BioinvestGPT

BioinvestGPT ApS · Research use only

Case study · ATHENA · Ziltivekimab

Before the first patient was enrolled, BVCT called NO-GO on ATHENA — from Drug MoA and the protocol alone.

BVCT locked its NO-GO on ATHENA (No. 1525, ziltivekimab) on , from the drug mechanism and the public registry protocol alone, 8 days before the first participant was enrolled. Novo Nordisk stopped ATHENA on after a data monitoring committee futility finding; the dashboard classifies that readout as True Negative (TN).

Drug MoA + the publicly registered protocol in. NO-GO out. No patient data. None needed.

The inputs on the lock date

On , BVCT locked No. 1525 on ATHENA. This is what the public record showed that day:

What the public record showed
RegistryNCT06200207 · posted
StatusNot yet recruiting
Study start, estimated
PopulationHFmrEF/HFpEF, hsCRP ≥ 2 mg/L
PrimaryKCCQ-CSS at 12 months
Enrolment680, planned
Drug MoAIL-6 inhibitor, monoclonal antibody (as recorded on the dashboard)
ATHENA dataNone — no one enrolled yet

What BVCT called

NO-GO, locked and sent to Novo Nordisk on : commercially insufficient additive benefit to standard of care on KCCQ-CSS and 6MWD.

The recorded calls, verbatim: commercial FAILURE (commercially insufficient additive clinical benefit to SoC with inferior-to-semaglutide efficacy in KCCQ-CSS and 6MWD); technical partial SUCCESS (statistically maybe significant yet weak additive clinical benefit to SoC in KCCQ-CSS and 6MWD). The dashboard classifies the readout as True Negative (TN).

What followed

  1. No. 1525 locked: NO-GO on ATHENA. The registry record live that day said Not yet recruiting, start estimated.
  2. First participant enrolled (registry, actual): 8 days after the lock.
  3. The call sent to Novo Nordisk.
  4. Further development discontinued (DMC futility): 897 days after the lock.

What patient-data-free means here

Study start date, in the registry’s own definition: the actual date on which the first participant was enrolled. ATHENA’s actual start is — after the lock.

NCT06200207 history: version 2 (, Not yet recruiting, start estimated) was live on the lock date; version 4 () posted the actual start.

BVCT simulates the causal mechanism inside virtual patients defined by the actual trial stratum — no patient-level data from the trial being predicted, and none from any prior trial. That is why the call could be made on regardless of the actual patient enrollment status. Nothing in this call depended on privileged access — only on simulating the mechanism based on public information.

The claim is the timestamp: the call existed before its readout.

No. 1525Correct Prediction

ATHENA · ziltivekimab

Sponsor
Novo Nordisk
Phase
Phase 3
Prediction Date
Readout Date
Prediction To Readout
897 days in advance
Prediction Classification
True Negative (TN)
Technical Prediction
partial SUCCESS (statistically maybe significant yet weak additive clinical benefit to SoC in KCCQ-CSS and 6MWD)
Commercial Prediction
FAILURE (commercially insufficient additive clinical benefit to SoC with inferior-to-semaglutide efficacy in KCCQ-CSS and 6MWD)

The ZEUS and HERMES calls locked on the same day are in the ziltivekimab case study.

How early do you want yours?

7-day GO / NO-GO verdict. Drug MoA + the drug to beat in, a Phase-3 head-to-head out — before the first patient is enrolled, first-in-class included.

Capital, years, programs, trial participants — with BVCT, all four get earlier clarity.

Start at bvct.bioinvestgpt.com →

Causal foresight · not statistical hindsight.

All solutionsReturn to the dashboard