In the May 12 update I wrote that the H1 2026 FDA accelerated approval meeting “remains the binary” and that management’s language was “the same hedge they were using in March.” That hedge just resolved — in the wrong direction on timing. The Q2 10-Q filed this morning states plainly: “We plan to meet with the FDA late in the fourth quarter of 2026 to discuss the INSPIRE DUCHENNE SGT-003 data package and to seek guidance on a potential accelerated approval pathway.”
That’s a roughly two-quarter slip from the guidance the stock has been priced against since my original deep dive. Two other lines in the same filing deserve attention: a second treatment-related serious adverse event now sits in the INSPIRE DUCHENNE safety table, and the runway language quietly trimmed from “into the first half of 2028” to “into mid-2028.”
The stock closed at $8.71, essentially flat (-1.7%) from the $8.86 print on June 25. The market has not digested this yet.
The Receipts
The FDA meeting moved to Q4 2026. The 10-Q language is specific and new: a meeting “late in the fourth quarter of 2026” to discuss the INSPIRE DUCHENNE data package. The implication: every prior filing said H1 2026 with an “update as discussions progress” hedge. That hedge has now been replaced by a hard date roughly six months later than the calendar in my original deep dive. Delays in scheduling a Type B/C interaction are not automatically a negative signal — they often reflect the sponsor waiting to assemble a fuller data package, and a company with a Phase 3 already dosing has less urgency to force the meeting. But it does mean the catalyst that anchors this entire thesis is now a 2027 story in terms of resolution, not a 2026 one.
A second treatment-related SAE, at N=53. As of August 4, 2026, SGT-003 has been dosed in 53 participants, with two previously disclosed treatment-related serious adverse events. In the Q1 10-Q, at N=47, there was one — the Grade 3 immune-mediated myositis that fully resolved. The risk-factor section confirms both events “were reviewed by the data and safety monitoring board with the recommendation to continue dosing without interruption, and both events have resolved.” One participant withdrew consent after Day 60 with no treatment-related AEs. Why it matters: this is the single most important number in the SLDB story, and it needs to be read carefully rather than reflexively. Two resolved SAEs in 53 dosed patients is a rate of ~3.8%, versus ~2.1% at the last count. Both resolved. The DSMB did not pause. For context on the bar: Elevidys carries a boxed warning for acute liver injury and acute liver failure following two reported deaths, and lost the non-ambulatory population from its Indications section. Solid’s low-burden, ~33-day steroid-only regimen continues to hold up in a way that competitor regimens requiring sirolimus or eculizumab do not. The clean-safety pillar of the thesis is intact — but it is one notch less pristine than it was in May, and in AAV gene therapy the safety database is the asset. At N=53 the confidence interval around a 3.8% SAE rate is still wide; what matters is whether the third event shows up at N=70.
Runway trimmed: “H1 2028” → “mid-2028.” Cash, cash equivalents and available-for-sale securities stood at $377.7M at June 30, down slightly from $380.7M at March 31 despite raising fresh equity in the quarter. The implication: Q2 operating expenses were $57.4M (R&D $44.3M, G&A $13.1M) against $41.7M in Q2 2025 — a 37.6% year-over-year increase. Net loss was $54.8M for the quarter and $111.5M for the half. Six-month cash used in operations was $99.8M. The burn I described in March as “the burn rate reality” is now running near $50M/quarter of actual cash out the door, and it’s still climbing as IMPACT DUCHENNE activates US and UK sites in H2.
SGT-003 is eating the budget. Segment disclosure shows $24.5M of Q2 spend on SGT-003 alone, up 99.0% from $12.3M a year ago, driven by manufacturing and clinical costs. Meanwhile SGT-601 (TNNT2 dilated cardiomyopathy) spend collapsed to $22,000 for the quarter from $2.3M, and SGT-501 fell 59.4% to $903K. Why it matters: this is a company concentrating chips behind the lead asset and letting the early cardiac bench idle. That’s rational capital allocation ahead of a BLA-relevant meeting, but it thins the “platform company, not a one-trick pony” argument I made in the original deep dive — at least on a cash-deployed basis.
The ATM got 2.4x bigger. The 424B5 filed today registers up to $200,000,000 of common stock through Jefferies. The prior facility, filed May 2025, was $85M — and it is fully exhausted: approximately $84,993,553 in gross proceeds sold, with $6,447 remaining. The implication: in the June 25 update I flagged that “the ATM is live and being worked methodically.” The 10-Q quantifies it: 8,515,987 shares sold in H1 2026 for $60.8M net, plus another 2,012,500 shares in July 2026 for $15.7M net. Share count went from 78,967,888 at January 1 to 107,225,238 as of August 4. That’s a 36% increase in shares outstanding in seven months.
The pipeline timelines also moved right. SGT-212 (Friedreich’s ataxia) initial FALCON data is now guided to Q1 2027, not H2 2026 — two participants dosed, well tolerated, no treatment-related SAEs. SGT-501 (CPVT) first patient dosing has slid to H2 2026 with safety data in H1 2027. Why it matters: the optionality assets I described as “lottery tickets” haven’t broken, but every one of them is now a 2027 event.
The Scientific Reality Check
Let me be precise about what has and hasn’t changed in the biology, because the filing contains no new clinical data on efficacy.
