Tissue case-controls: What is normal?

Matched case-control tissue studies are an important tool in epidemiologic research. Benign biopsies (BB) are often used as controls because they are plentiful and accessible, while reduction mammoplasty (RM) tissue is less available. The disadvantages are that BB may not be truly "normal." since the procedure was prompted by an abnormality. While RM patients have no detectable abnormalities, they are usually much younger than cancer patients. Unless RM specimens are age-matched, differences due to aging and menopausal status could be a confounding variable. This study was undertaken to compare tissue from BB and RM to determine if there are significant differences when controlling for age. Records of 504 consecutive women undergoing RM at Northwestern Memorial Hospital between January 1992 and March 2001 were reviewed. Of those. 84 were cancer patients undergoing RM on the contralateral breast. For each case, an age-matched RM control was identified by choosing the first same-age specimen from the same year. If an exact age-match was not possible, the next closest age was used. Age-matched BB controls were the first age-match from consecutive excisional biopsies from the same year. The highest-risk lesion in each specimen was classified using criteria by Page and Dupont. Normal + Nonproliferative Fibroadenoma Only Proliferative Atypical Cancer Case RM 52 (61.9%) 4 (4.8%) 13 (15.4%) 8 (9.5%) 7 (8.3%) Benign RM 64 (76.2% 3 (3.6%) 13 (15.4%) 4 (4.8%) 0 Benign Biopsy 25 (29.8%) 17 (20.2%) 33 (39.3%) 9 (10.7%) 0 Mean age and range in the BB group was 49.7 (26-73), 48.8 (25-68) for RM. 49.5 (25-73) for cases, and 36.2 (16-74) for the entire benign RM cohort. Normal/ nonproliferative lesions and fibroadenomas were pooled for analysis. Using the two-sample test of proportion, BBs were found to be proliferative significantly more often than RM (50% vs. 20%, p<.001). Seven incidental cancers were found in case RMs: 6 with foci of DCIS, 1 with a 2 cm invasive cancer, in this dataset, RM tissue is more "normal" than BB. and should be the preferred choice for case-control studies.