Recent studies show that morphology based analysis of ductal lavage specimens failed to detect many cancers in women with breast cancer. Such an observation raises doubts about the potential role of ductal lavage in an individual's risk assessment and early detection of breast cancer. We hypothesize that biomarker-based analysis using markers of malignancy field defects including DNA 5c exceeding rate (DNA 5cER) and G-actin might provide a more reliable test for breast cancer risk. The study was performed in 2 phases, the training and validation phase. For the training phase, 36 Chinese women were recruited (13 with breast cancer, 8 with intraductal papilloma and 15 with benign breast diseases). The validation phase included 10 women with cancer and 7 women without cancer. Ductal lavage samples were processed by the ThinPrep technique and evaluated by morphology followed by biomarker analysis using laser scan cytometry (LSC) for G-actin and DNA5cER. In the training phase, biomarker analysis was performed on the 67% (24 of 36) of samples that had over 100 epithelial cells. The sensitivity of DNA5cER was 90% with a specificity of 100%, and G-actin was 100% and 93%, respectively. By contrast, the sensitivity and specificity obtained by cytology alone were 67% and 93%, respectively. Similar results were obtained from the small validation study. Quantitative analysis of biomarkers for G-actin and DNA5cER is feasible and useful in distinguishing benign from malignant breast disease on archived ductal lavage slides. Further studies are warranted to determine the value of these biomarkers in prospective trials.
2006 Wiley-Liss, Inc.