Zosurabalpin attacks drug-resistant Acinetobacter through a new target. The test is whether that mechanism survives a pivotal trial.
25%likely
Well sourcedNext check
Watching
Why it is different
The opportunity depends on commercial neglect as well as the science: antibiotic economics can suppress attention even when a mechanism is novel.
Why it matters
Drug-resistant infection is a slow-moving crisis and the pipeline has been close to empty for decades.
Next check · 24 Oct 2026
Does Roche present zosurabalpin results from infected patients at IDWeek in October 2026?
How we scored it ·
from how often this kind of thing happens
If right, how big4/5How overlooked4/5Upside vs downside2/5Source trust4/5Independent sources3/5
The exact wording we score
On or before 31 Dec 2030, zosurabalpin reports a positive pivotal Phase 3 against carbapenem-resistant A. baumannii, meeting its pre-specified primary mortality or clinical-cure endpoint at non-inferiority-or-better versus standard of care. Resolves YES only on that qualifying result; NO on non-initiation, enrolment failure, endpoint miss, safety halt, deprioritisation, or no qualifying readout by the deadline. No other antibiotic can be substituted.
PROOF
Evidence and provenance
The links below are the evidence recorded when this forecast was scored. Each note states source quality or bias; a citation is not an endorsement.
Recorded sources · 4 citations / 3 independent origins
2026-07-28 | sell-side: little economically material coverage for Roche; mainstream press: novel mechanism is treated mainly as a science story; priced market: no clean instrument identified; specialist: infectious-disease outlets follow the clinical programme but registration remains an open gap | E4
Revision history5 entries
Probabilities and interpretations are never silently overwritten. This is the append-only record of what changed and why.
No second asset was actually named, so the “set” remained open to future enumeration. The claim returns to the only verified program, zosurabalpin, with an exhaustive outcome rule; P 30→25.
(3rd pass) — Claim raised to class level; unfalsifiable clause deleted; P 25→30. The edge is a claim about a category — that the market prices 'antibiotics don't matter commercially' and so ignores a new gram-negative mechanism as a near-term clinical event. Scoring one Roche asset's endpoint tested almost none of it. Also removed 'and is on track for regulatory filing/approval by 2030', which had no observable and could not be falsified as written; the falsifier now explicitly covers non-initiation and enrolment failure, which were silent failure modes. P moves only 25→30 — zosurabalpin is far ahead and no second qualifying asset was verified (~1.1 effective assets).
(2nd pass) — P 38→25. The claim is a three-way conjunction that priced timing at zero: (A) the pivotal enrols ~400 CRAB patients and reads out in time — but this file's own checkpoint puts only 72% on the trial being registered and recruiting by April 2027, in a population it calls 'operationally brutal' to enrol; (B) the endpoint is met, from a program that skipped Phase 2 efficacy de-risking; (C) 'on track for regulatory filing/approval by 2030', which has no observable and is unfalsifiable as written. 0.60 x 0.55 x 0.80 ≈ 0.26. For comparison, efruxifermin sits at 40 with a pivotal that is already enrolling — an unregistered trial was being priced 2 points below a running one. Impact held at 4: the anchor's 'major therapeutic class' clause carries it (first new-mechanism gram-negative class in decades), even though this file's own edge argument correctly denies that a ~$100B commercial subsector exists. Reclassified to refclass: phase3-attrition.
v2 rescore. P held at 38, trust 5→4 on a verification gap. No registered Phase 3 zosurabalpin trial could be located on ClinicalTrials.gov as of 2026-07-28, despite the May-2025 Roche announcement of a ~400-patient global Phase 3 to start 'late 2025 or early 2026'. Roche's HY-2026 results (2026-07-23) make no mention of zosurabalpin, antibiotics or Acinetobacter. Neither is proof of a problem — Roche discloses ID assets sparsely and the trial may be registered under a title not surfaced — but this file's premise that a global Phase 3 launched is currently unconfirmed in the public registry. The live uncertainty is therefore 'is the Phase 3 actually running', not 'will it succeed'. Sourcing fix: PubMed was being counted as a second source for the same Nature paper — indepsources 4→3.
created from fresh research sweep. Mechanism peer-reviewed (Nature); the Phase-1→Phase-3 jump is the main execution risk.
Method challenge
Adversarial review record
Conditionally admitted
Structured internal role review tied to this frozen claim. It is not independent human peer review. Independent specialist review not yet performed
clinical-domain specialist
biostatistics and trial-design reviewer
regulatory reviewer
forecasting-method reviewer
Strongest specialist challenge
The program appears to jump from Phase 1 into a difficult global pivotal population, and no Phase 3 registry record was found at review. Mechanistic novelty and urgent-pathogen status do not de-risk enrolment, comparator heterogeneity, mortality, or safety.
Outside view
Novel-antibiotic Phase 3 success is only a partial reference class because this program lacks conventional Phase 2 efficacy de-risking and targets critically ill CRAB patients. Schedule and trial execution lower the forecast to roughly one quarter.
Causal rival
Diagnostic delay, pathogen heterogeneity, changing standards of care, site selection, broken antibiotic economics, or resistance emergence can stop the program even if the LPS-transport mechanism is valid in vitro and in animals.
Measurement risk
A sponsor announcement is not a registered pivotal and non-inferiority depends on the pre-specified margin, population, and endpoint. Adjudication must use the registered primary mortality or clinical-cure analysis, not microbiologic clearance or a favorable subgroup.
Residual risk
Until a protocol is public, sample size, primary endpoint, non-inferiority margin, and schedule are not independently checkable. Conditional admission preserves the forecast while making that missing evidence a hard governance gate.
Admission condition
A Roche-sponsored pivotal Phase 3 protocol with a named primary endpoint must appear in a recognized public trial registry by 2027-04-30; otherwise void or sharply re-score the parent call.
COLDTELL