Erythropoietin as a cardioprotective agent: down but not out.

Erythropoietin (EPO) is an important regulator of the production red blood cells, and cloning of the gene led to the development of recombinant human EPO preparation, rhEPO. EPO caught the public eye when some leading athletes started to take it to increase the oxygen-carrying capacity of the blood and, thereby, to achieve enhanced athletic performance. Thus, they were conveniently able to achieve the polycythaemia that others acquired by living at high altitude. Better athletic performance did, however, comeat a serious price for some, as reports of increased heart attacks and strokes filtered in, so that this useofEPOfor enhanced sportsperformance is now banned. In contrast, the legitimate use of rhEPO as a cytokine in the treatment of anaemias of chronic renal disease or cancer has improved the quality of life of those seriously ill persons. The questions evaluated in this Editorial are whether the benefits of rhEPO extend beyond its haematological use for treatment of these chronic anaemias and whether it has cardiovascular protective qualities. Indeed, EPO is also clinically used for the treatment of the anaemia of chronic heart failure. However, strong experimental data do not translate into simple clinical benefit. Rather, despite increasing the haemoglobin, there are potential side effects that include worsening hypertension, thrombotic events and endothelin activation. The ever-present problem is whether the benefits of therapeutic rhEPO more than balance the side effects that have surfaced. In this regard, the present study by Ludman et al published in Heart is highly relevant (see page 1560). The experimental background is that EPO is a promoter of cardiac protective paths such as the reperfusion injury salvage kinase (RISK) and Janus Kinase/ Signal Transducer and Activator of Transcription factor (JAK-STAT) paths. 5 Thus, EPO should be able to promote protection from reperfusion injury, thereby slotting into the list of other protective modalities that act by preconditioning or postconditioning. To follow the logical basis for the Ludman study requires a deviation into lethal ischaemice reperfusion injury and protection from it by preconditioning or postconditioning. Although described as early as 1979 by Rona’s group, there was initial disagreement on the significance of reperfusion injury. Many researchers then believed that the injury merely reflected accelerated damage that would have occurred in any case had there been no reperfusion. Others proposed that there was a genuine added injury caused by reperfusion, as argued in my review in Circulation in 1989. Since then, Hausenloy and Yellon have emphasised the clear data for the entity of clinically relevant reperfusion injury, which can cause up to about 40% of the reperfused cells to die. This phenomenon is now termed lethal ischaemicereperfusion injury (IRI). The basic concept underlying preconditioning rests on the completely illogical finding that briefly and repetitively depriving the heart of its blood supply could protect and not damage the heart from further episodes of sustained ischaemia. In 1986, Robert Jennings and his group wrote the landmark article that described such protection as preconditioning. Thereafter, intense research has led to the establishment of at least two molecular pathways. The first and best studied is the RISK pathway, 12 and second is the novel survival activating factor enhancement (SAFE) pathway. These are not conflicting paths but are interactive and mutually supportative. As the basic principles of preconditioning and postconditioning are similar (box 1), and postconditioning is more readily tested in patients undergoing percutaneous coronary intervention (PCI) for acute coronary syndromes, the principles of postconditioning are the basis of the present study. In experimental postconditioning, repetitive short bursts of ischaemia at the onset of reperfusion limit the ultimate infarct size. Mechanistically, postconditioning activates the RISK and the SAFE paths. Since the first description of experimental postconditioning by the group of Vinten-Johansen in 2003, the jump from basic to human studies in only 2 years in 2005 by the group of Ovize 17 was very rapid and spoke strongly to clinicians. As a generalisation, almost every intervention that has been shown to be true in modifying rodent lethal IRI has also been tested in humans, generally with a positive result. In the particular case of rhEPO, there have been good background experiments. For example, in a major basic science study, EPO, administered at the point of reperfusion, reduced infarct size in an isolated perfused rat heart in an ERKand PI3-kinase-dependent manner. Despite the expectation that EPO should work when tested in humans, the present study unexpectedly showed that postconditioning by EPO failed to reduce infarct size in patients undergoing primary PCI. More seriously, and unexpectedly, EPO treatment doubled the incidence of microvascular obstruction (82% EPO vs 47% placebo; p1⁄40.02) and significantly increased indexed left ventricular (LV) enddiastolic and end-systolic volumes as well as LV mass index. Microvascular obstruction increased from 47.4% to 81.8% (p1⁄40.020). These results are a major disappointment. Published almost concurrently with the present report, the results of a similar study were also disappointing in that there was an overall neutral outcome with a negative result in the subgroup of those aged over 70 years. The failure of rhEPO to influence outcome in the present Ludman study certainly is in conflictwith the basic science data that provide strong evidence for the cardioprotective qualities of EPO. Specifically, there was up to 50% reduction in myocardial infarct size in rat, rabbit anddog models (but not sheep) when EPO was administered starting at the time of reperfusion. In pigs, although the infarct size was not reduced, there was a limited decrease in interstitial fibrosis, increased capillary area and regional functional improvement in darbepoetin-treatedpigs. EPO is able to lessen lethal IRI in the rodent heart by activating the RISK path. 19 Furthermore, EPO can experimentally up-regulate endothelial nitric oxide synthase (eNOS) activity in coronary artery endothelial cells in vitro and in vivo, thereby enhancing production of protective nitric oxide. However, EPO did fail to decrease infarct size in pigs, Correspondence to Lionel H Opie, Founding Director and Professor Emeritus, Hatter Cardiovascular Research Institute, Faculty of Health Sciences, University of Cape Town and Groote Schuur Hospital, Observatory, Cape Town 7925 South Africa; [email protected]