Radiomic differentiation of breast cancer molecular subtypes using pre-operative breast imaging - A systematic review and meta-analysis.

Abstract

INTRODUCTION:

Breast cancer has four distinct molecular subtypes which are discriminated using gene expression profiling following biopsy. Radiogenomics is an emerging field which utilises diagnostic imaging to reveal genomic properties of disease. We aimed to perform a systematic review of the current literature to evaluate the value radiomics in differentiating breast cancers into their molecular subtypes using diagnostic imaging.

METHODS:

A systematic review was performed as per PRISMA guidelines. Studies assessing radiomictumour analysis in differentiatingbreast cancer molecular subtypeswere included. Quality was assessed using the radiomics quality score (RQS). Diagnostic sensitivity and specificity of radiomic analyses were included for meta-analysis; Study specific sensitivity and specificity were retrieved and summary ROC analysis were performed to compile pooled sensitivities and specificities.

RESULTS:

Forty-one studies were included. Overall, there were 10,090 female patients (mean age of 47.6 ± 11.7 years, range: 21-93) and molecular subtypewas reported in 7,693 of cases, with Luminal A (LABC), Luminal B (LBBC), Human Epidermal Growth Factor Receptor-2 overexpressing (HER2+), and Triple Negative (TNBC) breast cancers representing 51.3%, 19.9%, 12.3% and 16.3% of tumour respectively. Seven studies provided radiomic analysis to determine molecular subtypes using mammography to differentiateTNBCvs.others (sensitivity: 0.82,specificity:0.79). Thirty-five studies reported on radiomic analysis of magnetic resonance imaging (MRI); LABC versus others(sensitivity:0.78,specificity:0.83),HER2+versusothers(sensitivity:0.87,specificity:0.88), andLBBCversusTNBC (sensitivity: 0.79,specificity:0.88) respectively.

CONCLUSION:

Radiomic tumour assessment of contemporary breast imaging provide a novel option in determining breast cancer molecular subtypes. However, amelioration of such techniques are required and genetic expression assessment will remain the gold standard.

• Authors: Matthew G Davey, Martin S Davey, Michael R Boland, Éanna J Ryan, Aoife J Lowery, Michael J Kerin
• Journal: European journal of radiology (Eur J Radiol) Vol. 144 Pg. 109996 (Nov 2021) ISSN: 1872-7727 [Electronic] Ireland
• PMID: 34624649 (Publication Type: Journal Article, Meta-Analysis, Systematic Review)
• Copyright: Copyright © 2021 The Authors. Published by Elsevier B.V. All rights reserved.

Topics

• Adult
• Breast
• Breast Neoplasms (diagnostic imaging, genetics)
• Female
• Humans
• Magnetic Resonance Imaging
• Mammography
• Middle Aged
• ROC Curve