The Gene
“Number one,” Anderson said, pointing elatedly at Ashi as he wheeled her down the hallway after the transfusion had been completed. A few of his colleagues at the NIH were waiting outside the door to witness the coming of the first human to be transfused with gene-modified cells, but the crowd thinned quickly and the scientists vanished back to their labs. “It’s like people say in downtown Manhattan,” Anderson groused. “Jesus Christ himself could walk by and nobody would notice.” The next day, Ashi’s family returned home to Ohio.
Did Anderson’s gene-therapy experiment work? We do not know—and perhaps we will never know. Anderson’s protocol was designed as a proof of principle for safety—i.e., could retrovirus-infected T cells be safely delivered into human bodies? It was not designed to test efficacy: Would this protocol cure ADA deficiency, even temporarily? Ashi DeSilva and Cynthia Cutshall, the first two patients on the study, received the gene-modified T cells, but they were allowed continued treatment with PEG-ADA, the artificial enzyme. Any effect of the gene therapy was thus confounded by that medicine.
Nonetheless, both DeSilva’s and Cutshall’s parents were convinced that the treatment worked. “It’s not a big improvement,” Cynthia Cutshall’s mother admitted. “But to give you an example, she just got over one cold. Usually her colds end up in pneumonia. This one didn’t. . . . That’s a breakthrough for her.” Ashi’s father, Raja DeSilva, concurred: “With PEG, we’d seen a tremendous improvement. But even with [PEG-ADA], she had runny noses and a constant cold, was on antibiotics all the time. But by the second infusion of genes, in December, it began to change. We noticed because we were not using up so many boxes of tissues.”
Despite Anderson’s enthusiasm, and the anecdotal evidence from the families, many proponents of gene therapy, including Mulligan, were far from convinced that Anderson’s trial had amounted to anything more than a publicity stunt. Mulligan, the most voluble critic of the trial from the very first, was particularly incensed by the claims of success when the data was insufficient. If the most ambitious gene-therapy trial attempted in humans was going to be measured in the frequency of runny noses and boxes of Kleenex, then it would be an embarrassment for the field. “It’s a sham,” Mulligan told a journalist when asked about the protocol. To test whether targeted genetic alterations could be introduced into human cells, and whether these genes would confer normal function safely and effectively, he proposed a careful, uncontaminated trial—“clean, chaste gene therapy,” as he called it.
But by then, the ambitions of gene therapists had been frothed to such frenzy that “clean, chaste,” careful experiments had become virtually impossible to perform. Following the reports of the T cell trials at the NIH, gene therapists envisaged novel cures for genetic diseases such as cystic fibrosis and Huntington’s disease. Since genes could be delivered into virtually any cell, any cellular disease was a candidate for gene therapy: heart disease, mental illness, cancer. As the field readied itself to sprint forward, voices such as Mulligan’s urged caution and restraint, but they were brushed aside. The enthusiasm would come at a steep price: it would bring the field of gene therapy, and human genetics, to the brink of disaster, and to the lowest, bleakest point in its scientific history.
On September 9, 1999, almost exactly nine years after Ashi DeSilva had been treated with genetically modified white blood cells, a boy named Jesse Gelsinger flew to Philadelphia to enroll in another gene-therapy trial. Gelsinger was eighteen years old. A motorcycling and wrestling enthusiast, with an easy, carefree manner, Gelsinger, like Ashi DeSilva and Cynthia Cutshall, was also born with a mutation in a single gene involved in metabolism. In Gelsinger’s case, the gene was called ornithine transcarbamylase, or OTC, which encodes an enzyme synthesized in the liver. OTC, the enzyme, performs a critical step in the breakdown of proteins. In the absence of the enzyme, ammonia, a by-product of protein metabolism, accumulates in the body. Ammonia—the chemical found in cleaning fluid—damages blood vessels and cells, diffuses past the blood-brain barrier, and ultimately results in the slow poisoning of neurons in the brain. Most patients with mutations in OTC do not survive childhood. Even with strictly protein-free diets, they are poisoned by the breakdown of their own cells as they grow.
Among children born with an unfortunate disease, Gelsinger might have counted himself as especially fortunate, for his variant of OTC deficiency was mild. The mutation in his gene had not come from his father or his mother, but had occurred spontaneously in one of his cells in utero, probably when he was still a young embryo. Genetically, Gelsinger was a rare phenomenon—a human chimera, a patchwork cellular quilt, with some cells lacking functional OTC, and some with a working gene. Still, his ability to metabolize proteins was severely compromised. Gelsinger lived on a carefully calibrated diet—every calorie and portion weighed, measured, and accounted for—and took thirty-two pills a day to keep his ammonia level in check. Despite such extreme cautionary measures, Gelsinger had still suffered several life-threatening episodes. At four, he had joyfully eaten a peanut butter sandwich that had precipitated a coma.
In 1993, when Gelsinger was twelve years old, two pediatricians in Pennsylvania, Mark Batshaw and James Wilson, began to experiment with gene therapy to cure children with OTC deficiencies. A former college-level football player, Wilson was a risk taker fascinated by ambitious human experiments. He had formed a gene-therapy company, named Genova, and an Institute for Human Gene Therapy at the University of Pennsylvania. Both Wilson and Batshaw were intrigued by OTC deficiency. As with ADA deficiency, that OTC is caused by the dysfunction of a single gene made the illness an ideal test case for gene therapy. But the form of gene therapy that Wilson and Batshaw envisioned was vastly more radical: rather than extracting cells, genetically modifying them, and injecting them back into children (à la Anderson and Blaese), Batshaw and Wilson imagined inserting the corrected gene directly back into the body via a virus. This would not be gene-therapy lite: they would create a virus carrying the OTC gene and deliver the virus into the liver through the bloodstream, leaving the virus to infect cells in situ.
The virus-infected liver cells would start synthesizing the OTC enzyme, Batshaw and Wilson reasoned, and thus correct the enzyme deficiency. The telltale sign would be a reduction of ammonia in the blood. “It wasn’t that subtle,” Wilson recalls. To deliver the gene, Wilson and Batshaw chose adenovirus, a virus that typically causes a common cold but is not associated with any severe disease. It seemed like a safe, reasonable choice—the blandest of viruses used as the vehicle for one of the boldest human genetic experiments of the decade.
In the summer of 1993, Batshaw and Wilson began to inject the modified adenovirus into mice and monkeys. The mouse experiments worked as predicted: the virus reached the liver cells, disgorged the gene, and transformed the cells into microscopic factories for the functional OTC enzyme. But the monkey experiments were more complicated. At higher doses of the virus, an occasional monkey raised a brisk immune response to the virus, resulting in inflammation and liver failure. One monkey hemorrhaged to death. Wilson and Batshaw modified the virus, shaving off many of the viral genes that might elicit immunity, to make it a safer gene-delivery vehicle. They also reduced the potential human dose by seventeenfold to doubly ensure the safety of the virus. In 1997, they applied to the Recombinant DNA Advisory Committee, the gatekeeper for all gene-therapy experiments, for approval for a human trial. The RAC was resistant at first, but it too had changed: in the decade between the ADA trial and Wilson’s, the once-fierce guardian of recombinant DNA had turned into an enthusiastic cheerleader of human gene therapy. The fizz of enthusiasm had even leached beyond the committee. Asked by the RAC to comment on Wilson’s trial, bioethicists argued that treating children with full-blown OTC deficiency might result in “coercion”: What parent wouldn’t want to try a breakthrough therapy that might work on a dying child? Instead, ethicists recommended a trial on normal volunteers and patients with mild variants of OTC, such as Jesse Gelsinger.
> In Arizona, Gelsinger, meanwhile, was chafing against the elaborate restrictions on his diet and medications (“All teenagers rebel,” Gelsinger’s father, Paul, told me, but teenage rebellion might feel particularly acute when it involves “a hamburger and a glass of milk”). In the summer of 1998, when he was seventeen, Gelsinger learned of the OTC trial at the University of Pennsylvania. Gelsinger was gripped by the thought of gene therapy. He wanted a respite from the grinding routine of his life. “But what got him even more excited,” his father recalled, “was the idea that he was doing it for the babies. How do you say no to that?”
Gelsinger could hardly wait to sign on. In June 1999, he contacted the Pennsylvania team through his local doctors to enroll in the trial. That month, Paul and Jesse Gelsinger flew to Philadelphia to meet Wilson and Batshaw. Jesse and Paul were both impressed. The trial struck Paul Gelsinger as a “beautiful, beautiful thing.” They visited the hospital and then wandered through the city in a haze of excitement and anticipation. Jesse stopped in front of the Rocky Balboa bronze outside the Spectrum Arena. Paul snapped a picture of his son, his arms raised in a boxer’s victory stance.
On September 9, Jesse returned to Philadelphia with a duffel bag filled with clothes, books, and wrestling videos to start the trial at University Hospital. Jesse would stay with his uncle and cousins in the city and admit himself to the hospital on the appointed morning. The procedure was described as so quick and painless that Paul planned to pick up his son one week after the therapy had been completed to bring him back home on a commercial flight.
On the morning of September 13, the day chosen for the viral injection, Gelsinger’s ammonia level was found to be hovering around seventy micromoles per liter—twice the normal level and at the upper edge of the cutoff value for the trial. The nurses brought news of the abnormal lab to Wilson and Batshaw. The protocol, meanwhile, was in full swing. The operating rooms were on standby. The viral liquid had been thawed and sat glistening in its plastic pouch. Wilson and Batshaw debated Gelsinger’s eligibility, but decided that it was clinically safe to continue; the previous seventeen patients had, after all, tolerated the injection. At about 9:30 a.m., Gelsinger was wheeled down to the interventional radiology suite. He was sedated, and two large catheters were snaked through his legs to reach an artery close to the liver. Around 11:00 a.m., a surgeon drew about thirty milliliters from a bag clouded with the concentrated adenovirus and injected the puff of virus into Gelsinger’s artery. Hundreds of millions of invisible infectious particles carrying the OTC gene streamed into the liver. By noon, the procedure was done.
The afternoon was uneventful. That evening, back in his hospital room, Gelsinger spiked a fever to 104 degrees. His face was flushed. Wilson and Batshaw did not make much of the symptoms. The other patients had also experienced transient fevers. Jesse called Paul in Arizona on the phone and said, “I love you,” before hanging up and drawing up his covers in bed. He slept fitfully through the night.
The next morning, a nurse noted that Jesse’s eyeballs had turned the palest shade of yellow. A test confirmed that bilirubin, a product made in the liver and also stored in red blood cells, was spilling into his blood. The elevated bilirubin meant one of two things: either the liver was being injured or blood cells were being damaged. Both were ominous signs. In any other human, the small bump in cellular breakdown or liver failure might have been shrugged off. But in a patient with OTC deficiency, the combination of these two injuries might spark a perfect storm: the extra protein leaching out from the blood cells would not be metabolized, and the damaged liver, deficient in protein metabolism even in the best of times, would be even less capable of processing the excess protein load. The body would intoxicate itself on its own poisons. By noon, Gelsinger’s ammonia level had climbed to a staggering 393 micromoles per liter—nearly ten times the normal level. Paul Gelsinger and Mark Batshaw were alerted. James Wilson heard the news from the surgeon who had inserted the catheter and injected the virus. Paul booked a red-eye to Pennsylvania, while a team of doctors swooped into the ICU to begin dialysis to avert a coma.
At eight o’clock the next morning, when Paul Gelsinger reached the hospital, Jesse was hyperventilating and confused. His kidneys were failing. The ICU team sedated him to try to use a mechanical ventilator to stabilize his breathing. Late that night, his lungs began to stiffen and collapse, filling up with fluids from the inflammatory response. The ventilator faltered, unable to push enough oxygen in, and so Jesse was hooked up to a device to force oxygen directly into the blood. His brain function was also deteriorating. A neurologist was called to examine him and noted Jesse’s downcasting eyes—a sign of damage to the brain.
The next morning, Hurricane Floyd struck the East Coast, battering the shores of Pennsylvania and Maryland with shrieking winds and torrents of rain. Batshaw was stuck on a train getting to the hospital. He ran down the last minutes of his cell phone’s battery talking to the nurses and doctors, then sat in the pitch darkness, stewing with anxiety. By the late afternoon, Jesse’s condition worsened again. His kidneys shut down. His coma deepened. Stranded in his hotel room, with no taxi in sight, Paul Gelsinger walked a mile and a half through the whistling storm to the hospital to see Jesse in the ICU. He found his son unrecognizable—comatose, swollen, bruised, yellowed with jaundice, with dozens of lines and catheters crisscrossing his body. The ventilator puffed ineffectually against his inflamed lungs with the flat, dull sound of wind slapping water. The room buzzed and beeped with hundreds of instruments recording the slow decline of a boy in desperate physiological distress.
On the morning of Friday, September 17, on the fourth day after the gene delivery, Jesse was found to be brain-dead. Paul Gelsinger decided to withdraw life support. The chaplain came into the hospital room and put his hand on Jesse’s head, anointing it with oil, and read the Lord’s Prayer. The machines were shut off, one by one. The room fell into silence, except for Jesse’s deep, agonal breaths. At 2:30 p.m., Jesse’s heart stopped. He was officially pronounced dead.
“How could such a beautiful thing go so, so wrong?” When I spoke to Paul Gelsinger in the summer of 2014, he was still searching for an answer. A few weeks earlier, I had e-mailed Paul about my interest in Jesse’s story. Gelsinger spoke to me on the phone, then agreed to meet me after my talk on the future of genetics and cancer at an open forum in Scottsdale, Arizona. While I stood in the lobby of the auditorium at the end of the talk, a man in a Hawaiian shirt with Jesse’s round, open face—a face that I remembered vividly from pictures on the Web—pushed his way through the crowd and extended his hand.
In the aftermath of Jesse’s death, Paul has become a one-man crusader against the overreach of clinical experimentation. He is not against medicine or innovation. He believes in the future of gene therapy. But he is suspicious of the hyperbaric atmosphere of enthusiasm and delusion that ultimately resulted in his son’s death. The crowd thinned, and Paul turned around to leave. An acknowledgment passed between us: a doctor writing about the future of medicine and genetics, and a man whose story had been etched into its past. There was the infinite horizon of grief in his voice. “They didn’t have a handle on it yet,” he said. “They tried it too quickly. They tried it without doing it right. They rushed this thing. They really rushed it.”
The autopsy of an experiment gone “so, so wrong” began in earnest in October 1999 when the University of Pennsylvania launched an investigation into the OTC trial. By late October, an investigative journalist from the Washington Post had picked up the news of Gelsinger’s death, and a tide of furor broke loose. In November, the US Senate, the House of Representatives, and the district attorney of Pennsylvania held independent hearings on Jesse Gelsinger’s death. By December, the RAC and the FDA had launched an investigation of the University of Pennsylvania. Gelsinger’s medical records, the pretrial animal experiments, consent forms, procedure notes, lab tests, and the records of all the other patients on the gene-therapy trial were pulled up from University
Hospital’s basement, and federal regulators plowed their way through the mounds of paper trying to exhume the cause of the boy’s death.
The initial analysis revealed a damning pattern of incompetence, blunders, and neglect, compounded by fundamental gaps in knowledge. First, the animal experiments performed to establish the safety of the adenovirus had been performed hastily. One monkey inoculated with the highest doses of the virus had died, and while this death had been reported to the NIH and the dose reduced for human patients, no mention of the death was found in the forms given to the Gelsinger family. “Nothing about the consent forms,” Paul Gelsinger recalled, “hinted clearly at the harm that the treatment could cause. It was portrayed like a perfect gamble—all upside and no downside.” Second, even the human patients treated before Jesse had experienced side effects, some striking enough to have halted the trial or triggered reevaluations of the protocol. Fevers, inflammatory responses, and early signs of liver failure had been recorded, but these too had been underreported or ignored. That Wilson had a financial stake in the biotechnology company that stood to benefit from this gene-therapy experiment only further deepened the suspicion that the trial had been put together with inappropriate incentives.
The pattern of neglect was so damning that it nearly obscured the most important scientific lessons of the trial. Even if the doctors admitted that they had been negligent and impatient, Gelsinger’s death was still a mystery: no could explain why Jesse Gelsinger had suffered such a severe immune reaction to the virus, while the seventeen other patients had not. Clearly, the adenoviral vector—even the “third-generation” virus shorn of some of its immunogenic proteins—was capable of inciting a severe idiosyncratic response in some patients. The autopsy of Gelsinger’s body showed that his physiology had become overwhelmed by this immune response. Notably, when his blood was analyzed, antibodies highly reactive to the virus were found, dating from even before the viral injection. Gelsinger’s hyperactive immune response was likely related to prior exposure to a similar strain of adenovirus, possibly from a common cold. Exposures to pathogens are well-known to incite antibodies that remain in circulation for decades (this, after all, is how most vaccines work). In Jesse’s case, this prior exposure had likely triggered a hyperactive immune response that had spiraled out of control for unknown reasons. Ironically, perhaps the choice of a “harmless,” common virus as the initial vector for gene therapy had turned out to be the trial’s key failing.