A morphological classification of breast in menstrual cycle

Studies of breast cancer risk in premenopausal women using archived normal/ benign breast epithelium are hampered by lack of reliable menstrual phase data.The knowledge of menstrual cycle phase (MCP) is important due to the cyclical variations observed in the breast through the menstrual cycle. We developed morphological criteria to classify H&E sections of the breast into follicular (F) and luteal (L) phase and used it in a blinded way to determine the MCP in the breast in a set of women who underwent breast biopsy for benign problems. Methods: 73 premenopausal women undergoing surgery were selected. The date of the last menstrual period, and the usual menstrual cycle length was ascertained pre-operatively. Normal lobules adjacent to benign lesions were reviewed by two pathologists (RR and SB). The parameters used in this classification included: distinction between the two layers of the acini, presence and the proportion of vacuolation of the myoepithelial layer in the acini, stromal edema, infiltrate, mitosis and apoptosis. Each of these was scored from 0 to 3, except the last two, which were scored from 0-2. The sum of the individual scores (total score) was used to assign 4 phases to represent early and late, follicular and luteal phases respectively (0-5;phase I, 6-9; phase II. 10-15, phase III and 16-19, phase IV). These were then correlated with the menstrual dates, which were standardized for a 28-day cycle. Serum levels for estradiol (E) and progesterone (P) were assayed in 34 of these patients at the time of surgery. Results: 40 and 39 patients were in F phase by morphology and dates respectively while 33 and 34 patients respectively belonged to L phase. 54 of the 73 (74%. p=0.001 ) patients were concordant for F and L phase by morphology and dates. The epithelial-myoepithelial layer distinction had the highest predictive value in determining the phase by morphology (R2 = 0.78). 25 (74%) of these were phase-concordant by morphology and P levels (p=0.01) and 25 (74%), were phase-concordant by dates and P levels (p=0.007).Women with a high score by morphology were 7 times as likely to be in luteal phase as women with a low score (OR 7.1, 95% CI 2.5-19.9). Conclusions: Histomophologic classification of breast tissue is a potentially useful way of determining MCP. When validated, this classification system may solve a thorny problem in the design of epidemiological investigations of breast cancer risk.