New variants of epithelial-myoepithelial carcinoma: oncocytic-sebaceous and apocrine.
Abstract: * Context.--Recently described variants of epithelial-myoepithelial carcinoma have not been well characterized but raise a distinct set of differential diagnostic considerations than the classic type.

Objective.--To report a detailed analysis of oncocytic-sebaceous epithelial-myoepithelial carcinoma (OEMCa) and a similar, but novel, variant, apocrine epithelial-myoepithelial carcinoma (ApEMCa).

Design.--Clinical, histologic, and immunophenotypic features of 5 OEMCas and 5 ApEMCas were analyzed. Ultrastructural examination was also performed on 3 OEMCa and 1 ApEMCa tumors.

Results.--The mean age for OEMCa (74.4 years; range, 58-82 years) was slightly higher than for ApEMCa (61.6; range, 46-79 years). All tumors arose in the parotid glands and demonstrated a multinodular pattern of growth with an average size of 3.3 cm (range, 2.3-6.5 cm). Available follow-up (n = 6; 3 OEMCas, 3 ApEMCas) shows a favorable course (no evidence of disease; mean, 17.4 months). Both were morphologically similar, but only OEMCa had sebaceous elements. Phosphotungstic acid hematoxylin staining, antimitochondrial antibody immunohistochemistry, and ultrastructural examination confirm the abundance of mitochondria in OEMCa but not in ApEMCa. The ductal component in ApEMCa was distinguished from that of OEMCa by apical snouts, intracytoplasmic vacuoles, nuclear pleomorphism, prominent nucleoli, and androgen receptor immunoreactivity.

Conclusions.--Oncocytic-sebaceous epithelial-myoepithelial carcinoma and ApEMCa should be considered in the differential diagnosis of oncocytic/oncocytoid salivary gland tumors. Oncocytic-sebaceous epithelial-myoepithelial carcinoma morphology may reflect a senescent phenotype, similar to other oncocytic lesions. The ductal component of ApEMCa shares some similarities with salivary duct carcinoma and supports the notion that epithelial-myoepithelial carcinoma can serve as the progenitor tumor for hybrid tumors.

(Arch Pathol Lab Med. 2009;133:950-959)
Article Type: Clinical report
Subject: Salivary gland tumors (Diagnosis)
Salivary gland tumors (Physiological aspects)
Immunohistochemistry (Usage)
Immunohistochemistry (Methods)
Carcinogenesis (Comparative analysis)
Morphology (Comparative analysis)
Authors: Seethala, Raja R.
Richmond, Jeffrey A.
Hoschar, Aaron P.
Barnes, E. Leon
Pub Date: 06/01/2009
Publication: Name: Archives of Pathology & Laboratory Medicine Publisher: College of American Pathologists Audience: Academic; Professional Format: Magazine/Journal Subject: Health Copyright: COPYRIGHT 2009 College of American Pathologists ISSN: 1543-2165
Issue: Date: June, 2009 Source Volume: 133 Source Issue: 6
Accession Number: 230152025
Full Text: Epithelial-myoepithelial carcinoma (EMCa) is an uncommon, biphasic salivary gland malignancy composed of ductal epithelial cells and myoepithelial cells with a broad morphologic spectrum. In the classic definition of EMCa, the myoepithelial component consists of polygonal cells with clear cytoplasm, whereas the ductal component is composed of small lumina lined by cuboidal, mildly eosinophilic cells reminiscent of intercalated ducts. (1-3) However in our previously reported series of 61 cases, (4) we had observed a much broader morphologic spectrum with respect to cytoplasmic characteristics, relative cell type proportion, and pattern of arrangement. As a result, several new and rediscovered histologic variants emerged.

Among these variants was the oncocytic or oncocytic-sebaceous EMCa (OEMCa), which was initially described by Savera and Salama (5) in 2005. This variant, defined by prominent oncocytic change in the ductal and/or myoepithelial component and sebaceous elements, populated 8% of all our EMCa. (4) The cells in this variant are true oncocytes because the prominent granular, eosinophilic cytoplasm is constituted by abundant mitochondria. Subsequent to that publication, we encountered several EMCa that were, at first glance, very similar to OEMCa in terms of the abundant granular, eosinophilic cytoplasm. However, unlike the oncocytes in OEMCa, these cells could only be considered "oncocytoid" because they did not quite show the same degree of granularity or abundant mitochondria by histochemical or immunohistochemical stains. Additionally, these cells showed periapical snouts, vacuolated cytoplasm, and nuclei with prominent central nucleoli. These cells were also positive for androgen receptor (AR), which is typically expressed in salivary duct carcinoma. We herein describe 5 cases of a new EMCa variant, which we dub apocrine EMCa (ApEMCa), and compare these tumors with an equal number of OEMCa.

MATERIALS AND METHODS

Case Selection

This study was approved by the University of Pittsburgh Medical Center Institutional Review Board (0601084). Five cases (4 consults and 1 in-house case) of OEMCa were retrieved from the Department of Pathology at University of Pittsburgh Medical Center (2004--2008). Four of these cases were extracted from our prior series of EMCa, (4) and detailed clinicopathologic parameters and extended follow-up are presented here. The major inclusion criterion was oncocytic cytoplasm in at least 50% of tumor (epithelial and/or myoepithelial cell components) on hematoxylineosin examination. Five cases of ApEMCa were also retrieved (4 consults and 1 in-house case). Inclusion criteria were ductal luminal cells with apocrine morphology (abundant granular/vacuolated oncocytoid cytoplasm with periapical snouts) and immunoreactivity for AR. Clinicopathologic parameters were obtained.

For OEMCa, the percentages of oncocytic and sebaceous components were assessed, whereas for ApEMCa, the percentage of apocrine components was assessed. Additional histopathologic features were evaluated as described previously, (4) including architectural growth patterns, presence of luminal microcalcifications, tumor necrosis, perineural invasion, and angiolymphatic invasion. Cytologic features of the epithelial and myoepithelial cells were evaluated, including the degree of nuclear atypia (mild, moderate, and severe), mitotic rate, and percentages of epithelial and myoepithelial components.

Ultrastructural Examination

Electron microscopy was performed on formalin-fixed paraffin-embedded material from 4 cases (3 OEMCa and 1 ApEMCa). A 3-mm3 portion of tissue was mechanically dissected from the paraffin block, minced into smaller portions, and then post-fixed in 2% osmium tetroxide. In each case, thick sections were cut and stained with toluidine blue to select suitable areas for ultrastructural examination under a Philips 410 transmission electron microscope (FEI Company, Hillsboro, Oregon).

Histochemistry and Immunohistochemistry

Single-label immunohistochemical stains were performed on all cases with adequate material. Table 1 summarizes the antibodies used. All stains were performed on a Ventana Benchmark autostainer (Ventana Medical Systems, Tucson, Arizona), except the antimitochondrial antibody, which was performed by hand. Staining was visualized using the Ventana iVIEW 2'-diaminobenzidine detection kit (Ventana Medical Systems) as the substrate chromogen (brown) for all antibodies, except S100 and antimitochondrial antibody. The S100 was visualized using the Ventana alkaline phosphatase red detection kit (Ventana Medical Systems) as the substrate chromogen (red). For the antimitochondrial antibody, traditional 2'-diaminobenzidine (Dako, Carpinteria, California) was used.

Double immunohistochemical staining for p63 and smooth muscle actin was also performed. For visualization of smooth muscle actin, nickel ammonium sulfate was added to the 2'-diaminobenzidine substrate chromogen to impart a black color. For p63 visualization, the proprietary Vector NovaRED chromogen (Vector Laboratories, Burlingame, California) was used to impart a dark red color.

To improve visual clarity of the double stains, the Nuance (CRI, Inc, Woburn, Massachusetts) multispectral imaging system was used as previously described.6 Briefly, a liquid-crystal tunable filter optically coupled to a charge-coupled device camera was used for acquisition of spectra, after acquisition of a reference ''image cube'' from the white background, and creation of a spectral library for hematoxylin and each chromogen. The un mixed spectra were reassigned to construct a "pseudofluorescent" composite image with a black background, using the following scheme: green, nickel-2'-diaminobenzidine; red, Vector NovaRED; and light blue, hematoxylin.

RESULTS

Gross and clinical features are summarized in Table 2. The mean age at presentation was 68 years (range, 46-82 years). The mean age for OEMCa (74.4 years; range, 58-82 years) was slightly higher than for ApEMCa (61.6 years; range, 46-79 years). The female to male ratio was even overall (OEMCa, 3:2; ApEMCa, 2:3). All tumors arose in the parotid glands with an average size of 3.3 cm (range, 2.3-6.5 cm). Oncocytic-sebaceous epithelial-myoepithelial carcinomas were slightly larger on presentation (mean, 3.8 cm; range, 2.5-6.5 cm) than ApEMCa (mean, 2.9 cm; range, 2.3-3.4 cm). All patients underwent surgical resection in the form of a parotidectomy. One patient with OEMCa received radiation therapy. Follow-up was available for 6 patients (3 OEMCa, 3 ApEMCa). These patients had no evidence of disease with a mean follow-up of 17.4 months (range, 4.3-38.6 months).

Table 3 summarizes the histologic features of OEMCa and ApEMCa. All tumors showed a multinodular pattern of invasion typical of all EMCa (Figure 1). Although the proportion of epithelial to myoepithelial components varied both between and within tumors, all tumors were decidedly biphasic even in areas of invasion. Papillary growth was noted in 3 of 5 OEMCa (60%; Figure 2, A) but not in ApEMCa. Additionally, luminal calcifications (Figure 2, B) were noted more frequently in OEMCa (all cases) than in ApEMCa (only 1 of 5 cases; 20%). The prevalence of perineural (OEMCa, 2 of 5, 40%; ApEMCa, 1 of 5, 20%) and angiolymphatic (OEMCa, 1 of 5, 20%; ApEMCa, 2 of 5, 40%) invasion were, however, similar, with the invasive areas retaining their biphasic nature, although some tumor emboli were purely myoepithelial. No angioinvasive or perineural invasive areas were purely ductal. For OEMCa, the mean oncocytic component percentage was 73% (range, 60%-85%), whereas for ApEMCa, the mean apocrine component percentage was 93% (range, 70%-95%). In OEMCa, the epithelial component was nearly uniformly oncocytic, but the myoepithelial component ranged from clear to oncocytic (Figure 3, A and B). In ApEMCa, the myoepithelial component ranged from clear to pale am phophilic, with 1 case showing oncocytic cytoplasm. Sebaceous components (Figure 3, C) were only seen in OEM Ca (mean, 26%; range, 15%-40%). In ApEMCa, most tumor (3 of 5; 60%) showed a ductal component with a tubular pattern (Figure 4, A), but 1 case showed solid overgrowth (Figure 4, B), whereas another showed cribriform overgrowth (Figure 4, C).

[FIGURE 1 OMITTED]

[FIGURE 2 OMITTED]

[FIGURE 3 OMITTED]

[FIGURE 4 OMITTED]

[FIGURE 5 OMITTED]

[FIGURE 6 OMITTED]

[FIGURE 7 OMITTED]

[FIGURE 8 OMITTED]

The morphology of the ductal component is what was most useful morphologically in the separation of OEMCa and ApEMCa. Key features separating oncocytic epithelium and apocrine epithelium in these lesions are shown in Figure 5. The lumina of OEMCa are smooth, whereas ApEMCa typically demonstrates apical snouts. The cytoplasm of OEMCa is uniformly granular and deeply eosinophilic, whereas ApEMCa ranges from eosinophilic and granular to pale and vacuolated, often coarsely vacuolated with intracytoplasmic lumina. An overall assessment of nuclear atypia showed that both lesions were similar, however, there was greater nuclear size variation in ApEMCa than in OEMCa. The luminal cells in ApEMCa tended to have more vesicular nuclei with more prominent nucleoli than OEMCa.

On ultrastructural examination, all 3 OEMCa examined showed abundant mitochondria in the epithelial components and, to a lesser degree, in the myoepithelial components. Organellar detail was difficult to ascertain because the material was originally processed for paraffin sections. Sebaceous cells were characterized by numerous vacuoles indenting the nucleus and scattered mitochondria. The lipid droplets were not evident because the examined material was derived from the paraffin blocks (Figure 6). The 1 ApEMCa examined demonstrated a luminal border with long irregular villi and scattered decapitation secretions (Figure 7). The cytoplasm showed varying degrees of vacuolar change. Mitochondria were noted as well, although not to the same extent as in OEMCa.

[FIGURE 9 OMITTED]

Immunohistochemical findings are summarized in Table 4. In all cases, the keratin cocktails (pankeratin, AE1/ AE3, CAM 5.2) selectively highlighted the luminal components, whereas the muscle markers (calponin, smooth muscle myosin heavy chain, and smooth muscle actin) as well as p63 and S100 highlighted the outer myoepithelial cells (Figure 8), although in 1 OEMCa, S100 was not selective for myoepithelial cells. Ki-67 proliferation index averaged 13% for OEMCa (range, 0%-30%;n = 3) and 6% for ApEMCa (range, 5%-10%; n = 5). p53 reactivity was low in all cases except 1 ApEMCa, where strong reactivity was noted in 30% of cells. None of the 4 OEMCa tested showed AR positivity, whereas all 5 ApEMCa showed strong reactivity in the apocrine components (Figure 9, C and D). Additionally, HER2/neu showed membranous positivity in 4 of 5 ApEMCa (80%). Gross cystic disease fluid protein (GCDFP) 15 was positive in all ApEMCa (Figure 10, A and B). Material was unavailable for testing OEMCa for HER2/neu and GCDFP 15. In the 4 OEMCa tested, phosphotungstic acid hematoxylin was diffusely positive in 3 cases (Figure 10, C) and focally positive in 1 case. Antimitochondrial antibody stain was strongly positive in all 4 cases (Figure 10, D). In ApEMCa, phosphotungstic acid hematoxylin was negative, and antimitochondrial antibody stain was only focally positive in all 5 cases.

COMMENT

Both OEMCa and ApEMCa expand the already wide differential diagnosis that must be considered when approaching the pink tumors of the head and neck, consisting of true oncocytic lesions, such as oncocytoma and oncocytic carcinoma, and prominently oncocytoid tumors, such as salivary duct carcinoma. (4,5) At first glance, the distinction between these 2 subtypes of EMCa appears to be equivalent to splitting hairs. But the morphologic and immunophenotypic differences on careful examination are reproducible and cannot be ignored. Oncocytic-sebaceous epithelial-myoepithelial carcinoma comprises true oncocytes, namely, cells in which the granular eosinophilia is a result of the abundance of mitochondria. Apocrine epithelial-myoepithelial carcinoma, on the other hand, consists of cells that are oncocytoid; their eosinophilia is likely a result of a combination of factors: protein content, secretory vacuoles, with perhaps a minor contribution from mitochondria. Additionally, the other morphologic features and immunopositivity for AR and GCDFP-15 are compatible specifically with apocrine differentiation. It is important not to mistake either of these EMCa variants for benign salivary gland lesions, such as oncocytoma/oncocytosis, although OEMCa is more likely to be confused with these entities. On the other hand, it is important not to overinterpret ApEMCa as salivary duct carcinoma, either de novo or ex pleomorphic adenoma, because the implied prognosis and treatment will be dramatically altered. An algorithmic approach is shown for distinguishing select entities in the differential diagnosis in Figure 11.

[FIGURE 10 OMITTED]

The key step in the decision-making tree is the recognition of a biphasic pattern. This is done both morphologically and immunophenotypically. A caveat to using immunohistochemistry alone is that oncocytomas/oncocytic carcinomas and salivary duct carcinoma in situ will have an outer delimiting basal or myoepithelial layer highlighted by immunostains. However, in oncocytoma/oncocytic carcinoma, the transition from this outer layer is gradual and does not result in 2 reproducibly distinct components, (7) and in salivary duct carcinoma in situ, this layer is indistinct and not considered part of the neoplastic process. (8) For these reasons, these entities are not considered truly biphasic.

The second step toward arriving at OEMCa and ApEMCa is the confirmation that the outer layer in a biphasic tumor is indeed myoepithelial. The outer myoepithelial cell layers are typically positive for p63 and a muscle marker, such as smooth muscle actin, calponin, or smooth muscle myosin heavy chain. S100 can be used, but we do not favor using this marker based on our previous experience with its infidelity as a myoepithelial marker. (4) This immunophenotype separates OEMCa and ApEMCa from other bilayered oncocytic lesions, namely Warthin tumor and oncocytic cystadenoma/cystadenocarcinomas, where the outer cell layers are only typically p63 positive, that is, basal. (9,10) In addition, the outer cell layers in Warthin tumors and oncocytic cystadenoma/cystadenocarcinomas are morphologically less prominent than those of the EMCa variants.

[FIGURE 11 OMITTED]

The final step in delineating OEMCa and ApEMCa is subtle and challenging. Entities still in the differential are sclerosing polycystic adenosis and pleomorphic adenoma (PA), with oncocytic/apocrine components or even salivary duct carcinoma arising in PA. Here, it is the morphology and arrangement of the outer myoepithelial layer that separates these lesions. In all EMCa variants, the outer myoepithelial cells have an activated appearance with cell and nuclear size that is more similar and often even larger than those of the ductal component. (3,4) This feature effectively separates these EMCa variants from sclerosing polycystic adenosis, where the outer layer is attenuated. (11) Key features in distinguishing OEMCa and ApEMCa from PAs with oncocytic or apocrine change are the presence of invasion in the EMCa variants, chondroid elements in PAs, and dispersion or streaming of the myoepithelial components into a myxoid stroma in PA. Another consideration with ApEMCa is the distinction from salivary duct carcinoma arising in PA. In the former, the biphasic nature is retained, even in the invasive areas, whereas in the latter, the salivary duct carcinoma is the only invasive constituent. Additionally, although showing atypia and even p53 reactivity in one case, the ductal component of ApEMCa had a low proliferation index and no evidence of necrosis.

The initial, and even the most recent, World Health Organization definition of EMCa requires that the myoepithelial component have clear cells. (1-3) Many tumors with a nonclear myoepithelial component exist that are, from a behavioral standpoint, no different from classic EMCa, just as spindled, plasmacytoid, and epithelioid-patterned myoepithelial carcinomas are no different from their clear-cell counterparts. (4,12) Thus, taxonomically speaking, it is logical to simply include biphasic invasive tumors without a clear-cell myoepithelial component as variants of EMCa.

The question still remains as to the similarities and differences in the tumorigenesis of all variants of EMCa. Because these are rare variants of a rare tumor, the underlying molecular events cannot be currently studied. Any proposed models of tumor development and progression are admittedly speculative. Nonetheless, we will make an attempt at this based on available morphologic and immunophenotypic data.

One point is that the mean age for OEMCa is about a decade higher than for EMCa overall, suggesting that, similar to other oncocytic lesions, (13,14) senescence may play a role in the pathogenesis of this variant. On the other hand ApEMCa is somewhat of a conundrum. The AR+, GCDFP-15+, HER2/neu-positive phenotype essentially defines the salivary duct carcinoma class of lesions. There is evidence to suggest that the immunophenotype precedes the development of malignancy. Recently, a small percentage of PA without overt salivary duct carcinoma, invasive or in situ, were noted to show AR and HER-2/ neu positivity, (15) This finding can be expected in PA because salivary duct carcinoma often evolves from PA, (16) but this phenotype in the ductal component of EMCa is rather unexpected.

However, salivary duct carcinomas arising from other tumors have been documented, and an EMCa-salivary duct carcinoma hybrid tumor has indeed been described. (17) In light of the paucity of mitoses, absence of necrosis, and low Ki-67 proliferation index in our ApEMCa, we did not label these tumors as EMCa-salivary duct carcinoma hybrid tumors. Nevertheless, these findings suggest that apocrine features preceding overt salivary duct carcinoma are possible in EMCa as well. Another argument supporting the feasibility of this phenomenon, is the observation that EMCa is the most common tumor partner in hybrid salivary gland carcinomas. (17,18) Because the other partner is typically higher grade, the implication is that the EMCa serves as the progenitor or base tumor.

When comparing OEMCa and ApEMCa, the morphologic differences are slight but reproducible, and the differential immunohistochemical profile and even ultrastructural features are quite useful in the distinction. Of interest, sebaceous elements were commonly seen in OEMCa but not ApEMCa. This is surprising when comparing these lesions to skin adnexal tumors, where apocrine lesions commonly display sebaceous elements. (19) It is difficult to draw conclusions based on such a low number of cases; we simply may not yet have encountered sebaceous differentiation in apocrine salivary lesions. However, one observation suggesting that this difference may have a biologic basis is that in the skin, both sebaceous and apocrine elements are likely driven androgenically because they both are AR positive. But as we noted previously, sebaceous elements in EMCa are negative4 for AR suggesting a different evolution. With this being the case, the coexistence of sebaceous elements with oncocytic rather than apocrine lesions can be considered a reflection of the different set of rules governing differentiation in salivary gland lesions.

In summary, we compare and contrast 2 relatively new variants of EMCa--OEMCa and ApEMCa. These lesions add to the already sizeable differential diagnostic litany for oncocytic and oncocytoid lesions of the head and neck. Based on limited follow-up, these variants appear to have the same multinodular growth pattern and indolent behavior as their classic EMCa counterparts. The mere existence of these morphologic variants raises several questions regarding tumorigenesis in EMCa, but their molecular pathogenesis will likely remain elusive for some time.

We thank Kim Fuhrer and Kathy Cieply for their excellent technical support. We also thank Dr Jennifer Reed, Tampa Bay, Florida; Dr Brent Benjamin, Renton, Washington; and Drs Zenon Gibas and Brett Moses, Langhorne, Pennsylvania, for providing follow-up data.

References

(1.) Donath K, Seifert G, Schmitz R. Diagnosis and ultrastructure of the tubular carcinoma of salivary gland ducts: epithelial-myoepithelial carcinoma of the intercalated ducts [in German]. Virchows Arch A Pathol Pathol Anat. 1972;356(1): 16-31.

(2.) Ellis GL, Auclair PL. Tumors of the Salivary Glands. Vol 17, 3rd ed. Washington, DC: Armed Forces Institute of Pathology; 1996:203-216.

(3.) Fonseca I, Soares J. Epithelial-myoepithelial carcinoma. In: Barnes EL, Eveson JW, Reichart P, Sidransky D, eds. World Health Organization Classification of Head and Neck Tumours. Lyon, France: IARC Press; 2005:225-226.

(4.) Seethala RR, Barnes EL, Hunt JL. Epithelial-myoepithelial carcinoma: a review of the clinicopathologic spectrum and immunophenotypic characteristics in 61 tumors of the salivary glands and upper aerodigestive tract. Am J Surg Pathol. 2007;31(1):44-57.

(5.) Savera AT, Salama ME. Oncocytic epithelial-myoepithelial carcinoma of the salivary gland: an underrecognized morphologic variant. Mod Pathol. 2005; 18(suppl 1):217A.

(6.) Levenson RM, Mansfield JR. Multispectral imaging in biology and medicine: slices of life. CytometryA. 2006;69(8):748-758.

(7.) McHugh JB, Hoschar AP, Dvorakova M, Parwani AV, Barnes EL, Seethala RR. p63 immunohistochemistry differentiates salivary gland oncocytoma and oncocytic carcinoma from metastatic renal cell carcinoma. Head Neck Pathol. 2007;1(2):123-131.

(8.) Cheuk W, Miliauskas JR, Chan JK. Intraductal carcinoma of the oral cavity: a case report and a reappraisal of the concept of pure ductal carcinoma in situ in salivary duct carcinoma. Am J Surg Pathol. 2004;28(2):266-270.

(9.) Seethala RR, LiVolsi VA, Zhang PJ, Pasha TL, Baloch ZW. Comparison of p63 and p73 expression in benign and malignant salivary gland lesions. Head Neck. 2005;27(8):696-702.

(10.) Prasad AR, Savera AT, Gown AM, Zarbo RJ. The myoepithelial immunophenotype in 135 benign and malignant salivary gland tumors other than pleomorphic adenoma. Arch Pathol Lab Med. 1999;123(9):801-806.

(11.) Gnepp DR, Wang LJ, Brandwein-Gensler M, Slootweg P, Gill M, Hille J. Sclerosing polycystic adenosis of the salivary gland: a report of 16 cases. Am J Surg Pathol. 2006;30(2):154-164.

(12.) Savera AT, Sloman A, Huvos AG, Klimstra DS. Myoepithelial carcinomaof the salivary glands: a clinicopathologic study of 25 patients. Am J Surg Pathol. 2000;24(6):761-774.

(13.) Brandwein MS, Huvos AG. Oncocytic tumors of major salivary glands: a study of 68 cases with follow-up of 44 patients. Am J Surg Pathol. 1991;15(6): 514-528.

(14.) Chang A, Harawi SJ. Oncocytes, oncocytosis, and oncocytic tumors. Pathol Annu. 1992;27(pt 1):263-304.

(15.) DeRoche T, Hoschar AP, Hunt JL. Immunohistochemical evaluation of androgen-receptor, HER-2-neu, and p53 in benign pleomorphic adenoma. Arch Pathol Lab Med. 2008;132(12):1907-1911.

(16.) Williams MD, Roberts D, Blumenschein GR Jr, et al. Differential expression of hormonal and growth factor receptors in salivary duct carcinomas: biologic significance and potential role in therapeutic stratification of patients. Am J Surg Pathol. 2007;31(1 1):1645-1652.

(17.) Croitoru CM, Mooney JE, Luna MA. Sebaceous lymphadenocarcinoma of salivary glands. Ann Diagn Pathol. 2003;7(4):236-239.

(18.) Nagao T, Sugano I, Ishida Y, et al. Hybrid carcinomas of the salivary glands: report of nine cases with a clinicopathologic, immunohistochemical, and p53 gene alteration analysis. Mod Pathol. 2002;15(7):724-733.

(19.) Kazakov DV, Calonje E, Rutten A, Glatz K, Michal M. Cutaneous sebaceous neoplasms with a focal glandular pattern (seboapocrine lesions): a clinicopathological study of 3 cases. Am J Dermatopathol. 2007;29(4):359-364.

Raja R. Seethala, MD; Jeffrey A. Richmond, MD; Aaron P. Hoschar, MD; E. Leon Barnes, MD

Accepted for publication October 15, 2008.

From the Department of Pathology and Laboratory Medicine, University of Pittsburgh Medical Center, Pittsburgh, Pennsylvania (Drs Seethala, Richmond, and Barnes); and Department of Anatomic Pathology, Cleveland Clinic, Cleveland, Ohio (Dr Hoschar).

The authors have no relevant financial interest in the products or companies described in this article.

Reprints: Raja R. Seethala, MD, Department of Pathology and Laboratory Medicine, A616.3 PUH, University of Pittsburgh, 200 Lothrop St, Pittsburgh, PA 15213 (e-mail: seethalarr@upmc.edu).
Table 1. Antibodies Used for Immunohistochemistry

            Antibody                             Company

Pankeratin cocktail
  AE1/AE3                          Dako, Carpinteria, California
  CAM 5.2                          Becton Dickinson, San Jose,
                                     California
  Wide-spectrum keratin            Dako
Pankeratin AE1/AE3                 Chemicon, Temecula, California
Pankeratin CAM 5.2                 Becton Dickinson
p63                                Dako
S100                               Dako
Smooth muscle actin                Dako
Calponin                           Dako
Smooth muscle myosin heavy chain   Dako
Antimitochondrial antibody         Biogenex, San Ramon, California
p53                                Dako
Ki-67                              Dako
Androgen receptor                  Dako
HER2/neu                           Novocastra, Newcastle,
                                     United Kingdom
Gross cystic disease fluid         Cell Marque, Hot Springs, Arkansas
  protein-15

            Antibody                 Clone          Dilution

Pankeratin cocktail
  AE1/AE3                          AE1/AE3      1:1000
  CAM 5.2                          CAM 5.2      1:80

  Wide-spectrum keratin            Polyclonal   1:2000
Pankeratin AE1/AE3                 AE1/AE3      1:100
Pankeratin CAM 5.2                 CAM 5.2      1:50 (prediluted)
p63                                4A4          1:500
S100                               Polyclonal   1:300
Smooth muscle actin                1A4          1:50
Calponin                           CALP         1:50
Smooth muscle myosin heavy chain   SMMS1        1:800
Antimitochondrial antibody         113-1        1:50
p53                                DO7          1:100
Ki-67                              ki-55        1:25
Androgen receptor                  AR411        1:100
HER2/neu                           CB11         1:200

Gross cystic disease fluid         23A3         1:25
  protein-15

Table 2. Clinical and Gross Parameters

Case                                             Size,
No.    Age, y/Sex   Diagnosis        Site         cm     Margins

1         73/F      OEMCa       R deep parotid    2.5       -
2         78/M      OEMCa       R parotid         6.5       -
3         58/M      OEMCa       R parotid         2.5       -
4         82/F      OEMCa       L parotid         4.3       +
5         81/F      OEMCa       L parotid         3.0       -
6         79/M      ApEMCa      R parotid         3.4       -
7         49/M      ApEMCa      R parotid         3.0       -
8         69/F      ApEMCa      Parotid, NOS      2.3
9         46/M      ApEMCa      L parotid         3.0
10        65/F      ApEMCa      R parotid         2.6       -

Case                 Local      Nodal     Distant     Outcome,
No.    Treatment   Recurrence   Status   Metastasis      mo

1          P
2        P, X          No        cNeg        No       NED 9.7
3          P           No                    No       NED 33.5
4          P                                          NED 38.6
5          P
6          P           No                    No       NED 22.7
7          P           No        cNeg        No       NED 11.6
8          P
9          P           No                    No       NED 4.3
10         P

Abbreviations: ApEMCa, apocrine epithelial myoepithelial carcinoma;
C, chemotherapy; cNeg, clinically negative (no histologic
assessment); ND, neck dissection; NED, no evidence of disease; NOS,
not otherwise specified; OEMCa, oncocytic epithelial myoepithelial
carcinoma; P, parotidectomy; X, radiation therapy.

Table 3. Histologic Parameters

Case               Nod
No.    Diagnosis   Scar   Papillae   Calcification   PNI   ALI

1      OEMCa        N        N             Y          Y     N
2      OEMCa        N        N             Y          N     Y
3      OEMCa        N        Y             Y          N     N
4      OEMCa        N        Y             Y          Y     N
5      OEMCa        N        Y             Y          N     N
6      ApEMCa       N        N             N          N     N
7      ApEMCa       N        N             N          N     Y
8      ApEMCa       Y        N             N          N     N
9      ApEMCa       N        N             N          N     N
10     ApEMCa       Y        N             Y          Y     Y

Case              Morphology      Morphology      Oncocytes,
No.     Atypia    Epithelial    Myoepithelial         %

1      Moderate   Oncocytic    Oncocytic              75
2      Moderate   Oncocytic    Clear/oncocytic        85
3      Moderate   Oncocytic    Clear/oncocytic        60
4      Moderate   Oncocytic    Clear/oncocytic        60
5      Moderate   Oncocytic    Oncocytic              85
6      Moderate   Apocrine     Oncocytic              --
7      Moderate   Apocrine     Clear/pale             --
                                 ampho
8      Moderate   Apocrine     Clear/pale             --
                                 ampho/spindled
9      Moderate   Apocrine     Oncocytic/pale         --
                                 ampho
10     Severe     Apocrine     Clear                  --

Case   Sebaceous,       Apocrine,
No.        %            %/Pattern

1        20 (a)            --
2          15              --
3          40              --
4          40              --
5          15              --
6          --       100 tubular
7          --       100 tubular
8          --       70 (a) solid
9          --       100 tubular
10         --       95 (a) cribriform

Abbreviations: ALI, angiolymphatic invasion; ampho, amphophilic;
ApEMCa, apocrine epithelial myoepithelial carcinoma; N, no; Nod,
nodular; OEMCa, oncocytic epithelial myoepithelial carcinoma; PNI,
perineural invasion; Y, yes.

(a) Cases with a conventional epithelial myoepithelial carcinoma
component as well.

Table 4. Immunohistochemical Findings

Case                 Keratin                       Muscle    Ki-67,
No.    Diagnosis   Cocktail (a)   p63     S100     Markers   % (b)

1      OEMCa           epi        myo
2      OEMCa           epi        myo   myo, epi     myo
3      OEMCa           epi        myo                myo       30
4      OEMCa           epi        myo     myo        myo       0
5      OEMCa           epi        myo                          10
6      ApEMCa          epi        myo     myo        myo       5
7      ApEMCa          epi        myo     myo        myo       10
8      ApEMCa          epi        myo     myo        myo       5
9      ApEMCa          epi        myo     myo        myo       5
10     ApEMCa          epi        myo     myo        myo       5

Case   P53,                                                Antimito-
No.    % (b)   AR    H2N (c)   GCDFP-15 (d)   PTAH (e)   chondrial (e)

1              Neg                              Pos           Pos
2
3              Neg                              fPos          Pos
4        5     Neg                              Pos           Pos
5        5     Neg                              Pos           Pos
6        5     Pos     Pos         Pos          Neg          fPos
7        0     Pos     Pos         Pos          Neg          fPos
8        5     Pos     Pos         Pos          Neg          fPos
9        5     Pos     Neg         Pos          Neg          fPos
10      30     Pos     Pos         Pos          fPos         fPos

Abbreviations: ApEMCa, apocrine epithelial myoepithelial carcinoma;
AR, androgen receptor; Epi, epithelial (luminal) component; fPos,
focally positive; GCDFP-15, gross cystic disease fluid protein-15;
H2N, HER2/neu; myo, myoepithelial (abluminal) component; Neg,
negative; OEMCa, oncocytic epithelial myoepithelial carcinoma; Pos,
positive; PTAH, phosphotungstic acid hematoxylin.

(a) Pankeratin, AE1/AE3, and/or CAM 5.2.

(b) Percentage based on overall expression in both epithelial and
myoepithelial components. Distribution between components was
roughly even.

(c) Weak to moderate, complete membranous staining in at least 10%
of cells was the minimal criterion for positivity. Staining mostly
limited to the epithelial component.

(d) Staining limited to epithelial component with occasional
bleeding into surrounding myoepithelial component and stroma.

(e) Staining was present in both components but was more pronounced
in the epithelial component.
Gale Copyright: Copyright 2009 Gale, Cengage Learning. All rights reserved.