Search

Abstract: Basal cell carcinoma (BCC) is the most common skin cancer, and it has numerous histologic mimics with variable prognoses and treatments. Although some immunohistochemical stains can be used for the differential diagnosis of BCC, variability and overlap in results can complicate their interpretation. Immunohistochemical staining for glioma-associated oncogene-1 (Gli-1) was performed on 26 nodular BCCs, 22 infiltrative BCCs, 9 basaloid squamous cell carcinomas, 12 desmoplastic trichoepitheliomas, 19 Merkel cell carcinomas, 11 sebaceous carcinomas, 10 cylindromas, 14 spiradenomas, 12 adenoid cystic carcinomas (AdCC), and 1 solitary trichoepithelioma. Strength of staining was scored as 0, 1+, 2+, or 3+, and distribution of staining was categorized as diffuse, multifocal, or focal. Strong, diffuse Gli-1 expression was seen in all tumors with basal epidermal-type differentiation, including BCC, trichoepithelioma, and basaloid squamous cell carcinoma. All examples of Merkel cell carcinoma were negative for cytoplasmic expression. Seven out of 11 sebaceous carcinomas were negative for Gli-1, and the remaining 4 showed 1+ expression. Cylindroma, spiradenoma, and AdCC, each an adnexal skin tumor, showed the most variable staining, but with cylindroma and spiradenoma demonstrating comparable labeling patterns. Overall, although Gli-1 may not distinguish between basal epidermal-type tumors, it may have a role in separating that group from lesions with adnexal differentiation, particularly sebaceous carcinoma, but also cylindroma, spiradenoma, and AdCC. Any cytoplasmic staining seems to exclude the diagnosis of Merkel cell carcinoma., Competing Interests: The authors declare no conflicts of interest., (Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved.)

Erythema multiforme (EM), Stevens-Johnson syndrome (SJS), and toxic epidermal necrolysis (TEN) comprise a family of mucocutaneous diseases associated with significant morbidity and mortality. Previous studies have confirmed the presence of autoantibodies to desmoplakin (Dp) I and II in patients with EM, SJS, and TEN. Truncated Dp production leads to characteristic changes visible on light microscopy: perinuclear clumping of keratin filaments and dyskeratotic keratinocyte. Based on these observations, the question arises as to whether a loss of Dp immunoreactivity in skin biopsies could serve as a diagnostic marker of EM, SJS, and TEN. This study analyzed Dp immunostaining patterns in 20 patients with EM or SJS/TEN. To assess the specificity of this approach, Dp immunostaining was also performed on specimens from patients with 5 potential histologic mimics of EM, SJS, and TEN. All of the samples from patients with EM, SJS, and TEN demonstrated absent or markedly diminished staining for Dp. A χ test demonstrated a statistically significant difference between the staining patterns in EM, SJS, and TEN and each of the other diagnostic groups that were investigated. This is the first report demonstrating statistically significant specificity of Dp staining patterns in EM/SJS/TEN as compared with other interface dermatitides.

Basaloid tumors of the skin pose a diagnostic challenge to pathologists, because the differential diagnosis is broad, sometimes with subtle differentiating features. We evaluated SOX10 expression in 120 primary cutaneous tumors with epidermal, sweat glandular, neuroendocrine/neuroectodermal, follicular, and sebaceous lineages. Our findings were compared with those of previous studies that evaluated SOX10 in tumors of the skin. SOX10 staining was seen in the majority of sweat gland tumors with the exception of syringoma and microcystic adnexal carcinoma. There were no immunoreactive cases among epidermal, neuroendocrine/neuroectodermal, follicular, or sebaceous tumors. These findings are comparable to reported in previous studies, and show SOX10 can be a useful adjunct in the differential diagnosis of nodular basaloid skin tumors. That marker has less utility in the assessment of sclerosing basaloid cutaneous neoplasms, because such tumors are almost uniformly nonreactive for it.

Indeterminate dendritic cell tumor (IDCT) is a cutaneous proliferation of histiocytes that share morphologic and immunophenotypic properties with Langerhans cells. IDCT was recently included in the updated WHO classification of tumors of hematopoietic and lymphoid tissues. Recent studies have shown that some cases of IDCT demonstrate an ETV3-NCOA2 translocation, supporting the idea that IDCT is a clonal neoplasm. We report 2 new cases of IDCT at our institution lacking the ETV3-NCOA2 translocation. We also present a comprehensive review of reported cases of IDCT in the medical literature. Eighty-five cases of IDCT were reported in the literature between 1985 and 2016. The median age at diagnosis was 45 years. In contrast to Langerhans cell histiocytosis, IDCT is limited to the skin in the majority of cases (88%) and generally follows an indolent clinical course. Most reported lesions are cured with complete excision. However, the histologic features of IDCT and langerhans cell histiocytosis are similar. Conjoint immunostaining for CD1a and langerin is necessary for optimal classification.

Sclerosing squamous cell carcinoma (SCC), also known as "desmoplastic" SCC, is a rare subtype of cutaneous malignancy. This variant is clinically significant because it is associated with an increased risk of local recurrence and metastasis. We herein present 16 examples of sclerotic SCC of the skin in 8 men and 3 women, with a median age of 66 years. The most common site of origin for this tumor is the skin of the head and neck, including the scalp (5 tumors in 2 different patients), forehead (3 cases), nasal ala (2 cases), neck (2 cases in the same patient), ear (2 cases), cheek (1 case), and chest (1 case). Microscopically, sclerosing SCCs are characterized by cellular cords, nests, and islands, as well as scattered single cells infiltrating densely desmoplastic and collagenized connective tissue. The differential diagnosis principally includes sclerosing basal cell carcinoma, microcystic adnexal carcinoma, and desmoplastic trichoepithelioma. The main goals of this study are to further characterize these lesions pathologically, and increase general awareness of this SCC subtype.

The immune checkpoint inhibitors targeting cytotoxic T-lymphocyte associated antigen 4 (CTLA-4) and the programmed death protein 1 (PD-1)/PD-L1 pathway have recently shown promising therapeutic results in patients with metastatic melanoma. Dermatologic side effects of these agents occur in ∼30-40% of cases. Here, we report the development of a biclonal cutaneous T-cell lymphoproliferative disorder in a patient being treated with ipilimumab (a CTLA-4 inhibitor) for metastatic melanoma. Nivolumab (a PD-1 inhibitor) had also been administered to him previously. An 8 mm reddish papule appeared on the skin of the left forearm. A biopsy of that lesion showed an atypical population of predominantly CD4-positive, CD30-positive T-cells that also expressed PD-1 and PD-L1 immunohistochemically. PCR studies for T-cell receptor rearrangements showed the presence of two distinct clonal T-cell populations. The lesion was completely excised and the patient had no local recurrences. There was also no subsequent evidence of a systemic lymphoproliferative process. Although the development of a lymphoid skin lesion in our patient may have only been coincidentally related to his treatment, immunostimulatory drugs could theoretically cause clonal expansion of a population of lymphocytes that leads to a lymphoproliferative disorder.

The transcription factor GATA3 is a recently described biomarker that is highly expressed in bladder and breast carcinomas. Although it has shown sensitivity as a marker of primary bladder carcinomas with purely urothelial differentiation, the ability of GATA3 to label primary bladder carcinomas with variant morphologic patterns has been incompletely assessed to date. The current study was designed to determine whether GATA3 staining is retained in "unconventional" bladder carcinomas. Eighty-eight cases of primary bladder cancers were retrieved from the authors' institutional archive, and they included the following histomorphologic types: 6 small cell carcinomas, 12 sarcomatoid carcinomas, 17 adenocarcinomas (both primary and urothelial variants with glandular differentiation), 24 micropapillary carcinomas, and 27 squamous cell carcinomas (both primary and urothelial variants with squamous differentiation). A tissue microarray was constructed and automated immunostaining for GATA3 (Clone L50-823, Biocare Medical, Concord, CA) was performed using standard technique. Among the 5 variants of unconventional bladder carcinoma, only the micropapillary and sarcomatoid forms exhibited consistent and strong immunolabeling for GATA3. Hence, the sensitivity of this determinant is diminished in several histologic forms of primary bladder carcinoma. That fact will affect the interpretation of GATA3 stains in the context of possible metastasis from primary bladder carcinomas with variant morphologic patterns, as well as their distinction from secondary bladder involvement by tumors of nonurothelial origin.

Dermatopathology lost a giant in the field with the death of Philip H. Cooper, MD, on Friday, January 30, 2015. The following obituary represents a celebration of his life and his contributions to our field.

Clear-cell renal cell carcinoma (CC-RCC) is the most common primary kidney malignancy, yet this morphology is not unique to renal primary tumors, as clear-cell variants of numerous nonrenal carcinomas of varying lineages exist. Therefore, because of CC-RCC's ability to metastasize to nearly any anatomic location, ancillary studies such as immunohistochemistry are often needed to establish the diagnosis. Despite CD10 and renal cell carcinoma monoclonal antibody (RCCma) being touted as sensitive and specific markers, some have suggested that more recent stains including PAX-2, PAX-8, or adipophilin (ADP) are more robust markers of CC-RCC. In this study, 26 cases of CC-RCC, and 51 nonrenal carcinomas with clear-cell histomorphology (CCM) were stained with CD10, RCCma, PAX-2, PAX-8, and ADP. CA-125 was also included to help distinguish CC-RCC from Müllerian clear-cell carcinomas, due the known expression of PAX-2 and PAX-8 in both these entities. RCCma highlighted 77% of CC-RCC and 27% of the CCM group, whereas CD10 was positive in 85% and 25%, respectively. ADP highlighted all CC-RCC and 45% of CCMs. PAX-2 was positive in 81% of CC-RCC and 24% of CCM, whereas PAX-8 stained 100% of CC-RCC and 39% of CCM. Müllerian-derived tumors (clear-cell carcinomas of the ovary, vagina, and cervix) were positive with PAX-2 and PAX-8 in 69% and 100% of cases, respectively. No cases of CC-RCC stained with CA-125, whereas 88% of the Müllerian-derived tumors were positive. In summary, although new markers such as PAX-2 and PAX-8 tend to be more sensitive markers of CC-RCC, they lose specificity when Müllerian tumors are included. Inclusion of a classic renal marker such as CD10 or RCCma in the immunohistochemical panel, as well as CA-125 obviates this difficulty.

Distinction between radial growth phase (RGP) and vertical growth phase (VGP) in cutaneous melanomas is prognostically significant. Despite established morphological criteria, molecular markers to separate RGP and VGP have not been well established. The goal of this study was to investigate associations of p16, WT1, and Fli-1 with RGP-to-VGP progression, by immunohistochemistry. The p16 is a tumor suppressor, whereas WT1 and Fli-1 are transcriptional activators. The authors hypothesized that entry into VGP would be associated with decreased p16 and increased WT1 and Fli-1. Paraffin sections from 18 RGP and 15 VGP melanomas were immunostained with well-characterized antibodies to p16, WT1, and Fli-1. Melanoma growth phases were determined using precodified morphological attributes. In RGP melanomas, p16 was expressed in 15 of 18 (83%), WT1 in 17 of 17 (100%), and Fli-1 at least focally in 6 of 18 (33%). The deep dermal component of VGP melanomas stained positively for Fli-1 in 9 of 14 (64%), strongly for WT1 in 10 of 14 (71%), and strongly for p16 in only 2 of 15 (13%). Observed patterns of WT1 immunopositivity did not support the authors' hypothesis; it is not likely to be a good indicator of VGP. On the other hand, Fli-1 staining trended toward more positive deep tumor compartment staining and p16 to weaker staining in the deep compartment. At present, application of histological criteria remains the best method for assignment of growth phase in melanomas; however, p16 and possibly Fli-1 immunostains may serve as useful adjuncts in morphologically indeterminate cases.

Melanocytic nevus rests in lymph nodes are a known diagnostic challenge, especially in patients with a history of melanoma. Reticulin and NM23 have been studied in this context. The pattern of reticulin staining in melanomas surrounds groups/nests of melanocytes but individual cells in benign nevi. NM23, a metastasis-suppressor gene, has an association with metastatic potential in melanomas and some carcinomas. Twenty-eight cases (14 cases of metastatic melanoma to lymph nodes and 14 cases of lymph node nevus rests, all confirmed with Melan-A staining) were stained with reticulin and NM23. The pattern of reticulin staining was reported as surrounding groups if staining was noted in approximately 5-10 melanocytes in greater than 50% of the lesion but was otherwise reported as surrounding individual melanocytes. Cytoplasmic staining was considered to represent reactivity for NM23. Reticulin staining around groups of melanocytes was identified in all 14 cases of metastatic melanoma. Regarding nodal nevus rest cases, 12 of 14 cases (86%) demonstrated staining around individual melanocytes, whereas in 2 cases, reticulin surrounded melanocytic groups. NM23 staining was equivocal in all cases. Reticulin staining reliably invests groups of melanocytes in cases of metastatic melanoma, whereas in nodal nevus rests, it predominantly surrounds individual melanocytes. NM23 demonstrated no discriminatory value in this analysis. In cases in which a collection of melanocytes is present within a lymph node, reticulin deposition around individual melanocytes supports a diagnosis of lymph nodal nevus rest.

CD10 is a cell surface-related neutral endopeptidase that is involved in cleaving cytokine peptides; it may also play a role in the proliferation of tumor cells and their acquisition of invasiveness. On account of its association with other overtly epithelial neoplasms, we hypothesized that CD10 might be preferentially expressed in the sarcoma-like components of sarcomatoid carcinomas as compared with true sarcomas. Immunohistochemical labeling for CD10 was assessed in various sarcomas and sarcomatoid carcinomas. An aggregate score was generated using both the intensity and extent of staining throughout the neoplasms. Overall, CD10 was expressed more often in true sarcomas (27 of 33 cases) as contrasted with sarcomatoid carcinomas (22 of 34 cases), but with no statistically significant difference between the 2 groups. Uterine "carcinosarcomas" expressed CD10 with accentuation in periglandular tumor cells. The sarcoma-like components in squamous cell carcinoma of the respiratory tract (larynx and lung), tongue, bladder, skin, and penis also expressed CD10 consistently. In the final analysis, there was no difference in CD10 expression between sarcomatoid carcinomas and true sarcomas. This marker seems to have little, if any, differential diagnostic value in the specified histopathologic context.

Proximal epithelioid sarcoma is an extremely uncommon lesion of the vulva, with the potential for aggressive behavior. We report a case of this entity and discuss its relationship to the epithelial-type "malignant rhabdoid tumor" (MRT) of the soft tissue. On the basis of a review of the pertinent information on both these entities, it is concluded that they likely represent biologically different but morphologically and immunohistochemically similar neoplasms. Both proximal epithelioid sarcoma and MRT comprise epithelioid cells with densely eosinophilic cytoplasmic inclusions, and they each lack the INI1 gene product. Nevertheless, the literature suggests that other selected genetic differences between the 2 lesions, and the more rapid and aggressive course of MRT distinguish these tumor types as separate clinicopathologic entities.

Oculocutaneous oncocytic tumors (OCOTs) are uncommon neoplasms that have been reported only rarely in the dermatopathology literature and whose immunophenotypic profile has not been well characterized. The clinical, histologic, and immunophenotypic features of 2 cases seen by the authors were assessed, and relevant publications in the literature were reviewed. Both patients with OCOTs were adult women with gradually enlarging, asymptomatic lesions involving the caruncle; they were locally excised. Histologically, the tumors were well-circumscribed nodules comprised of large oxyphilic cells arranged in confluent sheets and forming glandular spaces with secretory material. Microcystic areas and sparse intermingled goblet-cells were also apparent. Nuclear atypia and infiltrative growth were absent. Mitotic activity was absent in one case; a single mitotic figure was identified in the other. Immunostains demonstrated uniform expression of pankeratin and mitochondrial antigens. Both neoplasms were also labeled for markers associated with cutaneous adnexal, lacrimal, and minor salivary glandular tissue, including alpha-1-antitrypsin, gross cystic disease fluid protein-15, carcinoembryonic antigen, lysozyme and MUC1; each case expressed 4 of the 5 substances. Some cells expressed cytokeratins 5/6 and p63 consistent with the presence of basal-type differentiation in a subset of cells. No definite evidence of myoepithelial differentiation was demonstrated, as stains for smooth muscle actin, muscle-specific actin, and S100 protein were negative. Estrogen and progesterone receptor proteins were absent; strong cytoplasmic immunoreactivity for androgen receptor protein was evident, but nuclear staining was absent. The authors conclude that OCOTs show glandular differentiation. A review of the literature disclosed that none of these lesions arising in the caruncle behaved aggressively, in contrast to occasional tumors in other oculocutaneous sites.

There are no firmly established guidelines for the use of antibodies in immunohistology as individual tests or panels. Practicing pathologists must rely on information available in individual publications, review articles, books, and internet-based databases to develop diagnostic immunohistochemical algorithms for their individual practices. In contrast, other medical specialties have crafted many evidence-based practice guidelines (EBG) that are widely used; these have helped to augment standardization and cost effectiveness. In particular, the use of several "epithelial" and "mesothelial" antibodies has been proposed to distinguish epithelioid malignant mesothelioma from metastatic pulmonary adenocarcinoma. Other authors have previously done systematic literature reviews of this subject up through 2004 and integrated the results of 88 publications into summarized test-performance values for 15 preselected immunohistochemical markers. The results suggested that 7 tests provide optimal sensitivity and specificity (MOC-31, BG8, CEA, TTF-1, CK5/6, WT-1, and HBME-1), but they provide no guidance for integration of such data into EBG. Odds ratios (ORs) were employed to compare the effectiveness of any single test, and chosen combinations thereof, in the differential diagnosis of malignant mesothelioma and metastatic pulmonary adenocarcinoma. Surprisingly, selected single immunostains or antibody pairs yielded ORs (varying from 96.34 to 1233.19) that were equal or better in efficacy when compared with more comprehensive panels. These results support the potential value of systematic reviews, meta-analysis, and OR calculations for development of EBG in diagnostic immunohistology.

The immunohistochemical diagnosis of atypical epithelial proliferations in pleural fluid is a challenging topic in cytopathology and surgical pathology. Mesothelioma may be simulated clinically and radiologically by several other nonneoplastic and neoplastic disorders, mandating that strict histologic, histochemical, immunohistochemical, and ultrastructural guidelines be followed for its diagnosis. Because of its availability to most laboratories, immunohistochemistry has emerged as the most commonly used procedure for the diagnosis of pleural malignancies. This review considers the current status of that investigative modality, with particular attention to lesions that are suspected to be mesothelial.

Diagnostic immunohistochemistry (DIHC) is an extremely valuable technological development that has revolutionized the practice of surgical pathology over the past 25 years. Like any laboratory procedure, however, the precise concatenation of methodological details that are used by any given investigator has a major effect on final results. In this overview, the authors consider several of these variables and illustrate how they may be optimized in an effective and efficient approach to quality assurance in the DIHC laboratory. The great majority of this discussion concerns processing factors as they affect paraffin section-based immunohistology because that is still, by far, the most common adjunctive morphologic procedure (AMP) used in hospital pathology. However, information is also presented on elements of tissue processing that affect other analyses as well. The material presented here is a direct reflection of the conjoint experiences of the authors. Although we discuss our preferences regarding many technical aspects of this discipline, we certainly do not mean to imply that ours are the only methods that should be utilized. Nevertheless, it is hoped that they may serve to crystallize basic concepts surrounding fixation, processing, and optimal performance of AMPs.

Although it is classically a deep soft-tissue tumor of childhood, primitive neuroectodermal tumor (PNET) can occur at any age and may occasionally involve cutaneous sites. Merkel cell carcinoma (MCC) and basaloid neoplasms of cutaneous adnexa are the principal diagnostic alternatives to that tumor. The common expression of CD99 in PNET and cytokeratin-20 (CK20) in MCC suggests that these markers may be of value in this diagnostic setting, but they have not been rigorously examined in other small-cell and basaloid lesions of the skin. Accordingly, we evaluated CD99 and CK20 reactivity in formalin-fixed, paraffin-embedded sections of 30 MCC, five cutaneous metastases of pulmonary small-cell neuroendocrine carcinomas, 10 primary cutaneous adnexal carcinomas with basaloid features, 18 benign basaloid adnexal neoplasms of the skin (nine spiradenomas and nine cylindromas), and two cutaneous PNETs, using a standard immunohistologic technique and microwave-mediated epitope retrieval. Of the 30 MCC, 12 showed crisp membrane staining for CD99. Among the remaining tumors, only the two PNETs were positive for that marker. Although the majority of MCCs did not label for CD99, the pattern of reactivity in positive cases was indistinguishable from that observed in PNETs. Eighteen of 27 MCCs that were stained for CK20 were reactive for that protein, in contrast to metastatic small cell carcinomas, cutaneous PNETs, and appendageal skin tumors, which were uniformly negative for this marker. However, a subset of nine tumors, which were most consistent with MCC on clinical grounds, was CD99 positive and CK20 negative. Hence, reliance on CD99 alone as a diagnostic marker for PNET in this context cannot be recommended. Rather, careful assessment of the clinical presentation, together with extended immunophenotyping that includes other lineage markers and, when possible, cytogenetic analysis for characteristic chromosomal aberrations, remains the best means of separating MCC from PNET. Finally, the lack of CD99 reactivity in basaloid adnexal neoplasms of the skin suggests a utility in their differential diagnosis from cutaneous tumors with neuroendocrine or neuroectodermal differentiation.

Spindle cell proliferations of the skin are diverse, both morphologically and mechanistically. The authors have encountered four examples of a distinctive reactive/reparative cutaneous spindle cell lesion that shows homology with ones seen in the genitourinary tract and oral cavity and that is known as "postoperative/posttraumatic spindle cell nodule" (PSCN). These occurred in the skin of the face and scalp (2 cases), arm (1 case), and vulvar skin (one case), and were clearly related historically to prior episodes of trauma. The proliferations were characterized by variably-apposed and cytologically-bland spindle cells with numerous mitotic figures, set in a highly vascular stroma containing extravasated erythrocytes and inflammatory cells. All lesions were immunoreactive for vimentin, actin, and desmin, with no labeling for keratin. Postoperative/posttraumatic spindle cell nodule of the skin is a significant pseudoneoplastic lesion that is related (and perhaps identical pathogenetically) to nodular fasciitis; as such, it may be mistaken for a sarcoma or a spindle cell carcinoma. Careful attention to clinicopathologic and histologic details should result in its accurate recognition.

Four different combinations of monoclonal antibodies against rat ICAM-1, CD-11a, and CD-18 were utilized to determine the relative importance of LFA-1, Mac-1, and ICAM-1 in a rat model of severe lung allograft reperfusion injury. Negative control animals were given phosphate buffered saline (the carrier solution for the antibodies), while positive control animals were rendered neutropenic by the administration of a polyclonal mouse IgG. Antibodies were given with the donor lung flush, prior to left lung graft reperfusion, or both. Isolated graft function was determined 24 hr after implantation by arterial blood gas (ABG), and after sacrifice the native and transplanted lungs underwent bronchoalveolar lavage for alveolar protein quantitation, cell count and differential, and myeloperoxidase assay. Additionally, whole lung homogenates were assayed for myeloperoxidase activity. We found that the combination of anti-ICAM-1 (1 mg/kg) added to the donor lung flush, and anti-CD11a, anti-CD18, and anti-ICAM-1 (2 mg/kg i.v. of each) given to the recipient prior to reperfusion, resulted in significantly improved lung graft pAO2 by ABG, and decreased alveolar protein, cell count, and myeloperoxidase activity compared with control animals. Improvement was less than that seen in the neutropenic recipients, however. We conclude that LFA-1, Mac-1, and ICAM-1 are all important adhesion molecules in lung allograft reperfusion injury--yet even with antibody blockade of all three there are additional mechanisms allowing for neutrophil influx into the lungs.

Five cases of solitary syringoma (SS) were found by the authors in a search of adnexal neoplasms seen over an 18-year period. These tumors all occurred in adults as mobile, flesh-colored, circumscribed papules of the facial skin, measuring < or = 5 mm in greatest dimension. Each was cured by simple punch or shave biopsy excision. Comparison with nine histologically similar variants of microcystic adnexal carcinoma (MAC) showed that the latter lesions differed from SS in that they showed much deeper invasion, with regular involvement of the underlying subcutis and perineural or vascular-adventitial infiltration. Moreover, SS was wider than it was deep on microscopic examination, whereas the converse pertained to MAC. Knowledge of the typical clinicopathologic attributes of these two tumor types should prevent clinicians and pathologists from confusing them with each other.

The sclerotic fibroma (SF) is a tumor of the skin that may occur sporadically or in the context of Cowden's syndrome. The authors studied four examples of this tumor in an effort to understand better the nature of the fibrous matrix in SF and its significance. A novel antibody was utilized that is directed against the immunoreactive amino-terminal precuror domain of human type I procollagen (AP). This peptide is usually identified only at sites of active or recent collagen synthesis and deposition. All examples of SF showed strong cellular and stromal staining for AP. The matrix also demonstrated variable staining for type IV collagen and laminin. The overall immunophenotype of this lesion suggests that it may be a specialized fibroblastic ("dermal dendrocytic?") tumor. Despite its hypocellularity and hyalinized nature, SF is thought to exhibit ongoing matrical production that supports a neoplastic character for this lesion.

A 70-year-old Italian man with a history of squamous cell carcinoma of the lung presented with a nodular skin eruption. He had traveled extensively in India and Sri Lanka. The nodules were well demarcated and measured up to 3.5 cm in diameter. Histologically, there was a proliferation of spindled and polygonal cells with focal and relatively inconspicuous cytoplasmic vacuolation. A macrophage-monocyte lineage for the cells was confirmed by paraffin section immunohistochemistry, using the monoclonal antibodies anti-CD45, MAC-387, KP-1, UCHL-1, MT-1, L26, and MB2. Infiltrating borders, extension of the lesion into the subcutis, and involvement of small dermal nerves and eccrine glands initially suggested the possibility of a "histiocytic" neoplasm of indeterminate biological potential. However, air-dried and Giemsa-stained material from a fine-needle aspirate of one cutaneous nodule showed needle-shaped intracellular "negative images," and acid-fast stains revealed a large number of intracytoplasmic bacilli in virtually all of the vacuolated lesional cells. Furthermore, a second skin nodule that was excised 3 weeks after initial presentation showed the typical morphology of lepromatous leprosy. The clinicopathologic features of this case demonstrated several similarities with those of so-called "histoid" leprosy. Unusual morphologic variants of leprosy need to be considered in the interpretation of unusual "histiocytic" infiltrates in order to avoid a mistaken diagnosis of neoplasia, regardless of the geographic locale in which the patient is evaluated.

Objective: The deposition of submucosal fat in small and large bowel is known to occur in chronic processes such as inflammatory bowel disease, and has been assumed to reflect long disease duration. However, we have observed that submucosal fat deposition can occur in a short period of time., Materials and Methods: Four patients were evaluated in whom serial CT examinations showed intramural fat deposition occurring within six months of a normal study. Confirmation of fatty deposition was made by Hounsfield unit measurement in all patients and by pathologic examination of bowel specimens in two patients., Results: All four patients received cytoreductive chemotherapy for treatment of lymphoma or leukemia prior to the development of fat deposition. In all patients, serial CT examination documented the occurrence of fat-attenuation bowel wall thickening that developed in a short time period (12, 36, 67, and 186 days). Three of the four cases were initially misinterpreted as wall thickening from other causes (intussusception or colitis). Pathologic examination of bowel in the two patients with the shortest time intervals confirmed the presence of mature fat confined to the submucosa., Conclusion: These cases demonstrate that submucosal fat deposition in bowel wall is not limited to inflammatory bowel disease and other longstanding, chronic diseases; fatty infiltration can occur in a relatively short period of time, and is particularly likely to occur after cytoreductive therapy.

Pilomatrixoma, a neoplasm of hair germ matrix origin, is one of the most common cutaneous appendage tumors in patients 20 years of age or younger. Our review of cutaneous adnexal tumors in the first 2 decades of life showed that pilomatric lesions accounted for approximately 75% of cases. We examined 76 tumors from 69 patients between the ages of 8 months and 19 years (average age, 8 years). The female-to-male ratio was 3:1. We noted a predilection for the head and neck region, as has also been reported by other investigators. Sixty-nine tumors were examples of typical or classic pilomatrixoma; four of them showed the unusual histologic feature of transepidermal elimination or perforation. None of the tumors in the typical category recurred during a follow-up period that averaged 8.7 years. Seven of the lesions with atypical histologic features were examples of aggressive pilomatrixoma (three cases) and pilomatrical carcinoma (four cases). The average age of these seven patients was similar to that of the overall group. The pilomatrical carcinomas were characterized by invasive nests of tumor cells with irregular borders, large vesicular nuclei, prominent nucleoli, multiple mitotic figures, and focal necrosis. A desmoplastic stroma surrounded the infiltrating nests of tumor. Two of the four pilomatrical carcinomas recurred 2 and 6 months, respectively, after initial excision. The aggressive pilomatrixomas had a similar infiltrative growth pattern, but they did not exhibit the same degree of worrisome cytologic features. Single cell necrosis, identifiable mitotic figures, and prominent nucleoli differentiated the atypical from the typical pilomatrixomas. One of the three aggressive pilomatrixomas recurred after 2 months.(ABSTRACT TRUNCATED AT 250 WORDS)

Clinical and light-microscopic findings were reviewed in 60 benign histiocytic tumors of the skin arising in patients who were 19 years of age or younger. These lesions were placed in five categories, based on the presence or absence of Touton giant cells, the relative content of vacuolated polygonal cells and spindle cells, and the extent of stromal collagen in the proliferations. The resulting distribution included 16 cases of classic juvenile xanthogranuloma (JXG), four examples of xanthomatous JXG, 11 "transitional" JXG, 11 histiocytomas, and 18 dermatofibromas. The histological appearances of the lesions were related to the elapsed time between their clinical presentation and biopsy sampling. Intervals of 3.5, 3.5, 3.6, 7, and 13.4 months were noted for each histologic group, respectively. Hence, "early" lesions (JXG variants) showed no differences in clinical duration, but significantly longer evolutions were observed for histiocytomas and dermatofibromas. Immunohistochemical studies were conducted utilizing antibodies to several "histiocytic" markers (CD68, cathepsin B, MAC387), CD45 antigen, and S-100 protein. Cathepsin B was universally expressed by all tumors, but its intensity and scope were variable. In addition, only dermatofibromas failed to show CD68 positivity. Scattered mononuclear inflammatory cells in the lesions were CD45 reactive, whereas other cell types were not so labeled. MAC387 was absent in all cases. These findings suggest that these lesions do indeed show histiocytic differentiation, and that they are probably related to peripheral blood monocytes antigenically. Differences in the histologic appearances of these cutaneous histiocytic proliferations of childhood may simply reflect dissimilar periods of clinical growth before biopsies are performed.

Kaposi's sarcoma has achieved considerable notoriety in the past decade because of its association with the acquired immunodeficiency syndrome (AIDS). Nonetheless, this neoplasm continues to affect patients who do not have AIDS, and analysis of such cases may provide information that is useful in understanding the pathogenesis and treatment of Kaposi's sarcoma. Recent data indicate that non-AIDS-related Kaposi's sarcoma shows a predilection for individuals with the HLA-DR5 haplotype and that this genetic locus may affect immune function. In addition, other information suggests that infection with cytomegalovirus may represent the "trigger" for oncogenesis in patients with Kaposi's sarcoma. In view of the latter finding, therapies with such antiviral agents as zidovudine may be indicated in cases of non-AIDS-related Kaposi's sarcoma.

Adnexal carcinomas of the skin are rare lesions that encompass multiple histologic patterns. The authors have identified seven neoplasms in this category that showed combined morphologic features of at least two appendegeal tissues. Two manifested pilar and sebaceous differentiation, three showed pilar, sebaceous, and sudoriferous features, and two had sweat glandular and sebaceous characteristics. All tumors arose in the skin of the head and neck; four patients were men and three were women (age range, 21-84 years). One lesion had been present for 25 years with progressive enlargement over the last 3, and another two had slowly expanded over 1-2 years; the duration of growth for the remaining cases was uncertain. Follow-up revealed that six patients were free of disease for at least 1 year after complete excision. In the seventh case, complete excision was performed, but follow-up was not available. Immunohistochemical analysis showed reactivity for keratin and epithelial membrane antigen in each case. Carcinoembryonic antigen was seen in four tumors, and amylase was present in three; none expressed vimentin, S100 protein, or gross cystic disease fluid protein-15. It is concluded that mixed cutaneous adnexal carcinomas are recognizable entities that appear to have little potential for aggressive growth. Immunohistochemical studies may serve as adjuncts to conventional microscopy in the definition of cellular lineages pursued by such neoplasms.

Neuroendocrine differentiation in basal cell carcinoma (BCC) of the skin has been reported in the past on the basis of ultrastructural findings, argyrophilia, and the immunohistochemical detection of neuropeptides in such neoplasms. To assess further the relative frequency of neuroendocrine differentiation in BCC, paraffin sections of 53 randomly chosen cases were evaluated for Churukian-Schenk stain positivity and the expression of sensitive neuroendocrine markers. These included neuron-specific enolase (using a highly absorbed monospecific antiserum), chromogranin-A, synaptophysin, neurofilament protein, and Leu-7 antigen. Although 25% of cases demonstrated argyrophilia with the Churukian-Schenk method, suggesting a high frequency of possible neuroendocrine differentiation, immunohistochemical evidence of the same was observed in only two tumors (4%). These demonstrated immunoreactivity for chromogranin and neuron-specific enolase. The results of this analysis suggest that neuroendocrine differentiation in BCC is relatively uncommon, and that it is not reliably predicted by the results of argyrophil stains done on paraffin sections.

We herein report two cases of sweat gland carcinoma that arose in association with eccrine spiradenoma. These lesions presented as enlarging masses that previously had been stable for many years. One produced widespread metastasis and death 5 months after diagnosis. Immunohistochemical studies demonstrated an antigenic relationship between the benign and malignant components of sweat gland carcinoma ex eccrine spiradenoma, but ultrastructural analyses showed a paucity of specialized differentiation. This neoplasm appears to display a range of microscopic appearances and has proven fatal in 20% of reported cases.

Atypical fibroxanthoma belongs to the family of spindle-cell and pleomorphic neoplasms of the skin. The lineage of differentiation of this tumor and the means whereby it can be diagnostically separated from other similar morphologic entities have been controversial. In order to address these issues, the authors studied 30 spindle-cell and/or pleomorphic cutaneous tumors, including atypical fibroxanthomas (AFXs), superficial malignant fibrous histiocytomas (MFHs), dermatofibrosarcoma protuberans (DFSPs), sarcomatoid squamous-cell carcinomas (SCCs), spindle-cell malignant melanomas (MMs), and leiomyosarcomas, (LMSs). These cases were analyzed using a panel of eight antibodies and the immunoperoxidase technique. AFXs were reactive for vimentin, alpha-1-antichymotrypsin (AACT), alpha-1-antitrypsin (AAT), and cathepsin-B (CB) but failed to display cytokeratin (CK), epithelial membrane antigen (EMA), S-100 protein, and desmin. MFHs and DFSPs exhibited immunophenotypic profiles that were nearly identical to that just described. In contrast, SCCs were typified by positivity for CK and EMA; MMs exhibited uniform reactivity for S-100 protein; and LMSs contained desmin in four of five cases. A surprising result was the expression of S-100 by LMSs. Also, all tumors displayed at least one of the three proteolytic enzymes assessed in this study (AAT, AACT, and CB), demonstrating the relative diagnostic nonspecificity of these determinants. It is concluded that AFXs are "fibrohistiocytic" neoplasms, with substantial morphologic and immunohistochemical similarity to MFHs. The immunohistochemical classification of spindle-cell and pleomorphic tumors of the skin necessitates the use of antibody panels to assess the presence of markers that are characteristic of each diagnostic group.

The second-look laparotomy is the procedure of choice to obtain peritoneal and nodal biopsies from patients who have received a full course of chemotherapy for an ovarian malignancy. If there is pathologic evidence of persistent tumor, chemotherapy is continued. In most cases, the histopathologic interpretation is straightforward in terms of positive or negative results. However, atypical mesothelial reactions or glandular inclusions, or both, in lymph nodes are potential and real difficulties in differential diagnosis. This study evaluated a panel of immunohistochemical markers, namely carcinoembryonic antigen (CEA), Leu M1, human milk fat globule protein, cytokeratin, epithelial membrane antigen, and LN2 antibody for their usefulness in differentiating benign epithelium and mesothelium from carcinoma. Only CEA and Leu M1 were specific and sensitive, respectively, for malignant tumors in this study.

We report two cases of cutaneous malignant peripheral nerve sheath tumors. One occurred in a patient without neurofibromatosis and manifested as a recurrent scalp nodule. Histologically, the lesion began as a benign-appearing, dermal neurofibroma and progressed to a highly cellular, pleomorphic neoplasm with prominent mitotic activity and invasion of the subcutis. Immunohistochemical and ultrastructural examination confirmed the nerve sheath origin of this tumor. The other case presented as a large scalp mass in a patient with an established diagnosis of neurofibromatosis; its immunohistochemical characteristics were identical to those of case 1. Although malignant peripheral nerve sheath tumors are uncommon in the skin, they can be effectively separated from other spindle cell neoplasms by specialized pathologic studies.

An adenoid cystic carcinoma of the skin was compared with three similar neoplasms of salivary glands and with an adenoid basal cell carcinoma, from unrelated cases. The histological and immunocytochemical details of these tumors were analyzed in an attempt to determine whether or not their differing clinical behaviors would be reflected in pathologic dissimilarities. Although the single adenoid cystic carcinoma of the skin did not recur or metastasize over a 10-year follow-up period, its morphologic and biochemical features were identical to those of biologically aggressive salivary gland tumors. All four adenoid cystic carcinomas contained carcinoembryonic antigen, epithelial membrane antigen, salivary-type amylase, and alpha-lactalbumin, and all bound peanut agglutinin. Three of four expressed positivity for S100 protein, and two contained low-molecular-weight cytokeratin. In contrast, none was immunoreactive for beta-2-microglobulin, and only one displayed blood group isoantigen positivity. The adenoid basal cell carcinoma was negative for all immunological determinants, but it bound peanut agglutinin. Although these results should be regarded as preliminary, it appears that adenoid cystic carcinoma is a pathologically distinct and uniform entity, whether it occurs in the skin or in salivary glands. However, the clinical behavior of this tumor cannot be predicted on the basis of immunohistochemical or morphological studies. Finally, adenoid basal cell carcinoma is histopathologically and immunocytochemically separable from cutaneous adenoid cystic carcinoma.

The eruption of multiple lobular capillary hemangiomas (granulomata pyogenicum) is a previously documented but uncommon phenomenon. Reported cases have featured the recurrence of multiple satellite lesions after surgical removal of, or trauma to, a solitary capillary hemangioma. Sporadic cases of multifocal lesions arising de novo in both previously traumatized and uninjured skin have been described as well. We describe herein two unusual patients with disseminated lobular capillary hemangioma. One patient developed skin nodules in several sites, including the mouth, knee, thumb, and foot, after surgical removal of a solitary oculocutaneous neoplasm. The second patient had more than 700 skin lesions, distributed over the entire body. Unlike those in most previously reported examples of this condition, these two patients were elderly. Both were misdiagnosed clinically as having malignant vascular neoplasms, but the behavior of the lesions was benign, despite conservative surgical excisions in one patient and relative lack of therapeutic intervention in the other.

Although leukocyte common antigen is a valuable discriminant between lymphoreticular proliferations and other histologically similar neoplasms, several publications have described difficulties in adequately labeling this determinant in formalin-fixed, paraffin-embedded tissues. We studied 662 paraffinized cutaneous and noncutaneous neoplasms for the presence of leukocyte common antigen, using an extended avidin-biotin-peroxidase complex (ABC) method, a combined ABC-peroxidase-antiperoxidase technique, and a biotin-strept-avidin procedure. Each of the three immunohistochemical techniques produced effective results in the detection of leukocyte common antigen in all malignant lymphomas, but none of them produced false-positive staining in nonhematopoietic lesions. These results suggest that antibodies to leukocyte common antigen are highly specific and that the methods employed in this study allow for excellent diagnostic sensitivity in the recognition of cutaneous lymphomas.

Primary papillary serous carcinoma arising from the peritoneal surface (serous surface papillary carcinoma; SSPC) is a distinctive neoplasm with a histomorphologic resemblance to serous ovarian papillary carcinoma (SOPC). To determine if these tumors are similar antigenically, we studied 13 examples of SSPC and 31 of SOPC immunohistochemically. Antibodies to several determinants known to occur in the Müllerian epithelium were employed. Both neoplasms were uniformly positive for cytokeratin and epithelial membrane antigen (EMA); in addition, SSPC and SOPC were similar in quantitative and qualitative reactivity for B72.3 antigen, carcinoembryonic antigen, Leu M1, CA-125 antigen, LN1, LN2, MB2, S100 protein, placental alkaline phosphatase, and amylase. Residual nonneoplastic mesothelium failed to express any of these antigens except for cytokeratin, EMA, and CA-125. The clinical behavior of SSPC was similar to that of high-stage SOPC; all patients with adequate follow-up died of their tumors. These results suggest that SSPC and SOPC are analogous lesions, with respect to their cellular differentiation. Moreover, it would appear that both neoplasms display only a limited immunophenotypic homology to the mesothelium.