Highlights
- •Inter-observer bias of PD-L1 reading limits prediction of clinical outcome of NSCLC.
- •We developed an AI model to detect PD-L1 expression in tumour and to calculate TPS.
- •Discrepancy among pathologists on TPS interpretation was reduced by AI assistance.
- •AI-assisted TPS reading leads to better prediction of therapeutic outcome of ICI.
Abstract
Background
Manual evaluation of programmed death ligand 1 (PD-L1) tumour proportion score (TPS)
by pathologists is associated with interobserver bias.
Objective
This study explored the role of artificial intelligence (AI)-powered TPS analyser
in minimisation of interobserver variation and enhancement of therapeutic response
prediction.
Methods
A prototype model of an AI-powered TPS analyser was developed with a total of 802
non–small cell lung cancer (NSCLC) whole-slide images. Three independent board-certified
pathologists labelled PD-L1 TPS in an external cohort of 479 NSCLC slides. For cases
of disagreement between each pathologist and the AI model, the pathologists were asked
to revise the TPS grade (<1%, 1%–49% and ≥50%) with AI assistance. The concordance
rates among the pathologists with or without AI assistance and the effect of the AI-assisted
revision on clinical outcome upon immune checkpoint inhibitor (ICI) treatment were
evaluated.
Results
Without AI assistance, pathologists concordantly classified TPS in 81.4% of the cases.
They revised their initial interpretation by using the AI model for the disagreement
cases between the pathologist and the AI model (N = 91, 93 and 107 for each pathologist). The overall concordance rate among the pathologists
was increased to 90.2% after the AI assistance (P < 0.001). A reduction in hazard ratio for overall survival and progression-free survival
upon ICI treatment was identified in the TPS subgroups after the AI-assisted TPS revision.
Conclusion
The AI-powered TPS analyser assistance improves the pathologists’ consensus of reading
and prediction of the therapeutic response, raising a possibility of standardised
approach for the accurate interpretation.
Keywords
To read this article in full you will need to make a payment
Purchase one-time access:
Academic & Personal: 24 hour online accessCorporate R&D Professionals: 24 hour online accessOne-time access price info
- For academic or personal research use, select 'Academic and Personal'
- For corporate R&D use, select 'Corporate R&D Professionals'
Subscribe:
Subscribe to European Journal of CancerAlready a print subscriber? Claim online access
Already an online subscriber? Sign in
Register: Create an account
Institutional Access: Sign in to ScienceDirect
References
- Pan-Asian adapted ESMO clinical practice guidelines for the management of patients with locally-advanced unresectable non-small-cell lung cancer: a KSMO-ESMO initiative endorsed by CSCO, ISMPO, JSMO, MOS, SSO and TOS.Ann Oncol. 2020; 31: 191-201https://doi.org/10.1016/j.annonc.2019.10.026
- Pembrolizumab versus docetaxel for previously treated, PD-L1-positive, advanced non-small-cell lung cancer (KEYNOTE-010): a randomised controlled trial.Lancet. 2016; 387: 1540-1550https://doi.org/10.1016/S0140-6736(15)01281-7
- Nivolumab versus docetaxel in previously treated patients with advanced non-small-cell lung cancer: two-year outcomes from two randomized, open-label, phase III trials (CheckMate 017 and CheckMate 057).J Clin Oncol. 2017; 35: 3924-3933https://doi.org/10.1200/JCO.2017.74.3062
- Pembrolizumab versus chemotherapy for PD-L1-positive non-small-cell lung cancer.N Engl J Med. 2016; 375: 1823-1833https://doi.org/10.1056/NEJMoa1606774
- Atezolizumab versus docetaxel in patients with previously treated non-small-cell lung cancer (OAK): a phase 3, open-label, multicentre randomised controlled trial.Lancet. 2017; 389: 255-265https://doi.org/10.1016/S0140-6736(16)32517-X
- Predictive biomarkers for checkpoint inhibitor-based immunotherapy.Lancet Oncol. 2016; 17: e542-e551https://doi.org/10.1016/S1470-2045(16)30406-5
- PD-L1 expression as a predictive biomarker in advanced non-small-cell lung cancer: updated survival data.Immunotherapy. 2017; 9: 499-506https://doi.org/10.2217/imt-2016-0150
- Nivolumab versus docetaxel in advanced nonsquamous non-small-cell lung cancer.N Engl J Med. 2015; 373: 1627-1639https://doi.org/10.1056/NEJMoa1507643
- Atezolizumab versus docetaxel for patients with previously treated non-small-cell lung cancer (POPLAR): a multicentre, open-label, phase 2 randomised controlled trial.Lancet. 2016; 387: 1837-1846https://doi.org/10.1016/S0140-6736(16)00587-0
- Pembrolizumab for the treatment of non-small-cell lung cancer.N Engl J Med. 2015; 372: 2018-2028https://doi.org/10.1056/NEJMoa1501824
- PD-L1 immunohistochemistry in clinical diagnostics of lung cancer: inter-pathologist variability is higher than assay variability.Mod Pathol. 2017; 30: 1411-1421https://doi.org/10.1038/modpathol.2017.59
- Interobserver reproducibility of PD-L1 biomarker in non-small cell lung cancer: a multi-institutional study by 27 pathologists.J Pathol Transl Med. 2019; 53: 347-353https://doi.org/10.4132/jptm.2019.09.29
- Intra- and interobserver reproducibility assessment of PD-L1 biomarker in non-small cell lung cancer.Clin Cancer Res. 2017; 23: 4569-4577https://doi.org/10.1158/1078-0432.CCR-17-0151
- The practical implementation of artificial intelligence technologies in medicine.Nat Med. 2019; 25: 30-36https://doi.org/10.1038/s41591-018-0307-0
- Deep learning for prediction of colorectal cancer outcome: a discovery and validation study.Lancet. 2020; 395: 350-360https://doi.org/10.1016/S0140-6736(19)32998-8
- AI-based pathology predicts origins for cancers of unknown primary.Nature. 2021; 594: 106-110https://doi.org/10.1038/s41586-021-03512-4
- Artificial intelligence as the next step towards precision pathology.J Intern Med. 2020; 288: 62-81https://doi.org/10.1111/joim.13030
- Digital pathology and artificial intelligence.Lancet Oncol. 2019; 20: e253-e261https://doi.org/10.1016/S1470-2045(19)30154-8
- Artificial intelligence assistance significantly improves Gleason grading of prostate biopsies by pathologists.Mod Pathol. 2021; 34: 660-671https://doi.org/10.1038/s41379-020-0640-y
- Deep learning for diagnosis and survival prediction in soft tissue sarcoma.Ann Oncol. 2021; 32: 1178-1187https://doi.org/10.1016/j.annonc.2021.06.007
- A prospective validation and observer performance study of a deep learning algorithm for pathologic diagnosis of gastric tumors in endoscopic biopsies.Clin Cancer Res. 2021; 27: 719-728https://doi.org/10.1158/1078-0432.CCR-20-3159
- Impact of deep learning assistance on the histopathologic review of lymph nodes for metastatic breast cancer.Am J Surg Pathol. 2018; 42: 1636-1646https://doi.org/10.1097/PAS.0000000000001151
- Development and applications of computer image analysis algorithms for scoring of PD-L1 immunohistochemistry.Immuno-Oncol Technol. 2020; 6: 2-8https://doi.org/10.1016/j.iotech.2020.04.001
- Critical appraisal of programmed death ligand 1 reflex diagnostic testing: current standards and future opportunities.J Thorac Oncol. 2019; 14: 45-53https://doi.org/10.1016/j.jtho.2018.09.025
- Digital image analysis improves precision of PD-L1 scoring in cutaneous melanoma.Histopathology. 2018; 73: 397-406https://doi.org/10.1111/his.13528
- PD-L1 expression in gastric cancer determined by digital image analyses: pitfalls and correlation with pathologist interpretation.Virchows Arch. 2020; 476: 243-250https://doi.org/10.1007/s00428-019-02653-2
- Automated tumor proportion scoring for PD-L1 expression based on multistage ensemble strategy in non-small cell lung cancer.J Transl Med. 2021; 19: 249https://doi.org/10.1186/s12967-021-02898-z
- How can artificial intelligence models assist PD-L1 expression scoring in breast cancer: results of multi-institutional ring studies.NPJ breast cancer. 2021; 7: 61https://doi.org/10.1038/s41523-021-00268-y
- Deep semi supervised generative learning for automated tumor proportion scoring on NSCLC tissue needle biopsies.Sci Rep. 2018; 8: 17343https://doi.org/10.1038/s41598-018-35501-5
- Comparison of continuous measures across diagnostic PD-L1 assays in non-small cell lung cancer using automated image analysis.Mod Pathol. 2020; 33: 380-390https://doi.org/10.1038/s41379-019-0349-y
- A multi-institutional study to evaluate automated whole slide scoring of immunohistochemistry for assessment of programmed death-ligand 1 (PD-L1) expression in non-small cell lung cancer.Appl Immunohistochem Mol Morphol. 2019; 27: 263-269https://doi.org/10.1097/PAI.0000000000000737
- Automated tumor proportion score analysis for PD-L1 (22C3) expression in lung squamous cell carcinoma.Sci Rep. 2021; 11: 15907https://doi.org/10.1038/s41598-021-95372-1
- Artificial intelligence-assisted system for precision diagnosis of PD-L1 expression in non-small cell lung cancer.Mod Pathol. 2021; (Article in Press. https://doi.org/10.1038/s41379-021-00904-9)
- Towards real-time object detection with region proposal networks.IEEE Trans Pattern Anal Mach Intell. 2017; 39: 1137-1149https://doi.org/10.1109/TPAMI.2016.2577031
Agilent technologies, Inc. PD-L1 IHC 22C3 pharmDx interpretation manual – NSCLC. Accessed December 30, 2021. https://www.agilent.com/cs/library/usermanuals/public/29158_pd-l1-ihc-22C3-pharmdx-nsclc-interpretation-manual.pdf.
- New response evaluation criteria in solid tumours: revised RECIST guideline (version 1.1).Eur J Cancer. 2009; 45: 228-247https://doi.org/10.1016/j.ejca.2008.10.026
- High-performance medicine: the convergence of human and artificial intelligence.Nat Med. 2019; 25: 44-56https://doi.org/10.1038/s41591-018-0300-7
- On the interpretability of artificial intelligence in radiology: challenges and opportunities.Radiol Artif Intell. 2020; 2: e190043https://doi.org/10.1148/ryai.2020190043
- Digital pathology and PD-L1 testing in non small cell lung cancer: a workshop record.Cancers. 2020; 12: 1800https://doi.org/10.3390/cancers12071800
Article info
Publication history
Published online: May 14, 2022
Accepted:
April 4,
2022
Received in revised form:
March 10,
2022
Received:
January 11,
2022
Identification
Copyright
© 2022 Elsevier Ltd. All rights reserved.
