Bioethical Lessons From the Gelsinger Case: Three Myths of Human Experimentation
Jesse Gelsinger had just turned 18 when he volunteered to participate in a gene therapy experiment at the Institute for Human Gene Therapy at the University of Pennsylvania. When he appeared for the trial on September 9, 1999, he weighed 130 pounds and had a normal life expectancy, though he would always suffer from OTC, a genetic disease affecting the liver's ability to process ammonia. On September 13, Jesse was strapped down and sedated while a catheter was inserted into his groin. A doctor then injected 32 trillion particles of an adenovirus into his liver. Within 4 days, Jesse was in an irreversible coma. The virus had caused a massive immune response and release of fluids, which swelled his body by almost 40 pounds. When his parents agreed to a withdrawal of life support, their son was unrecognizable.
For the moment, put aside the numerous violations of FDA regulations committed by the Penn research team. Penn and Dr. Wilson have maintained that these FDA violations did not cause the death of Jesse Gelsinger. While not exactly accurate, focusing on those mistakes, as the FDA and the media have done, ignores the root cause of Jesse's death, which must be remedied to avoid further tragedies. The bioethical decisions made at the inception of the OTC trial are what killed Jesse Gelsinger, and the blame for those decisions reaches far beyond the principal investigators.
The first was the decision to permit James Wilson to conduct gene therapy experiments at Penn while maintaining a majority interest in Genovo, a biotech company that stood to make millions if the experiments were successful. This issue was debated at length by Penn's Conflict of Interest Standing Committee ("CISC"),whose job it is to make sure that university research is not burdened or compromised by conflicts of interest. In the minutes of one such committee meeting, members asked the critical question: "Since Dr. Wilson's research will be directed towards a solution in which he has a financial interest in the outcome, how can Dr. Wilson assure the University that he will not be conflicted when making decisions that could have an impact on ...the further development of his intellectual property?" The answer , of course, was he couldn't. The essence of the experiment was an attempt to prove that Dr. Wilson's vectors were the appropriate taxicabs for gene transfer. At the same meeting, the committee asked, "How can Dr. Wilson and the University avoid liability for any damages if a patient died from any products produced or studied at the University." Curiously, this prescient question is omitted from the final minutes of the CISC. The answer to that question was obvious. Hence,
MYTH 1: The purpose of a clinical trial is to benefit medicine and science at little or no increased risk to the subject.
The truth is every clinical trial is burdened by conflicting interests. Investigators and their institutions almost always have a financial stake in the experiment. Investigators can make a million dollars a year doing nothing but clinical trials. Pharmaceutical companies pay investigators up to $2500 for each subject recruited. In many trials, like the one in which Jesse Gelsinger died, both the institution and the investigator stand to profit greatly if the research is successful and if that success comes before anyone else's. It is too often a race not against a disease but against other experiments for the riches a victory may bring.
Investigators have other interests just as dangerous, because these interests like money divert the investigator's eyes off the patient as an autonomous being. The need and desire for fame, prestige, respect of their peers, and the necessity to publish or perish all exert pressure on the investigator to think of his work first, to bend the rules, and to complete the study at any and all costs.
But keep in mind, the absence of any self interest- manifested by the belief in the greater good, the quest for knowledge, the interests of mankind, the interests of the state-- these too are potential evils. The history of medicine and science is littered with subjects sacrificed for the greater good: the Nazi experiments, Unit 731, Tuskeegee, the US Radiation studies, Willowbrook. These experiments were not poisoned by self-interest. They were poisoned by the arrogance of men who believed they could make the choice of who would be martyrs for science or for the greater good. Not surprisingly throughout history the chosen ones, the subjects, were the prisoner, the mentally infirm, the Jew, the African American, the poor-- those who the poet called the abandoned and forsaked.
In any clinical trial, the well being of the subject must be the first priority. It is the only interest that matters. As the sages advised: a physician must do no harm; he must never see in the patient anything other than a fellow creature in pain; he must never do an experiment harmful to the patient even though advantageous to science or to others.
The second bioethical decision was the approval of the design of the experiment. Originally, Dr. Wilson and his team designed an experiment on infants born with OTC, facing almost certain death in the first months of life. The goal of the study was to see if gene transfer of a healthy OTC gene could allow these infants to live long enough to be put on the drug regimen designed by Dr. Batshaw, one of the principal investigators. Without the gene therapy, the babies would surely die so the benefits to the babies, even though remote, might outweigh the risks they would die in the experiment. In stepped Arthur Caplan, Penn's resident bioethicist and consultant to IHGT. That experiment is unethical, he said, because mothers distraught over the birth of a dying child could never make an informed decision about whether to allow the infant to participate in the study. I do not quarrel with that decision. Instead, Caplan proposed, conduct the experiment on healthy adults with mild forms of OTC. How is this a more ethical decision? The risks in this study were great; they were life threatening. The potential benefits to the participants were nonexistent. Such an experiment violated the Nuremberg Code, the Declaration of Helsinki, and any conceivable ethical norm governing science and medicine.
Yet this experiment was approved by three different bodies which should have rejected it as too risky and unethical. First, Penn's IRB approved the study, presumably without any understanding of the risks to the subjects, the nonexistent benefits, and the conflicts of interests of not only the principal investigator but of the University itself. Then it was approved by the Recombinant DNA Transfer Activities Committee, known as "The RAC," which was composed of the top scientists in the gene therapy field. A prominent member of the RAC later admitted that the members should have rejected the study as too risky but did not because of their own conflicts of interest. Finally, the FDA, the last gatekeeper, rubber-stamped the RAC's approval of the study, despite its clear departure from the bioethical rules governing such work. Hence,
MYTH 2: Somebody is watching; the investigator is watched by the IRB, the IRB by the OHRP, the OHRP by the HHS, the HHS by the FDA, the FDA by Congress.
The truth is no one is watching. There are 2-300,000 trials going on at any one time involving 7-8,000,000 participants. One major university was conducting 4,000 clinical trials at once; a small southern school 1200. How can an IRB do what it is supposed to do when some, like the one at the University of Oklahoma, meet once a month for an hour with dinner to follow. IRBs have certain critical duties under the law.
First, it is the duty to examine the design of the trials. The IRB must determine the importance of the knowledge to be gained by the result and the likelihood of gaining that knowledge by the experiment. The IRB must also make certain the experiment is based on competent animal studies and generally accepted scientific principles. The IRB must understand the risk-benefit calculus of the protocol and determine that the risks have been minimized. This means that, in a therapeutic trial, a state of clinical equipoise exists between the experimental treatment and the generally accepted course of treatment. In a nontherapeutic trial, the standard is minimal risk plus the attendant risk of the procedures necessary for the collection of data such as blood tests. For children or incompetents, the standard is nothing more than minimal risk.
Second, the IRB must evaluate the competence of the investigator. Those performing human studies should be at the top of their profession. In addition, the IRB must make certain clinically competent medical personnel are available for supervision and support.
Third, the IRB must make sure the selection process for subjects is equitable and just, that it does not take advantage of the minority, the poor, the imprisoned, the uninsured or the infirm.
Fourth, the IRB must review the informed consent document to determine if it describes the design and purpose of the trial, the procedures to be performed, the risks, the benefits, and the alternative treatments available.
Fifth, the IRB must monitor the trial. It must know if the trial falls out of clinical equipoise; it must make sure adverse events are being reported; and it must consider changes to the protocol on a prospective basis only.
How can IRBs possibly perform these critical functions given the amount of trials under their supervision?. They don't do that. They can't do that.
Government agencies are supposed to be doing all the same tasks and monitor the IRBs. They don't do that. They can't do that. In the 5 years before Jesse Gelsinger's death there were 39 adverse events reported in gene therapy trials. Four months after Jesse's death, 700 were reported. In the last 10 years, there have been 9 deaths reported as adverse events in clinical trials involving 70 to 80 million subjects. One study concluded the number should be at a minimum 20,000. I am not saying 20,000 caused by the experiment. That is not the definition of an adverse event. Adverse events are any serious health events, expected or unexpected, occurring during the course of the trial. It is not for the investigator to determine whether the experiment caused the event before it is reported. These numbers reveal the obvious: No one knows how safe these trials really are because no one is watching.
As I said, IRBs are supposed to look at the informed consent document. By most accounts that is what they spend most of their time on under the assumption that this document somehow provides their institution a release of liability. In the Gelsinger case, the informed consent document was misleading, incomplete and totally lacking in its ability to convey the complex nature of the experiment to the subjects. It was misleading because it described Gene Therapy as if it were an established medical procedure and not a speculative theory that has still not produced a single successful treatment for any disease or affliction. It was incomplete because it failed to include the fact that monkeys had died in the same experiment. But, more important, even if these faults had been corrected, the document could never produce an understanding in the lay reader of the purpose of the procedure, the substantial risks of the adenovirus injection, and the limited and speculative benefits to the advancement of medicine. Hence,
MYTH 3: The informed consent document evidences informed consent
The truth is it neither evidences that the subjects are informed or that they consent. The subjects are not informed because the document is not written in language they can understand. The subjects are not informed because, when the document is explained to them by the investigator, the wink of a trusted eye and the inflection in the voice trained to comfort tell the subjects what they want and need to hear: they are better off by participating; they are in good hands; and their interests are what matter. They are not informed because only the investigators and their fellow scientists understand the purpose of the experiment, the benefits to the subjects, and the risks of harm.
It does not evidence consent because it is not the subject's right to consent. A subject cannot consent to an unethical experiment. Anything more than minimal risk in a nontherapeutic trial, anything other than clinical equipoise in a therapeutic trial, is contrary to the ethical norms of Hippocrates, Maimonides, Nuremberg, and Helsinki. We as a society can not and do not consent to any experiment on any of us which has a greater risk of harm than the risks of everyday life. And it is our informed consent that is required.