Should I Continue an Experimental Drug?

October 15, 2014
Paul R. Helft, MD
Paul R. Helft, MD

Volume 28, Issue 10

Should I give a patient with a carcinoma of unknown primary another dose of a study drug, since she is requesting to continue it? My reading of the trial protocol suggests no absolute restriction on giving another dose of the drug.

Consultations in Ethics

My group has been following a 29-year-old patient who is a single mother of a 4-year-old child, and who was diagnosed with a carcinoma of unknown primary 9 months ago. She was treated with a platiunum-based regimen and initially had a very good response. She relapsed rapidly, however, and quickly developed serious, progressive, end-stage complications, including ascites, renal failure, refractory pain, and cachexia, and she is now confined to an ICU bed. Her last treatment regimen was administered in a phase I trial. Despite many conversations about her impending death, and our making recommendations to limit, withhold, and withdraw various treatments, including the phase I drug, the patient and her family have continued to desire any and all treatments, including the phase I drug she was previously receiving. My reading of the trial protocol suggests no absolute restriction on giving another dose of the drug. Should I give the patient another dose of this study drug, since she is requesting to continue it?

Dr. Helft Responds:

The tragedy inherent in the patient’s situation is palpable, and probably contributes to the complexity of the decision making. Clinicians desperately want to do something to try to fulfill a patient’s desire to get better. However, I would like to respond to your question first by separating out the purely clinical issues from the ethical issues.

Clinically, several dimensions of this patient’s situation are significant. For example, it is important to consider the chances of the phase I agent providing benefit. The historical solid tumor response rate in phase I clinical trials has been reported to be 4% to 6% (see, for example, Von Hoff DD, Turner J. Response rates, duration of response, and dose response effects in phase I studies of antineoplastics. Invest New Drugs. 1991;9:115-22). However, more recent analyses have suggested that response rates are closer to 11%, and higher for trials in which at least one cytotoxic chemotherapy drug is also included (see Horstmann E, McCabe MS, Grochow L, et al. Risks and benefits of oncology phase 1 clinical trials: 1991 through 2002. N Engl J Med. 2005;352:895-904). These estimates are also likely to be lower in patients who have a poor performance status, such as your patient. The second clinical issue is whether the research protocol actually allows another dose of the drug, which would be unusual, given the restrictive nature of most phase I protocols. The final issue I would like to raise is that of the possibility that another dose of the study drug will harm her. This depends partially on an assessment of whether previous doses have led to any toxicity, along with consideration of whether her current comorbidities increase the risk of toxicity.

The point of raising these clinical issues is that it is critically important to arrive at an assessment of the risk-benefit ratio of providing another dose of the study drug. In this case, I think it is likely that the risk-benefit ratio is unfavorable, mostly because of the very low odds of benefit, raising the ethical issue of whether further treatment with the study drug is more likely to harm than to help her.

On to the ethical issues. The first, in direct response to your question, is that-especially in the setting of a patient whose death is imminent-it is generally not consistent with principles of medical ethics to provide a treatment whose potential risk is greater than the potential benefit, even if patients desire it. That looks to be the case here, where you might argue that the chance of benefit for any treatment, let alone an experimental drug, is as close to zero as it can be. This aligns well with the principle of non-maleficence, which seeks to minimize harms when treatments no longer provide benefit. I like to refer to refusals to provide a requested treatment as “compassionate refusals,” since in order to minimize conflict with patients and families, they must be carried out with a strong sense that the clinician has the patient’s best interest at heart.

It is tempting at times to go ahead and provide such treatments for psychological reasons when the chances of benefit and of harm are both very low. The rationale is usually as follows: “It probably won’t help or harm, but it will help the patient feel that we are still doing something.” The illogical nature of such thinking should be apparent: treating psychological distress with anticancer drugs is neither rational nor clinically appropriate. The psychological distress that results from decisions to cease or to forgo further anticancer chemotherapy should be treated with empathic engagement with the distressing emotions.

Decisions to withhold or withdraw treatments, what I sometimes call “negative decisions,” depend fundamentally on trust in the person making the recommendation to withhold or withdraw the treatment. Such trust must be carefully cultivated, particularly through high-quality, empathic communication.

In a quandary about the ethical dimensions of a management decision?

Practicing oncologists regularly wrestle with a wide variety of ethical issues. If you have a case that you think would benefit from the advice of an oncologist with special training in ethics, email it to Consultations will be provided by Dr. Paul Helft, director of the Charles Warren Fairbanks Center for Medical Ethics. Cases that the Editors feel might be relevant and helpful to others may be published in a future issue of ONCOLOGY and/or on, with identifying details changed or removed to protect privacy.