Beneficial Effects of Erythropoietin in Models of Shock and Organ Failure-Nothing Is Simple and Easy

It is well-established that erythropoietin (Epo) allows treating anemia and reducing transfusion requirements (1). The fact that Epo may also improve outcome in the critical ill patient (2) suggests that-despite its well-known side effects (2)-it also has therapeutic qualities beyond mere hemopoiesis. In fact, Epo improved organ function in several shock models of various etiology (3), and these beneficial effects were referred to antioxidant and anti-inflammatory properties that were ultimately affiliated with attenuated tissue injury and apoptosis (4). Erythropoietin-induced organ protection was particularly present in I/R injury-like experiments, the most interesting target organs being the brain, the heart, and the kidney. Clinical data support this concept (5, 6). In this issue of Shock, Brendt et al. (7) and di Villa Bianca et al. (8) present two rather intriguing-because fairly contrary-articles on the effects of recombinant human Epo (rhEpo) and the long-acting Epo analog darbepoetin, respectively, during rodent endotoxemia. While di Villa Bianca et al. (8) focussed on the effects of rhEpo (300 U/kg administered 30 min prior as well as 1 and 3 h after the endotoxin challenge) on endothelial dysfunction and vascular reactivity, Brendt et al. (7) tested the hypothesis whether darbepoetin (25 μg/kg given either 24 h before or 2 h after endotoxin) would improve left ventricular function and myocardial tissue integrity in an isolated Langendorff heart preparation. In good agreement with the existing literature, rhEpo reduced the oxidative and nitrosative stress in isolated vessels and nearly completely restored the otherwise marked suppression of the antiapoptotic proteins Bcl-xl and Bcl-2, ultimately resulting in a pronounced improvement of both endothelial vasoconstriction and relaxation (8). In contrast, no matter whether it was administered in a pretreatment (24 h before endotoxin) or posttreatment (2 h after endotoxin) approach, darbepoetin affected neither the endotoxin-related changes in myocardial function, that is, the fall in left ventricular developed pressure, dp/dtmax and dp/dtmin, nor the gene expression of the apoptosis-regulating proteins Bcl-xl, Bcl-2, Bcl-xs, and Bax (7). How can we explain these controversial observations? First of all, one might question the use of darbepoetin per se. In fact, although Epo exerts its hematopoietic effects through a homodimeric Epo receptor complex (EpoR-EpoR), its organ-protective properties are referred to activation of an alternative receptor complex consisting of the EpoR and the common β receptor (EpoR-βcR) (9), stimulation of EpoR-βcR alone being devoid of the undesired Epo side effects (10). Darbepoetin, however, binds to both receptors (11) and thus should have organ-protective potential. Indeed, using darbepoetin, other authors reported reduced tissue damage in various organs (12-16). Second, in particular, because they did not simultaneously investigate rhEpo, it could be argued that Brendt et al. (7) used an inadequate amount of darbepoetin. Clearly, most studies reporting beneficial effects used high-dose (>1,000 IU/kg) Epo (17, 18), but doses as low as 300 IU/kg effectively reduced organ injury after hemorrhage or I/R (19, 20). In fact, the darbepoetin dose used by Brendt et al. (7) (25 μg/kg) was reported to be equipotent to 5,000 IU/kg of Epo (13), and markedly lower doses also attenuated heart, liver, and kidney damage (12, 16, 17). Moreover, optimal protection of the heart against experimental myocardial infarction was even manifested at 2.5 μg/kg only (14). In addition, in patients with chronic heart failure, low-dose (1.5 μg/kg every 20 days) darbepoetin treatment over 3 months improved biventricular function and plasma B-type natriuretic peptide (21). Finally, Brendt et al. (7) provided a "proof of principle" that darbepoetin effectively activated Epo receptors: both reticulocyte count and hemoglobin content increased 14 days after its administration. Third, one might raise the issue of the shock model per se; in fact, both rhEpo and darbepoetin proved to be beneficial in models characterized by an I/R-type of injury, that is, after vascular occlusion and/or hemorrhage and resuscitation (12-15, 18-20). By contrast, rhEpo failed to improve organ function in endotoxic shock. Using markedly higher rhEpo doses (4,000 IU/kg), however, other authors demonstrated renal protection in endotoxic mice (17), and 10 μg/kg darbepoetin, that is, much less than used by Brendt et al. (7), attenuated endotoxin-induced murine liver failure (16). Finally, 400 IU/kg rhEpo 18 h after cecal ligation and puncture improved the microcirculatory perfusion and tissue energy balance in murine sepsis (22). Consequently, it remains open why darbepoetin lacked any beneficial effects in the study by Brendt et al. (7). Thus, what can we conclude from the two controversial reports? Clearly, there is ample evidence originating that rhEpo has protective effects on many tissues and organs during shock states of various etiology. Nevertheless, before starting a clinical trial, there is still substantial need for data originating from clinically relevant large animal shock models, in particular taking in consideration the potential harm even after short-term administration (10). ACKNOWLEDGMENTS This work was supported by the Deutsche Forschungsgemeinschaft (SCHE 899/2-1, SCHE 899/2-2).