Cognition Therapeutics (CGTX) Q4 2025 earnings review
Pivoting to DLB Psychosis Amid Dilution and Cash Preservation
Cognition Therapeutics is betting its future on zervimesine (CT1812) for Dementia with Lewy Bodies (DLB) psychosis and early Alzheimer's, making the difficult but necessary choice to scrap its dry AMD program. This triage successfully extended the cash runway from late 2025 to Q2 2027. Q4 saw a significantly narrowed net loss of ~$3.3M (derived) and a surge in grant income. However, survival came at a steep cost to shareholders: the weighted-average share count nearly doubled year-over-year. The core narrative now hinges on a mid-2026 FDA meeting to legitimize a post-hoc efficacy signal into a registrational Phase 3 trial.
🐂 Bull Case
By cutting the MAGNIFY (dry AMD) study and leveraging its ATM facility, management secured cash through Q2 2027. This bridges the gap to the critical 2027 readout for the 545-patient START trial in Alzheimer's.
There are currently no approved therapies for DLB psychosis. Zervimesine's safety profile and oral administration give it a distinct advantage if the FDA accepts the Phase 2 efficacy signals.
🐻 Bear Case
To fund operations, the company heavily utilized direct offerings and its ATM. Weighted-average shares outstanding spiked 83% year-over-year to 72.8 million in FY25, crushing per-share value.
The 102% slowing of psychotic symptoms touted by management is based on a post-hoc analysis of the NPI-4 scale from the SHIMMER study. The FDA is historically skeptical of post-hoc endpoints for pivotal trial design.
⚖️ Verdict: ⚪
Neutral. Management made the right operational moves by killing the AMD program to extend the cash runway, but the investment case is heavily diluted and relies on the FDA accepting retrospective data cuts to greenlight a DLB Phase 3 trial.
Key Themes
Strategic Pivot to DLB Psychosis
Following end-of-Phase 2 feedback, Cognition is narrowing its primary focus to DLB psychosis—a condition affecting the majority of DLB patients with zero approved treatments. In the Phase 2 SHIMMER study, zervimesine demonstrated a 102% slowing of decline on the NPI-4 scale versus placebo. If a mid-2026 FDA meeting confirms this path, Cognition could secure a first-to-market advantage in a highly lucrative niche.
Rampant Share Dilution
The cost of staying alive in the clinical-stage biotech space has been brutal for equity holders. Weighted-average common shares outstanding skyrocketed from 39.7 million in FY24 to 72.8 million in FY25. While necessary to fund the pipeline, this level of dilution severely caps upside potential on positive clinical readouts.
Regulatory Foundation Relies on Post-Hoc Analysis
The entire DLB psychosis registrational plan hinges on an NPI-4 endpoint derived from a post-hoc analysis (evaluating patients regardless of baseline psychosis). While the 102% slowing looks impressive on slides, the Division of Psychiatry may demand a separate proof-of-concept study specifically targeting baseline-psychotic patients before allowing a Phase 3, introducing massive timeline risk.
START Trial Completes Enrollment
The Phase 2 START study in early Alzheimer's disease successfully completed enrollment (545 patients) in December 2025. Uniquely, this trial permits background therapy with approved anti-amyloid monoclonals (lecanemab, donanemab) in 15-20% of participants, positioning zervimesine as a potential combination therapy. The 18-month timeline pushes data release to 2027.
Disciplined Cost Controls
Management successfully decelerated spending to stretch the balance sheet. FY25 R&D expenses fell 11% YoY to $37.2 million, driven by the completion of the SHINE and SHIMMER trials. G&A expenses also dropped 14% to $10.6 million due to lower stock-based compensation.
Other KPIs
Stable. The company ended FY25 with $37.0M in cash and equivalents, plus an additional $35.7M in obligated grant funds remaining from the NIA. This combined liquidity pool underpins the operational runway projection.
Decelerating. A significant improvement from the $34.0 million net loss reported in FY24. This was achieved through a combination of higher full-year grant income ($23.4M vs $19.5M) and reduced operating expenses across both R&D and G&A.
Guidance
Accelerating. This is a massive improvement from the Q4 2025 timeline guided in earlier quarters. The extension was manufactured by concluding the dry AMD study, tapping the ATM facility, and enforcing strict expense control.
Stable. With enrollment completed in December 2025 and an 18-month treatment period, the timeline is firmly locked for 2027. This will act as the company's ultimate binary event.
Key Questions
FDA Receptiveness to Post-Hoc Data
Going into the mid-2026 meeting with the FDA's Division of Psychiatry, what specific feedback gives you confidence that the agency will accept a post-hoc NPI-4 analysis as the foundation for a pivotal Phase 3 design in DLB psychosis?
Partnership Prospects
With the cash runway extended to mid-2027, has the urgency for a non-dilutive partnership diminished, or are active discussions still underway to fund the potential Phase 3 DLB trial?
Future Capital Strategy
Given the dramatic increase in shares outstanding throughout 2025, what is the current capacity of the ATM facility, and how heavily will the company rely on it before the 2027 Alzheimer's data readout?