Apparent diffusion coefficient of breast cancer and normal fibroglandular tissue in diffusion-weighted imaging: the effects of menstrual cycle and menopausal status.

Abstract

The purpose of this study was to investigate prospectively whether the apparent diffusion coefficients (ADCs) of both breast cancer and normal fibroglandular tissue vary with the menstrual cycle and menopausal status. Institutional review board approval was obtained, and informed consent was obtained from each participant. Fifty-seven women (29 premenopausal, 28 postmenopausal) with newly diagnosed breast cancer underwent diffusion-weighted imaging twice (interval 12-20days) before surgery. Two radiologists independently measured ADC of breast cancer and normal contralateral breast tissue, and we quantified the differences according to the phases of menstrual cycle and menopausal status. With normal fibroglandular tissue, ADC was significantly lower in postmenopausal than in premenopausal women (P=0.035). In premenopausal women, ADC did not differ significantly between proliferative and secretory phases in either breast cancer or normal fibroglandular tissue (P=0.969 and P=0.519, respectively). In postmenopausal women, no significant differences were found between ADCs measured at different time intervals in either breast cancer or normal fibroglandular tissue (P=0.948 and P=0.961, respectively). The within-subject variability of the ADC measurements was quantified using the coefficient of variation (CV) and was small: the mean CVs of tumor ADC were 2.90% (premenopausal) and 3.43% (postmenopausal), and those of fibroglandular tissue ADC were 4.37% (premenopausal) and 2.55% (postmenopausal). Both intra- and interobserver agreements were excellent for ADC measurements, with intraclass correlation coefficients in the range of 0.834-0.974. In conclusion, the measured ADCs of breast cancer and normal fibroglandular tissue were not affected significantly by menstrual cycle, and the measurements were highly reproducible both within and between observers.