On May 16, 2024, the FDA withdrew the approval of a drug called infigratinib. It had reached the market three years earlier as Truseltiq, an oral inhibitor of a receptor family called FGFR, cleared under accelerated approval to treat a rare bile-duct cancer. Accelerated approval comes with a string attached: run the confirmatory trial that proves the early signal was real. That trial was never completed. According to the ASCO Post’s account of the withdrawal, the sponsor cited trouble enrolling the required study and concluded that keeping the drug on the market was “not commercially reasonable.” The cancer indication came off the shelves.
It did not stay gone for long.
This month, BridgeBio Pharma, the company behind infigratinib, reported positive phase 3 results for the very same compound, repurposed and re-dosed for an entirely different population: children with achondroplasia, the most common form of dwarfism. The logic is clean on paper. Achondroplasia is caused by an overactive FGFR3 receptor that slams the brakes on bone growth. Infigratinib dials FGFR signaling down. Take a drug built to choke FGFR signaling in tumors, drop the dose, and point it at the same receptor sitting jammed in the on position in a child’s growth plates, where its overactivity is exactly what stunts the bone. The cancer trial never got its confirmation. The growth trial hit its mark.
Here is the part regulators should be sweating, and mostly aren’t. The FDA has done this before. It approved another achondroplasia drug on the same surrogate five years ago and still does not have the one number that would tell anyone whether the approval was worth making. Now a second drug is arriving in the same currency, and the people it is meant for do not all agree it should exist.
What the trial measured, and what it didn’t
PROPEL 3 enrolled 113 children aged 3 to under 18 with open growth plates, randomized two-to-one to a once-daily pill or placebo for 52 weeks. The headline number is the kind that sells a drug, and it is not small: children on infigratinib grew at a least-squares mean of 5.96 centimeters per year against 4.22 on placebo, a difference of 1.74 cm/year with a p-value under 0.0001. The company called it the highest annualized height velocity ever reported in a randomized achondroplasia trial. The safety readout looked clean: no drug-related serious adverse events, no discontinuations, three transient cases of elevated blood phosphate.
Annualized height velocity is the centimeters a child gains in a year. It is a sensible thing to track over twelve months because it is the only thing you can track over twelve months. It is also a surrogate. The trial measured how fast these children grew. It did not measure whether they end up taller as adults, or whether the spinal stenosis and the sleep apnea and the daily friction of a world built for bigger people eased by a single millimeter. A family reading the press release is being sold a centimeter-per-year figure. The number that would tell them whether their child’s life gets measurably better is not in the document, because a one-year topline readout cannot contain it.
The proportionality claim is doing more work than the data
BridgeBio led its release with something sharper than raw speed: the “first statistically significant improvements in body proportionality” ever shown in achondroplasia. Proportionality matters because the condition lengthens limbs and trunk unevenly, and a drug that adds height without touching the ratio is just building a taller version of the same disproportion. The measure here is the upper-to-lower body segment ratio.
Read the fine print and the claim narrows fast. The ratio improved by a least-squares mean of 0.05 against placebo in children under eight, who made up more than half the trial, at a p-value under 0.05. Across the full enrolled population the difference shrank to 0.02 and landed at a p-value of 0.1849, which is to say it did not reach significance at all. A benefit that holds in the younger half and dissolves in the whole is a finding worth following, not a finished one. The press release put the half that holds in the headline.
We have been here before
There is already a drug on this exact path, and it is the reason the surrogate sounds familiar. BioMarin’s Voxzogo, a daily subcutaneous injection, won accelerated approval in 2021 on the same annualized growth velocity, with continued approval contingent on a confirmatory study to establish the effect on final adult height. A 2026 review of the real-world evidence notes that the effect on final adult height remains formally unestablished years later. Infigratinib’s entire commercial pitch is that it is a pill instead of a shot. That is a genuine advantage for a child facing years of daily needles. It is not evidence about adult height, which neither drug has delivered.
The same review documents what the press releases tend to skip. Little People of America, the country’s largest dwarfism-advocacy organization, has objected to this whole enterprise: chasing growth velocity, the group argues, does not address the health and quality-of-life needs people with achondroplasia actually name, and most of the hardship attached to short stature is social bias and a built environment that ignores them, not a medical defect in the child. BridgeBio had an answer ready, a quote from a Little People of America committee member calling the proportionality result meaningful to families. It is a real quote, and it slides past the actual objection. The complaint was never that better proportionality would be unwelcome. It was that growth velocity is the wrong thing to be chasing in the first place.
BridgeBio plans to file with the FDA and European regulators in the second half of 2026, carrying a Breakthrough Therapy designation it already holds. The agency will weigh a clean one-year safety readout, a real growth signal, and a proportionality claim that depends entirely on which half of the trial you look at. What it will not have, because no one has it, is the number that settles the question: how tall these children become, and whether their lives are better for the years on the drug. That is the number BioMarin still owes on a medicine approved five years ago. Now there will be two.
Sources
- BridgeBio – Positive Phase 3 PROPEL 3 topline results for oral infigratinib in achondroplasia (investor release, 2026)
- BridgeBio – PROPEL 3 topline results deck, February 2026 (PDF)
- FDA – Withdrawal of accelerated approval of infigratinib for metastatic cholangiocarcinoma (2024)
- ASCO Post – FDA withdraws approval of the FGFR inhibitor for cholangiocarcinoma (May 2024)
- FDA – Accelerated approval of vosoritide (Voxzogo), first drug to improve growth in children with achondroplasia (2021)
- PMC – Vosoritide (Voxzogo) for achondroplasia: clinical and real-world evidence, including final-adult-height status and community concerns (2026)
- MedPage Today – Oral FGFR1-3 inhibitor boosted height velocity in achondroplasia