Obsidian Therapeutics OBX Q3 2026: Big Miss, Bigger War Chest

Obsidian Therapeutics reported a Q3 2026 GAAP loss of $1.23 per share, well short of the $0.42 consensus estimate, a gap of nearly 193%. The headline miss is jarring, but a $15.1 million one-time acquired in-process R&D charge tied to its transaction with Galera Therapeutics accounts for a meaningful slice of it. Strip that out and the underlying quarterly spending picture looks closer to recent quarters.

The rest of the loss was driven by rising operating costs that come with being a newly public company. General and administrative expenses climbed to $10.3 million from $4.6 million in the same quarter a year earlier, reflecting the overhead of a Nasdaq listing that began August 4, 2026. R&D spending rose more modestly to $27.5 million from $25.0 million. As a clinical-stage biotech, Obsidian reported no revenue, which was fully expected.

The more consequential number is on the balance sheet. A $350 million private placement, which CEO Madan Jagasia described as oversubscribed, left the company with $324.4 million in cash and marketable securities. Management says that runway extends into the second half of 2028, comfortably past the most important near-term milestone.

That milestone is a topline readout from the registration-enabling cohort of the Agni-01 study for amsoki-cel, the company's lead TIL cell therapy targeting checkpoint inhibitor-resistant advanced melanoma. Enrollment is expected to wrap by the end of Q1 2027, with topline data due by year-end 2027. At the SITC 2026 conference, updated data from 15 patients showed a 67% objective response rate, a median progression-free survival of 8.3 months, and a median duration of response that had not yet been reached at a median follow-up of 14.3 months.

Five of those 15 treated patients remained in deep, ongoing response at the 12-month post-infusion landmark. The pooled safety data across 37 patients showed no dose-limiting toxicities, no treatment-related deaths, no ICANS, and no ICU transfers, with just one Grade 3 cytokine release syndrome event. Those figures, published in Nature Medicine, form the clinical foundation the company is now trying to replicate at scale in the registration cohort. A separate Phase 1 dataset in non-small cell lung cancer is expected in the first half of 2027.