The “first-generation” HBOCs were based on observations that cross-linking overcame hemoglobin subunit dissociation and renal toxicity. Experience with these solutions has shown that they can be vasoactive, sometimes increasing blood pressure, sometimes decreasing tissue perfusion, and sometimes both. Clinical trials were disappointing because of toxicity. A first-generation HBOC diaspirin cross-linked Hb (DCLHb, HemAssist) circumvented the safety concerns from dimerization of the hemoglobin tetramer by cross-linking the alpha chains chemically. However, clinical studies performed with human cross-linked hemoglobin (DCLHb) were stopped because of an increased mortality in two clinical trials in patients who received DCLHb after stroke and multiple injuries [34, 35]. Additional studies in cardiac and noncardiac surgery documented additional safety concerns with early study termination related to serious adverse events [36–38].

The “second-generation” HBOCs were based on a better understanding of the mechanisms of this vasoconstriction and specific modifications to reduce nitric oxide binding and resultant vasoconstriction. Hemopure is a polymerized bovine hemoglobin product with a p-50 of 30 mmHg that is closer to human hemoglobin than stroma-free hemoglobin. When infused, these linked hemoglobin molecules circulate in the plasma, are smaller, and have a lower viscosity and more readily release oxygen to tissues than allogeneic red blood cells [39]. A similar bovine hemoglobin substitute is used in veterinary medicine as Oxyglobin (HBOC-301), approved to treat canine anemia in 1998. Hemopure was approved for the treatment of perioperative acute anemia, with the intention of eliminating or reducing the need for allogeneic RBC transfusion, in elective adult surgical patients in South Africa in 2001. However, in the United States, phase III trials with Hemopure were put on hold due to safety issues. The main reason was the adverse effect profile of the compound, in particular an increase in the risk of stroke and myocardial infarction. Hemopure can be used for compassionate use for the treatment of severe anemia in Jehovah’s Witness patients but requires patient consent, institutional review board approval, and FDA emergency IND approval for use in individual patients.

Sanguinate is a bovine PEGylated carboxyhemoglobin developed to combine the beneficial functions of a carbon monoxide-releasing molecule (to promote antiinflammatory and anti-vasoconstriction effects) with an oxygen transfer agent. A phase I safety study was recently completed, and no serious adverse events were reported. A phase Ib trial in sickle cell patients is currently underway, and multiple phase II clinical trials are in development, including one for treatment of vasoocclusive crisis. Sanguinate is also available for use under an expanded access emergency IND program (as with Hemopure above) for treatment of Jehovah’s Witness patients with severe anemia.

PolyHeme is a first-generation pyridoxylated polymerized hemoglobin made from outdated human blood. It has a half-life of 24 h, a shelf life longer than 12 months when refrigerated, and a p-50 of 28–30 mmHg. PolyHeme is now no longer being produced and is not available for use. Multiple clinical trials were completed to test the safety and effectiveness of PolyHeme during resuscitation as well as both intraoperatively and postoperatively. In a phase II randomized trial, PolyHeme reduced the number of allogeneic RBC transfusions in acute trauma [40]. The US Multicenter PolyHeme Trauma Trial was the first trial in the United States of a HBOC in the prehospital setting that employed a waiver of informed consent. This was a 714-patient phase III trial in trauma patients randomized to receive either PolyHeme or standard of care at the time of injury [41]. There was no statistically significant mortality difference (day 1 and day 30) between the PolyHeme and control cohorts. Although there were more adverse events in the PolyHeme group, the benefit-to-risk ratio of PolyHeme was deemed favorable when blood is needed but not available [42]. PolyHeme failed to receive FDA regulatory approval.

Hemolink (hemoglobin raffimer) is a polymerized hemoglobin product manufactured from donated human blood and O-raffinose cross-linked to produce a polyHb. A phase III multicenter, double-blind clinical trial was performed to determine if intraoperative autologous donation (IAD) with 10 % pentastarch or with Hemolink immediately before cardiopulmonary bypass confers a reduction in blood transfusion compared with standard clinical practice in primary coronary bypass patients. Hemolink use was associated with significantly reduced number of allogeneic RBC units and non-RBC units administered and lower overall transfusion rates. The company elected to discontinue further development of Hemolink in 2004.

MP4/Hemospan is a PEG-conjugated human hemoglobin that underwent clinical trials in the United States and Europe [43]. To further increase the circulation time, hemoglobin can be linked to a macromolecule to increase its size. Human or bovine hemoglobin that is conjugated with polyethylene glycol (PEG) is protected from renal excretion. MP4 underwent evaluation in two pivotal phase III studies in Europe. One study (prevention trial) evaluated the ability of Hemospan to prevent acute hypotension in orthopedic surgery patients undergoing first-time hip replacement procedures under spinal anesthesia (NCT00421200). Although hypotensive episodes were significantly less in the MP4 group, more MP4 patients experienced adverse events, and no significant difference in the composite morbidity and ischemia outcome endpoints was identified [44]. The second study (treatment trial) evaluated the ability of MP4 to treat acute hypotension in orthopedic surgery patients undergoing first-time hip replacement procedures under spinal anesthesia (NCT00420277). Although the mean total duration of all hypotensive episodes was significantly shorter in the MP4 cohort compared to standard care, certain adverse events did occur more frequently in the MP4 group (nausea, bradycardia, hypertension, oliguria), and no difference in the composite morbidity and ischemia endpoints was identified [45]. A phase IIb study of MP4 in traumatic hemorrhagic shock patients (NCT01262196) aimed to evaluate the safety and efficacy of MP4 treatment in trauma patients suffering from lactic acidosis due to severe hemorrhagic shock. According to the annual report filed by the company with the SEC, MP4 did not achieve its primary endpoint, number of patients discharged and alive after 28 days as compared to normal standard of care treatment.

Recombinant human hemoglobin (rHb) is manufactured using E. Coli with recombinant technology that can be used to induce a variety of cell types to synthesize functional hemoglobin. Modifications of the hemoglobin molecular structure can alter the properties of the molecule, allowing researchers to create hemoglobins with improved functionality or enhanced safety. A major advantage of rHb is that it can be manufactured resulting in an unlimited supply. A modified recombinant human hemoglobin (rHb1.1, Optro) was under development during the 1990s. This product circumvented the safety concerns from dimerization of the hemoglobin tetramer by cross-linking the alpha chains through recombinant engineering with rHb1.1 [46]. rHb1.1 was a first-generation HBOC with a nitric oxide scavenging rate similar to that of native human hemoglobin. rHb2.0 was a second-generation HBOC, created via genetic manipulation of the distal heme pocket of both the alpha and beta subunits of hemoglobin leading to steric hindrance for nitric oxide entry with a nitric oxide scavenging rate 20- to 30-fold lower than rHb1.1 but maintenance of effective oxygen binding and release [47]. Preclinical animal studies were promising confirming decreased pulmonary hypertension, diminished capacity to scavenge nitric oxide, and lack of modulation of pulmonary vascular permeability [48–52]. Although rHb2.0 appeared promising, no clinical trials were performed and funding of this initiative was suspended.

A number of new advanced HBOCs are undergoing development [53]. A complete discussion of these is beyond the scope of this chapter. However, several interesting compounds deserve mention.

A human-derived pyridoxylated hemoglobin polyoxyethylene conjugate, which is polymerized with superoxide dismutase and catalase (PHP), was developed to reduce free-radical-mediated oxidative stress and scavenge excess nitric oxide in catecholamine-resistant septic shock [54]. A phase III multicenter randomized trial was initiated of PHP as a treatment to restore hemodynamic stability in patients with shock associated with systemic inflammatory response syndrome (SIRS). The study (NCT00021502) was terminated early due to difficulty with enrollment; however, analysis of the enrolled patients (n = 62) determined that PHP patients had quicker shock resolution, but the study was significantly underpowered for any meaningful analysis.

Cellular HBOCs consist of Hb molecules encapsulated inside oxygen carriers of different natures, aimed at mimicking red blood cell features. Cellular HBOC advantages consist of protecting the surrounding tissues and blood components from direct contact with potentially toxic tetrameric hemoglobin, avoiding the hemoglobin colloidal osmotic effect, prolongation of Hb circulation half-life, and not requiring the direct modification of the Hb molecule. With the application of nanotechnology, it is possible to achieve submicron-sized oxygen carrier and thus ensure oxygen availability to all body compartments. Two types of cellular HBOCs have been studied: liposome systems and polymeric microparticle/nanoparticle systems [55]. Liposomes appear to be retained in plasma for a significant period. By encapsulating hemoglobin with a stabilized phospholipid membrane, liquid-state preservation for 2 years is guaranteed at room temperature, and with dry powder, further prolonged preservation is possible. Modifying the method of preparing micro-dimension polymeric artificial cells can result in nano-dimension artificial cells. These nano-dimension artificial RBCs (80–150 nm in diameter) contain all the red blood cell enzymes. Recent studies show that using a polyethylene glycol-polylactide copolymer membrane, it is possible to increase the circulation time of these nano-dimension artificial red blood cells to double that of polyhemoglobin [56, 57].

RBCs can now be cultured in vitro from human hematopoietic, embryonic, or pluripotent stem cells, which represent a potentially unlimited source of RBCs. It also presents potential opportunities for development of a new generation of allogeneic transfusion products [58, 59].

Landmark studies showed that RBC obtained from an immortalized embryonic stem cell line can protect mice from lethal anemia [60] and that ten million RBCs (equivalent to 2 ml of blood) generated ex vivo from CD34 cells from an adult volunteer and transfused had in vivo survival comparable to that of native RBCs [61]. The recently established human-induced pluripotent stem (hiPS) cells represent potentially unlimited sources of donor-free RBCs for blood transfusion as they can proliferate indefinitely in vitro [62].

Adverse effects of HBOCs include hypertension, abdominal pain, skin rash, diarrhea, jaundice, hemoglobinuria, oliguria, fever, stroke, and laboratory anomalies such as an elevation in lipase levels. Although most of these side effects are transient and clinically asymptomatic, many clinical trials involving these agents have been discontinued or held due to the associated adverse effects. Current formulations appear to cause fewer severe effects compared to previous products; however, there remain concerns associated with HBOCs. It has been difficult to discern whether the adverse events that have been observed following the infusion of HBOCs in patients are related solely to the HBOCs or to other treatments administered to these patients during their routine care. A meta-analysis of 16 trials in adult patients (n = 3,711) involving 5 different HBOCs in varied patient populations reviewed data on death and myocardial infarction as outcome variables [63]. They reported a statistically significant increase in the risk of death (RR 1.30, 95 % CI 1.05–1.61) and the risk of MI (RR 2.71; 95 % CI 1.67–4.40). There are, however, many limitations to this analysis [64].