|
|
|
 |
| Ethics on trial |
17-May-2006 |
|
|
|
The near fatal collapse of six volunteers in a UK clinical trial raises questions about the risks trial participants face.
By Bianca Nogrady
PRIMUM
non nocere… First, do no harm. There is little room for ambiguity in the principle that governs all medical practitioners, but the same rule does not necessarily apply when it comes to clinical trials.
For example, in one recent trial the 73 patients in the placebo group were put under general anaesthetic and had a small femoral incision made for no reason other than to provide a control for those having a genuine surgical procedure. Some members of this control group had serious, albeit transient, adverse events — hardly what you expect from a placebo treatment.
Yet the trial, to determine whether closure of a patent foramen ovale reduced frequency of migraine, was approved by a UK ethics committee, and a second trial of similar design is already under way.
Some experts worry that clinical trials occasionally push the ethical envelope in the pursuit of knowledge, putting patients — volunteers — at what may be considerable risk of harm.
The issue came under global scrutiny earlier this year when a phase 1 trial of a new monoclonal antibody landed six healthy paid volunteers in hospital with multiple organ failure (see box, page 16).
The case was particularly alarming for researchers involved in their own phase 1 trials, such as Professor Paul Komesaroff, director of Monash University’s centre for the study of ethics in medicine and society.
“This case was extremely disturbing for those of us that work in this area because as far as we can tell there was nothing that would have enabled any of the participants to predict the outcome,”he says.
Phase 1 trials are those that test a drug in humans for the first time, an essential step in bringing a new treatment on to the market, but an inherently risky one because earlier trials in animals do not always predict the way a drug will affect humans.
The recent UK trial was a prime example, but it wasn’t the first to have disastrous consequences for patients.
In 1961, Walter Halushka, a student at the University of Saskatchewan in Canada, volunteered for a phase 1 clinical trial of a new anaesthetic. The consent form failed to disclose that the procedure involved cardiac catheterisation. When the drug was administered, Mr Halushka went into cardiac arrest and, although resuscitated, was left with brain damage.
A court found the researchers failed in their duty to inform and awarded damages, setting the benchmark for informed consent with its declaration that “the subject of medical experimentation is entitled to a full and frank disclosure of all the facts, probabilities and opinions which a reasonable man might be expected to consider before giving his consent”.1
Things are not always that clear-cut. A preliminary report by UK regulators into the recent UK trial suggests there was no obvious dereliction of duty by researchers. Professor Komesaroff believes the case shows disasters can happen, even when established procedures are followed meticulously.
However, some experts have raised concerns about whether consent forms signed by the trial’s participants adequately disclosed the drug’s risks. Four men are now arguing they were not told how revolutionary the drug was. Had they known it was so “cutting edge”, their lawyer told the media, they may not have participated in the trial.
Informed consent is only ever as solid as the information given to obtain it, says Mr Bill Madden, from Australian law firm Slater & Gordon.
“Although one view is that it’s simply a matter of consent, the consent would really have to be spot on … because the fact is that a court in those circumstances is going to take it apart line by line,” he says.
That’s all very well in theory, but researchers can’t provide all the information if they themselves don’t know what will happen, says Associate Professor Malcolm Parker, a GP who is associate professor of medical ethics at the University of Queensland.
“Because the whole object of the trial is to find out what the drug’s going to do, both in terms of efficacy and safety, you can’t give absolutely informed consent, because nobody knows what lies ahead,” he says.
When a trial is established specifically to examine a potential harm, it raises ethical issues too. In the wake of the global withdrawal of rofecoxib (Vioxx) due to concerns about a possible increased risk of MI and stroke, several studies have been launched to investigate the safety of other COX-2 inhibitors and NSAIDs in patients with osteoarthritis and particularly those at high cardiovascular risk.
Professor Komesaroff says better safety data on the drugs are undoubtedly needed, although the trials do raise ethical issues.
“The way in which one would assess a study like that … is to ask about what is actually occurring in clinical practice,” he says.
“At the early stage where everyone’s using the therapy and it is in widespread use, one could say there is an ethical obligation to go out and collect data.” But if new data accumulate showing increased risk, the threshold may pass where it is no longer ethical to proceed, he says.
In the case of the COX-2 studies, Professor Komesaroff believes there is an urgent need for more data on the cardiovascular effects of drugs such as celecoxib and meloxicam (Mobic).
“There should be a responsibility on the manufacturers to collect such data,” he says, while emphasising researchers should be cautious about testing the drugs in people at increased risk of cardiovascular events.
Participants in the COX-2 trials can at least expect to gain some therapeutic benefit in return for their exposure to possible risks posed by the drugs. The same cannot be said for all trials.
The MIST (Migraine Intervention with STARFlex Technology) trial subjected the placebo group to a sham procedure including a “light”general anaesthetic and a small femoral incision to simulate insertion of a catheter (Australian Doctor, 24 March).
The intervention group underwent closure of a patent foramen ovale (PFO) to determine if the procedure relieved migraine, after earlier research had revealed a higher frequency of PFO in people with migraine and a treatment effect on migraine following PFO closure.
The study failed to achieve its primary endpoint of complete cessation of migraine but found enough evidence of a therapeutic effect to justify further studies with longer follow-up. Researchers also said the study validated the need for placebo control.
But not everyone would agree that the end justified the means in this case.
Dr Kerry Breen, chairman of the Australian Health Ethics Committee, does not believe the trial would have got past an Australian ethics committee.
“Having untreated controls is only ethical if the people aren’t going to come to any harm,” he says. “To go as far as making cuts, I’m really astounded by that.”
MIST is not the first trial to make use of so-called placebo surgery.
In 2002, a controversial study examining the effectiveness of arthroscopic knee surgery in osteoarthritis gave the control group a short-acting tranquilliser and an opioid, then made several skin incisions.2 Researchers even went so far as to simulate the sounds of debridement and lavage in case the control patients did not have total amnesia.
An editorial in the New
England Journal of Medicinequestioned how informed the consent could really be, asking why a healthy, sane person would volunteer for a trial if they really understood they might have an unnecessary operation with no hope of benefit.3
But, although sham surgical procedures are always going to be controversial, some experts say it can be difficult to evaluate new procedures adequately without them.
With Australian ethics committees unlikely to approve placebo surgery, surgeons are often forced to find alternatives such as pharmaceutical intervention or comparing a new surgical technique with an older one.
“The real issue that it comes down to for the ethics committee is: is the question that’s being asked sufficiently important to warrant the risk of an adverse event in the placebo group?” says Professor Guy Maddern, professor of surgery at the University of Adelaide.
“[In] the arthroscopy trial, for example, the issue there would have been: is arthroscopy being done in such enormous numbers and at such enormous cost to the health system that it warrants putting patients through sham, and I think the answer is probably yes.”
Professor Maddern draws the line, though, at giving the placebo group a general anaesthetic without treatment. An alternative might be to give patients a memory-erasing drug such as midazolam, or do a “con job” on patients, he suggests.
For example, one study in the early days of laparoscopic gall-bladder operations compared open and laparoscopic surgery using fake dressings to hide the smaller incision of the laparoscopic procedure from patients.
So where does that leave the ‘first do no harm’ principle?
Some experts believe clinical research should be evaluated in a different ethical framework from that applied to clinical care, meaning the principle may not always apply.
The NEJM editorial said a randomised controlled trial was not a form of individualised medical therapy but a scientific tool for evaluating treatments in a group of research subjects. “Clinical research involves an inherent tension between the ethical values of pursuing rigorous science and protecting participants from harm.”
Professor Komesaroff believes research and clinical practice do inhabit different ethical landscapes.
“The scientist has to be loyal to the study itself, to the research, to the generation of knowledge,” he says. “The clinician has to be loyal to the patient.”
REAL-LIFE DISASTER
"THEY say he needs a miracle," a young woman told the BBC earlier this year as her critically ill boyfriend was treated in a London hospital for multiple organ failure.
The 28-year-old man was one of eight healthy volunteers paid £2000 ($A4900) to participate in a phase 1 trial of a new monoclonal antibody, TGN1412, designed to treat leukaemia and inflammatory diseases such as arthritis.
Within an hour of taking the drug all six men given the active treatment had collapsed in excruciating pain with massive organ failure, while the two given placebo watched in horror. UK regulators later found cytokines released when T cells were activated had produced a systemic inflammatory response.
The trial drug had been extensively tested in animals, with no sign of such a violent response. The doses were correct, ethical approval was legitimately gained, the volunteers were all suitably healthy before administration and the drug was not contaminated.
The one near-fatal flaw was that the drug was given to all six people within minutes of one another, a decision GP and ethicist Associate Professor Malcolm Parker says has been strongly criticised on both methodological and ethical grounds. "If they only gave it to one person they might have minimised harm."
All but one of the participants have been released from hospital, although lawyers say they are likely to have lifelong health effects. The final patient has complications that may lead to loss of fingers and toes.
References
1. Halushka v University of Saskatchewan 1965.
2. NEJM 2002; 347:81-88.
3. NEJM 2002; 347:137-39.
|
|
|
| | Related Articles - Research | |
|
|
|
|
Subscribe to our daily newsletter
|
|
|
|
|
|
| |
|
|
| |
|
Submit your feedback here: