By continuing to browse this site you agree to us using cookies as described in About Cookies
Notice: Please be advised that we experienced an unexpected issue that occurred on Saturday and Sunday January 20th and 21st that caused the site to be down for an extended period of time and affected the ability of users to access content on Wiley Online Library. This issue has now been fully resolved. We apologize for any inconvenience this may have caused and are working to ensure that we can alert you immediately of any unplanned periods of downtime or disruption in the future.
The value of fine-needle aspiration (FNA) in atypical proliferative to in situ to low grade invasive breast lesions remains limited due to the overlapping cytologic features of these entities. In the current study the authors review the FNA cytology of endocrine carcinoma and identify common cytologic features that allow for the diagnosis of this uncommon, low grade subtype of mammary malignancy.
The histopathology files from the medical practices of both authors were searched between January 1996 and May 1999 and yielded six cases of endocrine carcinoma. The clinical history and all previous FNA smears were reviewed.
All six patients were elderly women (mean age of 72 years). Four patients presented with breast masses, one patient presented with nipple discharge, and one patient presented with both a breast mass and nipple discharge. All six surgical specimens showed endocrine ductal carcinoma in situ (E-DCIS), with four specimens showing invasive endocrine carcinoma, two of which were labeled as mucinous carcinoma. All invasive components showed the same histomorphologic and immunohistochemical profiles as the in situ components. Cytology demonstrated common features of the cellular smears with clusters and single, monomorphic plasmacytoid tumor cells that possessed moderate amounts of eosinophilic, granular cytoplasm and eccentric nuclei with fine chromatin and inconspicuous nucleoli. In four cases, additional fragments of fine and elaborate papillary fronds also were present.
Mammary carcinoma with endocrine differentiation is an uncommon subtype of low grade carcinoma; to our knowledge only a few series have been reported in the literature.1–4 The authors of what we believe to be the largest series,2 comprising 34 cases, characterized the distinctive clinicopathologic features of the disease, including its common occurrence in elderly women (mean age of 70 years) and its presentation either as a breast mass or with nipple discharge. Histologically, endocrine carcinoma shows a distinctive morphology, together with the expression of at least one neuroendocrine marker, including neuron specific enolase (NSE), chromogranin, or synaptophysin. To our knowledge the cytologic features of endocrine carcinoma have not been described sufficiently in major textbooks5–7 and the literature, with only a few reports found regarding the fine-needle aspiration (FNA) findings of mammary endocrine carcinoma (including the controversial “carcinoid” tumor); these mostly were single case reports.8–12 In the current study we describe the FNA cytologic findings of a series of seven cases of endocrine carcinoma from six patients and review the previously reported cases in an attempt to evaluate any pertinent cytologic features common to these cases that allow for the distinction between endocrine carcinoma and other breast lesion aspirates.
MATERIALS AND METHODS
The files of the histopathology departments of the study institutions were searched for cases of carcinoma of the breast over a period of 3 years (January 1996 to May 1999). Among 526 cases of breast carcinoma, 7 cases of endocrine carcinoma were found. The tissue specimen was fixed in 10% formalin, routinely processed, paraffin embedded, and stained with hematoxylin and eosin (H&E). Immunostaining for NSE, synaptophysin, and chromogranin was performed on the samples using the standard avidin-biotin method. All seven cases had FNA samples, all of which were retrieved and reviewed. Of these, seven cases, one showed insufficient quantity and was excluded from this review. Of the remaining six cases, seven FNAs were performed (two aspirates for Case 4). Smears were made in 5 cases (Cases 1, 2, 3, 5, and 6) and were wet fixed in 95% alcohol; in 3 cases (1 sample from Case 1 and 2 samples from Case 4), cytospin preparations were made with Shandon Cytospin III (Shandon Life Sciences Int. Ltd., UK) and centrifuged at 600 revolutions per minute for 5 minutes. A cell block was made in 1 case (Case 6). All the slides were stained routinely with either H&E or Papanicolaou stains.
A 68-year-old woman presented with bloody right nipple discharge of 6 months' duration and a 2-cm retroareolar mass. FNA showed atypical epithelial proliferation and needle core biopsy showed endocrine ductal carcinoma in situ (E-DCIS). A subsequent simple mastectomy confirmed the diagnosis of E-DCIS. The patient was alive and well at last follow-up, 7 months after surgery.
A 79-year-old woman presented with a 5-cm right periareolar breast lump without discharge. Ultrasound examination was suspicious for malignancy but FNA was nondiagnostic. An excisional biopsy demonstrated invasive endocrine carcinoma with an extensive in situ component. Mastectomy confirmed the biopsy diagnosis, with 5 of 15 axillary lymph nodes positive for malignancy. The patient remained alive and well at last follow-up, 5 months after surgery.
A 70-year-old woman presented with a self-detected, 3-cm, left breast lump superior and lateral to the nipple without discharge. Mammography was highly suspicious for malignancy and FNA was suspicious for low grade malignancy. Needle biopsy showed ductal carcinoma, and modified radical mastectomy showed E-DCIS with a mucinous carcinoma. Twenty-three axillary lymph nodes all were benign. The patient received tamoxifen postoperatively. The patient was alive and well at last follow-up, 10 months after surgery.
A 69-year-old woman presented with bloody left nipple discharge but no mass was palpated. Two FNAs were performed, with the first showing a papillary lesion and the subsequent FNA demonstrating a carcinoma. A core needle biopsy showed an infiltrating ductal carcinoma and subsequent mastectomy confirmed the infiltrating endocrine carcinoma with E-DCIS. All 10 axillary lymph nodes were benign. The patients remained alive and well for > 3 years after surgery.
A 67-year-old woman presented with a 2-week history of a painless, 2-cm breast mass without nipple discharge. FNA was suspicious for mucinous carcinoma. Subsequent excisional biopsy confirmed an invasive mucinous carcinoma and E-DCIS. The patient was lost to follow-up.
An 81-year-old woman presented with a 5-cm, firm breast mass of 1 year's duration without nipple discharge. The FNA diagnosis was endocrine carcinoma because cell block material was available for immunostaining for endocrine markers. Subsequent mastectomy confirmed E-DCIS with foci of early stromal invasion. Two axillary lymph nodes were benign. At last follow-up, the patient remained alive and well after surgery.
The histologic features of all six cases were very similar. There was extensive DCIS with distention of large ducts. The cells were monomorphic, with oval, bland, and eccentric nuclei. The mitotic count was low (< 4 per 10 high-power fields) in all cases. The cytoplasm was eosinophilic and granular and, together with the eccentric nuclei, gave a plasmacytoid appearance. In smaller areas, the nuclei were more elongated and showed a palisading pattern. The majority of areas are solid, although a few areas showed arrangement in festoons and cords. No comedonecrosis was detected. Fibrous strands possessing fibrovascular cores were observed, raising a possibility of papillary lesions (Fig. 1). In all six cases the tumor cells expressed estrogen and progesterone receptors. NSE was positive in six cases, chromogranin was positive in five cases, and synaptophysin was positive in five cases. In one case (Case 2) there was pagetoid spread of the tumor cells into an adjacent intraductal papilloma. Two cases showed E-DCIS only (Cases 1 and 4) and four were associated with an invasive component, two of which were labeled as mucinous carcinoma. The histopathology of all the invasive components was similar to the in situ components, and in the two cases that were termed mucinous carcinoma, small clusters of tumor cells were noted within a loose myxoid stroma. The immunohistochemical profiles of these invasive components were the same as the corresponding in situ components, being positive for one or more endocrine markers.
Of the total seven FNA specimens (two FNAs from Case 4), three samples (Cases 2, 3, and 5) had smears only, two cases (two FNAs from Case 4) had cytospin preparations only, one case (Case 1) had both, and one case (Case 6) had smears and a cell block. Of these seven FNAs, four cases (Cases 3, 4, 5, and 6) were diagnosed as carcinoma (low grade), one case (Case 1) was termed atypical, and one case (Case 4) was diagnosed as a papillary neoplasm. One FNA (Case 2) was inconclusive due to low cellularity; review of this case showed several small clusters of epithelial cells in the smears and thus it was included in the current review. In all cases but one (Case 2) the overall cellularity of the smears was moderate to high, with groups and clusters of tumor cells present that were devoid of myoepithelium (Fig. 2). The regular honeycomb arrangement of the benign epithelium was lost. The tumor cells appeared to be plasmacytoid, as in the histologic sections. In the background, single tumor cells with similar morphology were noted (Figs. 3 and 4). Foamy histiocytes and sheets of benign ductal epithelium also were noted. In four cases (Cases 1, 3, 5, and 6), elaborate papillary structures were present, with delicate fronds of papillae containing fine fibrovascular cores. The tumor cells that were attached to the cores were slightly discohesive, giving a dilapidated appearance (Fig. 5). In one case (Case 6) there was adequate material during FNA for a cell block to be prepared and immunostaining of the cell block material showed expression of all neuroendocrine markers (NSE, synaptophysin, and chromogranin), allowing a definite diagnosis of endocrine malignancy to be made at the time of FNA.
Endocrine carcinoma of the breast is a complex entity without a unifying terminology or concept. Various terms have been used, including spindle cell, argyrophilic, mucin producing carcinoma;13 carcinoid tumor;14, 15 argyrophilic carcinoma;15 and carcinoma with neuroendocrine differentiation.16 Other authors2 have used the term E-DCIS to define the characteristic clinicopathologic features of the in situ component and stress the association with mucoid carcinoma as the invasive component, which possesses the same histomorphologic and immunohistochemical profile, echoing the previous terminology of “spindle cell, argyrophilic, mucin producing carcinoma.”
The histologic diagnosis of endocrine carcinoma in the six cases presented in the current study was based on previously used criteria.2 The clinical features are characteristic of the disease, with patients ranging in age from 67–81 years (mean age, 72 years), in contrast to other types of malignancies, which tend to occur in a younger age group. Four patients presented with palpable lumps, one with bloody nipple discharge and a palpable lump, and another patient with bloody nipple discharge only. The histology of all the cases was characteristic, including low grade nuclear features, nuclear palisading pattern, organoid appearance, the presence of a moderate amount of eosinophilic and granular cytoplasm, rosette and glandular formation, and fibrous septa. The immunohistochemical profile showed the expression of at least one neuroendocrine marker. One case demonstrated involvement of adjacent papilloma. Four cases possessed an invasive component, with two cases labeled as mucinous carcinoma.
Retrospective analysis of the corresponding seven FNA specimens from these six cases highlights the difficulty in establishing an accurate cytologic diagnosis because only four of the seven cases were diagnosed as malignant, with an additional case diagnosed as a papillary lesion and another case diagnosed as an atypical proliferative lesion. These results echo those of other authors who reported overlapping cytologic criteria for ductal hyperplasia (DH), atypical ductal hyperplasia (ADH), and low grade carcinoma.17 In that series, using cytologic criteria18 alone, including a monomorphic cell population of small to intermediate cells arranged in clusters or singly with small and round individual nuclei showing occasional small nucleoli, only 1 of 11 cases of DCIS and 3 of 9 cases of ADH were diagnosed correctly. In another study, “atypical” smears were shown to encompass lesions ranging from DH and ADH to a case of lobular carcinoma in situ,19 further illustrating the limited value of cytology in this group of lesions. In the former study,17 the addition of cytoarchitectural criteria improved the diagnostic accuracy considerably, with 8 of 11 cases of DCIS diagnosed.
In the current series of endocrine carcinoma, the smears fulfilled the cytologic criteria for the diagnosis of low grade malignancy with a monomorphic, intermediate sized epithelial cell population of oval cells. Myoepithelial cells were absent.19 The additional cytologic features of tumor cells possessing moderate amounts of eosinophilic and granular cytoplasm and eccentric nuclei, giving a plasmacytoid appearance, were observed in all cases, corresponding to the histologic characteristic of the poorly cohesive, bland-appearing carcinoma cells with rims of cytoplasm. The cytoarchitectural pattern of finely arborizing papillary fronds most likely arises from the in situ component that distends the large ducts. Dislodgment of the fibrous septa with vascular cores within the large ducts, together with the immediately surrounding tumor cells, during FNA would yield such papillary fronds. In optimal cases in which materials are available for cell block preparation (as in Case 6), immunostaining for neuroendocrine markers can be performed and a definitive diagnosis confirmed. It is interesting to note that although these features fulfill the cytologic criteria for low grade malignancy, the differentiation between in situ and invasive components was not possible based on the smears.
The distinction of endocrine carcinoma from high grade malignancy poses little difficulty by virtue of the absence of nuclear pleomorphism, individual cell necrosis, and a necrotic background as well as mitotic activity. Differentiation from benign papilloma also can be achieved by the absence of columnar cells, myoepithelial cells, apocrine cells, and foamy histiocytes.
More problematic would be the differentiation between other atypical proliferations and low grade malignancies because this cannot be achieved with certainty. Although the common theme within this group of diagnoses is the cytologic appearance of cellular smears comprised of monomorphic clusters of epithelial cells devoid of myoepithelium, theoretic subdivision into atypical proliferation, in situ carcinoma, or low grade invasive carcinoma must be based on subtle differences in cellularity and background status, which at best are unreliable. The identification of plasmacytoid morphology and the elaborate papillary fronds within this confusing entity could point toward the possibility of endocrine differentiation.
For lobular carcinoma, the aspiration material is sparsely cellular, with the tumor cells occurring in small loose clusters, singly or in short Indian files. Although anisonucleosis is minimal, the nuclear-cytoplasmic ratio would be high with intracytoplasmic vacuoles. In papillary carcinoma, the smears are more cellular with a hemorrhagic and necrotic background. Both isolated cells and small papillae or large clusters with three-dimensional projections can be found with more prominent cytologic atypia.
Another problem is distinguishing endocrine carcinoma from the so-called primary or metastatic carcinoid tumor,10, 20–22 with both having similar cytologic features of highly cellular smears with a uniform cell population, bland nuclei, absence of prominent nucleoli, low mitotic count, and moderate amount of eosinophilic granular cytoplasm with a plasmacytoid appearance. Judging from the illustrations, the cases reported as primary carcinoid10 or primary argyrophilic carcinoma11 most likely are cases of E-DCIS; such a concept also is favored by some authors22 who consider “primary breast carcinoid” a variant of conventional breast carcinoma. To our knowledge the few single case reports regarding FNA of a metastatic “carcinoid”20, 21 have not addressed its differentiation from endocrine carcinoma, but the arborizing vascular papillary pattern in the latter is of value because to our knowledge this feature has not been observed in the former. The absence of a primary carcinoid elsewhere, the appropriate clinical context of old age, and symptoms of breast mass or nipple discharge also may help to establish a diagnosis of endocrine carcinoma over metastatic carcinoid.
Last, the similarity of the tumor cells in the smears of endocrine carcinoma can simulate malignant lymphoma. Lymphoma cells are dispersed rather than kept in clusters, with variable nuclear irregularity and coarsely granular or clumpy nuclear chromatin. Careful cytologic evaluation therefore can lead to establishment of the diagnosis.
In the current study we describe the cytologic findings of a series of seven FNA cases of endocrine carcinoma from six patients. Although the established cytologic criteria for low grade malignancy applies, the additional characteristic of a plasmacytoid appearance with eccentric nuclei and granular cytoplasm as well as an arborizing vascular papillary pattern can allow for consideration of this entity. Given the known difficulty in making an accurate cytologic diagnosis of low grade invasive and in situ malignancies, the identification of such characteristic features, in the appropriate clinical setting of elderly patients with nipple discharge or a breast mass, can be of value in raising the suspicion of this uncommon subtype of mammary carcinoma. However, because the number of cases in the current series was limited, more experience is needed before a definite conclusion can be drawn.
The authors wish to thank Professor David L. Page of the Department of Pathology at Vanderbilt University Medical Center for many insightful suggestions and comments regarding the article.