by Nancy King, JD
I have just read your August 25 post on the EPA studies. While, I do not know that much about these particular exposure studies, I do know that EPA is taking the inquiry very seriously. I do want to question and clarify some of your discussion, principally with respect to Resnik’s 4th and 5th criteria.
You say: “Criteria 4, that the subject must “receive some benefits, such as medical evaluations” seems like a very light standard. If the medical evaluation were part of the study and data was being used for the study, would it be a benefit? I have always viewed a benefit as something that is aimed to help the subject, not the research. In the IRBs in which I have been involved, this is why receiving a study drug is not considered a benefit because it’s not intended to benefit the subject (but to gather data).”
You’re right that offering what John Lantos has called “inclusion benefits” — everybody gets a perk for joining the study — cannot be weighed against risks of harm to subjects in order to justify risk-bearing research. Nor can you offer more than a “light” benefit in order to balance greater risks of harm. That’s why the only benefits that count in the harm-benefit balance should be direct benefits to society from generalizable knowledge, and in some but not all research, benefits from the intervention under study. That works for phase III randomized clinical trials (RCT) where there is clinical equipoise — and it can reasonably apply to getting the study drug in a phase III RCT that is well-supported by data from earlier phases. But you’re right, it can’t be used outside that final study phase.
Then you say: “Criteria 5 is also concerning. Phase 1 trials are usually done with a small number of sick patients suffering from a disease who have often exhausted all approved treatment options. In the EPA studies, no potential therapy was being tested. Exposure to the pollutants were not intended to make people better and were given to see how much sicker they would get. Thus, this criterion is problematic because the analogy between a clinical trial and environmental exposure does not hold. And in a clinical trial there is a remote possibility (even if unintended) that a person might be helped whereas an exposure trial holds no such hope. An asthmatic or COPD patient has no potential benefit from exposure only risk.”
Phase 1 trials were originally drug studies with healthy volunteers, not patient-subject trials. Using patients as subjects is part of the change wrought by oncology research (because you can’t minimize the harms from cancer treatment), and phase 1 patient-subjects are supposed to be subjects for whom nonmaleficence, not beneficence, is the primary consideration. That is, these are subjects for whom nothing else has worked, and therefore it is okay to ask them to participate in a study because they will not be kept from anything that does work because they have enrolled in research. (In other phase 1 studies, patients may be able to postpone effective treatment without too much harm — think asking an adult taking an asthma or ADHD medication that works pretty well who is asked to test a new medication for 6 weeks. A study with a rescue medicine regimen and good stopping rules might be okay.)
You are absolutely right that elderly, ill, and otherwise vulnerable subjects need many well-tailored and closely monitored protections in all research and especially phase 1, but it is simply wrong to look for benefit to subjects in phase 1 studies. The intent to benefit is not for the subjects, it’s for future patients.
I am concerned that the meaning and intent of early-phase and first-in-humans research is getting warped, and that as a result, we are getting into a morass of therapeutic misconception. You and the IRBs you’ve worked with are far from the only folks making this mistake.
The cure for dangerous research is not to balance it with direct benefits. You’re totally right that some research should not be done, that risk minimization must be a real effort, and that the information given to all subjects must be much clearer. But we simply can’t paper over the risks of research with misplaced potential benefits. Pamela Sankar has written about different ways that investigators describe the dosing cohorts in phase 1 studies — emphasizing safety at lower doses and potential benefit at higher doses. Even that deceptive emphasis is simply not true. It’s not true for drugs, because in phase 1 you don’t have much of a clue about whether it’ll work, and it is especially untrue for biotechnological interventions, which don’t have pharmaceutical-style dose-response curves. What’s needed is honesty about the meaning of “maximum tolerated dose” and at how it is arrived.
I once heard Alex Capron say that there is an inevitable component of human sacrifice in research. He was right. And you’re right that the analogy between exposure studies and clinical trials doesn’t really work, but not for the reasons you state. Improvements in both caution and communication are increasingly necessary in human subjects research these days.