Copy number variations and clinical outcome in atypical spitz tumors

Atypical Spitz tumors (ASTs) are rare spitzoid neoplasms of uncertain biological behavior. Our study was designed to characterize genetic abnormalities that may help to differentiate ASTs from melanoma or Spitz nevi. We examined copy number variation in formalin-fixed, paraffin-embedded samples using an Agilent 44k array comparative genomic hybridization platform. Sixteen patients with AST (8 with positive sentinel lymph node biopsy, 1 with distant metastasis), 8 patients with Spitz nevi, and 3 patients with melanoma (2 spitzoid, 1 superficial spreading) were evaluated. Chromosomal aberrations were found in 7 of 16 ASTs, 1 with fatal outcome, 2 spitzoid melanomas, and 1 conventional melanoma. We found no difference in chromosomal instability between AST patients with positive and negative sentinel lymph node biopsies. Our patient with widely metastatic AST lacked the most frequent aberrations in melanoma involving chromosomes 6 and 11q that are loci targeted by fluorescence in situ hybridization (FISH) probes developed to distinguish malignant melanoma from benign melanocytic lesions. The vast majority of chromosomal abnormalities observed in ASTs are not commonly found in melanomas, suggesting that AST may be a distinct clinical entity and raising additional questions regarding their malignant potential, prognosis, and clinical management. The current FISH probes failed to detect 1 spitzoid melanoma, 1 fatal metastatic AST case, and the other chromosomally aberrant ASTs in our series, but detected 1 spitzoid melanoma and 1 conventional melanoma. Thus, a comprehensive, genome-wide approach to chromosomal abnormalities offered greater sensitivity and specificity than current FISH probes in identifying spitzoid lesions of uncertain malignant potential in this series.     

Proliferative nodules arising within congenital melanocytic nevi: a histologic, immunohistochemical, and molecular analyses of 43 cases

The histopathologic interpretation of proliferative nodules (PNs) in congenital melanocytic nevi can present significant challenges as some PNs may exhibit atypical features that make the distinction from melanoma difficult. We compared histologic features, Ki-67%, PHH3, and CD117% expression levels by immunohistochemistry in 18 benign and 25 atypical PNs (from 41 patients) with that of background congenital nevi (of these 43 cases), 10 congenital nevi, and 3 dermal melanomas arising in congenital melanocytic lesions. In addition, we evaluated the presence of BRAF, GNAQ, HRAS, KRAS, and NRAS mutations in all groups using the SNaPshot Multiplex System. Follow-up was available on 19 patients (9 benign and 10 atypical PNs) (range, 2 to 20 y; median, 8 y) and all were alive with no evidence of disease. The specific histologic features of atypical PNs, such as sharp demarcation (P<0.001), expansile growth (P<0.001), epidermal effacement (P<0.001), nuclear pleomorphism (P<0.001), and increased mitoses (P<0.001), differed significantly from those of benign PNs. Immunohistochemical results showed that Ki-67% and PHH3 scores, but not CD117% expression, were significantly higher (P<0.05) in atypical PNs. Molecular analyses showed that the PNs and background congenital melanocytic nevi of the giant congenital nevi possess more frequent NRAS mutations and infrequent BRAF mutations when compared with those of the remaining cases. These findings suggest that histologic features and Ki-67 and PHH3 expression levels are the strongest parameters to distinguish between benign versus atypical PNs. The immunohistochemical results suggest that atypical PNs are distinct borderline lesions residing between benign PNs and dermal melanomas. Although numerous mutations are detected in the samples, the diagnostic use of molecular analysis in this regard is limited.     

Detection of B-RAF and N-RAS mutations in human melanoma

BACKGROUND: It is now known that activating point mutations in components of the mitogen-activated protein kinase pathway commonly occur in melanoma. We previously described a method to detect point mutations in heterogenous tissues containing both wild-type and mutant B-RAF and N-RAS genes by using site-directed mutagenesis to introduce new restrictions sites in the cDNA sequence when the specific point mutations are present. We modified this technique to improve sensitivity and used it to determine the incidence of B-RAF and N-RAS mutations in human melanoma. STUDY DESIGN: We screened 115 melanoma samples for the most common B-RAF and N-RAS mutations found in melanoma using a site-directed mutagenesis-based detection technique. Southern blotting was used to increase sensitivity of the basic system. We also tested this method of genetic mutation detection in fine-needle aspiration specimens and paraffin-embedded tissues. RESULTS: Sixty-eight samples (20 of 36 primaries, 18 of 27 regional metastases, 16 of 40 nodal metastases, and 9 of 12 distant metastases) harbored the V599E B-RAF mutation (59%), 17 contained a Q61R N-RAS mutation, and 4 contained a Q61K N-RAS mutation. We were able to detect the V599E mutation in genomic DNA from paraffin-embedded melanoma samples and could routinely detect this mutation in fine-needle aspirations of melanoma tumors. This method of detection was sensitive and specific with no false positives. CONCLUSIONS: Activating mutations of B-RAF and N-RAS were present in approximately 60% and 18%, respectively, of samples tested. The site-directed mutagenesis system of mutation detection was both sensitive and specific in detecting these mutations and will likely prove very clinically useful in future studies.     

Consumption of the epidermis: a diagnostic criterion for the differential diagnosis of melanoma and Spitz nevus

The distinction between melanoma and its most important simulant, Spitz nevus, is usually made on microscopically. We point out "consumption of the epidermis" (COE) as an additional diagnostic criterion. We defined COE as thinning of the epidermis with attenuation of the basal and suprabasal layers and loss of rete ridges in areas of direct contact with neoplastic melanocytes. We analyzed 102 unequivocal melanomas and 125 unequivocal Spitz nevi for the presence of COE. COE had not been used in arriving at the diagnosis of these cases because we were unaware of the criterion at the time that the cases were first evaluated. COE was found in 88 of 102 (86%) of melanomas but only 12 of 125 (9.6%) of Spitz nevi (P < 0.001). We then looked for COE in an independent set of 61 ambiguous melanocytic lesions with overlapping histopathologic features that could not be classified unequivocally as Spitz nevus or melanoma. The cases were analyzed by comparative genomic hybridization (CGH) for aberration patterns suggesting a benign or a malignant process, based on previous studies. COE was found in only 6 of 42 (14%) of the ambiguous cases in which CGH suggested a benign process and 14 of 19 (74%) of the ambiguous cases in which CGH suggested melanoma (P < 0.001). Our data suggest that COE is a useful criterion in the evaluation of melanocytic neoplasms. Because COE was frequently found at the edges of ulcers in the majority of ulcerated melanomas, the thinning of the epidermis in COE may represent an early phase of ulceration. This may prove to be important in distinguishing ulceration due to an effect of the tumor from ulceration due to trauma, which would be expected not to have the same prognostic import. Future studies are required to analyze the prognostic value of COE itself.     

Loss of expression of protein kinase a regulatory subunit 1alpha in pigmented epithelioid melanocytoma but not in melanoma or other melanocytic lesions

Pigmented epithelioid melanocytoma (PEM) is a recently described entity comprising most cases previously described as "animal-type melanoma" and epithelioid blue nevus (EBN) occurring in patients with the multiple neoplasia syndrome Carney complex (CNC). Mutations of the protein kinase A regulatory subunit type 1alpha (R1alpha) (coded by the PRKAR1A gene) are found in more than half of CNC patients. In this study, we investigated whether PEM and EBN are related at the molecular level, and whether changes in the PRKAR1A gene status and the expression of the R1alpha protein may be involved in the pathogenesis of PEM and other melanocytic lesions. Histologic analysis of hematoxylin and eosin-stained sections and immunohistochemistry (IHC) with R1alpha antibody were performed on 34 sporadic PEMs, 8 CNC-associated PEMs from patients with known PRKAR1A mutations, 297 benign and malignant melanocytic tumors (127 conventional sections of 10 compound nevi, 10 Spitz nevi, 5 deep-penetrating nevi, 5 blue nevi, 6 cellular blue nevi, 2 malignant blue nevi, 3 lentigo maligna, and 86 melanomas of various types); in addition, 170 tissue microarray sections consisting of 35 benign nevi, 60 primary melanomas, and 75 metastatic melanomas, and 5 equine dermal melanomas, were examined. Histologic diagnoses were based on preexisting pathologic reports and were confirmed for this study. DNA studies [loss of heterozygosity (LOH) for the 17q22-24 locus and the PRKAR1A gene sequencing] were performed on 60 melanomas and 7 PEMs. IHC showed that R1alpha was expressed in all but one core from tissue microarrays (169/170), and in all 127 melanocytic lesions evaluated in conventional sections. By contrast, R1alpha was not expressed in the 8 EBN from patients with CNC and PRKAR1A mutations. Expression of R1alpha was lost in 28 of 34 PEMs (82%). R1alpha was expressed in the 5 equine melanomas studied. DNA studies correlated with IHC findings: there were no PRKAR1A mutations in any of the melanomas studied and the rate of LOH for 17q22-24 was less than 7%; 5 of the 7 PEMs showed extensive 17q22-24 LOH but no PRKAR1A mutations. The results support the concept that PEM is a distinct melanocytic tumor occurring in a sporadic setting and in the context of CNC. They also suggest that PEM differs from melanomas in equine melanotic disease, further arguing that the term animal-type melanoma may be a misnomer for this group of lesions. Loss of expression of R1alpha offers a useful diagnostic test that helps to distinguish PEM from lesions that mimic it histologically.     

The surgical management of Spitz naevi and atypical spitzoid neoplasms: A review of the literature

Melanocytic lesions with spitzoid differentiation represent a difficult management paradigm. Spitzoid lesions form a spectrum of distinct entities encompassing Spitz naevi, atypical Spitz naevi and spitzoid melanoma which range from benign to malignant. Differentiation between benign and malignant lesions can be challenging making surgical management difficult. In this article we highlight clinical conundrums regarding Spitz naevi. We review the literature in terms of the newer histological tools available to more explicitly classify these lesions which may have an impact on the clinical care that these patients receive. The article also examines the controversy which surrounds the role of sentinel lymph node biopsy (SLNB) and completion lymph node dissection in the management of melanocytic proliferations with Spitz differentiation.     

The Surgical Management of Spitz Nevi

BACKGROUND: The biologic behavior of Spitz nevi and atypical Spitz nevi ranges from completely benign to the rare malignant melanoma. Various recommendations for the surgical approach to these lesions have been proposed. OBJECTIVE: To determine any trend in the surgical management of Spitz nevi and atypical Spitz nevi among a community of dermatologists. METHODS: Retrospective review of the clinical features, surgical management and outcome of 89 patients with the diagnosis of Spitz nevus or atypical Spitz nevus. RESULTS: All biopsy techniques had a high incidence of involved margins: shave (67%), excision (28%), and punch (21%). Of the atypical Spitz nevi with positive margins on biopsy, there was a trend (7/9) toward reexcision with narrow margins (average 2.2 mm). CONCLUSION: The majority of atypical Spitz nevi incompletely removed by biopsy were excised with narrow uncontrolled margins. A stratified surgical approach depending on the clinical and histopathologic features of the Spitz lesion is proposed. More aggressive surgical management of Spitz lesions with atypical features may be warranted. Further studies to determine the biologic potential of these lesions are needed.     

Persistent (recurrent) Spitz nevi: a histopathologic, immunohistochemical, and molecular pathologic study of 22 cases

This report describes 22 Spitz nevi that seemed to have been clinically removed but persisted and clinically recurred at the biopsy site. These were evaluated in terms of histopathology, immunohistochemistry, and molecular pathology using comparative genomic hybridization (CGH) and fluorescent in situ hybridization. One of these 22 lesions was originally reported as an atypical melanocytic proliferation with some features of a Spitz nevus and was included in the study set at an early stage but was later recognized as melanoma after metastasis to regional lymph nodes 3 years after the local recurrence. We noted four histopathologic patterns in the recurrent lesions: 1) a predominantly intraepidermal pattern resembling "pseudomelanoma" as seen in recurrent "common" melanocytic nevi, 2) a compound, mostly nested pattern above or within a scar that was nearly identical to the originally biopsied Spitz nevus, 3) a nodular growth pattern that closely simulated invasive melanoma, and 4) a desmoplastic pattern resembling an intradermal desmoplastic Spitz nevus. Although the majority of recurrent lesions exhibited asymmetry and pagetoid spread, the dermal component usually had a low mitotic rate and retained architectural and cytologic maturation, which allowed distinction from invasive melanoma. Except for the metastasizing melanoma, the immunostaining pattern with S-100 and HMB-45 was identical to that previously reported for Spitz nevi. Ki67 revealed a very low proliferation rate in all cases, including the melanoma. CGH performed in 10 cases yielded results consistent with Spitz nevi in eight cases. The remaining two cases showed CGH profiles more typical of melanoma, and one of these was the above-referenced case of melanoma, proven by metastasis. Although ancillary molecular techniques such as CGH are of great help in distinguishing these from melanoma, until such techniques become widely available we advocate complete but conservative excision of any recurrent Spitz nevus.     

Melanocytic nevi with an atypical epithelioid cell component: clinical, histopathologic, and fluorescence in situ hybridization findings

Combined melanocytic nevi can contain a phenotypically distinct population of large atypical epithelioid cells in a background of smaller banal-appearing melanocytes. On the basis of the pattern of proliferation and degree of pigmentation, nevi with this pattern have been referred to as nevi with an atypical epithelioid cell component (N-AECC). When N-AECC display sheet-like or an expansile nodular growth pattern, notable cytologic atypia, and any level of mitotic activity, they can be difficult to distinguish from melanoma. The clinical history and appearance of these lesions may similarly raise concern for melanoma. In view of this diagnostic problem, we present 28 cases of N-AECC from our dermatopathology consultation and in-house practice. All 28 cases were found to be negative on the basis of fluorescence in situ hybridization (FISH) for imbalanced chromosomal aberrations commonly found in melanoma. The clinical outcomes showed a benign clinical course for all cases for which the outcome information was available. FISH analysis also revealed that, in 4 of 28 cases (14%), the AECC of the lesion demonstrated polyploidy localized to the large epithelioid cell component. This is likely more common among cases of N-AECC that have an atypical spitzoid epithelioid cell component, particularly those with obvious senescent nuclear changes. Care must be taken to avoid the pitfall of misinterpreting these FISH findings as changes consistent with melanoma. The use of ancillary testing methods including FISH may be beneficial in improving the diagnostic accuracy involved in making the distinction of N-AECC from melanoma. Further, we report a novel finding of polyploidy seen in certain cases of benign N-AECC.     

Enhanced detection of spitzoid melanomas using fluorescence in situ hybridization with 9p21 as an adjunctive probe

The use of molecular diagnostic methods such as fluorescence in situ hybridization (FISH) for challenging melanocytic neoplasms is becoming more widespread. In light of the diagnostic difficulty they pose, spitzoid melanocytic neoplasms constitute an area of greatest potential utilization. In this study we wished to evaluate the sensitivity of the currently used melanoma FISH probe assay in a group of unambiguous spitzoid melanomas. On the basis of comparative genomic hybridization data, copy number losses at chromosome 9 have long been recognized as a frequent event in melanoma. In this study we wished to evaluate the efficacy of 9p21 as a potential FISH target and then evaluate the added benefit of reflexing the standard melanoma FISH assay, with FISH targeting 9p21 and the centromere of chromosome 9 (Cep9). Cep9 was included as a control to exclude inadequate hybridization or truncation as a source of homozygous deletions. We first studied a training set of 85 melanomas and 58 nevi with FISH targeting 9p21 and Cep9. As per previous methodology, 30 cells were enumerated. In the training set, the nevi had a mean number of cells with homozygous 9p21 loss of 0.97, with a standard deviation of 1.26. The melanomas had a mean of 7.1 and a standard deviation of 6.76. This difference was significant (P=2×10(-12)). On the basis of the training set, we identified a cutoff of 10 homozygous deletions to distinguish between melanoma and nevi. In a subsequent validation set consisting of 76 melanomas and 88 nevi, we found this cutoff to have a sensitivity of 33% and a specificity of 100%. Finally, in our group of 43 unequivocal spitzoid melanomas, standard FISH against chromosomes 6 and 11 was 70% sensitive. Homozygous 9p21 loss was present in 11 of 27 (41%) cases tested. By combining the standard melanoma FISH assay with the 9p21 FISH assay, a combined sensitivity of 85% was found. Among these 27 cases tested with 9p21, there were 7 cases that were negative by the standard melanoma FISH assay but that were positive by 9p21, suggesting that the 9p21 assay may be highly complementary to the standard melanoma FISH assay. Hence, in this study, we validated the efficacy of 9p21/Cep9 as a diagnostic FISH assay in melanoma, and demonstrated its complementary effect to the standard FISH assay. 9p21 may be particularly helpful in lesions with spitzoid morphology.     

Acquired Melanocytic Lesions and the Decision to Excise: Role of Color Variegation and Distribution as Assessed by Dermoscopy

BACKGROUND: Because melanoma may sometimes be difficult to differentiate from nevi with clinical atypia, many benign lesions also undergo surgical removal. OBJECTIVE: To assess color type and distribution in dermoscopic melanocytic lesion images and to analyze the influence of color parameters on the diagnostic process and the decision to excise. METHODS: Overall, 603 images, referring to 112 melanomas and 491 nevi, were retrospectively subdivided into four groups: "clearly benign," "follow-up," "dermoscopic atypical nevi," and "dermoscopic melanomas," according to their dermoscopic aspects. The frequency of color type, number, and asymmetry were evaluated on digital images. RESULTS: With respect to lesions not eligible for excision according to dermoscopy (but excised for cosmetic reasons), those excised with a suspicion of malignancy showed a higher number of colors, whose distribution was also more asymmetric. Moreover, the frequency of the presence of black and blue-gray progressively increased from clearly benign lesions to atypical nevi and dermoscopic melanomas. CONCLUSION: In dermoscopic images, color parameters are essential elements for the diagnosis of atypical nevus, which can be differentiated from both a clearly benign lesion and a melanoma. Furthermore, pigmentation asymmetry and the presence of blue-gray represent the main color features, which should lead to the decision to excise.     

Combined melanocytic nevi: histologic variants and melanoma mimics

Combined melanocytic nevi are composed of 2 or more distinct populations of nevomelanocytes. Most commonly used to describe the combination of blue nevi with common nevi, it may also be applied to other combinations of benign melanocytic proliferations, including Spitz nevi and nevi with deep dermal pigmented nevomelanocytes. We report the incidence and distribution of these tumors at the Massachusetts General Hospital over the past decade and review guidelines for diagnostic criteria and nomenclature. Between 2000 and 2010 we identified 511 cases of combined nevi, represented by 4 histologically distinct diagnostic categories: (1) blue nevus, (2) nevi with deep dermal pigmented nevomelanocytes (plexiform/deep penetrating, inverted type A/clonal), (3) Spitz or pigmented spindled cell nevus, combined with another type of nevus (usually common or dysplastic), and (4) other combinations including 2 or more nevus types. Nearly one fifth of these tumors displayed atypical features; atypia was observed more often in combined nevi with Spitz or deep pigmented elements (26 of 55, 47%, and 25 of 98, 26%, respectively) than in combined common and blue nevi (37 of 336, 11%). Clinical follow-up data were available for 83% of the patients with atypical combined nevi; none developed recurrence or metastasis with a mean follow-up of over 4 years.     

Digital Dermoscopic Monitoring of Atypical Nevi in Patients at Risk for Melanoma

BACKGROUND: Atypical nevi are a common risk factor for melanoma. OBJECTIVES: The objective was to determine the utility of monitoring dermoscopic photographs of atypical nevi in a high-risk population. METHODS: Over a 4.5-year period, digital dermoscopic photographs were taken of clinically atypical nevi at initial and follow-up visits, such that side-by-side comparisons could be made. RESULTS: A total of 5,945 lesions were monitored in 297 patients over 3 to 52 months (median, 22 months), and 324 lesions were biopsied. Photographic (dermoscopic) changes were noted in 96 of 5,945 (1.6%) lesions, which included 64 dysplastic nevi (67%), 25 common nevi (26%), and 1 melanoma (1.0%). Of 6 melanomas biopsied during the follow-up period, only 1 was detected by dermoscopic photographic change at follow-up. CONCLUSIONS: Most clinically atypical melanocytic nevi are stable over time, and lesions exhibiting dermoscopic changes are most likely to be dysplastic nevi. Although dermoscopy is a useful tool for clinical examination, the sensitivity of dermoscopic monitoring is limited by melanomas that may arise in normal skin or in clinically benign nevi that were not initially photographed.     

Pediatric melanomas and the atypical spitzoid melanocytic neoplasms

Cutaneous malignancies in the pediatric population are rare. Melanocytic neoplasms have garnered increased attention as the incidence of melanoma rises and as published analyses of biologically indeterminate lesions become more commonplace. Pediatric melanomas have been studied in several large cohort series; still, most of our assumptions for treatment stems from research in the adult population. Many clinicians speculate that pediatric melanomas may be biologically different from the same histological entity in adults given observed differences in metastatic potential and overall outcomes in children. Even more confounding are the atypical spitzoid lesions, which continue to spark debate in the oncology and dermatopathology literature with respect to classification, malignant potential, and recommended treatment course. In this article, recent literature addressing both atypical spitzoid melanocytic neoplasms and melanoma in the pediatric population is discussed.     

Desmoplastic (sclerotic) nevus: an underrecognized entity that resembles dermatofibroma and desmoplastic melanoma.

Desmoplastic (sclerotic) nevus, a benign melanocytic neoplasm characterized by predominantly spindle-shaped nevus cells within a fibrotic stroma, can be confused with fibrous lesions and other melanocytic proliferations, including desmoplastic melanoma. We compared the histologic and immunohistochemical features of 16 desmoplastic nevi, nine desmoplastic melanomas, four hypopigmented blue nevi, and six dermatofibromas. The similarities between desmoplastic nevi and dermatofibromas included epidermal hyperplasia (12 of 16), presence of keloidal collagen (15 of 16), hypercellularity (16 of 16), and increased numbers of factor XIIIa-positive dendritic cells (12 of 12). The absence of adnexal induction (0 of 16), the rarity of lesions with multinucleated cells (3 of 16) or epidermal hyperpigmentation (2 of 16), and the presence of S-100 immunoreactivity (16 of 16) and melanocytic proliferation (9 of 16) helped differentiate desmoplastic nevi from dermatofibromas. The similarities between desmoplastic nevi and desmoplastic melanomas included the presence of atypical cells (16 of 16) and HMB-45 expression in the superficial portion of the lesions (11 of 16). The infrequent location on the head or neck (1 of 16), the absence of mitotic figures (0 of 16), a significantly lower number of Ki-67-reactive cells, and a decrease in HMB-45 expression in the deep area of the lesions (8 of 11) helped distinguish desmoplastic nevi from desmoplastic melanoma. Desmoplastic nevi had overlapping features with hypopigmented blue nevi, but features tending to favor the latter included a predominance of ovoid nuclei, higher numbers of atypical cells, and homogeneous staining with HMB-45. We conclude that a combination of histologic and immunohistochemical criteria facilitates the reliable diagnosis of desmoplastic nevus from its simulators.     

Immunohistochemical Expression of p16, p21, p27 and Cyclin D1 in Oral Nevi and Melanoma

The acquisition of abnormalities at G1/S is considered a crucial step in the genesis and progression of melanoma. The expression of cell cycle regulators has also been used in various neoplasms as an adjunct to diagnosis. The aim of this study was to compare the expression of p16, p21, p27 and cyclin D1 in oral nevi and melanomas. Expression of these cell cycle regulatory proteins was evaluated by immunohistochemistry in 51 oral melanocytic lesions, including 38 intramucosal nevi and 13 primary oral melanomas. p16 and p27 were highly expressed in intramucosal nevi, whereas p21 and cyclin D1 expression was higher in oral melanomas. The results indicate that p21 and cyclin D1 may be involved in the development of oral melanomas, and eventually they may be useful in the differential diagnoses of oral benign and malignant melanocytic lesions.     

A highly specific and discriminatory FISH assay for distinguishing between benign and malignant melanocytic neoplasms

The diagnosis of certain melanocytic proliferations remains one of the most challenging areas in pathology. In recent times, fluorescence in situ hybridization (FISH) has emerged as a promising diagnostic aid to conventional microscopy. We previously showed that a 4-probe FISH assay targeting 6p25 (RREB1), 6q23 (MYB), Cep6 (centromere 6), and 11q13 (CCND1) could discriminate between histologically unequivocal melanomas and benign nevi with a sensitivity of 86.7% and specificity of 95.4%. However, the sensitivity of the assay is approximately 70% in melanomas with spitzoid morphology. Furthermore, differentiating true gains from tetraploidy may cause difficulties in interpretation by inexperienced examiners. Here we refine the current probe set to better target spitzoid melanomas and more easily distinguish cells with imbalanced copy number aberrations from tetraploid cells. Using FISH data from 3 training sets of 322 tumors, including 152 melanomas and 170 nevi, we identified 9p21, 6p25, 11q13, and 8q24 as a probe set with improved discriminatory power in differentiating melanomas from nevi. In a validation set of 51 melanomas and 51 nevi this probe set had a sensitivity of 94% and specificity of 98%, compared with the original probe set that had a sensitivity of 75% and specificity of 96% in the same validation cohort. We propose that by incorporating 9p21 into the 4-probe FISH assay, with a new diagnostic algorithm, this new probe set would have improved discriminatory power in melanocytic neoplasms and improved sensitivity for detecting spitzoid melanomas, as demonstrated by our previous studies.