Daniel Visscher, MD Description This 35 year old woman presented with a palpable breast mass. The mass corresponded to an area of suspicious calcifications on mammogram, prompting a needle biopsy. What is the differential diagnosis? What is the appropriate follow up in view of the findings? Diagnosis and Discussion The core biopsy was diagnosed as “atypical apocrine metaplasia”. This interpretation prompted an open excision that showed apocrine DCIS arising in a complex sclerosing lesion characterized by extensive adenosis with apocrine “atypia” and stromal fibrosis. Apocrine metaplasia (APM) is such a ubiquitous finding in benign breast biopsies that it is often overlooked when examining histologic sections during the course of everyday practice. APM is so common that some authors have even gone so far as to suggest that apocrine phenotype is a variant of normal mammary differentiation. (1). Necropsy-based surveys, however, report 50% prevalence in normal breast tissue (2). The mechanism of APM is unclear; the cells are characterized by expression of androgen (but not estrogen) receptor and GCDFP-15 (prolactin inducible protein). The familiar cytoplasmic appearance reflects the presence of abundant coarse granules. APM is the cellular component in the hallmark lesion of fibrocystic mastopathy – Type 1 cysts, the fluid of which is rich in potassium and sex steroid conjugates. The pathologic spectrum of apocrine breast lesions, though, is quite diverse and may be summarized as follows: Simple and complex cysts Papillary apocrine change (PAC), including apocrine ADH Juvenile papillomatosis APM involving radial sclerosing lesion, adenosis or papilloma, including atypical apocrine adenosis Apocrine DCIS (low and high grade) Apocrine carcinoma, invasive This particular case, an example of atypical apocrine adenosis in a core biopsy, will be used to discuss and illustrate three unresolved controversies involving APM and its relationship to breast carcinoma. Is APM in cysts associated with increased breast cancer risk? There is broad consensus that biopsies showing simple cysts, lined by a single layer of apocrine cells, do not convey increased risk for subsequent breast carcinoma (i.e. in the absence of proliferative lesions such as duct hyperplasia). However, cysts are sometimes lined by multiple layers of apocrine epithelium, termed papillary apocrine change (PAC). Page et al are the only authors to have systematically studied PAC (3). They identified PAC in 27% of 5966 benign breast biopsies; it was architecturally “complex” or “highly complex” in only 6.5% and 1.0%, respectively. “Highly complex” PAC was described as “arches extending far into the central space and tending to cross or touch each other at least twice”. PAC was associated with increased breast cancer risk (1.4-3.1X) compared to the population overall. However, this largely disappeared after correction for concurrent ADH, which was present in 20% of biopsies with highly complex PAC. Further, 50/60 biopsies with highly complex PAC had other proliferative lesions such as usual ductal hyperplasia. Hence, these authors conclude that presence of highly complex PAC is a marker for severe proliferative disease and should prompt a search for high risk index lesions elsewhere in the biopsy. On occasion, as these authors note, the degree of architectural complexity in PAC can be rather worrisome and associated with nuclear atypism, raising the differential diagnosis of “apocrine atypia” or even DCIS, which is the subject of our next controversy. Is there such a thing as “apocrine ADH” and, if so, how is it distinguished from low grade apocrine DCIS? In their analysis of PAC, Page notes that these lesions could be architecturally quite complex, but were uniformly characterized by “nuclear cytological pattern(s) as seen in usual apocrine change”. O’Malley et al (4- 5) have attempted to define criteria to distinguish atypical from malignant apocrine lesions. Apocrine atypia is characterized by 1) three fold nuclear enlargement, 2) nucleolar enlargement or multiple nucleoli, and 3) “slightly irregular” nuclear membranes. Chromatin remains “fine”, however, and necrosis is absent. Importantly, the lesion is “usually of limited extent “(<4mm). Low grade apocrine DCIS, in addition to nuclear enlargement, also shows irregular nuclear membranes as well as coarse chromatin. Further, nuclear atypism is more widespread among the apocrine proliferation (>25% of cells). Finally, the lesion is >4mm and usually involves more than two lobular units and intervening ducts. Apoptosis may be observed, although necrosis is not present (as would be seen in high grade lesions). A truly cribriform architecture is more characteristic of low grade apocrine DCIS as opposed to apocrine atypia. It is important to note that these criteria, as the authors note, represent a proposal. They have never been tested in a long term follow up study. The clinical behavior of low grade apocrine DCIS, similarly, remains undefined. O’Malley et al. by using size and cytologic criteria do not distinguish intraductal apocrine proliferations from those arising in/associated with sclerosing lesions. This has the advantage of providing a unifying approach but does not address, our next controversy; that is, whether or not apocrine atypia in sclerosing lesions (i.e. as in our case) represent a special entity. Is “apocrine adenosis” a high risk or precancerous lesion? As noted above, apocrine metaplasia may involve or arise in papillomas, radial scars and sclerosing adenosis. (Parenthetically, clear distinction between these entities can be problematic at times owing to their frequent admixture in “sclerosing lesions”.) Even when not atypical, apocrine adenosis involving sclerosing lesions can impart a very busy and cellular morphologic picture that, when accompanied by stromal fibrosis and architectural distortion, can resemble a malignant lesion on casual inspection. The benign nature of these problematic biopsies, however, is readily appreciated by demonstrating the abundance of myoepithelium using special stains (actin, p63, calponin – whichever you happen to prefer). There is some evidence that atypical apocrine adenosis lesions are associated with increased breast cancer risk. First, similar to PAC, the frequency of ADH is increased in biopsies that have prominent apocrine adenosis (6). Second, Seidman et al (7) reported that patients with atypical apocrine adenosis were at a 5.5X increased relative risk for subsequent breast carcinoma development. This study, however, did not adjust for the effect of concurrent ADH. It also included a relatively large number of biopsies from older women, raising the question of whether some of their cases may have represented partially sampled examples of apocrine DCIS. (Note the analogy to our case, in which atypical apocrine cells from DCIS colonized adenosis in a core biopsy sample,) A smaller study by Carter and Rosen (8) found no elevated risk although their cohort included only 47 patients with short term follow up (35mo median). Hence, as with apocrine “ductal” lesions with atypia, the association of atypical apocrine adenosis with breast cancer risk is questionable at this time. Conclusions Apocrine metaplasia can be cytologically and/or architecturally atypical. Atypical apocrine lesions, although fortunately uncommon, can arise in histologically complex sclerosing proliferation, thereby creating a problematic histologic differential diagnosis with malignancy. The criteria for distinguishing apocrine atypia from low grade apocrine DCIS are different from those applied to conventional ductal lesions; they involve a different size threshold and are more cytologic (vs architectural) in character. The clinical significance of apocrine atypia in otherwise benign biopsies is not known but there is evidence that it may convey increased breast cancer risk, possibly via an association with other high risk index lesions. Apocrine atypia in a core biopsy should be evaluated carefully, recognizing that partially sampled apocrine DCIS is in the differential diagnosis and that the constellation of cellularity, sclerosis with distortion can mimic invasive malignancy. Open biopsy should be performed to rule out malignancy if apocrine atypia is present in a core biopsy (9). References Eusebi V, Damiami S, Losi L et al. Apocrine differentiation of breast epithelium. Adv Anat Pathol 1997; 4: 139-55. Wellings S, Alpers C. Apocrine cystic metaplasia: subgross pathology and prevalence in cancer-associated versus random autopsy breasts. Hum Pathol 1987: 18: 381-6. Page DL, Dupont WD, Jensen RA. Papillary apocrine change of the breast: Associations with atypical hyperplasia and risk of breast cancer. Cancer Epidem, Biomarkers and Prev. 1996; 5:29-32. O’Malley FP, Page DL, Nelson EH, et al. Ductal carcinoma in-situ of the breast with apocrine cytology: definition of a borderline category. Hum Pathol. 1994; 25: 164-8. O’Malley FP, Bane AL. The spectrum of apocrine lesions of the breast. Adv Anat Pathol 2004; 11: 1-9. Simpson JF, Page DL, Dupont WD. Apocrine adenosis – a mimic of mammary carcinoma. Surg Pathol. 1990; 3: 289-299. Seidman JD, Ashton M, Lefkowitz M. Atypical apocrine adenosis of the breast: a clinicopathologic study of 37 patients with 8.7-year follow up. Cancer. 1996; 77: 2529-2537. Carter DJ, Rosen PP. Atypical apocrine metaplasia in sclerosing lesions of the breast; a study of 51 patients. Mod Pathol. 1991; 4: 1-5. Wells CA, El-Ayat GA. Non-operative breast pathology: apocrine lesions. J Clin Pathol. 2007; 60: 1313- 1320.