Image analysis-based identification of high risk ER-positive, HER2-negative breast cancers

Abstract Background Breast cancer subtypes Luminal A and Luminal B are classified by the expression of PAM50 genes and may benefit from different treatment strategies. Machine learning models based on H&E images may contain features associated with subtype, allowing early identification of tumor...

Full description

Saved in:
Bibliographic Details
Main Authors: Dong Neuck Lee, Yao Li, Linnea T. Olsson, Alina M. Hamilton, Benjamin C. Calhoun, Katherine A. Hoadley, J. S. Marron, Melissa A. Troester
Format: Article
Language:English
Published: BMC 2024-12-01
Series:Breast Cancer Research
Subjects:
Online Access:https://doi.org/10.1186/s13058-024-01915-5
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1846136723264765952
author Dong Neuck Lee
Yao Li
Linnea T. Olsson
Alina M. Hamilton
Benjamin C. Calhoun
Katherine A. Hoadley
J. S. Marron
Melissa A. Troester
author_facet Dong Neuck Lee
Yao Li
Linnea T. Olsson
Alina M. Hamilton
Benjamin C. Calhoun
Katherine A. Hoadley
J. S. Marron
Melissa A. Troester
author_sort Dong Neuck Lee
collection DOAJ
description Abstract Background Breast cancer subtypes Luminal A and Luminal B are classified by the expression of PAM50 genes and may benefit from different treatment strategies. Machine learning models based on H&E images may contain features associated with subtype, allowing early identification of tumors with higher risk of recurrence. Methods H&E images (n = 630 ER+/HER2-breast cancers) were pixel-level segmented into epithelium and stroma. Convolutional neural network and multiple instance learning were used to extract image features from original and segmented images. Patient-level classification models were trained to discriminate Luminal A versus B image features in tenfold cross-validation, with or without grade adjustment. The best-performing visual classifier was incorporated into envisioned diagnostic protocols as an alternative to genomic testing (PAM50). The protocols were then compared in time-to-recurrence models. Results Among ER+/HER2-tumors, the image-based protocol differentiated recurrence times with a hazard ratio (HR) of 2.81 (95% CI: 1.73–4.56), which was similar to the HR for PAM50 (2.66, 95% CI: 1.65–4.28). Grade adjustment did not improve subtype prediction accuracy, but did help balance sensitivity and specificity. Among high grade participants, sensitivity and specificity (0.734 and 0.474, respectively) became more similar (0.732 and 0.624, respectively) in grade-adjusted models. The original and epithelium-specific images had similar performance and highest accuracy, followed by stroma or binarized images showing only the epithelial-stromal interface. Conclusions Given low rates of genomic testing uptake nationally, image-based methods may help identify ER+/HER2-patients who could benefit from testing.
format Article
id doaj-art-d6d419adfead42c48cb70ca307eb738a
institution Kabale University
issn 1465-542X
language English
publishDate 2024-12-01
publisher BMC
record_format Article
series Breast Cancer Research
spelling doaj-art-d6d419adfead42c48cb70ca307eb738a2024-12-08T12:50:06ZengBMCBreast Cancer Research1465-542X2024-12-0126111110.1186/s13058-024-01915-5Image analysis-based identification of high risk ER-positive, HER2-negative breast cancersDong Neuck Lee0Yao Li1Linnea T. Olsson2Alina M. Hamilton3Benjamin C. Calhoun4Katherine A. Hoadley5J. S. Marron6Melissa A. Troester7Department of Biostatistics, University of North CarolinaDepartment of Statistics and Operations Research, University of North CarolinaDepartment of Epidemiology, University of North CarolinaLineberger Comprehensive Cancer Center, University of North CarolinaDepartment of Pathology and Laboratory Medicine, University of North CarolinaDepartment of Genetics, University of North CarolinaDepartment of Statistics and Operations Research, University of North CarolinaDepartment of Epidemiology, University of North CarolinaAbstract Background Breast cancer subtypes Luminal A and Luminal B are classified by the expression of PAM50 genes and may benefit from different treatment strategies. Machine learning models based on H&E images may contain features associated with subtype, allowing early identification of tumors with higher risk of recurrence. Methods H&E images (n = 630 ER+/HER2-breast cancers) were pixel-level segmented into epithelium and stroma. Convolutional neural network and multiple instance learning were used to extract image features from original and segmented images. Patient-level classification models were trained to discriminate Luminal A versus B image features in tenfold cross-validation, with or without grade adjustment. The best-performing visual classifier was incorporated into envisioned diagnostic protocols as an alternative to genomic testing (PAM50). The protocols were then compared in time-to-recurrence models. Results Among ER+/HER2-tumors, the image-based protocol differentiated recurrence times with a hazard ratio (HR) of 2.81 (95% CI: 1.73–4.56), which was similar to the HR for PAM50 (2.66, 95% CI: 1.65–4.28). Grade adjustment did not improve subtype prediction accuracy, but did help balance sensitivity and specificity. Among high grade participants, sensitivity and specificity (0.734 and 0.474, respectively) became more similar (0.732 and 0.624, respectively) in grade-adjusted models. The original and epithelium-specific images had similar performance and highest accuracy, followed by stroma or binarized images showing only the epithelial-stromal interface. Conclusions Given low rates of genomic testing uptake nationally, image-based methods may help identify ER+/HER2-patients who could benefit from testing.https://doi.org/10.1186/s13058-024-01915-5Breast cancerCBCS3HistologyMultiple instance learningImage segmentationDistance weighted learning
spellingShingle Dong Neuck Lee
Yao Li
Linnea T. Olsson
Alina M. Hamilton
Benjamin C. Calhoun
Katherine A. Hoadley
J. S. Marron
Melissa A. Troester
Image analysis-based identification of high risk ER-positive, HER2-negative breast cancers
Breast Cancer Research
Breast cancer
CBCS3
Histology
Multiple instance learning
Image segmentation
Distance weighted learning
title Image analysis-based identification of high risk ER-positive, HER2-negative breast cancers
title_full Image analysis-based identification of high risk ER-positive, HER2-negative breast cancers
title_fullStr Image analysis-based identification of high risk ER-positive, HER2-negative breast cancers
title_full_unstemmed Image analysis-based identification of high risk ER-positive, HER2-negative breast cancers
title_short Image analysis-based identification of high risk ER-positive, HER2-negative breast cancers
title_sort image analysis based identification of high risk er positive her2 negative breast cancers
topic Breast cancer
CBCS3
Histology
Multiple instance learning
Image segmentation
Distance weighted learning
url https://doi.org/10.1186/s13058-024-01915-5
work_keys_str_mv AT dongneucklee imageanalysisbasedidentificationofhighriskerpositiveher2negativebreastcancers
AT yaoli imageanalysisbasedidentificationofhighriskerpositiveher2negativebreastcancers
AT linneatolsson imageanalysisbasedidentificationofhighriskerpositiveher2negativebreastcancers
AT alinamhamilton imageanalysisbasedidentificationofhighriskerpositiveher2negativebreastcancers
AT benjaminccalhoun imageanalysisbasedidentificationofhighriskerpositiveher2negativebreastcancers
AT katherineahoadley imageanalysisbasedidentificationofhighriskerpositiveher2negativebreastcancers
AT jsmarron imageanalysisbasedidentificationofhighriskerpositiveher2negativebreastcancers
AT melissaatroester imageanalysisbasedidentificationofhighriskerpositiveher2negativebreastcancers