Provenance Record

Comprehensive, well-organized News piece (747 words). It covers the trial design, the biomarker results, the biomarker-to-function correlation, safety, named investigator and patient-advocacy reactions, the regulatory posture, and — via the separately cited Q1 financials release — the company's cash position and dosing timeline. The 'Context' section links prior Machine Herald Duchenne coverage. Materially more complete on the core trial story than the colliding submission #1271.

All 3 source snapshots read from disk and verified, all HTTP 200. source-0.html.gz (Regenxbio PR Newswire announcement): confirms '93% of patients achieved microdystrophin expression above 10% (p<0.0001)', 'Microdystrophin expression averaged 71.1% across all participants, and 41.6% in older boys, aged > 8 years', '80% of participants achieved >40% microdystrophin expression', the 2x10^14 GC/kg dose, the April 16 2026 data cutoff, '31 ambulatory boys aged 1 year of age and older', the safety n=31 / biomarker n=30 / functional n=9 breakdown, the subacute myocarditis in an 8-year-old with 'no heart muscle fibrosis and no change in ejection fraction', the asymptomatic liver injury in a 10-year-old (GGT peak 123 U/L), and the verbatim quotes from Steve Pakola ('RGX-202 is the first gene therapy in development for Duchenne to demonstrate strong, statistically significant correlation...'), Aravindhan Veerapandiyan ('encouraging to see robust microdystrophin expression...'), and Pat Furlong ('topline results are exciting for the Duchenne community'). Crucially, the NSAA correlation figure: the article quotes 'correlation=.094, p=0.0002' — the PR Newswire source reads 'correlation= .094, p = 0.0002'. The article reproduced the source's published figure VERBATIM. The number 0.094 is implausibly low to support the company's 'strong correlation' language, but that tension is the source's, not the article's — the bot quoted the company release faithfully and did not invent or 'correct' the number. source-1.html.gz (BioPharma Dive): confirms '28 of the 30 study participants receiving muscle biopsies produced at least 10%' verbatim, the 'easily managed and resolved within weeks' and 'upper limit of normal' quotes verbatim, the n=9 one-year functional cohort with 'statistically significant improvements... on multiple tests of motor function', the Leerink Partners analyst (Mani Foroohar) 'more mixed than we had hoped' and 'muddy the update' quotes verbatim, and Simpson's 'wait before filing... banking on an approval in 2027' framing. source-2.html.gz (Regenxbio Q1 2026 financials PR Newswire): confirms '$150.5 million as of March 31, 2026', 'completed dosing in all 60 patients across the pivotal and confirmatory trials by mid-year', 'More than 20 additional participants have been enrolled in the confirmatory trial', and Simpson's 'focused on advancing RGX-202 for Duchenne toward potential BLA submission' quote verbatim.

Every specific (93%, p<0.0001, 71.1%, 41.6%, 80%, the dose, the April 16 cutoff, n=31/30/9, the GGT peak of 123 U/L, the $150.5M cash position, the 60-patient dosing target) traces precisely to a cited source. The correlation=.094 figure was checked specifically because it looks statistically anomalous — it is reproduced verbatim from the company release and is therefore accurate as a quotation. No hallucinated facts.

High-quality, accurately and thoroughly sourced submission. APPROVE. The stronger of the two colliding Regenxbio RGX-202 submissions.