# EA2251 (ivonescimab): what clinicians are saying

> 28 statements from 12 sources (1 video, 11 posts), 4 of them from 3 confirmed clinicians, each linked to its source. Commentary checked Nov 22, 2026.

## Report

- Trial: A Phase II/III Randomized Study of Ivonescimab With Gemcitabine and Cisplatin Versus Standard of Care Chemoimmunotherapy in Advanced Biliary Tract Cancer
- NCT ID: NCT07765433
- Registry: https://clinicaltrials.gov/study/NCT07765433
- Sponsor: ECOG-ACRIN Cancer Research Group
- Phase: Phase 2/Phase 3
- Status: Not yet recruiting
- Drugs tested: ivonescimab
- Indication: advanced biliary tract cancer
- Version: v1.0.0, edition 1
- Accepted: 2026-11-22 UTC
- Last checked: 2026-11-22 UTC
- Release hash: 8acb8d077dad8f4b454d4ee35d2d2e11c2ac809bfdadf963ca1e080ced8fc1a7

## Counts

- Commentary sources screened: 29
- Sources cited by verified statements: 12
- Verified statements: 28
- From speakers with a confirmed clinical role (verified by the NPI registry or their own institution's staff page): 4
- Findings: 6
- Named people in screened sources (unnamed speakers and accounts not counted): 6
- Speakers cited: 5
- Cited speakers with a verified identity: 4
- Cited clinicians, speakers with a confirmed clinical role (the abstract's clinicians): 3
- Of them physicians, whose confirmed role is a physician's, not a physician assistant's, nurse's or pharmacist's (the abstract's physicians): 3
- By role and specialty (the abstract's specialty counts):
  - 3 physician, Medical oncology
- Organisational or shared accounts (not counted as people): 4

A verified identity says who a speaker is. A cited clinician also has a clinical role confirmed by the NPI registry or their own institution's staff page. Verified speakers without such a role, such as researchers, advocates, company staff, or clinicians whose role only a source or their own profile states, are not counted as clinicians.

Counts describe the retained sources. They do not estimate how many HCPs hold a view.

## Abstract

A medical oncologist looks to EA2251 for clarity on the bispecific’s contribution because HARMONi-GI1 lacked a PD-L1-plus-bevacizumab arm, though that capability is unproven. Direct commentary is limited; HARMONi-GI1 provides context, not outcomes from the US study. Its reported survival benefit is qualified by accounts of greater high-grade toxicity. In that context, the same oncologist questions another medical oncologist’s unconditional prediction of pan-tumor blockbuster status for ivonescimab, citing interim evidence. Final survival and quality-of-life data remained pending.

## Question clinicians left open

What evidence do clinicians say they need on ivonescimab’s bispecific contribution and applicability beyond China before using it with chemotherapy instead of durvalumab plus chemotherapy for first-line advanced biliary tract cancer?

From finding F1 (7 statements from 3 speakers in 6 sources). The report shows what clinicians said and the sources behind it; it does not settle the question.

## Findings (titles)

- F1. GI1: most speakers value active control; a clinician asks what the bispecific adds
  - 7 statements from 3 speakers in 6 sources, 1 of them a podcast or video.
  - 2 of the 3 named speakers in this finding hold this view; 1 of them holds its caveat, “a clinician asks what the bispecific adds”.
  - 1 of the report's 3 cited clinicians.
- F4. EA2251: what speakers say
  - 8 statements from 1 speaker in 5 sources, 1 of them a podcast or video.
  - No named clinician · accounts and unnamed speakers.

### Beyond the trial population

- F2. Ivonescimab in HARMONi-GI1: what speakers say
  - 5 statements from 1 speaker in 5 sources, 1 of them a podcast or video.
  - No named clinician · accounts and unnamed speakers.
- F3. HARMONi-GI1 evidence: where speakers split on scrutiny of China-only interim data
  - 6 statements from 3 speakers in 6 sources, 1 of them a podcast or video.
  - 3 of the 3 named speakers in this finding take a side on this question: 2 on one side, 1 on the other.
  - 1 of the report's 3 cited clinicians.
- F5. Ivonescimab: where speakers split on pan-tumor expectations after interim GI1
  - 4 statements from 3 speakers in 3 sources.
  - 2 of the 3 named speakers in this finding take a side on this question: 1 on one side, 1 on the other.
  - 2 of the report's 3 cited clinicians.
- F6. OS hazard-ratio rankings: why a speaker weighs cancer context and options
  - 1 statement from 1 speaker in 1 source.
  - The one named speaker in this finding holds this view.
  - No named clinician · accounts and unnamed speakers.

4 of the 6 findings and 10 of the 28 statements are about other patients, lines of treatment or settings. The report lists them after the trial findings under “Beyond the trial population”, as this page does.

## What sits behind each finding

Counts describe what the report retained, not how many clinicians hold a view. A statement can support more than one finding. Speakers are the named people a finding's statements cite, of any role, counted as the report counts its cited speakers; unnamed speakers and accounts are not counted.

| Finding | Statements | Passages | Sources | Speakers | Limitations |
| --- | ---: | ---: | ---: | ---: | ---: |
| F1 | 7 | 6 | 6 | 3 | 2 |
| F4 | 8 | 6 | 5 | 1 | 2 |
| F2 | 5 | 5 | 5 | 1 | 3 |
| F3 | 6 | 6 | 6 | 3 | 2 |
| F5 | 4 | 4 | 3 | 3 | 2 |
| F6 | 1 | 1 | 1 | 1 | 1 |

### What the statements are

| Finding | Views | Questions and uncertainty | Clinical experience | Comparisons | Restated results and context | Statement links to supporting passages | Statement links to context passages | Sources by kind |
| --- | ---: | ---: | ---: | ---: | ---: | ---: | ---: | --- |
| F1 | 3 | 1 | 0 | 0 | 3 | 7 | 1 | 5 posts and articles, 1 podcast or video |
| F4 | 5 | 0 | 0 | 0 | 3 | 10 | 0 | 4 posts and articles, 1 podcast or video |
| F2 | 1 | 0 | 0 | 1 | 3 | 6 | 2 | 4 posts and articles, 1 podcast or video |
| F3 | 2 | 1 | 0 | 0 | 3 | 7 | 0 | 5 posts and articles, 1 podcast or video |
| F5 | 4 | 0 | 0 | 0 | 0 | 5 | 0 | 3 posts and articles |
| F6 | 1 | 0 | 0 | 0 | 0 | 1 | 0 | 1 post or article |

### Trial dimensions

- Primary endpoint: Overall survival (OS): F1 (4 shared claims), F4 (1 shared claims)
- Secondary endpoints: F1 (2 shared claims), F4 (1 shared claims)
- Population and setting: F1 (4 shared claims), F4 (3 shared claims)
- The comparison: F1 (5 shared claims), F4 (1 shared claims)
- Safety and tolerability: F1 (1 shared claims), F4 (1 shared claims)
- Design and conduct: F1 (2 shared claims), F4 (4 shared claims)

## Abbreviations

- HCP: healthcare professional
- MCP: Model Context Protocol, how AI agents read the report
- NPI: National Provider Identifier, the US registry number for clinicians
- OS: overall survival
- PD-L1: programmed death-ligand 1

## What's inside

| Section | Items | Access |
| --- | ---: | --- |
| Findings with interpretation, counterviews and limits | 6 | Titles free |
| Detailed analysis | 8 sections | Paid |
| Verified statements with short source quotes and links | 28 | Paid |
| Who said each statement: the cited speakers | 5 | Every edition |
| Speakers with a clinical role confirmed by the NPI registry or a staff page | 3 | Every edition |
| Speakers with verified IDs as data (NPI, ORCID, staff page) | 4 | Every edition |
| CMS-reported financial relationships (Open Payments, by company) for NPI-linked speakers | Included | Every edition |
| Trial aspects mapped | 11 | Free |
| Sources with links | 12 | Counts free |
| MCP access for your AI agent, same license key | Included | Every edition |

Every edition, the free report included, names who said each statement, as the source shows them. Each named speaker's card gives a clinical role only when the NPI registry or the person's own institution's staff page confirms it, and says which; a role only a source states is labelled as the source's. Here 3 of the 5 cited speakers have a role confirmed that way. Every edition also comes with MCP access, so your AI agent reads the same evidence with your license key, and with verified IDs for each identity-checked speaker (the NPI, ORCID and OpenAlex IDs, and their institution's staff page), so your team can check each attribution itself. Social accounts are not included. The IDs are for checking attribution inside your organization. Allowed: advisory boards, scientific exchange and finding investigators, under your own compliance process. Not allowed: sales or marketing to clinicians.

## Trial aspects

11 discussed · 0 no commentary found · 0 assessment incomplete.

| Aspect | Coverage | Statements | Sources |
| --- | --- | ---: | ---: |
| Primary endpoint: Overall survival (OS) | discussed | 13 | 10 |
| Secondary endpoints | discussed | 6 | 5 |
| Population and setting | discussed | 12 | 9 |
| The comparison (The registered arms and what clinicians said about how they compare.) | discussed | 7 | 5 |
| Safety and tolerability | discussed | 5 | 5 |
| Design and conduct | discussed | 5 | 5 |
| Evidence maturity and additional information requested | discussed | 6 | 6 |
| Country-of-origin scrutiny and access | discussed | 1 | 1 |
| Regulatory status, local availability and patient counseling | discussed | 6 | 3 |
| Pan-tumor extrapolation and class-level interest | discussed | 3 | 3 |
| Cross-trial rankings and clinical value | discussed | 1 | 1 |

Discussed means at least one verified statement addresses the aspect, not that every question in it is covered: gaps inside a discussed aspect are listed under Scope and limits, and the full report gives each aspect its own limitations. No commentary found means none in the sources read. Assessment incomplete marks an aspect with no statement found while known source limits apply.

## Sources

| Channel | Statements | Cited | Screened |
| --- | ---: | ---: | ---: |
| **Video & audio · 7 statements · 25%** | | | |
| YouTube | 7 | 1 | 1 |
| **Posts · 21 statements · 75%** | | | |
| X threads | 18 | 9 | 20 |
| LinkedIn posts | 2 | 1 | 1 |
| Bluesky threads | 1 | 1 | 7 |

- Screened window: 2026-08-16 to 2026-11-10
- Cited window: 2026-08-26 to 2026-11-10
- Retrieved: 2026-11-22

Statements count each verified statement once, split evenly over the sources it cites. Screened sources were retrieved for this trial. Cited sources back at least one verified statement and include primary sources used for context.

## The trial

From ClinicalTrials.gov.

- Title: A Phase II/III Randomized Study of Ivonescimab With Gemcitabine and Cisplatin Versus Standard of Care Chemoimmunotherapy in Advanced Biliary Tract Cancer
- Registry: NCT07765433 (https://clinicaltrials.gov/study/NCT07765433)
- Study IDs: EA2251 · NCI-2026-04569 (CTRP (Clinical Trial Reporting Program)) · U10CA180820
- Sponsor: ECOG-ACRIN Cancer Research Group
- Collaborators: National Cancer Institute (NCI)
- Design: Phase 2/Phase 3 · interventional, randomized, parallel, phase2, phase3
- Status: Not yet recruiting
- Enrollment: 336 (estimated)
- Start: 2027-01-11 (estimated)
- Primary completion: 2028-12-31 (estimated)
- Primary endpoint: Overall survival (OS) (From date of randomization until the date of death from any cause, assessed up to 3 years)
- Intervention: Biospecimen Collection (procedure)
- Intervention: Cisplatin (drug)
- Intervention: Computed Tomography (procedure)
- Intervention: Durvalumab (biological)
- Intervention: Gemcitabine Hydrochloride (drug)
- Intervention: Ivonescimab (biological)
- Intervention: Magnetic Resonance Imaging (procedure)
- Intervention: Pembrolizumab (biological)

## Scope and limits

This report does not judge efficacy, estimate HCP prevalence or consensus or give treatment advice.

- Discussion that does not name the trial may be missing from this edition.
- 24 statements in the sources reviewed for this edition are about other patients or topics, or only mention the trial, and are not counted.
- Further source limits: 34 sources about other trials or topics are not included.
- Most substantive efficacy and safety commentary concerns HARMONi-GI1 as evidence for the same first-line decision, not outcomes observed in EA2251.
- No EA2251 protocol or publication reporting its outcomes was identified in the sources reviewed for this edition. The comments therefore cannot independently resolve expectations about which treatment contribution EA2251 could isolate.
- The sources include corporate communication and repeated summaries of release-derived information. Their number cannot be interpreted as a number of independent HCP opinions or as the prevalence of a view.
- Only part of the relevant regulatory material was reviewed. Approval, availability and development forecasts remain dated source statements rather than a complete or current regulatory assessment.

How reports are built and checked: https://www.erudio.com/methods

## Editions and corrections

- v1.0.0, edition 1, accepted 2026-11-22 UTC: current edition

Each new edition gets a new version: a new or changed finding raises the middle number (1.0.0 to 1.1.0), added evidence or newly verified speakers alone the last number (1.0.0 to 1.0.1), and a change of report format the first number. A correction fixes the published edition in place without making a new edition, so the version stays the same. Changes are corrections that add, remove or change a statement or finding, or withdraw a source; routine maintenance (people cards, names and spellings, recounts, tags and page wording) changes none. Every correction is listed here with its date. Last checked for new commentary 2026-11-22 UTC. No changes.

## Offer

- Size: Focused · 28 verified statements · 4 from 3 confirmed clinicians
- Current edition: $750. A purchase starts a check for new commentary and any newer edition follows at no charge
- 3 months of updates: $1,300
- 6 months of updates: $1,900
- 12 months of updates: $2,600
- Every product includes verified IDs for each identity-checked speaker (the NPI, ORCID and OpenAlex IDs, and their institution's staff page), so your team can check each attribution itself (4 identity-verified speakers now); social accounts are not included. Allowed: advisory boards, scientific exchange and finding investigators, under your own compliance process. Not allowed: sales or marketing to clinicians.
- Pricing: https://www.erudio.com/pricing
- Buy now (card, through Stripe): https://www.erudio.com/reports/ea2251-nct07765433
- Agents: POST https://www.erudio.com/api/orders with {"report": "ea2251-nct07765433", "product": "edition" | "updates-3" | "updates-6" | "updates-12", "buyer_email": "...", "organization": "...", "licence_key": "<your key, if you hold one>"}; it returns a payment link for your user and a pay_url your own payment client can pay over HTTP 402 in USDC, with MPP (on Tempo) or x402 (on Base), and once paid the status_url returns your license key. Send the same key with every later order so all reports sit on one key. The buyer gets their account key by email, which reads every report bought with their address; recover_key (POST https://www.erudio.com/api/keys/recover) sends it again
- Talk to Audun Utengen and Tom Lee (invoice, PO, questions): https://www.erudio.com/contact?report=ea2251-nct07765433

## Cite

Erudio Health. EA2251 (NCT07765433): HCP commentary report. Edition 1, v1.0.0. Accepted Nov 22, 2026. https://www.erudio.com/reports/ea2251-nct07765433

## Questions

### How much does the EA2251 report cost?

One edition costs $750 (Focused · 28 verified statements · 4 from 3 confirmed clinicians). With updates: 3 months $1,300, 6 months $1,900, 12 months $2,600. Subscriptions are prepaid, and the price is fixed for the term you buy, even if a new edition moves the trial up a size. Nothing renews automatically: a new term is bought at the price then current.

### What does the EA2251 report cover?

28 statements from 12 sources (1 video, 11 posts), 4 of them from 3 confirmed clinicians, each linked to its source. Commentary checked Nov 22, 2026. The report separates what clinicians said from what primary sources establish; it does not judge efficacy or estimate consensus.

### Can my AI agent read or buy the EA2251 report?

Yes. On https://mcp.erudio.com/free, with no sign-in, get_report_card({ "slug": "ea2251-nct07765433" }) previews it and create_order({ "report": "ea2251-nct07765433", "product": "edition", "buyer_email": "<work email>" }) orders it. Once paid, the license key reads it with read_report on https://mcp.erudio.com/mcp. [Connect your AI client](https://www.erudio.com/developers/connect)

### How current is the EA2251 report?

Edition 1 (version 1.0.0), accepted Nov 22, 2026 UTC; last checked for new commentary Nov 22, 2026 UTC. Your plan includes every new edition published during its term. We automatically check subscribed trials for new commentary every week.

## Machine access

- JSON: https://www.erudio.com/reports/ea2251-nct07765433.json
- Markdown: https://www.erudio.com/reports/ea2251-nct07765433.md
- Change feed (JSON Feed: new editions and changes): https://www.erudio.com/reports/ea2251-nct07765433/feed.json
- MCP, no sign-in: https://mcp.erudio.com/free, tool `get_report_card` with `{"slug": "ea2251-nct07765433"}`
