Desmoplastic malignant melanoma: a clinicopathologic analysis of 113 cases

Desmoplastic melanoma (DM) is a rare variant of spindle cell melanoma, which usually develops in sun-damaged skin of elderly patients. Often the lesion is nonpigmented and frequently mistaken for a nonmelanocytic proliferation, which delays diagnosis and treatment and therefore worsens the prognosis. The spindle shape of neoplastic melanocytes, the prominent desmoplasia, and the frequent neurotropism of neoplastic melanocytes are its most characteristic histopathological features. We have studied the clinicopathologic features of 113 cases of DM. The mean age of the patients was 71.1 years; 48% of the patients were males and 52% were females. The neoplasm was located on the head in 72% of the cases. Malignant melanoma was the initial clinical diagnosis in only 27% of the cases. Histopathologically, all lesions appeared as poorly demarcated neoplasms that involved the entire dermis and often extended into the subcutaneous tissue. The neoplasms were composed of ill-defined fascicles of spindle cells. Desmoplasia was defined as the presence of spindle cells associated with a fibrotic stroma. Fifty-one cases (45%) were classified as "pure DM" when the lesion was entirely desmoplastic, and 62 cases (55%) were considered as "combined DM" when a recognizable desmoplastic component was seen in an otherwise conventional malignant melanoma. In 81% of the cases, an atypical intraepidermal melanocytic component (in situ malignant melanoma) was identified, whereas in the remaining 19% of the cases the intraepidermal component was lacking. Seventy-one percent of the cases were histologically amelanotic, 23% showed a small amount of pigment, and only 6% were heavily pigmented. Neural involvement was identified in 40/113 cases (35%), predominantly in the thickest tumors. Lymphoid nodules, found in 42/113 cases (37%), were significantly more frequent in pure DM than in combined DM (53% vs 24%). The null hypothesis of homogeneity of the "pure" and "combined" subgroups should be rejected (P < 0.002). Solar elastosis, with variable intensity, was seen in 82% of the cases. Mean Breslow thickness was 4.1 mm (4.6/3.7 mm, in the pure/combined subgroups, respectively), median was 4.0 mm (4.0/3.0 mm); Breslow thickness ranged from 0.3 to 11.0 mm, with half of the cases thicker than 4 mm. Only 4% of the cases showed Clark level below IV. The predominant neoplastic cells consisted of spindle-shaped melanocytes in 85% of the cases, whereas the remaining 15% of the cases demonstrated round neoplastic cells forming the main mass of the neoplasm. The mitotic rate of the neoplastic cells was low in 72% of the cases, 23% had an intermediate mitotic rate, and 5% showed a high mitotic rate. On follow-up, 55/113 patients (49%) (with an average of 55 months) demonstrated persistence of the disease. About 4% had local recurrences, 2% of lymph node invasion, 9% systemic metastases, and 12% died from the disease (2 cases of pure DM and 5 cases of combined DM). Although a better prognosis has been postulated for DM when compared with conventional cutaneous malignant melanomas of the same thickness, in most cases, a DM is diagnosed only in established long-standing and thick melanomas. Therefore, dermatologists and dermatopathologists should be more aware of this clinicopathologic variant of cutaneous malignant melanoma.     

The prognostic significance of basic fibroblast growth factor in cutaneous malignant melanoma

Basic fibroblast growth factor (bFGF) is a growth factor and an angiogenesis factor which may play a role in the evolution of cutaneous malignant melanoma (CMM). In this study, we evaluated the distribution of bFGF in CMM using immunochemical methods and correlated the pattern of bFGF expression with the clinical course. Formalin-fixed, paraffin-embedded sections of 46 CMMs were immunostained with a high-affinity purified antibody raised against human bFGF. CMM were categorized into lesions that exhibited subsequent recurrence (local, regional and/or systemic) or recurrence-free lesions. The minimum follow-up time was 5 years. Expression of bFGF within the tumors and in peritumoral and intratumoral blood vessels was similar in the two groups. Comparable results were attained when 8 recurring vs 8 non-recurring CMM, selected from the above tumors, were matched for age, gender, anatomic site and tumor thickness. These results suggest that the biologic behavior of CMM may not be predicted by immunoreactivity to bFGF in CMM cells or in the local tumor vasculature.     

Loss of claudin-1 expression in tumor-associated vessels correlates with acquisition of metastatic phenotype in melanocytic neoplasms

Claudins are a family of transmembrane proteins involved in cell-to-cell adhesion and are believed to be the main component of tight junctions. Recent studies have suggested that some metastatic solid tumors lack claudin expression. It is unknown whether claudins play a role in cutaneous melanoma. Immunohistochemical studies were performed on tissue microarrays containing 19 benign melanocytic nevi (BN), 21 dysplastic nevi (DN), 23 primary malignant melanomas (MMs), and 31 metastatic melanomas (MMMs) using a polyclonal anti-claudin-1 antibody. Immunoreactivity in tumor cells and associated vessels was graded by intensity and by percentage of reactive cells. Normal epidermis served as internal control (3+ labeling). Cases with at least 2+ labeling in more than 25% of the cells were considered positive. Claudin-1 expression was present in 37% of BN, 24% of DN, 26% of MM, and 3.2% of MMM. Tumor-associated vessels showed the following results: 11 of 19 (58%) in BN, 14 of 21 (67%) in DN, 17 of 23 (74%) in MM, and 6 of 31 (19%) in MMM. A significant loss of expression was noted between MMM and all other lesions in tumor cells and associated vessels. There was no significant difference between BN, DN, and MM. Within primary melanomas, there was a significant correlation between expression of claudin in tumor cells and Clark level/Breslow thickness. Also significant was a decreased expression of claudin in tumor vessels of lesions with higher Breslow thickness or Clark level. These data suggest that loss of claudin-1 may play a significant role in the acquisition of metastatic phenotype in cutaneous melanoma. Cohn ML, Goncharuk VN, Diwan AH, Zhang PS, Shen SS, Prieto VG. Loss of claudin-1 expression in tumor-associated vessels correlates with acquisition of metastatic phenotype in melanocytic neoplasms.     

Subclassification of desmoplastic melanoma: pure and mixed variants have significantly different capacities for lymph node metastasis

Background:  There is disagreement about the behavior and optimal management of desmoplastic melanoma (DM), particularly regarding the incidence of lymph node (LN) involvement. Recently, investigators have noted the frequently heterogenous histologic composition of DM and have found significant differences between pure desmoplastic melanoma (PDM) (≥90% comprised of histologically typical DM) and mixed desmoplastic melanoma (MDM) [≥10% DM and >10% conventional melanoma (CM)].
Method:  We reviewed 87 cases of DM comparing the histologic and clinical features of PDM (n = 44) to MDM (n = 43).
Results:  At surgical staging, there were LN metastases in 5 of 23 (22%) MDM patients, whereas all 17 PDM patients had negative LN biopsies (0%) (p = 0.04). PDM was less often clinically pigmented (36% vs. 67%) and had a lower mean mitotic index (1.3 vs. 3.0).
Conclusions:  There are differences between PDM and MDM, the most important of which is the incidence of LN involvement. Our findings support the clinical utility of classifying DM into pure and mixed subtypes because the negligible rate of nodal involvement in PDM does not support the routine performance of sentinel LN biopsy in this subgroup of melanoma patients. In contrast, the incidence of LN involvement in MDM is comparable to that of CM.     

Distinction of desmoplastic melanoma from non-desmoplastic melanoma by gene expression profiling

Desmoplastic melanoma (DM) is a variant of melanoma characterized by the presence of amelanotic fusiform melanocytes dispersed in a prominent collagenous stroma. DM behaves differently from conventional non-desmoplastic melanoma (NDM). It has a higher tendency for local recurrence and is less likely to metastasize to regional lymph nodes. In this study, we explored the possibility of distinguishing DM from NDM by gene expression profiling. RNA samples from ten primary cutaneous melanomas of similar depth of invasion were analyzed using the Affymetrix U133A oligonucleotide platform. Four tumors were DM, five were ND, and one tumor showed combined features of desmoplastic and conventional. Hierarchical cluster analysis clearly separated DM from NDM. The expression of a number of melanocyte differentiation genes was decreased in DM compared with NDM, which corresponded to immunohistochemical results. Various genes were upregulated in DM, including neurotrophic factors and genes involved in extracellular matrix production. A novel finding was the high expression of clusterin in DM, which was confirmed by immunohistochemical studies. Our results from gene expression profiling validate the distinction of DM from NDM. They also provide the opportunity to learn more about the biology of DM which had previously not yet been associated with this variant of melanoma.     

Cyclooxygenase-2 overexpression in human basal cell carcinoma cell line increases antiapoptosis, angiogenesis, and tumorigenesis

Cyclooxygenase-2 (COX-2) is critical for tumor formation, angiogenesis, metastasis, and prognosis. In this study, the role of COX-2 in antiapoptosis, tumorigenesis, and angiogenesis of human basal cell carcinoma (BCC) cells was investigated. Transfection of COX-2 constitutive expression vector into a BCC cell line yielded several overexpressing clones. All transfectants demonstrated remarkable resistance to ultraviolet B-induced apoptosis (confirmed by flow cytometry analysis, morphological change, and DNA fragmentation). Immunoblot analysis revealed marked increases in apoptosis-regulated genes Mcl-1 and Bcl-2. A 10-fold concentrated conditioned medium from COX-2-overexpressing BCC cells exhibited higher angiogenic activity in Matrigel plug and human umbilical vein endothelial cell tube formation assay. Cells exhibited increased levels of vascular endothelial growth factor-A (VEGF-A) mRNA and protein, and secreted VEGF-A and basic fibroblast growth factor (bFGF). COX-2-specific small interfering RNA markedly reduced the secreted species. After 7 weeks of inoculation, the tumor volume of COX-2-overexpressing cells in severe combined immunodeficient mice was significantly greater than that of vector control cells. Immunohistochemical analysis of CD31-positive vessels revealed a two-fold increase in microvessel density in COX-2 tumors, compared to control vector tumors. Our data indicate that Mcl-1 and Bcl-2, as well as VEGF-A and bFGF, are downstream effectors of COX-2-induced antiapoptosis and angiogenesis, respectively.     

The expression of CD23 in cutaneous non-lymphoid neoplasms

BACKGROUND: Cluster designation 23 (CD23) is generally used as a lymphoid marker. Its utility in cutaneous epithelial tumors has never been studied. In our routine practice, we observed that CD23 reacted strongly with eccrine and apocrine secretory coils. METHODS: Immunohistochemical staining of CD23 was performed in a total of 131 cases of apocrine, eccrine, follicular and other cutaneous non-lymphoid tumors. RESULTS: CD23 expression was detected in all benign apocrine tumors and in half of benign eccrine tumors, particularly those derived from secretory coils. CD23 staining was seen in 42% (8/19) of microcystic adnexal carcinoma (MAC), while no staining was observed in tumor cells of desmoplastic trichoepithelioma, morpheaform basal cell carcinoma and syringoma. All mammary and extramammary Paget's disease were labeled with CD23. In comparison, pagetoid Bowen's disease, melanoma in situ and sebaceous carcinoma exhibited negative staining. In addition, CD23 reacted diffusely with cutaneous mucinous eccrine carcinoma in a manner similar to breast or colonic adenocarcinoma. CONCLUSION: CD23 appears to be a reliable immunohistochemical marker of the eccrine/apocrine secretory coil and helpful in identifying sweat gland tumors of such origin. It is of ancillary value in differentiating MAC from its mimicker. CD23 is a useful addition to the diagnostic immunohistochemical panels for Paget's disease.     

A case of Stewart-Treves syndrome. An immunohistochemical and electron-microscopic study

A case of so-called postmastectomy lymphangiosarcoma occurring 14 years after radical mastectomy and radiation therapy for the left mammary carcinoma was studied histologically, electron-microscopically and immunohistochemically. The tumor cells formed cellular areas in the deeper dermis and vascular areas in the upper dermis. The tumor cells had round to spindle-shaped hyperchromatic nuclei and poorly delineated eosinophilic cytoplasm. In an immunohistochemical study for factor VIII relating antigen (F VIII RAG), positive staining was observed in the cytoplasm of some atypical tumor cells lining vascular channels and in that of a few tumor cells in cellular areas. Electron-microscopically, the tumor cells were similar to vascular endothelial cells. In these studies, this tumor appeared to mimic not only lymphatic vessels, but also blood vessels. For these reasons, the authors propose that the term “angiosarcoma” would be better suited than “lymphangiosarcoma”.     

Dermoscopy of desmoplastic melanoma: report of six cases

Background  Desmoplastic melanoma (DM) is a rare variant of cutaneous melanoma. Its diagnosis is often delayed by an unusual clinical presentation. The dermoscopic features of DM have not yet been described.
Objective  To define the dermoscopic features of DM.
Patients and methods  A single-institution register-based retrospective study of six cases of histology-proven desmoplastic melanoma for which dermoscopy data were available. The criteria we studied included: network, dots and globules, streaks, regression features, ulceration, number of colours, blue/white veil, and vascular pattern.
Results  Only three cases exhibited one classical feature for a melanocytic lesion; other cases were recognized on the basis of the presence of figures of regression (all six), i.e. white scar-like areas and 'peppering' (three of six), multiple (> 4) colours (five of six), and of melanoma-related vascular patterns (five) such as linear-irregular vessels (four) and milky-red areas (two).
Discussion  We believe that dermoscopy could help in the accurate diagnosis of this rare neoplasm. In the absence of a pigmented network, attention should be given to the identification of features of regression and to melanoma-associated vascular patterns.     

HLA-DR antigen expression in primary melanomas of the skin

Ninety-three primary malignant melanomas of the skin were typed immunohistologically for the expression of HLA-DR on tumor cells. HLA-DR-positive stroma cells were HLA-DR-negative or only locally positive in most cases. In 36 (39%) of the lesions more than 10% of the tumor cells were stained by two monoclonal antibodies against the nonpolymorphic portion of HLA-DR. HLA-DR-positive tumor cells were often accumulated at the advancing front of the melanoma. The occurrence of HLA-DR-positive tumor cells was related to tumor thickness and level of invasion. Substantial numbers of HLA-DR-positive tumor cells were found in half of the tumors thicker than 1.5 mm and in only 18% of flatter lesions. The highest percentage of HLA-DR-positive tumors was found in the group of melanomas invading the reticular dermis (level IV). The majority of tumors (18/24) that had metastasized within an observation period of 0-32 months were HLA-DR-positive. Regarding the mononuclear cell infiltrate, no correlation between the degree of overall infiltrate and the expression of HLA-DR by the tumor cells was found. The infiltrate within the tumor, however, was more often marked in HLA-DR-positive than in HLA-DR-negative melanomas.     

Vascular involvement in the prognosis of primary cutaneous melanoma

OBJECTIVE: To examine the role of vascular invasion as a prognostic factor in melanoma. DESIGN: Retrospective survival analysis. SETTING: Academic medical center. PATIENTS: A total of 526 patients with primary cutaneous melanoma from the University of California, San Francisco, Melanoma Center database with 2 years of follow-up or documented relapse. MAIN OUTCOME MEASURES: (1) Presence of vascular involvement defined as vascular invasion with tumor cells within blood or lymphatic vessels; or uncertain vascular invasion, with melanoma cells immediately adjacent to the endothelium. (2) Percentage with metastasis or death and relapse-free and overall survival. RESULTS: The presence of either type of vascular involvement significantly increased the risk of relapse and death and reduced the survival associated with melanoma. The impact of vascular involvement on these outcomes was similar to that of ulceration. In a multivariate analysis, vascular involvement was the second most important factor (after tumor thickness) in the primary tumor in predicting survival. CONCLUSIONS: Vascular involvement is an important independent predictor of metastasis and survival in melanoma. The phenomenon of uncertain vascular invasion describes an earlier step than definite vascular invasion in tumor progression.     

Immunohistochemistry of dermatofibromas and benign fibrous histiocytomas

Dermatofibromas (DF) are common, benign skin tumors composed predominantly of cells having elongated nuclei and very scant cytoplasm (i.e., fibroblasts) and capillaries in a collagenous stroma. Some authors distinguish DF from benign fibrous histiocytomas (BFH), which are composed of cells with round to oval nuclei and abundant cytoplasm (i.e., histiocytes). In general, this group of tumors expresses factor XIIIa but not the antigen recognized by MAC 387. However, immunohistochemical differences specifically between DF and BFH have not been reported.
We have studied the immunophenotype of 23 lesions having morphologic features predominantly either of DF (17 cases) or BFH (6 cases) using antibodies against desmin (muscle marker), α-smooth-muscle actin (muscle and myofibroblast marker), CD68 and HAM56 antigen (markers commonly expressed by macrophages, so called "histiocytic" markers), CD34 (a marker present in hematopoietic, vascular, and occasional dermal dendritic cells), and factor XIIIa (a transglutaminase present in many cells including dermal dendrocytes). Many spindle-shaped cells expressed a-smooth-muscle actin while many large, round cells expressed the histiocytic markers. However, most lesions expressed at least focally both α-smooth-muscle actin and "histiocytic" markers. Thus a clear-cut distinction between DF and BFH could not be made based on immunophenotype alone. Additionally, the prominent α-smooth-muscle actin immunoreactivity and desmin non-reactivity suggests myofibroblastic differentiation in the spindle-cell regions of these tumors, and indicates that expression of a-smooth-muscle actin cannot be used as definitive proof of muscle differentiation in spindle-cell tumors. We conclude that DF and BFH are not discrete entities, but represent polar expressions of one nosologic entity exhibiting both myofibroblastic and "histiocytic" differentiation.     

Primary invasive signet-ring cell melanoma

The histopathological variants of malignant melanoma include the common type (lentigo maligna, superficial spreading melanoma, nodular melanoma, acrolentiginous melanoma), spindle cell, desmoplastic, balloon cell, pleomorphic (fibrohistiocytic), myxoid, small cell melanoma and malignant blue nevus. Recently, signet-ring cell melanoma was introduced as an additional cytologic variant. We describe a 72-year-old patient with a primary signet-ring cell melanoma of the skin located on the upper arm. Histopathologic examination disclosed a melanocytic tumor extending from the epidermis to the deep reticular dermis. Numerous pleomorphic tumor cells showed large, intracellular vacuoles and oval to spindle-shaped nuclei at their periphery. Mitotic figures and multinucleated melanocytes were also observed. Some of the signet-ring cells exhibited cytoplasmatic periodic acid-Schiff (PAS)-positivity. Immunohistochemistry showed positive reaction of the tumor cells for S-100, HMB-45 protein and vimentin, confirming their melanocytic differentiation. Tumor cells were negative for cytokeratins, epithelial membrane antigen (EMA), and carcinoembryonic antigen (CEA). The signet-ring cell melanoma disclosed an invasion to Clark Level IV and tumor thickness of 2.2 mm. Signet-ring cell melanoma is a rare morphologic variant of melanoma. Its recognition is important for differentiation from other tumors featuring signet ring cells.     

Hpa-ASODN转染裸鼠黑素瘤移植瘤中VEGF和bFGF蛋白的检测

碱性成纤维细胞生长因子(bFGF)蛋白表达的影响。方法以人恶性黑素瘤A375细胞株及Hpa-ASODN转染A375细胞株建立裸鼠体内移植瘤模型,采用免疫组化检测肿瘤组织Hpa-ASODN,VEGF和bFGF蛋白的表达。结果Hpa-ASODN转染后裸鼠移植瘤VEGF和bFGF蛋白表达低于对照组(P<0.05)。结论Hpa-ASODN可下调裸鼠移植瘤组织中VEGF和bFGF蛋白的表达。     

The altered expression of α-smooth muscle actin in basal cell epithelioma and its surrounding stroma: with special reference to proliferating cell nuclear antigen expression and adenoid differentiation

Summary Altered expression of α-smooth muscle actin (α-SMA) is known to indicate the morphological, tumorigenic and immunological changes occurring in tumour and stromal cells. The purpose of this study was to analyse the dynamics of α-SMA expression in human basal cell epithelioma (BCE) cells and their surrounding stromal cells, in the process of differentiation towards cutaneous appendages such as hair, sebaceous, apocrine and eccrine glands. Using anti-α-SMA specific monoclonal antibody (MAb), 17 of 36 BCEs (47%) were shown to express α-SMA, despite the usual absence of α-SMA in all eukaryotic cells except muscle cells. Solid, adenoid and sclerosing types of BCE expressed α-SMA more frequently, and in greater amount, than cystic, keratotic and superficial types. Furthermore, the expression of α-SMA in BCE cells significantly paralleled the expression of proliferating cell nuclear antigen (PCNA) in these cells. Thus, the altered expression of α-SMA may reflect the growing properties of BCE cells under the specific cellular regulations for differentiation.
In addition, we have found anti-α-SMA MAb-positive fibroblasts with smooth muscle differentiation (myofibroblasts) in desmoplastic stroma surrounding BCE nests in 13 of 36 cases (36%). Coincidental expression of α-SMA in both BCE cells and stromal cells was found in nine of the 13 cases (69%), indicating the possibility of the induction of myofibroblastic stromal changes in surrounding tissues by cytokines secreted from BCE cells [e.g. basic fibroblast growth factor (bFGF)-like factor].     

p16 Expression differentiates between desmoplastic Spitz nevus and desmoplastic melanoma

Background:  Loss of p16 in melanomas reflects worse outcomes for patients. It is associated with depth of invasion, ulceration, vascular invasion, lymph node metastases, metastases, recurrence of melanoma and decreased 5-year survival. Desmoplastic melanoma is an insidious malignant melanoma subtype that commonly occurs on sun-damaged skin of the head and neck area in elderly patients. The diagnostic conundrum occurs with confusion of desmoplastic melanoma with scars, hyalinizing blue nevi, desmoplastic Spitz nevi and diffuse neurofibromas.
Methods:  The present study uses immunohistochemistry with a p16 antibody to differentiate desmoplastic Spitz nevi (n = 15 cases) from desmoplastic melanomas (n = 11). To date, no other studies have been published defining the expression pattern of p16 in desmoplastic melanoma.
Results:  81.8% of desmoplastic melanomas were negative for p16 and 18.2% were only weakly stained. In contrast, all desmoplastic Spitz nevi were moderately to strongly positive for p16.
Conclusions:  The staining pattern for p16 in desmoplastic melanomas and Spitz nevi in conjunction with the histopathologic features, S100 staining, Ki67 proliferation index and clinical scenario may aid in the difficult differential diagnosis between these two entities. Further confirmatory studies are indicated.     

A light microscopic and immunohistochemical evaluation of scars

BACKGROUND: Scars are commonly encountered by dermatopathologists and usually do not present a diagnostic challenge. However, in some cases, the pathologist may need to consider a broad differential diagnosis including fibrohistiocytic tumors, smooth muscle tumors, myofibroblastic proliferations and desmoplastic malignant melanoma. Although specific histologic aspects of scars have been well studied, a complete histochemical profile of scars, especially at various stages of evolution, has not been described. METHODS: Twenty-five cases of scars including 8 normal scars, 5 hypertrophic scars and 12 keloids were studied. Sections were examined with Verhoeff van Giesson, colloidal iron, Giemsa, smooth muscle actin (SMA), CD34, Factor XIIIa and S-100. RESULTS: All scars were negative for CD34 expression. Factor XIIIa immunostaining identified only rare dermal dendrocytes. S-100 was absent in 23 of 25 cases and stained scattered cells (less than 10%) in the other 2 cases. SMA was positive in 14 of 25 cases with 6 of these showing staining of up to 50% of spindled cells. Elastic fibers were markedly reduced or absent in all cases, mucin showed moderate or marked staining in three-fourths of the cases and dermal mast cells showed a moderate increase in 5 cases. CONCLUSIONS: These findings confirm prior reports that negative staining with CD34, Factor XIIIa and S-100 can help differentiate scars from dermatofibrosarcoma protuberans, dermatofibroma and desmoplastic malignant melanoma, respectively. SMA staining is much more variable and requires careful interpretation.     

IMP‐3 expression in melanocytic lesions

Background: Insulin‐like growth factor‐II mRNA‐binding protein 3 (IMP‐3 ), a member of the insulin‐like growth factor mRNA‐binding protein family, is expressed in several human malignancies, including melanomas. However, the expression of IMP‐3 has not been explored in melanoma in situ, various histologic subtypes of invasive melanomas and atypical Spitz tumors. Methods: IMP‐3 immunostain was performed in 157 melanocytic lesions. Results: Nearly all benign (8/8), dysplastic (8/8) and Spitz nevi (8/9) were negative for IMP‐3. Focal IMP‐3 positivity was observed in 5/12 melanoma in situ and 4/15 superficial melanomas (Breslow depth ≤1 mm). Half (10/20) of deep melanomas (Breslow depth >1 mm) and 25/52 metastatic melanomas demonstrated strong IMP‐3 staining. IMP‐3 expression differs significantly between non‐desmoplastic melanomas (superficial and deep) and benign or dysplastic or Spitz nevi (p = 0.0427, respectively). Four of 23 desmoplastic melanomas expressed IMP‐3 , which was significantly different from deep melanomas (p = 0.0109). IMP‐3 stained 7 of 10 atypical Spitz tumors. The difference between atypical Spitz tumors and Spitz nevi was statistically significant (p = 0.0256). Conclusion: A malignant circumstance, such as non‐desmoplastic melanoma or atypical Spitz tumor, can be inferred when IMP‐3 is expressed, suggesting potential diagnostic value of IMP‐3 in melanocytic lesions. Yu L, Xu H, Wasco MJ, Bourne PA, Ma L. IMP‐3 expression in melanocytic lesions.     

Protein expression of melanocyte growth factors (bFGF, SCF) and their receptors (FGFR-1, c-kit) in nevi and melanoma

BACKGROUND: Basic fibroblast growth factor (bFGF) and stem cell factor (SCF) are essential growth factors for melanocytes which carry the receptors FGFR-1 for bFGF and c-kit for SCF. Because both factors may be involved in melanoma development, the expression of bFGF/FGFR-1 and SCF/c-kit was investigated in melanocytic tumors of different progression stages. METHODS: Fifty primary melanomas and 44 melanocytic nevi were analyzed for the expression of SCF, c-kit, bFGF, and FGFR-1 by immunohistochemistry. RESULTS: In melanoma, SCF and c-kit were expressed in 76 and 84%, respectively, and coexpressed in 66%. bFGF and FGFR-1 were expressed in 45 and 86%, respectively, and coexpressed in 46%. In melanocytic nevi, SCF was expressed in 23% and c-kit in 70% while coexpression was more common in dysplastic (39%) than non-dysplastic subtypes (8%). bFGF and FGFR-1 were expressed in 55 and 67%, respectively, while coexpression was found in 47% but varied considerably between different histological subtypes. CONCLUSIONS: SCF and c-kit were frequently expressed by melanomas and dysplastic nevi suggesting an autocrine growth mechanism as described for bFGF. In both nevi and melanoma, c-kit was almost exclusively found in the epidermis while bFGF was more common in the dermis. Thus the growth factor/receptor expression seems to depend on the cutaneous localization of the melanocytic tumors.