Twenty-Year Outcome Following Central Duct Resection for Bloody Nipple Discharge

Abstract

Background:

Patients found with pathologic nipple discharge present a diagnostic dilemma to surgeons. No one diagnostic test, including cytology or radiologic imaging, has proved superior to any other in the differentiation of benign versus malignant sources of pathologic nipple discharge. Ductoscopy has been introduced as a way to assist with identification of potential sources of pathologic nipple discharge. Ductoscopy is also useful in the resection of deep or peripheral masses that may be missed with standard blind resection. This report evaluates the risk of missed malignancy following central duct resection (CDR).

Methods:

Records of 56 patients who underwent CDR for pathologic nipple discharge greater than 15 years previous were reviewed. Data including type of nipple discharge, future biopsy, and pathology reports were all examined.

Results:

Of the 56 patients, 36 had bloody, 18 serous, and 2 green initial discharges. Fifty-seven percent were found to have intraductal papilloma as the source of discharge, with fibrocystic disease and ductal ectasia providing the next most common causes. One woman was found to have ductal cancer and one lobular carcinoma in situ at time of CDR. Patients were followed for a mean period of 22 years. Fourteen women required future biopsy. Of these, 10 had benign disease and 4 had cancer, 3 on the ipsilateral side. The 3 women with breast cancer were 9, 13, and 17 years from initial resection.

Conclusions:

CDR for pathologic nipple discharge is an effective way to diagnose and treat pathologic nipple discharge without missing underlying cancers.

Ductoscopy is promoted over central duct resection (CDR) for discovery of peripheral or deep potentially malignant sources of pathologic nipple discharge. In a retrospective review of 56 patients who had undergone CDR for nipple discharge and who had > 15-year follow-up, CDR was determined to be safe, diagnostic, and therapeutically effective.

Pathologic nipple discharge is second only to breast masses as a cause for women presenting with breast complaints. Significant features of nipple discharge requiring surgical intervention include 4 types of discharge: clear, yellow, serosanguineous, or sanguineous. Four other characteristics of discharge have also been associated with malignancy: persistent, spontaneous, unilateral, and expressible from a mammary duct.¹

Cancer risk from nipple discharge is reported to be anywhere from 1.3% to 47%.^1–4 Patients with masses or abnormalities on mammography or sonography associated with their discharge are at higher risk for associated cancer.^2–5 The most common etiologies for their discharge are papilloma, fibrocystic disease, cancer, and ductal ectasia.¹ While percentages vary depending on the study, papilloma accounts for >50% of all causes of discharge.

Evaluation of patients with pathologic discharge includes history, physical examination, mammography, sonography, cytology, ductography, and most recently ductoscopy.^5–9 While the goal of all these diagnostic tests is to attempt to differentiate benign or malignant causes, none has been consistently predictive. Central duct resection (CDR) is offered to most women for both diagnosis and treatment of pathologic discharge.

With recent developments in technology, ductoscopy provides a means of visual examination within the ducts to evaluate the nipple discharge source. Although not useful in patients undergoing partial mastectomy to determine resection margins,¹⁰ ductoscopy assists with identifying the source of discharge and limits the area of resection. It is also reputed to be useful in identifying peripheral or deep masses that may otherwise be missed with a blind resection.^6,7

The purpose of this study was to evaluate a patient population that had undergone CDR for pathologic nipple discharge. By analyzing a group of patients that had undergone resection more than 15 years ago, we hoped to demonstrate whether or not blind CDR missed cancer.

METHODS

A retrospective chart review was performed of women with pathologic nipple discharge who underwent central duct excision at St. Joseph's Hospital in Marshfield, Wisconsin, between the years 1975 and 1985. Data collected included age, side of discharge, type of discharge, and pathologic findings after resection. Charts were then reviewed to determine if, in the subsequent years, patients developed new breast complaints that required surgical intervention. Any future biopsy or surgery was recorded along with the accompanying pathology.

Operative Technique

CDR was performed utilizing an infra-areolar incision with rising of the areola flap. All ducts were identified and removed from the dermal aspect, and resection was then carried down to the end of visible ductal tissue. All ducts were removed. The incision was then closed.

RESULTS

Sixty-eight patients who had undergone CDR for pathologic nipple discharge were identified. Twelve patients were lost to long-term follow-up. Their mean follow-up period was 7.6 years (range, 1–13 years). None of them had developed cancer up to the point of follow-up. Attempts to locate these individuals were unsuccessful.

Fifty-six patients were followed for more than 15 years. The mean duration of follow-up was 21.4 years (range, 15–31 years). The mean age was 50.2 years (range, 24–82 years). Eight of the 56 were deceased but included, as their follow-up was more than 15 years after resection. None had died of breast disease.

Of the 56 women, 36 (64.2%) had bloody discharge, 18 (32.2%) serous, and 2 (3.6%) had yellow/greenish discharge. Intraductal papilloma was identified in 32, fibrocystic disease in 7, ductal ectasia in 7, hyperplasia in 4, chronic inflammation in 2, epidermal cyst in 1, and fibroadenoma in 1 of the patients. Cancer was identified in 2 patients at the time of CDR, with one being a ductal carcinoma and the other lobular carcinoma.

Fifty-two (92.8%) patients underwent preoperative mammography or xerography prior to CDR. Forty-four of these had normal findings or fibrocystic changes on their radiologic studies. Eight (15.3%) patients were noted to have either masses or calcifications. Five had infra-areola masses that were intraductal papilloma, 2 had calcifications found to be ductal ectasia, and 1 had a mass elsewhere in the breast, which was fibrocystic disease. None of these was found to have cancer at the time of initial resection.

Regular follow-up, including mammography and physical examination, was performed in 48 (85.7%) patients. Four patients had no follow-up mammography but were seen by a primary care physician regularly, 2 had no mammography due to their age, and 2 patients had only one mammogram recorded in the follow-up period.

Of the 56 patients, 14 (25%) required future biopsy for suspicious lesions. Ten of the 14 biopsies proved to be benign, and no future intervention has been required. The benign diseases included fibrocystic disease, ductal hyperplasia, apocrine metaplasia, chronic inflammation, and fibroadenoma.

Four patients were found to have malignancy during the course of follow-up. Of the 4 patients who developed cancer, 3 (5.3%) occurred on the same side as the initial CDR and 1 on the contralateral side. Of the 3 patients with breast cancer on the ipsilateral side, the first was found to have lobular carcinoma 17 years after initial CDR, the second had invasive intraductal cancer found 9 years after CDR, and the third demonstrated ductal carcinoma in situ 13 years from resection. Of these patients with ipsilateral disease, 1 had calcifications and the other had an infra-areolar mass on mammography prior to CDR. The patient with contralateral disease developed ductal cancer at 3 years post-CDR.

DISCUSSION

Benign and malignant causes of pathologic nipple discharge continue to be difficult to distinguish. While the incidence of cancer is low, especially in today's setting of early screening, the risk of malignancy cannot be overlooked. Various methods have been developed to assist in determining the cause of the discharge, but none has been shown sensitive enough to dispense with the others. Mammography, physical examination, and sonography are all vital in the evaluation of such patients. Ductoscopy is now being actively pursued to better qualify the origins of nipple discharge. Ductoscopy has been shown to be safe and easy to use in patients with obstructed dilated ducts.³ It has been suggested that standard CDRs may miss peripheral lesions that are better visualized with ductoscopy.³ Ductoscopy has been reported to find proliferative abnormalities in up to 88% of their specimens, whereas central duct excision has only found 23% to 77% abnormalities.³

On evaluation of the data from a single institution over a 10-year period, 100% of our population had an identified abnormality on pathology. Follow-up for these patients was more than 15 years and mean follow-up was 22 years. Only 3 (5%) of the patients developed breast cancer: 1 lobular, 1 invasive ductal, and 1 ductal carcinoma in situ on the same side of resection. Those patients who developed cancers were diagnosed at 17, 9, and 13 years post-CDR, respectively. We presume that, if malignancy or premalignancy had been missed at the time of CDR, it would have been identified within the 15-year timeframe.

Preoperative evaluation with mammography revealed only 8 lesions. Of these, 5 were infra-areolar masses and 2 were calcifications. Two women who later developed ipsilateral breast cancer at 9 and 13 years were in this group. Preoperative mammography or xerography findings did not seem to help predict those who would develop cancer. Ultrasound, not used frequently in this time period, may have contributed more useful information in this group.

CDR for pathologic nipple discharge provided both diagnostic and therapeutic results for our patients. Of the 3 cancers that did develop, the earliest was at 9 years post-CDR. It is our opinion that central duct excision should still be strongly considered in patients with pathologic nipple discharge. It is safe, effective, diagnostic, and therapeutic.

ACKNOWLEDGMENTS

The authors thank the contributing surgeons of this project: Drs. Gail H. Williams, Jerry M. Hardacre, and Mark K. Swanson. The authors also thank Marshfield Clinic Research Foundation for its support through the assistance of Linda Weis and Alice Stargardt in the preparation of the manuscript.