CoChat

Development and validation of deep learning classifiers to detect Epstein-Barr virus and microsatellite instability status in gastric cancer: a retrospective multicentre cohort study

Open access

The DOI matches its Crossref record; 45 of the 45 references it cites were found in a registry or an index.

Abstract

BACKGROUND: Response to immunotherapy in gastric cancer is associated with microsatellite instability (or mismatch repair deficiency) and Epstein-Barr virus (EBV) positivity. We therefore aimed to develop and validate deep learning-based classifiers to detect microsatellite instability and EBV status from routine histology slides. METHODS: In this retrospective, multicentre study, we collected tissue samples from ten cohorts of patients with gastric cancer from seven countries (South Korea, Switzerland, Japan, Italy, Germany, the UK and the USA). We trained a deep learning-based classifier to detect microsatellite instability and EBV positivity from digitised, haematoxylin and eosin stained resection slides without annotating tumour containing regions. The performance of the classifier was assessed by within-cohort cross-validation in all ten cohorts and by external validation, for which we split the cohorts into a five-cohort training dataset and a five-cohort test dataset. We measured the area under the receiver operating curve (AUROC) for detection of microsatellite instability and EBV status. Microsatellite instability and EBV status were determined to be detectable if the lower bound of the 95% CI for the AUROC was above 0·5. FINDINGS: Across the ten cohorts, our analysis included 2823 patients with known microsatellite instability status and 2685 patients with known EBV status. In the within-cohort cross-validation, the deep learning-based classifier could detect microsatellite instability status in nine of ten cohorts, with AUROCs ranging from 0·597 (95% CI 0·522-0·737) to 0·836 (0·795-0·880) and EBV status in five of eight cohorts, with AUROCs ranging from 0·819 (0·752-0·841) to 0·897 (0·513-0·966). Training a classifier on the pooled training dataset and testing it on the five remaining cohorts resulted in high classification performance with AUROCs ranging from 0·723 (95% CI 0·676-0·794) to 0·863 (0·747-0·969) for detection of microsatellite instability and from 0·672 (0·403-0·989) to 0·859 (0·823-0·919) for detection of EBV status. INTERPRETATION: Classifiers became increasingly robust when trained on pooled cohorts. After prospective validation, this deep learning-based tissue classification system could be used as an inexpensive predictive biomarker for immunotherapy in gastric cancer. FUNDING: German Cancer Aid and German Federal Ministry of Health.

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1016/s2589-7500(21)00133-3

Cites

45 of 45 references verified0 citation errors · 0 not checkable

Statistics

Figures

Figure 1 Subgroup-dependent performance of deep learning-based classifiers for detection of microsatellite instability and EBV Subgroup-dependent AUROCs for detection of microsatellite instability (A) and EBV (B). AUGSB=samples from University Hospital Augsberg, Germany. AUROC=area under the receiver operator curve. BERN=samples from University of Bern, Switzerland. CLASSIC=samples from the CLASSIC trial in South Korea. EBV=Epstein-Barr virus. ITALIAN=samples from University of Siena, Italy. KCCH=samples from Kanagawa Cancer Center Hospital, Japan. KOELN=samples from University of Cologne, Germany. LEEDS=samples from Leeds Teaching Hospitals NHS Trust, UK. MAGIC=samples from the MAGIC trial in the UK. NA=not available. TCGA=samples from The Cancer Genome Atlas. TUM=samples from Technical University Munich, Germany. UICC=Union for International Cancer Control.
Figure 1 Subgroup-dependent performance of deep learning-based classifiers for detection of microsatellite instability and EBV Subgroup-dependent AUROCs for detection of microsatellite instability (A) and EBV (B). AUGSB=samples from University Hospital Augsberg, Germany. AUROC=area under the receiver operator curve. BERN=samples from University of Bern, Switzerland. CLASSIC=samples from the CLASSIC trial in South Korea. EBV=Epstein-Barr virus. ITALIAN=samples from University of Siena, Italy. KCCH=samples from Kanagawa Cancer Center Hospital, Japan. KOELN=samples from University of Cologne, Germany. LEEDS=samples from Leeds Teaching Hospitals NHS Trust, UK. MAGIC=samples from the MAGIC trial in the UK. NA=not available. TCGA=samples from The Cancer Genome Atlas. TUM=samples from Technical University Munich, Germany. UICC=Union for International Cancer Control.CC BY · PubMed Central
Figure 2 Effects of region-specific analysis and tumour-to-tissue ratio on classifier performance (A) Example tissue section, whole tumour annotation, and luminal surface annotation (ie, a virtual biopsy). (B) Microsatellite instability and EBV prediction scores for whole-slide images, tumour only, and virtual biopsy samples in the TUM, KCCH, and AUGSB cohorts. Prediction performance of the model for microsatellite instability (C) and EBV status (D) according to tumour-to-tissue ratio. Patients were stratified by the ratio between tumour-contianing area and total tissue area as follows: low was a tumour-to-tissue ratio of 0–0·33, medium was a ratio of 0·34–0·66, and high was a ratio of 0·66–1. AUROC=area under the receiver operating curve. AUGSB=samples from University Hospital Augsberg, Germany. BERN=samples from University of Bern, Switzerland. CLASSIC=samples from the CLASSIC trial in South Korea. EBV=Epstein-Barr virus. ITALIAN=samples from University of Siena, Italy. KCCH=samples from Kanagawa Cancer Center Hospital, Japan. LEEDS=samples from Leeds Teaching Hospitals NHS Trust, UK. MAGIC=samples from the MAGIC trial in the UK. TCGA=samples from The Cancer Genome Atlas. TUM=samples from Technical University Munich, Germany.
Figure 2 Effects of region-specific analysis and tumour-to-tissue ratio on classifier performance (A) Example tissue section, whole tumour annotation, and luminal surface annotation (ie, a virtual biopsy). (B) Microsatellite instability and EBV prediction scores for whole-slide images, tumour only, and virtual biopsy samples in the TUM, KCCH, and AUGSB cohorts. Prediction performance of the model for microsatellite instability (C) and EBV status (D) according to tumour-to-tissue ratio. Patients were stratified by the ratio between tumour-contianing area and total tissue area as follows: low was a tumour-to-tissue ratio of 0–0·33, medium was a ratio of 0·34–0·66, and high was a ratio of 0·66–1. AUROC=area under the receiver operating curve. AUGSB=samples from University Hospital Augsberg, Germany. BERN=samples from University of Bern, Switzerland. CLASSIC=samples from the CLASSIC trial in South Korea. EBV=Epstein-Barr virus. ITALIAN=samples from University of Siena, Italy. KCCH=samples from Kanagawa Cancer Center Hospital, Japan. LEEDS=samples from Leeds Teaching Hospitals NHS Trust, UK. MAGIC=samples from the MAGIC trial in the UK. TCGA=samples from The Cancer Genome Atlas. TUM=samples from Technical University Munich, Germany.CC BY · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Show all 45 references

Cited by

See all 127

Metadata

Authors
Muti, H. S., Heij, L. R., Keller, G., Kohlruss, M., Langer, R., Dislich, B., Cheong, J.-H., Kim, Y.-W., Kim, H., Kook, M.-C., Cunningham, D., Allum, W. H., Langley, R. E., Nankivell, M. G., Quirke, P., Hayden, J. D., West, N. P., Irvine, A. J., Yoshikawa, T., Oshima, T., Huss, R., Grosser, B., Roviello, F., d'Ignazio, A., Quaas, A., Alakus, H., Tan, X., Pearson, A. T., Luedde, T., Ebert, M. P., Jäger, D., Trautwein, C., Gaisa, N. T., Grabsch, H. I., & Kather, J. N.
Affiliations
  • Universitätsklinikum Aachen (DE)
  • Technical University of Munich (DE)
  • University of Bern (CH)
  • Yonsei University (KR)
  • National Cancer Center (KR)
  • Royal Marsden NHS Foundation Trust (GB)
  • Royal Marsden Hospital (GB)
  • MRC Clinical Trials Unit at UCL (GB)
  • University of Leeds (GB)
  • St James's University Hospital (GB)
  • National Cancer Center Hospital (JP)
  • Kanagawa Cancer Center (JP)
  • University Hospital Augsburg (DE)
  • University of Siena (IT)
  • University Hospital Cologne (DE)
  • University of Chicago (US)
  • Düsseldorf University Hospital (DE)
  • Heidelberg University (DE)
  • University Hospital Heidelberg (DE)
Corresponding author
Kather, Jakob Nikolas
Venue
The Lancet Digital Health
Publisher
Elsevier BV
Published
2021-10
Volume
3
Issue
10
Pages
e654-e664
ISSN
2589-7500 (ISSN-L)
OpenAlex
W3195171687
Cited by
126
References
44
Citations by year
2026: 14 · 2025: 18 · 2024: 27 · 2023: 28 · 2022: 32 · 2021: 7
FWCI
8.30
Type
journal article
Language
en
Update notices
erratum (2021-10-01)
Open access
Licence
Locations
  • The Lancet Digital Health (publishedVersion, cc-by)
  • Europe PMC (PubMed Central) (publishedVersion, other-oa)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • Research Publications (Maastricht University) (publishedVersion, cc-by)
  • White Rose Research Online (University of Leeds, The University of Sheffield, University of York) (acceptedVersion)
  • UCL Discovery (University College London) (submittedVersion)
  • YUHSpace (Yonsei University Medical Library) (submittedVersion, cc-by-nc-nd)
  • mediaTUM – the media and publications repository of the Technical University Munich (Technical University Munich) (submittedVersion)
  • mediaTUM – the media and publications repository of the Technical University Munich (Technical University Munich) (submittedVersion, other-oa)
  • RWTH Publications (RWTH Aachen) (submittedVersion, other-oa)
  • OPUS (Augsburg University) (submittedVersion, cc-by)
  • Use Siena air (University of Siena) (submittedVersion, cc-by)
  • Research Publications (Maastricht University) (submittedVersion, other-oa)
  • RWTH Publications (RWTH Aachen)
Funders
  • ERC (771083)
  • German Cancer Aid (70113864)
  • Deutsche Forschungsgemeinschaft (SFB-TRR57, GRK2727)
  • European Research Council (SFB-CRC1382, LU 1360/3-2)
  • Roche
  • Horizon 2020
  • Yorkshire Cancer Research (L394, L386)
  • Eli Lilly and Company
  • Cancer Research UK
Topics
  • Physical Sciences › Computer Science › Artificial Intelligence › AI in cancer detection
  • Health Sciences › Medicine › Oncology › Gastric Cancer Management and Outcomes
  • Health Sciences › Medicine › Radiology, Nuclear Medicine and Imaging › Radiomics and Machine Learning in Medical Imaging
Keywords
Epstein-Barr virus, microsatellite instability, gastric cancer, deep learning classifier, histopathology image analysis, immunotherapy response, mismatch repair deficiency, external validation, H&E staining, tumor-infiltrating immune cells, tumor-infiltrating lymphocytes, tumour-associated macrophages, tumor-associated neutrophils, cancer-associated fibroblasts
MeSH terms
Deep Learning, Aged, Female, Germany, United Kingdom, Histological Techniques, Humans, Italy, Japan, Male, Middle Aged, Retrospective Studies, Stomach Neoplasms, Switzerland, United States, Reproducibility of Results, Cohort Studies, Epstein-Barr Virus Infections, Microsatellite Instability, Republic of Korea

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Development and validation of deep learning classifiers to detect Epstein-Barr virus and microsatellite instability status in gastric cancer: a retrospective multicentre cohort study. https://app.cochat.ai/papers/10.1016/s2589-7500(21)00133-3
The paper
Muti, H. S., Heij, L. R., Keller, G., Kohlruss, M., Langer, R., Dislich, B., Cheong, J.-H., Kim, Y.-W., Kim, H., Kook, M.-C., Cunningham, D., Allum, W. H., Langley, R. E., Nankivell, M. G., Quirke, P., Hayden, J. D., West, N. P., Irvine, A. J., Yoshikawa, T., . . . Kather, J. N. (2021). Development and validation of deep learning classifiers to detect Epstein-Barr virus and microsatellite instability status in gastric cancer: a retrospective multicentre cohort study. The Lancet Digital Health. https://doi.org/10.1016/s2589-7500(21)00133-3

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.