The relationship between background parenchymal enhancement, amount of fibroglandular tissue and synchronous contralateral breast cancer in preoperative magnetic resonance imaging of newly diagnosed breast cancer patients.

objectiveThis study aims to evaluate whether the amount of fibroglandular tissue (FGT) and the level of background parenchymal enhancement (BPE) on magnetic resonance imaging (MRI) are associated with synchronous contralateral breast cancer (CBC) during preoperative MRI staging of newly diagnosed breast cancer patients. MATERIALS AND

methodsFrom January 2010 to December 2019, core needle biopsy-confirmed newly diagnosed breast cancer patients who underwent preoperative bilateral breast MRI were screened. 61 eligible patients with pathologically proven synchronous CBC were enrolled as the study group, while 122 matched primary breast cancer patients without synchronous CBC were selected as the control group.Odds ratios (ORs) and corresponding 95% confidence intervals(CIs) for BPE and FGT as predictors of synchronous CBC were analyzed by conditional logistic regression. The factors with significant differences between the study group and the control group were further analyzed by performing receiver operating characteristic (ROC) curve analysis to determine the area under the curve (AUC), sensitivity, and specificity. BPE and FGT were independently evaluated from MRI scans by Reader 1 and Reader 2, and their findings were statistically analyzed independently. The DeLong test was used to compare the statistical significance of AUC differences between Reader 1 and Reader 2.Kappa analysis was used to calculate the level of agreement between their findings.

resultsBPE level was associated with synchronous CBC. Compared to women with minimal or mild BPE, those with moderate or marked BPE were more likely to have synchronous CBC (Reader 1: OR = 2.90, P < 0.01; Reader 2: OR = 3.36, P = 0.02). ROC showed the effectiveness of BPE level in differentiating contralateral CBC patients from control subjects with reader 1 (AUC 0.66, sensitivity 67.2%, specificity 63.1%) and reader 2 (AUC 0.67, sensitivity 80.3%, specificity 53.3%). No statistically significant difference in AUC was detected between Reader 1 and Reader 2 (DeLong test: 95% CI: -0.03-0.06; Z = 0.823, P = 0.410).No correlation was found between the amount of FGT and synchronous CBC. Further, the kappa values indicated substantial agreement between Reader 1 and Reader 2 with regard to the BPE and FGT values.

conclusionHigh BPE level shows a correlation with an increased risk of synchronous CBC in newly diagnosed breast cancer patients; however, this association does not establish independence and could be influenced by unmeasured confounders. Thus, BPE alone cannot guide clinical decisions.