A baby named KJ was born with a disease that affects roughly 1 in 1.3 million newborns. His liver could not break down protein. Ammonia built up in his blood. His disease carries a 50 percent mortality rate in infancy; staying alive meant living in the hospital under constant monitoring. A liver transplant was the standard option. He was too young and too sick to receive one.
So a team at Children's Hospital of Philadelphia and Penn Medicine did something that had never been done. After identifying KJ's specific variant of CPS1 soon after his birth, they designed and manufactured a base editing therapy delivered via lipid nanoparticles to the liver to correct his faulty enzyme, all within six months. KJ received the first dose of his bespoke therapy in February 2025 between six and seven months of age. The treatment was administered safely, and he is now growing well and thriving.
That is not a clinical trial. That is a custom-built medicine, designed for one patient's exact mutation, manufactured and delivered in the time it takes most pharmaceutical companies to schedule a planning meeting. This is the future and it is happening now.
The debate about CRISPR in medicine has been framed wrong from the start. People ask: does it work? As if the answer is a single yes or no. The correct question is: which version are we talking about? Because right now, two very different approaches are advancing simultaneously, and understanding the gap between them tells you everything about where this is going.
What Standard CRISPR Already Proved
Casgevy, the first FDA-approved CRISPR therapy, is extraordinary. Do not let familiarity dull the numbers. Vertex's Casgevy has continued to show benefits in sickle cell disease patients for 5.5 years. Across three ongoing Phase III studies, all 45 SCD patients who received Casgevy achieved 12 consecutive months of freedom from hospital for an average of 36.1 months. In addition, 95.6 percent were free from vaso-occlusive crises for at least 12 months. For context: these are people who were spending, in some cases, 20 days a month in the hospital. Casgevy is not managing their disease. It is ending it.
Beyond hemoglobin disorders, NTLA-2001, a one-time intravenous CRISPR treatment for transthyretin amyloid cardiomyopathy, has advanced to a pivotal Phase 3 trial after showing more than 90 percent reduction in pathogenic TTR protein levels in patients. One infusion. Ninety percent reduction. As of early 2025, approximately 250 clinical trials involving gene-editing therapeutic candidates are being tracked, with more than 150 currently active.
The trendlines are unambiguous. Standard CRISPR, the kind that targets the same mutation across a defined patient population, works. Clinically, durably, dramatically. The problem is that it works for conditions where enough patients share the same underlying defect to make a clinical trial economically viable. Most genetic diseases do not qualify. Many doctors, patients and their families have been frustrated because pharmaceutical companies do not have a strong economic incentive to create gene-editing treatments for extremely rare disorders. That is the wall standard CRISPR hits. KJ's disease is on the other side of that wall.
The Regulatory Runway Just Opened
Here is what changed last week. The FDA unveiled draft guidance for an approval pathway that could see custom CRISPR therapies, like the one that treated critically ill baby KJ, formally embraced by the agency. The framework, called the "plausible mechanism pathway," is the regulatory infrastructure that makes bespoke medicine scalable. Fyodor Urnov of UC Berkeley put it plainly: "A disease with 100 causing mutations will no longer require 100 clinical trials." That sentence should stop you cold. It means the trial bottleneck, the reason pharmaceutical companies walk away from rare diseases, is being structurally dismantled.
A team of physicians and scientists created the bespoke in vivo CRISPR therapy for KJ, developed and delivered in just six months. This landmark case paves the way for a future with on-demand gene-editing therapies for individuals with rare, until-now untreatable genetic diseases, and sets precedent for a regulatory pathway for rapid approval of platform therapies in the United States. Platform is the key word. Kiran Musunuru, the Penn cardiologist who led KJ's treatment, said: "We realized we can do this over and over again, individualizing the therapy for many patients."
I want to be honest about one genuine tension here. The cost structure of CRISPR medicine is brutal right now. Casgevy was originally priced at $2.2 million per patient and the customized CPS1 treatment cost $2 million per dose. Although cell and gene therapies promised to revolutionize the pharmaceutical industry, uptake has been lower than predicted due to high costs and manufacturing challenges. Access is not solved. Equity is not solved. And I do not think you can simply wave that away by pointing to lifetime cost comparisons with standard of care, though those comparisons are real. The bespoke pathway has the potential to bring costs down through platform efficiencies, but that potential is not yet a fact.
What is a fact: KJ safely received three doses delivered by lipid nanoparticle, and each dose further reduced symptoms, suggesting additional editing with each dose. The therapy corrected the mutation, allowing KJ to tolerate high dietary protein, even while halving his nitrogen-scavenger drug dose, with no severe adverse effects. This is not a theoretical mechanism. This is a sick child getting better because a team of scientists built something specifically for him, in six months, and then the FDA reviewed the application in one week.
The standard CRISPR versus bespoke CRISPR framing sets up a competition that does not exist. Standard CRISPR proved the biology. Proved the delivery mechanisms. Prime editing, which does not create double-stranded breaks in DNA, may have a superior safety profile compared to traditional CRISPR editing, and it is already flowing into the next generation of both approaches. The two tracks are feeding each other. The teams working on Casgevy built the institutional knowledge that made KJ's treatment possible in six months instead of six years.
Thousands of researchers, clinicians, regulatory scientists, and engineers made this moment. Jennifer Doudna's lab. Rebecca Ahrens-Nicklas and Kiran Musunuru at CHOP and Penn. The IGI team at Berkeley. The FDA reviewers who cleared an IND application in a week because they understood what was at stake. The story of gene therapy is a story about accumulated human expertise converging on a problem, and the problem is losing.
We are going. The only question is when.
