Quantitative MRI radiomics in the prediction of molecular classifications of breast cancer subtypes in the TCGA/TCIA data set

Hui Li, Yitan Zhu, Elizabeth S. Burnside, Erich Huang, Karen Drukker, Katherine A. Hoadley, Cheng Fan, Suzanne D. Conzen, Margarita Zuley, Jose M Net, Elizabeth Sutton, Gary J. Whitman, Elizabeth Morris, Charles M. Perou, Yuan Ji, Maryellen L. Giger

Research output: Contribution to journalArticle

91 Citations (Scopus)

Abstract

Using quantitative radiomics, we demonstrate that computer-extracted magnetic resonance (MR) image-based tumor phenotypes can be predictive of the molecular classification of invasive breast cancers. Radiomics analysis was performed on 91 MRIs of biopsy-proven invasive breast cancers from National Cancer Institute’s multi-institutional TCGA/TCIA. Immunohistochemistry molecular classification was performed including estrogen receptor, progesterone receptor, human epidermal growth factor receptor 2, and for 84 cases, the molecular subtype (normal-like, luminal A, luminal B, HER2-enriched, and basal-like). Computerized quantitative image analysis included: three-dimensional lesion segmentation, phenotype extraction, and leave-one-case-out cross validation involving stepwise feature selection and linear discriminant analysis. The performance of the classifier model for molecular subtyping was evaluated using receiver operating characteristic analysis. The computer-extracted tumor phenotypes were able to distinguish between molecular prognostic indicators; area under the ROC curve values of 0.89, 0.69, 0.65, and 0.67 in the tasks of distinguishing between ER+ versus ER−, PR+ versus PR−, HER2+ versus HER2−, and triple-negative versus others, respectively. Statistically significant associations between tumor phenotypes and receptor status were observed. More aggressive cancers are likely to be larger in size with more heterogeneity in their contrast enhancement. Even after controlling for tumor size, a statistically significant trend was observed within each size group (P = 0.04 for lesions ≤ 2 cm; P = 0.02 for lesions >2 to≤ 5 cm) as with the entire data set (P-value = 0.006) for the relationship between enhancement texture (entropy) and molecular subtypes (normal-like, luminal A, luminal B, HER2-enriched, basal-like). In conclusion, computer-extracted image phenotypes show promise for high-throughput discrimination of breast cancer subtypes and may yield a quantitative predictive signature for advancing precision medicine.

Original languageEnglish (US)
Article number16012
Journalnpj Breast Cancer
Volume2
Issue number1
DOIs
StatePublished - Dec 14 2016

Fingerprint

Breast Neoplasms
Phenotype
Neoplasms
ROC Curve
Precision Medicine
Molecular Models
National Cancer Institute (U.S.)
Entropy
Discriminant Analysis
Progesterone Receptors
Estrogen Receptors
Area Under Curve
Magnetic Resonance Spectroscopy
Immunohistochemistry
Datasets
Biopsy

ASJC Scopus subject areas

  • Oncology
  • Radiology Nuclear Medicine and imaging
  • Pharmacology (medical)

Cite this

Quantitative MRI radiomics in the prediction of molecular classifications of breast cancer subtypes in the TCGA/TCIA data set. / Li, Hui; Zhu, Yitan; Burnside, Elizabeth S.; Huang, Erich; Drukker, Karen; Hoadley, Katherine A.; Fan, Cheng; Conzen, Suzanne D.; Zuley, Margarita; Net, Jose M; Sutton, Elizabeth; Whitman, Gary J.; Morris, Elizabeth; Perou, Charles M.; Ji, Yuan; Giger, Maryellen L.

In: npj Breast Cancer, Vol. 2, No. 1, 16012, 14.12.2016.

Research output: Contribution to journalArticle

Li, H, Zhu, Y, Burnside, ES, Huang, E, Drukker, K, Hoadley, KA, Fan, C, Conzen, SD, Zuley, M, Net, JM, Sutton, E, Whitman, GJ, Morris, E, Perou, CM, Ji, Y & Giger, ML 2016, 'Quantitative MRI radiomics in the prediction of molecular classifications of breast cancer subtypes in the TCGA/TCIA data set', npj Breast Cancer, vol. 2, no. 1, 16012. https://doi.org/10.1038/npjbcancer.2016.12
Li, Hui ; Zhu, Yitan ; Burnside, Elizabeth S. ; Huang, Erich ; Drukker, Karen ; Hoadley, Katherine A. ; Fan, Cheng ; Conzen, Suzanne D. ; Zuley, Margarita ; Net, Jose M ; Sutton, Elizabeth ; Whitman, Gary J. ; Morris, Elizabeth ; Perou, Charles M. ; Ji, Yuan ; Giger, Maryellen L. / Quantitative MRI radiomics in the prediction of molecular classifications of breast cancer subtypes in the TCGA/TCIA data set. In: npj Breast Cancer. 2016 ; Vol. 2, No. 1.
@article{2586a12a9a08499480a062b706621075,
title = "Quantitative MRI radiomics in the prediction of molecular classifications of breast cancer subtypes in the TCGA/TCIA data set",
abstract = "Using quantitative radiomics, we demonstrate that computer-extracted magnetic resonance (MR) image-based tumor phenotypes can be predictive of the molecular classification of invasive breast cancers. Radiomics analysis was performed on 91 MRIs of biopsy-proven invasive breast cancers from National Cancer Institute’s multi-institutional TCGA/TCIA. Immunohistochemistry molecular classification was performed including estrogen receptor, progesterone receptor, human epidermal growth factor receptor 2, and for 84 cases, the molecular subtype (normal-like, luminal A, luminal B, HER2-enriched, and basal-like). Computerized quantitative image analysis included: three-dimensional lesion segmentation, phenotype extraction, and leave-one-case-out cross validation involving stepwise feature selection and linear discriminant analysis. The performance of the classifier model for molecular subtyping was evaluated using receiver operating characteristic analysis. The computer-extracted tumor phenotypes were able to distinguish between molecular prognostic indicators; area under the ROC curve values of 0.89, 0.69, 0.65, and 0.67 in the tasks of distinguishing between ER+ versus ER−, PR+ versus PR−, HER2+ versus HER2−, and triple-negative versus others, respectively. Statistically significant associations between tumor phenotypes and receptor status were observed. More aggressive cancers are likely to be larger in size with more heterogeneity in their contrast enhancement. Even after controlling for tumor size, a statistically significant trend was observed within each size group (P = 0.04 for lesions ≤ 2 cm; P = 0.02 for lesions >2 to≤ 5 cm) as with the entire data set (P-value = 0.006) for the relationship between enhancement texture (entropy) and molecular subtypes (normal-like, luminal A, luminal B, HER2-enriched, basal-like). In conclusion, computer-extracted image phenotypes show promise for high-throughput discrimination of breast cancer subtypes and may yield a quantitative predictive signature for advancing precision medicine.",
author = "Hui Li and Yitan Zhu and Burnside, {Elizabeth S.} and Erich Huang and Karen Drukker and Hoadley, {Katherine A.} and Cheng Fan and Conzen, {Suzanne D.} and Margarita Zuley and Net, {Jose M} and Elizabeth Sutton and Whitman, {Gary J.} and Elizabeth Morris and Perou, {Charles M.} and Yuan Ji and Giger, {Maryellen L.}",
year = "2016",
month = "12",
day = "14",
doi = "10.1038/npjbcancer.2016.12",
language = "English (US)",
volume = "2",
journal = "npj Breast Cancer",
issn = "2374-4677",
publisher = "Nature Publishing Group",
number = "1",

}

TY - JOUR

T1 - Quantitative MRI radiomics in the prediction of molecular classifications of breast cancer subtypes in the TCGA/TCIA data set

AU - Li, Hui

AU - Zhu, Yitan

AU - Burnside, Elizabeth S.

AU - Huang, Erich

AU - Drukker, Karen

AU - Hoadley, Katherine A.

AU - Fan, Cheng

AU - Conzen, Suzanne D.

AU - Zuley, Margarita

AU - Net, Jose M

AU - Sutton, Elizabeth

AU - Whitman, Gary J.

AU - Morris, Elizabeth

AU - Perou, Charles M.

AU - Ji, Yuan

AU - Giger, Maryellen L.

PY - 2016/12/14

Y1 - 2016/12/14

N2 - Using quantitative radiomics, we demonstrate that computer-extracted magnetic resonance (MR) image-based tumor phenotypes can be predictive of the molecular classification of invasive breast cancers. Radiomics analysis was performed on 91 MRIs of biopsy-proven invasive breast cancers from National Cancer Institute’s multi-institutional TCGA/TCIA. Immunohistochemistry molecular classification was performed including estrogen receptor, progesterone receptor, human epidermal growth factor receptor 2, and for 84 cases, the molecular subtype (normal-like, luminal A, luminal B, HER2-enriched, and basal-like). Computerized quantitative image analysis included: three-dimensional lesion segmentation, phenotype extraction, and leave-one-case-out cross validation involving stepwise feature selection and linear discriminant analysis. The performance of the classifier model for molecular subtyping was evaluated using receiver operating characteristic analysis. The computer-extracted tumor phenotypes were able to distinguish between molecular prognostic indicators; area under the ROC curve values of 0.89, 0.69, 0.65, and 0.67 in the tasks of distinguishing between ER+ versus ER−, PR+ versus PR−, HER2+ versus HER2−, and triple-negative versus others, respectively. Statistically significant associations between tumor phenotypes and receptor status were observed. More aggressive cancers are likely to be larger in size with more heterogeneity in their contrast enhancement. Even after controlling for tumor size, a statistically significant trend was observed within each size group (P = 0.04 for lesions ≤ 2 cm; P = 0.02 for lesions >2 to≤ 5 cm) as with the entire data set (P-value = 0.006) for the relationship between enhancement texture (entropy) and molecular subtypes (normal-like, luminal A, luminal B, HER2-enriched, basal-like). In conclusion, computer-extracted image phenotypes show promise for high-throughput discrimination of breast cancer subtypes and may yield a quantitative predictive signature for advancing precision medicine.

AB - Using quantitative radiomics, we demonstrate that computer-extracted magnetic resonance (MR) image-based tumor phenotypes can be predictive of the molecular classification of invasive breast cancers. Radiomics analysis was performed on 91 MRIs of biopsy-proven invasive breast cancers from National Cancer Institute’s multi-institutional TCGA/TCIA. Immunohistochemistry molecular classification was performed including estrogen receptor, progesterone receptor, human epidermal growth factor receptor 2, and for 84 cases, the molecular subtype (normal-like, luminal A, luminal B, HER2-enriched, and basal-like). Computerized quantitative image analysis included: three-dimensional lesion segmentation, phenotype extraction, and leave-one-case-out cross validation involving stepwise feature selection and linear discriminant analysis. The performance of the classifier model for molecular subtyping was evaluated using receiver operating characteristic analysis. The computer-extracted tumor phenotypes were able to distinguish between molecular prognostic indicators; area under the ROC curve values of 0.89, 0.69, 0.65, and 0.67 in the tasks of distinguishing between ER+ versus ER−, PR+ versus PR−, HER2+ versus HER2−, and triple-negative versus others, respectively. Statistically significant associations between tumor phenotypes and receptor status were observed. More aggressive cancers are likely to be larger in size with more heterogeneity in their contrast enhancement. Even after controlling for tumor size, a statistically significant trend was observed within each size group (P = 0.04 for lesions ≤ 2 cm; P = 0.02 for lesions >2 to≤ 5 cm) as with the entire data set (P-value = 0.006) for the relationship between enhancement texture (entropy) and molecular subtypes (normal-like, luminal A, luminal B, HER2-enriched, basal-like). In conclusion, computer-extracted image phenotypes show promise for high-throughput discrimination of breast cancer subtypes and may yield a quantitative predictive signature for advancing precision medicine.

UR - http://www.scopus.com/inward/record.url?scp=85054102101&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85054102101&partnerID=8YFLogxK

U2 - 10.1038/npjbcancer.2016.12

DO - 10.1038/npjbcancer.2016.12

M3 - Article

AN - SCOPUS:85054102101

VL - 2

JO - npj Breast Cancer

JF - npj Breast Cancer

SN - 2374-4677

IS - 1

M1 - 16012

ER -