Dermoscopic Features of Mucosal Melanosis

ORIGINAL ARTICLE

Dermoscopic Features of Mucosal Melanosis F RANCESCA M ANNONE , MD, n V INCENZO D E G IORGI , MD, n A NTONIO C ATTANEO , MD, w D ANIELA M ASSI , MD, w A NGELINA D E M AGNIS , MD, z AND PAOLO C ARLI , MD n Department of Dermatology, zDepartments of Obstetrics and Gynecology, and wDepartment of Human Pathology and Oncology, University of Florence, Florence, Italy

n

BACKGROUND. Melanosis (lentiginosis, labial melanotic macula) is a benign pigmented lesion of mucosa characterized by pigmentation of basal keratinocytes with melanocytic normal or slightly increased in number. Melanosis, particularly when occurring on genitalia, can clinically mimic mucosal melanoma thus creating concern in both the patient and the physician. OBJECTIVE. In this study dermoscopic features from a series of clinically equivocal (n 5 11) or clinically typical (n 5 10) mucosal melanosis were analyzed. METHODS. All the women consecutively seen at the Vulva Clinic of the Department of Obstetrics and Gynecology, University of Florence, Italy, from May 1, 2002 to June 30, 2002, were examined. RESULTS. Three major dermoscopic patterns were identified: (1) a ‘‘structureless’’ pattern, predominantly found in clinically

equivocal vulvar melanosis, with a blue hue, associated with the presence of melanophages in the upper dermis, present in the majority of these lesions; (2) a ‘‘parallel pattern,’’ often found in clinically typical melanotyc macules of the lips and penis; and (3) a ‘‘reticular-like’’ pattern associated with clinically equivocal melanosis occurring at peculiar sites such as the areola (all the three cases occurred at that site) or, rarely, on the lip. CONCLUSIONS. Dermoscopy can play a role in the noninvasive classification of mucosal melanosis. The risk of misclassification with melanoma is probably dependent on dermoscopy pattern shown by the lesion. Prospective studies including early melanomas are needed to establish diagnostic performance of dermoscopy in pigmented lesions of the mucosa.

FRANCESCA MANNONE, MD, VINCENZO DE GIORGI, MD, ANTONIO CATTANEO, MD, DANIELA MASSI, MD, ANGELINA DE MAGNIS, MD, AND PAOLO CARLI, MD HAVE INDICATED NO SIGNIFICANT INTEREST WITH COMMERCIAL SUPPORTERS.

MELANOSIS (LENTIGINOSIS, labial melanotic macula) is a benign pigmented lesion that can occur in both the genital and the oral mucosa.1–3 Despite its benign behavior, on clinical grounds vulvar melanosis can show overlapping features with malignant melanoma at times being characterized by asymmetry, irregular borders, multifocality, variegated pigmentary patterns, and large size.4 In doubtful cases, which appear clinically alarming and are frequently misinterpreted as possible melanoma by nondermatologists, diagnosis of melanosis by clinical criteria alone may be unreliable.5,6 Therefore, in such cases, biopsy and histologic examination are considered necessary.5 In recent years, the use of dermoscopy (epiluminescence microscopy, surface microscopy) has been demonstrated to improve the diagnostic accuracy of nearly all pigmented skin lesions.7–11 Nevertheless, little is known about the possible role of dermoscopy in the diagnosis of mucosal melanosis.12 A previous study found that a diffuse pigmentation, without any other dermoscopic features related to melanocytic proliferAddress correspondence and reprint requests to: Francesca Mannone, MD, Department of Dermatology, University of Florence, Via degli Alfani, 37, 50121 Firenze, Italy, or e-mail: [email protected].

ation (‘‘structureless pattern’’), was the hallmark of melanosis.13 In practice, however, some cases do not fit the above-mentioned profile, showing additional features that remained to be classified. More recently, another dermoscopy pattern, depicted as a ‘‘parallel pattern’’ because of linear or partially curve pigmentary streaks arranged in a parallel manner, has been applied to mucosal melanosis.6,14 This study had two aims: first, to evaluate the frequency of genital melanosis in a group of patients consecutively seen in a specialistic vulva clinic, and second, to analyze the dermoscopic features of melanoses occurred at different anatomic sites (vulva, penis, lip, areola) with an attempt to investigate their histopathologic correlates.

Materials and Methods

Frequency of Genital Melanosis in a Specialistic Vulva Clinic All the women consecutively seen at the Vulva Clinic of the Department of Obstetrics and Gynecology, University of Florence, Italy, from May 1, 2002, to June 30, 2002, were examined by a last-year resident in dermatology (FM) experienced in the diagnosis of pigmented cutaneous lesions to investigate the presence

r 2004 by the American Society for Dermatologic Surgery, Inc.
Published by Blackwell Publishing, Inc. ISSN: 1076-0512/04/$15.00/0
Dermatol Surg 2004;30:1118–1123

Dermatol Surg

30:8:August 2004

MANNONE ET AL.: DERMOSCOPIC FEATURES OF MUCOSAL MELANOSIS

of genital and/or oral pigmented lesion. Also the presence of nail pigmentation, in particular the so-called melanonichia striata, was investigated. Predefined, clinical diagnosis included melanosis, melanocytic nevus, melanoma, seborrheic keratosis, and vascular lesion (hemangioma, angiokeratoma). Patients with clinically equivocal lesions were referred to the Pigmented Lesions Clinic of the Department of Dermatology of the University of Florence, for second-level examination by means of dermoscopy and if necessary, biopsy. The patients were questioned with regard to their symptoms, awareness of the lesions, and family history of melanoma.

1119

variable epidermal thickness (slightly atrophic, normal to slightly hyperplastic). In all cases the histopathologic features were diagnostic of melanosis. Dermoscopic photographs of each lesion (more than one image in the case of large lesions) were taken using a device mounted on Nikon camera (Dermaphot Pabish, Germany) and observed—for the purpose of the study—by the two examiners by means of a viewer. Dermoscopic features were examined and described in accordance with the standardized terminology (Hamburg Consensus Conference 1989).15 Table 1 summarizes the major dermoscopic criteria with their histologic correlates.

Dermoscopic Features of Mucosal Melanosis An investigation of dermoscopic features of mucosal melanosis was performed by two experienced observers (PC, FM) in a series of histologically confirmed melanosis. The entry criterion was that of a lesion which was considered equivocal from a clinical point of view (i.e., asymmetric, with irregular outline, intensively black or variegated pigmentation, larger than 6 mm). Eleven patients (median age 51 years, range 23–80 years) with clinically equivocal melanosis of the vulva and 3 patients with clinically equivocal melanosis of the nipple (women, 28, 32, and 40 years old, respectively) were included in the study. Two patients with vulvar melanosis belonged to the survey reported above. All the lesions were asymmetric in shape and larger than 6 mm in diameter and they showed irregular borders and intensively black or variegated pigmentation. In all cases, a 6-mm punch biopsy had been performed in the most heavily pigmented area to obtain a diagnostic confirmation. For completeness, also a series of clinically typical mucosal melanosis (small in diameter, symmetric, light brown, homogeneously pigmented macule, often multiple in number) was included (eight cases of melanosis of the lip and two cases of penile melanosis). Histologic examination was obtained in three of these cases (one of the lip, two penile melanosis). All the histologic slides were reviewed by an experienced pathologist (DM). Overall, histopathologic examination demonstrated prominent hyperpigmentation of basal keratinocytes with presence of scattered dermal melanophages. The number of melanocytes was normal or slightly increased. In some cases, however, melanocytes showed prominent dendritic processes containing melanin pigment. The epidermal thickness was variable. Some cases showed epidermal acanthosis, with preservation of the normal profile of the rete ridges, whereas other cases displayed lack of the normal profile of the rete ridges, associated with

Results

Prevalence of Vulvar Melanosis A total of 170 women were consecutively examined. The age ranged from teenage to elderly (19–92 years), but most were in the fifth and sixth decades of life (mean age 58  13.5 years). This group cannot obviously be considered as a sample of general population because visited at a tertiary referral center; moreover, such patients referred to the clinic complaining for lesions/symptoms related the vulvar region. According to the diagnosis previously made by the physician suggesting specialized consultation, the major reasons for referral were vulvar lichen sclerosus (44%), vulvovaginal infections (23%), lichen simplex (5.2%), vulvar intraepithelial neoplasia (4.7%), vulvodinia (3.5%), carcinoma of the vulva (first observation or follow-up) (3.5%), sebaceous cyst (1.7%), psoriasis (1%), genital warts (1%), and others (12.4%). Only two patients referred to the vulva clinic for the presence of a pigmented lesion. According to dermatologic examination, 33 women (19%) had pigmented lesions on the vulva. These lesions were a vulvar melanosis (13/33 cases), melanocytic nevi (6 cases), angiokeratomas or other benign vascular lesion (12 cases), and seborrheic keratoses (2 cases). Only 24% of the patients were aware of the presence of the lesions. Concerning patients with vulvar melanosis, none showed additional pigmentary alterations in other cutaneous sites, melanonichia, or other clinically evident abnormalities of pigmentation. The majority of melanoses (11/13) were small (o 5 mm), flat, evenly pigmented, and well circumscribed. Only 2 were equivocal from clinical examination and were referred for dermoscopy and histopathologic verification. Among the 6 melanocytic nevi, only 1 had clinical atypia (larger than 1 cm and variegated in pigmentation). This lesion was excised and histopathologic

1120

Dermatol Surg

MANNONE ET AL.: DERMOSCOPIC FEATURES OF MUCOSAL MELANOSIS

30:8:August 2004

Table 1. Epiluminescence Microscopy Criteria and Histologic Correlates for Pigmented Skin Lesions Epiluminescence Microscopy Criterion

Definition

Pigment network Diffuse pigmentation (blotches) Brown globules

Network of brownish lines over a background tan Pigmentation that preclude recognition of other criteria Round, oval, or spherical bodies

Black dots Radial streaming

Small, punctate and black structures Linear, brown to black streaks radiating from the Bulbous, often kinked projections, directly connected Whitish film overlying a more darkly pigmented area Circumscribed zones that have a gray and/or blue hue Depigmented areas that appear as dead white or light Areas of relatively lighter pigmentation Linear, dotted, or globular red structures

Pseudopods Whitish veil Gray-blue areas White areas Hypopigmented areas Vascular pattern Horny pseudocystis Pseudofollicular openings Red-blue areas Maple leaf-like areas Pseudopigment network

Histologic Correlates Pigmented rete ridges Melanin at all levels of the epidermis and/or the dermis Nests of pigmented melanocytes at the dermal–epidermal junction and/or in the papillary dermis or clusters of melanophages in the papillary dermis Focal collections of melanin in the stratum corneum Radially arranged pigmented nests border of the lesions Radially arranged pigmented nests to the body of the lesion or to the pigmented network Compact orthokeratosis and hypergranulosis Fibrosis and pigmented melanophages or melanocytes in a thickened papillary or reticular dermis Lack of melanin and fibroplasias pink patches

Circular whitish-yellow areas Comedo-like openings Red-blue, sharply demarcated areas Maple leaf-like, light to dark brown areas with branching or bud-like arrangement A grid of large, roundish, brown meshes

Reduced amount of melanin Neovascularization or vascularized nests of amelanocytic cells Intraepidermal horn globules underneath the surface Intraepidermal horn globules reaching the surface Dilated vascular spaces in the papillary dermis Pigmented epithelial nodules Melanotic pigment arranged around sebaceous follicles, quite numerous on the face

Table 2. Frequency of the Three Dermoscopic Patterns Observed, Compared to Different Mucosal Sites

Vulvar melanosis (clinically equivocal) Penile melanosis (clinically typical) Melanosis of the lip (clincally typical) Melanosis of the areola (clinically equivocal)

Parallel Pattern

Structureless Pattern

Reticular-like Pattern

Total

1 1 5 0

10 1 2 0

0 0 1 3

n 5 11 n52 n58 n53

examination showed a melanocytic proliferation with architectural disorder and cytologic atypia.

Dermoscopic Analysis Three major dermoscopic patterns were identified (Table 2). First was a parallel pattern, characterized by regularly distributed pigmentation linearly arranged according to skin profile (furrows and reliefs). This pattern was mainly found in clinically typical, small in diameter, melanosis of the lip (5/8) and 1 of 2 penile melanoses (Figure 1). It was also found in 1 of 11 vulvar melanosis (Figures 2A and 2B). A parallel pattern was characterized histologically by prominent

Figure 1. Typical melanosis of the penis showing dermoscopically a typical parallel pattern: linear or partially curved pigmentary streaks arranged in a parallel manner.

Dermatol Surg

30:8:August 2004

MANNONE ET AL.: DERMOSCOPIC FEATURES OF MUCOSAL MELANOSIS

1121

Figure 2. (A) Vulvar melanosis clinically mimicking mucosal melanoma. (B) Dermoscopic picture of the same lesion showing a parallel pattern characterized by linear or partially curved pigmentary streaks arranged in a parallel manner. Pigmentation appeared more irregular than that observed in clinically typical melanosis of the lip or of the penis. (C) Histopathologic examination showing a prominent hyperpigmentation of the basal epidermal layer with accentuation at the tips of elongated and clubbed rete ridges, in a pattern similar to that described for so-called ink spot lentigo.

hyperpigmentation of the basal epidermal layer with accentuation at the tips of elongated and clubbed rete ridges, in a pattern similar to that described for socalled ‘‘ink spot lentigo’’ (acquired reticulated lentigo) (Figure 1C). Second was a structureless pattern, characterized by diffuse pigmentation ranging from a light brown to a dark brown hue, seldom irregularly distributed, with possible presence of gray-blue color at the center of the lesion (Figure 3). Such a pattern has been mainly found in clinically equivocal, large in diameter melanosis of the vulva (10/11). In 3 of these cases, the pigment was forming a cobblestone-like pattern following the cutaneous profile in that particular site. In any case, neither a definite pigment network nor other dermoscopy parameters were found (Figures 4A and 4B). Histologically, a moderate to marked hyperpigmentation was observed along the basal cell layer, with no significant differences in distribution between the top and the bottom of the rete ridges (Figure 4C). The blue hue, present in the majority (6/11) of clinically equivocal melanosis of the vulva was associated histologically with the presence of melanophages in the upper dermis (Figures 4B and 4C). The third pattern, we called ‘‘reticular-like,’’ showed a honeycomb appearance similar to the well-established ‘‘pigment network’’ expected in lesions characterized by melanocytic proliferation we found this pattern in all the three melanoses of the areola and in one melanosis of the lip which appeared intensively black and occurred in an old men (Figures 5A and 5B). The finding of a prominent pigment network on histopathologic examination correlates with the presence of melanin pigment in the epidermal basal cells. The holes of the network corresponded to the tips of the dermal papillae whereby the lines of the network resulted from the projection of the pigmented rete ridges to the skin surface. The normal profile of the rete ridges was lost, the epidermis being acanthotic, with thick hyperpigmented epidermal crests coalescing at the base (Figure 5C). Such feature, indeed, was not

Figure 3. Example of a vulvar melanosis showing dermoscopically a structureless pattern: diffuse pigmentation without pigment network, globules, and streaks.

visible in other mucosal site probably to the flattened epidermis.

Discussion From our data, 19% of women consecutively observed at a specialized vulva clinic had pigmented lesions in the vulvar area. This percentage is slightly greater than the 12.3% of patients considered in the study conducted by Rock et al.16 The reason for this discrepancy is probably due to the different ethnic origin of examined populations: our case series was composed of white women, whereas the other group studied was constituted by 84% white women (11% black, 5% oriental) and the pigmented lesions occurred only in the group of white women.16 In both studies, interestingly, melanosis was the most common diagnosis The origin of mucosal melanosis is unclear, and its similarity to labial melanotic macules is striking. Many theories related to the possible cause of labial melanotic macules have been proposed, but the exact etiology and pathogenesis of these alterations remain to be known. The excessive pigmentation could be secondary to a chronic stimulus in the area or due to a defect

1122

MANNONE ET AL.: DERMOSCOPIC FEATURES OF MUCOSAL MELANOSIS

Dermatol Surg

30:8:August 2004

Figure 4. (A) Diagnostically equivocal vulvar melanosis (asymmetric lesion, irregular outline, variegated pigmentation, larger than 6 mm). (B) Dermoscopic picture of the same lesion showing a structureless pattern characterized by a irregularly distributed pigmentation forming a cobblestone-like pattern without any other dermoscopic features among those related to melanocytic proliferation (i.e., pigment network, globules, streaks). Blue areas are focally present in the center of the lesion. (C) Histopathologic examination showing hyperpigmentation of basal keratinocytes. The number of epidermal melanocytes was slightly increased and some of them showed prominent dendritic processes containing melanin pigment. Note the presence of dermal melanophages (histopathologic correlate of the blue areas).

Figure 5. (A) Melanosis of the areola clinically mimicking melanoma. (B) Dermoscopic picture of the same lesion showing the reticular-like pattern characterized by an honeycomb appearance similar to the well-established pigment network found in lesions characterized by melanocytic proliferation. (C) Histopathologic examination: the epidermis is markedly acanthotic, with thick hyperpigmented epidermal crests coalescing at the base. Scattered dermal melanophages are also present.

in the normal transport of melanine to suprabasal keratinocytes.17,18 A peculiar form of genital pigmentation, which however, can show atypical melanocytes, is the so-called psoralen and long-wavelength ultraviolet radiation-induced melanotic macule seldom reported on the penis.19 A clinically relevant finding is that mucosal melanosis may sometimes share clinical features with malignant melanoma representing a possible cause of diagnostic concern. This event happened in 2 of 13 cases of melanosis observed during our survey in a vulva clinic, which therefore required biopsy for diagnostic verification. In this particular field, the frequent lack of information about the history of the lesion may increase the diagnostic difficulties.17 In this view, dermoscopy may help in the proper management of these lesions, providing additional diagnostic information beyond those provided by clinical examination. Previous study showed that mucosal melanosis present a dermoscopy pattern defined as structureless,13 characterized by a diffuse, brown to black pigmentation, seldom irregularly distributed, without neither pigment network nor globules and/or streaks. Because melanosis is due to hyperpigmentation of the epidermis, with melanocytes normal or slightly increased in number,1–3 it is associated with functional abnormality more than with proliferation of melanocytes.19 Accordingly, neither pigment network nor brown globules and streaks are expected to be

found because all these variables require an increased number of melanocytes arranged in nests as specific histopathologic correlates.11 The lack of dermoscopic variables associated with melanocyte proliferation can be therefore considered peculiar of melanosis and it may help in differentiate melanosis from both nevi and melanoma.13 This study gives new insights in the attempt of noninvasive analysis of melanosis. To summarize, three well-established dermoscopic patterns of mucosal melanosis have been defined: 1. The structureless pattern, previously reported, frequently showed by vulvar melanosis. The clinician should keep in mind that gray-blue areas can be seldom found in the center of the lesion. The histopathologic correlate is the presence of dermal melanophages. Because of the well-known diagnostic value of the blue hue in the diagnosis of malignancy by dermoscopy,20 a careful examination of these lesions should be made to minimize any risk of misclassification with melanoma. 2. The parallel pattern, frequently found in clinically typical melanotic macules of the lips and penis. Both pattern do not show any of dermoscopy features that can be found in nevi and melanoma as a pigment network, globules, and streaks, whose histopathologic correlate is melanocyte proliferation.

Dermatol Surg

30:8:August 2004

MANNONE ET AL.: DERMOSCOPIC FEATURES OF MUCOSAL MELANOSIS

3. The reticular-like pattern. Interestingly, such feature has been only found in melanosis occurring at peculiar sites such as the areola (all the three cases) or, rarely, on the lip (one case only), whereas it was not visible in other mucosal sites probably owing to the flattened epidermis. Indeed, histologically, in the reported cases the normal profile of the rete ridges was lost, the epidermis being acanthotic, with thick hyperpigmented epidermal crests coalescing at the base. This is probably the histopathologic prerequisite for having the honeycomb design visible at mucosal surface. For diagnostic purposes, only the first two patterns can be confidently used in the noninvasive analysis of melanosis because of the low risk of misclassification with a melanoma. According to data obtained in ‘‘thin,’’ i.e., less than 1 mm in thickness, cutaneous melanomas, a pigment network is found in 95% of cases, brown globules in 78%, and streaks in 83%; these percentage changes with lesion’s thickness.21 Therefore, facing a mucosal melanosis with a structureless or parallel pattern, the lack of any other additional parameter helps clinician to exclude the diagnosis of malignancy. In contrast, dermoscopy cannot give a reliable classification of a melanosis once a reticular-like pattern is shown. Indeed, the honeycomb features observed in our cases looked rather similar to those found in melanoma, i.e., a large grid and a prominent network that abruptly ends at the periphery.7–10 Compared to a melanoma, however, the grid and holes distribution were more regularly distributed. In such a case, an incisional biopsy should be considered for diagnostic confirmation. In conclusion, melanosis is the most frequent pigmented lesion found in vulvar area. Some cases represent a diagnostic task because they show clinically equivocal features requiring biopsy for diagnostic verification. Dermoscopy may play a role in improving a noninvasive classification of mucosal melanosis. As previously made with cutaneous lesions, however, a formal assessment of diagnostic accuracy of dermoscopy in mucosal lesions including early melanomas should be awaited before basing the final lesion management on dermoscopy examination in routine practice.

1123

References 1. Barnhill RLJ, Albert LS, Shama SK, et al. Genital lentigginosis: a clinical and histopatologic study. Am Acad Dermatol 1990;22: 453–60. 2. Lenane P, Keane CO, Connell BO, Loughlin SO, Powell FC. Genital melanotic macules: clinical, histologic, immunohistochemical, and ultrastructural features. J Am Acad Dermatol 2000;42:640–4. 3. Kenneth K, Ho L, Dervan P, O’Loughlin S, Powell FC. Labial melanotic macule: a clinical, histopathologic, and ultrastructural study. J Am Acad Dermatol 1993;28:33–9. 4. Sison-Torre EQ, Ackerman AB. Melanosis of the vulva: a clinical simulator of malignant melanoma. Am J Dermatopathol 1985;7(Suppl):51–60. 5. Jaackson R. Melanosis of the vulva. J Dermatol Surg Oncol 1984;10:119–21. 6. Argenziano G, Soyer HP, De Giorgi V, et al. Atlante Interattivo di Demoscopia. Milan: EDRA Medical Publishing and New Media, 2001. 7. Rao BK, Marghoob AA, Stolz W, et al. Can early malignant melanoma be differentiated from atypical melanocytic naevi by in vivo techniques? Part I. clinical and dermoscopic characteristics. Skin Res Technol 1997;3:8–14. 8. Goldman L. Some investigative studies of pigmented nevi with cutaneous microscopy. J Invest Dermatol 1951;16:407–26. 9. Steiner A, Pehamberger H, Wolf K. In vivo epiluminescence of pigmented skin lesions II Diagnosis of small pigmented skin lesions and early detection of malignant melanoma. J Am Acad Dermatol 1987; 17:584–91. 10. Pehamberger H, Binder M, Steiner A, et al. In vivo epiluminescence microscopi: improvement of early diagnosis of melanoma. J Invest Dermatol 1993;100:356s–362s. 11. Yadav S, Vossaert KA, Kopf AW, et al. Histopathological correlates of strctures seen on dermoscopy (epiluminescence microscopy). Am J Dermatopathol 1993;15:297–305. 12. De Giorgi V, Massi D, Carli P. Dermoscopy in the management of pigmented lesions of the oral mucosa. Oral Oncol 2003;39:534–5. 13. Carli P, De Giorgi V, Cattaneo A, Giannotti B. Mucosal melanosis clinically mimicking malignant melanoma: non-invasive analysis by epiluminescence microscopi. Eur J Dermatol 1996;6:434–6. 14. Stolz W, Braun-Falco O, Bilek P, Cognetta AB. Color Atlas of Dermoscopy. Boston: Blackwell Scientific, 2002. 15. Bahmer F, Fritsch P, Kreusch J, et al. Terminology in surface microscopy. J Am Acad Dermatol 1990;23:1159–62. 16. Rock B, Hood AF, Rock JA. Prospective study of vulvar nevi. J Am Acad Dermatol 1990;22:104–6. 17. Veraldi S, Cavicchini S, Benelli C, Gasparini G. Laugier-hunziker sindrome: a clinical histopathological and ultrastructural study of 4 cases and review of the literature. J Am Acad Dermatol 1991;25: 632–6. 18. Chapel T, Taylor RM, Pinkus H. Volar melanotic macules. Int J Dermatol 1979;18:222–5. 19. Rhodes AR, Harrist TJ, Momtaz TK. The PUVA induced pigmented macule: a lentiginous proliferation of large, sometimes cytologgically atypical, melanocytes. J Am Acad Dermatol 1983;9:47–58. 20. Massi D, De Giorgi V, Carli P, Santucci M. Diagnostic significance of the blue hue in dermoscopy of melanocytic lesions: a dermoscopic-pathologic study. Am J Dermatopathol 2001;23:463–9. 21. Argenziano G, Fabbrocini G, Carli P, De Giorgi V, Delfino M. Epiluminescence microscopy: criteria of cutaneous melanoma progression. J Am Acad Dermatol 1997;37:68–74.