Is duct excision still necessary for all cases of suspicious nipple discharge?

Michael S. Sabel*, Mark A. Helvie, Tara Breslin, Alida Curry, Kathleen M. Diehl, Vincent M. Cimmino, Alfred E. Chang, Lisa A. Newman

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

40 Scopus citations


Despite the low likelihood of malignancy, it is recommended that all women with pathologic nipple discharge undergo duct excision based on the inadequate sensitivity of diagnostic modalities. However, these data originates prior to recent improvements in breast imaging. We performed a retrospective review of patients evaluated in the setting of modern diagnostic breast imaging. Of 175 women referred to our breast clinic with a primary complaint of nipple discharge, 142 (81%) had suspicious discharge. Of the 23 patients who opted for observation over duct excision, with a mean follow-up of 3.3 years, none have been diagnosed with cancer. Among patients who proceeded with surgery, cancer was diagnosed in seven patients (5%). Six of the seven patients had either an abnormal mammogram or ultrasound. Among 46 patients with suspicious nipple discharge, a normal physical exam and normal diagnostic mammogram/ultrasound, only one malignancy (2%) was identified in a 79-year-old patient with a personal history of breast cancer. In selected patients with suspicious nipple discharge, but normal physical exam and diagnostic imaging, short-term observation with repeat evaluation seems reasonable for patients who do not desire duct excision.

Original languageEnglish (US)
Pages (from-to)157-162
Number of pages6
JournalBreast Journal
Issue number2
StatePublished - Mar 2012
Externally publishedYes


  • breast neoplasm
  • galactography
  • mammography
  • nipple discharge
  • ultrasonography

ASJC Scopus subject areas

  • Internal Medicine
  • Surgery
  • Oncology


Dive into the research topics of 'Is duct excision still necessary for all cases of suspicious nipple discharge?'. Together they form a unique fingerprint.

Cite this