Background parenchymal enhancement in CEM and MRI: Is there always a high agreement?

Introduction

Background parenchymal enhancement (BPE) refers to the physiological enhancement of breast fibroglandular tissue. This study aimed to determine the agreement of BPE evaluation between contrast enhanced mammography (CEM) and magnetic resonance imaging (MRI) and investigate potential confounders.

Materials and methods

This retrospective, IRB-approved study included women recalled from screening or with inconclusive findings on mammography and/or ultrasound, who underwent both CEM and MRI between 2018 and 2022. Imaging protocols followed international recommendations. Two breast radiology fellows assessed density and BPE on CEM and MRI using BI-RADS, evaluating the early post-contrast acquisition (EP). In addition, BPE on MRI was assessed in the late post-contrast acquisition (LP). Data on menopausal status, breast thickness, and compression force were retrieved. Univariate and multivariate analyses were used.

Results

343 patients were consecutively enrolled. Post-menopausal patients showed lower BPE levels (p < 0.001). Agreement on BPE levels between CEM and MRI in the EP was fair (R1 κ = 0.342, R2 κ = 0.383). In the LP agreement was moderate for R1 (κ = 0.432) and fair for R2 (κ 0.368). There was a significant correlation between BPE and density in CEM (R1 ρ 0.399, R2 ρ 0.441; p < 0.001). A negative correlation was found on CEM between BPE and compression force (ρ −0.100, p = 0.001). Inter-reader agreement was high with both modalities (CEM κ = 0.678, MRI-EP κ = 0.690, MRI-EP κ = 0.692).

Conclusions

The agreement of BPE assessment between MRI and CEM was fair to moderate and reproducibility among readers was high. Compression force and mammographic breast density seem to influence BPE levels in CEM.