Progress in science-based medicine depends upon human experimentation. Scientists can do the most fantastic translational research in the world, starting with elegant hypotheses, tested through in vitro and biochemical experiments, after which they are tested in animals. They can understand disease mechanisms to the individual amino acid level in a protein or nucleotide in a DNA molecule. However, without human testing, they will never know if the end result of all that elegant science will actually do what it is intended to do and to make real human patients better. They will never know if the fruits of all that labor will actually cure disease. However, it is in human experimentation where the ethics of science tend to clash with the mechanisms of science. I refer to “science-based medicine” (SBM) as “based” in science, but not science, largely because medicine can never be pure science. There are too many other factors that affect choice of therapies, including patient preferences as well as what is available and feasible. Science has resulted in amazing medical advances over the last century, but if there is one thing that we have learned it’s that, because clinical trials involve living, breathing, fellow human beings, what is the most scientifically rigorous trial design might not be the most ethical.
About two weeks ago, the AP reported that experiments and clinical trials that resemble the infamous Tuskegee syphilis study and the less well known, but recently revealed Guatemala syphilis experiment were far more common than we might like to admit. As I sat through talks about clinical trial results at the Society of Surgical Oncology meeting in San Antonio a week ago, the revelations in this AP story reminded me that the intersection between science and ethics in medicine can frequently be a very tough question indeed. In fact, in many of the discussions, questions of what could or could not be done based on ethics were frequently mentioned, such as whether it is ethically acceptable or possible to do certain followup trials to famous breast cancer clinical trials. Unfortunately, it was not so long ago that such questions were answered in ways that bring shame on the medical profession.
More than Tuskegee and Guatemala
The most notorious of highly unethical human experiments outside of Nazi Germany and the Japanese empire during World War II is the infamous Tuskegee syphilis study. This study, conducted by our very own Public Health Service (PHS) was conducted between 1932 and 1972 and examined the natural progression of untreated syphilis in poor black men who received free health care from the government. In 1932, when this study was conceived it was not totally unethical. At the time there were precious few treatments for syphilis, and none of them worked very well. Consequently, observing the progression of syphilis, using the treatments available at the time, and following the subjects prospectively constituted a reasonable trial design. What did not constitute a reasonable trial design was not telling subjects’ wives. Worse, in the late 1930s and early 1940s, penicillin became available. By 1947 was the standard of care for treating syphilis. When campaigns to eradicate syphilis came to the county in which most of the subjects, study researchers prevented their subjects from participating. In essence, even after an effective treatment for syphilis had become widely available, study still researchers denied it to their subjects. By the end of the study in 1972, of the original 399 men in the study, 28 had died of syphilis; 100 were dead of related complications; 40 wives had been infected with syphilis; and 19 children had been born with congenital syphilis. The rationale for not providing effective treatment for these men and even discouraging them from undergoing such treatment? This:
Such individuals seemed to offer an unusual opportunity to study the untreated syphilitic patients from the beginning of the disease to the death of the infected person. An opportunity was also offered to compare the syphilitc process uninfluenced by modern treatment, with the results attained when treatment had been given.
Worse, there was no informed consent, and considerable inducements were offered to the men to join the study.
The Tuskegee syphilis study, unfortunately, is not the only abuse committed by the PHS. About six months ago, it was revealed that these sorts of experiments had been more widespread than commonly believed. Indeed, in the 1940s in Guatemala, the PHS had gone one better in that they had deliberately infected prison inmates in Guatemala with syphilis. As I described in a lot more detail when the revelations first hit the press, prostitutes who had tested positive for syphilis were recruited to visit the men in prison. The hypothesis to be tested was whether prophylactic penicillin treatment could prevent infection, and the other purpose was to define the response of syphilis to penicillin treatment. Again, there was no real informed consent. Worse, subjects were intentionally infected with a potentially fatal disease. True, they were treated, but treatment is not 100% effective, and one has to wonder if the prisoners, a vulnerable population completely under the control of prison personnel, understood the nature of the risks they were being induced to take.
Two weeks ago AP medical writer Michael Stobbe published a long article detailing the sordid history of medical research in the U.S. before the 1970s. His timing was not coincidental, because on the that followed there was a meeting of a presidential bioethics committee, the Commission for the Study of Bioethical Issues, triggered by the revelations last fall about the Guatemala syphilis experiment 65 years ago. Those revelations led the AP to do an exhaustive review of reports from medical journals and press clippings, and the AP found at least 40 studies similar to the Guatemala syphilis study in that patients were put at risk for serious disease or, even worse, healthy people were intentionally made ill to study disease. Some of these abuses are well known, others much less so.
Here are some examples from the AP article:
The AP review of past research found:
- A federally funded study begun in 1942 injected experimental flu vaccine in male patients at a state insane asylum in Ypsilanti, Mich., then exposed them to flu several months later. It was co-authored by Dr. Jonas Salk, who a decade later would become famous as inventor of the polio vaccine.
Some of the men weren’t able to describe their symptoms, raising serious questions about how well they understood what was being done to them. One newspaper account mentioned the test subjects were “senile and debilitated.” Then it quickly moved on to the promising results.
- In federally funded studies in the 1940s, noted researcher Dr. W. Paul Havens Jr. exposed men to hepatitis in a series of experiments, including one using patients from mental institutions in Middletown and Norwich, Conn. Havens, a World Health Organization expert on viral diseases, was one of the first scientists to differentiate types of hepatitis and their causes.
A search of various news archives found no mention of the mental patients study, which made eight healthy men ill but broke no new ground in understanding the disease.
- Researchers in the mid-1940s studied the transmission of a deadly stomach bug by having young men swallow unfiltered stool suspension. The study was conducted at the New York State Vocational Institution, a reformatory prison in West Coxsackie. The point was to see how well the disease spread that way as compared to spraying the germs and having test subjects breathe it. Swallowing it was a more effective way to spread the disease, the researchers concluded. The study doesn’t explain if the men were rewarded for this awful task.
- A University of Minnesota study in the late 1940s injected 11 public service employee volunteers with malaria, then starved them for five days. Some were also subjected to hard labor, and those men lost an average of 14 pounds. They were treated for malarial fevers with quinine sulfate. One of the authors was Ancel Keys, a noted dietary scientist who developed K-rations for the military and the Mediterranean diet for the public. But a search of various news archives found no mention of the study.
- For a study in 1957, when the Asian flu pandemic was spreading, federal researchers sprayed the virus in the noses of 23 inmates at Patuxent prison in Jessup, Md., to compare their reactions to those of 32 virus-exposed inmates who had been given a new vaccine.
- Government researchers in the 1950s tried to infect about two dozen volunteering prison inmates with gonorrhea using two different methods in an experiment at a federal penitentiary in Atlanta. The bacteria was pumped directly into the urinary tract through the penis, according to their paper.
The men quickly developed the disease, but the researchers noted this method wasn’t comparable to how men normally got infected — by having sex with an infected partner. The men were later treated with antibiotics. The study was published in the Journal of the American Medical Association, but there was no mention of it in various news archives.
Stobbe goes on to point out the “Holy Trinity” of news stories in the 1960s and early 1970s that brought to light the sorts of activities that we now consider abuses in medical research. The last of these was, of course, the Tuskegee syphilis study. The first of these occurred in 1963, when it came to light that researchers had injected cancer cells into elderly debilitated patients at the Jewish Chronic Disease Hospital in Brooklyn to discover whether their bodies would reject them. With our knowledge of tumor immunology now, we can look back on this experiment and know that the odds of any harm were quite small because tumors, with very, very rare exceptions, are not transplantable in humans. Our bodies recognize cells from another person to be foreign, whether they are cancer or not, and quickly destroy them. However, at the time, based on what was known, undoubtedly the scientists thought that there was at least a chance that these tumor cells would form cancers in the patients into whom they were injected, the denial of the hospital director who deemed the cells “harmless,” notwithstanding. (Indeed, the hospital director strikes me as either lying or deluded.) Otherwise, why seek to answer the question? Just as bad, there was no informed consent, the justification being that the cells were thought to be “harmless.” More details can be found here. The outcome was that the Board of Regents censured the researchers and suspended the licenses of two of the doctors involved. Later, however, they stayed the suspensions and instead put the doctors on probation for one year. There were no repercussions for the hospital or for Memorial Sloan-Kettering Cancer Center, where one of the investigators was on faculty.
The third of the “Holy Trinity” was an infamous experiment in Staten Island at the Willowbrook State School, which was a school for children with mental retardation. During the mid-1960s, children there were intentionally infected with hepatitis in order to determine whether gamma globulin could cure it. Besides the targeting of a vulnerable population (children and teens with profound mental retardation), this study demonstrated a number of problematic issues as well. First, the investigators rationalized infecting these children by rationalizing that hepatitis was so endemic in the facility due to the fact that most of the children there were incapable of being toilet trained that over 70% of new residents became infected within a year. This, of course, leads to the obvious question, namely: If that were the case then why not study the effect of gamma globulin on children who were infected normally? More disturbing, again investigators played fast and loose with informed consent, the form being worded in a vague and ambiguous manner that played down the fact that the children were going to be intentionally infected with hepatitis and implying that the serum they would be given would be an experimental vaccine. Finally, as is the case in many such studies, there was an element of coercion. Willowbrook at the time was very crowded, with long waiting lists for children to be admitted. At times, there was only room in the experimental wing. For parents who could not afford to take care of their children, this situation could bring considerable pressure to bear to “persuade” them to “do the right thing.”
In studying the history of medicine and clinical trials, what never ceases to amaze me is the different attitudes that physicians and scientists had towards their human subjects not all that long ago. Remember, it was primarily in the 1960s and 1970s when attitudes began to change. Before the 1970s, for instance, researchers thought almost nothing of using prisoners for experiments, even though prisoners are correctly considered a population for whom true informed consent without coercion is difficult to obtain without special effort. Indeed, in his news story Stobbe recounts an anecdote of a man at Holmesburg Prison in Philadelphia who agreed in exchange for cigarette money to have the skin peeled off of his back and searing chemicals painted on the open wounds in order to test a drug. Similarly, as the Willowbrook story shows us, it was not really all that long ago when scientists apparently felt justified in infecting profoundly mentally retarded children with hepatitis on the basis of at best dubious ethical justification.
Arguably, this willingness to experiment on children who were not normal and who were never going to be able to contribute to society was a holdover from the eugenics movement earlier in the 20th century. It’s important to remember that, however much eugenics was discredited by the Nazis, prior to the Holocaust Hitler was actually quite the admirer of American eugenics policies, drawing inspiration from them. Another factor that is frequently invoked as an explanation for the willingness of American scientists to flout ethical considerations is war, particularly World War II and then the Cold War, the resulting idea that it was “us against them,” and that for us to win would require shared sacrifice in the name of the nation. After all, the scientific primacy of the U.S. was correctly viewed as one of the most critical sources of our economic and military strength. Unfortunately, the lesson derived from that observation was that this primacy needed to be maintained at all costs. Moreover, the concept of the Cold War could be generalized to other “wars,” such as the “war on disease” or the current “war on cancer” that I’ve written about twice in the last month. As Stobbe put it:
Attitudes about medical research were different then. Infectious diseases killed many more people years ago, and doctors worked urgently to invent and test cures. Many prominent researchers felt it was legitimate to experiment on people who did not have full rights in society — people like prisoners, mental patients, poor blacks. It was an attitude in some ways similar to that of Nazi doctors experimenting on Jews.
“There was definitely a sense — that we don’t have today — that sacrifice for the nation was important,” said Laura Stark, a Wesleyan University assistant professor of science in society, who is writing a book about past federal medical experiments.
There was clearly also more than a little hubris at play as well:
It was at about this time that prosecution of Nazi doctors in 1947 led to the “Nuremberg Code,” a set of international rules to protect human test subjects. Many U.S. doctors essentially ignored them, arguing that they applied to Nazi atrocities — not to American medicine.
Finally, with the rise of large pharmaceutical companies in the 1940s and 1950s, increasingly there was more of a profit motive than a purely scientific one. Drugs needed to be tested, and prisoners provided a convenient source of young, healthy men upon which to test new products. Hubris, profit, and a wartime attitude that sacrificing for the good of the nation all swirled together into a mixture toxic to medical ethics during World War II and well into the postwar period. In having defeated the Nazis, we failed to learn a lesson from what had happened in Germany, where one of the most technologically and medically advanced societies then on the face of the earth did horrible things in the name of its ideology.
Yes, it is true that American scientists did not intentionally expose prisoners to freezing water in experiments designed to find better ways of rewarming pilots shot down over frigid waters or sailors who survived the sinking of their ship, as Nazi doctors did. It is also true that American scientists did not intentionally irradiate men’s testicles and women’s ovaries in order to develop a means of rapid sterilization, causing horrific bowel and bladder complications, especially in women, as Nazi scientists did, although American scientists did subject many to various radioactive substances in the name of research. Nor did American scientists inject dyes into the eyes of children in order to try to turn them blue, as Dr. Mengele did. On the other hand, American scientists did, as we have seen, intentionally infect prisoners and mentally retarded children (the same sort of children that the Nazis would have called Lebensunwertes Leben, or “life unworthy of life”) with diseases and then treat them, just as Nazi physicians intentionally infected concentration camp inmates with various diseases in order to determine the efficacy of different treatments or as Japanese physicians did when they intentionally broke the limbs of prisoners and contaminated them with bacteria-laden dirt. American offenses were different in scale and horror, but not significantly different in kind. Unfortunately, it was not until the 1970s, years after the international Helsinki Declaration was first published, until the Belmont Report was adopted and then not until the 1990s when The Common Rule became the basis of all federal regulations protecting human research subjects.
Could it happen again?
Fortunately, as one who now participates in clinical trials and clinical trial development, given the current level of regulation on human subjects research by the federal government, I have a hard time imagining how abuses such as the one’s I’ve described could happen again now. The amount of paperwork, regulation, and oversight of clinical trials has become so burdensome and complex that sometimes I wonder why I or anyone else would want to continue doing clinical research. Unfortunately, Stobbe doesn’t sound too optimistic. Actually, it’s not so much Stobbe, but rather the presidential Commission for the Study of Bioethical Issues, as Stobbe documents in a followup story:
Speakers noted that over the last several decades, as many as 1,000 rules, regulations and guidelines have been enacted worldwide to ensure the ethical conduct of medical research. In the United States, there are rules to protect people in every study done by federal scientists, funded by federal agencies or those testing a product requiring federal approval to be sold.
But that oversight is inconsistent — ethical rules can vary among federal agencies. What’s more, if federal funding or review is not involved, an unethical study could be done and no one in authority would ever know about it.
“We have a leaky system,” said Eric Meslin, director of the Indiana University Center for Bioethics.
Dr. Robert Califf, Duke University’s vice chancellor for clinical research, agreed there are weaknesses.
“It’s night and day and what you could do in the ‘good old days’ with no one knowing about it. But there’s no 100 percent guarantee. There still will be bad things that will happen,” he said.
In terms of pharmaceutical companies, there are clearly loopholes when it comes to overseas studies. Indeed, pharmaceutical companies have been doing more and more studies overseas. Although federal law states that such studies, if they are funded by the federal government or if they are to be used as part of an application for FDA approval of a drug, that is not always enough of a guarantee of oversight:
Last year, the U.S. Department of Health and Human Services’ inspector general reported that between 40 and 65 percent of clinical studies of federally regulated medical products were done in other countries in 2008, and that proportion probably has grown. The report also noted that U.S. regulators inspected fewer than 1 percent of foreign clinical trial sites.
Clearly, this is an unacceptable level of oversight, particularly outside of developed countries, such as those in Europe, where clinical trial oversight is comparable to that in the U.S.
Ironically, two examples come to mind of clinical trials that show the holes in our regulatory system for human subjects protection, both of which I have written about right here on SBM before. The first trial was a trial of homeopathic remedies for infants with infectious diarrhea in Honduras, as I wrote about here. At the time I couldn’t figure out how the investigators at the University of Washington managed to get this study through their IRB, but somehow they did, demonstrating that an IRB is not a guarantee against the approval of totally unethical and scientifically worthless experiments. Fortunately, as far as I can tell, no infant was injured, but the potential was definitely there. Then, let’s not forget the Gonzalez trial, a trial of a regimen of what can best be described as pure quackery consisting of up to 150 supplement pills a day, various nutritional pseudoscience, and daily (or more) coffee enemas. The results were devastating, in that subjects on the standard-of-care chemotherapy arm lived three times longer than those on the Gonzalez protocol arm. Such is the effect of the National Center for Complementary and Alternative Medicine (NCCAM) on research ethics.
The more disturbing example is Mark and David Geier, the father-son tag team of anti-vaccine activists who fervently believe that mercury in vaccines causes autism and somehow came up with an idea that can only be described as dangerously wacky, namely that by suppressing testosterone with a powerful drug (Lupron) they could make the quackery known as chelation therapy “work better” at chelating mercury from the brains of autistic children. The reason? Because “testosterone sheets” bind mercury and keep it from being chelated! In pursuing this research, the Geiers have created an IRB stocked with their cronies and fellow anti-vaccinationists to “oversee” the research, as Kathleen Seidel has so thoroughly documented. In the process, autistic children were subjected to a powerful drug that depresses their sex hormone levels, which is why its use is often referred to as “chemical castration.”
Predictably, Anne Dachel, Media Editor over at the anti-vaccine crank blog Age of Autism, has leapt all over this story as “evidence” that vaccines must be dangerous and that unethical scientists have been lying all along about the science showing tthat there is no evidence that vaccines cause autism:
Either Stobbe is a naÃ¯ve and trusting soul and can’t consider that the same government that allowed horrific medical experiments in the past also allowed our children to become vaccine guinea pigs, or he’s afraid of an issue that’s just too controversial to talk about here and now. It’s much safer to attack what went on in the last century.
Maybe 70 years from now, some enterprising reporter will bring up the ethics of injecting known neurotoxins in pregnant women, babies, and small children.
Maybe around 2080, they’ll ask why no one ever demanded independent studies on the cumulative effect of so many vaccines, so soon, on the health of a baby. Or why there was never a simple vax-nonvax comparison study looking at autism rates.
Or maybe in 2050, Dachel will cease laying down such burning stupid on a regular basis and realize that this is the sort of work that’s been replicated so many times and done in so many different countries that even if you were to throw out all the U.S. data it wouldn’t change the scientific conclusion that there is no good evidence to support the claim that vaccines cause autism. She also overlooks the fact that the vast majority of the studies that have failed to find a link between vaccines and autism were performed after the adoption of the Common Rule and much-increased federal oversight over clinical trials. Of course, stories like Stobbe’s make it easier for cranks to attack the entire U.S. clinical research enterprise as corrupt and unethical. However, that is not the reason why we need to close the loopholes in our current clinical trial regulations. We need to do it because it is the right thing to do.
When I write about SBM, I argue that medicine should be based on science, rather than be a science, because I realize that medicine can never be completely scientific. There are too many human variables, not the least of which are patient values, individual patient situations, and resources. Another reason is the clinical trial process itself. Sometimes the most scientifically rigorous clinical trial design is not the most ethical design; indeed, sometimes it might be downright unethical. One example is, as I have pointed out, the aforementioned study of vaccinated versus unvaccinated children that seems to be every anti-vaccine activist’s most fervent dream. The most scientifically rigorous design for such a study would be a randomized, double-blind, placebo-controlled trial. However, such a trial would leave half of its participants completely unprotected against potentially deadly childhood infectious diseases, making it totally unethical to perform, even if it could be scientifically and fiscally justified based on existing preliminary data, which it really cannot. This leaves population-based studies as the only alternative, despite their difficulties and shortcomings.
Perhaps a better example is how placebo-controlled trials have almost gone the way of the dodo in cancer chemotherapy trials. Most oncology trials are now designed to test a new drug against the current standard of care or the new drug plus the standard of care versus standard of care alone. This is because our ethical considerations have evolved such that we now no longer consider giving placebos to cancer patients to be ethical unless there truly is no existing effective treatment for their cancer or if we truly do not know if the proposed treatment is better than observation alone and observation alone is currently the standard of care. As I have described before, in clinical trials, there must be clinical equipoise; i.e., based on the scientific evidence as it is known at the time the trial begins, a reasonable scientific assessment of the risks and benefits must conclude that the risks to the experimental group are either minimal or outweighed by the potential benefits. Here’s another thought to chew on. Experiments in which people were intentionally exposed to infectious agents and then subjected to various treatments to cure the disease thus caused are potentially the most scientifically rigorous way of all to test such treatments in humans because they allow control of the start of the infection, the amount of bacteria injected, and many other variables that can’t be so easily controlled in “wild” cases of infectious disease. However, because such experiments violate the precept of, “First, do no harm,” they are utterly unethical and now properly condemned by any physician with a shred of ethics. That we should require laws, rules, and regulations to prevent such unethical experiments by scientists is unfortunately, but scientists are no different than any other person. Not all of them are ethical; some are completely unethical. Some can be corrupted.
There will always be unethical scientists, at least as long as there are unethical people. That’s why we need laws to protect human subjects. However, we must also remember that the protection of human subjects is a balancing act. Go too far in the direction of lax regulation, and incidents such as those described in Stobbe’s article will start to happen again. Go too far in the other direction, and the pace of discovery will grind to a halt. The key to finding the balance is to respect patient autonomy and to provide true informed consent that accurately balances risks versus benefits and to protect patients from any form of coercion. Doing so without making the clinical trial process so onerous that researchers flee the field while at the same time protecting patients from foreseeable harms will be the challenge.