Background: Tumor size is an important parameter in breast cancer staging, and may affect clinical decision-making. Definite tumor size is determined by measurement of the pathological specimen after surgery, and prior to surgery the size is assessed by imaging: mammography, ultrasound and MRI. Discrepancy between imaging assessed and pathological size is not infrequent. We suspected that breast density may affect image based tumor size assessment.
Purpose: To compare tumor size as measuresd by imaging modalities to the pathological size as determined on the specimen, and examine whether increased breast density affects measurement accuracy.
Material & methods: This was a retrospective study of all patients diagnosed and operated for primary breast cancer at our institution during the years 2015-2016. We determined maximal tumor size by each imaging modality and compared it to the tumor size on pathological report, as well as breast density as determined by mammography.
Results: After excluding patients who underwent neoadjuvant therapy, patients who were not operated, and patients in whom data were missing, a total of 183 patients with 198 tumors were included in the study. Mean age was 66, 89% (176) were over the age of 50, and 11% (22) were younger than 50. In patients over 50, 56% (99) had fatty breasts (density category A+B) and 43.8% (77) had dense breast (category B+C), in patients under 50, 22.7% (5) had fatty breasts (category A+B) and 77.3% (17) had dense breast tissue (category C+D). Comparison of mean tumor size in each group (pathology, mammography, US and MRI) revealed that US underestimated the tumor size in all density groups compared to pathology, and mammography was less accurate in very dense breasts, and overestimated the tumor size. MRI overestimated tumor size only in breast density category B, and was accurate in the remainder.
Conclusion: Breast density is an important factor that affects tumor size assessment by imaging, with US underestimating the size and mammography overestimating size in very dense breasts. These inaccurate estimations may affect treatment planning.