Quantitative Values from Synthetic MRI Correlate with Breast Cancer Subtypes

Abstract

Purpose To correlate quantitative T1, T2, and proton density (PD) values from synthetic MRI with breast cancer subtypes. Methods Twenty-eight breast cancer patients underwent MRI of the breast including synthetic MRI. T1, T2, and PD values were correlated with Ki-67. T1, T2, and PD values were compared between estrogen receptor (ER) positive and ER negative cancers, and between Luminal A and Luminal B cancers. The effectiveness of T1, T2, and PD in differentiating the ER-negative from the ER-positive group and Luminal A from Luminal B cancers was evaluated using receiver operating characteristic analysis.Results Mean T2 relaxation of ER-negative cancers was significantly higher than that of ER-positive cancers (p < .05). The T1, T2, and PD values exhibited a strong positive correlation with Ki-67 (Pearson's r = 0.75, 0.69, and 0.60 respectively; p < .001). Among ER-positive cancers (n=23), T1, T2, and PD values of Luminal A cancers were significantly lower than those of Luminal B cancers (p < .05). The areas under the curve (AUCs) of T1, T2, and PD for discriminating ER-negative from ER-positive cancers were 0.74 (95% confidence interval (CI): 0.54-0.88), 0.87 (95% CI: 0.69-0.97), and 0.62 (95% CI: 0.42-0.79), respectively. The AUCs of T1, T2, and PD values for discriminating Luminal A from Luminal B cancers were 0.83 (95% CI: 0.61-0.95), 0.75 (95% CI: 0.52-0.90), and 0.75 (95% CI: 0.53-0.91), respectively.Conclusion Quantitative values from synthetic MRI significantly correlate with subtypes of invasive breast cancers and may classify subtypes with reasonably good accuracy.