DTX401

DTX401 is Ultragenyx's one-time AAV8 gene therapy for Glycogen Storage Disease Type 1a (GSDIa), an ultra-rare inherited condition where the liver cannot release stored glucose into the blood. A single IV infusion delivers a working copy of the G6PC gene to liver cells, with the goal of replacing the around-the-clock cornstarch dosing that keeps these patients alive today [1][2]. The key Phase 3 trial hit its primary endpoint at Week 48 with a mean 41.3% reduction in daily cornstarch intake in the DTX401 arm (n=20) versus 10.3% on placebo (n=24), p<0.0001; 68% of treated patients reached a ≥30% reduction and 37% reached a ≥50% reduction [3]. The BLA is under FDA review with a PDUFA date of August 23, 2026, which would make DTX401 the first approved gene therapy for any glycogen storage disease [4].

Novel compound. AAV8 vector carrying the human G6PC gene, originally developed at Dimension Therapeutics and acquired by Ultragenyx in 2017. The key Phase 3 trial (NCT05139316) is listed as completed with n=49 adults and adolescents [2][3]. FDA designations include Orphan Drug, Fast Track, and RMAT (Regenerative Medicine Advanced Therapy - a category that gives the program intensive FDA engagement and possible priority review eligibility) [5]. The PDUFA goal date of 2026-08-23 sits roughly two months out from now; as of the latest disclosures the FDA has not announced an advisory committee, but a formal waiver has not been confirmed publicly either, so an AdCom remains theoretically possible inside the review window. A long-term safety follow-up study (NCT03970278, n=12) and a separate GSDIa disease monitoring program (NCT06636383, n=140) are running in parallel, the latter explicitly designed to capture serious adverse events attributed to DTX401 over time [6][7]. Approval would be Ultragenyx's first commercial gene therapy launch, after building a rare disease commercial engine with Crysvita, Mepsevii, Dojolvi, and Evkeeza [5][8].

GSDIa is caused by mutations in the G6PC gene, which codes for glucose-6-phosphatase. That enzyme is the last step in releasing glucose from the liver. When you eat, your liver stores glucose as glycogen. Between meals, it breaks glycogen back down, and glucose-6-phosphatase chops off a phosphate group to release free glucose into the blood. Without a working enzyme, patients can store glucose but cannot release it, so they crash into severe hypoglycemia within hours of eating. The current standard of care is uncooked cornstarch every 3 to 4 hours, including overnight, often via feeding tube. Even with perfect adherence, patients accumulate liver adenomas, develop kidney disease, and live with chronic lactic acidosis (a buildup of lactic acid in the blood that causes fatigue, muscle pain, and over time damages organs) [1]. DTX401 packages a functional G6PC gene inside an AAV8 capsid, a virus shell that naturally homes to liver cells. Once inside hepatocytes, the gene sits as an episome (a loop of DNA that functions independently inside the cell without integrating into the chromosomes) and the cell produces functional enzyme. The mechanism is genetically dispositive: GSDIa is a pure monogenic loss-of-function disease, so restoring the protein restores the missing biochemistry. The open question is not whether replacing G6PC helps, it is how much enzyme activity you get and how long it lasts [1][9].

NCT05139316 was a randomized, double-blind, placebo-controlled Phase 3 in adults and adolescents with confirmed GSDIa, with a primary endpoint of percent change from baseline to Week 48 in daily cornstarch intake [2]. That endpoint is hard and quantitative: cornstarch dosing is the disease-defining intervention, and a meaningful reduction translates directly into freedom from nighttime feeds and fewer hypoglycemic events. Patients on placebo cross over to active treatment after the primary analysis. The trial met its primary endpoint with a mean 41.3% reduction in daily cornstarch intake in the DTX401 arm versus 10.3% on placebo at Week 48 (p<0.0001); responder analyses showed 68% of treated patients with ≥30% reduction and 37% with ≥50% reduction, compared with 13% and 4% on placebo [3]. The BLA was filed on that basis [5]. Sample size of 49 is small in absolute terms but reasonable for a disease with an estimated US prevalence of roughly 1 in 100,000 births. The weak spots in the design are the 48-week window (durability beyond a year matters for a one-time therapy priced like a one-time therapy) and the use of cornstarch reduction rather than a direct hypoglycemia or quality-of-life endpoint. Longer-term follow-up data released in September 2025 reported maintenance of cornstarch reduction beyond Week 48 in the rollover cohort [3]. The follow-on monitoring study (NCT06636383) is the durability backstop [6]. Earlier Phase 1/2 work (NCT03517085) in adults established the dose ultimately advanced into Phase 3 and showed cornstarch reductions persisting over multiple years of follow-up in the highest dose cohort, which is the strongest precedent for AAV8-G6PC durability outside the key study itself [9].

This drug is under FDA review (NDA/BLA), with a PDUFA decision date of 2026-08-23. Our estimate of 93% is the historical filing-approval rate for its area, adjusted for its rejection history (no prior Complete Response Letters). At this stage the early-trial design model no longer applies - what matters is that it reached the FDA and whether it has been rejected before.

Durability is the structural risk. AAV vectors deliver a functional gene as a non-integrating episome that gets diluted as hepatocytes divide. The Phase 3 primary endpoint is at 48 weeks; the commercial story requires years. If cornstarch intake creeps back up at year 3 or 4, the value proposition collapses, and re-dosing is blocked by anti-AAV8 neutralizing antibodies that the first dose itself induces. Safety risk is on-target liver enzyme elevation. AAV liver-directed therapies routinely trigger transient ALT (alanine aminotransferase, a liver enzyme whose blood level spikes when liver cells are stressed or inflamed) elevations managed with prophylactic or reactive corticosteroids, and Ultragenyx has reported steroid use in the GSDIa program [1][10]. Severe or sustained liver inflammation has derailed dosing in adjacent programs. Regulatory risk is CMC (chemistry, manufacturing, and controls - FDA's review of how the drug substance and drug product are made and tested). FDA gene therapy reviews have produced complete response letters tied to manufacturing rather than clinical data, with both Sarepta's Elevidys and BioMarin's Roctavian seeing CMC or confirmatory-trial wrinkles during their reviews. Commercial risk centers on price and access. GSDIa affects roughly 6,000 patients across the US and EU. A one-time gene therapy in this category will price in the $2 to $3.5M range to be viable, and payers will demand outcomes-based contracts tied to sustained cornstarch reduction, which puts the durability question back at the center of the reimbursement conversation. The Roctavian precedent is sobering: BioMarin announced voluntary US withdrawal of Roctavian in February 2026 after sales of only ~$36M in 2025, taking ~$240M in write-offs, with payer hesitancy and durability uncertainty central to the failure [11]. FDA approval for a one-time liver-directed gene therapy does not guarantee reimbursement or uptake.

Worth watching closely. PDUFA on 2026-08-23 is a defined catalyst inside a 60-day window, with three possible outcomes: clean approval, approval with a REMS (Risk Evaluation and Mitigation Strategy - a mandatory safety monitoring program FDA can attach to an approval) or post-marketing commitments, or a CRL (Complete Response Letter - FDA's formal rejection or conditional non-approval, typically requiring resubmission). A CRL here would most likely be on CMC grounds rather than clinical. The signal to track between now and August is any FDA advisory committee announcement (none currently scheduled, which is a mild positive for a rare disease BLA, though a formal waiver has not been confirmed) and Ultragenyx earnings commentary on commercial readiness, payer engagement, and manufacturing inspection status [5]. The deeper scientific signal sits beyond PDUFA: long-term cornstarch intake data from the open-label extension and the disease monitoring program will tell you whether AAV8 gene therapy is a one-and-done cure or a multi-year bridge. For Ultragenyx (FY2025 total revenue of $673M, with Crysvita at $481M and Dojolvi at $96M; 2026 guidance reaffirmed at $730 to $760M total revenue) [8], a DTX401 approval is the first true gene therapy launch and a proof point for the broader AAV pipeline. The next domino is DTX301 for OTC deficiency (ornithine transcarbamylase deficiency - a different rare urea-cycle liver enzyme disorder, currently in Phase 3) [5]. For the field, this would be the first approved gene therapy for any glycogen storage disease, opening a regulatory template for GSDIb, GSDII (Pompe), and beyond [1].