Abstract Background: It has been reported that MRI-based density measurement is more sensitive than mammographic density to evaluate the small changes of breast density. This study analyzed the change of breast density in women receiving tamoxifen treatment using 3-D MRI.Material and Methods: Twenty-two breasts in 20 women were studied. Two MRI studies were performed at 1.5T, the baseline scan acquired after surgery and prior to receiving tamoxifen, and the follow-up scan at 10 months to 30 months after tamoxifen treatment. Non-fat-suppression T1-weighted images were used for the density measurement. The breast and the fibroglandular tissue were segmented using a computer-assisted algorithm recently published by our group. For each breast, the total fibroglandular tissue volume and the total breast volume were calculated, and then the ratio was calculated as the percent breast density (%BD). The changes in breast volume, fibroglandular tissue volume, and the percent density in the F/U scan compared to the B/L scan were evaluated. The reduction in the fibroglandular volume (ΔFV=FV2-FV1, ml), and in the percent breast density (Δ%BD= %BD2-%BD1) were analyzed.Results: The total breast volume did not show significant changes between F/U and B/L. For analysis of fibroglandular volume, 20 breasts showed volume reduction and 2 showed slight increase of volume. Overall, the mean ΔFV was 20.7ml, with 33% reduction. The ΔFV was correlated with the baseline FV (r=0.82) (Figure 1), indicating that when the fibroglandular tissue volume was higher, more reduction was found.For analysis of %BD, 19 breasts showed reduction, and 3 showed increased density, with a mean Δ%BD of 4.7%. The 22 breasts were further divided into three categories based on their B/L density as: fatty breast (%BD< 8%, N=4), moderately dense breast (%BD 8-20%, N=9), and dense breast (%BD >20%, N=9). The amount of density reduction was associated with the B/L density, showing 10.9, 17.2, and 28.6 ml for ΔFV; and 1.2%, 3.9%, and 7.2% for Δ%BD respectively in these 3 groups. Statistically, only Δ%BD between the fatty and the dense groups showed a significant difference (P=0.02).Discussion: MRI provides strong soft tissue contrast, thus it may have advantages for revealing detailed density distribution compared to mammography. The %BD can be measured by normalizing the FV to the total breast volume, analogous to the mammographic density. However, this will be subjective to the variations in the measurement of breast volume. The FV within the breast was a more robust parameter, which showed reduction in 20 of 22 breasts with a mean reduction of 33%, also a significant correlation with B/L FV. The reduction of density is expected to be through decreased hormonal level in the breast, and the reduction of fibroglandular tissue density should be correlated with the baseline density. Our results are consistent with this expectation, and demonstrate that MRI can be used to evaluate the density change induced by tamoxifen. Citation Information: Cancer Res 2009;69(24 Suppl):Abstract nr 4021.
Read full abstract