SAR442501 in achondroplasia: an anti-FGFR3 antibody dropped at 16 patients in a crowded field

What was tried

Sanofi ran upreACH-2 (NCT06067425), a Phase 2, open-label, sequential dose-escalation study of SAR442501 in children from birth to 12 years with achondroplasia confirmed by an FGFR3 mutation [1]. SAR442501 is a subcutaneous anti-FGFR3 monoclonal antibody designed to dampen the constitutive FGFR3 signaling that arrests chondrocyte proliferation in the growth plate [2]. The trial used two dose cohorts, with annualized growth velocity, growth velocity Z-score, and the cartilage turnover marker collagen X among its endpoints [1]. The study terminated after enrolling 16 children, 10 in the low-dose cohort and 6 in the high-dose cohort [1]. The posted reason for stopping reads "Sponsor decision. Not related to safety concern." [1].

The biological hypothesis

Achondroplasia is one of the cleanest gene-to-disease relationships in medicine. A recurrent gain-of-function variant in FGFR3, most often the G380R substitution, drives overactive receptor signaling that suppresses chondrocyte proliferation and differentiation in the growth plate [3]. Open Targets scores the FGFR3 association with achondroplasia at 0.82, with a genetic association component of 0.89, which places it among the most strongly anchored targets in the platform [4]. Three pharmacological strategies follow from that biology: restore the downstream brake on the MAPK pathway with a C-type natriuretic peptide analog, block the receptor directly, or trap the ligand [8]. SAR442501 took the direct-blockade route with an antibody to FGFR3, betting that receptor-level inhibition could match or exceed the growth benefit seen with downstream agents.

What actually happened

The trial did not fail on biology or safety. Of 16 enrolled children, none completed the planned treatment period because the study was terminated early [1]. In the low-dose cohort of 10, one participant had a serious adverse event, and in the high-dose cohort of 6, none did, with no deaths in either cohort [1]. No efficacy conclusion is possible from a dose-escalation cohort of this size stopped before its primary readout, and none was posted [1]. Sanofi listed the stop as a sponsor decision unrelated to safety, and outside reporting placed SAR442501 inside a broader rare-disease portfolio reduction rather than a program-specific signal [5]. There are no FAERS reports for SAR442501, consistent with an investigational antibody that never reached the market [6].

Failure mechanism, best guess

This was a portfolio termination in a crowded, genetically validated field, not a mechanistic defeat. By the time SAR442501 reached pediatric dosing, achondroplasia already had an approved daily therapy. Vosoritide, a CNP analog, raised annualized growth velocity by an adjusted mean of 1.57 cm per year versus placebo in its Phase 3 trial of 121 children [7]. Longer-acting CNP prodrugs and oral FGFR3 pathway inhibitors were advancing behind it, and an earlier soluble FGFR3 decoy from another sponsor had already been discontinued, which left SAR442501 as a fourth or later entrant into a maturing market [8]. An anti-FGFR3 antibody entering that landscape faced a steep differentiation bar. It had to beat or meaningfully simplify a once-daily injectable with an established effect size, while managing the broad physiology of FGFR3, which also governs bone, brain, and inner-ear development. When a late entrant cannot show a credible path to differentiated value, the rational move is to stop before a large pediatric program, which is what the enrollment of 16 and the sponsor-decision label together describe.

How to prevent this next time

Two quantitative tools would have surfaced this decision earlier. The first is a competitive red-team with an explicit differentiation power calculation. Sizing the program to detect a clinically meaningful improvement over the vosoritide benchmark of 1.57 cm per year, rather than over natural history, would have exposed how large and how long a trial needed to be to prove value [7]. The second is a decision-analytic go/no-go that expresses the probability of a differentiated, approvable result as a Bayesian predictive probability of success:

PP(success)=∫θ​P(final success∣θ,interim data)⋅π(θ∣interim data)dθ

Here the relevant final success is not any growth effect but a growth effect that clears an approved comparator, so the prior over the treatment effect should be centered on the incremental benefit, not the absolute one. Run that calculation against an approved drug and a strong genetic anchor of 0.82, and the expected value of a late third or fourth entrant compresses quickly [4]. The single highest leverage change would have been to run the competitive red-team and the differentiation power calculation against vosoritide before opening a pediatric dose-escalation, not after enrolling 16 children [7].

What this means for similar programs

A target can be fully validated and still be the wrong place to start a program. FGFR3 in achondroplasia carries one of the highest genetic association scores in Open Targets, yet that strength cuts both ways, because it invites many entrants and rewards the first to an approval [4]. For the next molecule against a saturated, genetically clean target, the binding question is differentiation, not validation. Modality matters here too. An antibody to a pleiotropic receptor carries a wider potential toxicity surface than a downstream peptide, which raises the differentiation bar further.

Open questions

Because the trial stopped before its readout, the biology stays untested rather than disproven. Did SAR442501 engage FGFR3 at the growth plate at tolerated doses, as measured by collagen X or growth velocity in the 16 dosed children? Would receptor-level blockade have produced a larger or more durable effect than CNP analogs if pushed to its primary endpoint? Does direct FGFR3 antibody blockade carry tissue-specific risks that only a longer pediatric exposure would reveal? These stay open for the next FGFR3 antibody, if one is attempted.

Sources

1. ClinicalTrials.gov. NCT06067425, upreACH-2 study record with posted results (16 enrolled, 10 low dose and 6 high dose, 0 completed, 1 serious adverse event in the low-dose cohort, 0 deaths, reason stopped "Sponsor decision. Not related to safety concern."). https://clinicaltrials.gov/study/NCT06067425.

2. PatSnap Synapse. SAR-442501, anti-FGFR3 monoclonal antibody, subcutaneous, Sanofi. https://synapse.patsnap.com/drug/83071a61c92641898901a175ef6667ef.

3. Legeai-Mallet L. C-Type Natriuretic Peptide Analog as Therapy for Achondroplasia. Endocr Dev 2015, 30:98-105.

4. Open Targets Platform. FGFR3 (ENSG00000068078) association with achondroplasia, overall 0.82, genetic association 0.89. https://platform.opentargets.org/target/ENSG00000068078/associations.

5. Taylor NP. Sanofi will drop phase 2 dwarfism drug as part of rare disease clear-out. FierceBiotech, 2026. https://www.fiercebiotech.com/biotech/sanofi-drops-phase-2-dwarfism-drug-part-rare-disease-clear-out.

6. OpenFDA FAERS. medicinalproduct SAR442501, no reports returned. https://api.fda.gov/drug/event.json.

7. Savarirayan R, et al. Once-daily, subcutaneous vosoritide therapy in children with achondroplasia: a randomised, double-blind, phase 3, placebo-controlled, multicentre trial. Lancet 2020, 396:684-692.

8. Wrobel W, et al. Advantages and Disadvantages of Different Treatment Methods in Achondroplasia: A Review. Int J Mol Sci 2021, 22:5573.