AVB-114

AVB-114 is Avobis Bio's autologous cell-seeded plug for complex Crohn's perianal fistulas - tunnels that form between the rectum and skin in Crohn's patients and resist conventional treatment. The Phase 2 STOMP-II trial (NCT04847739, n=60) completed its primary analysis in August 2025 and the data were presented at the American College of Gastroenterology Annual Scientific Meeting on October 29, 2025 [1][2][3]. FDA granted Fast Track designation in February 2025 and RMAT (Regenerative Medicine Advanced Therapy) designation in October 2025 - the two most material regulatory tailwinds available to a cell therapy at this stage [4][5]. This matters because the only other cell therapy to reach late-stage trials in this indication - Takeda's Alofisel (darvadstrocel) - failed its confirmatory US Phase 3 in 2023, leaving a wide opening for anyone with a working approach.

Novel investigational product, no approvals anywhere. Phase 2 STOMP-II (NCT04847739, n=60) primary analysis was completed in August 2025 and results were first presented publicly at ACG 2025 on October 29, 2025 [1][2][3]; peer-reviewed publication is pending. AVB-114 holds FDA Fast Track designation (February 27, 2025) and RMAT designation (October 3, 2025), granted on the basis of Phase 1 STOMP-I data (76% drainage cessation at 12 months in refractory patients) and the completed STOMP-II primary analysis [4][5]. RMAT unlocks intensive FDA collaboration on Phase 3 design, senior management engagement, and the option of rolling BLA review - materially de-risks the path that the ADMIRE-CD II Phase 3 failure made treacherous for this class. Avobis Bio is a private company built around this asset; no SEC filings constrain disclosure timing. Next milestone is the post-RMAT Type B meeting with FDA that should define Phase 3 design, followed by financing or partnership - autologous cell therapy is capital-intensive to commercialize, and a small private sponsor will need either a strategic partner or a meaningful raise before key start. Expect that conversation to surface in 2026-2027.

The stromal vascular fraction (SVF) is what you get when you take a patient's fat, digest it enzymatically, and spin out everything that isn't a mature fat cell - a mix of mesenchymal stromal cells, endothelial progenitors, pericytes, and regulatory immune cells. AVB-114 seeds this autologous cell mixture onto a bioabsorbable matrix plug, which a surgeon places into the fistula tract. The cells do two things in tandem: they dampen local inflammation (Crohn's perianal fistulas are driven by TNF-α and IL-1-rich chronic inflammation that prevents normal wound closure) and they recruit host tissue to fill in the tract as the matrix resorbs. The preceding open-label Phase 1 (STOMP-I) reported 76% drainage cessation at 12 months in patients with refractory perianal Crohn's, plus a ten-fold reduction in annualized exams under anesthesia and no treatment-related serious adverse events - this anchored the Phase 2 hypothesis and supported FDA Fast Track [4]. The mechanism has additional validation from a different angle - Alofisel (darvadstrocel), an allogeneic expanded adipose MSC product, hit its Phase 3 endpoint in Europe (ADMIRE-CD) and received EU approval in 2018, proving adipose-derived cells can heal these fistulas. But the US ADMIRE-CD II confirmatory trial missed in 2023 and Alofisel was withdrawn, so the cell therapy thesis is wounded, not dead [6]. One mechanistic hypothesis for why AVB-114 may outperform Alofisel: SVF retains regulatory T cells and vascular progenitors that are lost during the ex vivo expansion process required to make allogeneic MSC products, and autologous cells eliminate immune rejection risk that allogeneic products carry. Neither hypothesis has been clinically validated head-to-head, but both are plausible explanations and biomarker endpoints in Phase 3 (regulatory T cell engraftment, vascularization signatures on biopsy) would be the way to test them.

STOMP-II (NCT04847739) is a Phase 2, multicenter, randomized study comparing AVB-114 against standard of care in adults with complex Crohn's perianal fistulas who failed prior therapy (typically anti-TNF biologics, often vedolizumab or ustekinumab as well) and have quiescent rectal disease [2]. Primary endpoint: difference in combined remission rates between arms - combined remission means clinical closure of all treated external fistula openings plus absence of fluid or abscess pockets larger than 2 cm on MRI (the imaging test that shows whether the tract has truly closed or is just externally sealed). That's the same composite endpoint used in ADMIRE-CD, which makes cross-trial comparison meaningful. Enrollment was n=60, which is small for a randomized trial and means the readout is hypothesis-generating rather than statistically definitive for regulatory purposes - though RMAT designation provides a path to compress confirmatory requirements. The Schwartz et al. ACG 2025 presentation reported the randomized cohort showed clinical benefit [1][3]; the full effect size and durability data will determine whether Phase 3 is justified and how it should be powered. Quiescent rectal disease is an important inclusion criterion - active rectal inflammation independently prevents fistula healing, so this enriches for patients where the local intervention has a chance to work. Trial design is reasonable; concerns are the small n and the absence of a sham control. Without a sham procedure, patients in the standard-of-care arm know they didn't receive the intervention, which can deflate control arm response rates and artificially inflate the treatment effect - a known problem in prior fistula trials and a leading explanation for ADMIRE-CD II's geographic discordance.

Our model puts this drug's approval odds at 9%. It starts from the historical approval rate for Phase 2 drugs in this area (about 30%), then adjusts based on ten specific facts about the trial and its sponsor. The number is pulled down mainly by the sponsor's weak approval record and limited earlier-phase results, though having more secondary endpoints than usual pushes it back up. The remaining facts fall near average for this stage, so the final estimate lands well below the baseline.

Efficacy risk: the Alofisel precedent is the dominant one - ADMIRE-CD hit in Europe, ADMIRE-CD II missed in the US, and the leading explanations (placebo response inflation, geographic differences in surgical technique, patient selection drift) all apply to AVB-114's eventual Phase 3 [6]. Small Phase 2 effect sizes in this indication have repeatedly failed to replicate. RMAT mitigates this somewhat by giving FDA a seat at the Phase 3 design table. Safety risk: relatively low - autologous SVF has been used in other indications without major signals, and STOMP-I reported no treatment-related serious adverse events. Watch for infection at the placement site and ectopic tissue formation, both of which are mechanism-plausible. Execution risk: autologous means a fat harvest procedure and same-day or next-day surgical placement, which constrains commercial geography to centers with cell processing capability. Avobis is private and small; key trial financing is a real question even with RMAT tailwinds. Commercial risk: even if approved, payer coverage for a one-time procedural cell therapy in a non-life-threatening indication is uncertain - Alofisel struggled with reimbursement in EU markets despite approval. Pricing has to land above complex fistula surgery (~$15-30K per episode) without justifying itself against biologics patients are already failing. Market sizing is constrained by the strict eligibility criteria (quiescent rectal disease, multiple failed biologic lines).

Worth watching, position warranted before Phase 3 readout if financing/partnership materializes. The regulatory profile is materially better than the prior version of this writeup suggested. Regulatory Catalysts: RMAT designation (October 2025) [5] is the key de-risker - watch for the post-RMAT FDA Type B meeting outcome, which should clarify whether RMAT supports an accelerated Phase 3 endpoint, rolling BLA review, or both. This is the inflection that could compress the 2027-2028 timeline. Fast Track (February 2025) [4] supports rolling submission and is generally a quality signal. Data Catalysts: the signal that would change the assessment is the full STOMP-II publication - specifically the durable remission rate at 24 and 52 weeks, and whether the effect size is large enough to power a feasible Phase 3 at sites controlled for surgical technique. If the published data shows a >20 percentage point absolute difference in combined remission and durability holds past one year, this becomes a clear position. Corporate Catalysts: Avobis Bio is private; any disclosed financing round, strategic investor, or partnership announcement is a strong signal - a small private sponsor running a key autologous cell therapy trial alone is implausible, and the partner or buyer they pick will tell you how the rest of industry reads the data. Sequencing: (1) Type B meeting outcome, (2) full Phase 2 publication, (3) financing/partnership, (4) Phase 3 IND-enabling design. The clean read on this asset is late 2027 / early 2028 if Phase 3 starts on a reasonable timeline.