Pitfalls in machine learning-based assessment of tumor-infiltrating lymphocytes in breast cancer: a report of the international immuno-oncology biomarker working group
Research output: Contribution to journal › Review › Research › peer-review
Documents
- Fulltext
Final published version, 1.31 MB, PDF document
The clinical significance of the tumor-immune interaction in breast cancer is now established, and tumor-infiltrating lymphocytes (TILs) have emerged as predictive and prognostic biomarkers for patients with triple-negative (estrogen receptor, progesterone receptor, and HER2-negative) breast cancer and HER2-positive breast cancer. How computational assessments of TILs might complement manual TIL assessment in trial and daily practices is currently debated. Recent efforts to use machine learning (ML) to automatically evaluate TILs have shown promising results. We review state-of-the-art approaches and identify pitfalls and challenges of automated TIL evaluation by studying the root cause of ML discordances in comparison to manual TIL quantification. We categorize our findings into four main topics: (1) technical slide issues, (2) ML and image analysis aspects, (3) data challenges, and (4) validation issues. The main reason for discordant assessments is the inclusion of false-positive areas or cells identified by performance on certain tissue patterns or design choices in the computational implementation. To aid the adoption of ML for TIL assessment, we provide an in-depth discussion of ML and image analysis, including validation issues that need to be considered before reliable computational reporting of TILs can be incorporated into the trial and routine clinical management of patients with triple-negative breast cancer.
| Original language | English |
|---|---|
| Journal | Journal of Pathology |
| Volume | 260 |
| Issue number | 5 |
| Pages (from-to) | 498-513 |
| Number of pages | 16 |
| ISSN | 0022-3417 |
| DOIs | |
| Publication status | Published - 2023 |
Bibliographical note
Funding Information:
The authors would like to thank Jeannette Parrodi, PA assistant to Professor Sherene Loi, for het extensive help and administrative support for the International Immuno‐Oncology Biomarker Working Group (TIL working group). Without her, this working group would not even exist. Furthermore, the authors make the following acknowledgments regarding support and funding. GB: Funded by Gilead Breast Cancer Research Grant 2023. SV: Supported by Interne Fondsen KU Leuven/Internal Funds KU Leuven. BA: supported by the Swedish Society for Medical Research (Svenska Sällskapet för Medicinsk Forskning) postdoctoral grant, Swedish Breast Cancer Association (Bröstcancerförbundet) Research grant 2021. GC: Peer Reviewed Cancer Research Program (Award W81XWH‐21‐1‐0160) from the US Department of Defense and the Mayo Clinic Breast Cancer SPORE grant P50 CA116201 from the National Institutes of Health (NIH). CF‐M: Funded by the Horizon 2020 European Union Research and Innovation Programme under the Marie Sklodowska Curie Grant agreement No. 860627 (CLARIFY Project). SBF: NHMRC GNT1193630. WMG: Support by the Higher Education Authority, Department of Further and Higher Education, Research, Innovation and Science, and the Shared Island Fund [AICRIstart: A Foundation Stone for the All‐Island Cancer Research Institute (AICRI): Building Critical Mass in Precision Cancer Medicine, https://www.aicri.org/aicristart ]: Irish Cancer Society (Collaborative Cancer Research Centre BREAST‐PREDICT; CCRC13GAL; https://www.breastpredict.com ), the Science Foundation Ireland Investigator Programme (OPTi‐PREDICT; 15/IA/3104), the Science Foundation Ireland Strategic Partnership Programme (Precision Oncology Ireland; 18/SPP/3522; https://www.precisiononcology.ie ). SG: Partially supported by NIH grants CA224319, DK124165, CA263705, and CA196521. AG: Supported by Breast Cancer Now (and their legacy charity Breakthrough Breast Cancer) and Cancer Research UK (CRUK/07/012, KCL‐BCN‐Q3). TRK: Japan Society for the Promotion of Science (JSPS) KAKENHI (21K06909). UK: Funded by Horizon 2020 European Union Research and Innovation Programme under the Marie Sklodowska Curie Grant agreement 860627 (CLARIFY Project). JKL: This work is in part supported by NIH R37 CA225655 to JKL. AM: Research reported in this publication was supported by the National Cancer Institute under award numbers R01CA268287A1, U01CA269181, R01CA26820701A1, R01CA249992‐01A1, R01CA202752‐01A1, R01CA208236‐01A1, R01CA216579‐01A1, R01CA220581‐01A1, R01CA257612‐01A1, 1U01CA239055‐01, 1U01CA248226‐01, and 1U54CA254566‐01, National Heart, Lung and Blood Institute 1R01HL15127701A1, R01HL15807101A1, National Institute of Biomedical Imaging and Bioengineering 1R43EB028736‐01, VA Merit Review Award IBX004121A from the US Department of Veterans Affairs Biomedical Laboratory Research and Development Service the Office of the Assistant Secretary of Defense for Health Affairs, through the Breast Cancer Research Program (W81XWH‐19‐1‐0668), the Prostate Cancer Research Program (W81XWH‐20‐1‐0851), the Lung Cancer Research Program (W81XWH‐18‐1‐0440, W81XWH‐20‐1‐0595), the Peer Reviewed Cancer Research Program (W81XWH‐18‐1‐0404, W81XWH‐21‐1‐0345, W81XWH‐21‐1‐0160), the Kidney Precision Medicine Project (KPMP) Glue Grant, and sponsored research agreements from Bristol Myers‐Squibb, Boehringer‐Ingelheim, Eli‐Lilly, and Astrazeneca. SKM: Kay Pogue‐Geile, Director of Molecular Profiling at NSABP for her constant support and encouragement, Roberto Salgado, for initiating me into the wonderful subject of Immuno‐Oncology and its possibilities. FuAAM: Funding from EPSRC EP/W02909X/1 and PathLAKE consortium. FP‐L: Research grants from Fondation ARC, La Ligue contre le Cancer. RDP: The Melbourne Research Scholarship and a scholarship from the Peter MacCallum Cancer Centre. JSR‐F: Funded in part by the Breast Cancer Research Foundation, by a Susan G. Komen Leadership grant, and by the NIH/NCI grant P50 CA247749 01. JS: NIH/NCI grants UH3CA225021 and U24CA215109. ST: Supported by Interne Fondsen KU Leuven/Internal Funds KU Leuven. JT: Supported by institutional grants of the Dutch Cancer Society and the Dutch Ministry of Health, Welfare and Sport. EAT: Breast Cancer Research Foundation grant 22‐161. GEV: Supported by Breast Cancer Now (and their legacy charity Breakthrough Breast Cancer) and Cancer Research UK (CRUK/07/012, KCL‐BCN‐Q3). TW: Support by the French government under management of Agence Nationale de la Recherche as part of the Investissements d'avenir’ program, reference ANR‐19‐P3IA‐0001 (PRAIRIE 3IA Institute), and by Q‐Life (ANR‐17‐CONV‐0005). HYW: Funded in part by the NIH/NCI grant P50 CA247749 01. YY: Funding from Cancer Research UK Career Establishment Award (CRUK C45982/A21808). PS: Funding support from the National Health and Medical Research Council, Australia. SL: Supported by the National Breast Cancer Foundation of Australia (NBCF) (APP ID: EC‐17‐001), the Breast Cancer Research Foundation, New York [BCRF (APP ID: BCRF‐21‐102)], and a National Health and Medical Council of Australia (NHMRC) Investigator Grant (APP ID: 1162318). RS: Supported by the Breast Cancer Research Foundation (BCRF, grant 17‐194).
Publisher Copyright:
© 2023 The Authors. The Journal of Pathology published by John Wiley & Sons Ltd on behalf of The Pathological Society of Great Britain and Ireland.
- deep learning, digital pathology, guidelines, image analysis, machine learning, pitfalls, prognostic biomarker, triple-negative breast cancer, tumor-infiltrating lymphocytes
Research areas
ID: 387276018