For personal use. Only reproduce with permission from The Lancet Publishing Group. RESEARCH LETTERS These cases indicate the promising potential of the MSP- based method for early detection of breast malignancy, before the appearance of suspicious findings on mammography. MSP confirmed the cytological finding that led to the diagnosis of breast cancer in two women. In combination with cytology evaluation, MSP of ductal lavage could provide a useful adjunct to mammography in the early diagnosis of breast cancer. We thank Kyle Terrell and Heather Lewin, Indira Debchoudhury, and Dorian Korz for assistance; Bert Vogelstein, David Sidransky, Donald Coffey, and Alan Rein for reviewing the paper; and the Arthur and Rochelle Belfer Tissue Bank, Susan G Komen Foundation (BCTR 2000 577 to SS), The American Breast Cancer Foundation, and the NIH P50 CA88843 for grant support. 1 Elmore JG, Barton MB, Moceri VM, Polk S, Arena PJ, Fletcher SW. Ten-year risk of false positive screening mammograms and clinical breast examinations. N Engl J Med 1998; 338: 1089–96. 2 Evron E, Umbricht CB, Korz D, et al. Loss of cyclin D2 expression in the majority of breast cancers is associated with promoter hypermethylation. Cancer Res 2001; 61: 2782–87. 3 Sirchia SM, Ferguson AT, Sironi E, et al. Evidence of epigenetic changes affecting the chromatin state of the retinoic acid receptor beta2 promoter in breast cancer cells. Oncogene 2000; 19: 1556–63. 4 Herman JG, Graff JR, Myohanen S, Nelkin BD, Baylin SB. Methylation-specific PCR: a novel PCR assay for methylation status of CpG islands. Proc Natl Acad Sci USA 1996; 93: 9821–26. 5 Dooley WC. Endoscopic visualization of breast tumors. JAMA 2000; 284: 1518. Johns Hopkins University School of Medicine, Baltimore MD 21231, USA (E Evron, MD, W C Dooley MD, C B Umbricht MD, D Rosenthal MD, N Sacchi PhD, E Gabrielson MD, N E Davidson MD, S Sukumar PhD); UCSF School of Medicine, San Francisco, CA (B-M Ljung MD); and Pro·Duct Health, Menlo Park, CA (A B Soito BS, D T Hung MD) Correspondence to: Dr Saraswati Sukumar (e-mail: saras@jhmi.edu) to 20 mL of saline was introduced in incremental volumes to flush out epithelial cells from the ducts and lobules. The ductal fluid was placed immediately in cytology fixative and prepared with standard millipore filtration devices for cytology assessment and DNA extraction. We recruited 37 women with biopsy-proven cancer. Women underwent ROBE immediately before definitive surgery and after signing an informed consent form. DNA from both the ductal fluid cells and the matching surgical samples was tested with methylation-specific PCR (MSP) for Cyclin D2, RAR-β, and Twist. 2,3 Methylated alleles of at least one of three markers were detected in 17 of 20 irrigation fluid samples from patients with pathology-confirmed invasive carcinoma (table). Healthy breast tissue contained only unmethylated genes (zero samples of 20; table). Methylated alleles for RAR-β only were noted in two of 56 samples (table). By contrast, irrigation fluid from four patients who underwent re-excision, but were subsequently found to be tumour-free, contained only unmethylated markers (table). Irrigation fluid from two of seven patients with DCIS (Grade 1-3), and two of six patients with atypical ductal hyperplasias contained hypermethylated markers. DNA samples from 19 of the 20 excised tumour samples were positive by MSP for the presence of methylated markers. Analysis of the irrigation fluid thus missed two MSP-positive samples, presumably because of the low cell yields. Cytology analysis on this fluid was inconclusive in 23 samples due to inadequate cellularity, and no malignant cells were detected in the remaining samples. These results suggest that MSP is sensitive, as the techinique detected cancer cells in 85% of ductal fluid samples from patients with breast malignancy, including cases where the material was inadequate for cytology. We extended our analysis to 56 samples of ductal lavage fluid (obtained after informed consent) from women with non-suspicious mammograms and breast examinations, but at high risk for developing breast cancer (as defined by a Gail index 1·7, previous history of contralateral breast cancer, or BRCA1 and BRCA2 mutations). Using cytopathology, 50 samples were classified as benign or with mild changes, and six samples were classified as atypical with substantial changes or frankly malignant (figure 1). Among the cases with substantially abnormal cells or malignant cells, four of six samples were identified by MSP (67% sensitivity), whereas only five of 45 benign cases were positive (89% specificity; figure 2). Pathologically confirmed breast cancer was subsequently diagnosed in two women with abnormal cytological findings and MSP-positive ductal lavage fluid. A third patient in this category is undergoing further assessment. 1336 THE LANCET • Vol 357 • April 28, 2001 MUC 1: a genetic susceptibility to infertility? Andrew W Horne, John O White, Raul A Margara, Ross Williams, Robert M L Winston, El-Nasir Lalani In man and some animals regulation of embryo implantation by endometrial expression of the highly polymorphic MUC 1 mucin has been suggested. We assessed the polymorphism of MUC 1 in women known to be fertile and those with infertility due to suspected failure of embryo implantation. The median of the lower allele size in the infertile group was only 2⋅5 kb compared with 3⋅4 kb in the fertile group (p=0⋅0029, difference 0·9, [95% CI 0⋅1–1⋅3]). Women with unexplained infertility might have a genetic susceptibility to failure of embryo implantation due to small MUC 1 allele size. Despite thorough investigation many cases of infertility remain unexplained. Although morphologically normal embryos are transferred to the uterus in most in-vitro fertilisation (IVF) cycles, successful pregnancy only takes place in about one in five attempts. Since at least 50% of IVF embryos develop to the blastocyst stage in culture, failure of implantation is probably the reason for failure of treatment. 1 The essential cellular factors in endometrium that contribute to implantation are not fully understood. MUC 1 mucin, an oxygen-glycosylated (O) epithelial glycoprotein, could potentially modulate embryo attachment. It extends beyond the endometrial glycocalyx and is probably the first molecule that the embryo encounters on attachment. 2 Figure 2: MSP profiles of ductal lavage fluid U=unmethylated. M=methylated. 231=breast cancer cell line MDAMB 231. From women with invasive cancer and “malignant” cytology, and two women with ductal carcinoma in situ (DCIS1, DCIS2) and a cytology diagnosis of “atypical cells with marked changes”. MSP primer sequences were: RAR-β: U Forward-5’GGATTGGGATGTTGAGAATGT3’; U Reverse-5’CAACC- AATCCAACCAAAACAA3’; M Forward 5’GAACGCGAGCGATTCGAGT3’; M Reverse-5’GACCAATCCAACCGAAACG3’. Twist: U Forward-5’TTTGGATGGG- GTTGTTATTGT3’; U Reverse 5’CCTAACCCAAACAACCAACC3’ M Forward- 5’TTTCGGATGGGGTTGTTATC3’; M Reverse-5’AAACGACCTAACCCGAACG3’. Cyclin D2 analysis has been previously described. 2