Also known as hyperplasia of the usual type, UDH is a benign epithelial proliferation characterized by a tendency for the proliferating epithelial cells to bridge across and often fill and distend the involved spaces. Mild forms show epithelial proliferation two to four cell layers thick, but these do not appear to be associated with the same increase in breast cancer risk seen with the more florid examples of UDH.

The proliferation in UDH may have a solid, fenestrated, or micropapillary architecture. If lumens are present within the proliferation, they are irregular and variable in size and shape, often slit-like, and frequently
arranged around the periphery. When epithelial bridges are present, they appear stretched or twisted and commonly show central attenuation. Micropapillary projections, if present, are typically tuft-like or elongated and tapering, resembling the pattern of hyperplasia seen in gynecomastia. The cells comprising UDH are cytologically benign; vary in size, shape, and orientation; are arranged in a haphazard pattern; and have poorly defined borders that may result in a syncytial appearance. The cells do not polarize around lumens within the proliferation. In some instances, there is prominent streaming or swirling of the cells. The nuclei vary in size, shape, and contour and may show overlapping. Nuclear grooves and intranuclear cytoplasmic inclusions may be evident. These architectural and cytologic features of UDH are illustrated in Figs. 3.1, 3.2, 3.3, 3.4, 3.5, 3.6, 3.7, 3.8 and 3.9 and e-Figs. 3.1, 3.2, 3.3, 3.4, 3.5, 3.6, 3.7 and 3.8. Multiple cell types (including metaplastic cells with apocrine or, less often, squamous features) may be present (Figs. 3.10 and 3.11, e-Figs. 3.9 and 3.10). Foamy histiocytes (Fig. 3.12), as well as calcifications (Fig. 3.13), may be seen in association with the proliferating epithelial cells. Rarely, foci of necrosis are present (Fig. 3.14, e-Fig. 3.11). Alterations in the surrounding stroma, such as fibroblastic proliferation, elastosis, and mononuclear cell infiltrates, are uncommon. The key features of UDH are summarized in Table 3.1.

ADH is an epithelial proliferation confined to the mammary ductal-lobular system and is composed, at least in part, by a neoplastic cell population similar to that seen in low-grade DCIS. This population is characterized by relatively small, monomorphic cells with generally rounded nuclei that are evenly spaced and have well-defined borders. The cells may grow in arcades, rigid bridges or bars of uniform thickness, micropapillae that are typically broader at the tips than at the base (club-shaped), solid patterns, or fenestrated (cribriform) patterns in which the cells show polarization around extracellular lumens within the proliferation. The involved spaces may also contain a population of cells more characteristic of UDH or residual normal epithelium (Figs. 3.18, 3.19, 3.20, 3.21, 3.22 and 3.23, e-Figs. 3.13, 3.14, 3.15, 3.16, 3.17, 3.18 and 3.19).

A diagnosis of ADH should only be applied to lesions in which the diagnosis of low-grade DCIS is seriously considered but in which the features are not sufficiently developed for a definite diagnosis of DCIS. The extent of the atypical proliferation is the critical feature that distinguishes ADH from low-grade DCIS. Unfortunately, when spaces are completely involved, there is at present no universally accepted size or extent cutoff for this distinction. Page et al.6 originally proposed that all features of low-grade DCIS be present in at least two separate spaces before a diagnosis of DCIS is rendered and that anything less be categorized as ADH. Tavassoli and Norris7 subsequently proposed that lesions with all of the architectural and cytologic features of low-grade DCIS that are <2 mm in size be given the diagnosis of ADH and that larger lesions be categorized as low-grade DCIS (Fig. 3.24). The 2011 WHO Working Group concluded that it was not possible to recommend one of these approaches over the other and many experts, in fact, use combinations of both in their clinical practice. It should be noted that quantitative thresholds are meant to be practical guidelines that are useful in preventing the categorization of very small, low-grade lesions as DCIS and, in turn, in avoiding overtreatment of patients with minimal or equivocal lesions.8 The Working Group further recommended that a conservative approach be used particularly in core-needle biopsy specimens in which the differential diagnosis includes ADH and low-grade DCIS of limited extent. Categorization of such lesions as either ADH or as “atypical intraductal proliferative lesion” should be sufficient to prompt a surgical excision. Definitive categorization of such lesions should be based on evaluation of the subsequent surgical excision specimen. If no further lesion is found on the subsequent excision, the patient should be managed similarly to patients with ADH.

In our practice, we prefer to fall short of a diagnosis of overt DCIS for small (<2 mm) lesions with borderline features in order to avoid overdiagnosis (Figs. 3.24 and 3.25). In such cases, we render a diagnosis of “severely atypical intraductal proliferation bordering on low-grade DCIS.” If a lesion of this type is present at or near a margin of an excisional biopsy, we recommend a re-excision to exclude the possibility of DCIS. Rendering this diagnosis on a core-needle biopsy sample will prompt excision without labeling the patient as having DCIS in the event that no worse lesion is found on excision.

It should be emphasized that size or extent criteria apply only to the distinction of ADH from low-grade DCIS; intermediate- or high-grade DCIS should be diagnosed as DCIS, even if present in a single space or less than 2 mm in size.

The key features of ADH are summarized in Table 3.2.