Core Biopsy of Fibroepithelial and Stromal Lesions Michael Bilous MD, Department of Tissue Pathology Institute of Clinical Pathology and Medical Research Westmead Hospital, New South Wales

Clinical History: 55 year old woman with a left breast lesion at 11 o'clock detected by screening mammography. A core biopsy was performed under ultrasound guidance. Figure 1 Click to view with ImageScope Click to view with a Web-Based Viewer Core biopsy, whether conventional or vacuum-assisted, and whether performed for impalpable or palpable lesions, has an important role to play in the preoperative evaluation of this diverse range of lesions of the breast. Included in this category are the following three main lesional types adapted from the WHO classification [1]: Fibroepithelial tumours including fibroadenoma and its variants, tubular adenoma, phyllodes tumour and mammary hamartoma Myoepithelial lesions including adenomyoepithelioma Mesenchymal tumours and tumour-like lesions including myofibroblastoma, pseudoangiomatous stromal hyperplasia, nodular fasciitis, fibromatosis and a variety of tumours of stromal origin such as lipoma, and angiosarcoma Ideally the core biopsy should provide a definite diagnosis from which a clear management plan can be discussed with the patient together with information regarding prognosis. In the case of many of the fibroepithelial and stromal lesions, however, there are inherent problems in providing this information. Firstly the core biopsy findings may be equivocal, for example cellular fibroadenoma versus benign phyllodes tumour. Secondly even if a definite diagnosis can be given from the core biopsy, there may be no clearly established optimal treatment and/or uncertainty concerning prognosis, for example with phyllodes tumours in general. As is the case with core biopsies performed for any breast lesion, information concerning the imaging and clinical findings should always be made known to the reporting pathologist and these should be taken into account in reaching a diagnosis. Inevitably there will be changes seen in the core biopsy for which no definite diagnosis can be given. The biopsy result is then used as a triage together with imaging and any clinical findings in a multidisciplinary approach to determine whether surgical excision is required. This decision will take into account all the information available from the investigations of the patient and importantly, input from the patient herself. Representative lesions from the above classification that can cause diagnostic problems are discussed below. Fibroepithelial tumours Fibroepithelial tumours are derived from the stromal, myoepithelial and epithelial components of the terminal duct lobular unit (tdlu) and this explains their characteristic appearance and the close relationship between ductal and stromal components. The spectrum of tumours that is seen includes a range of lesions from the benign and common fibroadenoma through to the uncommon malignant phyllodes tumour. Overlapping features between the various types of fibroadenoma and the phyllodes tumours can make their differentiation on core biopsy material challenging. Fibroadenomas in younger patients typically have a myxoid stroma which becomes hyalinised, calcified and occasionally ossified in older women. All these forms are readily diagnosed by core biopsy and surgical excision can be avoided. Problems can arise in the diagnosis of juvenile fibroadenoma in which both stromal hypercellularity and epithelial hyperplasia may be seen, and the diagnosis of phyllodes tumour may be considered. The epithelial component in juvenile fibroadenomas usually forms oval or round glands and the stroma tends to have a more homogeneous distribution than is seen in a phyllodes tumour. In addition mitoses are few or absent. Although most commonly seen in women less than 20 years, they can also be seen in older women [2, 3]. A recent study by Shin and Rosen [4] suggested that using Ki-67 to evaluate the proliferative status of both the epithelial and stromal cells in core biopsies from fibroepithelial tumours could be used to differentiate between fibroadenomas and benign phyllodes tumours in a group of 45 women aged 25 years or younger. Complex fibroadenomas are those in which there is at least one of the changes of papillary apocrine hyperplasia, cysts, sclerosing adenosis or epithelial calcifications. This subgroup of fibroadenomas has been shown to increase the risk of subsequent invasive breast carcinoma [5]. Ductal carcinoma in situ (DCIS) and lobular carcinoma in situ (LCIS) may be seen within a core biopsy of a fibroadenoma. While the DCIS or LCIS may be confined to the fibroadenoma, the changes may also be part of a more widespread distribution in the breast. This uncertainty requires that the fibroadenoma with DCIS be excised with sufficient adjacent breast tissue to evaluate for the presence of further DCIS and/or invasive carcinoma. The presence of LCIS on the other hand is not necessarily an indication for surgical excision. However, pleomorphic LCIS or calcifying LCIS with comedo type necrosis in the fibroadenoma may be an exception to this rule but firm evidence to support a management strategy is currently lacking [6]. Probably the most difficult and significant differential diagnostic problem in the interpretation of core biopsies of fibroepithelial lesions is that between cellular fibroadenoma and phyllodes tumour. There are no defined clinical, imaging or pathology features to use in order to distinguish between fibroadenoma, cellular fibroadenoma and benign phyllodes tumour. Cellular fibroadenomas can be only loosely defined as those lesions in which there is stromal hypercellularity greater than that normally seen in a fibroadenoma such that a diagnosis of benign phyllodes tumour is considered by the pathologist. The most useful pathology features that have been used to distinguish between these lesions are stromal cellularity and atypia and stromal cell mitotic rate. MIB-1 and Topo immunohistochemistry have also shown some utility in the differential diagnosis [7, 8]. Our policy is to recommend surgical excision of lesions that are diagnosed as cellular fibroadenomas in order to exclude the possibility of a phyllodes tumour. This policy takes into account the heterogeneity of the stromal changes that can be seen in phyllodes tumours which are such that the core biopsy may fail to have sampled the more diagnostic areas of the tumour. A study by Jacobs et al [7] showed that the degree of stromal hypercellularity seen in the core biopsy (mild versus marked) could distinguish fibroadenomas (mild) from phyllodes tumours (marked). There was however a group with moderate stromal cellularity that on excision proved to be fibroadenomas (12 cases) and phyllodes tumours (8 cases) There is considerable difficulty in grading phyllodes tumours from examination of the surgically excised specimen into the categories benign, borderline and malignant, and thereby predicting the likely clinical course. This difficulty is compounded in the core biopsy specimen in which in addition to possible sampling bias, helpful diagnostic features such as the presence of an infiltrative tumour border, the degree if any of stromal overgrowth and even the mitotic rate may not be able to be adequately assessed. However even with the surgically excised tumour the role and weighting of these histological features in determining grade has not completely assessed or is the subject of conflicting reports. Guidelines exist but lack detail such as the mitotic rate per defined "high power field" [1]. The search for diagnostic markers that can be assessed by immunohistochemistry has not produced an unequivocal result although there are significant differences between benign and malignant phyllodes tumours seen in some series. Those that have been assessed and might be applicable to core biopsies as well as surgically excised specimens include Ki-67, p53, cyclin A, retinoblastoma protein, CD10, EGFR and c-Kit [9, 10, 11, 12, 13]. The latter is of particular interest because of possible treatment implications with Glivec® (imatinib mesylate) Mammary hamartomas, which can be regarded as circumscribed foci of breast tissue often with altered duct/stromal relationships, can be confused with fibroadenomas on core biopsy especially if the fibroadenoma has stromal adipose tissue and smooth muscle and a tubular or pericanalicular growth pattern. A confident imaging diagnosis of a circumscribed mass as well as assurance that the core biopsy was taken from the lesion and not adjacent breast tissue will help in making the diagnosis. Other lesions in which the core biopsy may present a difficult differential diagnostic problem include sclerosed intraduct papilloma and lactating nodule (lactating "adenoma"). Myoepithelial lesions There is a range of lesions in which myoepithelial cells are a prominent component. These include the common sclerosing adenosis which, when combined with atypical apocrine metaplasia, can present a diagnostic problem in a core biopsy. Myoepithelial cells may also be a prominent component of fibroadenomas and intraduct papillomas. Sharing pathology features in common with both of these tumours and also probably a common derivation is the adenomyoepithelioma. A core biopsy is likely to be diagnostic showing a proliferation of myoepithelial cells around small ductal lumina lined by epithelial cells. The tumour is often nodular but usually well defined and a distinct margin may be seen in the core biopsy specimen. Papillary areas are often present. The degree of myoepithelial cell hyperplasia is variable and although the clear cytoplasm of these cells may help differentiate them from the epithelial cells, a myoepithelial cell marker such as p63, calponin or smooth muscle actin helps to identify them. The myoepithelial cells may also show myoid features or be spindle-shaped. Other changes seen in adenomyoepitheliomas include apocrine metaplasia, squamous metaplasia, sebaceous differentiation and microcalcification. Significant cytological atypia and mitotic activity can be seen and this may lead to an overdiagnosis of invasive carcinoma in a limited core biopsy sample. This problem can be avoided by the use of immunohistochemistry and the correlation with imaging findings which will usually show a circumscribed mass lesion. Malignant change can be seen focally in an otherwise benign adenomyoepithelioma and this can be derived from the epithelial, myoepithelial or both cell types. For this reason surgical excision is recommended after a core biopsy diagnosis of adenomyoepithelioma. Mesenchymal tumours and tumour-like lesions A number of mesenchymal lesions seen in the breast correspond to those seen elsewhere in the body and the challenge is to be able to make a confident diagnosis on the limited material in a core biopsy. For example the differentiation between a haemangioma and a well differentiated angiosarcoma may be difficult with a surgically excised specimen and a confident diagnosis of haemangioma from examination of a core biopsy should be given with caution. Other lesions such as myofibroblastoma have a more distinctive appearance on core biopsy consisting of fascicles of spindle cells in a background of collagen. Variants of myofibroblastoma can be more difficult to diagnose. These include lesions with infiltrating margins, or epithelioid myofibroblasts, and those showing marked nuclear atypia. Occasional examples also show hyalinisation of the stroma in which the myofibroblasts are difficult to identify. "Stroma-only" lesions are a particular problem in diagnosis from core biopsy material. The problem is significant because the lesions range from benign and innocuous through locally aggressive to high grade malignant. Stroma-only lesions include nodular fasciitis, fibromatosis, desmoplastic fibroma, pseudoangiomatous stromal hyperplasia (PASH), metaplastic carcinoma, myofibroblastoma and malignant phyllodes tumour. In these cases the diagnosis is made with the help of immunohistochemistry which is used to assist in confirming or excluding a diagnosis. The differentiation between metaplastic carcinoma and malignant phyllodes tumour for example is assisted by the finding of cytokeratin-positive stromal cells in the former. Nodular PASH has a distinctive appearance and the diagnosis is assisted by the demonstration that the spindle cells lining the stromal clefts mark with CD34, calponin and actin but not with vascular endothelial markers such as Factor VIII or CD31. In all of the lesions discussed and especially the fibroepithelial tumours, core biopsy can be of invaluable assistance in patient management both as a guide to the extent of surgery required and also in the avoidance of the need for any surgery. References Tavassoli FA., Devilee P. Tumours of the breast and female genital organs. WHO Classification of Tumours. 2003 IARC Press Pike A., Oberman HA. Juvenile (cellular) adenofibromas. A clinicopathologic study. Am J Surg Pathol 1985; 9:730-736 Mies C., Rosen PP. Juvenile fibroadenoma with atypical epithelial hyperplasia. Am J Surg Pathol 1987; 11:184-190 Shin SJ., Rosen PP. Ki-67 index is diagnostically useful in distinguishing benign fibroepithelial lesions in young females. Lab Invest 2006; 86 Suppl 1: Abstr 182 Dupont WD., Page DL., Parl FF., et al. Long term risk of breast cancer in women with fibroadenoma. N Engl J Med 1994;331:10-15 Sneige N., Wang J., Baker BA., et al. Clinical histopathologic, and biologic features of pleomorphic lobular (ductal-lobular) carcinoma in situ of the breast: a report of 24 cases. Mod Pathol 2002:15(10):1044-50 Jacobs TW., Chen YY., Guinee DG., et al. Fibroepithelial lesions with cellular stroma on breast core needle biopsy: are there predictors of outcome on surgical excision? Am J Clin Pathol 2005; 124(3):342-354 Ridgway PF., Jacklin RK., Ziprin P. Perioperative diagnosis of cystosarcoma phyllodes of the breast may be enhanced by MIB-1 index. Journal of Surgical research 2004; 122:83-88 Kersting C., Kuijper A., Schmidt H., et al. Amplifications of the epidermal growth factor receptor gene (egfr) are common in phyllodes tumours of the breast and are associated with tumour progression. Lab Invest. 2006; 86(1):54-61 Phyllodes tumour of borderline malignancy: seven year follow up with immunohistochemical study. Pathol. Int. 2005; 55(9):585-9 Kuijper A., de Vos R., Lagendijk JH. et al. Progressive deregulation of the cell cycle with higher tumour grade in the stroma of breast phyllodes tumours. Am J Clin Pathol. 2005; 123(5):690-8 Tse GM., Tsang AK., Putti TC., et al. Stromal CD10 expression in mammary fibroadenomas and phyllodes tumours. J Clin Pathol. 2005; 58(2):185-9 Stromal expression of vascular endothelial growth factor correlates with tumour grade and microvessel density in mammary phyllodes tumours: a multicenter study of 185 cases. Human Pathol 2004; 35(9):1053-1057