Breast cancer tumor size assessment with mammography, ultrasonography, and magnetic resonance imaging at a community based multidisciplinary breast center

Sarah Ramirez, Max Scholle, Jennifer Buckmaster, Robert Hunter Paley, Gopal Chandru Kowdley

Research output: Contribution to journalArticlepeer-review

26 Scopus citations

Abstract

Paramount to staging and patient management is accurately measuring the size of invasive breast cancers. We assessed the accuracy of mammography (MG), ultrasonography (US), and magnetic resonance imaging (MRI) at our community-based hospital in which multiple radiologists and imaging machines are used in the care of our patients. We performed a retrospective analysis of a prospectively maintained database of 277 patients seen at our breast center from 2009 to 2010.We tabulated MG, US, and MRI-reported tumor sizes in 161 women with pathology-proven invasive breast cancer and compared the preoperative size measurements with final pathologic tumor size. In the 161 patients, 169 lesions were identified. Imaging using all three modalities was available in 47 patients. When compared with final pathology, MRI had a correlation of r = 0.75 to mean tumor size as compared with US (r = 0.67) and MG (r = 0.76). Mean tumor size was 1.90 cm by MG, 1.87 cm by US, 2.40 cm by MRI, and 2.19 cm by pathology. We were able to achieve an excellent correlation of pathologic tumor size to preoperative imaging. The absolute differences in size between the modalities were small. MRI, in select patients, added to the assessment of tumor size based on US and MG.

Original languageEnglish (US)
Pages (from-to)440-446
Number of pages7
JournalAmerican Surgeon
Volume78
Issue number4
StatePublished - Apr 2012

All Science Journal Classification (ASJC) codes

  • Surgery

Fingerprint Dive into the research topics of 'Breast cancer tumor size assessment with mammography, ultrasonography, and magnetic resonance imaging at a community based multidisciplinary breast center'. Together they form a unique fingerprint.

Cite this