Soon after William Harvey described the circulation of blood in the early 17th century, Christopher Wren unsuccessfully experimented with replacing wine for blood. During a cholera outbreak in the late 19th century, Gaillard Thomas also failed in his attempts to substitute milk for blood. It was not until 1933 that the first successful blood substitution experiments were performed. William Ruthrauff Amberson of the University of Tennessee Medical School used hemolyzed red blood cells for an exchange transfusion in a cat and demonstrated that he could keep the animal alive for 36 hours. Unfortunately, infusion of a similar product into humans produced oliguria and bradycardia. These toxicities were initially thought to be due to the lipids within erythrocyte membranes. Yet, when Dr. Amberson infused hemoglobin free of red-cell membranes into a hemorrhaging post-partum woman who had depleted her hospital’s inventory of cross-matched blood, she developed bradycardia and hypertension, and ultimately died from renal failure. These pioneering studies that spanned three centuries have demonstrated that the ideal blood substitute is an elusive goal.

Today, the blood supply in the United States is safe and usually sufficient to meet the needs of our patients. However, the current system does have two large fundamental problems. First, our reliance on altruistic blood donation creates seasonal shortages during the summer and winter holidays. Second, evolving and emerging infections will always endanger the safety of transfused human products. In theory, a red blood cell substitute would obviate both of these issues. A high-quality substitute would also be helpful for use in patients who are difficult to cross match, for individuals who will not accept transfusions of human products, and for emergency infusions at the scenes of trauma (civilian and military).

Most modern blood substitutes are either perfluorocarbons or hemoglobin-based oxygen carriers. Perfluorocarbons are non-water soluble, biologically inert, artificial, organic fluids with a high solubility for oxygen. Gas molecules are not chemically bound to perfluorocarbons, but instead are absorbed and released by simple diffusion. A large phase III trial using the perfluorocarbon, Oxygent, was halted early because of an increase in stroke rates in patients who were undergoing cardiopulmonary bypass. Hemoglobin-based oxygen carriers have held more promise. Four different methods have been used to avoid the toxicities induced by free hemoglobin: 1) cross-linking of the alpha chains, 2) polymerization of the hemoglobin chain tetramers, 3) conjugation of the hemoglobin to a larger molecule such as polyethylene glycol, and 4) encapsulating hemoglobin within liposomes. Since 1996, at least a dozen trials have been performed that analyzed the utility of these agents in a variety of clinical settings.

In Brief

In a systematic review of the available literature on purified hemoglobin-based blood substitutes published since 1980, Natanson, et al. identified 70 trials, focusing only on the 16 randomized controlled trials involving 3,711 patients who received one of five cell-free hemoglobin products. One product, HemAssist, was cross-linked hemoglobin; three products, Hemopure, Hemolink, and PolyHeme, were polymerized hemoglobin; and one product, Hemospan, contained hemoglobin conjugated to polyethylene glycol. Disappointingly, but not surprisingly, the use of any of these products was associated with an almost three-fold increased risk of myocardial infarctions. Overall mortality was only mildly worse (relative risk 1.30) in subjects exposed to the blood substitutes. Further analysis did not indicate that any one hemoglobin product or any one indication for therapy was particularly worse than any of the others. These results demonstrate that the use of the available cell-free hemoglobin products is associated with too much morbidity, and an unacceptable rate of morbidity, to be of any clinical benefit.


Natanson C, Kern SJ, Lurie P, et al. Cell-free hemoglobin-based blood substitutes and risk of myocardial infarction and death: a meta-analysis. JAMA. 2008. [Epub ahead of print]

Competing Interests

Dr. Abrams indicated no relevant conflicts of interest.