Provenance Record

Well-structured News piece (644 words) using an Overview / What We Know / Regulatory Path / What We Don't Know format. Technical specifics (dose, delivery route, cohort sizes, endpoint definitions) are precise and correctly attributed. The headline and summary accurately convey the actual news: dosing complete in the pivotal trial plus longer-term Phase 1/2 data — not a pivotal efficacy readout.

All 3 source snapshots read from disk (gunzip-decompressed) and verified against the article's claims. source-0.html.gz (GlobeNewswire, status 200, primary source): CONFIRMED — completion of dosing of 17 females in the overenrolled REVEAL pivotal trial; developmental plateau population; single-arm, open-label, single intrathecal high-dose TSHA-102 (1x10^15 total vg), ages 6 to <22; self-complementary AAV9 delivering a functional form of MECP2 via miRARE technology; Part A N=12 ages 6-21, 100% gained/regained >=1 developmental milestone by 12 months; 310 total functional gains (~26/patient) = 31 milestones + 279 additional skill gains; milestone gains up 69% (6-12 mo) and 94% (6 to >=12 mo); Rett caused by MECP2 mutations, 15,000-20,000 patients across U.S./EU/U.K., no approved disease-modifying root-cause therapies; ASPIRE to dose four females aged 2 to <4 years; interim analysis after all 17 complete 6 months follow-up; Nolan quote ('In early 2027, we plan to engage with the FDA to review the interim data and discuss next steps toward submitting the BLA, with topline results and regulatory feedback expected in the first half of 2027') is VERBATIM; Rossignol quote ('TSHA-102 consistently drove early, durable functional gains across the core domains of the disease, including communicating with words or phrases, eating with utensils without assistance and walking with support, which continue to accumulate over time') is VERBATIM and her title/affiliation match the source. source-1.html.gz (BioSpace, status 200, verbatim mirror of the press release): CONFIRMED — 'no treatment-related SAEs or DLTs reported as of the June 2026 data cutoff across REVEAL Phase 1/2 and pivotal trials (N=29)' matches the article's cited claim exactly. source-2.html.gz (CGTLive, status 200, OFF-ALLOWLIST): CONFIRMED — first patient dosed in Q4 2025 ('initial dosing occurred in the fourth quarter of 2025' = article's 'late 2025'); Breakthrough Therapy Designation from FDA (announced October 2025); 'less than 6.7% likelihood' of gaining/regaining a milestone without treatment per natural-history data (article's '<6.7% of comparable patients'). No snapshot failed; no WebFetch fallback needed.

No hallucinations, no fabricated specifics, no misattributed quotes. Every number, name, code, date, and endpoint traces to a cited snapshot. The article correctly attributes the newer figures (17 dosed, 15,000-20,000 patients) to GlobeNewswire rather than CGTLive's older figures (15 planned, ~10,000 patients), avoiding a stale-number error.

High-quality, accurately sourced News submission. All quotes verbatim, all specifics traceable, and the sensitive clinical distinction between completed pivotal dosing and Part A efficacy is drawn cleanly. The lone off-allowlist source is credible and corroborated with nothing to correct, so this is a clean APPROVE rather than APPROVE_WITH_CORRECTIONS.