To the best of our knowledge, there have not been published studies specifically examining the accuracy of CNB for breast non-malignant papillary lesions in specific ethnic groups. However, clinical and biological differences are known to exist between the ethnic groups. For example, African-American women with breast cancer are reported to be more likely to have advanced disease at diagnosis, estrogen receptor negative, a high S-phase fraction, higher risk of recurrence and poorer prognosis than Caucasian women . Insulin-like growth factor II receptor expression is higher in breast tumor of African-American women than those of any other ethnic group . Compared to Caucasian women, African American women carry more risk factors for breast cancer: younger age at first live birth, less and shorter breast feeding, higher BMI, less moderate physical activity . While alcohol consumption increase breast cancer risk in Caucasian women, recent alcohol consumption apparently does not affect breast cancer risk in African American women, while early age drinking in this ethnic group seems to decrease the risk . Association between 5p12 genomic markers and breast cancer susceptibility was identified in Caucasians and East-Asians, but not in Africans or African-Americans . Results in our study show a large percentage (25%) of benign or atypical papillary lesions diagnosed on CNB will be upgraded in the final excisional examination in a primarily African-American patient population. Our upgraded rate is slightly higher than the median percent of underestimation (16.6%) based on a meta-analysis of 34 studies from mixed ethnic groups . While this is an interesting trend, the small patient number of our study warrants further investigation of this question.
Pathological classification of breast papillary lesions on CNB, especially on H & E preparation alone, can be challenging. Classical intraductal papilloma shows obvious myoepithelial cells covering fibrovascular cores, a polymorphic epithelial cells and normochromatic nuclei; on the other hand, papillary DCIS has no or scant myoepithelial cells, its neoplastic epithelial cells are monomorphic with rigid architecture, and hyperchrmomatic nuclei . Atypical papillary lesion is defined as papilloma with focal population of monotonous cells with the cytological and architectural features of low grade ductal neoplasia, scant or absent myoepithelial cells in these foci, and the atypical epithelial cells usually show lack of staining for high-molecular weight keratin and uniform positivity for estrogen receptor. Over the years, many investigators have suggested immunohistochemical stains on CNB enhance diagnostic accuracy [14–17]. In this sense, it is interesting to note that in a meta-analysis of published studies, a decreasing trend of underestimation was obvious over time between CNB and surgical excision . Furthermore, a recent study from Toronto, Ontario points out that correlation between CNB and excision diagnoses for breast papillary lesions is significantly greater for pathologists specialized in breast than for other pathologists . As part of current study, we performed and re-evaluated immunostains (CK5/6 and CK903) on all seven CNBs that were upgraded to intraductal papillary carcinoma on subsequent specimens. While the discrepancy on some cases was due to sampling, the diagnosis of three core biopsies should be upgraded to malignant according to current diagnostic criteria using immunohistochemistry. Immunostains, as well as other molecular investigative procedures , can be very helpful in borderline cases. Another potential source of discrepancy between CNB and excision comes from the heterogeneous nature of the papillary lesion even within a single lesion. Different approaches have been proposed to evaluate lesions of great intrinsic variability, including extended biopsy , powered biopsy system , and assessment aided by automated intelligent system [22, 23]. Application and evaluation of theses novel approaches in breast CNB may help to more accurately classify papillary lesions. When and if possible, more advanced continued education of breast pathology should be encouraged for all practicing pathologists.
Several clinical and radiological variables, including patient’s age, lesion size, distance from nipple, BI-RADS category, have been examined in relations to their predictive value of pathological upgrading from CNB to excisional biopsy. Great efforts have been made, but conclusions from different reports vary significantly [24–27]. In our cohort, the average age of these upgraded cases was 65 years, compared to an average of 48 years among those patients without upgrading. We also evaluated the relationship between upgrading and lesion size, distance from nipple, microcalcification, and BI-RADS category, none of which achieved statistical significance. Hence, in our study of primarily African American population in an urban setting, older age appeared to be an important factor in predicting diagnostic upgrade to a more severe lesion.