Desmoplastic malignant melanoma. An immunocytochemical study of 25 cases

Primary Cutaneous Desmoplastic Melanoma With Collagen Rosettes and Pseudoglandular Features

ABSTRACT

Primary cutaneous desmoplastic melanoma (DM) is a group of rare melanocytic tumors arising on severely sun-damaged skin, histologically characterized by the proliferation of spindled melanocytes in a prominent desmoplastic stroma, with a range of morphological presentations. In this article, we report a unique case of primary cutaneous DM composed of a nodular proliferation of highly pleomorphic spindled and epithelioid cells, pseudoglandular structures, clear cell change, and unusual collagen rosettes. Immunohistochemical analysis showed a strong and diffuse positivity for S-100 protein, SOX-10, nestin, p75 (nerve growth factor receptor), WT1, and p53. Molecular analysis detected a mutation in the NF1 gene [c.4084C > T, p.(Arg1362Ter)], 2 different pathogenic mutations in TP53 [c.742C > T, p.(Arg248Trp), AF:12%, COSM1640831 and c.528C > G, p.(Cys176Trp), AF:12%, COSM11114], and a mutation in GNAS [c.601C > T, p.(Arg201Cys), AF: 9%, COSM123397]. To the best of our knowledge, this is the first case reporting collagen rosettes and pseudoglandular features in primary cutaneous DM.

PDGFB Rearrangements in Dermatofibrosarcoma Protuberans of the Vulva: A Study of 11 Cases Including Myxoid and Fibrosarcomatous Variants

Dermatofibrosarcoma protuberans (DFSP) is a low-grade fibroblastic sarcoma that tends to arise in young to middle age adults and involve the trunk and proximal extremities. Rare examples of vulvar DFSP have been reported, including myxoid, myoid, and fibrosarcomatous variants, but detection of the characteristic t(17;22)(q22;q13) that produces COL1A1-PDGFB gene fusion has not been evaluated in a large series of primary vulvar tumors. The clinical, morphologic, immunohistochemical, and molecular cytogenetic features of 11 cases were examined. Patient age ranged from 29 to 75 yr (mean, 46 yr; median, 43 yr). Seven tumors were purely classic DFSP, 1 was purely myxoid DFSP and the remaining 3 had varying quantities of fibrosarcomatous DFSP. All cases of classic DFSP had diffuse expression of CD34 and low-level p53 immunoreactivity. Myxoid variants had strong, but reduced expression of CD34. Fibrosarcomatous DFSP showed focal CD34 expression and increased p53 reactivity. Nine of 11 tumors (82%) had rearrangement of PDGFB by fluorescence in situ hybridization. The 2 nonrearranged tumors were a classic DFSP and a myxoid DFSP with fibrosarcomatous transformation. Follow-up was available for 9 patients (82%) and ranged from 1 to 108 mo (mean, 30 mo; median, 21 mo). Eight patients had tumors with positive margins, one of which developed local recurrence after no further therapy. No patient developed metastasis. The high frequency of PDGFB rearrangement in vulvar DFSP provides a useful exploit in diagnostically challenging cases and genetic evidence of probable clinical response to targeted therapeutics in cases of locally advanced or metastatic tumors.

Desmoplastic Melanoma of the Head and Neck: Histopathologic and Immunohistochemical Study of 28 Cases

Studies of the immunohistochemical profiles and clinical course of desmoplastic melanoma have produced conflicting results. We identified 28 cases of desmoplastic melanoma after a search of our files for spindle cell neoplasms of the head and neck from 1960 through 1995. The 17 male (61%) and 11 female (39%) patients averaged 65 years of age. The cheek was the most common location (12 cases, 43%). The average length of follow-up was 5 years. Overall 5-year survival rate was 46%. Melan A and tyrosinase positivity (P = 0.0195), smooth muscle actin positivity (P = 0.0328), tumor size (P = 0.0297), and tumor thickness (P = 0.0419) were significantly associated with local progression-free survival. No histologic or immunohistochemical marker was associated with overall or metastasis-free survival.

Diagnostic Immunohistochemistry in Cutaneous Neoplasia: An Update

Immunohistochemistry (IHC) is an important adjunct in the diagnosis of neoplastic skin diseases. In addition to the many established IHC markers currently in use, new markers continue to emerge, although their general acceptance and routine application requires robust validation. Here, we summarize the most well-established and commonly used biomarkers along with an array of newer ones reported in the past several decades that either demonstrate or hold high clinical promise in the field of cutaneous pathology. We also highlight recent applications of novel IHC markers in melanoma diagnosis including genetic mutation status markers [e.g. BRAF (v-raf murine sarcoma viral oncogene homolog B) and NRAS (neuroblastoma RAS viral oncogene homolog)] and an epigenetic alteration marker (e.g. 5-hydroxymethylcytosine). We specifically focus on the role of IHC in the differential diagnosis of cutaneous lesions that fall under the following categories: melanoma, epidermal tumors with an intraepidermal epitheliomatous pattern, spindle cell lesions of the dermis, small round blue cell tumors of the dermis, and cutaneous adnexal tumors. While IHC is a valuable tool in diagnostic dermatopathology, marker selection and interpretation must be highly informed by clinical context and the histologic differential diagnosis. With rapid progress in our understanding of the genetic and epigenetic mechanisms of tumorigenesis, new IHC markers will continue to emerge in the field of diagnostic dermatopathology.

Value of melanocytic-associated immunohistochemical markers in the diagnosis of malignant melanoma: a review and update

Since the identification of S100 protein as an immunohistochemical marker that could be useful in the diagnosis of melanoma in the early 1980s, a large number of other melanocytic-associated markers that could potentially be used to assist in the differential diagnosis of these tumors have also been investigated. A great variation exists, however, among these markers, not only in their expression in some subtypes of melanoma, particularly desmoplastic melanoma, but also in their specificity because some of them can also be expressed in nonmelanocytic neoplasms, including various types of soft tissue tumors and carcinomas. This article reviews the information that is currently available on the practical value of some of the markers that have more often been recommended for assisting in the diagnosis of melanomas, including those that have only recently become available.

Mohs surgery for melanoma: rationale, advances and possibilities

Mohs surgery (MS) is an effective technique for the removal of a variety of cutaneous neoplasms by virtue of its thorough assessment of margins. It has yet to become widely accepted for melanoma because recognizing melanocytes histologically in frozen section can be problematic. Recently, 'rapid' methods of immunohistochemistry have been developed that resolve this issue by staining the melanocytes in frozen section. In addition, some of the immunohistochemistry protocols that previously required up to 1 h now take 19 min or less. These technological enhancements for MS have removed some of the obstacles towards the acceptance of MS as a legitimate option for removal of melanomas, especially poorly demarcated lesions and lesions from the head and neck, the distal extremities and the genitalia. Experience thus far with MS for melanoma has shown lower recurrence rates and improved disease-specific survival compared with historical controls for standard excision, while at the same time minimizing the sacrifice of normal tissue.

Blue Nevi and Variants: An Update

Context.—Blue nevi are a subset of melanocytic proliferations containing cells reminiscent of the embryonal neural crest–derived dendritic melanocytic precursors. They are common specimens in a general pathology practice, but some of their rare variants may pose diagnostic difficulty. Recent molecular studies provide new insights into genetics of blue nevi.

Objective.—To critically review clinical and histologic features of blue nevi with emphasis on diagnostic problems and rare variants, as well as to provide an update on the pathogenesis of blue nevi.

Data Sources.—Published peer-reviewed literature and personal experience of the authors.

Conclusions.—Challenging areas in diagnosis of blue nevi include recognition of amelanotic, desmoplastic, atypical, and malignant variants of blue nevus. Recent data show that mutations in genes responsible for common nevi or melanomas such as BRAF, NRAS, or c-kit are rare in blue nevi. Benign and malignant blue nevi harbor frequent mutations in the Gαq class of G-protein α subunits, Gnaq and Gna11 proteins.

P75 nerve growth factor receptor staining is superior to S100 in identifying spindle cell and desmoplastic melanoma

BACKGROUND

Spindle cell melanoma (SCM) including desmoplastic melanoma (DM) is a rare variant of malignant melanoma that may present diagnostic difficulties particularly when staining with S100 is negative, weak, focal, or a combination of these. Conventional melanocytic markers in SCM are usually negative.

OBJECTIVE

We sought to compare the staining of p75 nerve growth factor receptor (NGF-R) and S100 in SCMs.

METHODS

We evaluated the staining of p75 NGF-R and S100 in 13 cases of SCMs: 3 SCMs without desmoplasia, 5 pure DMs, and 5 combined DMs with a conventional component.

RESULTS

Staining with p75 NGF-R was positive in 13 of 13 (100%) cases of SCMs. In 3 cases the intensity of staining and the percentage of cells staining with this marker were greater than those with S100. One case of SCM was negative for S100 but demonstrated strong expression of p75 NGF-R. One case was focally and weakly positive with S100 but expressed strong positive staining with p75 NGF-R. Absence of staining with p75 NGF-R was noted in the conventional round cell component of two of 5 (40%) combined DMs whereas the same areas were strongly positive for human melanoma black (HMB)-45 and Melan-A. In 5 of 5 (100%) cases of combined DMs the desmoplastic component stained positive with p75 NGF-R, demonstrating an inverse relationship with the staining of conventional melanocytic markers.

LIMITATIONS

Small study size was a limitation.

CONCLUSION

p75 NGF-R exhibits superior staining characteristics and greater sensitivity in identifying SCM and DMs than S100. P75 NGF-R may be a useful diagnostic and ancillary stain in addition to S100.

Immunostaining in Mohs micrographic surgery: a review

BACKGROUND

With the advent of incorporating the immunoperoxidase staining technique into the processing of frozen tissue, the use of Mohs micrographic surgery (MMS) has been expanded to include several high-risk tumors such as lentigo maligna, malignant melanoma, and dermatofibrosarcoma protuberans.

OBJECTIVES

To thoroughly review the English medical literature pertaining to the use of immunohistochemical staining techniques on frozen sections during MMS and to summarize the basic relevant outcomes from the different relevant studies.

MATERIALS AND METHODS

Medline search was conducted, with the following words used in the search criteria: "Mohs surgery,""staining,""immunostaining," and "immunoperoxidase." RESULTS Generally, all immunostains showed advantage over the traditional hematoxylin and eosin approach. Studies of MART-1 in melanoma chemosurgery indicated that it is typically crisp and has less background staining than MEL-5 and better staining consistency than HMB-45. In cases of desmoplastic melanomas, S100 is the stain of choice.

CONCLUSION

Immunostaining offers an advantage in MMS. Large, randomized, prospective studies comparing the different immunostains are still lacking in the literature. The authors have indicated no significant interest with commercial supporters.

KBA.62: a useful marker for primary and metastatic melanomas

We previously described a novel antimelanoma antibody, designated KBA.62. However, review of the literature showed that only a few studies have reported on this antibody. We report our experience in the diagnosis of melanoma using KBA.62 and its value in both primary and metastatic conditions. In addition to conventional melanomas, we included desmoplastic and spindle cell melanomas, whose diagnosis can be challenging. We also focus on identification of malignant cells in sentinel lymph node biopsies using KBA.62, by comparison with anti-S-100 protein and HMB-45 antibodies, because discrepancies in sensitivity and specificity of melanoma markers have given rise to numerous studies aiming to determine the best panel for the evaluation of sentinel lymph node from patients with melanoma. Overall, KBA.62 was positive in 93% of primary and metastatic melanomas. Interestingly, the 12 cases of desmoplastic and spindle cell melanomas had strongly positive results. In addition, the results of our study performed on a series of 215 sentinel lymph nodes showed that the sensitivity of anti-S-100 protein and KBA.62 antibodies in detecting occult metastasis was similar. Moreover, KBA.62 identified 6 patients (3%) who had confirmed sentinel lymph node metastasis but were negative for HMB-45. The resolution was higher with KBA.62 than that observed with anti-S-100 protein antibody, as the nonmelanocytic positive cells for KBA.62 in lymph nodes were only represented by endothelial cells, which therefore constituted an intrinsic positive control. We conclude that KBA.62 antibody is a useful additional marker for melanoma, specifically in desmoplastic/spindle cell cases and in the context of micrometastasis in sentinel lymph node.

Divergent differentiation in malignant melanomas: a review

The aim of this review was to document and discuss diagnostic problems associated with divergent differentiation ('metaplastic change') in malignant melanomas, defined as the development in these tumours of morphologically, immunohistochemically and/or ultrastructurally recognizable non-melanocytic cell or tissue components. Types of divergent differentiation reported in malignant melanoma include: fibroblastic/myofibroblastic, Schwannian and perineurial, smooth muscle, rhabdomyosarcomatous, osteocartilaginous, ganglionic and ganglioneuroblastic, neuroendocrine and probable epithelial. Divergent differentiation is certainly a rare phenomenon and, when it occurs, can be missed by unwary pathologists and lead to diagnostic uncertainty. A carefully chosen immunohistochemical panel and the input of electron microscopy can help to clarify the nature of the cellular differentiation of these tumours and lead to a correct final diagnosis. The clinical significance of such aberrations is uncertain, nor are the underlying mechanisms as yet well defined.

Selected Diagnostic Problems in Neoplastic Dermatopathology

Context.—Selected cutaneous neoplasms share features with benign counterparts or have subtle morphologic features that could be overlooked by the pathologist.

Objective.—To present clues to the diagnosis of potentially deceptive malignancies, including desmoplastic malignant melanoma, nevoid malignant melanoma, subcutaneous lymphoma, metastatic breast carcinoma, and epithelioid sarcoma.

Data Sources.—Published literature and personal experience.

Conclusions.—Knowledge of commonly misdiagnosed cutaneous neoplasms will help the general surgical pathologist avoid these potential pitfalls in neoplastic dermatopathology.

Malignant melanoma in the 21st century, part 1: epidemiology, risk factors, screening, prevention, and diagnosis

Malignant melanoma is an aggressive, therapy-resistant malignancy of melanocytes. The incidence of melanoma has been steadily increasing worldwide, resulting in an increasing public health problem. Exposure to solar UV radiation, fair skin, dysplastic nevi syndrome, and a family history of melanoma are major risk factors for melanoma development. The interactions between genetic and environmental risk factors that promote melanomagenesis are currently the subject of ongoing research. Avoidance of UV radiation and surveillance of high-risk patients have the potential to reduce the population burden of melanoma. Biopsies of the primary tumor and sampling of draining lymph nodes are required for optimal diagnosis and staging. Several clinically relevant pathologic subtypes have been identified and need to be recognized. Therapy for early disease is predominantly surgical, with a minor benefit noted with the use of adjuvant therapy. Management of systemic melanoma is a challenge because of a paucity of active treatment modalities. In the first part of this 2-part review, we discuss epidemiology, risk factors, screening, prevention, and diagnosis of malignant melanoma. Part 2 (which will appear in the April 2007 issue) will review melanoma staging, prognosis, and treatment.

NGFR-positive desmoplastic melanomas with focal or absent S-100 staining: Further evidence supporting the use of both NGFR and S-100 as a primary immunohistochemical panel for the diagnosis of desmoplastic melanomas

The histologic diagnosis of desmoplastic melanoma can be challenging in circumstances in which biopsy specimens are small, or there are unsuspected clinical settings, re-excision scars, and unusual sites. This is particularly true when the overlying junctional component is absent or the spindle cells lack melanin pigment. In these instances, the importance of immunohistochemistry cannot be overemphasized. S-100 protein is the primary immunohistochemical stain used for this purpose, with a sensitivity approaching 90%. HMB-45, although a more specific marker for melanocytes, is less sensitive and often negative in these cases. In addition, NGFR, a marker of Schwannian differentiation, has been shown to be a useful confirmatory stain for desmoplastic melanoma, with staining intensity comparable with, or better than that of S-100 protein. We report 2 cases of desmoplastic melanomas that stained only focally and weakly with S-100 protein, but showed diffuse and intense staining with NGFR. In both cases, S-100 staining could have been interpreted as non-confirmatory, thus misguiding the diagnosis. We suggest that NGFR can be a useful complementary marker to S-100 in those desmoplastic melanomas in which staining for S100 protein is only focal or weak.

Unusual variants of malignant melanoma

A potential diagnostic pitfall in the histologic assessment of melanoma is the inability to recognize unusual melanoma variants. Of these, the more treacherous examples include the desmoplastic melanoma, the nevoid melanoma, the so-called 'minimal-deviation melanoma,' melanoma with prominent pigment synthesis or 'animal-type melanoma,' and the malignant blue nevus. Also problematic are the unusual phenotypic profiles seen in vertical growth phase melanomas; these include those tumors whose morphological peculiarities mimic cancers of nonmelanocytic lineage and those melanomas that express aberrant antigenic profiles not commonly associated with a melanocytic histogenesis. Metaplastic change in melanoma, balloon cell melanoma, signet-ring cell melanoma, myxoid melanoma, small cell melanoma and rhabdoid melanoma all have the potential to mimic metastatic and primary neoplasms of different lineage derivations. Abnormal immunohistochemical expression of CD 34, cytokeratins, epithelial membrane antigen, and smooth muscle markers as well as the deficient expression of S100 protein and melanocyte lineage-specific markers such as GP100 protein (ie HMB-45 antibody) and A103 (ie Melan-A) also present confusing diagnostic challenges. In this review, we will discuss in some detail certain of these novel clinicopathologic types of melanoma, as well as the abnormal phenotypic expressions seen in vertical growth phase melanoma.

Unusual histological variants of cutaneous malignant melanoma with some clinical and possible prognostic correlations

Malignant melanoma is known for the wide range of histological patterns it can assume mimicking other malignant tumors. We present a review of most of the unusual histological variants of cutaneous melanoma and describe their immunohistochemical features, associate clinical findings, and possible behavior related to the histological subtype. In addition, we propose their classification into four groups corresponding to the (1) architectural patterns; (2) cytologic features; (3) stromal changes; and (4) the possible association of these findings (i.e. architectural + cytologic features). Although most of these unusual variants have the same prognosis as conventional melanomas, with Breslow thickness and ulceration, being the most important predictor of survival in clinical stage I, some of them have a peculiar biologic behavior that the clinicians and the dermatopathologists should know in order to give melanoma patients all educational information available.

Neurotropic melanoma invading the inferior alveolar nerve

BACKGROUND

Desmoplastic neurotropic melanoma (DNM) and neurotropic melanoma (NM), rare lesions of the head and neck, often present as a benign-appearing nodule which later progresses to cranial nerve involvement.

OBJECTIVE

To discuss treatment and outcome of 3 cases of DNM/NM of the lower lip.

METHODS

Three case reports with 12- to 54-month follow-up and literature review.

RESULTS

The first case had an initial excision of melanoma of his lower lip and presented 16 years later with severe bilateral atrophy of his muscles of mastication with NM and underwent intensity modulated radiation therapy but no further resection. We also report two other cases of DNM of the lip, one of which received gamma knife radiosurgery for intracranial extension.

LIMITATIONS

Small retrospective case series.

CONCLUSIONS

Locoregional recurrences from DNM are more common than distant metastasis warranting vigilant surveillance following resection of the initial lesion. Radiation therapy and gamma knife radiosurgery may be used for cases of unresectable recurrences.

New Phenotypical and Ultrastructural Findings in Spindle Cell (Desmoplastic/Neurotropic) Melanoma

Spindle cell melanoma is a rare and distinctive variant of malignant melanoma that is composed of spindled neoplastic cells and includes desmoplastic and neurotropic melanoma. The lack of expression of several melanoma markers may result in a delayed or wrong diagnosis. In this study, we have analyzed in detail the phenotype of the tumor cells in 9 spindle cell melanomas on both paraffin-embedded and frozen material, using melanocytic, neural, and mesenchymal markers. The neoplastic cells expressed the melanocytic markers S-100, Mel-CAM, and NKIC3, but lacked gp100 and Melan-A; tyrosinase and c-Kit were expressed in 2 of 7 cases. Most cases expressed the neural markers p75-nerve growth factor receptor, neural cell adhesion molecule, and NSE. All cases expressed vimentin but lacked the mesenchymal markers CD34 and alpha-smooth muscle actin. Remarkably, all spindle cell melanomas strongly and diffusely expressed the fibroblastic markers Thy1 (CD90) and aminopeptidase N (CD13) and variably expressed the enzyme prolyl-4-hydroxylase, involved in procollagen formation. The coexpression of melanocytic, neural, and fibroblastic markers suggests bidirectional differentiation of neoplastic melanocytes toward (myo)fibroblasts and Schwann cells, a feature that was confirmed by electron microscopy. Furthermore, the lack of CD90 and CD13 staining in a wide range of melanocytic lesions suggests specificity of these markers for spindle cell melanoma.

Efficacy of radiation therapy in the local control of desmoplastic malignant melanoma

BACKGROUND

Desmoplastic malignant melanoma (DMM) is a rare variant of malignant melanoma with high local recurrence rate after surgical excision. We performed a retrospective review to address the role of radiation therapy in local control of this tumor.

METHODS

Between 1976 and 1997, 44 patients with the pathologic diagnosis of DMM were registered at our tumor registry. Fourteen patients received postoperative RT, and one patient received preoperative RT. Three of the irradiated lesions had gross residual or positive surgical margins. Doses ranged from 44 to 66 Gy.

RESULTS

Sixty-eight percent of DMM lesions occurred in the head and neck region. Forty-eight percent (21 of 44) of patients experienced a local recurrence after initial excision (mean time to recurrence, 12 months). Local failure in head and neck was 46% (14 of 30). Clark level, primary site, and neurotropism did not predict local recurrence; the Clark level predicted distant metastasis. No viable tumor was found in the surgical specimen of the patient who received preoperative RT. None of 15 patients who received adjuvant irradiation had any additional recurrences (mean follow-up, 64.7 months). By contrast, four of seven patients with history of recurrence who did not receive RT had local relapse (p =.005). The incidence of distant metastasis did not reach statistical significance between the irradiated and nonirradiated groups.

CONCLUSIONS

The high rate of local recurrence of DMM after surgical resection is dramatically reduced by adjuvant radiation therapy. We recommend adjuvant postoperative radiation therapy as a part of treatment of DMM.

Desmoplastic neurotropic melanoma of the head and neck: the role of radiation therapy

BACKGROUND

Desmoplastic neurotropic melanoma (DNM) is an uncommon cutaneous melanoma variant with pronounced neurotropism. In contrast to ordinary melanomas, locoregional recurrences are common and distant metastasis are uncommon in patients with DNM. Local control with surgery and radiation therapy may assume a more important role in this variant of melanoma. We present a case of an unresectable skull base recurrence of DNM that was controlled using radiation therapy alone and review the literature.

METHODS

Case report with 36-month follow-up.

RESULTS

The patient is a 68-year-old with multiple recurrences of a DNM that originated on the forehead. After extensive surgery, including total parotidectomy and temporal bone resection, the patient had an unresectable recurrence of the skull base develop. This was treated with definitive radiation therapy, resulting in a complete response. The patient has had no evidence of recurrence at 3 years.

CONCLUSIONS

DNM is a locally aggressive type of melanoma with a high risk of local recurrence that can be radioresponsive. The incidence of distant metastasis is low, so aggressive treatment to control local disease is warranted. This may include surgery plus adjuvant radiation therapy or definitive radiotherapy for unresectable recurrences.

Current concepts in the management of patients with melanoma

Melanoma is a significant health problem. Despite public education and free cancer screenings, the incidence and mortality of melanoma continues to rise; however, many currently diagnosed melanomas are thin lesions, suggesting that education and awareness is having an impact. In addition, there are still subsets of patients who need increased surveillance in order to increase their survival. Although large congenital nevi may be precursors of melanoma, small and medium congenital nevi have an insignificant risk for melanoma development. Large congenital nevi, which are axial in location, appear to be more likely to develop melanoma and are associated with melanocytosis and melanoma of the CNS, both of which portend a poor prognosis. Recently, the recommended margins of excision have become more conservative so that many of the surgical defects can be closed primarily. Lymphoscintigraphy and sentinel node biopsy have replaced elective node dissections, thus decreasing the morbidity associated with the surgical management of melanoma. Although controversy still exists as to whether or not sentinel lymph node biopsy alters a patient's prognosis, it has been shown to be a powerful prognostic indicator. Although most melanomas are managed by routine surgical excision, other modalities are sometimes employed. For example, cryosurgery or radiation therapy may be indicated in the frail, elderly individual with a large facial lentigo maligna. Mohs surgery is the treatment of choice for head and neck melanomas and those located in areas where maximum preservation of tissue is required and for desmoplastic and acral lentiginous melanomas. Much more work remains in the area of adjuvant therapy, chemotherapy, and immunotherapy. Dacarbazine remains the drug of choice in disseminated melanoma, but remissions are usually short lived. Interleukin and biochemotherapy has yielded good results but the percentage benefiting is small. Although high dose interferon increases disease-free and overall survival in some patients, it remains a controversial drug which is not easily tolerated. In the new staging system for melanoma, ulceration is second only to Breslow's thickness. In transit (satellite) lesions have also been included in this new system. The new system also recognizes that patients with only microscopic metastatic nodal disease fare better than patients with clinically enlarged metastatic nodes and that it is the number of nodes involved with metastases, not their size, that determines the patient's prognosis. Except for lesions <1mm thick, the Clark's level of invasion has been de-emphasized.

Desmoplastic and spindle cell melanomas express protein markers of the neural crest but not of later committed stages of Schwann cell differentiation

BACKGROUND

The rare desmoplastic and spindle cell variants of malignant melanoma exhibit histological and biochemical features suggestive of early Schwann cell differentiation. These features include a spindle-shaped morphology, neurotropism, and the expression of the low affinity nerve growth factor receptor (p75NGFR).

METHODS

We evaluated by immunohistochemistry (using formalin-fixed, paraffin-embedded tissues) nine desmoplastic and three spindle cell melanomas for the expression of peripherin, p75NGFR, neural cell adhesion molecule (CD56/N-CAM), and growth-associated phosphoprotein-43 (GAP-43). Peripherin is expressed in the neural crest and in neurons, but not in cells committed to the Schwann cell lineage. p75NGFR and CD56/N-CAM also are expressed in early neural crest cells, but persist in unmyelinated and early premyelinating Schwann cells. GAP-43 is expressed in unmyelinated Schwann cells, but is downregulated in the later premyelinating to promyelinating stages of cells committed to the Schwann cell lineage.

RESULTS

Peripherin was expressed in 7/12 (58%), p75NGFR in 4/12 (33%), and CD56/N-CAM in 6/12 (50%) of the desmoplastic and spindle cell melanomas. GAP-43 was not expressed (0%) in any of the 12 melanomas (chi2, p = 0.05).

CONCLUSIONS

Desmoplastic and spindle cell melanomas express protein markers common to cells of the neural crest and to neurons similar to the immunophenotype previously reported for epithelioid cell melanomas. The expression of peripherin and the lack of expression of GAP-43 further define that these rare subtypes of melanoma do not recapitulate the later committed stages of Schwann cell differentiation.

S100-positive spindle cells in scars: a diagnostic pitfall in the re-excision of desmoplastic melanoma

Distinguishing desmoplastic melanoma (DM) from scar tissue on routine microscopy can be difficult, especially in re-excision specimens, and S100 immunohistochemistry has been recommended as a useful adjunct. The purpose of this study is to evaluate the extent and nature of S100 positivity in scars. In this study, formalin-fixed paraffin archival tissues were evaluated with immunohistochemistry. Ten re-excision specimens of previously biopsied nonnevomelanocytic lesions were immunostained with the S100 and CD57 (Leu 7) antibodies. In 9 of the 10 cases, the scars contained S100-positive spindle cells, but there were no cases with CD57+ cells. Ten re-excised atypical nevi and 10 re-excised melanomas were also immunostained for the S100 protein, and all 20 cases contained S100-positive spindle cells within the scars. There was a trend toward quantitatively more S100-positive spindle cells in these nevomelanocytic re-excisions. To evaluate the nature of the spindle cells, scars from two of the nonnevomelanocytic re-excisions were further analyzed utilizing immunostains for glial fibrillary acidic protein, HMB-45, Melan-A, CD1a, factor XIIIa, and neuron specific enolase. In both scars, neuron specific enolase diffusely stained the fibroblast population, but the remaining immunostains were negative in the scar. The presence of S100-positive spindle cells in scars represents a potential diagnostic pitfall, particularly in the evaluation of re-excision specimens of DM.

Tyrosinase expression in malignant melanoma, desmoplastic melanoma, and peripheral nerve tumors

CONTEXT

Pathologists may encounter problems in the differential diagnosis of malignant melanoma, spindle and epithelioid neoplasms of peripheral nerves, and fibrohistiocytic tumors. Tyrosinase has been demonstrated to be a sensitive marker for melanoma.

OBJECTIVE

To determine the specificity of tyrosinase expression in the differential diagnosis of melanoma, desmoplastic melanoma, and peripheral nerve sheath tumors.

DESIGN

Immunoreactivity for tyrosinase, HMB-45 (anti-gp100 protein), S100 protein, CD34, and vimentin was studied in 70 tumors, including 15 melanomas (5 desmoplastic, 4 amelanotic, 6 melanotic), 13 malignant peripheral nerve sheath tumors; 10 schwannomas (1 pigmented), 12 neurofibromas (4 pigmented), and 20 fibrohistiocytic tumors (10 dermatofibrosarcoma protuberans and 10 dermatofibromas). Microwave-based antigen retrieval was performed in 10mM citrate buffer, pH 6.0, for 20 minutes at 121 degrees C.

RESULTS

All melanomas demonstrated positive immunostaining for tyrosinase, HMB-45, and S100 protein. Immunoreactivity for HMB-45 was generally stronger than that for tyrosinase in amelanotic lesions and significantly stronger in 1 of the desmoplastic lesions. The 4 pigmented neurofibromas were focally positive for tyrosinase, but did not stain for HMB-45. The pigmented schwannoma was focally positive for both tyrosinase and HMB-45. The malignant peripheral nerve sheath tumors, dermatofibrosarcoma protuberans, and dermatofibromas were nonreactive for tyrosinase and HMB-45.

CONCLUSIONS

Our results support the sensitivity of tyrosinase expression and demonstrate the relative specificity of tyrosinase as a marker for melanocytic lesions, including desmoplastic melanoma, although pigmented peripheral nerve tumors may demonstrate focal positive staining. Immunoreactivity for tyrosinase and HMB-45 may have been enhanced by the microwave-based antigen-retrieval technique used in this study.

The clinical behavior of desmoplastic melanoma

BACKGROUND

Desmoplastic melanoma (DM) is a rare variant of malignant melanoma. A better understanding of the clinical course of DM will impact on its treatment.

METHODS

We reviewed the medical records of 59 patients with DM seen at the Mayo Clinics Scottsdale and Rochester since 1985.

RESULTS

Thirty-seven (63%) patients were male with a mean age of 62.8 years. The mean DM thickness was 6.5 mm. A total of 23 patients (39%) experienced local recurrence (LR). LR correlated with positive, unknown, or <1 cm margins. Fifty percent of patients who locally recurred subsequently developed metastatic disease. No patients were found to have positive nodal disease during ELND (16) or SLN biopsy (12). Only 1 patient (2%) developed delayed regional node metastases. Sixteen patients developed metastatic disease. The most common site was the lungs (81%).

CONCLUSIONS

LR is a significant problem and correlates with an increased risk of systemic metastatic disease. With the rare occurrence of lymphatic spread, we recommend patients undergo SLN biopsy only. DM appears to preferentially metastasize to the lungs and should be targeted when evaluating the patient for metastatic disease.

Analysis of microphthalmia transcription factor expression in normal tissues and tumors, and comparison of its expression with S-100 protein, gp100, and tyrosinase in desmoplastic malignant melanoma

Microphthalmia transcription factor (Mitf) is a nuclear protein involved in the development of melanocytes and the regulation of melanin synthesis. Recent studies have suggested that Mitf may be a more sensitive and specific melanocyte marker than S-100 protein and gp100. However, there is insufficient knowledge on the specificity of Mitf, and a systematic examination of its use for the recognition of desmoplastic melanoma has not yet been performed. In this study, we compared the expression of Mitf with S-100 protein, gp100, and tyrosinase in 20 desmoplastic melanomas by using the antibodies D5 (anti-Mitf), anti-S100P, HMB-45 (anti-gp100), and T311 (anti-tyrosinase). All 20 melanomas were positive for S-100 protein, 7 were positive for Mitf, 6 for gp100, and 11 for tyrosinase. To examine the specificity of Mitf, a panel of normal tissue and 386 samples of miscellaneous tumors, including dermal and subcutaneous spindle cell lesions relevant for the differential diagnosis of desmoplastic melanoma, were examined by immunohistochemistry. Furthermore, normal tissue samples were tested for Mitf mRNA by reverse transcriptase polymerase chain reaction (rt-PCR). Immunoreactivity for Mitf was seen not only in melanocytes of normal skin, but also in macrophages, lymphocytes, fibroblasts, Schwann cells, and smooth muscle cells at various sites, and tumors derived thereof. Our results indicate that the antibody D5 lacks sufficient sensitivity and specificity for widespread diagnostic use. Especially in re-excisions, when immunohistochemistry is often needed to distinguish an inflamed scar tissue from tumor, the presence of immunopositive inflammatory cells and fibroblasts limits the diagnostic use of D5.

S100 expression in cutaneous scars: a potential diagnostic pitfall in the diagnosis of desmoplastic melanoma

AIMS

The histological distinction of desmoplastic melanoma from cutaneous scar tissue, particularly in the context of re-excision specimens or possible recurrence, may be very difficult. Immunostaining for S100 protein is often used to discriminate although there are little data on S100 expression in scar tissue. The aim of this study was to assess whether S100-positive cells are present in dermal scars and, if so, their extent, distribution and nature.

METHODS AND RESULTS

Twenty-two re-excision specimens of previously biopsied nonmelanocytic skin lesions were reviewed. Formalin-fixed paraffin-embedded sections containing dermal scars were stained by a standard ABC immunoperoxidase technique for S100 protein, CD1a and neurofilaments. The distribution and morphology of positive cells within the dermal scar tissue were documented. Cells expressing S100 protein were identified within the scars of 21 of the 22 cases. The number of S100-positive cells varied between cases but in four specimens was substantial. They displayed a variety of morphological appearances but the majority were spindle-shaped. A few showed mild cytological atypia. It is suggested that the majority represent Schwann cells with a minority of Langerhans cells and cells of uncertain lineage.

CONCLUSION

S100-positive cells, including spindle cells showing mild atypia, are found in cutaneous scars. S100 staining of re-excision specimens or putative recurrences of desmoplastic melanoma should be interpreted with caution.

Microphthalmia transcription factor and melanoma cell adhesion molecule expression distinguish desmoplastic/spindle cell melanoma from morphologic mimics

Desmoplastic/spindle cell melanoma is a rare variant of melanoma. A number of factors complicate the diagnosis of desmoplastic/spindle cell melanoma, including the variable absence of a lentiginous component, its spindle cell morphology, and its many morphologic mimics, including scars, malignant peripheral nerve sheath tumor, neurofibroma, atypical fibroxanthoma, and spindled carcinoma. The immunohistochemical confirmation of desmoplastic/spindle cell melanoma may also be difficult, because the majority of tumors are negative for specific melanocytic markers such as HMB-45 and Melan-A, despite their usual expression of S-100 protein. Two new and potentially promising melanocytic markers, microphthalmia transcription factor (MiTF) and melanoma cell adhesion molecule (Mel-CAM), have been shown to be sensitive markers of epithelioid melanoma, but have not been tested in desmoplastic/spindle cell melanoma or in other rare melanocytic neuroectodermal tumors such as clear cell sarcoma. We immunostained 79 tumors (20 desmoplastic/spindle cell melanomas, 10 scars, 10 neurofibromas, 12 malignant peripheral nerve sheath tumors, 10 atypical fibroxanthomas, 10 clear cell sarcomas, 3 melanotic schwannomas, and 4 cellular blue nevi) for MiTF and Mel-CAM. MiTF expression was seen in 11 of 20 desmoplastic/spindle cell melanomas, 0 of 10 scars, 2 of 10 neurofibromas, 0 of 12 malignant peripheral nerve sheath tumors, 1 of 10 atypical fibroxanthomas, 7 of 10 clear cell sarcomas, 3 of 3 melanotic schwannomas, and 3 of 4 cellular blue nevi. Mel-CAM expression was present in 14 of 17 desmoplastic/spindle cell melanomas, 0 of 10 scars, 4 of 10 neurofibromas, 3 of 11 malignant peripheral nerve sheath tumors, 0 of 10 atypical fibroxanthomas, 9 of 10 clear cell sarcomas, 3 of 3 melanotic schwannomas, and 0 of 4 cellular blue nevi. MiTF and Mel-CAM were coexpressed in 6 of 17 desmoplastic/spindle cell melanomas and in no other tumor. Regarding desmoplastic/spindle cell melanoma, scar, neurofibroma, malignant peripheral nerve sheath tumor, and atypical fibroxanthoma, the sensitivity and specificity of MiTF for desmoplastic/spindle cell melanoma were 55% and 91%, respectively. For this same group of tumors, Mel-CAM had a sensitivity of 82% and a specificity of 83%. We conclude that the sensitivity and specificity of MiTF for desmoplastic melanoma equals or exceeds that of such markers as HMB-45 or Melan-A, and that MiTF should be part of the initial immunohistochemical panel for the work-up of such cases. Mel-CAM, while very sensitive, is relatively nonspecific, because it is also expressed in a variety of mesenchymal tumors and carcinomas. Mel-CAM is best reserved for cases morphologically suspected to be desmoplastic/ spindle cell melanoma, in which S-100 is positive and MiTF and other melanocytic markers are negative. These markers may also be helpful in certain other differential diagnoses, such as distinguishing clear cell sarcomas from epithelioid malignant peripheral nerve sheath tumors.

Microphthalmia transcription factor expression in cutaneous benign, malignant melanocytic, and nonmelanocytic tumors

The protein encoded by the microphthalmia (mi) gene is a transcription factor essential for the development and survival of melanocytes. Using a monoclonal antibody generated against human Mi transcription factor protein (Mitf) the authors previously demonstrated that Mitf expression is conserved in primary and metastatic malignant melanomas, and appears to be a highly sensitive and specific melanocytic marker. Mitf expression in various cutaneous nevi and cutaneous nonmelanocytic tumors has not been documented systematically. The authors evaluated Mitf immunostaining in 62 benign nevi, 58 primary cutaneous melanomas, and 53 nonmelanocytic tumors. Mitf immunostaining was conserved in all benign nevi, with Spitz nevi and neurotized nevi demonstrating decreased staining intensity. With the exception of desmoplastic melanomas, all primary cutaneous melanomas were immunopositive regardless of the cell type. Only one of 14 desmoplastic melanomas was Mitf positive. None of the nonmelanocytic tumors was immunopositive, including those lesions that may resemble melanoma histologically (spindle cell carcinomas, atypical fibroxanthomas, and leiomyosarcomas). The results demonstrate that Mitf antibody expression is conserved in the majority of benign and malignant melanocytic lesions, and that it may be helpful in the diagnosis of primary melanocytic skin lesions. Its use in desmoplastic melanomas is limited and is reflective of other melanocyte-associated antigens. Mitf discriminates between spindle cell nonmelanocytic tumors and melanomas with a spindle cell morphology, and is useful in a panel with other appropriate antibodies.

Mohs micrographic excision of melanoma using immunostains

BACKGROUND

Mohs excision of melanoma remains controversial, in part because of concerns regarding evaluation of frozen section margins. Several immunohistochemical stains are available for melanoma that can be used on frozen sections.

OBJECTIVE

To review our experience with Mohs micrographic excision of melanoma using immunostains.

METHODS

Sixty-eight patients were treated, including 46 with melanoma in situ and 22 with invasive melanoma, 62 of which were on the head or neck. HMB-45, MEL-5, Melan-A (A-103), and S-100 stains were employed.

RESULTS

Sixty-seven of 68 tumors were excised to clear margins, requiring an average of 2.0 layers. Immunostains greatly enhanced detection of melanoma on frozen sections. The average margin required for clearance of in situ melanoma was 8.3 mm and of invasive melanoma was 11.1 mm. Only 23 of 46 (50%) in situ melanomas were clear with < or =6 mm margins; 15 mm margins were required to clear 96% of the tumors. Eleven of 22 (50%) invasive melanomas were clear with < or =6 mm margins; 26 mm margins were required to clear 95% of the tumors. Melan-A (A-103) was the most consistently crisp and easily interpreted immunostain.

CONCLUSIONS

Mohs excision of melanoma using immunostains can be useful, especially for tumors on the head and neck. For routine excision, margins wider than those currently recommended may be required to ensure tumor clearance. We recommend that (1) biopsies be stained preoperatively for Melan-A and/or HMB-45, (2) a debulking layer be obtained for permanent sections prior to Mohs layers, and positive and negative control specimens from the tumor and distant skin should be employed for comparison of staining patterns. Large-scale prospective studies of in situ and invasive melanoma on the head and neck are necessary.

Comparison of immunohistochemical labelling of melanocyte differentiation antibodies melan-A, tyrosinase and HMB 45 with NKIC3 and S100 protein in the evaluation of benign naevi and malignant melanoma

A panel of three melanocyte differentiation antibodies has been compared with anti-S100 protein and NKIC3 in an assessment of benign and malignant melanocytic lesions. Anti-polyclonal S100 protein labelled all cases of primary cutaneous malignant melanoma, metastatic melanoma, desmoplastic melanoma and myxoid melanomas. In addition all benign and dysplastic naevi were positive. Conversely, HMB 45 was the least sensitive marker, labelling 24/31 primary cutaneous melanomas, 14/24 metastatic melanomas and only 1/6 desmoplastic melanomas. In the case of naevi, only junctional forms labelled consistently. Results for anti-melan-A and anti-tyrosinase were similar, although anti-tyrosinase proved slightly more sensitive in cases of malignant melanoma. NKIC3 revealed similar results to anti-tyrosinase, but had the disadvantage of reduced selectivity. It is concluded that anti-tyrosinase and anti-melan-A are useful additions to the panel of melanocytic monoclonal antibodies. In addition, both antibodies appear to have greater sensitivity for malignant melanoma than the conventionally used HMB 45 and could be considered as supportive markers to polyclonal anti-S100 protein in the diagnosis of malignant melanoma.

Morphological and immunophenotypic variations in malignant melanoma

A variety of cytomorphological features, architectural patterns and stromal changes may be observed in malignant melanomas. Hence, melanomas may mimic carcinomas, sarcomas, benign stromal tumours, lymphomas, plasmacytomas and germ cell tumours. Melanomas may be composed of large pleomorphic cells, small cells, spindle cells and may contain clear, signet-ring, pseudolipoblastic, rhabdoid, plasmacytoid or balloon cells. Various inclusions and phagocytosed material may be present in their cytoplasm. Nuclei may show bi- or multi-nucleation, lobation, inclusions, grooving and angulation. Architectural variations include fasciculation, whorling, nesting, trabeculation, pseudoglandular/pseudopapillary/pseudofollicular, pseudorosetting and angiocentric patterns. Myxoid or desmoplastic changes and very rarely pseudoangiosarcomatous change, granulomatous inflammation or osteoclastic giant cell response may be seen in the stroma. The stromal blood vessels may exhibit a haemangiopericytomatous pattern, proliferation of glomeruloid blood vessels and perivascular hyalinization. Occasionally, differentiation to nonmelanocytic structures (Schwannian, fibro-/myofibroblastic, osteocartilaginous, smooth muscle, rhabdomyoblastic, ganglionic and ganglioneuroblastic) may be observed. Typically melanomas are S100 protein, NKIC3, HMB-45, Melan-A and tyrosinase positive but some melanomas may exhibit an aberrant immunophenotype and may express cytokeratins, desmin, smooth muscle actin, KP1 (CD68), CEA, EMA and VS38. Very rarely, neurofilament protein and GFAP positivity may be seen.

Comparison of S-100 versus hematoxylin and eosin staining for evaluating dermal invasion and peripheral margins by desmoplastic malignant melanoma

Desmoplastic malignant melanoma (DMM) is a rare and locally aggressive variant of malignant melanoma, which is difficult to diagnose clinically and microscopically. A retrospective study of 21 DMM were collected during 7 years, representing 1.7% of melanomas. By comparing S-100 and hematoxylin and eosin (H & E) staining in DMM, we sought to determine whether S-100 staining offered a more accurate means of assessing dermal and neural invasion, tumor thickness, and peripheral margins. Eleven cases were excluded because the tumor extended past the deep margin. Six cases that were stained with S-100 showed a greater tumor thickness than by H & E (difference of 0. 13 to 2.79 mm). In two cases, the tumor thickness was greater by using H & E than S-100, and there was no difference in the remaining two cases. Of the eight cases in which the peripheral margins were positive by S-100, neoplastic cells were only apparent at those margins in four cases when examined with H & E. The remaining two cases showed negative margins by both stains. Clinical follow-up was obtained from nine cases, and none had recurrence of melanoma. Particularly for hypocellular and amelanotic tumors, S-100 staining proved to be a valuable adjunct in determining the extent of the tumor at the peripheral margins.

Cutaneous melanoma with myxoid features: twelve cases with differential diagnosis

Substantial myxoid change can occur in malignant melanoma, but its importance in primary disease has not been systematically evaluated. This report describes the clinical, microscopic, histochemical, and immunohistochemical findings in 12 patients with primary cutaneous malignant melanoma with myxoid features. The tumors presented as solitary lesions situated on the limbs (six lesions), trunk (four lesions), and head and neck (two lesions). The patients included six women and six men, whose ages ranged from 26 to 95 years, with a mean of 63 years. Breslow thickness varied from 0.48 mm to more than 12 mm, with a mean of more than 3.2 mm. Clinical follow-up for an average of 22 months showed one local recurrence, but no evidence of metastases yet. In all cases, there was a combination of myxoid and nonmyxoid areas. A minimum of 15% myxoid cross-sectional area was required for inclusion in the study, and up to 80% was observed. The pale blue mucin identified on hematoxylin and eosin staining was sensitive to hyaluronidase and positive for alcian blue in the 10 cases stained. Immunohistochemical staining was positive for S-100 in all 9 cases stained, positive for HMB-45 in 9 (90%) of 10, and negative for cytokeratin in all 9 cases in which myxoid melanoma remained in the block after previous sections. The presence of myxoid stroma did not define a biologically significant subgroup of melanoma. Only in cases with extensive (>50%) myxoid stromal effacement of the melanoma was there a major diagnostic hurdle. The diagnosis of primary cutaneous melanoma with myxoid features was seldom as problematic as metastatic myxoid melanoma. Positive S-100 stains, negative cytokeratin immunohistochemical stains, and hyaluronidase-sensitive alcian blue staining assisted in the diagnosis of this entity.

Metastatic malignant melanoma resembling malignant peripheral nerve sheath tumor: report of 16 cases

We report 16 cases of metastatic malignant melanoma presenting clinically as lymphadenopathy or a soft-tissue mass and histologically resembling malignant peripheral nerve sheath tumor (MPNST). In two cases, the metastatic malignant melanoma was preceded by a primary cutaneous malignant melanoma; in four cases, it presented synchronously with such a tumor; and in 10 cases, there was no evidence of a previous or concomitant malignant skin lesion. Histologically, the tumors were characterized by a malignant-appearing spindle cell proliferation arranged in fascicles, often accompanied by a peritheliomatous growth pattern, alternating hypercellular and hypocellular areas, numerous mitoses, and foci of necrosis. In nine cases, there was residual lymph node tissue. In none of the cases was there evidence of an anatomic connection with a nerve, a coexistent neurofibroma, or the stigmata of neurofibromatosis. Fourteen of the cases showed strong and generally diffuse immunoreactivity for S-100 protein, and five cases showed positivity for HMB-45. Four of eight patients with follow-up information died of the disease. Tumors with a microscopic appearance compatible with MPNST but showing strong diffuse S-100 protein staining and featuring remnants of lymph node may represent metastatic malignant melanomas and should elicit a search for a previous or concomitant tumor in the skin and other sites. The similarities these tumors share with MPNST are probably related to their common neuroectodermal histogenesis.

Recent developments in the pathology of melanocytic neoplasia

This article selectively discusses clinically relevant aspects of the pathology of cutaneous melanocytic neoplasms, from the literature of the past recent years. Topics include the changing role of immunohistochemistry in diagnosis, the controversies over dysplastic nevi, description of other specialized variants of melanocytic nevi, diagnosis of melanocytic neoplasms of acral skin, and melanoma occurring in childhood. Several variants of melanoma including desmoplastic and spindle-cell types, verrucous melanoma, epidermotropic melanoma, and melanoma of the female genitalia are reviewed. The issue of appropriate margins of resection for melanoma and the use of Mohs Micrographic surgery for this tumor are considered. Finally, a review of the sentinel node biopsy technique and of nodal nevi is presented.