Ductal Lavage: Sensitivity and Specificity in Women with Proven Breast Cancer

Abstract & Commentary

Synopsis: The technique of ductal lavage was examined in women who immediately thereafter underwent mastectomy for breast cancer. A physiologic solution was injected into cannulated breast ducts and then aspirated and examined for atypical or frankly malignant cells. Although the presence of such cells had demonstrably high specificity, the sensitivity was disappointing (only 5 of 38 [13%] of cancerous breasts had marked atypia or frankly malignant cells). Thus, ductal lavage remains an interesting research parameter but is likely to not be useful in cancer screening.

Source: Khan SA, et al. J Natl Cancer Inst. 2004;96: 1510-1517.

Ductal lavage, a technique which involves the cannulation of breast ducts and the infusion and subsequent aspiration of a physiological solution for cytological examination has theoretical appeal as a screening tool for women at high risk of breast cancer. Included would be those women with more dense breasts, as these have proven more difficult to examine by routine mammography, magnetic resonance imaging (MRI), and digital mammography.1,2 To assess the use of ductal lavage as a cancer diagnostic test, Khan and associates at Northwestern University, the University of California at San Francisco and the Fox Chase Cancer Center investigated the association between ductal lavage cytologic findings and histological findings in women with known breast cancer undergoing mastectomy.

They performed ductal lavage in the operating room prior to the removal of 44 breasts from 32 women with known cancer and on eight breasts from seven women undergoing prophylactic mastectomy. When possible, they also injected ducts with dye for analysis on subsequent mastectomy specimens. Associations between cytologic and histologic results were compared.

Although ductal lavage was not always possible, at least one duct was lavaged in 36 breasts (mean of 1.4 ducts per breast). Markedly atypical or malignant cytology was found in five cancer containing breasts. In 39 ducts with complete cytologic and histologic data, ductal fluid cytology was shown to have a sensitivity of 43% (95% confidence interval [CI] = 23%-72%), specificity was 96% (95% CI = 86%-100%) and accuracy was 77% (95% CI = 63%-89%). When including mild atypia as well as marked atypia or the presence of frankly malignant cells, sensitivity was calculated at 79% (95% CI = 57%-96%), specificity was 64% (95% CI = 46%-83%) and accuracy was 69% (95% CI = 55%-83%). When all 31 cytologically evaluable breasts were analyzed, sensitivity was 17% (95% CI = 7%-35%), specificity was 100% (95% CI = 5%-100%) and accuracy was 19% (95% CI = 9%-38%). Khan et al concluded that for breasts with proven cancer, ductal lavage appears to have low sensitivity and high specificity for cancer detection.

Comment by William B. Ershler, MD

These findings must be considered discouraging for those who had high hopes for ductal lavage as an answer for the screening dilemma observed in high risk patients with more dense breasts for whom mammography and other imaging techniques have proven unsatisfactory. Of the 38 cancerous breasts, only 5 (13%) had marked atypia or malignant cells present in the ductal effluent. And, if including those with mild atypia, the number of positives was 16 (42%). Thus, as pointed out by Fabian, Kimler and Mayo in the accompanying editorial,3 by either definition, the sensitivity (13% to 42%) is too low for the technique to be considered as a screening procedure for the majority of women at risk of developing breast cancer. In fact, the sensitivity may well be lower than mammography or MRI in the same population. Nevertheless, as an adjunct to these imaging studies there may ultimately be proven value to ductal lavage.

The technique, however, may well have value as a marker of biological importance in some women with high risk for breast cancer. Current cancer prevention trials have incorporated nipple aspiration fluid cytology as an outcome of interest, and ductal lavage may be similarly incorporated in such research schemes. However, for the time being, ductal lavage cytology can not be considered a sensitive screening tool and its use remains to be established in the clinical setting.


1. Masood S, et al. Mod Pathol. 2002;16:40A.

2. Fabian CJ, et al. J Natl Cancer Inst. 2000;92: 1217-1227.

3. Fabian CJ, et al. J Natl Cancer Inst. 2004;96: 1488-1489.

William B. Ershler, MD, INOVA Fairfax Hospital Cancer Center, Fairfax, VA; Director, Institute for Advanced Studies in Aging, Washington, DC, is Editor for Clinical Oncology Alert.