When I was a medical resident at Harvard Medical School's cancer center, I had an experience that was both unnerving and illuminating. In the wee hours of the morning, I was about to give a patient her scheduled dose of an experimental chemotherapeutic agent, but I had trouble getting the air bubbles out of the syringe. So there I was, muttering and tapping on the syringe, when the patient suddenly sighed deeply ... and died. Had all gone smoothly, I'd have given the drug 20 seconds earlier, she'd have died immediately thereafter, and the death would likely have been recorded as drug related.

Henry I. Miller

I learned from this incident that just because two events are closely related in time doesn't necessarily mean that the first caused the second. (In science, reasoning this way is called the post hoc, ergo propter hoc fallacy: "after the fact; therefore, because of the fact.") Later experience taught me another lesson: If the drug had been controversial, made with a novel technology, or unusual in some other way and had been suspected of killing the patient, politics would have determined the aftermath. These lessons came into play recently when the unfortunate death of a young man treated with the relatively new and experimental medical intervention called human gene therapy elicited a heavy-handed overreaction by federal regulators.

Gene therapy is the insertion of normal or modified genes into a patient, to correct genetic or acquired disorders via the synthesis in the body of missing, defective, or insufficient gene products. Six thousand patients in three dozen countries are currently undergoing gene therapy for diseases ranging from cystic fibrosis to cancer and AIDS.

Eighteen-year-old Jesse Gelsinger, who suffered from a rare metabolic disorder (deficiency of ornithine transcarbamylase, or OTC) caused by subnormal levels of a liver enzyme, and marked by toxic levels of ammonia in the blood, was quite healthy when he arrived at the Hospital of the University of Pennsylvania to begin an experimental gene therapy regimen last September. But his condition began to deteriorate within hours of receiving the first treatment, an intravenous infusion of a preparation containing the normal gene he lacked, encased in an enfeebled virus called adenovirus; gene therapy uses the weakened virus as a kind of shuttle to get the genes to the right tissues in the patient. (Fully competent adenoviruses cause a number of illnesses, including conjunctivitis and colds.) Gelsinger died four days later of multiple organ failure.

In spite of more than 20 man-years of intensive study in recent months by the researchers at the University of Pennsylvania and Children's National Medical Center in Washington, D.C., who carried out the trial, the exact cause of Gelsinger's multiorgan deterioration remains unknown. It is even possible, though most unlikely, that like my experience two decades ago, the death was unrelated to the therapy.

The response of the Food and Drug Administration (FDA), which oversees the clinical testing of all new medicines, to this unforeseen, tragic event has been precipitous and inappropriate. Before and during a three-day conference convened by the FDA outside Washington, D.C.--and without knowing the cause of the problem--agency officials publicly blasted the University of Pennsylvania researchers. They accused them of various kinds of misconduct--having admitted Gelsinger into the trial even though he did not meet eligibility requirements, having failed to immediately report information about two other patients who (long before the death) had experienced serious side effects, and having omitted information in the patient consent form about the death of monkeys that had received a similar but much higher-dose treatment. The first of these accusations is untrue. The second was misleading, in that although the toxicity in other patients had not been reported immediately after it occurred, the FDA had had that information long before the Gelsinger incident. The third was well within the usual, acceptable standards of clinical research: The results of animal studies, especially those that use a much higher dose than would be administered to humans, is seldom mentioned in the patient consent form; and the fact that 17 OTC-deficient patients had been treated in the University of Pennsylvania trial before Gelsinger without unexpected or irreversible problems also argues against the importance of the monkey data. (Moreover, the consent form had received the required approval from the hospital's institutional review board [IRB], whose sole responsibility is to protect the rights of human subjects in clinical research.)

The FDA insisted on public self- flagellation by the gene therapy researchers. On the first day of the meeting, James Wilson, the lead researcher for the study, maintained that he was "fully comfortable with the clinical decision" to treat Gelsinger. Thereafter, the FDA's senior gene therapy regulator demanded Wilson and his colleagues perform a public mea culpa--"or else." The next day saw a remarkable transformation. "I really regret" the oversights, said a contrite Wilson. "We could have done better ... and we apologize," said one of his collaborators. "We now realize we should have contacted FDA," said another.

Subsequently, the FDA went public with another round of lurid accusations of misconduct, but these amounted to nothing more than procedural, paperwork deficiencies and two instances of late reporting of patients' liver enzyme abnormalities.

Certainly, concern by FDA officials about Gelsinger's death is warranted. But what action is appropriate? Before all the data are in, certainly not a rush to judgment. Certainly not public humiliation--a trial by press conference and leaks to the media--of one of the most expert and reputable gene therapy teams in the world. And certainly not the FDA's impulsive tightening of manufacturing and quality control requirements for academic researchers--which are traditionally (and appropriately) more relaxed than for drug companies, or the FDA's sudden freeze on all eight University of Pennsylvania gene-therapy protocols (not all of which even involve adenovirus).

Worse still, the FDA halted two gene therapy experiments being conducted by drug company Schering-Plough that used adenovirus--one for the treatment of liver cancer, the other for colorectal cancer that has metastasized to the liver.

The FDA's actions can hardly be seen as anything but political, in view of the testimony at the conference of two clinical researchers from the University of California, San Francisco, and Cornell University that in their respective trials with similar adenovirus preparations, in a total of 180 patients they had each seen one instance of serious toxicity and no deaths.

The FDA's harassment of the University of Pennsylvania's exemplary gene therapy program will markedly inhibit the pace of clinical studies and even the willingness of academics to undertake gene therapy research. One eminent clinician speculated that FDA has already set the field back five years.

The FDA's show of toughness may have been an attempt to deflect attention from its own culpability. If there was an identifiable mistake, it was in the choice of patients for these first attempts at gene therapy for OTC deficiency. Rather than stable adult patients, it would have been more prudent to treat OTC-deficient babies who were comatose and had a dire prognosis. That was, in fact, the original intention of the OTC researchers, but they were dissuaded by their bioethicist, Arthur Caplan, who felt that parents of dying infants are "coerced by the disease of their child" and are, therefore, incapable of giving informed consent. In other words, the protocol treated and placed at risk a group that did not need the therapy, because the patients who might have benefited from it could not give genuine informed consent and were declared ineligible. FDA officials or the hospital's IRB could and should have reversed that decision.

The realities of clinical testing are sometimes harsh. One patient described the process as being "like willing your body to science while you're still alive," and serious diseases sometimes require aggressive therapies. Also, while a treatment is still unproven and unfamiliar--the stage at which gene therapy is now--therapeutic success may be elusive. When I was a medical student in the 1970s, for example, bone marrow transplantation was highly experimental. It was performed at only a handful of medical centers in the United States, and success rates were abysmal. But clinical research has refined the technique and identified diseases for which the technique is useful: In a genetically determined disease of red blood cells called thalassemia major, for example, more than 80 percent of the patients are now cured.

Regulators' decisions are difficult. They affect human lives and the pace and productivity of medical research. Often they need to be made on the basis of incomplete data, under time pressure, and in the glare of media attention. The challenge is to assure patients' safety, but with minimum interference in the conduct of individual trials and of medical research overall. In this case, the FDA has thrown its weight around unnecessarily and prematurely. The result has been the intimidation of gene therapy researchers throughout the country, and palpable costs to future generations of Jesse Gelsingers.