Hemoglobin-based Solutions

Introduction : why were the hemoglobin solutions developed?



The sophisticated techniques used by modern medicine permit successful interventions in heavily injured patients, treatments of ischemic patients (chronic anemia, stroke, septic shock) and complex transplantation surgery, but they have increased the need of human blood. At the same time, the HIV epidemic rose suspicion on the safety of allogenic blood transfusion. Together with the risk of errors in blood transfusion, the risk of transmission of HIV and other viral diseases such as hepatitis, and the insufficiency of palliative treatments (blood predonation, pre- and perioperative hemodilution, perioperative blood sparing, lowering of transfusion trigger) accelerated the development of blood substitutes as alternatives to human blood, with the ideal characteristics of oxygen transport and the availability for immediate use with no typing and cross-matching <|[1-3]|>.

Intense efforts in the field of blood substitutes started up and were largely supported since 1980s by investment of the US military, which was concerned by the need of a reliable resuscitation solution that could be used immediately after injury and did not need special storage conditions. The currently developped hemoglobin solutions appear clearly unable to meet emergency requirements, and as an expected consequence, a fall of military interest and support is observed <|[4]|>. Nevertheless, modified hemoglobin solutions have been produced, that have been used in animal models with success and are now at the stage of clinical assays. Their use in humans has led to better understanding of oxygen delivery physiology, but also revealed sequelae requiring complementary research.



Structure and catabolism of natural hemoglobin



Hemoglobin is a tetrameric molecule formed by 2 a and 2 b monomers, each of them containing a heme. The monomers are joinded together by a weak bond between the 2 a and the 2 b subunits, and by a tighter bond between the a and b monomers. The ferrous iron (Fe2+) of the heme is bound to the hemoglobin chain by a proximal histidine. Hemoglobin is found under 2 forms oxyhemoglobin with a high affinity for oxygen and deoxyhemoglobin with a low affinity for oxygen <|[5]|>. The affinity of hemoglobin for oxygen is dependent on the pO2 (sigmoid curve of oxygen saturation) and on the pH (Bohr effect) values of the blood. The arterial partial pressure corresponding to 50 saturation of hemoglobin is 26-27 mm Hg. The transition from oxy- to deoxyhemoglobin is regulated by allosteric effectors like 2,3 diphosphoglycerate (2,3-DPG). Binding and release of oxygen result in conformational modifications of the heme and protein moieties (the R and T states). When
oxygen is bound to Fe2+, the iron and porphyrin ring are coplanar and iron is in a low reactive state. At deoxygenation, a slight rotation of the monomers moves the iron out of the plane of the porphyrin ring, rendering it more sensitive to oxidation to
Fe3+ (methemoglobin). This iron oxidation produces superoxide anion and hydrogen peroxide, which further react with methemoglobin to produce ferrylglobin radicals, the first step to lipoperoxidation damage (on neuronal membrane and on lipoproteins for example) <|[6, 7]|>. Inside the erythrocytes, antioxidant molecules and specific enzymes (superoxide dismutase, catalase, GSH peroxidase, methemoglobin reductase) inhibit this dangerous cascade <|[8]|>.



Free hemoglobin in plasma is rapidly broken into dimers which are bound by haptoglobin and transported to the liver. Hemopexin and albumin dissociate Fe3+-heme from the monomers and transport it to the liver, where heme is recycled into the synthesis of new heme proteins, and Fe3+ is transferred to apotransferrin <|[8]|>. When the binding capacity of plasma proteins is overwhelmed, the hemoglobin tetramers and dimers reach the kidney where they are filtered in the glomeruli and reabsorbed in the proximal tubules. Hemoglobinuria indicates that the absorptive capacity of the kidney is overloaded, leading to nephrotoxicity. It has also become evident now that the tetramer is able to extravasate owing to its positive electric potential <|[9]|>. Large doses of free hemoglobin thus cross to microvessel barrier, reaching interstitial spaces and resulting in jaundice-like syndrome. Moreover, it appears that free heme is incorporated into the endothelial cell membranes where it may potentiate oxidant damage, but that chronic exposure of endothelial cells to free heme induces the iron-binding protein ferritin and the heme oxygenase, a protective enzyme degrading heme into bilirubin and CO <|[10]|>.



Lysed erythrocytes: the first solution used to replace blood



The first hemoglobin solutions were obtained by lysis of erythrocytes and used in humans as soon as the end of the 19th century. In 1898, von Stark used a solution of lysed erythrocytes to treat patients with chronic anemia <|[11]|>. Rapidly, it appeared that free hemoglobin was harmful, leading to disseminated intravascular coagulation, cardiac failure and renal toxicity. In 1916, low doses of free hemoglobin were administered to human to study its renal clearance <|[12]|>. In the 1930s, a systemic and pulmonary vasopressor effect of free hemoglobin was described in animals, that could not be attributed to volemic expansion <|[13]|>. This special effect was used with success to resuscitate a patient with hemorrhagic shock by administration of low doses of a hemoglobin solution <|[14]|>. More recently, Rabiner et al <|[15]|> treated hemorrhagic shock patients with 180 to 300 mg/kg stroma-free hemoglobin, and Savitsky et al <|[16]|> administered 250 ml of hemoglobin solution to volunteers with minor side effects.

These assays led to the conclusion that the toxicity of free hemoglobin solutions was due to the presence in these solutions of residual fragments of erythrocyte membranes and lipids that were nephrotoxic and caused hemolysis. These remnants also activated intravascular coagulation, complement, platelets, and white blood cells, leading to the liberation of inflammatory mediators <|[15]|>. A careful purification of hemoglobin after erythrocyte lysis would thus produce a safe blood substitute. But, despite careful purification, it appeared that the administration of free hemoglobin solutions remained an at risk intervention, due to the variability in the techniques of purification, to the instability of the solution (methemoglobin production), to the increase of the oncotic pressure (limiting the concentration of hemoglobin to 7 g/dl), to the excessive affinity of free hemoglobin for oxygen (impairing oxygen delivery to tissue) and to its rapid catabolism leading to renal toxicity.



Modified hemoglobin solutions



From the early studies with natural free hemoglobin, three major problems were brought to light acute toxicity attributed to erythrocyte membrane remnants, renal toxicity, and excessive affinity for oxygen attributed to the absence of the allosteric effector 2, 3-DPG. The modifications that were introduced in natural hemoglobin essentially aimed at reducing or eliminating these problems. Purification of natural hemoglobin was improved, beginning in 1970, to discard the membrane fragments and the lipidic residues, producing stroma free hemoglobin (SFH) <|[17]|> that had no longer acute toxicity. Chemical modifications of the hemoglobin molecule were designed to suppress the rapid break of the tetrameric molecule in plasma and to avoid or reduce renal toxicity. The following modifications have been used internal stabilisation of the tetrameric molecule by cross-linking of dimers, pyridoxylation, surface modification by conjugation with large molecules, polymerisation, and encapsulation in synthetic liposomes. These various solutions of modified hemoglobin are currently undergoing phase I, II, and III clinical trials.



Sources of stroma free hemoglobin



The modifications of natural hemoglobin are realized on highly purified hemoglobin (stroma free hemoglobin, SFH), which is extracted mainly from human or bovine erythrocytes, but can also be recombinant hemoglobin. The preparation of human hemoglobin needs special care for the elimination of viruses. Special methods are needed for the purification of recombinant hemoglobin because endotoxins tightly bind to hemoglobin. For bovine hemoglobin, special care is needed to rule out the presence of prions.



>I>Human hemoglobin



Natural free hemoglobin is obtained from lysis of the erythrocytes contained in expired units of banked blood. Solutions of natural human free hemoglobin have certain undesirable effects, which include vasomotor effects, activation of the complement, kinin and coagulation systems, nephrotoxicity, interference with macrophage function, antigenic effects, histamine release, and iron deposits. Most of these effects, attributed to stromal remnants, are now eliminated or largely limited by the high degree of hemoglobin purification. Free (Fe2+)-hemoglobin in its deoxygenation state is easily oxidized to methemoglobin (Fe3+), and must thus be stored in an anaerobic environment. Extracellular hemoglobin has a high colloid oncotic pressure, which limits its concentration in solution to 7 g/dl. Its affinity for O2 is greater than that of intracellular hemoglobin, because its lacks

2, 3-DPG. This allosteric effector normally interacts with a cluster of 8 positive charges of the b -subunit end of the central cavity of hemoglobin, and facilitates the lowering of O2 affinity <|[18, 19]|>.

Stroma free human hemoglobin has a reduced P50 of 12-14 mm Hg compared to the P50 of 27 mm Hg for intracellular hemoglobin <|[Figure 1]|>, and thus delivers less O2 to the tissues. Recently a mutant form of hemoglobin, the Presbyterian hemoglobin, has been described, with replacement of one amino acid (b -asparagine 108 fi b -lysine) by central cavity of the a-b dimer that results in changes in the allosteric control mechanism with a lower affinity for O2 <|[20]|>. This mutant hemoglobin would be a good basis for further chemical modification to produce a modified hemoglobin solution that presents a convenient affinity for oxygen.



Bovine hemoglobin



Unlike human hemoglobin, bovine free hemoglobin does not require 2, 3-DPG for the control of its affinity for O2, and thus has a P50 of approximately 30 mm Hg. This value favours O2 delivery to the tissues. Bovine hemoglobin thus appeared to be an interesting potential alternative to the human molecule, especially given its abundant availability and low cost. However, other problems currently limit its use the risk of transmission of bovine spongiform encephalopathy, difficulties with purification causing persistence of membrane fragments and consequent possible immune responses and complement activation, and the possible production of antibodies due to infusion of large quantities of bovine proteins.



Recombinant hemoglobin



Scientists at Somatogen, using recombinant technology, demonstrated that recombinant human hemoglobin could be produced by Escherichia coli whose genome was modified by addition of the human genes coding for the globin molecule <|[21]|>. Subsequently, E. coli were designed to produce a di-a-globin molecule in which the 2 a chains are fused, head to tail, and cannot dissociate in plasma <|[22]|>.



To resolve the problem of the affinity of free hemoglobin for O2, the Presbyterian mutation on b chains was reproduced by genetic engineering technology. This final hemoglobin variant is designed rHb 1.1 (Optro, Somatogen) and has a higher P50 than the normal molecule this results in improved O2 delivery at the tissue level. The rHb 1.1 has an oxygen-binding curve similar to that of normal human blood, a P50 of between 30 and 33 mm Hg <|[Figure 1]|>, a plasma half-life four times greater than that of free hemoglobin, a storage half life that is indefinite when frozen, greater than 24 hours at 4C, and five hours at room temperature. For large scale production, the recombinant molecule has to be produced at a high yield, with a high level of gene expression, correct protein folding and assembly, and purification from the other products of E. coli, while maintaining reasonable cost. Second generation recombinant molecules are now cloned, expressed and characterized, and are on the way to be tested. Attempts have been made to produce recombinant hemoglobin by yeast and pigs, but seem to have been stopped.



Modifications of stroma free hemoglobin



An overview of the main techniques and reagents used to modify free hemoglobin with the most important consequences of these modifications are presented on<|[ Figure 2]|> and in <|[Table I]|>.



Intramolecular cross-linking of hemoglobin



The creation of a covalent bond firmly linking the constituant dimers prevents or at least delays the direct dissociation of the tetrameric hemoglobin molecule <|[23]|>. Furthermore, by reaction of the cross-linked hemoglobin with an analogue of
2, 3-DPG, the affinity for O2 can be reduced, even outside the erythrocytes, thus improving tissue O2 delivery.

The intramolecular cross-linking is generally obtained through use of polyanionic cross-linking reagents at specific cationic binding loci (between a Lys 99 or at b subunits at position b Lys 82). Organic phosphates have been shown to secure a linkage between the amino-terminal groups of the b chains, and bifunctional cross-linkers react with the amino acids at the interface between the b chains that are normally involved in the binding of 2, 3-DPG. These cross-linkers influence the functional properties of hemoglobin, changing the affinity for oxygen and carbon dioxide by shifting the hemoglobin towards the R or the T states <|[24]|>.

Many reactions have been proposed and used for cross-linking of SFH. An easy way is the acetylation of hemoglobin at physiologic pH by acetylsalicylic acid (aspirin), or by bis (3, 5 dibromosalicyl) fumarate (DBBF), the diester of dibromo actelylsalicylic acid <|[25]|>. This reaction creates a chemical bond between the a monomers. A typical example of a-a cross-linked hemoglobin is Diaspirin Cross-Linked Hemoglobin (DCLHb) or HemAssist (Baxter), which was prepared by reaction of natural human hemoglobin with DBBF <|[26]|>. A cross-linking between b monomers is obtained by reaction with bis-pyridoxal phosphate. Intramolecular cross-linking was also obtained with open ring-adenosine triphosphate (o-ATP) <|[9]|>. The life of these cross-linked hemoglobins in circulation varies from a few hours to 30 hours, depending on the dose administered and the species of the animal. These new hemoglobins are characterised by a P50 of 30-35 mm Hg.



Intermolecular cross-linking (surface modification) and polymerization of hemoglobin.



The hemoglobin molecule, free or already intramolecularly cross-linked, can be further stabilized by external cross-linking with macromolecules such as hydroxy-ethyl starch, polyethyleneglycol (PEG), Dextran 20 or with artificial support (nanocrystalline beads), or by polymerization with cyanate or glutaraldehyde reagents. Particular intermolecular cross-linking has been designed to obtain changes in the surface electric charges of hemoglobin, with additional reduction of the hemoglobin negative charge by incorporation of an anionic peptide such as glutathione, which is furthermore endowed with antioxidant properties <|[9]|>. This surface change would reduce the extravasation of hemoglobin, and thus increase its plasma life. Polymerization is obtained with cross-linkers that react on surface amino groups and link adjacent molecules resulting in a variety of molecular sizes. External cross-linking and polymerization furnish modified hemoglobins with molecular weight ranging from 64,000 to 400,000 daltons, increasing the risk of immunogenicity but decreasing the renal toxicity, increasing the lifetime in blood and allowing the administration of important concentrations without increase of the oncotic pressure and with a correct delivery of O2 to tissues even when infused at low doses. Pyridoxalated hemoglobin polyoxyethylene (PHP Apex Bioscience) is an example of human hemoglobin first modified with pyridoxal phosphate to lower its affinity for O2, then conjugated with polyoxyethylene. Examples of polymerized hemoglobins are PolyHeme (Northfield) prepared by polymerization of human hemoglobin with glutaraldehyde, HemoLink (Hemosol, Ltd) prepared by cross-linking of human hemoglobin and polymerization with ring-opened raffinose
(o-raffinose) and Hemopure (Biopure Corp) prepared from bovine hemoglobin polymerized with glutaraldehyde <|[4]|>.



Liposome-encapsulation of tetrameric hemoglobin



An alternative method to obtain modified hemoglobin is the encapsulation of stroma free hemoglobin (modified or not) into a synthetic, non-antigenic vesicle prepared from cholesterol and saturated phospholipids <|[27]|>. Encapsulation increases the intravascular half-life, and attenuates the vasoactive effects of free hemoglobin. These packaged hemoglobin molecules have values of P50 and of the Hill coefficient similar to those of blood. The co-encapsulation of protective and regulator molecules such as 2, 3-DPG yields a P50 of 30 mm Hg, and allows a valuable kinetics of binding and off-loading of O2, even faster than that of the erythrocytes <|[28]|>. The major problem with encapsulated hemoglobin remains its short circulation time resulting from rapid phagocytosis and uptake by reticulo-endothelium system, resulting in hepatic overload. Modifications of the surface properties of the artifical membranes of the microcapsules (addition of polysaccharides) could significantly increase the circulation time, but also increased the size of the microcapsules (from 1 to 5 mm). A new promising technique of encapsulation is to use biodegradable nanocapsules (less than 0.2 mm) made of polylactic acid which is degraded in vivo into water and carbon dioxide <|[29, 30]|>. These nanocapsules are stronger and more porous than lipid vesicles, and can contain more hemoglobin molecules, which can be coencapsulated with protective enzymes (superoxide dismutase, catalase, methemoglobin reductase). These nanocapsules are also permeable to glucose and small hydrophilic molecules (reducing molecules for example) allowing a metabolic activity and a correct function of the encapsulated enzymes, so that these nanocapsules could be really named artificial red blood cells.

Among all the blood substitutes based on hemoglobin, the liposome or nanocapsules-encapsulated hemoglobin solutions are the most like native erythrocytes, but they are complex to manufacture, and consequently, they are expensive, what has somewhat slowered their development in clinical studies.



DCLHB a model of intramolecular cross-linked hemoglobin



DCLHb is prepared from human erythrocytes that have been shown to be negative for viruses (in particular HIV, HBV, and HCV). After washing, the cells are lysed to yield hemoglobin, which is filtered to remove stromal elements. After deoxygenating the hemoglobin, 3, 5-dibromosalicyl fumarate is added, in order to create a covalent bond (a fumarate bridge) between lys a1 99 lys a2 99 of the 2a subunits. The product is then pasteurised at 70C, for viral inactivation, and for the denaturation and precipitation of uncrosslinked hemoglobin and other contaminating proteins <|[31]|>. This is followed by re-oxygenation and purification by ion-exchange chromatography.

The final product is added with an electrolyte solution and adjusted at physiologic pH to produce a sterile and nonpyrogenic solution which has the following composition DCLHb 10 g/dl pH (37C) 7.4 Na 146 mM K 4 mM Ca 1.15 mM Mg 0.45 mM Cl 116 mM lactate 34 mM, osmolarity 290 mOsm-L-1 colloid-oncotic pressure (37C) 44 mm Hg. It is frozen at 20C. At this temperature the DCLHb is stable for one year, with minor oxidation leading to the formation of 0.3 methemoglobin (metHb) per month. The solution can be stored for one month in a refrigerator, and one day at room temperature. DCLHb is non-antigenic and does not require typing and cross-matching <|[32]|>. Because the oxy-hemoglobin dissociation curve is shifted to the right, its affinity for O2 is reduced <|[Figure 1]|>. The intramolecular bridge also affects the transport of CO2. DCLHb binds less CO2 than the native molecule, regardless of the concentration of CO2. Only 50 of the binding sites for CO2 are occupied <|[33]|>.



Modified hemoglobin solutions are not substitutes of blood



Modified hemoglobins are not blood substitutes but oxygen carriers, as they do not possess the multiple metabolic functions of erythrocytes <|[34, 35]|>. To succeed in the specific function of oxygen carrier, a modified hemoglobin solution must fulfill a lot of criteria. They firstly must be devoid of toxicity and antigenicity, and must therefore be purified for endotoxins, membrane remnants and lipid removal. Hemoglobin has no antigenic properties even in conjugated forms and recent clinical trials demonstrated that cross-linked, polymerized and encapsulated hemoglobins were not antigenic. Modified hemoglobins must present a high oxyphoric capacity, transporting an O2 quantity at least equivalent to that of 10g hemoglobin/100 ml blood, but also a convenient delivery of this oxygen to tissues, it is a P50 close to that of natural hemoglobin (26 to 27 mm Hg). They must have a lifetime in circulation sufficient to avoid repeated administration and present a catabolism that does not lead to renal toxicity or excessive extravasation. Their viscosity, oncotic, osmotic and rheologic properties must be similar to that of blood. They may neither react wtih O2 and other plasma compounds, nor catalyse the reaction of O2 with plasma compounds. They also must be stable during sterilization and storage, and easy to obtain in large quantities at low or moderate cost.



Preclinical use of modified hemoglobin solutions



Before starting clinical studies, preclinal studies were carried out on hundreds of numerous animal species (rats, rabbits, dogs, pigs, sheeps, monkeys), essentially in models of hypovolemic shock, and isovolomic blood exchanges, but also in traumatic and septic shock. The results were generally encouraging, and opened the way to clinical trials. New modified hemoglobins, and particularly recombinant molecules, continued to be tested in preclinical assays before their use in humans.



Hypovolemic shock and isovolemic blood exchange models



In these models, all types of hemoglobin solutions (cross-linked, polymerized, recombinant, encapsulated, from human and bovine origin) have been used with minor side effects and excellent results, particularly the return to normal cardiac frequency, the return to normal blood lactate levels and acid-base equilibrium, an improvement of tissular extraction of O2, the return to normal blood flow in most organs, a convenient oxygenation of peripheral tissues and a rate of survival not lower than 80 <|[34, 36]|>. However, the return to normal PaO2 values was generally of short duration. No live threatening side effects have been described absence of antigenicity, no complement or polymorphonuclear leukocytes activation, no renal toxicity and no overload of reticulo-endothelial system, minor or no alterations of organs (heart, kidney, central nervous system).

With most of cross-linked hemoglobins, a rapid and sustained increase of mean arterial pressure was generally described. This hypertensive effect has been largely documented for animal studies with DCLHb and other cross-linked hemoglobins <|[37, 38]|>. DCLHb was more effective than infusion of large volumes of lactated Ringers solution in re-establishing and maintaining arterial blood pressure and mixed venous O2 saturations in hemorrhagic shock models. It was as effective as blood, even when the quantities administered reached 50 of those of infused blood. But the return to normal venous O2 values was short-lived, and one frequently noted effect was an early and sustained increase in mean arterial pressure, with most often a decreased heart rate. This pressure increase was dose-dependant, but plateaued quickly and was easily controlled using anti-hypertensive agents. The hypertensive effect could lead to alterations of the regional blood flow in intestines, kidneys and adrenal glands, but in a cat model of exchange-transfusion, there was no evidence of renal dysfunction <|[39]|>. This pharmacologic property of cross-linked hemoglobin would be mediated by three elements of the endogenous vasomotor autoregulatory system inhibition of NO by direct NO uptake or NO synthase inhibition, stimulation of the production of endothelin or of the endothelin binding to specific receptors, and sensitisation and/or potentialisation of a1 and a2 adrenergic receptor responses to catecholamines <|[40, 41]|>. These hypothesis were partially confirmed by using specific molecules acting at the level of NO synthase, of endothelin or adrenergic receptors <|[Figure 3]|>.



Polymerized bovine hemoglobin has also been demonstrated to improve oxygenation when administered after complete isovolemic blood exchange in dogs. The histological examination indicated that no changes had occured in kidney, indicating sufficient tissue oxygenation and lack of renal toxicity. In liver, a slight increase was observed in single cell necrosis as well as the presence of ultrastructural changes of hepatocytes (cytoplasmic protrusions, partial loss of microvilli), swelling of endothelial cells, and marked diminution of glycogen-granula in hepatocytes. However, hepatocyte damage was very discrete, excluding major hepatotoxicity <|[42]|>. Bovine intramolecularly o-ATP and intermolecularly o-adenosine cross-linked hemoglobin with increased surface negative charges to reduce extravasation was tested in 40 of total blood exchange in rats, and demonstrated the absence of nephrotoxic reactions (no transglomerular passage) and pro-oxidant and pro-inflammatory response on isolated human endothelial cells <|[9, 43]|>.

Hemoglobin vesicles were used in animal models of blood exchange showing an oxygen transport capability equivalent to that of red blood cells, and minor effects on platelet number <|[44]|>.



Septic and traumatic shock models



More recently, modified hemoglobin solutions were used with ambivalent results in animal model of septic shock. In septic states, tissue O2 delivery is inadequate in relation to demand. Systemic vascular resistance is low, leading to low systolic and diastolic blood pressures. Modified hemoglobin attenuates the systemic arterial hypotension induced by injection of endotoxin, without affecting regional blood flow to other organs. When low doses of pyridoxalated hemoglobin polyoxyethylene conjugate were used in an ovine model of hyperdynamic sepsis, pulmonary arterial pressure and pulmonary vascular resistance increased, but mean arterial pressure and systemic vascular resistance were normalized, without decreasing blood flow to splanchnic organs and kidney, and without interaction with the host response to sepsis <|[45]|>. But, these results are in contrast with other studies demonstrating that PHP infusion cause death in experimental E. coli peritonitis in mice <|[46]|>, and that modified hemoglobin solution significantly worsened the pulmonary arterial hypertension and the arterial hypoxemia in a pig model of sepsis after administration of endotoxin <|[47]|>.

In traumatic lung injury in pigs, DCLHb led to pulmonary hypertension, worsened lung compliance, and greater pulmonary contusion lesion size <|[48]|>, what indicated that further studies are necessary to assure that the use of modified hemoglobin solution will be of some benefit in thoracic trauma. In a rodent model of severe head injury, DCLHb was demonstrated to improve intracranial pressure without reduction of cerebral blood flow <|[49]|>, but further studies are necessary to assure the safety of DCLHb in brain trauma, particularly for high doses, and the consequences of a transcapillary passage of the free hemoglobin.



Ischemia models



In cases of myocardial ischemia, modified hemoglobins can improve tissue perfusion because of their low viscosity and the small size of the hemoglobin molecules compared to erythrocytes. Modified hemoglobin solutions could, therefore, constitute a treatment for various ischemic states. In the animal, DCLHb has proven to be particularly efficacious in supporting cardiac function during coronary angioplasty. Perfusion of DCLHb through the catheter during balloon occlusion has been demonstrated to improve the oxygenation of the myocardium <|[50]|>. In animal models of cerebral ischemic lesions, isovolemic hemodilution with DCLHb increases cerebral blood flow and oxygenation <|[51]|>.

In a hamster model, DCLHb has also been proven to increase venular red blood cell velocity under postischemic conditions, without inducing enhanced leukocyte-endothelium interaction in the microcirculation of striated skin muscle <|[52]|>.



Clinical studies



The potential uses of modified hemoglobin solutions are the treatment of acute hemorrhage, the perioperative hemodilution (acute normovolemic hemodilution), the treatment of anemia and ischemic states (stroke, acute myocardial infarction, coronary angioplasty), the preservation of organ before transplantation. The use of modified hemoglobins is also planned in the treatment of septic shock where their vasopressor effect is expected to be beneficial, and in cancer to increase local oxygenation during radiotherapy. But, most of the clinical studies that have been performed until now were devoted to blood replacement during surgery, with as major endpoint to avoid or delay autologous blood transfusion <|[35, 38]|>.

Phase I studies on healthy volunteers were used to determine the safety of the material and maximal allowable dosage. These studies have ended and the results were published, indicating that modified hemoglobins no longer have the renal toxicity of natural free hemoglobin. An increase of the blood pressure and gastrointestinal effects were generally described with tetrameric cross-linked and polyhemoglobins, limiting their use to low doses. These effects were not described for pyridoxalated human polyhemoglobin.

<|[Table II]|> resumes the clinical trials that have been already performed, are still ongoing or will start soon
<|[4, 8, 30, 38]|>. For polyhemoglobins, pyridoxalated human hemoglobin is in phase III, pyridoxalated bovine hemoglobin in phase II and o-raffinose human hemoglobin in phase II. For intramolecularly modified hemoglobins, DCLHb reached phase III studies, which were stopped in May 1998, and recombinant human hemoglobin with modified affinity for oxygen (amino acid modification) is in phase I and II clinical trials as a single injection in surgical patients and in acute normovolemic hemodilution. Among the conjugated hemoglobins, pyridoxalated polyoxyethylene hemoglobin and polyethyleneglycol bovine hemoglobin are now in phase II. The results of these studies are far from being totally analysed and published in their complete form. But, from the data of clinical trials published until now, it appears that the intravascular lifetime of modified hemoglobins ranged from 10 to 30 hours. This short intravascular life limits their use to temporary replacement of blood, reducing autologous blood transfusion or delaying it until it could be performed in completely safe conditions. From our experience in orthopedic and cardiac surgery, the use of DCLHb effectively succeeded in this purpose since it could avoid blood transfusion in 20 to 30 of cases <|[38, 53]|>.



Clinical studies with intramolecularly cross-linked hemoglobin



The most studied intramolecularly cross-linked hemoglobin is DCLHb, which received the tradename of HemAssist (Baxter Inc., USA). In a phase I study of 24 healthy conscious volunteers receiving doses of from 25 to 100 mg-kg-1, the most frequently noted complication was mild and transitory abdominal discomfort <|[54]|>. At the same time, arterial hypertension and a dose dependent increase in total creatine phosphokinase (CPK) and iso-LDH5 were seen. A multicentre trial of patients with severe hypovolemic shock consisted of randomisation (within 4 hours of the diagnosis of the shock state) to receive either 50 or 100 mL of 10 DCLHb or normal saline solution. This study showed a dose-dependent reduction in mortality, in complications, and in the incidence of multiple organ failure, without compromise in renal function.

DCLHb (100 to 500 mL of a 10 solution) was used for its vasopressor effect in a phase I study in critically ill patients (septic shock and systemic inflammatory response syndrome) with secondary organ dysfunction. This preliminary study demonstrated an immediate and potent vasopressor effect of DCLHb that allowed reduction in the amounts of pressor drugs administered. These results indicated that intramolecular cross-linked hemoglobins might have future clinical applications with minor side effects in septic shock <|[55]|>.



An investigation of tolerance of DCLHb, using randomisation and a double blind construction was carried out in elective surgery in 82 patients having total hip arthroplasty. Patients received from 25 to 200 mg-kg-1 10 DCLHb or a control of Ringers lactate solution prior to induction of anesthesia. This study again showed an immediate increase in arterial blood pressure of from 10 to 20, peaking at the end of the infusion, and occurring simultaneously with a reduction in heart rate. The vasopressor effect was not dependent on the administered dose and was not associated with increased blood loss. After induction of anaesthesia, arterial pressure decreased in both treatment and control groups, but the 4 groups receiving DCLHb consistently had higher blood pressures and better hemodynamic stability than the control group over the 6 hours following the infusion. The volumes of crystalloid and/or colloid administered were comparable in the 2 groups <|[35, 38]|>.

A randomised, single blind, multicentre phase II study included 70 patients having elective surgery on abdominal aortic aneurysms. These patients were treated with doses of 50, 100, or 200 mg-kg-1 of 10 DCLHb or an equivalent volume of a control infusion of Ringers lactate. The infusions were started after the induction of anesthesia, and lasted 15 minutes. The
2 higher doses of DCLHb significantly increased arterial blood pressure for six hours following the infusion <|[53]|>. This increase did not cause higher blood loss. A phase II study in hemodialysis patients showed improved hemodynamic stability with DCLHb, possibly because the molecule did not transfer into the dialysate renal function in these patients remained stable <|[56]|>. DCLHb was also used in phase II study in acute ischemic stroke in man it increased the mean arterial pressure in correlation with an increase in plasma concentration of endothelin-1 <|[57]|>.

DCLHb has reached phase III studies in orthopedic surgery, general surgery, cardiac surgery, hypovolemic and hemorrhagic shock. In these studies, 750 mL (three 250 mL sacs) were administered. We participated in 2 randomised, single blind, human studies, in orthopedic (n = 24) and cardiac surgery (n = 209 multicentre study). As reported in other studies, we observed an effect of DCLHb infusion on hemodynamic parameters (increase of systolic and diastolic arterial pressure, of systemic vascular resistance with concomitant decrease of heart rate), but these effects plateaued after the first infusion of 250 ml of DCLHb. The main observation of these studies was the effect of DCLHb on the blood saving. In the first postoperative day, 59 of cardiac patients and 92 of orthopedic patients did not need blood transfusion, and blood savings after 7 days were of 33 and 19 respectively. No serious side effects were observed. These results are under way to be published <|[38, 53]|>.

On basis of these first encouraging results, which generally confirmed the rapid, dose-independent vasopressor effect of DCLHb with minor side effects (total plasma bilirubin increase, reduction of platelet count), phase III human studies with infusion of more than 1000 mL of DCLHb were planned (in surgery and in trauma patients), but these clinical trials were suspended in May 1998 for preliminary data evaluation, and are now definitively stopped. One of the suspected reasons for this arrest was the systemic and pulmonary hypertension of this modified hemoglobin, that can aggravate the ischemia already present in many organs (gastro-intestinal tract) after shock. Analysis of the results have led to stop the preparation of DCLHb and to shift attention to develop new generations of modified hemoglobins, especially recombinant hemoglobins, devoid of hypertensive effects.



Clinical studies with intermolecularly cross-linked or polymerized hemoglobins



Phase II clinical trials for surgery and hemodilution, trauma and emergency are planned with HemoLink (Hemosol), a polymerized and cross-linked human hemoglobin prepared with o-raffinose-derived dialdehyde to modify the 2, 3-DPG pocket. This modified hemoglobin has a good P50 without addition of pyridoxal phosphate.

Polymerised bovine hemoglobin (HBOC-201 or Hemopure, Biopure) was tested in healthy volunteers and was subjected to phase I and II studies in orthopedic, cardiac, and urologic surgery. Its intravascular persistance was 20h, and its administration was accompanied by an increase of oxygen diffusion capacity, but also an hypertensive effect <|[58]|>. It was also tested in trauma patients and in the treatment of sickle cell crisis without producing side effects <|[59]|>. Phase III clinical trials are planned with this hemoglobin <|[30]|>, but the recent discovery of bovine spongiform encephalitis transmission might affect the progress of Hemopure development.



Like DCLHb, most of the conjugated and polymerized hemoglobins had hypertensive effects, which were dose-dependent <|[9, 60]|>. An exception would be PolyHeme (Northfield Laboratories). This pyridoxalated, glutaraldehyde-cross-linked human hemoglobin was used in a phase I trial in volunteers and in patients with no evidence of kidney toxicity, blood pressure increase, vasoconstriction, fever or organ dysfunction with doses equivalent to one unit of blood (63 g modified hemoglobin). The product is in phase II and III trials in trauma and emergency surgery patients, with high doses (up to 5 liters) <|[30]|>.



Clinical studies with recombinant hemoglobin



Pilot studies of tolerance were carried out on volunteers with doses ranging from 15 to 320 mg/kg <|[61]|>. No renal, hepatic, or pulmonary toxicity was noted, but an hypertensive effect was observed, together with a decrease of heart rate. There was no renal excretion. Fever, higher than 38C, occured between 3 and 8 hours after infusion, with headache, myalgia, chills and gastrointestinal effects <|[61, 62]|>. These symptoms resolved either spontaneously or after ibuprofen (400 or 600 mg). By increasing the purification process, no further episodes of fever were noted. A first phase I multicenter trial was centered on the intraoperative use of rHb 1.1 (Optro, Somatogen) at doses ranging from 25 to 100 g to restore blood volume and oxygen delivery, and showed minor adverse effects such as elevated lipase and amylase in the 12 to 24 hours, returning to normal at 48 h, and without clinical signs of pancreatitis. A phase II study has been performed with 12.5 to 50 g Optro in acute normovolemic hemodilution in patients with total hip replacement and found no adverse effects. Phase II studies with Optro are underway in North America for coronary artery bypass surgery. Further studies, for normovolemic hemodilution and in oncology, have been planned. Optro has also received approval for use in conjugation with erythropoietin in patients with end-stage renal disease or refractory anemia. In these patients, erythropoietin appears to act synergistically in stimulating erythropoiesis by still non-understood mechanisms <|[4, 30]|>. But, it appears that the clinical trials with Optro were recently stopped, since new more efficient recombinant modified hemoglobins are under study.



Clinical studies with encapsulated hemoglobins


Initial tolerance studies in the animal with first generation encapsulated Hb revealed side effects such as bradycardia, leukopenia, thrombopenia, increases in the values of transaminases and bilirubin, complement activation, hypertension, and decreases in cardiac output, especially with isovolemic exchange <|[63]|>. A lipid contaminant, lysolecithin, capable of activating complement, was the cause of these effects. The second generation of liposomes contain synthetic phosphatidylcholine and an antagonist of the activation of tissue platelet factor. Third generation liposomes (a lyophilised preparation) are beneficial in the treatment of hemorrhagic shock, where they increase PaO2, improve hemodynamic indices, and survival <|[64]|>. Nonetheless, among their undesirable side effects, it should be noted that these liposomes bind endotoxin, and that lipopolysaccharide (LPS) and encapsulated hemoglobin can exacerbate the manifestations of septic shock. In terms of the elimination of these particles from the circulation, the reticulo-endothelial systems of the liver and spleen are the primary areas for this function some degree of overload of these organs could thus be expected after administration of liposomal hemoglobin.



Limitations and unsolved problems with hemoglobin solutions



Limitations



Because modified hemoglobin solutions do not require compatibility testing, have low viscosity, do not pose an infectious risk, and have favourable O2 transport properties, their clinical use would appear to be promising. Studies to date have shown an absence of toxicity and immunogenicity <|[65]|>, and only minor side effects, the most consistent of which is a rapid but transitory increase of systemic arterial blood pressure, and gastrointestinal complaints. Modified hemoglobins favour tissue perfusion and oxygenation, and could reduce the incidence of ischemic phenomenon. Nonetheless, the modified hemoglobin solutions do not fulfil the numerous other roles of the blood, including regulatory, metabolic, and defence functions. Further, their metabolic pathways are poorly characterised, and their plasma half-life is short, ranging from 12 hours for cross-linked hemoglobins to about 2 days for polymerized-hemoglobins, compared to the 120 days of human erythrocytes. They are thus best seen as an emergency substitute, useful in the short term they cannot replace transfusions in certain pathologic situations such as chronic anemia, but can postpone (or even eliminate) the need to transfuse homologous blood. The use of modified hemoglobin solutions also poses a certain number of practical problems. Extracellular hemoglobin can stimulate, or mask, post-transfusion hemolysis. The hematocrit value after use of these solutions no longer faithfully reflects O2 transport capacity. SpO2 and mixed venous O2 saturations are still measured correctly, but the corresponding values of PO2 no longer have the same meaning, given the differences in P50 values between intra-erythrocytic and extracellular hemoglobins. The presence of plasmatic hemoglobin can lead to false values from machines that measure concentrations optically, such as that of bilirubin. Similarly, the functioning of machines designed to wash red blood cells, such as blood recovery devices, is disturbed by the presence of plasmatic hemoglobin, and the measurement of several blood enzymes and compounds is perturbed by the red colour of plasma, at least in the 24 hours following the administration of the blood substitute <|[66]|>.



Hypertensive effect



The vasoconstrictive effect of cell-free hemoglobin remains the first problem to resolve before starting large developments of hemoglobin-based blood substitutes. If the scavenging of NO is at the origin of vasoconstriction, it will be difficult to construct hemoglobins with limited capacities for NO without limiting simultaneously the binding of O2. If vasoconstriction is a result of nitrosohemoglobin formation, recombinant hemoglobins without this binding capacities may be designed. The hypertensive effect of modified hemoglobin can also be linked to the ability of the tetrameric molecule to cross the endothelial barrier and to enter the interstitial space where it binds and removes nitric oxide. The endothelial permeability of intramolecularly cross-linked hemoglobins is higher than that of conjugated hemoglobins, and was intensified when endothelium was pretreated with IL6 or endotoxin <|[67]|>. Little permeability was observed with encapsulated hemoglobin.



This tissular NO scavenging would also explain the gastrointestinal effects since NO affects the nerve plexus and smooth muscles resulting in esophageal spasm and gastrointestinal effects. The modification of the surface electric charges of modified hemoglobin towards negative charges would limit the interstitial passage of free hemoglobin and its NO-scavenging effects. Surface modifications of hemoglobins by polymerization and conjugation were designed in this purpose, and new generations of hemoglobin molecules appeared to have succeeded in decreasing or suppressing the hypertensive effect of free hemoglobins. The hypertensive effect of free hemoglobin has also been attributed to excessive delivery of oxygen to arterioles leading to autoregulatory vasoconstriction <|[68]|>. This hypothesis would perhaps better explain the kinetics of the hypertensive effect, which is observed in the minutes following the infusion, even at low doses, and is not clearly dose-dependent. The hypertensive effect has also been explained by direct effect on endothelin secretion, an hypothesis that was only partially sustained by the use of an endothelin receptor antagonist, which only attenuated the hypertensive effect of DCLHb <|[69]|>. Whatever the explanation to this hypertensive effect, attempts are now made to design new generations of blood substitutes that could control vasoactivity. The preparation of S-nitrosothiols-hemoglobin (transporting NO) is a promising new way of research <|[70]|>.



Intravascular reactivity



Hemoglobin is a reactive molecule that can generate oxidant species. Inside the erythrocytes, a redox equilibrium is supplied by the presence of enzymes (superoxide dismutase, catalase, methemoglobin reductase, glutathione peroxidase) and antioxidant molecules. Out of the red blood cell, hemoglobin can react with plasma compounds, leading to oxidations. An important problem that could arrive with free hemoglobin is linked to the susceptibility of deoxyhemoglobin to oxidation, leading to the production of methemoglobin, which has a peroxidative activity and forms further reactive O2 species <|[7, 8]|>. This oxidation of hemoglobin into methemoglobin also easily releases hemin, which rapidly associates with membranes, leading to cytotoxicity. The cross-linking does not decrease the peroxidative activity of hemoglobin. Inside the erythrocyte, enzymes and specific compounds protect the hemoglobin molecule, but these compounds are absent in the solutions of modified hemoglobin <|[Figure 4]|>. To mimick the natural antioxidant environment of hemoglobin as it exists in erythrocytes, studies have been carried out to cross link catalase and superoxide dismutase to polyhemoglobin or to bind other molecules that could reduce the oxidant activity of hemoglobin, particularly its capacity to form ferryl radicals <|[7]|>. Attempts are made to prepare hemoglobins co-encapsulated with erythrocyte enzymes such as catalase, carbonic anhydrase, superoxide dismutase



The autoxidation of cross-linked hemoglobin and the release of the heme would be more rapid than with native hemoglobin, so that cross-linked hemoglobin would cause a higher rate of induction of heme oxygenase in endothelial cells <|[10]|>, and contribute indirectly to oxidative stress on the endothelium. The formation of methemoglobin thus, not only lowers the effectiveness of administered modified hemoglobin, but is the source of potentially toxic ferryl hemoglobin, hemin, bilirubin and free iron <|[7, 8, 71, 72]|>. It also seems that free hemoglobin could mediate reactions leading to the oxidation of low density lipoproteins (dityrosine formation). In consequence, LDL is no transformed and accumulates in blood <|[73]|>.



Other risks and unsolved problems



Several other risks must be considered. The initial concerns that polymerization of hemoglobin to produce macromolecules might alter antigenicity have proven unfounded with modified human hemoglobin. Modified hemoglobin would complex endotoxins, at least in vitro, increasing the biological activity of these compounds, and would increase lethality in gram-negative sepsis by increasing the TNF production by mononuclear and Kupffer cells <|[74]|>. Large doses of free hemoglobin would also have a bacterial growth-enhancing effect by releasing iron. Liposome-encapsulated hemoglobin appears to modulate the TNF-a production by mononuclear phagocytes, playing a possible modulating role on immune function that would need further studies <|[75]|>.

Finally, it has to be underlined that little is known about the interactions of the modified hemoglobins with haptoglobin and about their catabolism (possible toxicity of large amounts of bilirubin. More studies in this way are needed before to consider that modified hemoglobin solutions are absolutely safe and useful blood substitutes, taking also into account that hemoglobin solutions are only effective for 24 hours, with a cost that will probably be higher than that of packed red cell preparations.



Conclusions



The clinical use of cell free hemoglobin solution has so far been limited by acute toxic side effects until ultrapurification and chemical modifications have been established in the production process of hemoglobin-based oxygen carriers. Free-hemoglobin is now purified to a high degree. Nevertheless, it still has unfavorable side effects, that have been predicted for a long time.

But, the variety of research work and scientific papers that appeared on the role of free hemogmlobin in oxidative reactions seems to have not influenced the tentatives to use modified free hemoglobin solutions as oxygen carriers. Moreover, added to this oxidative capacity of modified free hemoglobins, little is known about their ways of degradation in vivo. Outside the erythrocyte, hemoglobin is subject to degradation and heme loss, and is able to react with plasma compounds. It readily diffuses in the plasma space were it degrades in tissue leading to a jaundice-like effect lasting for several days in patients. The mechanisms of these side-effects are not clearly understood so that they cannot be controlled.



The use of modified free hemoglobin solutions is subjected to other problems such as nitric oxide capture leading to hypertensive effects and interactions with monocytes/macrophages leading to inflammation reactions. Unless these problems are solved and the ways of degradation of free hemoglobin are explained, the use of hemoglobin-based products as oxygen carrier will not be safe in human, especially in situations of acute hemorrhagic shock and sepsis <|[76]|>.

The solution to blood substitute for clinical use would perhaps be found in new recombinant molecules which are being designed with specific mutations and are on the way to be used in clinical trials. Another solution would be to succeed in the production of complete erythrocytes by three-dimensional cultures or hemoglobin co-encapsulated with molecules,
closely mimicking erythrocytes.

Appendix: 

1. Tomasulo P. Transfusion alternatives. Impact on blood banking worlwide. In Winslow RM, Vandegriff KD, Intaglietta M (eds). Blood substitutes physiological basis of efficacy. Boston Birkhuser, 1995: 1-19.

2. Winslow RM. Blood substitutes- a moving target. Nat Med, 1995 1 1212-1215.

3. Dietz NM, Joyner MJ, Warner MA. Blood substitutes fluids, drugs or miracle solutions Anesthes Analg, 1996 82 390-405.

4. Winslow RM. New transfusion strategies red cell substitutes. Ann Rev Med, 1999 50 337-353.

5. Perutz MF. Stereochemistry of cooperative effects in haemoglobin. Nature, 1970 228 726-739.

6. Winterbourn CC. Oxidative reactions of hemoglobin. Meth Enzymol, 1999: 186 265-272.

7. Everse J, Hsia N. The toxicities of native and modified hemoglobins. Free Radic Biol Med, 1997 22 1075-1099.

8. Deby-Dupont G, Pincemail J, Lamy M. Hemoglobin-based red cell substitute preliminary human studies. In Vincent JL, (ed). Yearbook of Intensive Care and Emergency Medicine. Berlin Springer-Verlag, 1994: 264-275.

9. Simoni J, Simoni G, Hartsell A, Feola M. An improved blood substitute. In vivo evaluation of its renal effects. ASAIO J, 1997 43: M714-M725.

10. Balla J, Nath KA, Balla G, Juckett MB, Jacob HS, Vercellotti GM. Endothelial cell heme oxygenase and ferritin induction in rat lung by hemoglobin in vivo. Am J Physiol, 1995 268: L321-L327.

11. Von Stark. Die Resorbarheit des haimatins und die bedeutung der hemoglobin-preparate. Deutsche Medizinische Wochenschrift, 1998 24 805-808.

12. Sellards AW, Minot GR. Injection of hemoglobin in man and its relation to blood destruction with especial reference to the anemias. J Med Res, 1916 34 469-494.

13. Amberson WR. Blood substitutes. Biol Rev, 1937 12: 48-52.

14. Amberson WR, Jennings JJ, Rhode CM. Clinical experience with hemoglobin-saline solution. J Appl Physiol, 1949 1 469-489.

15. Rabiner SF, OBrien K, Peskin GW, Friedman LH. Further studies with stroma-free hemoglobin. Ann Surg, 1970 171 615-622.

16. Savitsky JP, Doczi J, Black J, Arnold JD. A clinical safety trial of stroma-free hemoglobin. Clin Pharmacol Ther, 1978 23 73-80.

17. Feola M, Simoni J, Tran R, Canizaro P. Mechanisms of toxicity of hemoglobin solutions. Biomater Artif Cells Artif Organs, 1988 16 217-226.

18. Benesch R, Benesch RE. Intracellular organic phosphates as regulators of oxygen release by haemoglobin. Nature, 1969 221 618-622.

19. Gottfried DS, Juszczak LJ, Fataliev NA, Acharya S, Hirsch RE, Friedman JM. Probing the hemoglobin central cavity by direct quantification of effector binding using fluorescence lifetime methods. J Biol Chem, 1997 272 1571-1578.

20. Gottfried DS, Manjula BN, Malavalli A, Acharya AS, Friedman JM. Probing the diphosphoglycerate binding pocket of HbA and HbPresbyterian (beta 108Asn fi Lys). Biochemistry, 1999 38 11307-11315.

21. Hoffman SJ, Looker D, Roehrich JM, Cozart PE, Durfee SL, Tedesco JL, Stetler GL. Expression of fully functionnal human hemoglobin in Escherichia Coli. Proc Nat Acad Sci USA, 1990 87 8521-8525.

22. Looker D, Abbott-Brown D, Cozart P, Durfee S, Hoffman S, Mathews AJ, Miller-Roehrich J, Shoemaker S, Trimble S, Fermi G. A human recombinant haemoglobin designed for use as a blood substitute. Nature, 1992 356 258-260.

23. Yang T, Olsen KW. Thermal stability of Hb cross-linked in the T state by bis (3,5 dibromosalicyl) fumarate. Biochem Biophys Res Comm, 1991 174 518-523.

24. Johnson MB, Adamson JG, Mauk AG. Functional comparison of specifically cross-linked hemoglobins biased toward the R and T states. Biophys J, 1998 75 3078-3084.

25. Walder JA, Zangg RH, Walder RY, Steele JM,
Klotz IM.
Diaspirins that cross-link beta chains of hemoglobin bis (3,5 dibromosalicyl) succinate and bis (3,5 dibromosalicyl) fumarate. Biochemistry, 1979 18 4265-4270.

26. Przybelski RJ, Daily EK. The pressor/perfusion effect of Diaspirin Cross-linked Hemoglobin (DCLHb). In Vincent JL, (ed). Yearbook of Intensive Care and Emergency Medicine. Berlin Springer Verlag, 1994: 252-263.

27. Beissinger MC, Farmer RL, Gossage JL. Liposome encapsulated Hb as a red cell surrogate. Trans Am Soc Artif Internal Organs, 1986 32 58-63.

28. Rudolph AS. Encapsulation of hemoglobin in liposomes. In Winslow RM, Vandegriff KD, Intaglietta M, (eds). Blood Substitutes Physiological Basis of Efficacy. Boston Birkhaser, 1995: 90-104.

29. Chang TMS. Modified hemoglobin blood substitutes present status and future perspectives. Biotechnol Annu Rev, 1998 4 75-112.

30. Chang TMS. Future prospects for artificial blood. Trends Biotechnol, 1999 17 61-67.

31. Farmer M, Ebeling A, Marshall T, Hauck W, Sun CS, White E, Long Z. Validation of virus inactivation by heat treatment in the manufacture of Diaspirin Cross-linked Hemoglobin. Biomater Artif Cells Immobil Biotechnol, 1992 20 (2-4): 429-433.

32. Estep TN, Gonder J, Bornstein I, Aono F. Immunogenicity of Diaspirin Cross-linked human hemoglobin solutions. Biomater Artifi Cells Immobil Biotechnol, 1992 20 (2-4): 603-609.

33. Vandegriff KD, Le Tellier YC, Winslow RM. Determination of the rate and equilibrium constants for oxygen and carbon monoxide binding to R-state human hemoglobin cross-linked between the a subunits at lysine 99. J Biol Chem, 1991 266 17049-17059.

34. Remy B, Deby-Dupont G, Lamy M. Apports et perspectives des drivs de lhmoglobine. Schweiz Med Wochenschr J, 1997 127 1088-1096.

35. Remy B, Deby-Dupont G, DAns V, Ernest P, Lamy M. Substituts des globules rouges mulsions de fluorocarbures et solutions dhmoglobine. Ann Fr Anesth Ranim, 1999 18 211-224.

36. Gulati A, Sen AP. Dose-dependent effect of diaspirin cross-linked hemoglobin on regional blood circulation of severely hemorrhaged rats. Shock, 1998 9 65-73.

37. Przybelski RJ, Malcolm DS, Burris DG, Winslow RM. Cross-linked hemoglobin solution as a resuscitative fluid after hemorrhage in the rat. J Lab Clin Med, 1991 117 143-151.

38. Remy B, Deby-Dupont G, Lamy M. Red blood cell substitutes fluorocarbon emulsions and haemoglobin solutions. Brit Med Bull, 1999 1 277-298.

39. Ulatowski JA, Nishikawa T, Matheson-Urbaitis B, Bucci E, Traystman RJ, Koehler RC. Regional blood flow alterations after bovine fumaryl bb-crosslinked hemoglobin transfusion and nitric oxide synthase inhibition. Crit Care Med, 1996 24 558-565.

40. Schultz SC, Grady B, Cole F, Hamilton J, Burhop K, Malcolm DS. A role of endothelin and nitric oxide in the pressor response to Diaspirin Cross-linked Hemoglobin. J Lab Clin Med, 1993 122 301-308.

41. Gulati A, Rebello BS. Diaspirin Cross-linked Hb Involvement of adrenergic mechanisms in the pressor effect. Artif Cells Blood Substit Immobil Biotechnol, 1994 22 603-612.

42. Lipfert B, Standl T, Dullmann J, Helmchen U,
Schulte am Esch J, Lorke DE.
Histology and ultrastructure of liver and kidney following blood exchange with ultrapurified, polymerised bovine hemoglobin in comparison with hydroxyethyl starch. Lab Invest, 1999 79 573-582.

43. Simoni J, Simoni G, Martinez-Zaguilan R, Wesson DE, Lox CD, Prien SD, Kumar RV. Improved blood substitute. Evaluation of its effects on human endothelial cells. ASAIO J, 1998 44 M356-M367.

44. Izumi Y, Sakai H, Hamada K, Takeoka S, Yamahata T, Kato R, Nishide H, Tsuchida E, Kobayashi K. Physiologic responses to exchange transfusion with hemoblobin vesicles as an artificial oxygen carrier in anesthetized rats changes in mean arterial pressure and renal cortical tissue oxygen tension. Crit Care Med, 1996 24 1869-1873.

45. Bone HG, Schenarts PJ, Fischer SR, McGuire R, Traber LD, Traber DL. Pyridoxalated hemoglobin polyoxyethylene conjugate reverses hyperdynamic circulation in septic sheep. J Appl Physiol, 1998 84 1991-1999.

46. Griffiths EA, Cortes N, Gilbert P, Stevenson P, Macdonald S, Pepper D. Haemoglobin-based blood substitutes and sepsis. Lancet, 1995 345 158-160.

47. Aranow JS, Wang H, Zhuang J. Effect of human hemoglobin on systemic and regional hemodynamics in a porcine model of endotoxemic shock. Crit Care Med, 1996 24 807-814.

48. Cohn SM, Zieg PM, Rosenfield AT, Fisher BT. Resuscitation of pulmonary contusion effects of a red cell substitute. Crit Care Med, 1997 25 484-491.

49. Piper JR, Garrioch MA, Souter MJ, Andrews PJD, Thomson D. Effects of diaspirin cross-linked hemoglobin on post-traumatic cerebral perfusion pressure and blood flow in a rodent model of diffuse brain injry. Br J Anaesth, 1998 80 639-643.

50. McKenzie JE, Cost EA, Scandling DM, Ahle NW, Savage RW. Effects of diaspirin cross-linked hemoglobin during coronary angioplasty in the swine. Cardiovasc Res, 1994 28 1188-1192.

51. Cole DJ, Drummond JC, Patel PM. Effects of viscosity and oxygen content on cerebral blood flow in ischemic and normal rat brain. J Neurol Sci, 1994 124 15-20.

52. Nolte D, Botzlar A, Pichelmann S, Boukela E,
Messmer K.
Effects of diaspirin-cross-linked hemoglobin (DCLHb) on the microcirculation of striated skin muscle in the hamster a study on safety and toxicity. J Lab Clin Med, 1997 130 314-327.

53. Lamy ML, Daily EK, Brichant JF, Larbuisson RP, Demeyere RH, Vandermeersch EA, Lehot JJ, Parsloe MR, Berridge JC, Sinclair CJ, Baron JF, Przybelski RJ. A randomized trial of Diaspirin Cross-linked Hemoglobin (DCLHb) solution as an alternative to blood transfusion following cardiac surgery. Anesthesiology (in press).

54.Pzybelski RJ, Daily EK, Kisicki JC, Mattia-Goldberg C, Bounds MJ, Colburn WA. Phase I study of the safety and pharmacologic effects of diaspirin cross-linked hemoglobin solution. Crit Care Med, 1996 24 1993-2000.

55. Reah G, Bodenham AR, Mallick A, Daily EK, Przybelski RJ. Initial evaluation of diaspirin cross-linked hemoglobin (DCLHb) as a vasopressor in critically ill patients. Crit Care Med, 1997 25 1480-1488.

56. Swan S, Halstenson C, Collins AJ, Colburn WA, Blue J, Przybelski RJ. Pharmacologic profile of diaspirin cross-linked hemoglobin in hemodialysis patients. Am J Kidney Dis, 1995 26 918-923.

57. Saxena R, Wijnhoud AD, Man in't Veld AJ, van den Meiracker AH, Boomsma F, Przybelski RJ, Koudstaal PJ. Effect of dispirin cross-linked hemoglobin on endothelin-1 and blood pressure in acute ischemic stroke in man.
J Hypertens, 1998 16 1459-1465.

58. Hughes GS, Antal EJ, Locker PK, Francom SF,
Adams WJ, Jacobs EE.
Physiology and pharmacokinetics of a novel hemoglobin-based oxygen carrier in humans. Crit Care Med, 1996 24 756-764.

59. Feola M, Simoni J, Angelillo R, Luhruma Z, Kabakele M, Manzombi M, Kaluila M. Clinical trial of hemoglobin based blood substitute in patients with sickle cell anemia. Surg Gynecol Obstet, 1992 174 379-386.

60. Kasper SM, Walter M, Grune F, Archer-Paik LE,
Scoggin CH.
Effect of a hemoglobin-based oxygen carrier (HBOC-201) on hemodynamics and oxygen transport in patients undergoing preoperative hemodilution for elective abdomnal aortic surgery. Anesth Analg, 1996 83 921-927.

61. Shoemaker SA, Gerber MJ, Evans GL,
Archer-Paik LE, Scoggin CH.
Initial clinical experience with a rationally designed genetically engineered recombinant human Hb. Artif Cells Blood Substi Immobil Biotechnol, 1994 22 (suppl. 3): 457-465.

62. Viele MK, Weiskopf RB, Fisher D. Recombinant human hemoglobin does not affect renal function in humans analysis of safety and phamacokinetics. Anesthesiology, 1997 86 848-858.

63. Rudolph AS. Transient changes in the mononuclear phagocyte system following administration of the blood substitute, liposome encapsulated hemoglobin. Biomaterials, 1994 15 796-804.

64. Rabinovici R, Rudolph AS, Feuerstein G. A new salutary resuscitative fluid Liposome Encapsulated Hemoglobin (LEH) - hypertonic saline solution. J Trauma, 1993 35 121-127.

65. Patel MJ, Webb EJ, Shelbourn TE, Mattia-Goldberg C, George AJ, Zhang F, Moore EG, Nelson DJ. Absence of immunogenicity of diaspirin cross-linked hemoglobin in humans. Blood, 1998 91 710-716.

66. Ma Z, Monk TG, Goodnough LT, McClellan A, Gawryl M, Clark T, Moreira P, Keipert PE, Scott MG. Effect of hemoglobin- and perflubron-based oxygen carriers on common clinical laboratory tests. Clin Chemistry, 1997 49 (9): 1732-1737.

67. Nakai K, Sakuma I, Ohta T, Ando J, Kitabatake A, Nakazato Y, Takahashi TA. Permeability characteristics of hemoglobin derivatives across cultured endothelial cell monolayers. J Labo Clin Med, 1998 132 313-319.

68. Intaglietta M, Johnson PC, Winslow RM. Microvascular and tissue oxygen distribution. Cardiovasc Res, 1996 32 632-643.

69. Gulati A, Sharma AC, Singh G. Role of endothelin in the cardiovascular effects of diaspirin crosslinked and stroma reduced hemoglobin. Crit Care Med, 1996 24 137-147.

70. Jia L, Bonaventura C, Bonaventura J, Stamler JS. S-nitrosohaemoglobin a dynamic activity of blood involved in vascular control. Nature, 1996 380 221-226.

71. Van Dyke BR, Saltman P. Hemoglobin a mechanism for the genration of hydroxyl radicals. Free Radic Biol Med, 1996 20 985-989.

72. Svistunenko DA, Patel RP, Voloshchenko SV,
Wilson MT.
The globin-based free radical of ferryl hemoglobin is detected in normal human blood. J Biol Chem, 1997 272 (11): 7114-7121.

73. Ziouzenkova O, Asatryan L, Akmal M, Tetta C, Wratten ML, Loseto-Wich G, Jurgens G, Heinecke J, Sevanian A. Oxidative cross-linking of ApoB100 and hemoglobin results in low density lipoprotein modification in blood. Relevance to atherogenesis caused by hemodialysis.
J Biol Chem, 1999 274 18916-18924.

74. Su D, Roth RI, Levin J. Hemoglobin infusion augments the tumor necrosis response to bacterial endotoxin (lipopolysaccharide) in mice. Crit Care Med, 1999 28 771-778.

75. Rudolph AS, Cliff R, Kwasiborski V, Neville L, Abdullah F, Rabinovici R. Liposome-encapsulated hemoglobin modulates lipopolysaccharide-induced tumor necrosis factor a production in mice. Crit Care Med, 1997 25 460-468.

76. Zimmerman JJ. Deciphering the dark side of free hemoglobin in sepsis. Crit Care Med, 1999 27 685-686.