Rosen's Breast Pathology, 4e

xix

Introduction

carcinoma. We are not convinced that there is merit to the proposed change. As noted in Chapter 12, unless otherwise stated, the term “invasive ductal carcinoma” as used in this book refers to invasive ductal carcinoma, NOS. This flawed proposal does raise an important issue about the precise microanatomical origin of various le- sions included under the broad heading of invasive ductal carcinoma, NOS. The ductal system is a complex series of branching tubules extending from intralobular duct- ules to the major lactiferous ducts that terminate in the nipple. While the endpoint in the nipple is demarcated histologically by a squamocolumnar junction, the point at which terminal ductular epithelium ends and the secretory glandular epithelium begins is less clearly defined. In fact, this junction appears to be somewhat labile and subject to changes induced by physiologic and/or proliferative factors. Although it is possible that a small percentage of lesions cur- rently categorized as invasive ductal carcinoma, NOS, arise from epithelium that may have attributes associated with secretory lobular epithelium, no reliable basis for distin- guishing this subset of carcinomas has been demonstrated. This issue is confounded by the ability of some in situ carcinomas that are classified as DCIS to grow into the glandular compartment of lobules, a process referred to as “lobular cancerization.” Conversely, in situ carcinomas that arise in the glandular epithelium of lobules, LCIS, may ex- tend into ducts either as the solid proliferation that charac- terizes florid and pleomorphic LCIS, or as the dispersed cells found in “pagetoid spread.” When either of these situations is encountered, classification of the in situ carcinoma (and invasive carcinoma if present) is based on the histologic features of the lesion, not on its microanatomic distribution. The foregoing arguments notwithstanding, the fact that approximately 75% of mammary carcinomas are classi- fied as invasive ductal carcinoma, NOS, is troublesome since within this broad category are subsets of tumors with similar histologic appearances that display diverse clinical attributes. The genetic heterogeneity of these tumors has become the subject of intense investigation, especially in the past decade, leading to the identification of molecular subtypes based of gene expression profiles for estrogen receptor (ER), HER2, epidermal growth factor receptor (EGFR), cytokeratins (CKs), and other markers. At pres- ent, the subtypes based on the expression of these markers are referred to as luminal, basal-like, HER2-rich, molecular apocrine, and claudin-low. The complexity of this approach to classification is manifested when a subset of invasive duc- tal carcinomas, NOS, is classified according to more than one molecular subtype. This is illustrated by the report by Lu et al., 45 who studied the expression of claudin subtypes in high-grade invasive ductal carcinomas. Claudins are a group of proteins that play an important role in maintaining tight intercellular junctions. The investigators found a significant correlation between the expression of specific claudins and particular molecular subtypes of breast carcinomas (basal- like, luminal, etc.). In addition, low levels of expression for

claudins 1, 3, 4, 7, and 8 were detected in 30 of 226 (14%) of tumors (referred to as “claudin low”), 77% of which were basal-like. When compared to patients with “non-claudin low” basal-like carcinomas, those with “claudin-low carci- nomas” had a significantly worse recurrence-free survival. It is noteworthy that the attributes which define some of these molecular subtypes among tumors classified as invasive ductal carcinoma, NOS, are also associated with special types of breast carcinoma that are defined by their distinctive histologic and clinical characteristics. 46 For ex- ample, the basal-like subtype characterized as ER(−), PR(−), HER2(−), CK5/6(+), and EGFR(+), which is associated with a relatively unfavorable prognosis among invasive ductal carcinomas, NOS, is also found in prognostically favorable adenoid cystic, medullary, and secretory carcinomas. The relative nonspecific nature of subclassifications of invasive ductal carcinoma, NOS, based on these gene expression profiles is highlighted by the association of the same basal- like attributes with clinically more aggressive metaplastic and pleomorphic lobular special types of breast carcinoma. Despite these overlapping patterns of gene expression between subsets of invasive ductal carcinoma, NOS, and some special histologic types of mammary carcinoma illus- trated by the foregoing examples, gene expression profiling reveals significant and distinctive genomic differences such as the upregulation of genes involved in immune response in medullary carcinomas 47 and the downregulation of genes involved in cell proliferation and migration in adenoid cys- tic carcinomas. 48 As Steensma 49 observed, “This is the age of massive ge- nome surveys- at least for a little while longer.” Gene expres- sion profiling is really in its infancy and will most likely lead to important, lasting discoveries in the future. The recent finding that carcinomas arising in different organ systems share certain genomic alterations 50 has important thera- peutic implications. It is reasonable to predict that genomic studies will eventually lead to stratification of invasive ductal carcinomas NOS, into clinically meaningful subgroups. In the meantime, changing the name of this group of neo- plasms without a sound scientific basis will only cause con- fusion without improving our understanding of the disease. With a little patience, advances in the genomics of breast carcinoma will probably take care of the problem. Paul P. Rosen, MD REFERENCES 1. Roy JE, Hunt JL. Detection and classification of diagnostic discrepan- cies (errors) in surgical pathology. Adv Anat Pathol 2010;17:359–365. 2. Smith ML, Raab SS. Directed peer review in surgical pathology. Adv Anat Pathol 2012;19:331–337. 3. Raab SS, Swain J, Smith N, et al. Quality and patient safety in the di- agnosis of breast cancer. ClinBiochem2013 , http://dx.org/10.1016/j. clinbiochem.2013.04.024 . 4. Renshaw AA, Gould EW. Measuring errors in surgical pathology in real-life practice. Defining what does and does not matter. Am J Clin Pathol 2007;127:144–152.