Data or life? Ethical obligations to present and future patients.

By Jahel Queralt-Lange

Each year 10.9 million new cases of cancer are diagnosed worldwide, and 6.7 million people die. The good news is that better drugs are developing faster. We all want to hear about “wonder drugs” and the scientific and medical communities feel the urge (and sometimes the pressure) to provide them. However, some ethical problems might appear in our way to this breakthrough. Before being released into the market, any drug has to undergo a trial in which its benefits are tested and weighed up against its adverse effects. Of course, the desirability of increasing our knowledge in order to improve our health is beyond question. What poses problems is that those individuals who participate in the trials are the means by which we achieve that knowledge. Past experiences, like the Tuskegee Syphilis experiment that took place in the US showed that it’s morally outrageous to trade the life of some individuals for the sake of benefiting society. Nowadays the issue is regulated but moral dilemmas persist.

Weeks ago The New York Times reported the story of two cousins with melanoma cancer, who were both enrolled in the last stage of a clinical trial . At that point, the benefits of the drug that was tested, PLX4032, had been proved to be significant. As the article says, “The standard chemotherapy used in melanoma, dacarbazine, slowed tumour growth in 15% of patients for an average of two months. By contrast, PLX4032 had altered tumor growth in 81% of patients for an average of eight”. What was uncertain, however, was the efficacy of the drug in extending life. The tumors of those who took it shrunk but there was still no evidence of how long they were going to live. As it happens in these cases, a computer decided by lottery the treatment each cousin was going to get. Luck was not equally merciful to both and only one of them had the chance to try the new drug, the other one was assigned to the control group. He got the standard chemo and the doctor’s apology for that. The health of the unlucky cousin declined tremendously and he asked to switch to the other group. His plea was declined. Roche, the pharmaceutical company that manufactured the drug, considered that in order to ensure a clean comparison they had to withhold PLX4032 from patients that had been treated with chemotherapy. So, patients had to cope with the decision of the randomizing computer. The unlucky cousin died whereas the lucky one is still alive. Putting aside complex methodological issues, the striking moral question posed by this case seems to be this: should we stop a trial early because of apparent benefit? If we have evidence that the benefits of a drug out-weigh its bad-effects, is it morally justified to carry on with a lottery that gives patients 50% chance of getting a treatment that is less good?

At least two reasons can be offered against stopping the trial early. The first one is that it (still) hasn’t been proved that those who took the drug will live longer. This is, in fact, what the doctors who support the trial have argued. But why should we assign so much weight to extending life as opposed to improving it? At this phase of the experiment the good results of the drug regarding the latter seem to be solid. One of the things that has been shown is that the drug can ameliorate the condition of the patients who are about to die. On what ground can we deny it to those who, like the unlucky cousin, have an advanced melanoma? Wouldn’t be a good thing to improve the quality of their lives even if only for a few months? Is this result good enough to commercialize the drug? This relates to the second reason. Scientific evidence shows that interrupted trials tend to overestimate treatment effects. If doctors use these incomplete results to treat their future patients, the whole of society will be exposed to harm. That is, by allowing present patients to benefit from a drug, we might be neglecting our obligations to future patients. Morality seems to require avoiding unnecessary suffering to both.

Two more things can be said in favour of stopping the trial – both have to do with conditions required for a clinical trial to be ethical. Firstly, at this stage of the experiment we could question whether there is, still, an “honest null hypothesis”. That is, if there is uncertainty among the scientific community about which one of the two drugs is superior. It’s true that we still don’t know how effective the new drug is at extending people’s lives, but that doesn’t make it worse than the standard treatment because the latter is certainly not good at that (nowadays, only 10% of the patients with a metastatic melanoma survive). Second, because of the difference between the two drugs, patients and doctors know who belongs to the control group and who is in the experimental group. Therefore, the trial isn’t double blind as the rules of scientific validity prescribe. The reasons behind this requirement are well known. In a non-blind experiment there is always the risk that the results might be influenced by the psychological attitudes – e.g. enthusiasm or pessimism- of doctors and patients. Should we then stop enrolling patients in an experiment that can be questioned on ethical and also scientific grounds? Surely, there is no easy answer but there are some measures that can bring us closer to fulfilling all our moral requirements. Perhaps, the less controversial one would be to offer the new drug to patients treated with chemotherapy that are with certainty a few weeks from death. At this stage the relevant data regarding their case has been collected and we can avoid them suffering further.  The second measure would be to stop randomizing and adopt one of these procedures instead: a) maintain the two groups but give the patients the possibility to make an informed choice of which drug they want to take; or b) take as a control group not actual patients but those who already had cancer, dead or alive. The latter is called a “historically controlled trial” and, as John Worrall from LSE, has argued, they can be used without sacrificing scientific validity in cases where, as it happens to be the case, the newly treated group is in any significant way different from the historically treated group ('Why There's No Cause to Randomize', The British Journal for the Philosophy of Science 2007; 58(3):451-488). Using past patients could be after all the best way to protect both present and future patients.

  • Facebook
  • Twitter
  • Reddit

Comments are closed.

Authors

Subscribe Via Email

Affiliations