this regard, we are still in the starting blocks.