Hemoglobin-based Oxygen Carriers
2009; Lippincott Williams & Wilkins; Volume: 111; Issue: 5 Linguagem: Inglês
10.1097/aln.0b013e3181ba3c2c
ISSN1528-1175
AutoresToby Silverman, Richard B. Weiskopf,
Tópico(s)Hemoglobinopathies and Related Disorders
ResumoSEMISYNTHETIC or synthetic preparations of hemoglobin, now termed hemoglobin-based oxygen carriers (HBOCs), have been in development as an alternative to erythrocytes for several decades. Nonclinical and clinical studies of HBOCs have raised questions about their safety. Similarity of the serious adverse event profiles among these products has raised questions regarding the possibility of common underlying mechanism(s) of toxicity despite differences among these molecular preparations. The observed serious adverse events have presented an obstacle to development because they raise concerns about the relative benefit/risk of these biologics.A conference sponsored by the Food and Drug Administration (FDA) and National Institutes of Health (NIH) was held April 29–30, 2008 to (1) review the existing publicly available information about the characteristics and clinical profiles of the products that are or were in development; (2) discuss the potential mechanisms of toxicity both for the whole organism or by specific organ system(s); and (3) consider the feasibility of, obstacles to, and ethical issues related to future clinical trials of HBOCs.This report summarizes the scientific presentations and opinions of the speakers at the conference. A transcript of the meeting is available.‡§Appendix 1contains a list of the members of the Planning Committee. Appendix 2contains a list of the Speakers. Appendix 3summarizes information about conflict of interest available to the FDA at the time of the workshop.The workshop was divided into four sessions. Session I provided a forum for a scientific overview of oxygen physiology, general biochemical and physiologic characteristics of HBOCs, nitric oxide physiology, and characteristics of nonclinical studies of HBOCs. Presentations in Session II provided an overview of current information about the safety and efficacy of HBOCs that have been evaluated in clinical trials, the regulatory and ethical framework for clinical trials of HBOCs, and unresolved issues related to publicly available information from commercially sponsored studies of HBOCs. FDA and industry representatives discussed data from clinical trials of HBOCs that are publicly available from peer-reviewed literature, press releases, or presentations at public meetings. Session III was divided into two parts. In the first part, functional aspects of HBOCs as a class of therapeutic agents were discussed, with particular emphasis on common clinical findings, need for additional research into mechanisms of toxicity, clinical trial design issues, and clinical settings where benefit might outweigh safety issues. In the second portion, organ-specific aspects of HBOC safety were addressed. Panelists discussed possible mechanisms for observed safety issues and the pros and cons of sensitive and specific biomarkers for evaluation of organ-specific toxicity. Session IV was devoted to seeking a way forward in development of HBOCs. Presentations were made on new modifications of HBOCs, effects of HBOCs on the microvasculature, molecular biologic reengineering of HBOCs to achieve desired physiologic effects, development of preclinical models, and focused clinical trial designs were discussed. Panelists then discussed ethics, the feasibility of designing new molecules, various clinical trial designs, and the need for HBOCs in civilian and military settings and in ex-U.S. sites.Moderator: Joseph Fratantoni, M.D., Vice President Medical and Clinical Development, Maxcyte, Inc., Gaithersburg, Maryland.Speakers noted that regulation of oxygen transport/oxygen physiology, including physiologically appropriate oxygen affinity and cooperativity, and hypoxic vasodilatation appear to be fundamental to appreciating the complexities of HBOCs and to understanding how a patient's organ and whole body physiology adjust to hypoxia. Hypotheses with varying degrees of experimental validation have been published to explain the phenomenon of hypoxic vasodilatation including (1) release of adenosine triphosphate (ATP) and binding of ATP to an endothelial cell receptor, resulting in increased circulating nitric oxide and vasodilatation, (2) allosteric release of nitric oxide by sulfhydryl-linked nitric oxide that occurs as the red blood cell (RBC) deoxygenates,1(3) partially oxygen-saturated hemoglobin acting as a nitrite reductase,2and (4) metabolic autoregulation of blood flow.Nitric oxide, a major systemic vasodilator, is a short-lived molecule that is produced by, among other cell types, endothelial cells, and it is released and diffuses both luminally and abluminally. The importance of nitric oxide in modulating downstream effects of acellular hemoglobin was discussed. Some investigators have challenged long-held assumptions that nitric oxide scavenging by HBOC results in impaired blood flow and that oxygen binding of HBOC should match that of RBC.3,4They hypothesize that the presence of low-oxygen-affinity hemoglobin intraluminally disrupts the oxygen gradient from red blood cell to vessel wall and leads to vasoconstriction. Other researchers have emphasized the extent to which the heme moiety in acellular hemoglobin is subject to oxidation. Spontaneous oxidation of hemoglobin within the protective environment of the red blood cell is effectively reversed by naturally occurring intracellular reducing agents. Outside of the red blood cell, however, hemoglobin undergoes spontaneous oxidation and is subject to the effects of oxidant agents such as hydrogen peroxide and the effects of interaction with nitric oxide. Chemical modifications to the hemoglobin can lead to unexpected and significant changes in susceptibility to oxidation and thereby to destabilization of the product. Protective antioxidant strategies have recently been a focus of the HBOC community.Conventional single and repeated-dose safety and toxicity testing of the various HBOCs in preclinical animal models have not been able to predict the adverse outcomes frequently observed in clinical trials. Evaluation of HBOCs to date has been performed in two environments: (1) conventional safety testing in healthy normal animals under good laboratory practice-compliant conditions; (2) "academic" or efficacy testing settings that are performed under non-good laboratory practice and non-International Conference on Harmonzation (ICH) standards.5–8The effects observed in conventional studies conducted in healthy, normal (inbred) young animals do not mimic those of clinical tests of HBOCs, many of which have been conducted in populations with elderly, diabetic, atherosclerotic, hypertensive patients with significant comorbid conditions who are undergoing elective surgical procedures. A prominent feature of these populations and conditions is the presence of endothelial dysfunction, including impaired nitric oxide response and exaggerated endothelin response.9–11Introduction of an HBOC into the circulation likely worsens preexisting endothelial dysfunction, probably by exacerbation of the impaired nitric oxide availability. Limitations of conventional animal toxicity studies include feasibility and applicability of dose escalation and repeat dosing, failure to detect infrequent events, and failure to establish possible synergy of HBOCs with preexisting disease.Moderator: Barbara Alving, M.D., M.A.C.P., National Center for Research Resources, National Institutes of Health, Bethesda, Maryland.Some unresolved issues that have hampered the development of HBOCs have included difficulty in defining clinical benefit and defining and assessing clinically meaningful, readily measurable efficacy endpoints, understanding certain safety parameters such that a safe dose of products can be administered, and defining an acceptable benefit:risk profile for each clinical indication in studies where subjects provide informed consent and in studies where informed consent cannot be obtained. Table 1contains publicly available information about the physicochemical characteristics of eight commercially sponsored HBOCs and a listing of clinical situations in which these HBOCs were evaluated.Publicly available information (peer-reviewed journals, press releases, meeting presentations, etc .) on the safety of HBOC products studied in a variety of clinical settings was presented. Of the eight commercial products, data were available in the public domain for six. The safety data (table 2) formed the basis for subsequent discussions during the workshop. Important caveats regarding these data are described in the footnote.To place existing clinical trial data into context for further discussion, the concept of risks and benefits was introduced in the framework of general ethical considerations and two FDA regulations (Code of Federal Regulations, CFR) governing investigational use of an unapproved drug product (21 CFR 312) and investigational use of an unapproved drug product in situations when obtaining informed consent is not possible (21 CFR 50.24). The latter recognizes the need for such research when it is necessary to obtain the required data. Under one ethical framework devised by Drs. Emanuel, Wendler, and Grady, there are a number of hierarchically important and universal requirements that must be fulfilled for clinical research to be ethically acceptable.12These include the need first for the research to have social (scientific and medical) value and scientific validity and to be performed under conditions where subject selection is fair. Other considerations include the need for a favorable benefit:risk ratio, informed consent, respect for potential and enrolled participants, and independent review of the research. The requirement for prospect of direct benefit to the subjects enrolled in a trial with exception from informed consent is explicitly stated under 21 CFR 50.24; however, all elements of the assessment (conceptual support, preclinical evidence, and clinical studies in other settings and other populations) generally apply to all clinical research. The considerations stated explicitly in 21 CFR 50.24 reflect the complex assessment that must be made, an assessment that requires experienced judgment in conjunction with rigorous scientific evidence about the prospect for benefit.Presentations of clinical data were made by representatives of five commercial sponsors; Sangart, Northfield, Prolong Pharmaceuticals (reporting data from Enzon), Biopure, and Apex. In addition, Tim Estep (President, Chart Biotech Consulting, LLC, Erie, CO) presented publicly available data from Baxter/Somatogen, though not as a representative of the companies. Most presenters emphasized that each product had unique characteristics and properties and should be considered individually rather than collectively.Sangart13–16is developing a polyethylene glycol (PEG)-conjugated hemoglobin, MP4 (MalPEG-hemoglobin). This larger molecule is characterized by (1) high oxygen affinity that limits oxygen release in arterioles and decreases rates of oxidation; (2) increased oncotic pressure; and (3) a formulation of lower hemoglobin concentration. The development of this molecule is based on the theory of autoregulation, wherein high oxygen unloading in arterioles leads to vasoconstriction. Based on preclinical data, Sangart hypothesizes that MP4 might preserve functional capillary density in anemic and shock states and reduce the incidence of hypertension and other adverse events, arguing that nitric oxide scavenging does not explain all HBOC-related adverse events.Six clinical trials in elective surgery, chronic critical limb ischemia, and prevention of hypotension have been and are being performed. Safety results of a phase 2 trial in orthopedic surgery14showed an imbalance in cardiac and vascular events, including bradycardia and blood pressure elevation; gastrointestinal events including nausea; cardiac rhythm, including bradyarrhythmias; and pancreatic enzymatic activity (lipase and amylase) with greater frequency after administration of product than placebo (Ringer's acetate). Data were presented suggesting that the small increases in mean arterial pressure may be attributable to the hyperoncotic volume-expansion properties of the product rather than nitric oxide scavenging.Northfield manufactures Poly-SFH-P, a glutaraldehyde polymerized stroma-free human hemoglobin product (PolyHeme®) and intends to seek an indication for when blood is not available, as might occur at the scene of injury, during transport to definitive care, or in the hospital, because of religious objection, blood incompatibility, or shortage. Northfield has conducted six clinical trials in 1,133 subjects, 674 of whom have received the product. The results of a phase 2 trial in 171 subjects retrospectively compared against historical control data from Jehovah's Witness patients undergoing elective surgery and who had sustained significant blood loss leading to low or extremely low red blood cell hemoglobin concentrations,17,18and the randomized pivotal phase 3 trauma trial, conducted under the provisions of 21 CFR 50.24, were presented. The phase 3 study was designed to assess through extrapolation the benefit of PolyHeme® to the intended population for whom blood would not be available for prolonged periods of time; it was not designed to demonstrate that PolyHeme® could be used in place of blood.19The study had a dual superiority/noninferiority design.20Results of mortality assessments in the Intent-to-Treat (ITT; as randomized), As-Treated (AT; as treatment received), and Per Protocol (PP; without major protocol violations) populations were presented. In the prespecified ITT population, the mortality for the test group was 13% (47 of 350) and 10% (35 of 364) for the control group, failing the prespecified 7% noninferiority boundary. Northfield expressed the opinion, however, that the appropriate analysis was the PP analysis because this analysis excludes both subjects who received a treatment other than that assigned and those who were enrolled in violation of entry criteria. Subjects in the PolyHeme® group were reported to have had on average lower blood pressure before randomization and more severe neurologic findings, more severe coagulopathy at randomization, and more severe injuries than subjects in the control arm.Serious adverse events in this trial were reported for 40% (141 of 349) test subjects and 35% (126 of 365) control subjects. The most common serious adverse events that occurred in excess in test subjects included pneumonia, multiple organ failure, hemorrhagic shock, respiratory failure, hypercoagulable state, coagulopathy, and myocardial infarction. Post hoc adjudication of cardiac events by a blinded subcommittee of the data monitoring committee suggested that myocardial infarction in victims of trauma is much more common overall than previously believed and did not differ between treatment groups.Enzon developed a PEG-HBOC (polyethylene glycol-conjugated hemoglobin) bovine product, with the goal of development to increase tumor oxygenation to enhance radiosensitivity of susceptible tumors. The most common side effects were reported to be mild hypertension in 24%, dysphagia in 24%, nausea in 24%, and vomiting in 12%. Esophageal spasm was effectively prevented/well-managed with an atropine-like antispasmodic agent Levesinex. Clinical development of this PEG-hemoglobin was halted in the late 1990s. Subsequently, Prolong has been developing PEGylated drugs hypothesizing that attaching one or more PEG (polyethylene glycol moieties) to a protein molecule could improve therapeutic and functional characteristics of biopharmaceuticals, including acellular bovine hemoglobin.The Naval Medical Research Center (NMRC) and FDA have presented results of clinical trials of Biopure's candidate acellular hemoglobin product, HBOC-201, a glutaraldehyde crosslinked bovine hemoglobin, at the December 2006 Blood Products Advisory Committee meeting (Bethesda, MD). At the April 2008 workshop, Biopure presented a reanalysis of safety signals arising out of two pivotal phase three studies in elective surgery noting that data from the larger of the two studies, HEM-0115,21are representative of all safety signals seen in the pooled safety database arising out of 21 clinical trials. Biopure acknowledged that there was an imbalance against HBOC-201 for adverse events and serious adverse events and presented an analysis of the clinical and preclinical data with respect to mechanisms. Biopure concluded that the safety signals did not reflect product toxicity per se and that the serious adverse events were the result of a combination of patient comorbidity, clinical management differences, and underdosing with HBOC-201 with ongoing anemia and ischemia on the one hand and overdosing with HBOC-201with volume overload and heart failure on the other.21In the HEM-0115 trial, 59% of patients allocated to the HBOC arm avoided red cell transfusion; however, their total hemoglobin concentration was less than that of the group given red cells, which might have affected both efficacy and safety results.Biopure maintained that vasoconstriction was not the primary factor in the emergence of serious adverse events. They presented preclinical animal studies evaluating vital organ and skeletal muscle blood flow and tissue oxygen tension in animals undergoing up to 50% exchange transfusion to show that there was increased vascular resistance in skeletal muscle but no evidence of vasoconstriction in heart, brain, and kidney. Biopure also presented studies in pigs and a phase 2 study in humans undergoing percutaneous coronary intervention to show that HBOC-201 affects mean arterial pressure but not coronary function. Coronary blood flow, left ventricular end-diastolic pressure, electrocardiographic findings, and cardiac output were not affected by use of HBOC-201, and left ventricular function, as assessed by pressure-volume loops, was maintained with infusion of oxygenated HBOC-201 directly into the coronary artery.Biopure stated, although no data were presented, that there have been no issues regarding management of hypertension or about myocardial infarctions and other adverse events/serious adverse events in over 480 patients treated in South Africa.Biopure concluded that given the adverse events observed in clinical trials of HBOC-201 compared with red blood cells in elective surgeries, further clinical trials against red blood cells are not warranted. They further concluded that HBOC-201 should be evaluated in trials where blood is not immediately accessible or an option.Baxter manufactured a diaspirin αα-crosslinked human tetrameric hemoglobin product, DCLHb, and later recombinant tetrameric hemoglobin products based in part or whole on molecules previously developed by Somatogen. Baxter presented a history of the development of DCLHb focusing on properties and factors potentially affecting the outcome of human phase 3 clinical trials.During preclinical testing of DCLHb, Baxter identified several safety concerns, including the development of cardiac lesions in some animal species, transient liver pathology, gastrointestinal effects, jaundice, and vasoactivity. In primates, the myocardial lesions were described as focal or multifocal, minimal to moderate myocardial degeneration characterized by cytoplasmic swelling and vacuolization of myofibers occurring primarily in the left ventricle and septum. The lesions, which resolved with time, exhibited a threshold and a plateau maximum with dose. Morphometry studies in rhesus monkeys suggested that an average of 1.3% of the myocardium was susceptible to damage. The lesions were not associated with changes in cardiac enzymes, and there were no functional deficits detected in pigs with the lesion. Many tested interventions, including pharmacologic interventions, had no effect on the development of the cardiac lesions. However, based on work with the recombinant molecules, polymerization of the hemoglobin product and reduction of the rate of nitric oxide scavenging mitigated lesion development. The relationship of these cardiac lesions to human pathology is unclear.Vasoactivity was overtly manifested as an increase in systemic blood pressure that was highly correlated with hemoglobin extravasation and nitric oxide scavenging and tended to manifest and maximize rapidly at low doses.Baxter conducted two phase 3 trials evaluating the use of DCLHb in the treatment of traumatic hemorrhagic shock.22–24One trial was conducted in the prehospital setting with use of product in lieu of blood. The second trial was an in-hospital evaluation of DCLHb administered in addition to standard of care including red blood cells. The in-hospital phase 3 trial was halted because of excess mortality in the test arm. The prehospital study was halted early with a higher, but not statistically significant, point estimate for mortality for DCLHb. Pancreatitis was noted as one important safety finding for DCLHb. Increased blood pressure and uncontrolled bleeding did not systematically occur among trauma subjects administered DCLHb in the in-hospital trial.Unpublished data from subsequent preclinical studies in traumatic hemorrhagic shock showed that administration of relatively small volumes of DCLHb after resuscitation with large volumes of crystalloid may have an adverse impact.rHb1.1, first developed by Somatogen and later acquired by Baxter, had clinical properties and toxicities similar to DCLHb. To reduce the incidence of cardiac lesions, vasoactivity, and gastrointenstinal effects, a second generation recombinant product, rHb2.0, was developed with reduced interaction with nitric oxide. Subtherapeutic doses administered to subjects enrolled in early phase 1 studies resulted in complement activation. Baxter abandoned development of HBOCs.Estep concluded that HBOC solutions have multiple properties that are important to tissue perfusion and oxygenation, including oxygen transport, vasoactivity, oncotic pressure, viscosity, and fluid volume, and that these properties might have very different dose response characteristics.Apex manufactures a pyridoxylated chemically modified human hemoglobin that is conjugated with polyoxyethylene as a nitric oxide scavenger intended for the treatment of distributive shock. Nitric oxide is hypothesized to be the final common mediator in nitric oxide induced shock independent of etiology and redundant cytokine pathways, with direct toxic and pathophysiologic actions.Results of a small study of continuous infusion of pyridoxylated polyoxyethylene-conjugated hemoglobin in the treatment of subjects with volume refractory, pressor-dependent systemic inflammatory response syndrome were presented. Mean arterial pressure increased, and heart rate decreased with administration of product, and median times to first withdrawal of conventional vasopressors among survivors was shorter in pyridoxylated polyoxyethylene-conjugated hemoglobin-treated subjects than in control subjects. Mortality rates at days 10 and 28 were not different among treatment groups. Although perhaps not statistically different for this small study, the absolute difference in mortality on day 10 was 18% (48% for control vs . 30% for pyridoxylated polyoxyethylene-conjugated hemoglobin); however, on day 28 the mortality was similar for the two groups (57.6% for pyridoxylated polyoxyethylene-conjugated hemoglobin and 58.6% for control). There was a statistically insignificant increase in days alive and free from cardiovascular dysfunction and mechanical ventilation favoring pyridoxylated polyoxyethylene-conjugated hemoglobin, without an increase in the need for certain medical interventions for liver, kidney, coagulation, or central nervous system complications. Blinded review and adjudication using prespecified criteria did not confirm the excess of cardiac events identified by investigators who were not blinded to treatment allocation. Apex concluded that a blinded, adjudicated review of all cardiac events will be performed using prospective definitions in future studies, and that the patient population should be at high risk of dying due to shock unresponsive to standard of care such as dopamine or norepinephrine.During the question and answer session, Tom Fleming, Ph.D. (Department of Biostatistics, University of Washington), at the request of the session moderator, provided a biostatistical critique of the commercial presentations. He noted the following beliefs. (1) Absence of an adverse safety signal in a single clinical trial does not equate to an absence of harm; there must be sufficient data to rule out a major morbidity occurring at a clinically unacceptable rate, or excess mortality. (2) Any noninferiority margin must be justified; the effect of the active comparator on outcome must be known so that the level of benefit that can be lost (noninferiority margin) without also losing clinical significance of the outcome with the test product can be determined. (3) Whether to use the ITT population or the PP population for analysis is an area of debate for noninferiority trials because a difference in adherence to the protocol between the active comparator and the experimental treatment can affect observation of the true difference. (4) Any post hoc analyses, although interesting and hypothesis-generating, must be viewed with great caution; an ITT analysis is the most reliable indicator of causality even though ITT analyses do not take into account all the clinical exigencies of each individual patient.Session III was divided into two sessions. The first session was devoted to discussion of clinical findings and mechanisms of toxicity. The second session was devoted to discussion of organ-specific aspects of safety findings. Panel members were permitted to present brief position papers. The sessions concluded with question and answer periods.Moderator: Harvey Klein, M.D., Chief, Transfusion Medicine, National Institutes of Health, Bethesda, Maryland.Harvey Klein, M.D., summarized four overarching questions for the panel. (1) Can information about the safety and efficacy obtained from clinical trials in one setting be applied to another? (2) Given what is known about the biochemistry and pharmacology of the current and previous HBOCs, can safety information obtained from the study of one HBOC be used to inform safety and risk assessments for a different HBOC? (3) Are there toxicities or harmful interactions between these molecules and a patient's underlying disease(s) that are common to all of these molecules? (4) Are there lessons from what was heard in Sesion II for designing future trials?Differences in mechanism of death and prognosis in blunt trauma and penetrating trauma were discussed by Demetrios Demetriades, M.D., Ph.D. (Professor of Surgery, Department of Surgery, Division of Trauma and Critical Care, University of Southern California, Los Angeles, California). He noted that blunt trauma rarely results in hemorrhagic death within 1 h. Excluding head injury, the mortality in blunt trauma for patients with systolic blood pressure less than 90 mmHg is approximately 20%, whereas the mortality rate in patients with penetrating trauma and systolic blood pressure less than 90 mmHg is approximately 33%. Segregating blunt and penetrating trauma into two separate clinical studies because of the difference in prognosis, enrolling victims of penetrating trauma (excluding head injured patients) with a systolic blood pressure less than 80 mmHg, and using HBOCs on a "compassionate" basis were discussed.Daniel Freilich, M.D., C.D.R., M.C., U.S.N. (Naval Medical Research Center, Combat Casualty Directorate, Silver Spring, Maryland) discussed strategic considerations in the design of clinical trials to evaluate the current generation of HBOCs for trauma. He noted that preclinical data should steer HBOC clinical trial design for trauma and summarized 16 studies performed in preclinical models of controlled hemorrhage, uncontrolled hemorrhage, traumatic uncontrolled hemorrhage with or without concomitant traumatic brain injury and concluded that there was a significant potential benefit of HBOC (specifically HBOC-201 from Biopure) to reduce mortality in severe hemorrhage.‖Optimization of study design and practice guidelines, inclusion of a target population with severe hemorrhagic shock and high mortality, and risk minimization strategies should enable clinical evaluation of the current generation of HBOCs with reasonable risk.John Holcomb, M.D. (Commander, US Army Institute of Surgical Research, Brooke Army Medical Center, San Antonio, Texas) identified trauma-induced coagulopathy as an independent predictor of increased mortality and discussed an algorithm using physiologic parameters of heart rate, systolic blood pressure, pH, and hematocrit that has relatively high predictive ability for massive transfusion among patients with all four predictors. He noted that massively transfused patients were at particular risk of dying necessitating rapid diagnosis and intervention. Dr. Holcomb presented data suggesting that increasing the plasma and platelet to red blood cell ratio decreases early death due to hemorrhage and improves overall survival from 41% to 74% at day 30.25These results were seen despite equal numbers of units (21) of RBC transfusions. Dr. Holcomb noted that, although significant attention has been paid to rapid restoration of oxygen delivery with transfused hemoglobin, one must not neglect the plasma and platelet transfusions as well.Steven Cohn, M.D. (University of Texas Health Science Center, Houston, Texas) discussed the potential use of HBOCs when blood is not available in both prehospital and hospital settings and the need to consider the risk of mortality and morbidity of not using an HBOC in these circumstances.Ed Norris, M.D., M.B.A., F.A.H.A. (Associate Professor, Department of Anes
Referência(s)