A cited biomarker strategy for every programme, with each gap owned and dated before the board meets
A cited biomarker strategy for every programme in days, with each gap owned and dated before the board meets.
See one case, screen by screen ↓Before a programme moves to its next stage gate, the translational board wants one document: which biomarkers show target engagement, which show the biology changing downstream, which decide who is enrolled, and which watch for harm — each with an assay that is fit for its context of use, and a route to the regulator.
The evidence for that document is scattered. Performance numbers sit in papers and preprints, the Phase 1 biomarker report and the lab's stability study sit in the data lake and the lab notebook, competitor cut-offs sit in registry records and congress posters, and validation status sits in vendor specifications and qualification records. A translational lead spends weeks pulling it together by hand, and the gaps — an assay that failed stability, a pre-screen whose device route is not documented — are often found by the board, not before it.
Estimated split for one programme’s strategy, by hand and with the solution. Most of the time that remains is people checking values and co-signing.
Four readers gather the evidence in parallel, a writer drafts and a reviewer challenges the draft; a scientist checks the values and the translational board signs.
NWB-3140, an anti-amyloid antibody in early Alzheimer’s disease, is moving from Phase 1b to Phase 2 and goes to the translational board on Oct 21. Here is its biomarker strategy, screen by screen, in the working solution.
Maya opens the solution to 10 programmes and 52 biomarkers: 3 strategies for the Oct 21 board, 13 biomarkers with open gaps, 9 values waiting for a check. The readiness ladder lays out NWB-3140’s nine biomarkers by role — target engagement, pharmacodynamic, patient selection, safety — and by how ready each one is, from candidate to regulatory precedent.
Needs you: “NWB-3140 biomarker strategy v3 — All four co-signs are in · send to the board by Oct 14.”
The evidence watch has seen that the leading competitor Phase 3, HLV-210, now lists plasma GFAP at week 78. Maya refreshes the dossier. The literature scout reads 64 full texts, 2 new since v3; the internal data reader takes the Phase 1b biomarker report, 4 lab reports and stability study SS-118; the trial design analyst compares 14 competitor designs; the assay and precedent reader sets 9 validation statuses against the 2026 FDA biomarker validation guidance.
Scientist check: “Values at or above 0.90 already confirmed by Sam Patel · 2 routed to him.”
Version 4 lands in the eight house sections — context of use, target engagement, pharmacodynamics, patient selection, safety, assay development plan, regulatory route, gap analysis — with 114 cited claims from 49 sources and 0 uncited claims. Each section shows its status — checked by Sam Patel, co-signed by Lena Ortiz, or drafted or updated by an agent. Pharmacodynamics is marked “Updated by agent”, with the new GFAP finding added and cited.
“The leading competitor Phase 3 now also measures plasma GFAP at week 78, which raises the value of fixing GFAP stability.”
In patient selection, the pre-screen claim cites a 1,042-participant memory-clinic cohort. One click opens the paper at the passage, with the numbers highlighted: AUC 0.93 (95% CI 0.91–0.95) against amyloid PET, 91% sensitivity, 89% specificity with two cut-offs, and 18% of results in an intermediate zone.
“Values from this passage feed Plasma p-tau217 (NWB-3140).”
Plasma p-tau217 sits at rung 5, regulatory precedent: the same analyte was cleared as a blood test in 2025 for cognitive decline. Clinical and analytical performance follow — LLOQ 0.03 pg/mL, precision ≤ 6.2% intra- and ≤ 9.8% inter-assay, 3 freeze–thaw cycles and 24 months at −80 °C — each value extracted with its confidence and checked by Sam. Values below 0.90 wait for him.
NWB-3140-201 side by side with HLV-210: the same plasma p-tau217 pre-screen, but two cut-offs where they use one; intermediate results go to CSF or PET where they are excluded; APOE ε4 homozygotes included with extra MRI; GFAP exploratory because of the stability gap, where they measure it at weeks 26, 52 and 78; a tau PET sub-study at 9 sites against their n = 400.
Public sources only, with the registry update date recorded.
The reviewer reads the draft like a board member. Plasma GFAP recovered 84% after 9 months at −80 °C against an 85% criterion, and Phase 2 samples will be stored up to 18 months — re-validate with stabiliser tubes, Sam Patel, Nov 30. The p-tau217 pre-screen decides who proceeds to PET, so its device route needs documenting — Omar Haddad, Nov 14. Tau PET tracer at only 9 of 40 sites — Kenji Mori, Dec 15.
Only a person can accept or close a gap.
The assay development plan runs to first participant in, May 2027: the CLIA transfer of the p-tau217 pre-screen at Kestrel Assay Labs, two-platform bridging in 120 Phase 1b samples and GFAP re-validation at Tamsin Bioanalytical, APOE genotyping set-up, a partial validation of the CSF drug assay, and the central imaging charter at Lumen Imaging Core.
Critical path: “the CLIA transfer of the p-tau217 pre-screen must finish before the protocol is final (Feb 15); GFAP re-validation ends in mid-March, just before the sample-collection plan is locked — two months before first participant in.”
Maya signed as author on Oct 6 at 16:42. Sam confirmed the extracted values at 09:15 the next morning; Lena Ortiz, the clinical pharmacologist, co-signed target engagement at 11:02; Omar Haddad, the diagnostics lead, co-signed patient selection at 13:30. With all co-signs in, Maya sends the strategy to the translational board.
“All co-signs are in. Send the strategy to the translational board — the pack closes Oct 14.”
Dana chooses between approve, approve with conditions and return for rework. The conditions come pre-filled from the two High gaps — re-validate GFAP with stabiliser tubes or analyse within 6 months and keep it exploratory; document the risk determination and complete the CLIA transfer by January. She signs with her password, and the decision is recorded on the strategy.
“Electronic signature · meaning: approval of the biomarker strategy · recorded in the audit trail.”
The same strategy, seen by the five people who write, check, co-sign and sign it.
Four readers gather the evidence in parallel, a writer drafts and a reviewer challenges the draft; a scientist checks the values and the translational board signs.
Finds and reads papers and preprints for each candidate biomarker and extracts sensitivity, specificity, AUC, cut-offs and cohort details.
Reads your own Phase 1 biomarker reports, laboratory validation reports and stability studies from the data lake and lab notebooks.
Reads competitor trial registry records and congress posters and records how each biomarker is used — pre-screen, cut-off, endpoint — then compares it with your design.
Reads vendor specifications, validation reports and regulatory records, and sets the validation status and regulatory precedent for each biomarker against its context of use.
Drafts the biomarker strategy in the house template, section by section, with every sentence cited.
Reviews the draft like a board member: unsupported or single-source claims, missing validation, lead times and site feasibility — each gap with a severity, an owner and a date.
Signs, with conditions on record. The agents propose; a named person decides.
Anyone on the team can ask in plain words — or tell it what to change. Answers cite their sources.
Why is GFAP a gap?
Plasma GFAP recovered only 84% after 9 months at −80 °C, below the 85% acceptance criterion, and Phase 2 samples will be stored up to 18 months. It matters more now: the leading competitor Phase 3 measures GFAP at weeks 26, 52 and 78. Proposed: re-validate with stabiliser tubes — owner Sam Patel, due Nov 30.
Which programmes lack a validated patient-selection assay?
Three: Veltrimab (c-Met IHC — trial assay, companion test not contracted), NWB-8840 (B7-H3 IHC — assay not validated) and NWB-9152 (interferon signature — exploratory only). Veltrimab is the urgent one: co-development needs about 24 months and Phase 3 starts Q4 2027.
What does the CDx route need?
A test that decides who is treated is a medical device. NWB-3140 needs a CLIA lab and a documented device risk determination for the p-tau217 pre-screen; Veltrimab needs a companion IHC co-developed with the drug — about 24 months from contract, and in the EU a class C device under IVDR. Omar Haddad owns both.
Add a rule: flag any key claim supported by only one source
Done — the rule is on in Settings → Evidence rules and applies to every draft. In the NWB-3140 strategy it flags 3 claims: the CSF:serum ratio, tau PET at 9 of 40 sites, and NfL showing no treatment difference. The literature scout will look for a second source for each on its next run.
13 screens from the working solution, on its sample data. Pick one to see it large.
The template, the evidence rules the agents follow on every run, and who signs what are settings.
| Setting | Default | Choose from |
|---|---|---|
| House template | Biomarker strategy v4 — 8 sections | v4 — 8 sections · v3 — 6 sections · Early-development plan — 5 sections |
| Citation style | Numbered | Numbered · Author–year |
| Independent sources per key claim | 2 | Any number; claims with fewer are flagged |
| Flag single-source claims | Off | On · Off |
| Scientist check below this confidence | 0.90 | Any threshold |
| Recency window | 5 years | 3 · 5 · 10 years |
| Preprints | Allowed, always labelled | On · Off |
| Biomarker values confirmed by | Sam Patel | Sam Patel · Priya Raman |
Every figure is labelled: a target the solution is built to, an estimate, a typical published result, or a proven one.
“demo” = seen in the working solution, on its sample programmes · “target” = the design goal, measured in the live solution · “estimated” = our estimate · “typical” = published figures (lack of efficacy ≈ 40–50% of clinical failures). Context: FDA biomarker assay validation guidance (2026), FDA–NIH BEST. People, programmes, companies and laboratories named on this page are characters in the working solution.
The document a translational board signs before a programme’s next stage gate: the target-engagement, pharmacodynamic, patient-selection and safety biomarkers, the assay for each and its validation status, the regulatory route, and the gaps. This solution drafts it in your house template, with a biomarker evidence table and an assay development plan alongside.
Each biomarker sits on a five-rung ladder: candidate, assay available, fit-for-purpose validated, clinically shown, regulatory precedent. The assay and precedent reader sets the validation status and regulatory precedent behind each rung from vendor specifications, validation reports and regulatory records, applying the context-of-use principle of the FDA biomarker validation guidance.
The agents extract sensitivity, specificity, AUC, LLOQ, precision and stability from papers, your internal reports and vendor specifications, exactly as written, each with a confidence and its source passage. Values below the threshold — 0.90 by default — wait for a biomarker scientist, and the strategy uses only checked values.
It flags any test that decides treatment as device-relevant and writes the regulatory route: for example, a CLIA laboratory and a documented device risk determination for a pre-screen. It never claims a qualification without a record, and cites existing qualifications and cleared tests where they exist.
Yes. The trial design analyst reads registry records and congress posters, records each biomarker’s use — selection, enrichment, stratification, endpoint, safety — with cut-offs, and compares it with your design. A weekly evidence watch re-reads new papers and registry changes and tells the owners what changed.
Yes. A scientist checks the values, section owners co-sign, and the translational board chair signs or returns the strategy with an electronic signature. The reviewer proposes gaps, but only a person can accept a risk or close a gap, and every step is on the activity trail.
Yes. The house template, citation style, export format, sources needed per key claim, the single-source flag, the confidence threshold for a scientist check, the recency window, preprints and who co-signs and signs are all settings.
The Agentic Solution Engine builds and deploys it from your requirements — your house template, your sources and a past strategy or two — and it goes live once every quality gate has passed. We will walk you through it on one of your own programmes first.
Ask the literature in plain words and get a sourced answer in seconds.
1-min filmPipeline WatchEvery rival trial, readout and approval lands in a sourced brief for each programme, every Monday.
1-min filmPreclinical ReportsRaw study data becomes a draft report in hours, with every number traced to its source.
1-min filmWe’ll build a Biomarker Strategy Dossier for one of your own programmes.