Colloid - an overview | ScienceDirect Topics (2023)

Use of Colloids

The problem of “fluid creep” as a result of crystalloid resuscitation was one of the early drivers of the use of colloid resuscitation. In fact, the use of plasma for burn resuscitation was described in the early 1940s by Henry Harkins, who recommended 1 L of plasma for each 10% TBSA burned for patients with burns to more than 10% TBSA. The erroneous hypothesis for the use of colloids was based on the observation of hemoconcentration after a major burn, and the deduction that this must have resulted from a loss of plasma volume; therefore, colloid solutions containing macromolecules would theoretically increase plasma colloid osmotic pressure and act as better intravascular volume expanders than crystalloid fluid alone. Unfortunately, measurement of colloid oncotic pressure and resuscitation based on those measurements proved unsuccessful in improving outcomes of many forms of shock, including major burn injury.

During the 1970s, resuscitation tactics using balanced salt solutions arose, leading to a shift away from early provision of colloids. Moyer and others conducted resuscitation studies with lactated Ringer’s solution and concluded that burn shock likely arose from an extravascular sodium deficiency that was best treated with balanced salt solution. Baxter and Shires also found that extracellular sequestration of sodium and water resulting from decreased transmembrane potential, at the expense of decreased intracellular sodium and water occurred early in burn injury. They advocated successfully for resuscitation only with balanced salt solutions in the first 24 hours. Pruitt and others found that use of colloid early in burn resuscitation did not appear to augment intravascular retention of fluid because increased extracellular oncotic pressure caused by capillary leak syndrome of sufficient magnitude results in loss of membrane integrity: the membrane becomes permeable to albumin and other macromolecules until equilibrium is reestablished along the new concentration gradient that promotes transmembrane flux of albumin. Without treatment, the clinical consequences are hypovolemia and tissue edema that may persist until microcirculatory integrity is restored and the patient begins to mobilize and excrete the fluid administered for resuscitation. As a result of these works, colloid-free resuscitation in the first 24 hours post burn injury became the standard of care in burn resuscitation.

Colloid use in burn resuscitation might have become obsolete had Pruitt not described the problem of fluid creep. The phenomenon refers to the consequences of aggressive crystalloid resuscitation. Edema-related complications were noted, such as massive facial and airway swelling requiring prolonged endotracheal intubation, extremity compartment syndromes (including of unburned extremities), and respiratory and cardiac failure. Other resuscitation-related complications now reported as consequential to overresuscitation include abdominal compartment syndrome and rarely permanent blindness (as a result of prolonged elevated intraocular pressure).

Benefits

Colloid solutions produce vascular volume expansion with less interstitial expansion than do crystalloid solutions. They support colloid osmotic pressure and are useful in patients who are symptomatic from their hypoalbuminemia. In theory, the larger molecular weight colloids, because of their molecular size and configuration, might seal capillary leaks in patients with capillary leak syndrome.¹² Some have postulated that capillaries have both small pores, which reflect colloids, and large pores, which do not. The capillary leak syndrome that occurs in critical illness causes a change in the number of these pores but not in their size. For this reason, the molecular weight distribution of hydroxyethyl starch would not be a determining factor.¹³ Some studies have supported the benefits of colloids in capillary leak,^14,15 others have not.^13,16

Crystalloids or Colloids for Intravascular Plasma Volume Expansion

Although crystalloids are the most rational choice of fluid for replacement of evaporative losses, maintenance fluid requirements, and expansion of the entire extracellular fluid volume, the choice of crystalloid or colloid for plasma volume replacement in the perioperative phase is not clear. This underlines the lack of adequately powered perioperative studies directly comparing the two fluid types when administered in a similar fashion. Although most GDT studies use colloid for their intravascular volume expansion, these studies are also comparisons of fluid administration guided by physiologic endpoints with nonguided therapy. Whether the same level of benefit can be achieved by crystalloid-based GDT merits further investigation. Crystalloid may be effective in plasma volume expansion (PVE) at amounts less than previously reported, although typically 40% to 50% more crystalloid than colloid must be administered to obtain the same clinical volume effect.³ When combined with the increased propensity of crystalloid to filter across the capillary membrane, more extravascular volume expansion occurs, potentially causing tissue edema. When compared with colloids, crystalloids may lead to increased GI mucosal edema¹⁵⁹ and the potential for delayed postoperative GI function and bacterial translocation. The differential effect of crystalloids when compared to colloids on tissue O₂ tension has no clear consensus.⁵² The limited data comparing colloid with crystalloid for perioperative PVE may lead clinicians to extrapolate findings from studies in critical care. A Cochrane review highlighted the lack of improvement in all-cause mortality when colloids were used for intravascular volume expansion in unselected critical care populations.¹⁶⁰ In studies specific to patients with sepsis, starch-based colloids, including smaller MW versions, are associated with an increased requirement for renal replacement therapy, blood transfusion, and an increase in severe adverse events.⁸³ These data should be interpreted cautiously. First, some critical care studies compared starches with saline in the control group, which may itself be associated with renal problems.⁵¹ Second, the surgical population has a different physiologic phenotype to those in critical care; meta-analysis suggests that starch is not associated with excess mortality or kidney injury in surgical patients, although the available trials are limited.⁸⁶ Despite these limitations, it is reasonable to avoid starch colloids in perioperative patients with severe sepsis or at increased risk of renal failure, pending the necessary large-scale trials assessing their safety in the perioperative setting. This is reflected in license restrictions for these fluids in the United States. In the United Kingdom, the use of starches has been suspended in all settings. The potential toxicities of colloids must be weighed against the potential fluid overloading effects of PVE using crystalloids until more data are available informing the debate on crystalloids versus colloids for perioperative PVE.

2.1 Synthetic Procedure of Colloidal Solution of Exfoliated Titanate Layers with Micrometric Dimensions

Colloidal solutions of exfoliated single oxide layers are adequate as raw materials to sandwich large enzyme molecules between thin oxide layers with less than 1nm in thickness. We can find a number of studies regarding synthesis of colloidal solutions of exfoliated layered metal oxides. In general, parent powder of a layered metal oxide with anionic oxide layers and inlayer cations is prepared by a solid-state reaction at high temperature. Subsequently, the interlayer cations are exchanged for bulky tetraalkylammonium cations in a basic solution to exfoliate to single oxide layers (nanosheets). The obtained colloidal solution is mixed with a solution of the intended enzymes, resulting in formation of nanohybrids between the layers and the enzymes. Here, a fabrication procedure of Fe-doped titanate layers with potassium ions in the interlayer (K_0.8Fe_0.8Ti_1.2O₄) is introduced in detail (Harada, Sasaki, Ebina, & Watanabe, 2002).

Colloids

Human albumin (4%–5%) in saline is considered to be the reference colloidal solution. It is fractionated from blood and heat treated to prevent transmission of viruses. It has many theoretical advantages, especially in animal studies, but clinical studies have not shown outcome differences. Its main theoretical advantage is that, compared with crystalloids, it is less inflammatory. This may be because it is a natural molecule. Other than its dilutional effect, albumin is associated with minimal coagulopathy. No clinical evidence has shown that albumin is better than other colloids, but the (SAFE) study in Australia has shown 4% albumin to be safe, compared with normal saline, in ICU patients.⁶⁰ The SAFE study, whose main intent was to show equivalency, found no difference in the primary outcome (28-day mortality rate) or in any secondary outcome. The Committee on Tactical Combat Casualty Care has recommended a low-volume resuscitation fluid using 500 mL of Hextend for tactical reasons. The reason for that choice was that 7.5% HTS is not commercially available. The adoption of damage control or hemostatic resuscitation has been thought to result in improved outcome, decreased blood use, and decreased incidence of ARDS.⁶¹ ARDS and MODS still occur, but at a much lower rate than previously seen.

There are still other advantages of 25% albumin over artificial colloids. Albumin has a proven immunologic antiinflammatory effect and five times less volume than current artificial colloids. Unlike artificial colloids, it does not potentially lead to coagulopathic side effects. It has been proven to be safe from infectious and clinical standpoints. The volume of fluid that has to be carried is obviously much less (Fig. 4.20). Albumin costs approximately 30 times more than crystalloids and three times more than dextran or Hextend, but those comparisons were made against 5% human albumin. The cost of 100 mL of 25% albumin, compared with 500 mL of Hextend on a physiologic basis, is only approximately three times as much. During the Vietnam War, 25% albumin was first made available, and it seemed to have worked well. It was packaged in a green can that could be transported without damage, had a long shelf life, and was easy to use.

The commonly used synthetic colloids are plasma, albumin, dextran, gelatin, and starch-based colloids. Hetastarch solutions are produced from amylopectin obtained from sorghum, maize, or potatoes. Extensive randomized controlled trials have examined the safety and efficacy of 5% albumin, 6% hetastarch, and 6% dextran. However, no evidence has shown that one colloid is superior to another or that colloids are better or worse than crystalloids.⁶² Colloids such as hetastarch can have proinflammatory effects similar to those of crystalloids. In some cases, colloids will do more harm in large volumes than crystalloids, but not all colloids should be considered the same. It is well known that artificial colloids can perpetuate coagulopathy; dextran is used specifically to help prevent clotting after vascular surgery. The inflammatory system is tightly interwoven with the coagulation process. Hetastarch, particularly the high–molecular-weight preparations, is associated with alterations in coagulation, specifically resulting in changes in the viscoelastic measurements and fibrinolysis. Studies have questioned the safety of concentrated (10%) hetastarch solutions with a molecular weight of more than 200 and a molar substitution ratio of more than 0.5 in patients with severe sepsis, citing increased rates of death, acute kidney injury, and use of renal replacement therapy. To prolong intravascular expansion, a high degree of substitution on glucose molecules protects against hydrolysis by nonspecific amylases in the blood. However, this results in accumulation in reticuloendothelial tissues such as skin, liver, and kidneys. Because of the potential for accumulation in tissues, the recommended maximal daily dose of hetastarch is 33 to 55 mL/kg/day. Thus, it would be prudent to limit the use of Hextend to 1 L in trauma patients who are often harmed if they have coagulopathy from increased bleeding. Studies in trauma patients have shown an association between acute kidney injury and death after blunt trauma. Patients with severe sepsis assigned to fluid resuscitation with hydroxyethyl starch 130/0.4 had an increased risk of death at day 90 and were more likely to require renal replacement therapy compared with those receiving Ringer acetate. In animal models, albumin seems to be better for preventing inflammation, whereas hetastarch and dextran in high doses appear to cause inflammation and coagulopathy.

Colloids

Colloid solutions contain large, oncotically active molecules in a base solution of either 0.9% sodium chloride or a buffered, balanced electrolyte solution. Colloid molecules are too big to traverse gap junctions, so more of the water in these solutions tends to be retained within the plasma space. Theoretically, assuming a plasma volume of 3 liters in a 70-kg patient, 1 liter of an isooncotic colloid solution increases plasma volume by 1 liter, which is 4 to 5 times the plasma volume expansion achieved by the same volume of an isotonic sodium-based crystalloid. However, with critical illness, the vascular endothelium “leaks,” allowing colloid molecules to pass into the interstitium, where they exert an osmotic pressure effect. This probably explains the observation that when resuscitating critically ill patients with 0.9% sodium chloride only 1.3 times the volume is required (not four times as predicted) compared with 4% albumin to achieve the same hemodynamic end points.¹

Commonly used natural colloids include albumin and fresh-frozen plasma. Albumin is available as 4%, 5%, and 20% preparations. Both 4% and 5% solutions are approximately isooncotic with plasma; 20% albumin is hyperoncotic and therefore expands the plasma volume by about four times its volume. Commonly used artificial colloids include modified gelatins (e.g., Gelofusine) and hydroxyethyl starch compounds. Hydroxyethyl starch comprises a family of colloids that are categorized on the basis of their average size into high (>400 kD), medium (200 kD), and low (70 kD) molecular weight preparations.² Two commonly used hydroxyethyl starches are hetastarch (average molecular weight 480 kD) and pentastarch (average molecular weight 200 kD). By comparison, Gelofusine has an average molecular weight of about 35 kD. Most artificial colloids are isooncotic or slightly hyperoncotic. The plasma half-time of artificial colloids varies among preparations but is typically on the order of 4 to 6 hours. Thus, the colloid effect has largely dissipated by 24 hours. Large volumes of hydroxyethyl starch can cause impaired hemostasis, mainly because of reduced effectiveness of the factor VIII/von Willebrand factor complex (i.e., acquired von Willebrand disease).² High molecular weight compounds such as hetastarch appear to cause greater impairment of hemostasis than do medium-sized compounds such as pentastarch.^3,4 Impaired hemostasis may also occur with gelatin-based colloids.⁵ To avoid hemostatic problems, the dosages of these artificial colloids should be kept below 20 ml/kg. All artificial colloids can cause allergic (including anaphylactic) reactions.

15.2.1 Colloids

Colloid solutions include albumin solutions, available in two different preparations (the isooncotic 4% or 5% and the concentrated 20%–25% solutions) and synthetic solutions such as hydroxyethyl starches (HES) derived from maize or potato, gelatins derived from bovine collagen, and dextrans (polysaccharide) solutions.

Despite major modifications in their structure (molecular weight, degree of molar substitution that is the number of hydroxyethyl residues per glucose unit) over time, leading to faster clearance from the body,⁷ starches should certainly be avoided as fluid therapy. Indeed, there are at least 4 large randomized controlled trials that specifically confirmed an increased risk for acute kidney injury and need for renal replacement therapy with starch solutions when compared to balanced or unbalanced crystalloid solutions in either severe sepsis patients or general ICU patients. In a French study involving 129 patients with severe sepsis, the use of HES was found, in a multivariate analysis, to be an independent risk factor for renal failure when compared to a gelatin solution.⁸ In the German VISEP study, the use of HES was associated with a significantly higher rate of acute renal failure (34.9% vs 22.8%, P=.002), a significantly higher number of days on renal replacement therapies (18.3% vs 9.2%) and a trend toward increased mortality at day 90 compared to Ringer lactate solution in patients with severe sepsis.⁹ A dose effect was clearly demonstrated in this study, with higher dose of HES leading to a significant increase in mortality when compared to lower dose. Likewise, in the Scandinavian 6S study, the primary outcome of death or need for dialysis at day 90 was significantly higher in the group of septic patients receiving HES as compared to Ringer lactate solution.¹⁰ Similarly to the German study, the use of HES was also associated with an increased need for renal replacement therapy during the ICU stay. Finally, the Australian and New Zealand CHEST study confirmed, in a large general ICU population (about 7000 patients), a significantly increased need for renal replacement therapy in those patients who received HES as compared to those treated with normal saline solution (NS).¹¹ However, there was no difference in mortality in that particular study. Those data on deleterious side effects on kidney function were confirmed in various metaanalyses, and some authors proposed alternate volume replacement should be used in place of HES products.¹² With those data in mind, both the United States Food and Drug Administration and the European Medicines Agency issued, in 2013, warnings on the use of HES in septic patients, leaving the debate open for the perioperative setting. This led to various publications with some conflicting results. In the particular setting of kidney transplantation, Legendre et al. already raised concerns, back in the early 1990s, on the use of HES in the management of brain dead kidney donors, as it could induce osmotic nephrosis-like lesions.¹³ To confirm the impact of these histopathological findings, the same group prospectively studied 47 kidney recipients.¹⁴ They demonstrated a potential for nephrotoxicity with HES as there were significantly more dialysis requirements in those patients receiving a kidney from an HES-treated donor and the evolution of the creatinine value after transplantation was better in those patients receiving a kidney from a gelatin-treated donor. However, the long-term effect at 5 years was not different between groups in terms of serum creatinine value.¹⁵ Other authors did not confirm the relationship between osmosis-like lesions in the kidney recipient and the use of HES as other confounding factors could have influenced the results.¹⁶ Some authors also did not confirm any difference in outcome between patients receiving a kidney from a donor treated with HES or another type of solution such as crystalloid^17,18 while others even observed an improved outcome with 3rd-generation HES solutions on renal function of kidney recipients.¹⁹ However, those studies were retrospective and the latest one¹⁹ compared two different types of HES solutions without any other control solution. Finally, a large prospective observational study to assess the effect of HES, given to the kidney donor, on graft function in the recipient was conducted in Los Angeles.²⁰ Data were completed for 986 kidneys transplanted from 529 donors. In this study, HES use was associated with a 41% increased risk of delayed graft failure in the kidney transplant recipient. Those latest data certainly reinforce the statement by the EBPG, which already raised caution on the use of starches in kidney donor management and kidney recipient patients, 2 years before this latest study was published.^3,4

Other synthetic colloid solutions were certainly not studied to the extent of starch. Gelatins exist in three different solutions (succinilated, urea-linked, or oxypolygelatin), are eliminated by glomerular filtration, and are known to induce allergic reactions. In a well-known animal model of peritonitis (cecal ligature and punction or CLP) gelatins were shown to increase markers of renal failure (urea and creatinine) and kidney damage (NGAL), and induce structural changes as seen with starches.²¹ In humans, data on the use of gelatins in adequately powered randomized trials are scarce. Such a trial is planned in Germany (the GENIUS study for gelatin in ICU and sepsis; EudraCT 2015-000057-20), but this study will evaluate the safety of gelatins in critically ill septic patients, as compared to a crystalloid solution. In a study comparing the use of crystalloids and colloids, there was a trend toward increased kidney injury (as assessed by the RIFLE score R), a significant increase in the use of renal replacement therapy, and a prolonged ICU stay with gelatins,²² confirming to a large extent a previous trial.²³ Likewise, in a similar large before-and-after study, gelatins were shown to increase the need for renal replacement therapy as compared to crystalloids without achieving any better hemodynamic outcome in cardiac surgery patients.²⁴ Hence, the use of gelatins at the present time remains questionable.

Dextrans are polysaccharides that are mainly eliminated by glomerular filtration and the digestive tract system, while cellular uptake and storage has also been described. In a canine model of kidney transplantation, the use of dextran was associated with decreased renal function and histopathological features of osmotic nephrosis-like lesions.²⁵ Some case reports, reviewed elsewhere²⁶ also describe such lesions associated with kidney failure in humans. On top of those potential kidney side effects, dextrans may also induce severe allergic reactions and compromise hemostasis, which makes their use difficult to justify in the ICU setting or in kidney transplant patients.

From a theoretical point of view, albumin contained in natural colloid albumin solutions would be of great interest and show major advantages such as the major determinant of plasma colloid pressure, an ability to bind and transport various molecules including drugs, free radicals scavenging, and the protection of the endothelial surface layer. However, they are not equivalent in terms of sodium and chloride content; indeed the 4% or 5% solution displays sodium and chloride concentrations at 140–160mmol/L and 100–130mmol/L respectively while the 20% and 25% solutions contain 100–125mmol/L sodium and 20–25mmol/L chloride. Since this difference could account for different effects on the kidney (see the balanced–unbalanced solution debate later in the chapter), results from a particular study would probably not extend to other studies. In the critical care setting, albumin was not demonstrated to be superior in term of renal failure, need for renal replacement therapy, or mortality when compared to crystalloids.^27,28 In the Saline versus Fluid Evaluation or SAFE study,²⁷ 4% albumin was compared to 0.9% saline and did not result in any better outcome, including for the kidney. In the Albumin Italian Outcome Study or ALBIOS study,²⁸ a 20% albumin solution was prescribed to target a plasma albumin level of 30g/L. Again, the various outcomes including the incidence of acute kidney injury were not affected by the treatment. Various factors such as the high chloride content of the albumin solution in the SAFE study or the high oncotic pressure in the ALBIOS study could explain the nonpositive results since those two characteristics could lead to renal failure (see previous). Furthermore, the cost of albumin solutions (in the absence of clear benefit in the general ICU population) prevents its generalized use.

The situation is slightly different for kidney transplant patients. Indeed, in this context, observational studies showed albumin solution was associated with increased urine output, graft renal function and 1-year graft survival.^29–31 In a large series of 438 patients transplanted with kidneys from cadaveric donors, there was even a linear dose effect with albumin, the higher dose (1.2–1.6g/kg body weight) achieving the better outcomes in term of urine volume, serum creatinine and glomerular filtration rate, delayed graft failure, and graft function at 1 year.³² Because of the design of those studies, other factors (such as concomitant medications and the use of mannitol) may have affected the results. In a recent retrospective study on about 2000 kidney-transplanted patients over a 20-year period, the use of albumin was found to be an independent factor of protection for acute rejection and chronic graft dysfunction.³³ Beyond the fact this study was retrospective, the importance of the results are weakened by the very limited number of patients (91) treated with albumin. Controlled clinical data investigating the effect of albumin in kidney transplant patients were not available until recently. A first study on 44 patients assessed the outcome of patients receiving intraoperative 20% albumin and 0.9% saline versus 0.9% saline alone to target a CVP between 10 and 15mmHg.³⁴ There was no difference between groups at baseline. The volume infused during surgery was equivalent (roughly 4L). Neither urine output nor the various creatinine values at days 1, 3, and 5 were different between groups. Another study analyzed intraoperative 20% albumin versus 0.9% saline, to target a CVP of 12–15mmHg, on the outcome of kidney graft function in 80 patients.³⁵ Groups were well balanced apart from a trend towards an increased kidney donor age in the albumin group (49 vs 45; P=.051). None of the outcomes (hemodynamic parameters, creatinine serum value and urine output) was affected by the treatment group. Hence, the authors concluded albumin solution should be used rationally, considering its risk–benefit and cost–benefit ratios.

Synthetic Colloids

Synthetic colloid solutions include hydroxyethyl starches (HES) in which rapid degradation by α-amylase is prevented by hydroxyethylation of glucose subunits (Figure 36-5). The molar replacement ratio indicates the proportion of glucose molecules replaced with hydroxyethyl units (e.g., 0.4 = 40% replacement). The C2/C6 ratio indicates the number of hydroxyethyl units at C2 relative to C6. HES with higher molar replacement and C2/C6 ratios is retained longer due to slower metabolism (Table 36-4). HES is excreted by the kidney after degradation.

Differences between albumin and HES in efficacy and safety are controversial. HES products are available in iso-oncotic (6%) or hyperoncotic (10%) solutions. The 6% HES solutions with the average molecular weight of 600 to 670kDa (Hespan, Hextend) are most commonly used, but low molecular weight HES (130kDa, Voluven) has recently become available in the United States.⁴⁶ HES products are as effective as albumin as fluid replacements, but large doses of HES (particularly Hespan) can adversely affect coagulation (fibrin polymerization) and exacerbate renal dysfunction in sepsis.^47,48 Excess HES can falsely elevate turbidimetric fibrinogen measurements.⁴⁹

Synthetic Colloids

Synthetic colloid solutions include hydroxyethylstarches (HESs), in which rapid degradation by α-amylase is prevented by hydroxyethylation of glucose subunits (Fig. 44.3). The molar replacement ratio indicates the proportion of glucose molecules replaced with hydroxyethyl units (e.g., 0.4 = 40% replacement). The C2/C6 ratio indicates the number of hydroxyethyl units at the C2 relative to the C6 position. HES with higher molar replacement and C2/C6 ratios is retained longer owing to slower metabolism (Table 44.2). HES is excreted by the kidneys after degradation.

Differences between albumin and HES in efficacy and safety are controversial. HES products are available in iso-oncotic (6%) or hyper-oncotic (10%) solutions. The most commonly used are 6% HES solutions, with an average molecular weight of 600 to 670 kDa (Hespan, Hextend, B. Braun, Bethlehem, PA), but low-molecular-weight HES (130 kDa, Voluven, Fresenius Kabi, Halden, Norway) has recently become available in the United States.³⁰ HES products are as effective as albumin as fluid replacements, but large doses of HES (particularly, Hespan) can adversely affect coagulation (fibrin polymerization)³¹ and exacerbate renal dysfunction in sepsis.³² Excess HES can falsely elevate turbidometric fibrinogen measurements.³³Recently, the European Medical Association has recommended suspending marketing authorization for HES colloids in patients with sepsis or at risk for renal insufficiency. Colloid replacement therapy might still be indicated in early volume resuscitation after acute blood loss (e.g., trauma).

Blood Substitutes

Crystalloid or colloid solutions provide volume and oncotic pressure but do not provide the oxygen transport of RBCs or other acellular fluids commonly referred to as ‘blood substitutes.’ It is unclear as to whether initial resuscitation of critically ill patients should begin with crystalloid or colloid solutions (e.g., gelatin and hydroxyethyl starch). Nonetheless, data suggest that use of colloid solutions may be associated with impaired renal function. Bayer et al. found that renal injury and dialysis were significantly lower in their ICU when 6% hydroxyethyl starch and 4% gelatin were not utilized during a colloid-free time period as part of a sequential study.

Second-generation HBOCs have been produced in an effort to reduce nitric oxide binding and resultant vasoconstriction. One HBOC, Hemopure (OPK Biotech, Cambridge, MA), is approved for treating perioperative anemia in adult surgical patients in South Africa and in veterinary medicine in the United States as Oxyglobin. Questions have been raised about safety, particularly because of the similarity in serious adverse event profiles among HBOCs. These events (e.g., hypertension, stroke, and myocardial infarction) raise concerns about the relative risk/benefit of HBOC, especially as allogeneic blood is currently in adequate supply and relatively safe. Nonetheless, there are certain settings, such as religious objection to transfusion or alloimmunized patients, for whom availability via compassionate use could be justified.