Hemoglobin‐based oxygen carriers: Biochemical, biophysical differences, and safety

Abbreviations

EU

European Union

FDA

Food and Drug Administration

Hb

hemoglobin

HBOC

hemoglobin‐based oxygen carrier

HBOC‐201

Hemopure

IND

Investigational New Drug Application

kNO

second‐order rate constant of NO induced oxidation of hemoglobin

MW

molecular weight

n50

cooperativity among heme centers of hemoglobin

NADH

nicotinamide adenine dinucleotide

NO

nitric oxide

OECs

oxygen equilibrium curves

p50

oxygen affinity when Hb is half saturated

PEG

polyethylene glycol

pRBC

packed red blood cells

Redox

reduction–oxidation

rHb

recombinant hemoglobin

SDS‐Page

sodium dodecyl‐sulfate polyacrylamide gel electrophoresis

SFH

stoma‐free hemoglobin

SOD

superoxide dismutase

1. INTRODUCTION

Hemoglobin‐based oxygen carriers (HBOCs) have a long history of development, yet none are currently licensed for use in the US. Early HBOC studies were interested in the toxicity of infused hemoglobin, followed by the studies of Amberson and colleagues from the University of Maryland. ¹ As early as 1934, Amberson purified bovine hemoglobin and infused it into cats. In 1949, he described purified human hemoglobin infused into anemic parturients with hemorrhage after childbirth. ² Development continued when the US Army manufactured a tetrameric cross‐linked hemoglobin (α‐α cross‐linked hemoglobin), which later was produced by the Baxter Corporation (Deerfield, IL), as 2,3‐diaspirin cross‐linked hemoglobin (HemAssist). ² However, it failed in human studies because of decreased cellular perfusion and increased morbidity and mortality. Recombinant HBOC was developed but failed due to bacterial endotoxin concerns (Optro, Somatogen, Boulder, CO). ² During a resurgence of activity in the mid‐1980s, manufacturers developed second‐generation HBOCs, including HBOC‐200 (Oxyglobin; HbO₂ Therapeutics, Souderton, PA), approved in the US and the EU for canine anemia; Hemoglobin‐glutamer‐201 (Hemopure; HbO₂ Therapeutics, Souderton, PA) studied in the largest completed clinical trial of an HBOC, approved in South Africa in 2001 and Russia in 2006 and allowed for expanded access use in the US by the FDA. ² , ³ , ⁴ PolyHeme by Northfield Laboratories (Evanston, IL) and Hemolink by Hemosol, Inc. (Mississauga, Ontario, Canada) were discontinued due to a question of efficacy or to safety concerns. ² ErythroMer (Kalocyte; Baltimore, MD), an encapsulated HBOC, is being developed as a replacement for red blood cells. ⁵ Most recently, Hemarina (Morlaix, FR) has developed HemO2Life, approved in the EU in 2022 for ex‐vivo perfusion of kidneys prior to transplantation and orphan designation in the EU for patients undergoing hematopoietic stem cell transplantation. ⁶ Other HBOCs are in various preclinical stages of development. The efficacy of various HBOCs is dependent on biophysical and biochemical characteristics, results of preclinical and clinical studies, and the indication(s) being proposed. ²

The challenge for earlier generation HBOCs was safety due to multiple factors (see Table 1, ⁵ , ⁶ , ⁸ , ⁹ , ¹⁰ , ¹¹ , ¹² , ¹³ , ¹⁴ , ¹⁵ , ¹⁶ , ¹⁷ , ¹⁸ ). The purpose of this Commentary is to define these safety issues related to biophysical and biochemical characteristics. We compare and contrast first, second, third, and new‐generation HBOCs based on these characteristics, then  with the information learned and correlating to published clinical studies, begin to study how to manage the safety concerns when alternatives are unavailable. First‐generation products are referred to as HBOCs that have undergone clinical trials in the US (phase I–III) whereas HBOCs that have yet to undergo clinical trials are referred to as second‐generation HBOCs. However, in a recent chapter on classifications of HBOCs, an alternative nomenclature basing the generations on chemical structure has also been proposed. ⁷ Finally, we provide insights into newer generation HBOC development to avoid these issues and derive patient benefit from the efficacies of these products.

TABLE 1.

Biochemical reaction parameters and clinical trials and their outcome.

HBOC class: Classification 7	Product name	Hemoglobin type	Molecular weight (kDa)	SDS‐PAGE analysis (sodium dodecyl‐sulfate polyacrylamide gel electrophoresis)	p50 (mmHg) (oxygen affinity when Hb is half saturated)	n50 (cooperactivity)
Cross‐linked: First Generation	Diaspirin cross‐linked Hb DCLHb (HemAssist) (alpha‐alpha cross‐linked Hb) 6	Tetramer (bi3,5‐dibromosalicyl fumarate cross‐linked deoxy SFH between two alpha subunits)	64	14 kDa band (beta chains) and 32 kDa band (2 alpha alpha cross‐linked chains)	31.10 ± 0.27	2.46 ± 0.1
	rHb recombinant (Optro) 4	cross‐linked tetramer (Hb Presbyterian and glycine used to bridge two alpha subunits)	64		35 ± 2.8	1.92 ± 0.91
	Hemolink (O‐R‐PolyHbA0) 6	Polymerized Hb (intra and intermolecularly cross‐linked product of purified HbA0)	<64 to >500, with 90% being 64–500	Broad bands, progressively larger, representing cross‐linked subunits	34.4 ± 1.29	1 ± 0.05
Polymerized: Second Generation	Polyheme (PolyHHb) 6	Polymerized human Hb	130–250	Broad bands, progressively larger, representing cross‐linked subunits	31.34 ± 0.8	2.14 ± 0.04
	Hemopure (HBOC‐201) (PolyBvHb) (Biopure/HbO2 Therapeutics) 6	Polymerized bovine Hb	130–500	Broad bands	34.28 ± 0.77	2.1 ± 0.04
	Oxyglobin 6	Polymerized bovine Hb, more heterogenous than Hemopure	130–500	Broad bands, large tetrameric 64 kDa fraction	34.59 ± 0.67	2.05 ± 0.08
	OxyVita 14	Polymeric Hb‐based carrier, alteration in sulfo‐NHS can be used to manipulate extent of polymerization	17,453 (solution)		6.4 ± 02	1.2 ± 0.08
Conjugated: Third Generation	Hemoximer 15	Stroma‐free human Hb, chemically modified	109		23.4
	Hemospan 15	Uses Hb derived from outdated human Hb	90		8.18 ± 0.21	1.56 ± 0.01
	Sanguinate (PEGBvHb) CO 6	Bovine Hb conjugated with PEG on surface lysines	120	Monomer 16 kDa band with progressively larger MW bands (conjugated alpha or beta subunits)	9.56 ± 0.28	1.56 ± 0.01
Encapsulated: New Generation	ErythroMer 4	Nanoparticles with toroidal morphology			pH responsive (19 mmHg – 30 mmHg in Ph 7.6–7.2)	2.5
	Hb vesicles 16	Hb vesicles	Smaller than RBCs
Naturally Polymerized: New Generation	Hemarina M101 5	Extracellular Hb derived from sea worms	~3600		7	2.5

HBOC class: Classification 7	Catalase activity (unit/mL)	K autoxidation (h−1) (rate constant)	kNO (μM‐−1 s−1) NO‐induced oxidation of hemoglobin	Heme loss (heme dissociation from ferric Hb) (h−1) (fast phase)	Clinical trials	Outcome
Cross‐linked: First Generation	0.09	0.081 ± 0.003	41.5 ± 2.5	7.3 ± 0.05	Increased mortality, HTN, MI, arrhythmia, GI issues, pancreatitis, hemorrhage, pancreatic enzyme increase, compared to control with pRBC. Also noticed myocardial lesions, which resolved with time. Vasoactivity correlated with Hb extravasation and NO scavenging 8	Product was discontinued by Baxter in 1999 due to traumatic hemorrhagic shock clinical trial data showing increased mortality 9 , 10
	0	0.095 ± 0.01	59.6 ± 1.3	7.2 ± 0.45	HTN, chest pain, GI concerns, liver abnormalities, and elevated pancreatic enzymes. Sub therapeutic doses in early Phase II trials led to complement activation 8	Development was discontinued in 2003 by Baxter due to negative hypertensive effects 10
	0.16	0.13 ± 0.01	37.1 ± 1.7	7.3 ± 0.04	Increased HTN, MI, GI issues, liver abnormalities, pancreatic enzyme increase, compared to hetastarch or Ringer's lactate controls. 8	Product was discontinued due to cardiac toxicity 11
Polymerized: Second Generation	4.01	0.26 ± 0.01	41.3 ± 1.2	11.7 ± 0.08	Increased death, MI, CVA, respiratory distress, coagulation defects, hemorrhage, and sepsis, compared to controls treated with crystalloid and pRBCs 8	A trauma trial found that found mortality of 13% for the test group vs. 10% for controls with 40% SAEs in the test group vs. 35% of controls, after which Northfield discontinued the product. Due to lack of FDA approval, the product was discontinued in 2009 12 , 13
	0.19	0.22 ± 0.02	44 ± 1.1	3.47 ± 0.91	Increased death, HTN, pulmonary HTN, chest pain, CHF, cardiac arrest, MI, arrhythmia, CVA, PNA, respiratory distress, acute renal failure, hypoxia, hypovolemia, GI issues, liver abnormalities, coagulation defects, hemorrhage, sepsis, and pancreatic enzyme release, compared to controls treated with pRBCs 8	The manufacturer contended that SAEs in Phase III trials were a result of patient comorbidity, patient management differences, and underdosing the product with ongoing anemia and ischemia and overdosing with CHF. There is no evidence of vasoconstriction in the heart, brain or kidney. This product affects the MAP, SVRI while not affecting LVEDP, coronary output, or coronary perfusion. It is approved for human use in acute anemia in South Africa 6 , 8
	0.15	0.21 ± 0.02	43.9 ± 1	4.5 ± 0.95		US FDA approved for veterinary use 6
		0.018 ± 0.00023	30 ± 3.3	Fast phase = 0.59 ± 0.04		Oxyvita remains in development 13
Conjugated: Third Generation						It was investigated as a vasopressor in shock due to its NO scavenging abilities. It failed in Phase III clinical trials in 2011 and was discontinued 14
	0.41	0.070 ± 0.002	44.5 ± 1.3	14.80 ± 0.06	Increased mortality, HTN, MI, arrhythmia, GI issues, and elevated pancreatic enzymes, compared to controls treated with Ringer's lactate 8	Product was discontinued in 2015 by Sangart, following failed Phase III clinical trials 13
	0.24	0.19 ± 0.02	49.7 ± 1	16.7 ± 0.21		This product is undergoing prolonged development, with unpublished Phase II clinical trials complete 13
Encapsulated: New Generation		inhibits Hb autoxidation by incorporating leuco‐methylene blue	equal to fresh RBC	0.3 M % heme	No clinical studies to date
						This product is currently undergoing Phase I clinical trials 13
Naturally Polymerized: New Generation	3.5 U/mg Hb				There was better graft function and improved recovery of serum creatinine in renal transplant patients whose transplanted kidneys were preserved with Hemarina M101 ex vivo. There were also low rates of acute rejection, no capillary thrombosis, or alteration of microcirculation, proving its safety 17	This product is approved in the EU for extracorporeal kidney perfusion and orphan designation 5

Note: Definitions: Cooperativity, When an oxygen atom binds to one of hemoglobin's four binding sites, the affinity to oxygen of the three remaining available binding sites increases and is expressed as n50; Rates of autoxidation, Refers to the accessibility of the heme pocket to exogenous reagents that are able to induce oxidation through electron transfer to bind oxygen to heme and is expressed as h⁻¹; NO‐induced oxidation, Also known as nitric‐oxide dioxygenase (NOD), it refers to the efficient conversion of NO and O₂ to nitrate by Hb and rates are expressed as μM‐⁻¹ s⁻¹; Heme loss, Heme, the prosthetic group of hemoglobin, may be released due to an intrinsic instability of hemoglobin and accumulate in the red blood cells. Experimentally oxidized hemoglobin releases its heme to a receptor molecule such as albumin and rates are expressed as h⁻¹; Catalase activity, In the presence of oxygen, hemoglobin catalyzes the decomposition of hydrogen peroxide to water and oxygen and is expressed in units/ml; Oxygen equilibrium curves, A graphical representation of the relationship between the amount of oxygen bound to hemoglobin and the partial pressure of oxygen in the blood.

Abbreviations: CHF, congestive heart failure; CVA, cerebrovascular accident; EU, European Union; GI, gastrointestinal;Hb, hemoglobin; HTN, hypertension; kDa, kilo Daltons; kNO, nitric oxide‐induced oxidation; LVEDP, left ventricular end‐diastolic pressure; MAP, mean arterial pressure; mg, milligram; MI, myocardial infarction; mmHg, millimeters of Mercury; MW, molecular weight; NO, nitric oxide; pRBCs, packed red blood cells; RBCs, red blood cells; SVRI, systemic vascular resistance index; U, units.

While packed red blood cell (pRBC) transfusions remain the mainstay of treatment for patients who are severely anemic, require increased oxygen delivery to tissues, or are acutely bleeding, HBOCs have been in development for decades as an alternative. These products hold potential for scenarios in which pRBCs may not be a feasible option, such as in severe trauma and combat medicine in remote locations (due to limited storage, the requirement for crossmatching, and potential difficulties with cold storage), when there is scarcity in donated blood, and for patients who cannot receive pRBCs, either due to rare blood types/compatibility issues or religious objection. HBOCs, using acellular Hb derived from bovine or human Hb or genetically engineered, have been explored as an option since the 1930s when Amberson used purified human and bovine Hb. HBOCs have been created using different methods, often involving cross‐linking, the addition of chemical modifiers, PEGylation, and polymerization (Figure 1). ²

FIGURE 1.

Development of hemoglobin solutions as oxygen carriers. Stroma‐free hemoglobin (SFH) derived from outdated human or bovine blood is either cross‐linked, conjugated with macromolecules, or polymerized to stabilize the hemoglobin tetramer and enhance its function. SFH or cross‐linked SFH are, in some cases, encapsulated within liposomes (adopted from Alayash AI, with modifications; Nature Biotechnology Vol 17 June 1999, 545–549). [Color figure can be viewed at wileyonlinelibrary.com]

Hb within RBCs is found at high concentration which favors stabilization of the Hb tetramer. Additionally, 2, 3‐diphosphoglycerate (2,3‐DPG), a natural allosteric modifier of Hb adds additional stability to the molecule. When Hb is extracted from RBCs in a dilute solution, it tends to dimerize into α/β dimers, which, if infused, will be cleared rapidly and which could damage kidneys and other organs. Manufacturers treat Hb with chemical reagents that specifically crosslink the dimers and/or polymerize a number of tetramers in a large polymer. This will stabilize Hb in circulation and in some cases improve its function.

Despite the variety of chemical approaches used in creating HBOCs, there have been persistent concerns related to serious adverse events throughout their development. ² , ⁸ , ¹⁹ , ²⁰ , ²¹ In particular, concerns have arisen relating to cellular oxidative stress, nitric oxide (NO) scavenging, methemoglobinemia, and the inflammatory response seen when heme is lost from Hb. ² , ⁸ , ¹⁹ , ²⁰ , ²¹

Various mechanisms have been proposed for the causes of serious adverse events with the use of HBOCs. Among the major biochemical mechanisms that explain HBOC toxicity includes (a) scavenging of endothelial nitic oxide and consequent hemodynamic changes, (b) oversupply of oxygen and autoregulatory responses, and (c) heme‐mediated oxidative reactions. These pathways, singularly or collectively, may have contributed to the vascular responses and subsequent injuries. ¹⁹ Cellular and sub‐cellular models using human kidney endothelium, mouse lung epithelium, and alveolar cells have been used to demonstrate the deleterious effects of Hb oxidation and heme loss. ²¹ Free heme is itself a damage‐associated molecular pattern (DAMP), which leads to an inflammatory response. ²¹ DAMPs are molecules within cells that are a component of the innate immune response released as danger signals from damaged or dying cells due to trauma or an infection. Heme released from Hb was recently recognized as a DAMP molecule as it triggers a cascade of inflammatory responses. ²⁰

Free heme is also a critical mediator of endothelial dysfunction, kidney and lung injury in sepsis, and poor systemic outcomes of hemolytic disorders. It likely also plays a role in poor outcomes from polytrauma. ²² Hb undergoes a redox (reduction–oxidation) transition, leading to the formation of higher oxidation intermediates, such as ferryl Hb (HbFe⁴⁺), which can attack other biological entities and ultimately self‐destruct, leading to heme loss. ²¹ There has been considerable research activity focused in recent years on using haptoglobin (protein scavenger) and hemopexin (heme scavenger). ²³ , ²⁴ However, haptoglobin exhibits weak binding to some HBOCs and the cost of using highly purified scavenger as an additive may have prohibited further development of these important reagents.

Newer and current‐generation HBOCs have been designed to a certain extent to overcome some of the reported toxicities associated with first generation HBOCs. However, very little is known about possible toxicities, as some of these HBOCs have only recently been produced or reported in the literature. ¹⁴ , ¹⁵ , ¹⁷ , ²⁵ , ²⁶

Among the most promising intervention strategies designed to control or minimize HBOC toxicity besides haptoglobin and hemopexin, are antioxidants, and reducing agents such as ascorbic acid were used. ²¹ Utilizing the power of genetic engineering, new‐generation HBOC prototypes have been engineered with built‐in superior oxidative stability, which may pave the way for genetically engineered HBOCs. ²⁷

Despite the diverse nature of chemical and genetic modifications during HBOC development, very little attention was given to the impact of these alterations on the efficacy and safety of HBOCs. Recently, a comprehensive biochemical and biophysical characterization was carried out on all HBOCs that have undergone clinical evaluation, allowing for the first‐time direct comparison. ⁸ Here, we revisit the documented HBOCs' chemistries and attempt to correlate these unique chemical entities with their published clinical outcomes. This Commentary is divided into the following sections: Introduction (above), Evaluation, Findings, Discussion, and Conclusions.

2. EVALUATION

The goal of this work was to synthesize available data to address concerns around the serious adverse events of HBOCs, to aid in the use of pharmacological strategies to address adverse effects. Concretely, we reviewed recent literature on the subject, examining the structure, activity, and biochemical relationships of some discontinued and actively developing HBOCs. We have also analyzed available literature on the serious adverse events of these products. For each of these products, we identified Hb type, molecular weight, SDS‐PAGE analysis, p50, n50, catalase activity, autoxidation rate, NO‐induced oxidation (k_NO), and heme loss kinetics. We then synthesized this data into a table to allow for comparison of these factors. Finally, we interpreted our findings, with the goal of creating a path forward for further innovation and development of HBOCs.

3. FINDINGS

In collating the available biochemical data on these HBOC products, various trends were identified (see Table 1). One critical variable common among HBOCs is the different starting materials. Almost all HBOCs began as stroma‐free Hb (SFH) or stroma‐poor Hb. This means that these Hbs retained some red cell proteins prior to chemical modifications. One particular HBOC, Hemolink, in which all of these contaminants were eliminated, started with almost 99% pure Hb known as HbA₀ by using extensive anionic and cationic chromatography; however, the product eventually was discontinued due to adverse effects noted in a number of clinical trials. ² , ²⁸

Many HBOCs are derived from human outdated blood prior to chemical modifications with minimal purification steps that remove other red cell proteins to produce SFH or stroma‐poor Hb. After treatment with the modifying chemical reagent(s), this results in HBOC solutions that are contaminated with these proteins. Other products derive from bovine blood, and one uses a sea worm hemoglobin, but all require some chemical modification. ⁶

As HBOCs evolved from small, cross‐linked molecules with lower molecular weight to larger, polymerized molecules, heme loss kinetics trended lower, with the exception of Sanguinate. ⁸ This is in fitting with the goal of larger HBOCs, which were created with a purpose to reduce protein unfolding, thus decreasing heme exposure and loss. ¹⁵

Oxygen affinities, as measured by oxygen equilibrium curves (OECs) for HBOCs, range from sigmoidal to non‐sigmoidal almost linear, and in some cases with reduced cooperativity (communication among the 4 heme centers). The ideal p50 (oxygen affinity when Hb is half saturated) of HBOCs is still unknown, with earlier cross‐linked and/or polymerized products targeting a p50 of approximately 30–35 mmHg, close to the p50 of fresh human blood (27–29 mmHg). Conjugated, encapsulated, and naturally polymerized HBOCs, however, have sought a lower p50, decreasing oxygen offloading to tissues, except in extremely hypoxic conditions by increasing its affinity to Hb. Although some within the HBOC community believe that HBOCs should have higher oxygen affinity (decreased p50) as HBOCs with low oxygen affinity (larger p50) may trigger autoregulatory responses (vasoconstriction) due to the immature offloading of oxygen. Both Hemarina and Sangart (PEGHHb) have lower p50s. One notable exception is the ErythroMer encapsulated product, which is pH responsive, thus allowing for a dynamic p50. ⁵ , ²⁵

Cooperativity, (n50; when an oxygen atom binds to one of hemoglobin's four binding sites, the affinity to oxygen of the three remaining available binding sites increases) has not evolved drastically across HBOC products. It varies from 1.0 in Hemolink (O‐R‐PolyHbA0) to 2.5 in ErythroMer and HEMO2Life. ² , ⁸ , ¹⁴ , ²⁵ In addition to the unique oxygen affinity characteristics of HBOCs discussed above and listed in Table 1, we included some other equally important characteristics that defined these HBOCs further. Collectively, these properties may have some bearing on their safety as well as efficacy. Besides their unique electrophoretic properties, HBOCs differ considerably in their oxidative and redox properties that include autoxidative (spontaneous) oxidation kinetics that describe the trends of iron to oxidize and subsequent oxidative changes. In addition, antioxidative properties are indicated by their own catalase activities, that is, the ability to remove oxidants such as peroxide. Altogether, these reactions are profoundly different from those characteristics of normal unmodified HbA which dissociates into dimers rapidly after transfusion and exhibits non‐physiological oxygen binding as well as oxidative behavior. Polyheme (PolyHHb) has increased catalase activity, at 4.01 units/mL, due possibly to the contaminating catalase in SFH. All other products had low catalase activity, between 0 and 0.24 units/mL. ⁸

Cross‐linked HBOCs, which were ultimately discontinued, showed lower rates of autoxidation, with rate constants between 0.081 and 0.095 hr⁻¹. All other products, meanwhile, had autoxidation rate constants between 0.19 and 0.26 h⁻¹. ⁸

Most HBOCs had rate constants of NO‐induced oxidation between 40 and 45 μM⁻¹ s⁻¹. Recombinant Hb (Optro) showed very high rates of NO‐induced oxidation (59.6 μM⁻¹ s⁻¹), which is notable given its discontinuation because of hypertensive effects. ⁸ , ¹⁴

Polymerized bovine Hb, HBOC‐201 (Hemopure) and HBOC‐200 (Oxyglobin), showed the least heme loss, with levels of 3.47 and 4.5 h⁻¹, respectively. ⁸ Conversely, the PEGylated Hbs, Hemospan and Sanguinate, showed increased heme loss, with levels of 14.8 and 16.7 h⁻¹, respectively, due possibly to the fact that these HBOCs lack intermolecular crosslinking. ⁸ , ¹⁹

4. DISCUSSION

Because of the proprietary nature of manufactured HBOCs, very little information was made available in the open literature on manufacturing, preclinical, and clinical assessments. It was only recently that most manufacturers were willing to share their products with independent researchers and then only after the termination of their manufacturing programs. This culminated in the publication of a seminal paper comparing side by side all HBOCs that had been clinically tested in humans. ⁸ In that publication, the authors attempted to correlate some of these unique features with their clinical outcome and attempted to draw some general lessons that will enable the design of a safe and effective product and possibly enable the use of earlier products. Based on this publication, clinical trials, and clinical outcomes were added to the biophysical and biochemical characteristics of the various products to be able to make more definitive statements regarding products.

When critically reviewing the trends, it may make sense to organize the products based on formulation (HBOC Class) and by increasing molecular weight. ⁷ In this case, the smallest molecules, the cross‐linked and cross‐linked polymerized, are notable for increased rates of NO oxidation, with the greatest being rHb. The least may be the largest molecules, OxyVita and HEMO2life, as manufacturers believe that they are too large to be extravasated through endothelial barriers. ErythroMer may have resolved this issue with similarity to fresh red blood cells by allowing for modulating p50 based on the pH of the immediate cellular environment. ⁶ , ¹⁵ However, some products may be able to overcome this concern with concurrently administered nitrite, as the hypertension noted ²⁹ , ³⁰ is a result of a decrease in circulating NO. Another study by the same group ³⁰ showed that a combination of nitrite and HBOC‐201 appeared to increase the risk of pulmonary complications in a dose‐dependent manner, independently of hemodilutional effects on hemostatic components. ³⁰

Another issue with earlier generation HBOCs is methemoglobin (oxidized Hb) production. This is dependent on the level of NADH diaphorase and cytochrome b5 reductase, which is in turn dependent on cellular Hb concentrations. ³¹ Since the reductase for converting methemoglobin to oxyhemoglobin is found in the red cell membrane, ³¹ with severe anemia there is insufficient reducing ability. Use of ascorbic acid ⁴ , ³² and methylene blue can be administered to ameliorate the adverse effects of methemoglobinemia when above 10%, as this level interferes with oxygenation. However, in cases where the hematocrit is extremely low, this may be an unresolvable issue, and could potentially require pRBCs in the extreme, as this would be the only way to increase red cell membrane diaphorase and reductase. Also, it has been documented that early in hemorrhage maintenance of Hb concentrations above 6–7 g/dL may improve survivability. ³² Hence, an HBOC with higher Hb concentration, such as HBOC‐201 (Hemopure, 13 g/dL), may be advantageous, or OxyVita, which was available as a lyophilized product, and therefore could be diluted with only modest amounts of fluids to maintain higher Hb, and which could be advantageous. ³³ The importance of the identified trends can only be speculated upon in the context of each product's clinical outcomes.

The second large product, HEMO2Life, naturally polymerized and derived from a marine worm, has recently been studied in and approved in the EU for ex‐vivo perfusion of transplanted kidneys. In this context, its use in the preservation solution for ex‐vivo kidneys has shown less graft dysfunction and faster recovery of serum creatinine post‐transplant. ³⁴

Heme loss is a result of HBOCs prematurely losing heme (due to oxidative changes within the Hb molecule) or hemolysis that can potentially be mitigated using haptoglobin or hemopexin. ²⁰ HBOCs have increased rates of autoxidation, roughly two to three times higher than rates present in human or bovine Hb within red blood cells. ⁸

There may be a link between NO scavenging and increased myocardial infarctions with HBOCs due to NO inhibition of platelet function that would be diminished by scavenging. ⁹ Patients who had nitrite co‐infused with HBOCs actually showed increased blood flow due to the nitrite reductase action of HBOCs, generating increased NO. ⁹ However, it is important not to link all HBOCs together with serious adverse effects, as all HBOCs are not chemically identical, as we have pointed out. Each has different effects and must be studied specifically. For example, HBOC‐201 (Hemopure) has been demonstrated to not have any effect on platelet function in a randomized, controlled, double‐blind, single‐site study. ³⁵ It is unclear if other products have specific effects on platelet function. ³⁶ Additionally, PolyHeme, a product no longer under study, has been demonstrated to avoid renal glomerular and tubular responses due to a designed anti‐inflammatory property. ³⁷

While HBOCs have been historically compared to pRBCs, the authors posit that this is not the necessary comparison to make. It seems clear that HBOCs have been plagued by valid concerns of serious adverse events due to multiple mechanisms; however, with some necessary changes in their development, we would contend that HBOCs may be lifesaving in situations where pRBCs are either not available or not an option. For example, approved use of HBOC‐201 (Hemopure) for FDA expanded access indications has been documented in almost 100 cases in multiple case reports. In many cases, it may have been lifesaving, when standard care had either failed or would have been ineffective or deleterious. ⁴ Requirements for expanded access approval have been documented. ³⁸

Despite the long history of HBOC development without the introduction of a product licensed in the US, this remains an area of active interest. ³⁹ Of note, there was a decrease in the publication rate in about 2011, which may have been fallout from a meta‐analysis claiming serious adverse events. ⁴⁰ , ⁴¹

The paper by Natanson et al., ⁴⁰ was a pivotal point in the development of HBOCs. Mackenzie et al. ⁴¹ began to address the issue that all HBOCs in development are quite different but may have wrongly claimed that the FDA demanded that HBOCs be equated to pRBCs: in fact, the FDAs “point to consider” in 1991 referred to them appropriately as HBOCs. ⁴²

We are aware that some in the “HBOC community” are under the misapprehension that all HBOC studies may have been put on hold by the FDA. The FDA is not allowed to make public any existing or pending submissions, including any toxicity or safety concerns of efficacy (or lack thereof) that are not in the public domain. Thus, we cannot comment on whether there are any ongoing trials with an HBOC and whether there have been any clinical trials with an HBOC over the past decade or two. What nonclinical data are required prior to extending into clinical studies may vary with the product. However, the very basic requirements are the same as for any biological prior to a first‐in‐human trial: the demonstration of purity/potency/toxicology/pharmacology, animal husbandry practices if the HBOC is animal‐derived, viral clearance, safety, and efficacy. A full range of nonclinical studies would be required. Nonhuman primate studies would not specifically be required. The human population to be included in clinical studies will depend on the proposed indication, and the benefit–risk approach will also vary based on the proposed indication: higher potential benefit might make higher potential risk acceptable. Although some products might go “head‐to‐head” against packed red blood cells, the potential limitations of HBOCs (safety, efficacy, and Hb concentration) may make this an impossible hurdle to overcome for some products, requiring a more limited indication. All HBOCs are not the same, and each new product would need to meet the criteria for advancing to human trials.

Given the relatively good outcomes of the small case reports and the longstanding availability of HBOC for expanded access use, one might ask why the randomized trials have been unsuccessful. Space does not allow a full discussion but there are several potential reasons, primarily having to do with the relative inadequacies of single‐arm observational studies versus randomized control trials. As for the expanded access experience, only approximately 76 subjects have been reported to date in the literature. There are several other potential confounders including population heterogeneity, varying comorbidities and procedural risk (e.g., all‐comers vs. surgery vs. trauma), and the potential for an assortment of biases (selection bias, treatment bias, withdrawal bias, and survivor bias). Additionally, the return of data by practitioners, who may only handle a single expanded access request in their practice lifetime, may be incomplete.

There are multiple hurdles HBOC products will need to overcome to allow for licensure in the US as indicated in the preceding paragraph.

We are unable to address potential indications that may be sought by current manufacturers. However, to take one example, when blood is not an option, it will require studying what is in fact a relatively small and geographically dispersed population. Additionally, HBOC manufacturers will need to identify an appropriate control or comparator. If no product is appropriate as a comparator, would there be adequate clinical equipoise given the history of these products to allow a comparison of standard of care with an HBOC product? Questions like these may be difficult to address fully as these products progress from nonclinical to clinical development and hopefully licensure. An expanded access experience, even a substantial one, would not suffice alone for licensure, although it might provide supportive data.

Our commentary is limited by the lack of examination of all relevant variables of the newest HBOCs, for example, HbV (hemoglobin vesicles), OxyVita, ErythroMer, and Hemarina, and there are additional products in various stages of preclinical and clinical, investigation (Sanguinate, Prolong, South Plainfield, NJ). ²⁶ , ⁴³ Head‐to‐head clinical trials are needed to compare definitively current and future HBOCs.

AUTHOR CONTRIBUTIONS

Jonathan S. Jahr, Khrystia MacKinnon, Victor C. Baum, and Abdu I. Alayash designed, wrote, reviewed, and revised manuscript, and have no conflicts of interest or financial conflicts with this work in any way.

CONFLICT OF INTEREST STATEMENT

The authors have disclosed no conflicts of interest.

Jahr JS, MacKinnon K, Baum VC, Alayash AI. Hemoglobin‐based oxygen carriers: Biochemical, biophysical differences, and safety. Transfusion. 2025;65(2):386–396. 10.1111/trf.18116