The side effect, a buildup of fluid in the brain, led to the death of one of the children and presents a grave setback for a class of personalized medicine.Weeks after Valeria Schenkel took an experimental drug named after her, the daily seizures that had afflicted her from birth became less frequent. But the drug caused fluid to build up in her brain, and a year later, she died at age 3.The drug was given to only one other child, and she experienced the same side effect and nearly died last year.The drug contained snippets of genetic material tailor-made to turn off the mutated gene causing the extremely rare form of epilepsy that these children were born with. A handful of researchers and nonprofit organizations have raised millions of dollars to make these treatments, known as antisense drugs, for at least 19 children and adults with severe diseases that are too rare to garner interest from pharmaceutical companies. The treatments have helped some of these patients, raising hopes that the personalized approach might one day save thousands of lives.But the brain side effect, known as hydrocephalus, reported on Sunday at the American Neurological Association meeting in Chicago, is a blow for the niche medical field that has made rapid progress over the past five years. Hydrocephalus happens when too much fluid fills cavities in the brain, increasing pressure on brain tissue and risking lethal damage if untreated.“I think it’s worth saying: No question that encountering hydrocephalus has been a setback, sobering and important,” said Dr. Timothy Yu, the neurologist and genetics researcher at Boston Children’s Hospital who developed the drug, known as valeriasen.But traditional drug companies, he added, are not helping patients with thousands of rare, untreatable and rapidly progressing diseases that cause death and severe disabilities. Personalized genetic treatments may be their only hope.“We have to learn as much as we can from each and every one, because they’re just so incredibly valuable in every sense,” Dr. Yu said.Scientists first imagined creating “antisense oligonucleotide” drugs — pieces of custom-made DNA or RNA designed to correct for genetic errors in cells — in the 1960s. But it took decades to make stable and effective versions of such drugs.Unlike older gene therapies, which replace defective genes, antisense drugs can act like chemical dampers, turning off the mutated genes so they cannot be used by cells.In 2016, the Food and Drug Administration approved the first successful antisense drug, nusinersen, to treat spinal muscular atrophy, a muscle wasting disease. The next year, Dr. Yu and his team made an antisense drug for Mila Makovec, a 6-year-old who had a rare, fatal form of the neurological illness Batten disease.The researchers made the drug, called milasen, in just 10 months, far faster than the decade or more it often takes to make traditional drugs. Before milasen, no one had ever made a drug for a single person, let alone so quickly.From left: Mario Schenkel; Leonard Kaczmarek, a researcher at Yale who collaborated on the valeriasen trial; Dr. Timothy Yu holding Valeria Schenkel; and Alexandra Schenkel.Mario and Alexandra SchenkelIn 2018, Mario and Alexandra Schenkel of Lucerne, Switzerland, asked Dr. Yu to help their daughter. Valeria was born in 2018 with an error in one of her genes, KCNT1, that caused a rare and devastating form of epilepsy. She had dozens of seizures every day, and could not sit up or speak. Half of children with her condition die by age 3.The Schenkels raised more than $1 million; started a foundation, the Valeria Association; and moved to Massachusetts to support Dr. Yu’s efforts to make an antisense drug to shut down Valeria’s flawed gene. In September 2020, valeriasen was injected into the fluid around her spinal cord.At first, Valeria improved. But 11 months after she started the drug, doctors detected hydrocephalus in her brain. Her family withdrew from the valeriasen trial, and she died on Sept. 5, 2021.“She was such a fighter,” Ms. Schenkel said. “She was the one who paved the way for KCNT1 research and all the children who will hopefully get the chance for a better quality of life.”One of them, Lucy Greenblott, is the only other patient who has taken valeriasen.Lucy started taking the drug on June 17, 2021, at age 2½. Her seizures dropped drastically; for more than a month, she had no or only a handful of seizures each day.But by early August, Lucy would have episodes of whimpering and odd limb movements, as if she was in pain. By late August, every muscle in her body went rigid, her eyes rolled up above her eyelids and she was crying out in agony.Doctors discovered that she had hydrocephalus, sometimes called “water on the brain,” and implanted a shunt, a device used to drain fluid from the brain.“The scariest thing was wondering if we hurt her worse,” said Lucy’s father, Seth Greenblott, a lawyer and business consultant from Hopkinton, N.H. “Did she have to suffer more than she already does?”Lucy Greenblott in November 2021, three months after she stopped taking the drug valeriasen.Seth GreenblottDr. Yu is investigating why valeriasen caused hydrocephalus in both girls, and whether the side effect could be prevented with adjusted doses of the drug.In 2021, the pharmaceutical company Roche said that three patients who received an experimental antisense drug in a Huntington’s disease clinical trial had hydrocephalus, raising the possibility that the problem could plague many applications of such therapies. Some patients on nusinersen have also developed hydrocephalus. But the fluid buildup is not uncommon in people with neurological diseases, and it can often be difficult to determine whether the complication was caused by a drug or the underlying illness.The new reports underscore the need for researchers to share data on experimental drugs that are tested in only one or a few people, doctors said.The F.D.A. must approve each clinical trial of an antisense drug or other type of personalized medicine, even if it is used in only a single patient. But the scientists running these trials are not required to share information with one another, meaning that they may not know about safety risks seen by other investigators.“This is a new field of medicine,” said Liza-Marie Johnson, a pediatric oncologist and bioethicist at St. Jude Children’s Research Hospital who was not involved in the antisense trials. “How do we share data and monitor so that we can learn from events and make this safer and potentially benefit future patients?”Other doctors said the setback in Dr. Yu’s trial should not prevent scientists from testing customized medicines.“There are other kids out there with fatal diseases,” said Dr. Arthur Krieg, a rheumatologist at the University of Massachusetts RNA Therapeutics Institute and a consultant to biotechnology companies, one of which makes antisense drugs. “How are we going to know if the treatment can help them unless we give it to them?”Valeria’s father agreed.“Neurologists shouldn’t be scared of being bold and trying these kinds of drugs and helping patients with rare diseases,” Mr. Schenkel said.As for Lucy Greenblott, since she stopped taking valeriasen in August 2021, she has had as many as 15 seizures per day. Her parents are considering putting her back on the drug.“We know our time with her is limited, and this may be our best hope to improve the quality and length of her life,” Mr. Greenblott said. “But we also know we could shorten her life and make it harder. I don’t know how anyone makes that decision.”