«PacifiCare’s medical management guidelines represent the recommendation of the PacifiCare Medical Management Guideline (MMG) committee. They are ...»
ductal carcinoma in situ (DCIS) or lobular carcinoma in situ (LCIS); 5-year breast cancer risk according to Gail index 1.7%; BRCA1 or BRCA2 mutation-positivity; abnormal nipple discharge; or previous mastectomies for breast cancer or for breast cancer risk reduction (prophylactic mastectomies). Outcomes assessed include: adequacy of the cytological specimens; cell yield; number of cells per duct; detection of abnormal cells (mild atypia, marked atypia, or malignancy);
endoscopic and surgical biopsy findings; incidence of breast precancers and cancers; diagnostic accuracy; and complications. DL specimens are generally categorized as: 1) insufficient material for analysis, 2) benign cells, 3) atypical cells non-suspicious for malignancy, 4) atypical cells suspicious for malignancy, and 5) malignant cells.
Summary of the Evidence
One large, manufacturer-sponsored, multi-center, prospective clinical trial compared the efficacy of breast DL with nipple aspiration for the collection and cytological detection of cellular atypia in 507 women (700 breasts) at high risk for breast cancer. DL was significantly more likely than nipple aspiration to provide sufficient cells for cytological diagnosis (78% versus 27% of patients), resulted in the collection of a greater number of cells per breast, and resulted in significantly more diagnoses of cellular atypia or breast cancer (17% versus 4% for paired samples and 24% versus 10% for paired and unpaired samples combined). The number of epithelial cells per breast duct obtained during DL correlated significantly with the severity of the cytopathological changes found. The median numbers of cells collected per duct were as follows: benign, 4000; mild atypia, 13,400; marked atypia, 40,000; and malignant, 83,000. DCIS was diagnosed by histopathological examination in 4 of 11 (36%) patients with cellular abnormalities on DL that had imaging and surgery; however, no data are available on the 7 remaining patients. Interpretation of the data from this study is hampered by weaknesses in study design and execution such as the lack of follow-up and histopathological confirmation of the Cytological Examination of Breast Fluids for Cancer Screening- Commercial Medical Management Guideline findings for all patients as well as a lack of information on how the test results influenced health outcomes. (Dooley et al., 2001) In a cohort study, Carruthers et al. evaluated if ductal lavage could predict the occurrence of breast cancer as well as further stratify patients at high-risk for developing breast cancer. Ductal lavage was performed in 116 high-risk patients (Gail Risk score or = 1.7%, previous breast cancer, strong family history, previous suspicious biopsy specimen). If atypia or papillary cells were identified, a standard protocol of evaluation was initiated. Two hundred twenty-three lavages were performed on 116 patients. Twenty-seven lavages in 25 patients yielded atypical or papillary-like cells. The 15 patients who underwent further evaluation for atypia had no evidence of cancerous or precancerous lesions. All patients were followed-up: 2 developed breast cancer, both of whom had had normal previous lavage. No patient with abnormal lavage developed cancer during follow-up. The authors concluded ductal lavage to be of limited value in the screening of high-risk patients. (Carruthers et al., 2007) In a small cross-sectional study, Brogi et al. evaluated the correlation between cytological diagnoses obtained by DL and histopathological findings in 30 mastectomy specimens from 26 breast cancer patients and 4 patients undergoing prophylactic mastectomy. Twenty-nine DL samples were satisfactory for cytological examination. Of these, 27 were obtained from 24 breasts with CIS; 20 samples showed invasive breast cancer. Among the 29 satisfactory DL samples, 10 (34%) showed mild atypia, 4 (14%) showed marked atypia, 15 (52%) were benign, and 0 (0%) showed cancer cells.
While interobserver agreement was fair (kappa value = 0.52), the authors concluded that DL lacks sufficient sensitivity for the diagnosis of breast cancer. (Brogi et al., 2003) In a pilot study, Hartman et al. evaluated the efficacy of DL and magnetic resonance imaging (MRI) versus mammography and clinical breast exam (CBE) for breast cancer detection in women at high risk for the disease who were BRCA mutation carriers or who had a 10% risk of developing breast cancer within 10 years according to the Claus model. DL detected atypia in specimens from 7 (23%) patients including a high-grade atypia in 1 patient with a normal mammogram and normal MRI results. Six other patients who had atypia on DL had normal mammographic results. The data suggest that DL might detect lesions that are otherwise missed; however, longer-term follow-up is needed to determine if the detection of cellular atypia on DL accurately predicts the risk of breast cancer and affects patient outcomes. (Hartman et al., 2004) Francescatti et al. evaluated the results of attempted ductal lavage in 120 patients at high-risk for breast cancer. Thirtytwo patients were excluded because 29 patients did not produce nipple aspirate fluid and the surgeon was unable to cannulate the effluent-producing duct in 3 patients. Of the remaining 88 patients, 15 (17%) had insufficient epithelial content for diagnosis, 51 (58%) had benign cytologic results, and 22 (25%) had abnormal cells. Of the 25%, 20 patients had mild atypia, 1 had marked atypia and 1 had malignant changes. (Francescatti et al., 2005) Kahn et al. studied the association between ductal lavage cytologic findings and histologic findings in women with known breast cancer. Ductal lavage was performed on 44 breasts in 32 women with known cancer and on 8 breasts in 7 women undergoing prophylactic mastectomy, two with occult malignancy. In 39 ducts with complete cytologic and histologic data and when marked atypia or malignant cells defined a positive cytologic test, sensitivity of ductal lavage was 43%, specificity was 96%, and accuracy was 77%. When mild or marked atypia or malignant cells defined a positive cytologic test, sensitivity was 79%, specificity was 64%, and accuracy was 69%. Analysis of all 31 cytologically evaluable breasts showed sensitivity was 17%, specificity was 100%, and accuracy was 19%. The investigators concluded that ductal lavage appears to have low sensitivity and high specificity for cancer detection. (Kahn et al., 2004) Specimens obtained by DL might be suitable for evaluation by techniques such as fluorescent in situ hybridization (FISH) or cytogenetic analysis. Preliminary studies have demonstrated the feasibility of analyzing ductal epithelial cells for chromosomal abnormalities, which could potentially assist in the definitive diagnosis of breast cancer. However, these diagnostic techniques are in the preliminary stages of development and it remains unclear how they would impact the Cytological Examination of Breast Fluids for Cancer Screening- Commercial Medical Management Guideline diagnostic accuracy of DL or its role in risk stratification. (Yamamoto et al., 2003, Evron et al., 2001, King et al., 2003) Overall, the published evidence regarding DL suggests that the procedure is feasible and well tolerated, is less invasive and yields a greater number of breast duct epithelial cells for cytological analysis than FNA, and is capable of detecting abnormal cells in NAF specimens from individual breast ducts. However, there are minimal data on the diagnostic sensitivity or specificity of DL, on the clinical significance of the presence of atypical cells in DL specimens, on whether the demonstration of no atypical or malignant cells correlates with a decreased risk for breast cancer, or on the use of either positive or negative test results in patient management or outcomes. Furthermore, there is speculation that breast cancer is not a field defect throughout the breast but can occur in one ductal system, thus if that ductal system is not lavaged, a false negative would be yielded. (Lindsay et al., 2004)
Most of the published evidence on FDS is limited to preliminary cross-sectional studies evaluating the technical success of intraductal visualization and the diagnostic accuracy of the technique or the feasibility of intraoperative breast endoscopy.
Shen et al. studied the role of FDS in 415 women with abnormal nipple discharge. FDS identified an intraductal papilloma (IDP) in 166 patients (40%) including 10 with atypical papillomas and 156 with typical papillomas. DCIS was confirmed by histopathological examination in 11 patients with IDPs; 6 (55%) of these patients had normal findings on mammography and CBE. The sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) for FDS were 73%, 99%, 80%, and 98%, respectively. For FDS and DL together, the corresponding figures were 64%, 100%, 100% and 97%. The results suggest that FDS can diagnose precancerous lesions of the breast that are not detected by conventional means. It was unclear how or whether patients with normal findings by FDS were followed up to confirm the absence of disease (to confirm the specificity and NPV values), or how the test results impacted clinical decision-making. (Shen et al., 2001) In a study of 65 patients with abnormal nipple discharge, FDS identified intraductal abnormalities in 38 patients; the results of histopathological examination were positive in 37 of 38 (97.4%). The PPV of FDS was 97.4% versus a PPV of 89.2% for ductography, a statistically significant difference. The sensitivity, specificity, and accuracy of DL were 50%, 94.3%, and 89.7%, respectively. The authors concluded that FDS had good PPV for detection of intraductal lesions;
however, the sensitivity of DL was low for the diagnosis of breast cancer in this population. (Yamamoto et al., 2001) A small clinical study evaluated the efficacy of intraoperative FDS for the management of symptomatic bloody nipple discharge in 27 patients with normal mammographic findings within 3 months before surgery. Breast endoscopy was technically successful in 26 patients, and allowed for visualization of all of the proximal ducts. A lesion associated with the bleeding was identified in 26 of the 26 (100%) patients who were cannulated, with 19 (70%) having multiple intraluminal lesions. In 2 patients, DCIS was located 2 to 4 cm deeper than large retroareolar papillomas within the duct.
Breast cancer was detected in 2 (7.4%) patients, and in both there was a more proximal papilloma within the same ductal system. One breast cancer found by FDS was not visible on a mammogram or ultrasound image. Atypical ductal hyperplasia (ADH) was the cause of bleeding in 33% of the patients with benign lesions. Since FDS identified a relatively high incidence of multiple lesions that might otherwise have been missed, the authors concluded that standard surgery to treat patients with bloody nipple discharge, i.e. blind resection of the first 2 to 3 cm of the duct, might miss DCIS or ADH located deeper within the breast duct that are the true bleeding source. Further research is needed to determine whether there are situations in which FDS-guided resection would serve as a useful adjunct to or if it could replace open surgical excision. (Dooley et al., 2002) In a small preliminary study of 49 patients, FDS identified intraductal lesions that were not observed on ductography or mammography in 8 (16%) of the 49 patients with nipple discharge or suspected breast cancer (the reasons for suspicion of breast cancer were not provided). Intraductal papillomas in 2 patients and DCIS in 6 patients that were detected by FDS and DL were confirmed by histopathology, as the 8 patients proceeded to have surgery. The authors noted that the Cytological Examination of Breast Fluids for Cancer Screening- Commercial Medical Management Guideline test results aided in surgical planning but added that additional studies are needed to determine the optimal role of FDS and DL in treatment decision-making. (Hunerbein et al., 2003) Some researchers have challenged the utility of FDS and DL for the detection and diagnosis of breast cancer. In a retrospective analysis of the presence and type of involvement of the nipple and central duct area in 801 mastectomy specimens performed for invasive breast cancer, DCIS, or both, 17% of the invasive cancers had no demonstrable intraductal component defined as atypical proliferation or atypical cells. Furthermore, only 22% of cases showed nipple and central duct involvement. These findings lead to questions regarding the adequacy of these methods for breast cancer detection since their accuracy depends upon the presence and accessibility of precursor lesions such as ADH or intraductal carcinomas. Since FDS and DL examine only 1 or 2 ducts among a total of 15 to 20 breast ducts that open at the nipple, these techniques might also miss focal abnormalities or those occurring in ducts that are not examined.
(Badve, 2004, Badve et al., 2003) Evidence from the available studies suggests that FDS is feasible, that it yields useful clinical information on intraductal disease and that it allows for the performance and cytological examination of DL washings for detection of cellular atypia. However, there is minimal evidence on how this technique would be used in the management of the patient, eg.
Determining the need for other diagnostic tests, such as mammography or ductography, assessing the need for surgical biopsy or excision, or planning the extent of surgical excision.
Limitations of the available studies of DL and FDS include histopathological examination only for selected cases, i.e.