- Email: [email protected]
Is melanocytic nevus with focal atypical epithelioid components (clonal nevus) a superficial variant of deep penetrating nevus?
DERMATOPATHOLOGY
Is melanocytic nevus with focal atypical epithelioid components (clonal nevus) a superficial variant of deep penetrating nevus? Whitney A. High, MD,a Kenneth W. Alanen, MD,b and Loren E. Golitz, MDa Denver, Colorado, and Edmonton, Alberta, Canada Background: We compare features of two similar nevi which may be confused with melanoma: deep penetrating nevus (DPN), and nevus with focal atypical epithelioid component (NFAEC). Methods: Clinical/demographic and histologic information regarding 146 DPN and 81 NFAEC were compared. Patient outcomes were ascertained via a questionnaire solicited to the referring physicians. Results: Clinical features were similar for each type of nevus. Histologically, all lesions demonstrated identical aggregates of epithelioid melanocytes. DPN more often extended into the deep dermis or subcutaneous tissue, while NFAEC was typically confined to the superficial dermis. NFAEC more often demonstrated a junctional component (P \ .001) and coexistent common nevus with congenital features (P = .035). DPN more often demonstrated adnexal spread (P \.001). A subgroup with overlapping features was identified. With an average of 6.4 years follow-up, there were just two recurrences, and no metastases. Limitations: This retrospective study utilized tissue specimens from a single reference laboratory; ergo, some inherent selection bias exists. Specimens were randomly selected for P53 immunostaining, leading also to potential sampling error. Conclusions: DPN and NFAEC show considerable similarity, differing mainly in the depth of extent for the lesion. It is possible that NFAEC represents a more superficial variant of DPN. A subgroup with overlapping features, but intermediate depth, supports such a relationship. ( J Am Acad Dermatol 2006;55:460-6.)
n 1989, Seab et al1 described 70 cases of a distinctive melanocytic lesion, for which they coined the term deep penetrating nevus (DPN). DPN demonstrates a predilection for the face, proximal extremities, and trunk, and typically occurs in the first four decades of life.1,2 In most respects, DPN is similar if not identical to a lesion Barnhill et al3 termed ‘‘plexiform spindle-cell nevus.’’ Histologically, DPN is symmetrical and demonstrates an inverted, wedge-shaped architecture with
I
From the Departments of Dermatology & Pathology,a University of Colorado Health Sciences Center, and the Division of Dermatology,b Department of Medicine, University of Alberta. Drs High and Alanen contributed equally to this work. Funding sources: None. Conflicts of interest: None identified. Reprints not available from the authors. Correspondence to: Whitney A. High, MD, Assistant Professor, Dermatology & Dermatopathology, University of Colorado Health Sciences Center, PO Box 6510, Mail Stop F703, Aurora, CO 80045-0510. E-mail: [email protected]. 0190-9622/$32.00 ª 2006 by the American Academy of Dermatology, Inc. doi:10.1016/j.jaad.2006.04.054
460
deep extension into the dermis or subcutis. Tracking of melanocytes along adnexal and neurovascular structures is characteristic. Spindled cells and epithelioid components may be present.4 While the nuclei of epithelioid cells in DPN are typically smaller and more hyperchromatic than true epithelial cells, they maintain a polygonal shape that has led to use of terminology similar to other melanocytic lesions, such as epithelioid blue nevus. In epithelioid blue nevus, melanocytes are often dispersed as single cells between collagen bundles, providing partial distinction from classic DPN.5 Significant overlap exists, and many have hypothesized that that before the description by Seab et al, many DPN were diagnosed as cellular blue nevi.1,2 Because of the deep extension, and the presence of epithelioid melanocytes with associated melanophages, DPN may also be confused with melanoma. Nevus with focal atypical epithelioid components (NFAEC) is likely synonymous with inverted type A nevus.3,6,7 This entity has also been referred to as a ‘‘clonal nevus,’’ akin to ‘‘clonal seborrheic keratoses.’’ In fact, the original series of NFAEC was submitted as
High, Alanen, and Golitz 461
J AM ACAD DERMATOL VOLUME 55, NUMBER 3
‘‘clonal nevi,’’ but the term NFAEC was suggested during peer review.6 While clonality is expected in nearly all common nevi,8,9 it has not been rigorously established within NFAEC. Currently, our research group is performing such an analysis to confirm or refute the appropriateness of this terminology, and to compare the apparent biphasic populations of melanocytes. Clinically, NFAEC are characterized by a focal dark area in an otherwise homogenous nevus.6,7 Such pigmentation, especially when noted abruptly, may raise clinical concern for melanoma. The hyperpigmented area corresponds histologically to a relatively well-circumscribed collection of heavily pigmented melanophages, in association with slightly atypical epithelioid melanocytes. These melanocytes have dusty brown cytoplasm and irregular nuclei, and are usually located within the superficial dermis. Despite some minor cytologic atypia, or more aptly a departure in morphology from surrounding banal nevus cells, metastases with NFAEC have not been reported.6 The events involved in development of NFAEC are largely unknown. Herein, we reviewed all cases diagnosed as DPN or NFAEC over a 13-year period at our referral laboratory. Demographic, clinical, and histologic features were identified, compiled, and analyzed. Retrospectively, we identified a subgroup of nevi that exhibited overlapping histologic features of DPN and NFAEC. We propose that NFAEC and DPN may be closely related entities. Our report comprises the largest study to date of both DPN and NFAEC, and confirms important findings from previous series, further documenting the apparently benign nature of these lesions.
Archived hematoxylin-eosin (H&E) stained slides were reviewed, without knowledge of the original diagnosis or clinical data. Immunohistochemical staining for p53 (Calbiochem, San Diego, Calif) was performed on 20 randomly selected nevi (10 NFAEC, 10 DPN) using the avidin-biotin complex method. Appropriate positive and negative controls were included. To provide an additional point of comparison, 227 consecutive nevi diagnosed at our laboratory were assessed for a congenital pattern, specifically that of deep extension and/or nests approximated to cutaneous adnexal structures, such as pilosebaceous units, sweat glands, neurovascular bundles, and arrector pili muscles. Data analysis was performed using SAS (Statistical Analysis System, SAS Institute Inc, Cary, NC). Pearson x2 tests were used to test for differences between demographic and histologic characteristics. Categories that contained only a few observations (\5) but with essentially identical percentages were not analyzed. For continuous variables, such as age at diagnosis or depth of extent, a standard T-test analysis was utilized. A significance level of .05 was used in all analyses.
RESULTS
All cases diagnosed as DPN* or NFAEC at our academic dermatopathology referral practice between 1986 and 1999 were reviewed and included. Cases demonstrating features of a dysplastic (Clark’s) nevus, blue nevus, or cellular blue nevus were excluded. Clinical data was obtained from referral letters and/or standardized forms that accompanied the consultation material. Questionnaires seeking follow-up information were mailed to the referring physicians. Specific inquiries regarding recurrence of the lesion and/or development of melanoma were included.
Clinical data The average age of patients with NFAEC was 25.4 years (range: 4-77 years). There were 48 women and 33 men (81 lesions in 81 persons). The most common clinical impressions for cases diagnosed as NFAEC were: atypical nevus (28%), melanoma (23%), benign nevus (23%), and ‘‘clonal nevus’’ (12%). The average age of patients with DPN was 28.3 years (range: 4-64 years). There were 85 women and 61 men (146 lesions in 146 persons). The most common clinical diagnoses for cases diagnosed as DPN were: blue nevus (28%), melanocytic nevus (27%), atypical nevus (23%), and melanoma (7.5%). The head, neck, and back were the most common locations for both types of nevi (Table I). Follow-up data was obtained via questionnaire for 82% of patients. The average follow-up was 6.4 years (range: 6 months-13 years). Two patients had developed a recurrence (one DPN, one NFAEC). Histologic review revealed that both lesions had involved surgical margins upon biopsy. No patient developed metastases and/or malignant melanoma.
*DPN was formally described in 1989 by Seab et al1 at the Armed Forces Institute of Pathology (AFIP) in Washington, DC. One of us (L. E. G.) had been introduced to the concept of DPN during fellowship training at the AFIP and therefore cases with this pattern had been coded for retrieval from the database prior to the published description.
Histologic data Deepest extent of lesion. One hundred fortysix lesions were initially diagnosed as DPN. Ninetyone DPN (62%) extended only into the middle to deep reticular dermis. Forty-one DPN (28%)
METHODS
462 High, Alanen, and Golitz
J AM ACAD DERMATOL SEPTEMBER 2006
Table I. Anatomical distributions of DPN and NFAEC* DPN Location
Head and neck Back Arm and shoulder Abdomen Leg and thigh Chest Buttock Foot Totals
Number
Table II. Histologic features of DPN and NFAEC Histologic feature
NFAEC %
Number
%
52 34 28 12 11 5 2 2
35 23 19 8 8 3 1 1
29 20 17 7 5 1 1 1
35 24 20 8 5 1 1 1
146
100
81
100
DPN, Deep penetrating nevus; NFAEC, nevus with focal atypical epithelial component. *Statistical analysis using a Pearson x 2 failed to demonstrate significant differences in the anatomical distribution of lesions, with P values ranging from 0.74 to 0.98.
Fig 1. Low-power photomicrographs demonstrate the overall architecture of a DPN (A) with symmetrical, roughly-wedge shaped architecture, and central bluntended penetrating lobules containing dermal melanophages, and a NFAEC (‘‘clonal’’ nevus) (B), manifesting as a symmetric and superficial silhouette with focal aggregates of larger, epitheloid melanocytes with dusty cytoplasm and associated dermal melanophages. (A and B, Hematoxylin-eosin stain; original magnification: A, 340; B, 3100).
extended into the subcutaneous fat. In 14 cases (10%), the deepest extent of the DPN could not be determined. Eighty-one lesions were diagnosed initially as NFAEC. In six NFAEC (7.4%), the deepest extent of the lesion could not be evaluated, but in all
DPN (N = 146)
NFAEC (N = 81)
Number with junctional 85 (58) 71 (88) component (%) Average depth of extent 2.38 0.74 in mm* (SD) Adnexal spread (%) 129 (88) 32 (39) Neurovascular spread 110 (75) 53 (65) Number with co-existing 123 (84) 76 (94) common nevus with congenital features (%)
Statistical significancey
P \ .001 — P \ .001 NS P = .035
DPN, Deep penetrating nevus; NFAEC, nevus with focal atypical epithelial component; NS, not significant (P [ .05); SD, standard deviation. *Depth of extent refers to the deepest extent of atypical epithelioid nevus cells within the tissue. y Statistical significance was calculated using a standard T-test for average depth of extent and Pearson x2 analysis for the remainder of the variables.
others, the nests were confined to the superficial dermis. Average depth of extent for DPN versus NFAEC was 2.38 mm versus 0.74 mm, respectively (P \ .001). Figure 1 demonstrates the overall architecture a typical DPN and NFAEC, respectively. Junctional activity. A junctional nevomelanocytic component was significantly more common in NFAEC (71/81; 88%) than in DPN (58/146; 58%; P \ .001). When a junctional component was present within a DPN, it was minimal in extent (Table II). Adnexal spread. One hundred twenty-nine DPN (88%) demonstrated the spread of epithelioid nevus cells along adnexal structures, while such a finding was present in only 32 NFAEC (39%; P \ .001). Neurovascular spread was identified within 110 DPN (75%) and 53 NFAEC (65%), a difference which was not statistically significant. Adnexal spread within DPN and NFAEC is illustrated in Figure 2. Associated melanocytic nevi with ‘‘congenital features.’’ One hundred twenty-three DPN (84%) and 76 NFAEC (94%) had a co-existent common nevus that exhibited ‘‘congenital features’’ consisting of deep extension and/or nests closely approximated to cutaneous adnexal structures, such as pilosebaceous units, sweat glands, neurovascular bundles, and arrector pili muscles (P = .035; Table II). Mitotic activity. One NFAEC contained a single mitotic figure. Nine DPN (6%) had at least a single mitotic figure, and two of these cases contained two mitotic figures. Atypical mitoses were not observed within any lesion. Cytologic features. Both NFAEC and DPN were characterized by loosely-arranged aggregates of
High, Alanen, and Golitz 463
J AM ACAD DERMATOL VOLUME 55, NUMBER 3
Fig 2. Adnexal spread was identified in both (A) DPN and (B) NFAEC, but was decidedly more common in the former (P \ .001). (A and B, Hematoxylin-eosin stain; original magnification: A and B, 3400.)
epithelioid and occasionally spindled melanocytes, which were often intermixed with heavily pigmented melanophages (Fig 3). The cytology of the epithelioid melanocytes from all types of nevi was virtually indistinguishable. Nuclei were slightly enlarged, with irregular contours, small or inconspicuous nucleoli, mild to moderate nuclear hyperchromasia, and an amorphous rather than coarse chromatin. Nuclear vacuoles (pseudoinclusions) were evident focally within both types of nevi. Overlap group. When reviewed retrospectively, 43 of 227 cases (19% of total) demonstrated overlapping histologic features of DPN and NFAEC. Within this intermediate group, 23 lesions were diagnosed originally as DPN, and 20 lesions were diagnosed originally as NFAEC. Interestingly, nevi originally diagnosed as DPN but now placed within the intermediate group had a more obvious junctional component compared to the more classic DPN lesions, and were, on average, not as thick as the remaining group of more classic DPN (1.60 mm vs 2.38 mm; Fig 4). Conversely, the 20 cases initially interpreted as NFAEC but now placed within the intermediate group were, on average, thicker than more classic NFAEC lesions (0.95 mm vs 0.74 mm). The NFAEC lesions within the intermediate group more often demonstrated adnexal spread of the atypical epithelioid melanocytes (90% vs 39%). In fact, there were no significant differences between the ‘‘intermediate DPN’’ or ‘‘intermediate
Fig 3. The individual cytology of (A) DPN was essentially identical to that of (B) NFAEC. Both lesions demonstrated an admixed population of epithelioid and occasional spindled melanocytes with slightly enlarged nuclei, small nucleoli, and amorphous chromatin surrounded by dusty cytoplasm. (A and B, Hematoxylin-eosin stain; original magnification: A and B, 3600.)
NFAEC’’ groups, with respect to junctional activity, adnexal spread, or the coexistence of a common nevus with congenital features. Table III summarizes the important characteristics of this intermediate group. p53 immunohistochemistry. Using the p53 immunohistochemical stain, diffuse nuclear staining of the atypical epithelioid cells was seen in 2/10 NFAEC (20%) and 2/10 DPN (20%). In these cases, fewer than 5% of the epithelioid nevus cells demonstrated staining. There was no staining of the adjacent cytologically bland nevus cells.
DISCUSSION This is the largest series to date of both DPN and NFAEC lesions. These entities are of particular importance because of clinical and histologic confusion with malignant melanoma. Indeed, melanoma was the clinical impression in 23% of the cases subsequently interpreted as NFAEC, and in 8% of the cases subsequently interpreted as DPN. The anatomical distribution of NFAEC and DPN in this study was strikingly similar. Nearly 80% of each type of nevus was located on the upper portion of
464 High, Alanen, and Golitz
J AM ACAD DERMATOL SEPTEMBER 2006
Table III. Histologic features of the intermediate group Original diagnosis* DPN (n = 23)
NFAEC (n = 20)
Statistical significancez
Number with junctional 16 (70) component (%) 1.60 Average depth of extent in mmy (SD) Adnexal spread (%) 21 (91) Number with co-existing 22 (96) common nevus with congenital features (%)
15 (75)
NS
0.95
—
18 (90) 19 (95)
NS NS
Histologic feature
Fig 4. Lesions identified as being in an ‘‘overlap group’’ possessed features of (A) DPN, such as moderately deep involvement, but also features of (B) NFAEC, such as surrounding nests of more banal-appearing melanocytes and junctional activity. (A and B, Hematoxylin-eosin stain; original magnification: A, 3100; B, 3200.)
the body, with a particular predilection for the head and neck (Table I). Such results are in keeping with those of other studies.1-4,6,7 Patients with NFAEC were only slightly younger than those with DPN (average age 25.4 years vs 28.3 years). Clearly, a limitation of our study is that we were unable to establish the temporal evolution of the lesion prior to biopsy. Mitotic activity was minimal or absent in NFAEC and DPN. Nevertheless, both types of nevi can be confused with malignant melanoma because they are composed of intradermal collections of atypical epithelioid melanocytes. In fact, it has been suggested by some that NFAEC are actually melanomas.10 We disagree with this interpretation, for if some of our cases were actually melanomas, with an average of 6.4 years of follow-up, it seems likely metastases would be detected in an appreciable fraction of cases. A minor limitation of the study is that clinical outcomes were reported in a respectable, but incomplete, 82% of cases. Nonetheless, it seems likely that tragic and litigious outcomes, such as a missed melanoma, would be preferentially recalled and reported. NFAEC and DPN are symmetrical lesions, without cytologic atypia within the junctional component (when present). The lesions are typically devoid of significant mitotic activity. Neither pagetoid spread
DPN, Deep penetrating nevus; NFAEC, nevus with focal atypical epithelial component; NS, not significant (P [ .05); SD, standard deviation. *Original diagnosis prior to reclassification into intermediate group upon retrospective blinded review. y Depth of extent refers to the deepest extent of atypical epithelioid nevus cells within the tissue. z Statistical significance was calculated using a standard T-test for average depth of extent and Pearson x2 analysis for the remainder of the variables.
within the epidermis, nor a desmoplastic response within the dermal collagen is common in either type of nevus. Moreover, the nuclear features of NFAEC and DPN differ from melanoma. While DPN and NFAEC have amorphous chromatin and small nucleoli, melanoma typically has more vesicular and hyperchromatic nuclei with enlarged nucleoli. Documentation of just two local recurrences (one occurring in a DPN and another in a NFAEC) and no metastases over a follow-up period averaging 6.4 years supports the presumed benignity of these lesions. Seab et al1 had reported that the epithelioid melanocytes of DPN often extended between collagen bundles without evoking a significant desmoplastic response. In our study, the majority of cases of DPN and NFAEC did not demonstrate desmoplastic stroma, however, focal areas of sclerosis may be observed on occasion. Importantly, we did identify a subgroup with histologic features that overlapped between classic DPN and NFAEC. Appropriately, within this overlapping group nevi were intermediate in thickness between classic DPN and classic NFAEC, and had a more obvious junctional component than did typical DPN. More than 95% of the lesions within this intermediate group were associated with a congenital-type melanocytic nevus. Such intermediate lesions may be misdiagnosed as melanoma, as they are uncommon and cannot be readily classified as DPN or NFAEC. While we do not believe it essential that these ‘‘intermediate’’ nevi be treated as a distinct
High, Alanen, and Golitz 465
J AM ACAD DERMATOL VOLUME 55, NUMBER 3
entity, we assert that it may provide evidence that NFAEC and DPN lie upon a continuum; the chief determinant being depth of extension. On occasion, distinction may prove difficult, but would appear to be without clinical or prognostic significance. Interestingly, 94% of NFAEC and 84% of DPN had a coexistent common melanocytic nevus that exhibited ‘‘congenital features’’ (deep extension and/or nests in close approximation to adnexal structures, such as pilosebaceous units, sweat glands, neurovascular bundles, and arrector pili muscles). Such ‘‘congenital features’’ were present to varying degrees, but were nearly always readily identified at scanning magnification. To determine if this observation was a chance phenomenon, we reviewed 227 consecutive ‘‘ordinary’’ compound or intradermal nevi from our files, and found that only 37% (84/227) showed such ‘‘congenital features.’’ In this respect, most NFAEC and DPN may represent biphasic nevi. Previously, Cooper2 described the existence of a preexisting common nevus in two-thirds of the DPN in his study, although he did not specify how many of these nevi exhibited ‘‘congenital features.’’ Use of the histologic term ‘‘congenital features,’’ except when corroborated by the clinical history, has been objected to by some. Admittedly, it is not dispositive of a truly congenital lesion.11 Nevertheless, it is a utilitarian term in that it describes a detectable pattern of nevus cells within the lower two-thirds of the dermis and subcutis, often with splaying of collagen bundles, adnexal extension, and infiltration of the arrector pilori. It is likely that the ‘‘congenital features’’ documented both for DPN and NFAEC represents a growth or organizational tendency inherent to the nevus cells, and one that is present to a lesser degree in common nevi. The etiology of minor cytologic atypia present in some DPN and NFAEC is unclear, although alterations in the p53 pathway may be involved.3 In the present investigation, p53 staining atypical epithelioid cells were demonstrated in 20% of NFAEC and in 20% of DPN. In both types of nevi, the associated common nevus cells were uniformly negative for p53 staining. It had been reported previously, by our own institution, that 50% (9/18) of NFAEC demonstrated immunohistochemical expression of p53.6 It is presumed that simple sampling error may account for the discrepancy in staining patterns within these sample populations. We are unaware of any other studies investigating p53 expression in DPN. Alternatively, it is possible that any minor cytological atypia is reactive in nature, as a mild to moderate lymphocytic infiltrate was present in most cases of DPN and NFAEC in our study. Lymphocytic
infiltrates have been reported in all series of DPN, ranging from 34% to 75% of cases.1,2,12 It is also possible that the cytologic atypia observed is degenerative in nature. This may be supported by the amorphous quality of the chromatin and the paucity of mitotic activity. Mehregan et al13 found fewer than 5% of the cells in DPN marked with antibodies to proliferating cell nuclear antigen. Similarly, it has been reported that proliferating cell nuclear antigen labeling in NFAEC does not differ from that of ordinary acquired melanocytic nevi.6 Persuasive evidence to support our proposed association between NFAEC (‘‘clonal nevi’’) and DPN has already been noted, albeit not for the purpose of supporting an etiologic relationship. In 2004, Kazakov et al14 examined anti-MAGE expression among 106 benign melanocytic lesions using the monoclonal antibody B57.14 The B57 antibody detects expression of the MAGE family of genes. Prior to the aforementioned study, MAGE had been demonstrated only in the testes, placenta, and in 25% to 42% of melanomas.15-17 Interestingly, while antiMAGE staining was noted in only 49% of all the lesions, and sample sizes within diagnostic categories were small, the authors specifically noted that the most consistent staining occurred in NFAEC (28/30) and DPN (6/6).14 Though not specifically addressed by the authors, and in light of our own evidence, it is possible this consistent and shared anti-MAGE expression among NFAEC and DPN serves as unrecognized testimony to a relationship between the entities.
CONCLUSION In summary, we report the largest series to date of both DPN and NFAEC, and confirm the benign behavior of these lesions. Despite minor histologic differences, we propose that DPN and NFAEC may be related entities upon a continuum, differing mainly in the proportion of and deepest extension of atypical epithelioid cells. This is supported by: (1) the identical cytologic appearance in both types of nevi, (2) the high frequency of coexisting common nevi with congenital features, (3) a similar age and anatomical distribution, and (4) the identification of a group of nevi with overlapping histologic features. Although difficult to prove, it is quite possible that NFAEC may represent an abortive DPN; the behavior of which is somehow influenced by the genetic expression of the cells, or the cytokine milieu of the surrounding tissue. REFERENCES 1. Seab JA Jr, Graham JR, Helwig EB. Deep penetrating nevus. Am J Surg Pathol 1989;13:39-44.
466 High, Alanen, and Golitz
J AM ACAD DERMATOL SEPTEMBER 2006
2. Cooper PH. Deep penetrating (plexiform spindle cell) nevus. A frequent participant in combined nevus. J Cutan Pathol 1992;19:172-80. 3. Barnhill RL. Pathology of melanocytic nevi and malignant melanoma. Newton, MA: Butterworth-Heinemann Publishers; 1995. 4. Barnhill RL, Mihm MC Jr, Magro CM. Plexiform spindled cell nevus: a distinctive variant of plexiform melanocytic naevus. Histopathology 1991;18:243-7. 5. Groben PA, Harvell JD, White WL. Epithelioid blue nevus: neoplasm sui generis or variation on a theme? Am J Dermatopathol 2000;22:473-88. 6. Ball NJ, Golitz LE. Melanocytic nevi with focal atypical epithelioid cell components: a review of seventy-three cases. J Am Acad Dermatol 1994;30:724-9. 7. Huynh PM, Glusac EJ, Bolognia JL. The clinical appearance of clonal nevis (inverted type A nevi). Int J Dermatol 2004;43: 882-5. 8. Hui P, Perkins AS, Glusac EJ. Assessment of clonality in nevi. J Cutan Pathol 2001;28:140-4. 9. Robinson WA, Lemon M, Elefanty A, Harrison-Smith M, Markham N, Norris D. Human acquired naevi are clonal. Melanoma Res 1998;8:499-503. 10. Okun MR. Melanocytic nevi with focal atypical epithelioid components. J Am Acad Dermatol 1996;34:861-2.
11. Tannous ZS, Mihm MC Jr, Sober AJ, Duncan LM. Congenital melanocytic nevi: clinical and histopathologic features, risk of melanoma, and clinical management. J Am Acad Dermatol 2005;52:197-203. 12. Mehregan DA, Mehregan AH. Deep penetrating nevus. Arch Dermatol 1993;129:328-31. 13. Mehregan DR, Mehregan DA, Mehregan AH. Proliferating cell nuclear antigen staining in deep-penetrating nevi. J Am Acad Dermatol 1995;33:685-7. 14. Kazakov DV, Kutzner H, Rutten A, Michal M, Requena L, Burg G, et al. The anti-MAGE antibody B57 as a diagnostic marker in melanocytic lesions. Am J Dermatopathol 2004;26: 102-7. 15. Jungbluth AA, Busam KJ, Kolb D, Iversen K, Coplan K, Chen YT, et al. Expression of MAGE-antigens in normal tissues and cancer. Int J Cancer 2000;85:460-5. 16. Hofbauer GFL, Schaefer C, Noppen C, Boni R, Kamarashev J, Nestle FO, et al. MAGE-3 immunoreactivity in formalin-fixed, paraffin embedded primary and metastatic melanoma. Am J Pathol 1997;151:1549-53. 17. Busam KJ, Iversen K, Berwick M, Spagnoli GC, Old LJ, Jungbluth AA. Immunoreactivity with the anti-MAGE antibody 57B in malignant melanoma: frequency of expression and correlation with prognostic parameters. Mod Pathol 2000;13: 459-65.
Is melanocytic nevus with focal atypical epithelioid components (clonal nevus) a superficial variant of deep penetrating nevus? Whitney A. High, MD,a Kenneth W. Alanen, MD,b and Loren E. Golitz, MDa Denver, Colorado, and Edmonton, Alberta, Canada Background: We compare features of two similar nevi which may be confused with melanoma: deep penetrating nevus (DPN), and nevus with focal atypical epithelioid component (NFAEC). Methods: Clinical/demographic and histologic information regarding 146 DPN and 81 NFAEC were compared. Patient outcomes were ascertained via a questionnaire solicited to the referring physicians. Results: Clinical features were similar for each type of nevus. Histologically, all lesions demonstrated identical aggregates of epithelioid melanocytes. DPN more often extended into the deep dermis or subcutaneous tissue, while NFAEC was typically confined to the superficial dermis. NFAEC more often demonstrated a junctional component (P \ .001) and coexistent common nevus with congenital features (P = .035). DPN more often demonstrated adnexal spread (P \.001). A subgroup with overlapping features was identified. With an average of 6.4 years follow-up, there were just two recurrences, and no metastases. Limitations: This retrospective study utilized tissue specimens from a single reference laboratory; ergo, some inherent selection bias exists. Specimens were randomly selected for P53 immunostaining, leading also to potential sampling error. Conclusions: DPN and NFAEC show considerable similarity, differing mainly in the depth of extent for the lesion. It is possible that NFAEC represents a more superficial variant of DPN. A subgroup with overlapping features, but intermediate depth, supports such a relationship. ( J Am Acad Dermatol 2006;55:460-6.)
n 1989, Seab et al1 described 70 cases of a distinctive melanocytic lesion, for which they coined the term deep penetrating nevus (DPN). DPN demonstrates a predilection for the face, proximal extremities, and trunk, and typically occurs in the first four decades of life.1,2 In most respects, DPN is similar if not identical to a lesion Barnhill et al3 termed ‘‘plexiform spindle-cell nevus.’’ Histologically, DPN is symmetrical and demonstrates an inverted, wedge-shaped architecture with
I
From the Departments of Dermatology & Pathology,a University of Colorado Health Sciences Center, and the Division of Dermatology,b Department of Medicine, University of Alberta. Drs High and Alanen contributed equally to this work. Funding sources: None. Conflicts of interest: None identified. Reprints not available from the authors. Correspondence to: Whitney A. High, MD, Assistant Professor, Dermatology & Dermatopathology, University of Colorado Health Sciences Center, PO Box 6510, Mail Stop F703, Aurora, CO 80045-0510. E-mail: [email protected]. 0190-9622/$32.00 ª 2006 by the American Academy of Dermatology, Inc. doi:10.1016/j.jaad.2006.04.054
460
deep extension into the dermis or subcutis. Tracking of melanocytes along adnexal and neurovascular structures is characteristic. Spindled cells and epithelioid components may be present.4 While the nuclei of epithelioid cells in DPN are typically smaller and more hyperchromatic than true epithelial cells, they maintain a polygonal shape that has led to use of terminology similar to other melanocytic lesions, such as epithelioid blue nevus. In epithelioid blue nevus, melanocytes are often dispersed as single cells between collagen bundles, providing partial distinction from classic DPN.5 Significant overlap exists, and many have hypothesized that that before the description by Seab et al, many DPN were diagnosed as cellular blue nevi.1,2 Because of the deep extension, and the presence of epithelioid melanocytes with associated melanophages, DPN may also be confused with melanoma. Nevus with focal atypical epithelioid components (NFAEC) is likely synonymous with inverted type A nevus.3,6,7 This entity has also been referred to as a ‘‘clonal nevus,’’ akin to ‘‘clonal seborrheic keratoses.’’ In fact, the original series of NFAEC was submitted as
High, Alanen, and Golitz 461
J AM ACAD DERMATOL VOLUME 55, NUMBER 3
‘‘clonal nevi,’’ but the term NFAEC was suggested during peer review.6 While clonality is expected in nearly all common nevi,8,9 it has not been rigorously established within NFAEC. Currently, our research group is performing such an analysis to confirm or refute the appropriateness of this terminology, and to compare the apparent biphasic populations of melanocytes. Clinically, NFAEC are characterized by a focal dark area in an otherwise homogenous nevus.6,7 Such pigmentation, especially when noted abruptly, may raise clinical concern for melanoma. The hyperpigmented area corresponds histologically to a relatively well-circumscribed collection of heavily pigmented melanophages, in association with slightly atypical epithelioid melanocytes. These melanocytes have dusty brown cytoplasm and irregular nuclei, and are usually located within the superficial dermis. Despite some minor cytologic atypia, or more aptly a departure in morphology from surrounding banal nevus cells, metastases with NFAEC have not been reported.6 The events involved in development of NFAEC are largely unknown. Herein, we reviewed all cases diagnosed as DPN or NFAEC over a 13-year period at our referral laboratory. Demographic, clinical, and histologic features were identified, compiled, and analyzed. Retrospectively, we identified a subgroup of nevi that exhibited overlapping histologic features of DPN and NFAEC. We propose that NFAEC and DPN may be closely related entities. Our report comprises the largest study to date of both DPN and NFAEC, and confirms important findings from previous series, further documenting the apparently benign nature of these lesions.
Archived hematoxylin-eosin (H&E) stained slides were reviewed, without knowledge of the original diagnosis or clinical data. Immunohistochemical staining for p53 (Calbiochem, San Diego, Calif) was performed on 20 randomly selected nevi (10 NFAEC, 10 DPN) using the avidin-biotin complex method. Appropriate positive and negative controls were included. To provide an additional point of comparison, 227 consecutive nevi diagnosed at our laboratory were assessed for a congenital pattern, specifically that of deep extension and/or nests approximated to cutaneous adnexal structures, such as pilosebaceous units, sweat glands, neurovascular bundles, and arrector pili muscles. Data analysis was performed using SAS (Statistical Analysis System, SAS Institute Inc, Cary, NC). Pearson x2 tests were used to test for differences between demographic and histologic characteristics. Categories that contained only a few observations (\5) but with essentially identical percentages were not analyzed. For continuous variables, such as age at diagnosis or depth of extent, a standard T-test analysis was utilized. A significance level of .05 was used in all analyses.
RESULTS
All cases diagnosed as DPN* or NFAEC at our academic dermatopathology referral practice between 1986 and 1999 were reviewed and included. Cases demonstrating features of a dysplastic (Clark’s) nevus, blue nevus, or cellular blue nevus were excluded. Clinical data was obtained from referral letters and/or standardized forms that accompanied the consultation material. Questionnaires seeking follow-up information were mailed to the referring physicians. Specific inquiries regarding recurrence of the lesion and/or development of melanoma were included.
Clinical data The average age of patients with NFAEC was 25.4 years (range: 4-77 years). There were 48 women and 33 men (81 lesions in 81 persons). The most common clinical impressions for cases diagnosed as NFAEC were: atypical nevus (28%), melanoma (23%), benign nevus (23%), and ‘‘clonal nevus’’ (12%). The average age of patients with DPN was 28.3 years (range: 4-64 years). There were 85 women and 61 men (146 lesions in 146 persons). The most common clinical diagnoses for cases diagnosed as DPN were: blue nevus (28%), melanocytic nevus (27%), atypical nevus (23%), and melanoma (7.5%). The head, neck, and back were the most common locations for both types of nevi (Table I). Follow-up data was obtained via questionnaire for 82% of patients. The average follow-up was 6.4 years (range: 6 months-13 years). Two patients had developed a recurrence (one DPN, one NFAEC). Histologic review revealed that both lesions had involved surgical margins upon biopsy. No patient developed metastases and/or malignant melanoma.
*DPN was formally described in 1989 by Seab et al1 at the Armed Forces Institute of Pathology (AFIP) in Washington, DC. One of us (L. E. G.) had been introduced to the concept of DPN during fellowship training at the AFIP and therefore cases with this pattern had been coded for retrieval from the database prior to the published description.
Histologic data Deepest extent of lesion. One hundred fortysix lesions were initially diagnosed as DPN. Ninetyone DPN (62%) extended only into the middle to deep reticular dermis. Forty-one DPN (28%)
METHODS
462 High, Alanen, and Golitz
J AM ACAD DERMATOL SEPTEMBER 2006
Table I. Anatomical distributions of DPN and NFAEC* DPN Location
Head and neck Back Arm and shoulder Abdomen Leg and thigh Chest Buttock Foot Totals
Number
Table II. Histologic features of DPN and NFAEC Histologic feature
NFAEC %
Number
%
52 34 28 12 11 5 2 2
35 23 19 8 8 3 1 1
29 20 17 7 5 1 1 1
35 24 20 8 5 1 1 1
146
100
81
100
DPN, Deep penetrating nevus; NFAEC, nevus with focal atypical epithelial component. *Statistical analysis using a Pearson x 2 failed to demonstrate significant differences in the anatomical distribution of lesions, with P values ranging from 0.74 to 0.98.
Fig 1. Low-power photomicrographs demonstrate the overall architecture of a DPN (A) with symmetrical, roughly-wedge shaped architecture, and central bluntended penetrating lobules containing dermal melanophages, and a NFAEC (‘‘clonal’’ nevus) (B), manifesting as a symmetric and superficial silhouette with focal aggregates of larger, epitheloid melanocytes with dusty cytoplasm and associated dermal melanophages. (A and B, Hematoxylin-eosin stain; original magnification: A, 340; B, 3100).
extended into the subcutaneous fat. In 14 cases (10%), the deepest extent of the DPN could not be determined. Eighty-one lesions were diagnosed initially as NFAEC. In six NFAEC (7.4%), the deepest extent of the lesion could not be evaluated, but in all
DPN (N = 146)
NFAEC (N = 81)
Number with junctional 85 (58) 71 (88) component (%) Average depth of extent 2.38 0.74 in mm* (SD) Adnexal spread (%) 129 (88) 32 (39) Neurovascular spread 110 (75) 53 (65) Number with co-existing 123 (84) 76 (94) common nevus with congenital features (%)
Statistical significancey
P \ .001 — P \ .001 NS P = .035
DPN, Deep penetrating nevus; NFAEC, nevus with focal atypical epithelial component; NS, not significant (P [ .05); SD, standard deviation. *Depth of extent refers to the deepest extent of atypical epithelioid nevus cells within the tissue. y Statistical significance was calculated using a standard T-test for average depth of extent and Pearson x2 analysis for the remainder of the variables.
others, the nests were confined to the superficial dermis. Average depth of extent for DPN versus NFAEC was 2.38 mm versus 0.74 mm, respectively (P \ .001). Figure 1 demonstrates the overall architecture a typical DPN and NFAEC, respectively. Junctional activity. A junctional nevomelanocytic component was significantly more common in NFAEC (71/81; 88%) than in DPN (58/146; 58%; P \ .001). When a junctional component was present within a DPN, it was minimal in extent (Table II). Adnexal spread. One hundred twenty-nine DPN (88%) demonstrated the spread of epithelioid nevus cells along adnexal structures, while such a finding was present in only 32 NFAEC (39%; P \ .001). Neurovascular spread was identified within 110 DPN (75%) and 53 NFAEC (65%), a difference which was not statistically significant. Adnexal spread within DPN and NFAEC is illustrated in Figure 2. Associated melanocytic nevi with ‘‘congenital features.’’ One hundred twenty-three DPN (84%) and 76 NFAEC (94%) had a co-existent common nevus that exhibited ‘‘congenital features’’ consisting of deep extension and/or nests closely approximated to cutaneous adnexal structures, such as pilosebaceous units, sweat glands, neurovascular bundles, and arrector pili muscles (P = .035; Table II). Mitotic activity. One NFAEC contained a single mitotic figure. Nine DPN (6%) had at least a single mitotic figure, and two of these cases contained two mitotic figures. Atypical mitoses were not observed within any lesion. Cytologic features. Both NFAEC and DPN were characterized by loosely-arranged aggregates of
High, Alanen, and Golitz 463
J AM ACAD DERMATOL VOLUME 55, NUMBER 3
Fig 2. Adnexal spread was identified in both (A) DPN and (B) NFAEC, but was decidedly more common in the former (P \ .001). (A and B, Hematoxylin-eosin stain; original magnification: A and B, 3400.)
epithelioid and occasionally spindled melanocytes, which were often intermixed with heavily pigmented melanophages (Fig 3). The cytology of the epithelioid melanocytes from all types of nevi was virtually indistinguishable. Nuclei were slightly enlarged, with irregular contours, small or inconspicuous nucleoli, mild to moderate nuclear hyperchromasia, and an amorphous rather than coarse chromatin. Nuclear vacuoles (pseudoinclusions) were evident focally within both types of nevi. Overlap group. When reviewed retrospectively, 43 of 227 cases (19% of total) demonstrated overlapping histologic features of DPN and NFAEC. Within this intermediate group, 23 lesions were diagnosed originally as DPN, and 20 lesions were diagnosed originally as NFAEC. Interestingly, nevi originally diagnosed as DPN but now placed within the intermediate group had a more obvious junctional component compared to the more classic DPN lesions, and were, on average, not as thick as the remaining group of more classic DPN (1.60 mm vs 2.38 mm; Fig 4). Conversely, the 20 cases initially interpreted as NFAEC but now placed within the intermediate group were, on average, thicker than more classic NFAEC lesions (0.95 mm vs 0.74 mm). The NFAEC lesions within the intermediate group more often demonstrated adnexal spread of the atypical epithelioid melanocytes (90% vs 39%). In fact, there were no significant differences between the ‘‘intermediate DPN’’ or ‘‘intermediate
Fig 3. The individual cytology of (A) DPN was essentially identical to that of (B) NFAEC. Both lesions demonstrated an admixed population of epithelioid and occasional spindled melanocytes with slightly enlarged nuclei, small nucleoli, and amorphous chromatin surrounded by dusty cytoplasm. (A and B, Hematoxylin-eosin stain; original magnification: A and B, 3600.)
NFAEC’’ groups, with respect to junctional activity, adnexal spread, or the coexistence of a common nevus with congenital features. Table III summarizes the important characteristics of this intermediate group. p53 immunohistochemistry. Using the p53 immunohistochemical stain, diffuse nuclear staining of the atypical epithelioid cells was seen in 2/10 NFAEC (20%) and 2/10 DPN (20%). In these cases, fewer than 5% of the epithelioid nevus cells demonstrated staining. There was no staining of the adjacent cytologically bland nevus cells.
DISCUSSION This is the largest series to date of both DPN and NFAEC lesions. These entities are of particular importance because of clinical and histologic confusion with malignant melanoma. Indeed, melanoma was the clinical impression in 23% of the cases subsequently interpreted as NFAEC, and in 8% of the cases subsequently interpreted as DPN. The anatomical distribution of NFAEC and DPN in this study was strikingly similar. Nearly 80% of each type of nevus was located on the upper portion of
464 High, Alanen, and Golitz
J AM ACAD DERMATOL SEPTEMBER 2006
Table III. Histologic features of the intermediate group Original diagnosis* DPN (n = 23)
NFAEC (n = 20)
Statistical significancez
Number with junctional 16 (70) component (%) 1.60 Average depth of extent in mmy (SD) Adnexal spread (%) 21 (91) Number with co-existing 22 (96) common nevus with congenital features (%)
15 (75)
NS
0.95
—
18 (90) 19 (95)
NS NS
Histologic feature
Fig 4. Lesions identified as being in an ‘‘overlap group’’ possessed features of (A) DPN, such as moderately deep involvement, but also features of (B) NFAEC, such as surrounding nests of more banal-appearing melanocytes and junctional activity. (A and B, Hematoxylin-eosin stain; original magnification: A, 3100; B, 3200.)
the body, with a particular predilection for the head and neck (Table I). Such results are in keeping with those of other studies.1-4,6,7 Patients with NFAEC were only slightly younger than those with DPN (average age 25.4 years vs 28.3 years). Clearly, a limitation of our study is that we were unable to establish the temporal evolution of the lesion prior to biopsy. Mitotic activity was minimal or absent in NFAEC and DPN. Nevertheless, both types of nevi can be confused with malignant melanoma because they are composed of intradermal collections of atypical epithelioid melanocytes. In fact, it has been suggested by some that NFAEC are actually melanomas.10 We disagree with this interpretation, for if some of our cases were actually melanomas, with an average of 6.4 years of follow-up, it seems likely metastases would be detected in an appreciable fraction of cases. A minor limitation of the study is that clinical outcomes were reported in a respectable, but incomplete, 82% of cases. Nonetheless, it seems likely that tragic and litigious outcomes, such as a missed melanoma, would be preferentially recalled and reported. NFAEC and DPN are symmetrical lesions, without cytologic atypia within the junctional component (when present). The lesions are typically devoid of significant mitotic activity. Neither pagetoid spread
DPN, Deep penetrating nevus; NFAEC, nevus with focal atypical epithelial component; NS, not significant (P [ .05); SD, standard deviation. *Original diagnosis prior to reclassification into intermediate group upon retrospective blinded review. y Depth of extent refers to the deepest extent of atypical epithelioid nevus cells within the tissue. z Statistical significance was calculated using a standard T-test for average depth of extent and Pearson x2 analysis for the remainder of the variables.
within the epidermis, nor a desmoplastic response within the dermal collagen is common in either type of nevus. Moreover, the nuclear features of NFAEC and DPN differ from melanoma. While DPN and NFAEC have amorphous chromatin and small nucleoli, melanoma typically has more vesicular and hyperchromatic nuclei with enlarged nucleoli. Documentation of just two local recurrences (one occurring in a DPN and another in a NFAEC) and no metastases over a follow-up period averaging 6.4 years supports the presumed benignity of these lesions. Seab et al1 had reported that the epithelioid melanocytes of DPN often extended between collagen bundles without evoking a significant desmoplastic response. In our study, the majority of cases of DPN and NFAEC did not demonstrate desmoplastic stroma, however, focal areas of sclerosis may be observed on occasion. Importantly, we did identify a subgroup with histologic features that overlapped between classic DPN and NFAEC. Appropriately, within this overlapping group nevi were intermediate in thickness between classic DPN and classic NFAEC, and had a more obvious junctional component than did typical DPN. More than 95% of the lesions within this intermediate group were associated with a congenital-type melanocytic nevus. Such intermediate lesions may be misdiagnosed as melanoma, as they are uncommon and cannot be readily classified as DPN or NFAEC. While we do not believe it essential that these ‘‘intermediate’’ nevi be treated as a distinct
High, Alanen, and Golitz 465
J AM ACAD DERMATOL VOLUME 55, NUMBER 3
entity, we assert that it may provide evidence that NFAEC and DPN lie upon a continuum; the chief determinant being depth of extension. On occasion, distinction may prove difficult, but would appear to be without clinical or prognostic significance. Interestingly, 94% of NFAEC and 84% of DPN had a coexistent common melanocytic nevus that exhibited ‘‘congenital features’’ (deep extension and/or nests in close approximation to adnexal structures, such as pilosebaceous units, sweat glands, neurovascular bundles, and arrector pili muscles). Such ‘‘congenital features’’ were present to varying degrees, but were nearly always readily identified at scanning magnification. To determine if this observation was a chance phenomenon, we reviewed 227 consecutive ‘‘ordinary’’ compound or intradermal nevi from our files, and found that only 37% (84/227) showed such ‘‘congenital features.’’ In this respect, most NFAEC and DPN may represent biphasic nevi. Previously, Cooper2 described the existence of a preexisting common nevus in two-thirds of the DPN in his study, although he did not specify how many of these nevi exhibited ‘‘congenital features.’’ Use of the histologic term ‘‘congenital features,’’ except when corroborated by the clinical history, has been objected to by some. Admittedly, it is not dispositive of a truly congenital lesion.11 Nevertheless, it is a utilitarian term in that it describes a detectable pattern of nevus cells within the lower two-thirds of the dermis and subcutis, often with splaying of collagen bundles, adnexal extension, and infiltration of the arrector pilori. It is likely that the ‘‘congenital features’’ documented both for DPN and NFAEC represents a growth or organizational tendency inherent to the nevus cells, and one that is present to a lesser degree in common nevi. The etiology of minor cytologic atypia present in some DPN and NFAEC is unclear, although alterations in the p53 pathway may be involved.3 In the present investigation, p53 staining atypical epithelioid cells were demonstrated in 20% of NFAEC and in 20% of DPN. In both types of nevi, the associated common nevus cells were uniformly negative for p53 staining. It had been reported previously, by our own institution, that 50% (9/18) of NFAEC demonstrated immunohistochemical expression of p53.6 It is presumed that simple sampling error may account for the discrepancy in staining patterns within these sample populations. We are unaware of any other studies investigating p53 expression in DPN. Alternatively, it is possible that any minor cytological atypia is reactive in nature, as a mild to moderate lymphocytic infiltrate was present in most cases of DPN and NFAEC in our study. Lymphocytic
infiltrates have been reported in all series of DPN, ranging from 34% to 75% of cases.1,2,12 It is also possible that the cytologic atypia observed is degenerative in nature. This may be supported by the amorphous quality of the chromatin and the paucity of mitotic activity. Mehregan et al13 found fewer than 5% of the cells in DPN marked with antibodies to proliferating cell nuclear antigen. Similarly, it has been reported that proliferating cell nuclear antigen labeling in NFAEC does not differ from that of ordinary acquired melanocytic nevi.6 Persuasive evidence to support our proposed association between NFAEC (‘‘clonal nevi’’) and DPN has already been noted, albeit not for the purpose of supporting an etiologic relationship. In 2004, Kazakov et al14 examined anti-MAGE expression among 106 benign melanocytic lesions using the monoclonal antibody B57.14 The B57 antibody detects expression of the MAGE family of genes. Prior to the aforementioned study, MAGE had been demonstrated only in the testes, placenta, and in 25% to 42% of melanomas.15-17 Interestingly, while antiMAGE staining was noted in only 49% of all the lesions, and sample sizes within diagnostic categories were small, the authors specifically noted that the most consistent staining occurred in NFAEC (28/30) and DPN (6/6).14 Though not specifically addressed by the authors, and in light of our own evidence, it is possible this consistent and shared anti-MAGE expression among NFAEC and DPN serves as unrecognized testimony to a relationship between the entities.
CONCLUSION In summary, we report the largest series to date of both DPN and NFAEC, and confirm the benign behavior of these lesions. Despite minor histologic differences, we propose that DPN and NFAEC may be related entities upon a continuum, differing mainly in the proportion of and deepest extension of atypical epithelioid cells. This is supported by: (1) the identical cytologic appearance in both types of nevi, (2) the high frequency of coexisting common nevi with congenital features, (3) a similar age and anatomical distribution, and (4) the identification of a group of nevi with overlapping histologic features. Although difficult to prove, it is quite possible that NFAEC may represent an abortive DPN; the behavior of which is somehow influenced by the genetic expression of the cells, or the cytokine milieu of the surrounding tissue. REFERENCES 1. Seab JA Jr, Graham JR, Helwig EB. Deep penetrating nevus. Am J Surg Pathol 1989;13:39-44.
466 High, Alanen, and Golitz
J AM ACAD DERMATOL SEPTEMBER 2006
2. Cooper PH. Deep penetrating (plexiform spindle cell) nevus. A frequent participant in combined nevus. J Cutan Pathol 1992;19:172-80. 3. Barnhill RL. Pathology of melanocytic nevi and malignant melanoma. Newton, MA: Butterworth-Heinemann Publishers; 1995. 4. Barnhill RL, Mihm MC Jr, Magro CM. Plexiform spindled cell nevus: a distinctive variant of plexiform melanocytic naevus. Histopathology 1991;18:243-7. 5. Groben PA, Harvell JD, White WL. Epithelioid blue nevus: neoplasm sui generis or variation on a theme? Am J Dermatopathol 2000;22:473-88. 6. Ball NJ, Golitz LE. Melanocytic nevi with focal atypical epithelioid cell components: a review of seventy-three cases. J Am Acad Dermatol 1994;30:724-9. 7. Huynh PM, Glusac EJ, Bolognia JL. The clinical appearance of clonal nevis (inverted type A nevi). Int J Dermatol 2004;43: 882-5. 8. Hui P, Perkins AS, Glusac EJ. Assessment of clonality in nevi. J Cutan Pathol 2001;28:140-4. 9. Robinson WA, Lemon M, Elefanty A, Harrison-Smith M, Markham N, Norris D. Human acquired naevi are clonal. Melanoma Res 1998;8:499-503. 10. Okun MR. Melanocytic nevi with focal atypical epithelioid components. J Am Acad Dermatol 1996;34:861-2.
11. Tannous ZS, Mihm MC Jr, Sober AJ, Duncan LM. Congenital melanocytic nevi: clinical and histopathologic features, risk of melanoma, and clinical management. J Am Acad Dermatol 2005;52:197-203. 12. Mehregan DA, Mehregan AH. Deep penetrating nevus. Arch Dermatol 1993;129:328-31. 13. Mehregan DR, Mehregan DA, Mehregan AH. Proliferating cell nuclear antigen staining in deep-penetrating nevi. J Am Acad Dermatol 1995;33:685-7. 14. Kazakov DV, Kutzner H, Rutten A, Michal M, Requena L, Burg G, et al. The anti-MAGE antibody B57 as a diagnostic marker in melanocytic lesions. Am J Dermatopathol 2004;26: 102-7. 15. Jungbluth AA, Busam KJ, Kolb D, Iversen K, Coplan K, Chen YT, et al. Expression of MAGE-antigens in normal tissues and cancer. Int J Cancer 2000;85:460-5. 16. Hofbauer GFL, Schaefer C, Noppen C, Boni R, Kamarashev J, Nestle FO, et al. MAGE-3 immunoreactivity in formalin-fixed, paraffin embedded primary and metastatic melanoma. Am J Pathol 1997;151:1549-53. 17. Busam KJ, Iversen K, Berwick M, Spagnoli GC, Old LJ, Jungbluth AA. Immunoreactivity with the anti-MAGE antibody 57B in malignant melanoma: frequency of expression and correlation with prognostic parameters. Mod Pathol 2000;13: 459-65.