Case 3A -

Microinvasive carcinoma (AJCC pT1mic)

History: This is an excisional biopsy from a 63 year old woman with extensive ductal carcinoma in situ measuring 3.0 cm in greatest extent. In two microscopic fields the following changes were identified. What is the appropriate staging classification for this lesion? What is its clinical significance? Case 3A - Figure 1 Case 3A - Figure 2 Diagnosis: Microinvasive carcinoma (AJCC pT1mic). Discussion: The literature contains a number of recent reports detailing the pathologic features and outcome of patients with microinvasive carcinomas. This is defined, per AJCC criteria (T1mic), as unequivocal invasion of stroma by neoplastic cells not exceeding 1 mm in any single focus. Many experts have added the additional, more restrictive, definition that the invasion must involve "non-specialized" interlobular stroma. (In many cases this distinction is difficult if not impossible owing to architectural distortion and periductal/perilobular fibrosis.) Under these fairly restrictive criteria T1mic lesions are relatively uncommon, accounting for 1-3% of breast carcinomas in unselected series. Microinvasion generally arises in a characteristic histologic setting. In most cases there is a background of extensive (greater than 2 cm) DCIS having high nuclear grade/comedo features. These lesions are commonly associated with extensive periductal/perilobular inflammation, consisting primarily of lymphoid infiltrates. Microinvasion is typically multifocal; most authors report an average of three separate foci. We frequently encounter questions about small nests of cells within the fibrous "capsule" that surrounds intracystic papillary carcinomas. There are few, if any, studies that compare these lesions to intracystic tumors that lack such foci. Most authors would probably agree that there is little evidence to suggest that such foci appreciably increase likelihood of metastasis (which, in intracystic papillary lesions, is <1%). Hence, diagnosis of invasion in the setting of intracystic lesions requires the presence of infiltrative growth into surrounding fibrous-fatty stroma. Microinvasive carcinomas may be confused with T1a tumors (i.e. 1-5mm of invasion) having extensive DCIS in which invasion was not appreciable on gross examination. The maximal dimension of invasive growth in all cases with grossly inapparent invasion should be carefully measured on microscopic sections. (A clear ruler works wonders for this.) It is important to distinguish T1a lesions from microinvasive (i.e.T1mic) carcinomas since the former have been shown to have significantly more aggressive behavior (i.e. in terms of the likelihood of metastasis). Recent studies that have carefully defined microinvasive lesions have demonstrated that their clinical behavior is difficult to distinguish from high grade DCIS. Axillary metastases are identified in less than 3% of patients (vs 12% for T1a) and distant relapse occurs in less than 2%. The importance of recognizing microinvasion, thus, consists chiefly in distinguishing these lesions from more advanced stage cases. Some authors have noted that microscopic foci of invasion may be identified both in biopsies and in subsequent lumpectomy/mastectomy specimens. In such cases it should be remembered that microinvasion (that is, Stage T1mic) applies only to cases in which no individual focus of invasion is greater than 1 mm (i.e., the invasive foci should not be summed). In patients with core biopsies showing DCIS with microinvasion, subsequent lumpectomy shows additional invasive disease in about two thirds of cases. Most of these patients have T1a lesions in their lumpectomy. Cases with suspected microinvasion should be subjected to careful sampling, particularly if there is a large field of DCIS. Evaluation of sections with microscopically invasive foci at multiple levels is also useful. This may demonstrate more extensive invasion, in which case the tumor would be upstaged to T1a. In cases with equivocal microinvasion, where leveling of the block demonstrates absence of histologically diagnostic areas, most authors would recommend classification as in situ carcinoma. Some authors advocate employing special stains for myoepithelium (e.g. smooth muscle actin, calponin or p63) since invasive growth, by definition, occurs without a continuous lining of myoepithelium. These may be quite helpful in several settings (see below). There is a relatively low level of interobserver concordance for microinvasive lesions; in our view this reflects in part the difficulty in identifying inconspicuous foci amidst areas of DCIS, with its attendant inflammatory changes and altered architecture. There are also a number of benign or in situ lesions that closely resemble invasive carcinoma, especially in core biopsy material. These are listed below. Lesions that Resemble Microscopically Invasive Carcinoma "Lobular cancerization" Epithelial "displacement" Sclerosing papillomas Radial scars LCIS/DCIS involving adenosis Florid or microglandular adenosis In contrast to lobular cancerization, microinvasion is characterized by the presence of angular (not rounded) nests of cells; these are generally admixed with "single cells". Second the overall contour of lobular cancerization is circumscribed, in contrast to the "non-organoid" growth pattern of invasion, particularly if such foci are identified outside of specialized intralobular stroma. Foci of microinvasion are often accompanied by fibroblast proliferation and collagen deposition; these would be less pronounced in lobular cancerization. Involvement of adenosis by LCIS or DCIS creates a particularly worrisome histologic pattern owing to extreme cellularity with numerous small crowded nests and abnormal cytology. However, desmoplasia is generally lacking and a vaguely lobular pattern can usually be discerned at low magnification. Areas of florid adenosis are generally present, either in continuity with the in situ lesion or in other areas of the slide. Although this pattern becomes recognizable with experience, it is readily detected with stains for myoepithelium. Similarly, these special stains are also quite useful in radial scars/complex sclerosing lesions that may resemble malignant lesions due to the presence of asymmetrically distributed "entrapped glands". Careful observation in these cases, however, will reveal a peripheral rim consisting of benign, proliferative lesions (i.e. not malignant glands). Special stains for myoepithelium, however would not be particularly useful in the diagnosis of microglandular adenosis (MGA). MGA is a (fortunately) rare lesion characterized by proliferation of small glands having a highly infiltrative pattern (i.e. without a lobular outline). It closely simulates low grade invasive carcinomas that form well formed tubules without stromal desmoplasia and it may occur in post-menopausal patients. Two microscopic features are critical in diagnosis of MGA. First, the gland lumens contain highly characteristic eosinophilic secretory material; this finding is extremely rare in carcinomas. Second, although basal lamina is present, some cases of MGA lack a continuous myoepithelial layer. On the other hand, there are a few invasive malignant tumors (listed below) that may be mistaken for benign or in situ lesions. Examples are shown in the accompanying CD. Note that the problems in this setting are twofold. They involve either confusing low grade tumors, such as tubular carcinoma, with adenosis or confusing large nest/confluent invasive patterns with DCIS. Invasive Lesions that Resemble DCIS or Benign Proliferations Bland invasive carcinomas (incl tubular) Invasive cribriform carcinoma Confluent invasive growth patterns Adenoid cystic carcinoma References: de Mascarel I, MacGrogan G, Mathoulin-Pelissier S, Soubeyran I, Picot V, Coindre J. Breast ductal carcinoma in situ with microinvasion. Cancer 2002; 94(8):2134-2142. Ellis, Lee, Elston, Pinder. Microinvasive carcinoma of the breast: Diagnostic criteria and clinical relevance. Histopathology 1999; 35(5):470-472. Padmore R, Fowble B, Hoffman J, Rosser C, Hanlon A, Patchefsky A. Microinvasive breast carcinoma. Cancer 2000: 88(6):1403-1409. Eusebi V, Foschini MP, Betts CM, et al. Microglandular adenosis, apocrine adenosis, and tubular carcinoma of the breast: an immunohistochemical comparison. Am J Surg Pathol 1993;17:99-109.