Skip to main content

Lund University Publications

LUND UNIVERSITY LIBRARIES

A multiparametric approach to predict triple-negative breast cancer including parameters derived from ultrafast dynamic contrast-enhanced MRI

Ohashi, Akane LU ; Kataoka, Masako ; Iima, Mami ; Honda, Maya ; Ota, Rie ; Urushibata, Yuta ; Dominik Nickel, Marcel ; Toi, Masakazu ; Zackrisson, Sophia LU and Nakamoto, Yuji (2023) In European Radiology 33(11). p.8132-8141
Abstract

Objective: Triple-negative breast cancer (TNBC) is a highly proliferative breast cancer subtype. We aimed to identify TNBC among invasive cancers presenting as masses using maximum slope (MS) and time to enhancement (TTE) measured on ultrafast (UF) DCE-MRI, ADC measured on DWI, and rim enhancement on UF DCE-MRI and early-phase DCE-MRI. Methods: This retrospective single-center study, between December 2015 and May 2020, included patients with breast cancer presenting as masses. Early-phase DCE-MRI was performed immediately after UF DCE-MRI. Interrater agreements were evaluated using the intraclass correlation coefficient (ICC) and Cohen's kappa. Univariate and multivariate logistic regression analyses of the MRI parameters, lesion size,... (More)

Objective: Triple-negative breast cancer (TNBC) is a highly proliferative breast cancer subtype. We aimed to identify TNBC among invasive cancers presenting as masses using maximum slope (MS) and time to enhancement (TTE) measured on ultrafast (UF) DCE-MRI, ADC measured on DWI, and rim enhancement on UF DCE-MRI and early-phase DCE-MRI. Methods: This retrospective single-center study, between December 2015 and May 2020, included patients with breast cancer presenting as masses. Early-phase DCE-MRI was performed immediately after UF DCE-MRI. Interrater agreements were evaluated using the intraclass correlation coefficient (ICC) and Cohen's kappa. Univariate and multivariate logistic regression analyses of the MRI parameters, lesion size, and patient age were performed to predict TNBC and create a prediction model. The programmed death-ligand 1 (PD-L1) expression statuses of the patients with TNBCs were also evaluated. Results: In total, 187 women (mean age, 58 years ± 12.9 [standard deviation]) with 191 lesions (33 TNBCs) were evaluated. The ICC for MS, TTE, ADC, and lesion size were 0.95, 0.97, 0.83, and 0.99, respectively. The kappa values of rim enhancements on UF and early-phase DCE-MRI were 0.88 and 0.84, respectively. MS on UF DCE-MRI and rim enhancement on early-phase DCE-MRI remained significant parameters after multivariate analyses. The prediction model created using these significant parameters yielded an area under the curve of 0.74 (95% CI, 0.65, 0.84). The PD-L1-expressing TNBCs tended to have higher rim enhancement rates than the non-PD-L1-expressing TNBCs. Conclusion: A multiparametric model using UF and early-phase DCE-MRI parameters may be a potential imaging biomarker to identify TNBCs. Clinical relevance statement: Prediction of TNBC or non-TNBC at an early point of diagnosis is crucial for appropriate management. This study offers the potential of UF and early-phase DCE-MRI to offer a solution to this clinical issue. Key Points: • It is crucial to predict TNBC at an early clinical period. • Parameters on UF DCE-MRI and early-phase conventional DCE-MRI help in predicting TNBC. • Prediction of TNBC by MRI may be useful in determining appropriate clinical management.

(Less)
Please use this url to cite or link to this publication:
author
; ; ; ; ; ; ; ; and
organization
publishing date
type
Contribution to journal
publication status
published
subject
keywords
Biomarkers, Breast, Breast neoplasms, Magnetic resonance imaging, Triple-negative breast neoplasms
in
European Radiology
volume
33
issue
11
pages
10 pages
publisher
Springer
external identifiers
  • pmid:37286791
  • scopus:85161333068
ISSN
0938-7994
DOI
10.1007/s00330-023-09730-w
language
English
LU publication?
yes
id
1db46198-cece-463f-9b0f-bc11436a1a30
date added to LUP
2023-08-24 15:17:07
date last changed
2024-04-20 01:37:24
@article{1db46198-cece-463f-9b0f-bc11436a1a30,
  abstract     = {{<p>Objective: Triple-negative breast cancer (TNBC) is a highly proliferative breast cancer subtype. We aimed to identify TNBC among invasive cancers presenting as masses using maximum slope (MS) and time to enhancement (TTE) measured on ultrafast (UF) DCE-MRI, ADC measured on DWI, and rim enhancement on UF DCE-MRI and early-phase DCE-MRI. Methods: This retrospective single-center study, between December 2015 and May 2020, included patients with breast cancer presenting as masses. Early-phase DCE-MRI was performed immediately after UF DCE-MRI. Interrater agreements were evaluated using the intraclass correlation coefficient (ICC) and Cohen's kappa. Univariate and multivariate logistic regression analyses of the MRI parameters, lesion size, and patient age were performed to predict TNBC and create a prediction model. The programmed death-ligand 1 (PD-L1) expression statuses of the patients with TNBCs were also evaluated. Results: In total, 187 women (mean age, 58 years ± 12.9 [standard deviation]) with 191 lesions (33 TNBCs) were evaluated. The ICC for MS, TTE, ADC, and lesion size were 0.95, 0.97, 0.83, and 0.99, respectively. The kappa values of rim enhancements on UF and early-phase DCE-MRI were 0.88 and 0.84, respectively. MS on UF DCE-MRI and rim enhancement on early-phase DCE-MRI remained significant parameters after multivariate analyses. The prediction model created using these significant parameters yielded an area under the curve of 0.74 (95% CI, 0.65, 0.84). The PD-L1-expressing TNBCs tended to have higher rim enhancement rates than the non-PD-L1-expressing TNBCs. Conclusion: A multiparametric model using UF and early-phase DCE-MRI parameters may be a potential imaging biomarker to identify TNBCs. Clinical relevance statement: Prediction of TNBC or non-TNBC at an early point of diagnosis is crucial for appropriate management. This study offers the potential of UF and early-phase DCE-MRI to offer a solution to this clinical issue. Key Points: • It is crucial to predict TNBC at an early clinical period. • Parameters on UF DCE-MRI and early-phase conventional DCE-MRI help in predicting TNBC. • Prediction of TNBC by MRI may be useful in determining appropriate clinical management.</p>}},
  author       = {{Ohashi, Akane and Kataoka, Masako and Iima, Mami and Honda, Maya and Ota, Rie and Urushibata, Yuta and Dominik Nickel, Marcel and Toi, Masakazu and Zackrisson, Sophia and Nakamoto, Yuji}},
  issn         = {{0938-7994}},
  keywords     = {{Biomarkers; Breast; Breast neoplasms; Magnetic resonance imaging; Triple-negative breast neoplasms}},
  language     = {{eng}},
  number       = {{11}},
  pages        = {{8132--8141}},
  publisher    = {{Springer}},
  series       = {{European Radiology}},
  title        = {{A multiparametric approach to predict triple-negative breast cancer including parameters derived from ultrafast dynamic contrast-enhanced MRI}},
  url          = {{http://dx.doi.org/10.1007/s00330-023-09730-w}},
  doi          = {{10.1007/s00330-023-09730-w}},
  volume       = {{33}},
  year         = {{2023}},
}