Solid papillary ductal carcinoma in situ versus usual ductal hyperplasia in the breast: a potentially difficult distinction resolved by cytokeratin 5/6

Summary

The solid papillary variant of ductal carcinoma in situ is an uncommon entity, which usually presents in the seventh or eighth decade and may be associated with invasive mucinous carcinoma. Solid papillary ductal carcinoma in situ (SP-DCIS) shares many morphological features with usual ductal hyperplasia (UDH) involving a papilloma: papillary architecture, solid growth, cellular streaming, and low-grade nuclear features. These similarities can make the distinction between these 2 entities challenging. Recent studies have demonstrated that immunohistochemical staining for cytokeratin 5/6 can distinguish UDH from conventional forms of ductal carcinoma in situ. Most of the epithelial cells of UDH express cytokeratin 5/6, but the tumor cells of ductal carcinoma in situ do not. We tested the hypothesis that the results of staining for cytokeratin 5/6 can distinguish UDH from the solid papillary variant of ductal carcinoma in situ. Immunohistochemical staining of 14 cases of SP-DCIS and 9 cases of UDH (4 involving papillomas) was performed using cytokeratin 5/6 antibody clone D5/16 B4. Strong cytoplasmic or membrane staining was considered positive. The hyperplastic cells in all cases of UDH showed strong staining for cytokeratin 5/6. The percentage of positive cells ranged from 50% to 80%. None of the SP-DCIS tumor cells stained for cytokeratin 5/6; however, many cases did show staining of occasional entrapped, benign epithelial, and myoepithelial cells. We conclude that the absence of strong cytokeratin 5/6 expression by SP-DCIS distinguishes it from its morphological mimic, UDH. Pathologists must guard against misinterpreting SP-DCIS as UDH in those cases in which the carcinoma cells engulf cytokeratin 5/6-expressing residual, native epithelial cells.

Introduction

Solid papillary carcinoma is an unusual variant of low-grade ductal carcinoma in situ, which usually arises in the seventh or eighth decade and may be associated with invasive mucinous carcinoma [1], [2]. This type of carcinoma in situ shares many morphological characteristics with papillomas complicated by usual ductal hyperplasia (UDH). Both lesions have a papillary architecture and a skeleton of dense collagenous stroma; both demonstrate a solid or fenestrated proliferation of epithelial cells, which often display a streaming pattern, and both consist of cells with irregular nuclei, granular chromatin, and small nucleoli. These morphological similarities create well-documented problems distinguishing the 2 entities. In one of the earliest studies of solid papillary ductal carcinoma in situ (SP-DCIS), Azzopardi [3] reported several cases originally interpreted as “adenoma,” “epitheliosis,” and “papillomatosis.” Because the distinction between SP-DCIS and UDH has obvious clinical relevance, adjunctive diagnostic tools such as immunohistochemical staining are desirable for the evaluation of problematic cases.

Since the enumeration and cataloguing of the major cytokeratins of human tissues by Moll [4], a large body of research has revealed that the 2 types of cells that compose the epithelium of the normal breast (luminal cells and basal cells) show distinctive patterns of cytokeratin expression. The former contain cytokeratins 7, 8, 18, and 19, and the latter contain high-molecular-weight cytokeratins 5 and 14, and very low levels of cytokeratins 7 and 19 in certain large ducts [5], [6], [7], [8]. Diagnostically, cytokeratin profiles of certain pathologic processes of the breast can help identify the nature of the pathologic cells. In UDH, which represents a proliferation of both luminal- and basal-type cells, most cells stain with antibodies to cytokeratin 5/6. In contrast, conventional ductal carcinoma in situ has the cellular characteristics of luminal cells; therefore, the carcinoma cells stain for cytokeratin 8/18/19, but fail to stain for cytokeratin 5/6 [9]. This difference in cytokeratin profiles has been successfully exploited by using immunohistochemical staining for cytokeratin 5/6 to distinguish UDH from conventional types of ductal carcinoma in situ [8], [9], [10], [11], [12], [13], [14], [15], [16], [17]. We hypothesized that immunohistochemical stains for cytokeratin 5/6 might also be useful for distinguishing UDH from SP-DCIS, which shows greater morphological similarity to UDH than most forms of ductal carcinoma in situ. To the best of our knowledge, this is the first report to specifically examine the expression pattern of cytokeratin 5/6 in SP-DCIS and to evaluate whether its staining profile can reliably distinguish SP-DCIS from UDH.