Semi-automated analysis of HER2 immunohistochemistry in invasive breast carcinoma using whole slide images: Utility for interpretation in clinical practice Journal Article

   

Authors:	Liao, C. H. C.; Bakoglu, N.; Cesmecioglu, E.; Hanna, M.; Pareja, F.; Wen, H. Y.; D’Alfonso, T. M.; Brogi, E.; Yagi, Y.; Ross, D. S.
Article Title:	Semi-automated analysis of HER2 immunohistochemistry in invasive breast carcinoma using whole slide images: Utility for interpretation in clinical practice
Abstract:	Human epidermal growth factor receptor 2 (HER2) gene amplification and subsequent protein overexpression is a strong prognostic and predictive biomarker in invasive breast carcinoma (IBC). ASCO/CAP recommended tests for HER2 assessment include immunohistochemistry (IHC) and/or in situ hybridization (ISH). Accurate HER2 IHC scoring (0, 1+, 2+, 3+) is key for appropriate classification and treatment of IBC. HER2-targeted therapies, including anti-HER2 monoclonal antibodies and antibody drug conjugates (ADC), have revolutionized the treatment of HER2-positive IBC. Recently, ADC have also been approved for treatment of HER2-low (IHC 1+, IHC 2+/ISH-) advanced breast carcinoma, making a distinction between IHC 0 and 1+ crucial. In this focused study, 32 IBC with HER2 IHC scores from 0 to 3+ and HER2 FISH results formed a calibration dataset, and 77 IBC with HER2 IHC score 2+ and paired FISH results (27 amplified, 50 non-amplified) formed a validation dataset. H&E and HER2 IHC whole slide images (WSI) were scanned. Regions of interest were manually annotated and IHC scores generated by the software QuantCenter (MembraneQuant application) by 3DHISTECH Ltd. (Budapest, Hungary) and compared to the microscopic IHC score. H-scores [(3×%IHC3+) +(2×%IHC2+) +(1×%IHC1+)] were calculated for semi-automated (MembraneQuant) analysis. Concordance between microscopic IHC scoring and 3DHISTECH MembraneQuant semi-automated scoring in the calibration dataset showed a Kappa value of 0.77 (standard error 0.09). Microscopic IHC and MembraneQuant image analysis for the detection of HER2 amplification yielded a sensitivity of 100% for both and a specificity of 56% and 61%, respectively. In the validation set of IHC 2+ cases, only 13 of 77 cases (17%) had discordant results between microscopic and MembraneQuant images, and various artifacts limiting the interpretation of HER2 IHC, including cytoplasmic/granular staining and crush artifact were noted. Semi-automated analysis using WSI and microscopic evaluation yielded similar HER2 IHC scores, demonstrating the potential utility of this tool for interpretation in clinical practice and subsequent accurate treatment. In this study, it was shown that semi-automatic HER2 IHC interpretation provides an objective approach to a test known to be quite subjective. Copyright © 2024 Liao, Bakoglu, Cesmecioglu, Hanna, Pareja, Wen, D’Alfonso, Brogi, Yagi and Ross.
Keywords:	immunohistochemistry; adult; controlled study; human tissue; protein expression; overall survival; sensitivity and specificity; clinical practice; metabolism; in situ hybridization, fluorescence; gene amplification; epidermal growth factor receptor 2; pathology; breast neoplasms; tumor marker; in situ hybridization; fluorescence in situ hybridization; breast tumor; breast carcinoma; scoring system; receptor, erbb-2; tissue microarray; progesterone receptor; image processing, computer-assisted; image processing; predictive value; her2; carcinoma, ductal, breast; whole slide imaging; copy number variation; erbb2 protein, human; diagnostic test accuracy study; procedures; cancer prognosis; humans; prognosis; human; female; article; biomarkers, tumor; breast ductal carcinoma; semi-automated analysis
Journal Title:	Pathology & Oncology Research
Volume:	30
ISSN:	1219-4956
Publisher:	Springer
Date Published:	2024-08-28
Start Page:	1611826
Language:	English
DOI:	10.3389/pore.2024.1611826
PUBMED:	39267995
PROVIDER:	scopus
PMCID:	PMC11390455
DOI/URL:	https://www.scopus.com/inward/record.uri?eid=2-s2.0-85203847831&doi=10.3389%2fpore.2024.1611826&partnerID=40&md5=698d0b6c09fff529f4ab686e9e645637
Notes:	The MSK Cancer Center Support Grant (P30 CA008748) is acknowledge in the PDF -- Corresponding authors is MSK author: Dara S. Ross -- Source: Scopus

https://synapse.mskcc.org/synapse/works/has_related_data_chk/219216