Breast ImagingE3264. Qualitative and Quantitative Computer Imaging Analysis of Breast MRI BPE and BPU in FDG PET/CT
Ha R, Mema E, Guo X, Wynn R, Zhao B. Columbia University Medical Center, New York, NY
Address correspondence to R. Ha (firstname.lastname@example.org)
Objective: Our objective is to study the possible association between breast MRI background parenchymal enhancement (BPE) and metabolic activity assessed by background parenchymal uptake (BPU) on FDG PET/CT and its potential as another biomarker of breast cancer risk.
Materials and Methods: An institutional review board–approved study of 30 patients with no abnormal findings in one breast was performed. MRI studies were performed in sagittal orientation at 3-T field strength using a dedicated breast coil. Qualitative MRI BPE assessments were performed on a 4-point scale by consensus of three fellowship-trained breast radiologists who were blinded to history in accordance with BI-RADS categories. Quantitative BPE analysis was performed using a published and validated in-house technique. For FDG PET/CT, after image reconstruction, a 3D ROI was carefully drawn around the glandular breast tissue, determined by visual inspection by a breast radiologist. From these ROIs, the maximum standardized uptake value (SUVmax) was calculated. Statistical tests were used to evaluate the association between BPE and BPU including t test, Spearman correlation test, and linear regression analysis.
Results: The BPU SUVmax values strongly correlated with both the qualitative and quantitative measurements of BPE (r = 0.59 [ p< 0.001] and r = 0.50 [p < 0.001], respectively]. In addition, quantified mean signal intensity (SI) and median SI values of BPE moderately correlated with BPU SUVmax (r = 0.41 [p < 0.033] and r = 0.42 [p < 0.032], respectively]. The qualitatively assessed low BPE group (BI-RADS category 1 or 2) had a mean BPU SUVmax of 1.18 (SD = 0.32). The high BPE group (BI-RADS category 3 or 4) had a mean BPU SUVmax of 1.96 (SD = 0.44). The two groups differed significantly (high BPE group, t = 14.64 [p < 0.001] vs low BPE group, t = 13.13 [p < 0.001]). On linear regression analysis, BPU SUVmax significantly predicted qualitative and quantitative measurements of BPE (B = 1.25, t = 3.85 [p < 0.001]; B = 18.89, t = 3.59 [p < 0.001], respectively). BPU SUVmax also explained a significant proportion of variance in qualitative and quantitative measurements of BPE (R2 = 0.34, F (1, 28) = 14.83 [p < 0.001]; R2 = 0.31, F (1, 28) = 12.66 [p < 0.001], respectively).
Conclusion: There is a significant association between breast MRI BPE and PET BPU. The increased risk for breast cancer in patients with high BPE could be related to elevated metabolic activity of the normal breast tissue. In addition, SUVmax value may serve as another biomarker of breast cancer risk.