C. J. Granger2, E. L. Ryon3, A. R. Hogan1, H. L. Neville1, C. M. Thorson1, E. A. Perez1, J. E. Sola1, A. Brady1 1University Of Miami,Division Of Pediatric And Adolescent Surgery/Department Of Surgery/Miller School Of Meidcine,Miami, FL, USA 2University of Miami,Leonard M. Miller School Of Medicine,Miami, FL, USA 3University of Miami,Division Of Surgical Oncology/Department Of Surgery/Miller School Of Medicine,Miami, FL, USA
Introduction: Ultrasound (US) imaging is an adjunct to clinical exam (CE) in the assessment of pediatric breast lesions. We sought to investigate the accuracy of CE and US in determining maximum diameter (Ø) of breast lesions versus final pathology (P).
Methods: A single institutional retrospective analysis of patients < 25 years of age who underwent breast mass resection (CPT 19120, 19301) from Feb 2011 to Sept 2015 was performed. Data was collected and analyzed using SPSS.
Results: 67 patients underwent breast resection with a mean age of 16 ± 2 years. The mean PØ (MPØ) for all lesions was 4.2 ± 2.7 cm. Lesions encountered were fibroadenoma (88%, MPØ 3.8 ± 1.8 cm), juvenile fibroadenoma (7.5%, MPØ 8.1 ± 6.8 cm), low-grade phyllodes sarcoma (3%, MPØ 7.5 ± 3.5 cm), and fibrous hamartoma of infancy (1%, the only male patient). 51% of lesions were right sided, 37% were left sided, and 12% were bilateral. 34% of patients had no surgical indication documented. Documented indications included: 28=increasing size, 11=mastodynia, 8=large size, 1=malignant US findings, and 1=recommended by primary physician. 13% of patients developed recurrent and/or new lesions. The complications from surgery included: 3=mastodynia, 2=seromas, 3=local skin reactions, and 1=numbness. Of the two patients with low-grade phyllodes sarcoma, one was lost to follow-up and the other underwent re-excision for a positive margin.
50 patients had documented CE measurements while 48 patients underwent US imaging. Of these, 28 patients had both CE and US measurements and were included in the sub-group analysis. Both groups were normally distributed by Shapiro-Wilk’s test (p=0.107 for CEØ; p=0.373 for USØ). Paired t-test comparing PØ to CEØ found the groups to be the same with an underestimation on CE of 0.4 ± 1.2 cm, p=0.87. When comparing PØ to USØ there was a statistically significant underestimation on US of 0.6 ± 1.2 cm, p=0.01. The difference between CEØ and USØ was not statistically significant (0.2 ± 0.8 cm, p=0.227).
Conclusion: In this single institutional retrospective study, CE estimates were equivalent to final P, while US significantly underestimated the size of breast lesions. The underestimation in each case was between 0.4 cm for CE and 0.6 cm for US which is likely not clinically significant and possibly biased by time to surgery. Nevertheless, given the accuracy of CE, the utility of US in measurement of pediatric breast lesions is limited and should be individualized.