MétaCan
Menu
Back to cohort

Pan-tumor harmonization of pathologic response assessment for standardized data collection in neoadjuvant IO trials (PATHdata): Final analysis of a multi-institutional reproducibility study.

2024· article· en· W4399737887 on OpenAlexaff
Julie S. Deutsch, Tricia R. Cottrell, Krista Y. Chen, Carlos E. de Andrea, Ezra Baraban, Pierre Fiset, Jaroslaw Jedrych, Christine Orr, Roberto Salgado, Christian M. Schürch, Richard A. Scolyer, Raja R. Seethala, Lynette M. Sholl, Sabina Signoretti, Michael T. Tetzlaff, Annikka Weissferdt, Xiaowei Xu, James Ziai, Ashley Cimino‐Mathews, Janis M. Taube

Bibliographic record

VenueJournal of Clinical Oncology · 2024
Typearticle
Languageen
FieldMedicine
TopicRadiomics and Machine Learning in Medical Imaging
Canadian institutionsMcGill UniversityQueen's University
FundersSociety for Immunotherapy of Cancer
KeywordsMedicineReproducibilityHarmonizationMedical physicsOncologyInternal medicineNuclear medicineStatistics

Abstract

fetched live from OpenAlex

2515 Background: Immunotherapeutic agents are now being investigated for treating earlier-stage cancers. Radiographic assessment by RECIST, widely used to assess treatment response in clinical trials for advanced cancers, has limitations in the neoadjuvant setting; and pathologic response assessment is increasingly being used as a primary and/or secondary endpoint. To that end, a pan-tumor scoring system for assessing pathologic response was developed (1,2). This scoring system allows for the quantitative assessment of residual viable tumor (RVT) in multiple locations: i.e. primary and lymph node (LN) or distant metastases, akin to RECIST. %RVT scored using this system been associated with patient outcomes after treatment with anti-PD-1-based therapies. Additionally, %RVT in LN has been shown to have additive value to %RVT in the primary tumor when predicting patient survival (3). As a result, pathologists are now being asked to score pathologic response in the primary tumor and LN as a part of ongoing clinical trials and routine clinical care. Here, we evaluated the reproducibility of %RVT scoring using pan-tumor immune-related pathologic response criteria (irPRC). Methods: A multi-institutional, international study led by the Society for Immunotherapy of Cancer was initiated to assess the concordance of pathologic response assessment in resection specimens from patients treated with anti-PD-1-based therapies. Online lecture-based modules for irPRC scoring were developed, and 14 pathologists from multiple institutions, including academic and industry partners, were trained to score H&E-stained slides. The pathologists have scored n=37 pathology cases from resection specimens and on-treatment biopsies from >10 different tumor types, in part derived from phase II/III clinical trials. %RVT in the primary tumor and LN from patient specimens were scored separately (total of n=374 slides scored by each pathologist). Results: At the first interim analysis, scoring of pathologic response using irPRC was shown to be highly reproducible, irrespective of disease location (i.e. primary tumor vs lymph node metastasis). The second half of the study is nearing completion, and these reproducibility numbers will be finalized and presented in the final abstract. Extended analyses will also be presented that include subset analyses by tumor type. Conclusions: The results will be interpreted and presented in the context of the larger field for pathologic response assessment. A post-study survey completed by the participating pathologists will be used to refine irPRC training materials prior to dissemination to the wider immuno-oncology community. 1. Cottrell et al. Ann Oncol2018. 2. Stein et al. Clin Can Res 2020. 3. Deutsch, et al. Nat Med 2023.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.

How this classification was reachedexpand

Full frame machine prediction

Teacher imitation

Not calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.

metaresearch head score (Codex)0.336
metaresearch head score (Gemma)0.355
Version: metacan-v3-hybrid-931329e0061cValidation status: machine_predicted_unvalidated
Candidate categoriesMetaresearch
Consensus categoriesMetaresearch
DomainCandidate signal: Reproducibility · Consensus signal: none
Study designCandidate signal: Observational · Consensus signal: Observational
GenreCandidate signal: Empirical · Consensus signal: Empirical
Teacher disagreement score0.664
Threshold uncertainty score0.819

Distilled classifier scores by category (both heads)

CategoryCodexGemma
Metaresearch0.3360.355
Meta-epidemiology (narrow)0.0010.001
Meta-epidemiology (broad)0.0030.005
Bibliometrics0.0040.007
Science and technology studies0.0020.002
Scholarly communication0.0040.002
Open science0.0030.007
Research integrity0.0010.002
Insufficient payload (model declined to judge)0.0060.001

Machine scores (provisional)

The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.

Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.

Opus teacher head0.384
GPT teacher head0.591
Teacher spread0.207 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it

Classification

machine, unvalidated

Machine predicted; the direct Gemma label and the distilled Codex classifier agree on what is shown here.

Study designObservational
DomainReproducibility
GenreEmpirical

How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".

Quick stats

Citations1
Published2024
Admission routes1
Has abstractyes

Explore more

Same venueJournal of Clinical OncologySame topicRadiomics and Machine Learning in Medical ImagingFrench-language works237,207