Role of erythropoietin in acute kidney injury: what does the evidence say?

2598 2
Since the discovery of Erythropoietin in 1977 and subsequent cloning of the gene in 1985, recombinant Erythropoietin (EPO) has been widely used in the management of anemia due to cancer, chemotherapy, and chronic kidney disease.
Although the endocrine effects of EPO in stimulating maturation and differentiation of erythroid precursor cells in the bone in response to hypoxia is widely known, the anti-apoptotic and anti-inflammatory properties particularly in the critical organs such as kidney, heart and brain were discovered only in the last decade.
Researchers from the United States demonstrated that EPO receptors are expressed in the renal tubular epithelial cells, proximal tubular cells, and mesangial cells. EPO induced activation of these receptors leads to activation of Janus activated kinase 2(JAK-2) pathway which in turn stimulates several other signaling pathways (MAPK,NFKB,STAT 3/5), all of which promote anti-apoptotic and proliferative proteins that increase cell survival (read this article).
With this knowledge, researchers from Australia demonstrated that EPO administration both in vitro and in vivo in doses up to 5000 U/kg in animals with ischemic AKI hastened cell recovery and prevented cell death. Several other researchers have demonstrated this effect in different injury models (hypoxia, ischemia-reperfusion, nephrotoxins, sepsis etc).
As would be expected, these promising results in animal studies led to a randomized, double blind, placebo controlled trial (EARLYARF) in 529 ICU patients who were identified at risk of AKI by urinary biomarkers(c-glutamyl transpeptidase and alkaline phosphatase ≥46.3). Administration of 100,000 units of EPO IV in 2 divided doses over 24 hrs did not reduce the risk of developing AKI, quite contrary to what was seen in animal studies.
In my opinion, likely causes for this discrepancy included:
1) Dosing in animals had been much higher (equivalent to up to 350000 units in humans).
2) Optimal timing of EPO administration might be an issue (within 6 hrs of injury)
3) Unclear, if the ideal urinary biomarkers were used.
Further studies with proper dosing and timing of EPO administration and early identification of patients with risk of AKI using ideal urinary biomarkers may answer this question best. Besides, the utility of administering EPO to ICU patients to reduce transfusions has already been refuted, is not cost effective and may be associated with 40% increased risk of thrombotic events (read this blog). This leaves us with no good indication to consider EPO in critically ill patients with AKI (not ESRD).
If you are a big believer of basic research, you may still want to consider EPO for its favorable actions, but with recent negative publicity EPO has been receiving, it appears like the clock is quickly moving towards “may be not or even no”.


  1. Thanks for the comment.
    3966 patients were screened and 529 were enrolled. There were 84 patients in the EPO arm, 78 placebo and 345in the observation arm.

    As far as adverse effects were concerned, no major complications noted. Authors concluded that they were not adequately powered to study this.

  2. Nice post.

    Regarding EARLYARF, if you screen 529 patients but only randomize 129, you should refere to the study size as 129, not 500+.

    Also, from the abstract, it is not clear if the study cohort suffered much actual ARF. The authors seem to hedge on this. If the coohort did not have much actual AKI, it would decrease the power of the study.

    It is interesting that despite the very high doses of epo, they didn't have an increase in adverse effects that are found in other ICU-associated anemia studies. Did the authors address this?

Leave a Reply