Anne Hamilton Dougherty, University of Texas Medical School, Houston
rod since its inception, the PolyHeme trial is a randomized
evaluation of an investigational oxygen-carrying blood substitute in severely
injured trauma victims utilizing waiver of consent. The Boston University Institutional Review
Board (IRB) concluded that it did not meet the conditions set forth in 21 CFR
50.24 for exception from informed consent and shared its determination with the
Food and Drug Administration (FDA), Northfield Laboratories and with other
IRB’s reviewing the trial. More
recently, Drs. Kipnis, King and Nelson have
posted an open letter to participating IRB’s imploring them to impose
limits on the trial, arguing that waiver of informed consent is unjustified (Kipnis,
King and Nelson 2006).
32 IRB’s have reviewed the protocol and, after thoughtful consideration of its
risks and benefits in the context of regulatory requirements, voted
approval. The Committee for the
Protection of Human Subjects of the
three-and-one-half hours of full committee debate, along with countless hours
of subcommittee research, review and negotiation. Community consultation in the
included press releases to nine daily newspapers and 41 non-daily newspapers,
16 radio stations, internet postings, 12 advertised community group hearings,
and two radio station call-in shows.
How can caring
and informed individuals disagree? The
primary objection to the trial lies in continued randomization upon hospital arrival
where banked blood, the standard of care, is available. Exception from informed consent requires that
“human subjects are in a life-threatening situation, available treatments are
unproven or unsatisfactory, and obtaining informed consent is not feasible . .
.” (21 CFR 50.24). Critics argue that, despite acknowledged limitations of
banked blood, its general acceptance in resuscitation is prima facie evidence of its “satisfactory” nature. Once blood is available, they profess,
randomization should be reserved for consenting subjects.
efficacy and safety of transfused blood “proven”? Despite its widespread use and acceptance,
the performance of banked blood has never been subjected to the level of
scrutiny imposed on investigational new drugs.
Transfusions fail to improve oxygen consumption in critically ill
patients (Fernades et al. 2001). Attendant
risks of transfusion reaction and transmission of infection are universally
recognized. Few accept transfusion
absent life-threatening anemia.
transfusion of banked blood “satisfactory”?
Clearly, it is an effective means of immediate resuscitation in
hemorrhagic shock, but at what price?
Clinical trials have demonstrated that transfusion increases the
incidence of multiorgan failure (MOF) and increases mortality (Kao 2000;
Johnson et al. 2001). In particular,
transfusion of six or more units within the first 12 hours is the primary risk
factor for MOF, independent of injury severity (Moore, Moore and Sauaia
1997). MOF is the leading cause of
post-injury death, conferring a mortality of 35% and prolonging ICU stays an
average of one month with mechanical ventilation and dialysis. The adverse effect on outcomes is not limited
to trauma victims, but is particularly pertinent to this population because of
the pathophysiology of transfusion-related injury (Corwin et al. 2004; Herbert
and Fergusson 2004; Napolitano 2004; Rao et al 2004). Many thoughtful clinicians feel treatment
that increases mortality and organ failure is unsatisfactory.
blood be bad? It’s a universal natural
product that serves us well. Banked
blood, however, is artificially preserved and gradually degrades with its
42-day storage, releasing from red cell membranes lipid mediators and
cytokines. These factors combine with
traumatized tissue in a two-step priming of neutrophils and oxygen radical
production, each of which is highly correlated with development of MOF (Moore,
Moore and Sauaia 1997; Johnson et al. 2001).
Transfused units are necessarily the oldest in the blood bank with the
highest concentration of toxins.
Furthermore, trauma patients requiring massive transfusions first
receive unmatched blood. Type-specific
blood requires up to 20 additional minutes to deliver and fully cross-matched
blood can require 45 minutes. The risk
of immunologic complications and medical error increases under the pressure of
emergency delivery. Blood’s failure to
improve oxygen consumption in critically ill patients apparently stems from
changes in red cell deformability with storage.
Such rheologic changes can produce red cell entrapment, microvascular
obstruction and tissue ischemia (Fernandes et al. 2001; Berezina et al.
PolyHeme has a long shelf life and is compatible with all blood
types. The risk of viral and bacterial
transmission is extremely low. Extensive
pre-clinical and earlier clinical studies have shown very few adverse
effects. Deleterious vasoconstrictive
effects that have plagued earlier blood substitutes are noticeably absent with
aborted trial of PolyHeme as adjunct to the practice of acute
normovolemic hemodilution (ANH) in elective abdominal aortic aneurysm repair
has sparked renewed concern. Ten
subjects in the PolyHeme treatment arm experienced myocardial
infarction with two fatalities, compared to none in control subjects. The two treatment arms, however, differed
more substantially than in the administration of PolyHeme. In this population with established
cardiovascular disease, ANH is used to minimize intraoperative blood loss. Phlebotomy (up to 2-3 units, 1.5 liters on
average) targets a hemoglobin 9 g/dl and blood volume is replaced with an
equivalent amount of preoperative colloid; patient blood is stored for
perioperative return. The PolyHeme
group in the ANH trial was bled up to six units (2.8 liters on average), and
received two units of colloid and four of PolyHeme in an attempt to
further enhance the benefit of ANH. Not
only was the PolyHeme group
bled almost twice as aggressively as the control group, it also received larger
volumes of autologous and allogeneic blood in the perioperative return
phase. Cardiac events were concentrated
in low enrolling centers where fluids were administered more liberally. Massive phlebotomy and fluid shifts in this
vulnerable population, rather than the test article, were felt to be responsible
for excessive cardiac morbidity and mortality.
The trial was curtailed for futility (Northfield Laboratories 2006).
Excessive cardiac adverse events have not been observed in any other PolyHeme
King and Nelson (2006) criticize PolyHeme for its absence of
clotting factors. Packed red cells also
offer far fewer clotting factors than whole blood. The PolyHeme trial provides for both
treatment groups to receive supplemental fresh frozen plasma, replacing
clotting factors in both groups. Since
transfusion-related toxins are associated with red cell membranes, plasma administration
does not appear to participate in the development of MOF.
twelve-hour clinical phase of the study has also been the subject of suspicion
in the press. Some have argued that it
is designed to simulate evacuation times under battlefield conditions rather
than those in urban areas, speculating that the product is destined strictly
for military use; its development in
civilian subjects thus would violate the justice principle. In fact, the interval mirrors pre-clinical
data on the temporal relationship between injury and neutrophil priming as well
as the observation that transfusion of six units or more over the first twelve
hours increases risk of MOF, factors as relevant to
have argued that the trial should reflect an intended use of PolyHeme
as a bridge to blood only. In fact, the
trial is designed as a superiority trial, testing the hypothesis that the use
of PolyHeme as the initial resuscitative fluid in severely injured
patients with hemorrhagic shock reduces mortality, compared to the standard of
care (crystalloid in the field, then blood when available). That hypothesis cannot be tested if all
groups receive blood at the first opportunity.
I support the
action of our IRB in approving the PolyHeme trial. In the wake of increasing public criticism we
have revisited and affirmed our initial decision. Although some critics feel there is an
ethical flaw in trial design, those arguments themselves are flawed and
caring individuals have a right to disagree.
Whereas the definition of “unsatisfactory treatment” is imprecise and
open to debate, few would argue that allogenic blood transfusion has very
significant limitations, its risk being far more complex and ominous than
simple immunologic incompatibility or infection transmission.
Our system of
human subjects protections depends on the authority of the local IRB system,
charged with weighing the risks and benefits of research in the context of
community standards, needs and resources.
Nowhere is the value of the local review system more evident that in
studies requiring waiver of consent which
are particularly difficult to orchestrate and demand sincere commitment from
the investigative team, the IRB and the community alike. Protocols that may be unsuitable for one
medical system may be entirely appropriate for others with different resources. A conscientious local IRB, knowledgeable and
passionate about the needs and concerns of the community it serves, is in the
best position to review the risk and benefit of proposals in this setting and
to interact with the potential subject population in community consultation.
L., S. B. Zaets, C. Morgan, C.R. Spillert, M. Kamiyama, Z. Spolarics, E.A.
Deitch, and G.W. Machiedo. 2002.
Influence of storage on red blood cell rheologic properties. Journal
of Surgery Research 102: 6-12.
A. Gettinger, R.G. Pearl, M.P. Fink, M.M. Levy, E. Abraham, N.R. MacIntyre,
M.M. Shabot, M.S. Duh, and M.J. Shapiro.
2004. The CRIT study: Anemia and blood transfusion in the
critically ill-current clinical practice in the
Care Medicine 32:39-52.
J., Jr., N. Akamine, F. V. De Marco, J.A. De Souza, S. Lagudis, and E.
Knobel. 2001. Red blood cell transfusion does not increase
oxygen consumption in critically ill septic patients. Critical
Care 5: 362-367.
and D.A. Fergusson. 2004. Do transfusions get to the heart of the
matter? Journal of the American Medical Association 292(13): 1610-1612.
E.E. Moore, P.J. Offner, D.A. Partrick, D.Y. Tamura, G. Zallen, and C.C.
Silliman. 2001. Resuscitation with a blood substitute
abrogates pathologic postinjury neutrophil cytoxic function. Journal
of Trauma 50:449-56.
2000. Mechanisms and new approaches for
the allogeneic blood transfusion-induced immunomodulatory effects. Transfusion
Medicine Review 14:12-22.
N.M.P. King, and R. M. Nelson.
2006. An open letter to IRBs
considering Northfield Laboratories’ PolyHeme Trial. American
Journal of Bioethics 6(3): 18-21.
E.E. Moore, and A. Sauaia. 1997. Blood transfusion: An independent risk factor for postinjury
multiple organ failure. Archives of Surgery 132: 620-625.
L.M. 2004. Current status of blood component therapy in
surgical critical care. Current Opinion in Critical Care 10:
Laboratories, Inc. 2006. Press release:
Laboratories releases summary observations from its elective surgery trial.
March 20, 2006.
J.G. Jollis, R.A. Harrington, C.B. Granger, L.K. Newby, P.W. Armstrong, D.J.
Moliterno, L. Lindblad, K. Pieper, E.J. Topol, J.S. Stamler, and R.M.
Califf. 2004. Relationship of blood transfusion and
clinical outcomes in patients with acute coronary syndromes. Journal
of the American Medical Association 292(13): 1555-1562.
Bookmark the permalink.