In 1949, a team in Linus Pauling’s lab compared hemoglobin from sickle cell patients to the healthy kind and named something no one had described before: a disease caused by a single defective molecule. One wrong letter in the gene for beta-globin, one amino acid swapped, and a red cell that should slip through a capillary hardens into a crescent and jams it. They called it the first “molecular disease,” four years before anyone had even solved the structure of DNA. The molecular fix took another seventy-four years. When it finally arrived it came with a $2.2 million price tag, a round of chemotherapy that can leave a patient infertile, and a patient count you could seat in a lecture hall. Two years after approval, 64 people have been infused.
The proof it works has a name: Victoria Gray. On July 2, 2019, in a Nashville hospital, doctors reinfused her own bone marrow cells through the same catheter that used to carry her transfusions, except these cells had been pulled out, edited with CRISPR, and put back. She has not had a pain crisis since. Seven years, no crisis, from a woman who once described the pain as being struck by lightning and hit by a freight train at the same time. Her cells became the clinical foundation for Casgevy, the first CRISPR medicine the FDA ever approved, cleared in December 2023.
So it is worth noticing that this week the field’s own flagship journal, The Lancet, framed Gray’s story not as a cure but under an editorial titled “Gene-editing therapies: delivering safely for patients.” Not curing. Delivering safely. When the establishment reaches for that phrase, the ribbon-cutting is over, and what it papers over is the conditioning chemo, the price, and the empty appointment slots.
A bone marrow transplant from a matched sibling had long been able to cure sickle cell, but only a minority of patients ever find a donor who matches, so for most people the disease had no fix that did not depend on someone else’s marrow. That is what Casgevy was supposed to change, because it is built from a patient’s own cells. And the mechanism is clever enough to explain both the promise and the fine print. Casgevy does not repair the sickle mutation. It uses CRISPR to knock out a switch, an enhancer of the gene BCL11A, that normally silences fetal hemoglobin after birth. Switch it off and the body starts making fetal hemoglobin again, the kind that does not sickle. The edit is done to cells outside the body, and no virus ferries the change into the genome. That matters more than it sounds, because the ghost that has haunted gene therapy for twenty years is the delivery vehicle.
Here is what that ghost looks like when it materializes. Older gene therapies use a virus to splice a working gene into a cell’s DNA, and the virus does not pick its landing spot carefully. Land near the wrong gene and you can switch on cancer. That is not hypothetical. FiercePharma, reporting findings from the New England Journal of Medicine, described seven boys who developed blood cancers, acute myeloid leukemia and myelodysplastic syndrome, after receiving bluebird bio’s Skysona for a rare brain disease, and the FDA restricted the therapy to patients with no other option. Around the same time the agency forced a class-wide boxed warning onto every approved CAR-T cell therapy after T-cell cancers turned up in treated patients. Neither of those is Casgevy, and its viral-free design is built to sidestep exactly this failure mode. But this corner of medicine already has a cancer record, and “delivering safely” is the phrase you reach for once you have read it.
Casgevy’s design does dodge that specific danger. Read the fine print on how the edited cells get back in, though. Before reinfusion the patient’s own marrow has to be wiped out to make room, and the tool for that is busulfan, a myeloablative chemotherapy. Busulfan is gonadotoxic; it can leave patients infertile, which turns a cure into a cruel choice for a young adult who wants children. And it casts its own long shadow, because the FDA’s clinical review notes busulfan is associated with myeloid malignancies and solid tumors, in a sickle cell population already at elevated risk for them. The CRISPR edit may be clean. The chemotherapy that clears the runway for it is the same brute-force conditioning oncology has leaned on for decades, and it swaps one lifelong risk for a fresh set. Seven years of follow-up on Victoria Gray is a strong signal. It is not thirty years, and safety signals in this field have a habit of surfacing late.
Then there is the wall almost no one clears. Casgevy lists at $2.2 million, and once you add the months of conditioning, the apheresis, the hospital stay and the monitoring, payers are looking at something closer to $3 million per patient. The efficacy behind that number is hard to argue with: in the pivotal trial, 96.7 percent of evaluable patients went at least 12 consecutive months with no vaso-occlusive crisis, and every one of them stayed out of the hospital over that stretch. Now look at how many people have actually gotten it. As of Vertex’s own full-year 2025 report, 147 patients had started the process by having cells collected, and just 64 had been infused, across both approved conditions, sickle cell and transfusion-dependent beta thalassemia. A one-time cure that reaches 64 people in two years is a press release, not a public-health event. And in July 2026 Vertex won an expanded label covering children as young as two, widening the market while the price wall stays exactly where it was. The beneficiaries are named and specific: Vertex Pharmaceuticals and CRISPR Therapeutics, the companies whose science this is and whose stock this moves.
None of this makes the science fake. It makes the marketing incomplete. The honest version of the Victoria Gray story is that a disease medicine sidelined for generations finally has a one-time treatment that works, and that treatment costs three million dollars, runs through a chemotherapy that can cause infertility and cancer, and has so far reached a rounding error of the roughly 100,000 Americans who live with sickle cell disease, most of them Black, most underserved for decades by the research money that skipped this condition.
What to watch is not another triumphant anniversary. Casgevy’s approval came with a mandated long-term study, CLIMB-131, that follows patients for up to 15 years, and the numbers that will actually answer the “delivering safely” question are the ones that come due years from now: whether CRISPR’s off-target edits stay silent, whether busulfan-linked malignancies show up in a population already prone to them, and whether the FDA’s post-market monitoring catches a signal before a patient does. The next data point of consequence belongs to the regulators and the manufacturer, not the launch team. When it lands, it will tell us whether the first molecular disease finally got a safe cure, or an expensive down payment on one.
Sources
- The Lancet – Editorial, “Gene-editing therapies: delivering safely for patients” (2026)
- Vertex / CRISPR Therapeutics – FDA approval of Casgevy for sickle cell disease, with pivotal-trial efficacy (Dec 2023)
- BioSpace – Sickle cell gene therapies Casgevy and Lyfgenia still lacking traction two years in (uptake: 147 collected, 64 infused)
- FDA – Approves first gene therapies to treat patients with sickle cell disease (~100,000 Americans, most Black)
- Journal of Medical Economics – Cost-effectiveness of exa-cel in sickle cell disease, US (2026)
- FDA / CADTH Clinical Review – Exagamglogene autotemcel (Casgevy): busulfan conditioning and malignancy risk
- STAT – New gene therapies confront sickle cell patients with a choice: a cure or fertility (2023)
- Vertex – FDA approval of expanded Casgevy use for ages 2 and older (July 2026)
- FiercePharma – FDA restricts bluebird bio’s Skysona after blood cancer reports
- CancerNetwork – Communicating CAR-T secondary malignancy risks and the FDA boxed warning
- Innovative Genomics Institute – Meet Victoria Gray, the first CRISPR sickle cell patient