Positive predictive value of the breast FNAB diagnoses of epithelial hyperplasia with atypia, papilloma, and radial scar

Indeterminate and erroneous fine-needle aspirates of breast with focus on the 'true gray zone': a review

OBJECTIVE

To review our experience and the literature on inconclusive/erroneous fine-needle aspirates (FNAs) of breast with the focus on the 'true gray zone'. To describe the cytology, differential diagnosis, pitfalls and limitations of common and rare lesions.

STUDY DESIGN

We conducted a literature search focusing on breast FNAs with statistical data of C3 and C4 categories including false-positive and false-negative cases. Similar data from 2003 to 2009 was obtained from our institution.

RESULTS

C3 and C4 categories account for 3-17% of breast FNAs. Contributing factors are technical difficulties, inexperienced pathologists interpreting FNAs of breast and overlap of cytologic features of certain benign and malignant conditions; this last, 'true gray zone' accounts for 2% of cases. Fibroadenoma, proliferative breast lesions, gynecomastia, infiltrating and in situ low-grade adenocarcinomas and tubular, cribriform, lobular and mucinous carcinomas are the most common problematic lesions. Granular cell tumor, adenomyoepithelioma, pregnancy-related lesions, fat necrosis, inflammatory and radiation changes, adenoid cystic carcinoma, spindle-cell lesions and Phyllodes tumor are less common.

CONCLUSION

Inconclusive/erroneous FNAs of breast due to the 'true gray zone' are rare. Most are due to the overlapping cytologic features of some benign and malignant conditions. Practical features that may help arrive at the correct diagnoses are elucidated.

The fine needle aspiration biopsy diagnostic criteria of proliferative breast lesions: A retrospective statistical analysis of criteria for papillomas and radial scar lesions

This study aimed to analyze statistically the accuracy of fine needle aspiration biopsy cytological criteria in diagnosing epithelial hyperplasia with atypia (EHA), papillary lesions (PAP), and radial scar/complex sclerosing lesions (RS/CSL). The 42 criteria studied were extracted from a literature review and those developed and used in our department. Cytological cases with diagnoses of EHA, PAP, and RS/CSL for the period of 1997-2001 were correlated with relevant histological follow-up and the positive predictive power of these diagnoses have been presented in an earlier publication. Some 77 cases with definite, specific histological diagnoses of atypia or a more severe lesion, PAP and RS/CSL, were reviewed and scored using 55 cytological criteria. As the group of EHA and RS/CSL yielded a smaller number of cases, these were grouped together in the statistical analysis and compared to PAP. The cytological features, which were most diagnostic for PAP, were stellate (Odds ratio 1.75) and meshwork (Odds ratio 3.29) tissue fragments, while the presence of tubular structures was inversely proportional to the histological outcome of PAP. True papillary fragments were uncommon and not statistically significant in diagnosing PAP of the breast.

A prospective study of the diagnostic accuracy of cytological criteria in the FNAB diagnosis of breast papillomas

Seventy-four fine needle aspiration biopsies (FNAB) of breast were selected from the 133 cases with surgical biopsy follow up, from a total of 1,154 consecutive breast FNAB received in a 6-month-period. These 74 cases were reviewed and scored using all relevant cytological criteria for proliferative breast lesions used in our recent retrospective study, without reference to the original cytological and surgical biopsy diagnoses. Of the 42 criteria scored, 13 had a statistically significant association between the cytology score and the presence or absence of a papilloma (PAP), and the sensitivities, specificities and positive predictive values (PPV) of these individual criteria, or a combination of criteria, were derived. It was found that stellate and meshwork tissue fragments and papillary fragments were all highly specific (0.98) for the diagnosis of PAP, with meshwork fragments having the highest PPV (0.93). Stellate fragments (0.41) and papillary fragments (0.24) were less sensitive not only because they occurred less often, but also because they were present in smaller numbers. The presence of a proteinaceous background with macrophages and siderophages or a moderate to marked number of apocrine sheets were useful indicators only when coexisting with at least one of the above three features.