Denne siden ble automatisk oversatt og nøyaktigheten av oversettelsen er ikke garantert. Vennligst referer til engelsk versjon for en kildetekst.

Benchmarking Large Language Models Against Tumour Boards for Oncology Treatment Recommendations (BEACON)

28. juli 2026 oppdatert av: Assistance Publique - Hôpitaux de Paris

Benchmarking AI for Clinical Oncology decisioNmaking (BEACON): A Prospective, Multicentre, Blinded Evaluation of Frontier Large Language Models Against Multidisciplinary Tumour Board Recommendations in Oncology Treatment Planning

BEACON (Benchmarking AI for Clinical Oncology decisioNmaking) is a prospective, multicentre, comparative, blinded, non-interventional benchmark evaluating the treatment recommendations of five frontier large language models (LLMs) against the recommendations of multidisciplinary tumour boards (RCP) in oncology treatment planning. One hundred standardised synthetic cases (20 per localisation, across breast, lung, urological, digestive and gynaecological cancers) are submitted as identical structured input to two independent tumour boards per localisation and to five frontier LLMs. Each recommendation - human or model - is decomposed into five predefined decision domains (intent, surgery, radiotherapy, systemic therapy, work-up and biomarkers) and scored 0/1/2 for concordance against a two-tier reference: the consensus of the two tumour boards, complemented by an a priori locked guideline matrix (ESMO, NCCN). The primary endpoint is domain-level concordance between LLM and RCP consensus, expressed as a linearly weighted Cohen's kappa. A co-primary safety endpoint captures the proportion of recommendations carrying serious harm potential, because concordance alone can conceal dangerous errors. Because expert boards may disagree with one another on identical cases, model performance is always interpreted against the human consensus. BEACON is designed as reusable, openly licensed, pre-registered infrastructure: all synthetic cases, evaluation rubrics, the locked guideline matrix, scoring algorithms and verbatim prompts are released for full reproducibility.

Studieoversikt

Status

Rekruttering

Forhold

Intervensjon / Behandling

Detaljert beskrivelse

BEACON is a prospective, multicentre, blinded benchmark using automated, criteria-based scoring. It is built on three design decisions that distinguish it from the existing literature: (i) synthetic, standardised cases remove the record-completeness variability that confounds retrospective comparisons and allow the identical input to be given to every board and every model; (ii) two independent tumour boards per localisation let human-human agreement be measured rather than assumed; and (iii) a guideline matrix, locked a priori, provides an objective anchor applied identically to human and model recommendations.

Reference standard. For each case-domain, a guideline matrix (guideline-recommended / acceptable / unsupported options per case-domain; ESMO, NCCN), locked and time-stamped before data collection, is applied identically to boards and models.

Five decision domains. Every recommendation is decomposed into D1 Intent, D2 Surgery, D3 Radiotherapy, D4 Systemic therapy (class + line), and D5 Work-up & biomarkers before any comparison.

Studietype

Observasjonsmessig

Registrering (Antatt)

100

Kontakter og plasseringer

Denne delen inneholder kontaktinformasjon for de som utfører studien, og informasjon om hvor denne studien blir utført.

Studiekontakt

Studer Kontakt Backup

Studiesteder

  • Frankrike
      • Paris, Frankrike
        • Rekruttering
        • Hôpital Europeén Georges Pompidou
        • Ta kontakt med:

Deltakelseskriterier

Forskere ser etter personer som passer til en bestemt beskrivelse, kalt kvalifikasjonskriterier. Noen eksempler på disse kriteriene er en persons generelle helsetilstand eller tidligere behandlinger.

Kvalifikasjonskriterier

Alder som er kvalifisert for studier

  • Voksen
  • Eldre voksen

Tar imot friske frivillige

Nei

Prøvetakingsmetode

Ikke-sannsynlighetsprøve

Studiepopulasjon

100 synthetic oncology treatment-planning cases (20 per localisation) across five localisations: breast, lung, urological (prostate, bladder / upper-tract urothelial, kidney), digestive and gynaecological. Each case is a structured JSON input specifying UICC 8th-edition stage, biomarkers, ECOG performance status, comorbidities and a standardised clinical question. No human participants, no patient data and no identifiable individuals. Recommendations are produced by two independent tumour boards per localisation and by five frontier LLMs (queried May 2026).

Beskrivelse

Inclusion Criteria:

  • Synthetic oncology case within one of the five predefined localisations (breast, lung, urological, digestive, gynaecological).
  • Complete structured schema: UICC 8th-edition stage, biomarkers, ECOG performance status, comorbidities and a standardised clinical question.
  • A clinically answerable treatment-planning question that is mappable to the locked guideline matrix.

Exclusion Criteria:

  • Case outside the five predefined localisations.
  • Incomplete, internally inconsistent or ambiguous schema.
  • Duplicate or near-duplicate of an existing case in the set.
  • Question not resolvable by current guidelines.

Studieplan

Denne delen gir detaljer om studieplanen, inkludert hvordan studien er utformet og hva studien måler.

Hvordan er studiet utformet?

Designdetaljer

Kohorter og intervensjoner

Gruppe / Kohort
Intervensjon / Behandling
Breast cancers
Annen: Multidisciplinary tumour boards
Two independent tumour boards per localisation (10 boards in total) issue a categorical recommendation for every synthetic case. Where both boards agree, their consensus defines the reference standard; where they differ, the case-domain is classified as EQUIPOISE and analysed separately.
Annen: Frontier large language models
Five frontier LLMs (GPT-5.6, Claude Fable 5, Gemini 3.1 Pro, DeepSeek V4 Pro, Llama 4 Maverick) each receive the identical structured input for every case, three times in independent sessions, under locked prompts, versions and settings.
Lung cancers
Annen: Multidisciplinary tumour boards
Two independent tumour boards per localisation (10 boards in total) issue a categorical recommendation for every synthetic case. Where both boards agree, their consensus defines the reference standard; where they differ, the case-domain is classified as EQUIPOISE and analysed separately.
Annen: Frontier large language models
Five frontier LLMs (GPT-5.6, Claude Fable 5, Gemini 3.1 Pro, DeepSeek V4 Pro, Llama 4 Maverick) each receive the identical structured input for every case, three times in independent sessions, under locked prompts, versions and settings.
Urological cancers
Annen: Multidisciplinary tumour boards
Two independent tumour boards per localisation (10 boards in total) issue a categorical recommendation for every synthetic case. Where both boards agree, their consensus defines the reference standard; where they differ, the case-domain is classified as EQUIPOISE and analysed separately.
Annen: Frontier large language models
Five frontier LLMs (GPT-5.6, Claude Fable 5, Gemini 3.1 Pro, DeepSeek V4 Pro, Llama 4 Maverick) each receive the identical structured input for every case, three times in independent sessions, under locked prompts, versions and settings.
Digestive cancers
Annen: Multidisciplinary tumour boards
Two independent tumour boards per localisation (10 boards in total) issue a categorical recommendation for every synthetic case. Where both boards agree, their consensus defines the reference standard; where they differ, the case-domain is classified as EQUIPOISE and analysed separately.
Annen: Frontier large language models
Five frontier LLMs (GPT-5.6, Claude Fable 5, Gemini 3.1 Pro, DeepSeek V4 Pro, Llama 4 Maverick) each receive the identical structured input for every case, three times in independent sessions, under locked prompts, versions and settings.
Gynaecological cancers
Annen: Multidisciplinary tumour boards
Two independent tumour boards per localisation (10 boards in total) issue a categorical recommendation for every synthetic case. Where both boards agree, their consensus defines the reference standard; where they differ, the case-domain is classified as EQUIPOISE and analysed separately.
Annen: Frontier large language models
Five frontier LLMs (GPT-5.6, Claude Fable 5, Gemini 3.1 Pro, DeepSeek V4 Pro, Llama 4 Maverick) each receive the identical structured input for every case, three times in independent sessions, under locked prompts, versions and settings.

Hva måler studien?

Primære resultatmål

Resultatmål
Tiltaksbeskrivelse
Tidsramme
Domain-level performance between LLM recommendations and the locked guidelines.
Tidsramme: Assessed once at central scoring, after data collection (~October 2026)
For each recommendation domain and each LLM, proportion of LLM recommendation concordant with locked guidelines
Assessed once at central scoring, after data collection (~October 2026)

Sekundære resultatmål

Resultatmål
Tiltaksbeskrivelse
Tidsramme
Proportion of recommendations carrying serious harm potential ( LLM and tumour boards)
Tidsramme: Up to October 2026
Up to October 2026
Domain-level recommendation concordance between LLM and tumour-boards
Tidsramme: Up to October 2026
Each recommendation domain, decomposed into the five decision domains and scored per domain on an ordinal scale (2 = complete concordance; 1 = partial concordance; 0 = discordance).
Up to October 2026
Inter-tumour board domain-level recommendation concordance
Tidsramme: Up to October 2026
Agreement between the two independent tumour boards scored per recommendation domain
Up to October 2026
Equipoise rate
Tidsramme: Up to October 2026
Proportion of case-domains where the two tumour boards give different categorical recommendations
Up to October 2026
Completeness
Tidsramme: Up to October 2026
Proportion of required domains addressed (LLM and tumour boards)
Up to October 2026
Missingness
Tidsramme: Up to October 2026
Proportion of critical omissions (LLM and tumour boards)
Up to October 2026
Intensity bias
Tidsramme: Up to October 2026
Proportion of recommendation corresponding to over- or under-treatment
Up to October 2026

Samarbeidspartnere og etterforskere

Det er her du vil finne personer og organisasjoner som er involvert i denne studien.

Sponsor

Assistance Publique - Hôpitaux de Paris

Studierekorddatoer

Disse datoene sporer fremdriften for innsending av studieposter og sammendragsresultater til ClinicalTrials.gov. Studieposter og rapporterte resultater gjennomgås av National Library of Medicine (NLM) for å sikre at de oppfyller spesifikke kvalitetskontrollstandarder før de legges ut på det offentlige nettstedet.

Studer hoveddatoer

Studiestart (Faktiske)

1. mai 2026

Primær fullføring (Antatt)

1. oktober 2026

Studiet fullført (Antatt)

1. oktober 2026

Datoer for studieregistrering

Først innsendt

28. juli 2026

Først innsendt som oppfylte QC-kriteriene

28. juli 2026

Først lagt ut (Faktiske)

31. juli 2026

Oppdateringer av studieposter

Sist oppdatering lagt ut (Faktiske)

31. juli 2026

Siste oppdatering sendt inn som oppfylte QC-kriteriene

28. juli 2026

Sist bekreftet

1. juli 2026

Mer informasjon

Begreper knyttet til denne studien

Nøkkelord

Ytterligere relevante MeSH-vilkår

Andre studie-ID-numre

  • APHP261032

Plan for individuelle deltakerdata (IPD)

Planlegger du å dele individuelle deltakerdata (IPD)?

UBESLUTTE

Legemiddel- og utstyrsinformasjon, studiedokumenter

Studerer et amerikansk FDA-regulert medikamentprodukt

Nei

Studerer et amerikansk FDA-regulert enhetsprodukt

Nei

Denne informasjonen ble hentet direkte fra nettstedet clinicaltrials.gov uten noen endringer. Hvis du har noen forespørsler om å endre, fjerne eller oppdatere studiedetaljene dine, vennligst kontakt register@clinicaltrials.gov. Så snart en endring er implementert på clinicaltrials.gov, vil denne også bli oppdatert automatisk på nettstedet vårt. .

Abonnere