CRISPR at the Clinic: How Gene Editing Moved From Promise to Patient in 2025–2026

The history of gene therapy is a history of near-misses. Treatments that looked transformative in mouse models failed in humans. Clinical trials were halted by immune reactions. Patients died, regulatory agencies tightened scrutiny, and the field spent years rebuilding trust. CRISPR was supposed to be different — more precise, more versatile, more controllable than anything that came before it. The question, through the years of preclinical excitement and early trials, was always whether the technology would hold up when it met real patients with real disease.

The evidence now accumulating suggests it does. Not uniformly, not without complications, and not yet at a scale that reaches most patients who could benefit. But the clinical data from 2025 and early 2026 marks a transition from "we believe this can work" to "we have proved it works, in humans, in multiple disease settings."

Casgevy: the commercial proof point

Casgevy — developed jointly by Vertex Pharmaceuticals and CRISPR Therapeutics — was approved by the FDA in December 2023 for sickle cell disease and transfusion-dependent beta-thalassemia. It is the world's first approved CRISPR medicine, and the results presented at the American Society of Hematology meeting in December 2025 have been striking.

In 45 evaluable sickle cell patients, 100% achieved freedom from vaso-occlusive crises — the agonising pain episodes that define life with the disease — for at least twelve consecutive months, with a mean VOC-free duration of over 35 months. In beta-thalassemia patients, 98.2% achieved transfusion independence for at least twelve months, with a mean duration of over 41 months. Preliminary data from children aged 5 to 11 showed the same pattern: complete VOC freedom, full transfusion independence.

Casgevy generated $115.8 million in 2025 revenue — modest by blockbuster standards, but significant as validation that the treatment pipeline can function commercially. About 64 patients received infusions across the full year, with uptake accelerating: 30 of those came in the final quarter alone. Approximately 90% of eligible US patients now have insurance coverage. Vertex projects revenue to nearly triple in 2026.

The numbers also illuminate the challenge. At a list price of $2.2 million per patient, and requiring months of preparation including myeloablative chemotherapy conditioning at one of approximately 65 specialised treatment centres worldwide, Casgevy is not yet a therapy that scales to the global sickle cell burden. Over 90% of people with sickle cell disease worldwide live in sub-Saharan Africa. None of the current treatment infrastructure exists there.

The first in vivo Phase 3 success

Casgevy edits cells outside the body: stem cells are harvested, modified in a laboratory, and reinfused. The next frontier — editing genes directly inside living patients with a single infusion — cleared a critical milestone in April 2026.

Intellia Therapeutics reported Phase 3 results for lonvoguran ziclumeran, its in vivo CRISPR therapy for hereditary angioedema. The drug, delivered via lipid nanoparticles that carry the editing machinery to liver cells, produced an 87% reduction in HAE attacks compared to placebo over six months, with 62% of patients entirely attack-free. No serious adverse events were reported. Intellia has begun rolling BLA submission to the FDA and is targeting a US launch in the first half of 2027.

This is the first in vivo CRISPR therapy to succeed in a Phase 3 trial — a category of milestone the field has been working toward for years. The lipid nanoparticle delivery approach, which encases the CRISPR components in fatty particles that fuse with liver cells after intravenous infusion, had already been validated by the mRNA COVID vaccines. What Intellia demonstrated is that the same delivery mechanism can carry gene-editing machinery with sufficient precision and safety to meet regulatory standards for a chronic disease treatment.

Personalised medicine at unprecedented speed

The most remarkable individual story from 2025 came in May, when a team at Children's Hospital of Philadelphia reported treating an infant born with carbamoyl phosphate synthetase I deficiency — a rare metabolic disorder that prevents the liver from processing protein and is lethal without continuous medical management. The team designed, received FDA clearance for, manufactured, and delivered a personalised in vivo CRISPR therapy for this specific patient within six months of diagnosis. The child showed improved protein tolerance and reduced dependence on medication.

No other medical technology has demonstrated the ability to design and deploy a patient-specific molecular intervention on that timeline. The CPS1 case is a single patient, not a clinical trial, and replicating this pipeline at scale would require regulatory frameworks and manufacturing infrastructure that do not yet exist. But it establishes a precedent: the machinery for on-demand, individual gene editing is real and has been used.

Base editing and prime editing enter the clinic

Classic CRISPR-Cas9 works by cutting both strands of DNA, relying on the cell's own repair mechanisms to make the desired change. Those repair mechanisms are imperfect, sometimes introducing unintended insertions or deletions. Two newer approaches — base editing and prime editing — sidestep this by making targeted chemical modifications to individual DNA letters without breaking the double strand.

Base editing entered clinical evidence in December 2025, when results from a UCL and Great Ormond Street Hospital trial of BE-CAR7 were published. The treatment uses base-edited T cells to target T-cell acute lymphoblastic leukemia, a blood cancer that is extremely difficult to treat because conventional CAR-T therapies cannot be used — T cells cannot target other T cells without destroying themselves. BE-CAR7 achieved 82% deep remission in patients, with 64% remaining disease-free at follow-up of up to three years.

Prime editing — the most precise of the three approaches, capable of making any of the twelve possible base-pair changes plus small insertions and deletions — reached its first human clinical proof in May 2025. Prime Medicine's PM359, targeting a rare immune disorder called chronic granulomatous disease, showed restoration of the defective protein in the treated patient. The results were published in the New England Journal of Medicine in December 2025. Prime editing in humans is no longer theoretical.

What remains genuinely hard

The honest accounting of where CRISPR stands in 2026 requires acknowledging what has not been solved. Delivery to tissues other than the liver remains largely unsolved for in vivo approaches. Lipid nanoparticles, after intravenous infusion, accumulate primarily in the liver — which makes them excellent for metabolic and cardiovascular diseases but inadequate for muscle disorders, neurological conditions, or cancers outside the bloodstream. Brain delivery is blocked by the blood-brain barrier; the most promising approaches remain preclinical.

Off-target editing effects — molecular edits occurring at unintended sites in the genome — are measurably lower with base and prime editing than with classic Cas9, but are not zero. Regulatory agencies now require genome-wide experimental assays for every new guide RNA sequence used in a clinical product. The standards are appropriate; meeting them adds time and cost to every new programme.

And cost remains the largest structural barrier to impact. At $2.2 million per patient, even a curative therapy reaches only a tiny fraction of those who need it. The clinical science is advancing faster than the economic and logistical infrastructure needed to deploy it equitably. That gap is the next problem the field needs to solve.