The sensitivity of intraductal approaches to detect or predict breast cancer is a compound measurement that reflects both the sampling and analysis of fluids. Although cytology has been used as the 'gold standard' in early studies to evaluate breast fluids for risk assessment and cancer detection, it has been proven to have low sensitivity for detecting malignancy and reveals a high degree of prognostically ambiguous atypia. In an effort to address these issues, the analysis of breast fluids has been extended to a large variety of biomarkers.
Some investigators have taken an in situ
approach in order to colocalize markers and cytology to the same cells. Savitri Krishmanurthy (MD Anderson Cancer Center, Houston, TX, USA) described the use of interphase fluorescent in situ
hybridization (FISH) to measure chromosomal aneusomy on previously Pap-stained NAF cytology specimens [17
]. In previous studies she detected aneusomy in all of the cytologically malignant but only a subset of the cytologically atypical samples. She discussed the potential utility of FISH as an adjunct for confirming benign and malignant cytologic findings, and for further stratifying the classification of cytologic atypia. In a similar strategy to preserve cytomorphologic analysis, JianYu Rao (UCLA, Los Angeles, CA, USA) used quantitative fluorescence image analysis to analyze markers in ductal lavage thin prep specimens. Measurement of DNA 5cER (exceeding rate) content and G-actin overexpression in a small validation study resulted in 100% sensitivity and 100% specificity, suggesting that biomarker analysis may have better potential for separating cancer and noncancer samples than cytology.
Breast fluid specimens typically contain large numbers of infiltrating cells as well as benign epithelial cells, creating a 'needle in a haystack' challenge for biomarker development. Methylation-specific PCR is particularly well suited to this task because of its ability to detect rare cells containing cancer-associated DNA methylation events [18
]. Sara Sukumar (Johns Hopkins University, Baltimore, MD, USA), who pioneered the application of these assays to ductal fluids, reported on her latest work to transition this approach from a subjective, gel-based method into an objective, automated, quantitative assay, termed QM-MSP (quantitative multiplex methylation-specific PCR). This approach combines the power of multiplex and real-time PCR to detect methylation events simultaneously at multiple loci, and increases the sensitivity of the approach to 1 in 104
copies of DNA [20
]. In a pilot study with Savitri Krishnamurthy and Henry Kuerer (MD Anderson Cancer Center, Houston, TX, USA), QM-MSP analysis was applied to Pap-stained NAF cytology slide cell scrapings. Analyzing a panel of 10 loci, methylated alleles were detected in all of the malignant, a subset of the atypical, and none of the normal samples. Dr Sukumar is currently working to apply QM-MSP to breast fluids collected by NAF, ductal lavage, and fine needle aspiration, as well as plasma.
Edward Sauter (University of Missouri, Columbus, MO, USA) also described the application of hypermethylation assays to DNAs extracted from 22 matched pairs of NAF and breast tumor tissue samples collected from breast cancer patients [19
]. Using a panel of six loci, he detected hypermethylation in all of the malignant tissues and identical methylation patterns in 82% of the matched NAF specimens. No hypermethylation was detected in the normal tissues. He also used a panel of 11 microsatellite markers to evaluate matched specimens for loss of heterozygosity (LOH), and microsatellite instability [21
]. Identical alterations were identified in 43% of fluids from matched cancer tissues harboring changes. Luciane Cavalli (Georgetown University, Washington, DC, USA), working to develop early detection strategies for BRCA1
mutation carriers, reported on the development of protocols to isolate free DNA from the acellular fraction of ductal lavage fluids for molecular evaluation, an approach that leaves the cellular fraction intact for cytologic analysis [22
]. Using PCR-based approaches on these DNAs, she demonstrated LOH at the BRCA1
locus in four out of nine known BRCA1
mutation carriers. Gillian Mitchell and Imogen Locke (The Royal Marsden NHS Foundation Trust, London, UK) also reported detection of LOH at the BRCA1
loci in ductal lavage fluids from BRCA
-positive women. Finally, Sauter and Cavelli also described the detection of DNA mutations in mitochondrial DNA extracted from breast fluids – an advantageous approach because of the relative abundance of mitochondrial versus nuclear DNA [22
There were also two presentations describing the application of SELDI-TOF/MS (surface-enhanced laser desorption and ionization–time of flight/mass spectrometry) to nipple aspirate fluids. Timothy Pawlik (MD Anderson Cancer Center, Houston, TX, USA) analyzed the proteomic profiles of fluids collected from the diseased and contralateral breasts of 23 breast cancer patients, and compared these with the profiles of fluids collected from 10 healthy control inidividuals [24
]. Significant differences in the profiles between the fluids collected from patients and control inidividuals were detected, but interestingly the proteomic profiles of the cancerous and contralateral breast fluids of breast cancer patients were highly conserved. Sascha Dua (Royal Marsden Hospital Foundation Trust, London, UK) reported on her experience to date with a small pilot set of nipple aspirate fluids. In keeping with the observations presented by Pawlik, they also identified differentially expressed peaks between cancer and control fluids, but no significant differences between fluids from the healthy and cancerous breasts of the same patient. Both groups are working on identifing the protein peaks and will follow up with validation studies.
Seema Khan summarized a body of work describing the pathophysiology of the estrogen receptor (ER) in the human breast, exploring its potential significance as a marker of proliferation, and is currently performing studies to evaluate ER in breast fluid cells. Highlighting the potential use of intraductal approaches in chemoprevention trials, Arun Banu (MD Anderson Cancer Center, Houston, TX, USA) summarized preliminary results from a phase II prevention trial designed to test whether celecoxib induces changes in proliferation and apoptosis in the breast tissue of high-risk women. Using fine needle aspiration and ductal lavage samples, she observed significant downregulation of ER expression within 6 months of treatment. She is also in the process of analyzing other markers of proliferation and apoptosis.
Collectively, these studies demonstrate the successful application of biomarkers to detect breast cancer associated genetic, epigenetic, and proteomic abnormalities in breast fluids collected by a variety of intraductal techniques. In keeping with a growing body of published studies, markers appear to have superior sensitivity for breast cancer detection compared with cytology [14
]. Ed Sauter concluded his talk with the comment that no single marker has adequate sensitivity to detect all malignancies, and an important goal will be to create systematically a panel of markers with high sensitivity and specificity. The successful evaluation and validation of breast cancer detection biomarkers in breast fluids will require the accurate sampling of ducts associated with lesions.