'Crispr babies' versus 'bubble boy'

NEW YORK – The news that scientists may have finally used gene therapy to cure the “bubble boy” immune disorder, SCID-X1, came as a surprise not because it happened so fast, but because it took so long that it had begun to seem impossible.

Scientists were talking about revolutionizing medicine with gene therapy back in the 1980s, and the first child with a different form of the disease, called SCID-ADA, was given gene therapy in 1992. By 2000, doctors were treating the first kids with SCID-X1. But there were problems. Some of them developed leukemia.

There’s a belief that became pervasive in the 1990s that medicine is moving so fast that ethics can’t keep up. Science stories in the news would refer to “Brave New World” or Frankenstein’s monster.

But now that we’re living in that long-imagined future, it looks like science isn’t keeping pace with the hype, which over the years has included promises of cures tied to the human genome project, the expectation that gene therapy would be commonplace and even the weird belief that cloning would replace sex as the preferred method of human reproduction.

Things haven’t quite panned out that way. To better understand why, I talked with Jonathan Kimmelman, a medical ethicist at McGill University in Canada and an expert in human experimentation. He said that despite all the hype, medical technology doesn’t leap forward with every new idea the way other kinds of tech can. The ethics of human research slows things down.

Not that medical ethics is easy. The challenge for ethicists, and for society, is to judge research decisions based on what the scientists knew at the time, not the outcome. Unethical researchers might get lucky, and good ones might get very unlucky. By those standards, he said, the researchers who accidentally caused some SCID patients to get leukemia were still taking an ethically acceptable risk, given the scale of the potential benefits, but researchers at the University of Pennsylvania whose experiment killed an 18-year-old subject were not.

In that 1999 case, Jesse Gelsinger died from an experimental gene therapy aimed at curing a different genetic disorder — one less life-threatening than SCID. His immune system mounted a deadly reaction to the virus used to insert the gene into his cells — a deactivated cold virus called an adenovirus.

In retrospect, there were problems with that trial — financial conflicts of interest, worrisome signs in animal studies that were ignored and some irregularities in the way the human subjects were treated, said Kimmelman, who has written a book about the case. After the death, lots of people claimed to have seen these problems, but, sadly, none of them took the initiative to blow the whistle.

SCID gene therapy trials progressed more carefully, even though the disease was claiming lives with each passing year. A defective gene prevents the bone marrow from creating working immune cells, so kids with the disease have essentially no immune system. This came to public attention in the 1970s, when doctors found a way to keep the famous “bubble boy,” David Vetter, alive until the age of 12 by sealing him into a sterile plastic enclosure.

Gene therapy seemed like a promising solution. Doctors knew which genes were damaged, and they knew that they need to get working copies into the patient’s bone marrow.

But there’s another layer of precision needed: It can matter where newly introduced genes get incorporated into the person’s chromosomes. Viruses can’t be programmed to put them in any specific place. Scientists knew, said Kimmelman, that getting the working versions of these genes into the wrong places might trigger leukemia. They thought it was very unlikely, but realized only after the fact that the viruses tended to preferentially place the genes in locations where they increased risk. In 2002, the SCID-X1 trial was stopped after the disease affected four children.

Over the years, scientists have examined other, safer vectors and, counterintuitively, found that for SCID-X1 their best bet was a deactivated human immunodeficiency virus (HIV). These latest experiments, done in Saint Jude’s Children’s Research Hospital in Memphis and published in the New England Journal of Medicine, took steps to prevent leukemia. It’s still early, but the researchers say that so far the results look promising.

A similar standard should apply to the claimed gene-edited babies allegedly born in China late last year. The ethics have to be judged on the risks that were taken at the time, not the outcome, which may never be known given the secrecy surrounding the research. The babies — twin girls — were essentially human guinea pigs. The only disease involved was the father’s HIV-positive status, but there are safe ways to make sure a father’s virus isn’t passed to his offspring.

The risks of this are still relatively unknown, and the fact that the SCID researchers misjudged the risk of giving their subjects leukemia should serve as a warning. Once again, in the case of the “Crispr babies,” the ethical principles were there, but they were broken — maybe by a rogue scientist but possibly by one whose experiments were known and funded by the Chinese government. Kimmelman points out that people have been debating the ethics of genetic engineering on unborn children since the 1970s, soon after the debut of genetic engineering.

In the medical community, there was almost universal agreement that the experiment was unethical because the twin girls were subject to unnecessary risk. The main problem with genetic technology isn’t the need to prevent the birth of Frankenstein’s monster, but to follow the ethical principles that Hippocrates wrote about more than 2,000 years ago. The needs of patients have to come first, even if it slows down the pace of progress.

Science writer Faye Flam is a Bloomberg Opinion columnist. She has written for The Economist, The New York Times, The Washington Post, Psychology Today, Science and other publications. She has a degree in geophysics from the California Institute of Technology.