Dermatoscopy of “dysplastic nevi”: A beacon in diagnostic darkness

ARTICLE

Auteur(s) : Alexander Roesch¹, Walter Burgdorf², Wilhelm Stolz³, Michael Landthaler¹, Thomas Vogt^1,*

¹Department of Dermatology, University of Regensburg, Franz-Josef-Strauss-Allee 11, D-93053 Regensburg, Germany
²Department of Dermatology, Ludwig Maximilian University, Munich, Germany
³Clinic for Dermatology and Allergology, Hospital Munich-Schwabing, Munich, Germany

accepté le 6 Juin 2006

Clinical morphology: The daily diagnostic dilemma

After recognition of a suspicious lesion, the next question is what does the clinical morphology imply and what consequences have to be drawn. Such clinical borderline lesions may represent (1) true MM precursors with a “dysplastic” abnormal histologic appearance (2) early MM with clear histologic indicators of MM (3) clinically suspicious benign lesions with completely innocuous histology, and (4) clinically suspicious lesions fulfilling Clark and Elder’s histological features of dysplasia. In our opinion, the latter mostly represent an endpoint of nevocytic development and never progress into a malignant tumour. But, which one will? A history of rapid growth can be helpful to recognize early MM or precursors, since dysplastic benign nevi are characteristically stable in size [3, 9, 10]. However, patients’ histories are often vague. ( Figure 1 ) displays a number of examples. Five (histologic) MM are hidden among the benign nevi. They are hard to find with the naked eye.

Dermatopathology: Not really a gold-standard

When evaluating epithelial neoplasms, dysplasia is defined as the architectural and cytological deviation from the normal configuration of the epithelium. In epithelial carcinogenesis, a continuous progression from minor dysplasia to severe dysplasia and finally to epithelial carcinoma can often be observed in one single specimen. This applies to the epidermis and other epithelia as well, e.g. the best-known example being the adenoma-carcinoma sequence in colon cancer [12]. For melanocytic proliferations, the situation is not as straightforward as it is in epithelial skin cancer [13]. The main problem with the histological diagnosis of melanocytic “dysplasia” is – very similar to the clinical situation – its lack of commonly accepted and reproducibly measurable criteria.

According to Clark and Elder, histologic melanocytic dysplasia is characterized by the following features [1, 2, 4]:

• (1) lentiginous melanocytic hyperplasia (confluence of melanocytic cells in the junctional zone; melanocytic cells bridging across papillary tips ( (figure 2G) ); shoulder phenomenon, i.e., peripheral extension of the junctional component beyond the dermal component (( figure 2 ) H);
• (2) epithelioid melanocytic atypia (large melanocytic cells with raised quantities of cytoplasm and finely distributed pigmentation resulting in “dusty” or “milky” appearance; ( (figure 2 ) G, H);
• (3) lamellar fibrosis, i.e., elongated fibroblasts positioned next to each other, separated by layers of condensed extracellular matrix, and/or concentric eosinophilic fibrosis around the rete ridges (( figure 2 G ));
• (4) perivascular lymphocytic infiltrate in the papillary dermis ( (figure 2H) ).

However, Ackermann pointed out that the nevus subtype defined by these criteria represents the most common subtype at all, because some of Clark and Elder’s criteria are found in almost any excised melanocytic nevus [14]. Most nevi which are excised have been found clinically suspicious. In this dilemma, some dermatopathologists generously use the diagnosis “dysplastic nevus”, often just to indicate “I have seen the problem, don’t blame me if it is already an early MM”. Sometimes even re-excision of “dysplastic” nevi with larger safety margins is recommended by pathologists, which is inconsequent and means “I don’t know”. Others, including our group, raise the threshold of melanocytic “dysplasia”, restricting the diagnosis to lesions that already partially fulfil MM criteria, in particular those showing a certain asymmetry. This latter approach is more likely to identify MM precursors, but it will also miss a few early MM. ( Figure 2 ) displays examples of clear MM (I-M), common nevi (A-D) and questionable “DN” (E-H).

The degeneration of melanocytic nevi is one point that may have not received enough attention as a factor contributing to “dysplastic” cytological appearance. In other neuroectodermal tumours such as schwannomas, it is well accepted that degenerative changes may lead to impressive cytological atypia with the occasional formation of bizarre, hyperchromatic giant nuclei, sometimes denoted as ancient schwannoma [13]. Some of the cellular features interpreted as melanocytic “dysplasia” may also be the result of homologous degenerative processes. Accordingly, an “ancient nevus” subtype has been described by Kerl et al. as one possible end point of nevocytic development [15]. The impressive cellular atypia of these cases was regarded as a degenerative change; the histological findings suggesting malignancy were contrary to the benign clinical behaviour confirmed by long-time follow-up.

In contrast to the definition of DN, for MM there are more commonly accepted histological criteria [16, 17]:

• (1) asymmetry of the tumour architecture, pagetoid scattering (suprabasal infiltration of the epidermis by melanocytic cells; ( figure 2L ));
• (2) nuclear pleomorphism, heterogeneous chromatin distribution with relative hyperchromasia, prominent nucleoli, mitoses, lack of maturation at the base (( figure 2 ) M);
• (3) fibrosis and regression;
• (4) asymmetric lymphocytic infiltration.

Some authors emphasize cyto-morphologic deviations in the diagnosis of DN. For instance Barnhill suggested that discontinuous nuclear atypia and abnormal, intraepidermal proliferation of melanocytic cells (lentiginous or junctional) should be regarded as a prerequisite for the diagnosis of intermediary melanocytic lesions, i.e., true MM precursors [3, 18, 19]. Previously, Mihm and Barnhill tried to distinguish six grades of melanocytic dysplasia: (i) common melanocytic nevi, (ii) melanocytic nevi with features of DN, (iii) DN with slight cytologic atypia, (iv) DN with moderate cytologic atypia, (v) DN with severe cytologic atypia, and (vi) primary MM. However, as with all schemes for melanocytic dysplasia, this concept also lacked interobserver reproducibility. In this study, experienced dermatopathologists had a concordance in grading DN ranging from 35% to 58% (kappa value 0.38-0.47), while that of less experienced dermatopathologists ranged from 16% to 65% (kappa value 0.05-0.24) [20]. Regardless of how refined the histological criteria become, the distinction between DN and true initial MM remains blurred and a gold standard remains elusive. The bad thing about it is not only the imponderability from the perspective of the patient, it also challenges any research effort on DN, also the molecular analyses. The question is always what did researchers investigate, if they say DN were included [21]?

In summary, similarly to the clinical dilemma, histopathology has problems in discriminating the dysplastic portion of tumors from the true malignant ones and the true benign ones. And, histopathology also does not solve the question whether the DN is an end point or a precursor lesion of MM. The histopathological detection of DN (according to Clark and Elder’s criteria) within MM biopsies was taken as a proof for Clark’s model of melanoma progression. However, current data suggest that a considerable proportion – around 60–75% – of MM develop de novo without any precursor lesion [14, 22]. It became clear that only a relatively small proportion of melanomas as a whole are associated with preexisting nevi. The size of this subset has been studied extensively [23]. In one larger study Harley and Walsh found, for instance, that only 23% of the melanomas arose in association with preexisting nevi of which 55% were acquired, 28% were congenital (small) and in 17% a distinction could not confidently be made. Of the acquired nevi, the majority were of the dysplastic type according to Clark’s definition along with common nevi. They concluded, similarly to other authors, that only a small subset of melanomas suggests a DN to MM sequence, and a similar role of small congenital nevi and common nevi should not be underrecognized [24]. The reported proportions again challenge the true role of DN as a precursor. For the clinician this means that the high incidence of both small congenital and common acquired nevi, and DN versus the incidence of melanomas should be kept in perspective in devising strategies for early detection and prevention.

Epidemiology: Limited predictive value of dysplastic melanocytic nevi

For instance, Tucker et al. tried to add more details to the concept described by Clark and Elder. They showed that the presence of only one clinically suspicious nevus, according to the clinical ABCD rule, doubles the risk for the development of MM. Individuals with ten or more nevi filtered by the ABCD rule had a twelve-fold increased risk of MM in their study [35]. From this point of view, the postulate that MM develops from pre-existing DN and the conclusion that there is a nevus-melanoma sequence with DN as intermediate lesions seems logical. However, contradictory data were published by Grob et al. [29], who found a significant increase in MM risk in patients with an increased total number of melanocytic skin lesions, including clinically innocuous nevi. In their study, the presence of DN was not linked to an increased MM risk. Instead, the presence of more than 120 pigmented nevi with a diameter of less than 5 mm indicates a 19.6- fold increase in risk, as compared to subjects with less than ten nevi. They assigned particular importance to the frequency of lesions with diameters > 5 mm, independently from other morphological criteria. Five nevi with a diameter > 5 mm increased the risk of MM by a factor of 10. A further current large evaluation of more than 20,000 patients with MM in the German Central Malignant Melanoma Registry points into the same direction, showing that the total number of clinically inconspicuous nevi on the lower extremities is a good predictor for the development of MM in both males and females [37]. Weiss, Bertz and Jing also found a correlation between the total number of pigmented skin tumors including small common nevi and the development of MM [38]. In their study, the number of benign nevi turned out to be the best predictive parameter, with a relative MM risk of 14.9 for patients with fifty nevi or more. Bataille et al. added that the presence of common nevi in atypical locations (back of the foot, buttocks, dorsal capillitium) might also indicate an increased risk of MM development [34].

Kanzler and Mraz-Gernhard have suggested three theoretical possibilities to explain these epidemiological circumstances [39]: (i) subjects with multiple pigmented nevi, whether typical or not, have more melanocytic cells than other individuals. This increases the probability of a deleterious genetic defect per melanocytic cell; (ii) the formation of multiple benign tumors may already be the result of a genetic predisposition for proliferative changes, and consequently also for the occurrence of malignant cells in the pigment cell system, and (iii) previous exposure to environmental factors, such as sunlight in childhood, may generally foster the development of melanocytic tumors, benign and malignant. This concept implies that nevus and MM development are independent without an obligate stepwise transition. Furthermore, it should be noted that there is a weak correlation between the topographic distribution of MM and DN which also contradicts the paradigm of DN as MM precursors [40].

Taken together, it becomes obvious that any melanocytic tumour has a certain propensity to progress to MM. In this regard, the DN are not different from other pigmented tumors such as common benign nevi, developing (growing) melanocytic nevi in children/adults, irritated melanocytic nevi, sun-exposed melanocytic nevi, regressive melanocytic nevi, melanocytic nevi under chemotherapy, melanocytic nevi in pregnancy, or ancient melanocytic nevi. They all contribute to the set of nevus-associated MM ( (figure 4) ).

Molecular biology and genetics: A mirror of the dilemma, but no solution

Table 1( Table 1 ) provides an orientation and overview over the varying molecular aspects that have been studied in DN. Major issues in molecular DN research have been cell cycle control (loss of suppressor gene functions), genomic instability (mismatch repair deficits), chromosomal aberrations, oncogen activating mutations and telomerase activation, and signal transduction and signalling kinase activation levels [64].

Among the complex and heterogeneous findings listed in table 1, the possible role of p16 tumour suppressor gene deserves a closer look, since it is deleted or mutated in various human tumors including MM, particularly in MM families (inherited MM). Only limited conflicting results have been reported in sporadic DN [48, 49, 54, 55]. The p16-gene, INK4a (CDKN2A), is located on chromosome 9p21 and codes for a 16 kDa protein that inhibits pro-proliferative, cyclin-dependent kinases (CDK 4/6) in the G1 phase of the cell cycle. About 50% of patients from melanoma-prone families have a germline mutation in INK4a. 25-40% of sporadic MM also show mutations leading to a failure of INK4a function [65]. Since MM and the tumour suppressor INK4a seem to be causally linked, chromosome analyses of the 9p21 locus have been carried out particularly carefully on individuals with DNS. Patients with mutations in INK4a had a significantly higher total number of nevi, but there was no significant correlation with the number of DN [66]. In summary, while the linkage of familial MM and the 9p21 locus seems to be consistent in many studies, the linkage of DN and this locus remains questionable [42].

Another potential MM susceptibility region with significant linkage scores in DN-families is 1p36 with the PITSLRE gene coding for a p58 protein kinase, which also contributes to the control of the cell cycle. Based on genetic linkage analyses on FAMMM syndrome and familial MM, Goldstein et al. suggested that 1p36 is a tumour-susceptibility locus that is not only relevant for MM but also for the formation of DN [42]. But, further studies substantiating a possible role of p58 in the development of DN are not available and later studies did not unequivocally confirm this [64].

According to Knudson’s two-hit hypothesis, both alleles of a tumour suppressor gene have to be deactivated for a loss of activity. Loss of the healthy allele (loss of heterozygosity, LOH) is one of the most frequently observed mechanisms for a second hit, with the first hit being a germline mutation in a tumour suppressor gene. Using LOH analyses of tumour tissue, these allelic losses (for example, as a result of chromosomal deletion or unbalanced translocation) can be detected [64]. Allelic losses of 1p, 9p and/or 17p (the coding region for the p53 suppressor gene) could not be found in tissue samples from common nevi, but LOH was present in DN [45-47]. Similar patterns of losses seem to exist in MM and DN, but in DN allelic losses are much less frequent [46].

A further hallmark of cancer genomics is the frequent occurrence of microsatellite instability (MSI) – length variations in non-coding, short, repetitive DNA sections, which flank the coding regions of genes. These regions may act as regulatory DNA sections (for example, as topoisomerase binding sites, sites of recombination). Their sequential uniformity can result in mismatch pairing during chromosome replication and this may also lead to subsequent microsatellite length mutations. Under normal conditions, mismatch repair (MMR) proteins prevent the mismatching of bases during DNA replication. The main human MMRs are hMSH2, hMLH1, hPMS1 and 2 as well as GTBP. MSI has been documented in DN and its incidence seems to increase with the degree of histological atypia in DN [55]. Since no MSI has been reported in normal nevi, DN seem to be intermediate between common nevi and MM, concerning mismatch repair competence. However, MSI is uncommon in both DN and MM. Over 80% of sporadic MM are MSI negative [56, 64, 67-70]. Therefore, both the diagnostic value and the biological significance of MSI and mismatch repair in DN are still doubtful.

Most recently, another new candidate gene has emerged that is linked to MM development, BRAF. As a serine/threonine kinase within the MAPK-modulated signal transduction cascade, BRAF transmits proliferative signals to the nucleus. Since a high mutation rate of BRAF can be found in common nevi, DN and MM [60], BRAF-analysis is also unsuitable to achieve the desired discriminative power or tell us anything about the MM precursor potential or end point differentiation of a DN.

Finally, telomeres – TTAGG repeats at the end of the chromosomes – represent a promising candidate to elucidate the nature of DN. Telomeres are lost during cell cycling and represent a limiting factor concerning the number of possible cycles. Telomerase is an enzyme that adds TTAGG repeats to the ends of chromosomes. It is suppressed in somatic cells which limits their growth. In contrast, telomerase is active in many cancers including MM. Also during the postulated sequence from normal nevi to MM, a progressive increase of telomerase activity has been postulated by several studies. But strikingly, telomerase activity was only increased in DN when tissue specimens were histologically selected according to cytological and nuclear atypia [61, 62]. If Clark and Elder’s criteria were used, no differences in telomerase activity were found between common nevi and DN [63].

Taken together, the molecular studies demonstrate that a few lesions among the DN studied always seem to be transitional between common nevi and MM, but the majority are not. New and yet not fully exploited is the potential of applying “omics”-tools, i.e. genome wide analysis of the genome itself, the transcriptome and the proteome by chip-CGH, cDNA array analysis and mass spectrometry of the proteome (e.g. by MALDI-TOF MS) [71-73]. There is hope that the growing possibilities of laser catapulted microdissection of DN and chip based “omics”-tools together with the growing possibilities of data mining, cluster and pathway analysis will finally resolve the enigmas of melanocytic dysplasia, too [74-77]. Doubtless, much more work is needed to expand the molecular knowledge of DN. For that purpose however, reproducible definitions of DN would be highly desirable.
Table 1 Molecular biological features of dysplastic nevi

Dermatoscopy and computer-assisted image analysis: Making melanocytic tumors with uncertain malignant potential a managable disease

Although other systems such as the pattern analysis [80] and the 7-point checklist [81] are available, the ABDC rule of dermatoscopy [82, 83] is most often applied. Stolz and co-workers developed the ABCD rule of dermatoscopy, an algorithm that also forms the basis for some commercially available computerized systems. Four dermatoscopic criteria are analyzed for evaluation of a pigmented lesion: Asymmetry (in no, one or two axes), Border (sharp versus blurred demarcation in eight segments), Colour (number of colours: dark brown, light brown, black, red, grey, white), and Differential structures (number of micorarchitectural features of melanocytic lesions: network, branching streaks, structure-free zones, globules, and dots). ( Figure 6 ) displays the basic principles of the ABCD rule of dermatoscopy [22, 84].

For example, when analysing ( figure 5 ), the point value would be calculated as follows: A: 2 points, since one can not find an axis to mirror one half to the other when considering form and structural features. B: 0 points, since the border is blurred all around the eight segments. C: 3 points for colour, since one can discriminate light brown, dark brown and black. D: 4 points for structure-free zones, branching streaks, dots, globules. Based on empirical data the ABCD point values need to be weighted to express the relative power of each of the criteria [22, 84]. Since asymmetry turned out to be a much stronger discriminator between MM and nevus than colours or structural features and the latter are stronger discriminators than the border, asymmetry goes into the total dermatoscopic score (TDS) multiplied with 1.3, colour and structural feature multiplied with 0.5 and border with 0.1. Hence, the TDS for ( figure 5 ) would be:differential structures, divide the result by two and add 1.3 per axis of asymmetry and then score gradually up 0.1 increments for any of the eight segments with sharp demarcation of the structures. Such TDS scoring certainly needs practice. However, it turned out to be particular useful for guidance of less-experienced investigators, and increased the accuracy of MM diagnosis to over 90% in experts [85]. ( Figure 7 ) gives an impression of the leap in resolution by dermatoscopy. The hidden MMs are now more easily to detect if one guides oneself through the images according to the ABCD rule of dermatoscopy.

Another great advantage is that with the advent of digital computer-assisted dermatoscopy, the TDS can be established by computed image analysis. Therefore, objectivity and reproducibility is guaranteed. Next to its reproducibility, the main advantage of computer-aided dermatoscopy is the rapid and exact digitized recording of skin lesions. Even minor structural changes can be followed over time, e.g. the development of small peripheral pseudopods or asymmetrical growth, two criteria that are highly specific for malignant melanocytic progression [86]. Morphological changes such as increase of size, changes in shape and colour, signs of regression, and the appearance of other differential structures can also be sensitively monitored. Binder and colleagues demonstrated by computerized dermatoscopy that typical and dysplastic nevi may undergo a subtle increase in size, but they keep their symmetrical form, and rarely develop new dermatoscopic structures [87]. Figures 8 and 9 give examples of suspicious melanocytic lesions. One was stable over time, probably representing an end point of nevocytic evolution ( (figure 8) ), whereas the other one showed asymmetrical growth over time, a sign of malignancy ( (figure 9) ). Both were histologically “proven”.

In summary, video documentation systems with an integrated, computerized colour image analysis can guide clinicians in the assessment of pigmented skin lesions but do not solve the problem of “dysplastic” lesions [22, 88]. Surprisingly, the practical net effect of such objective scoring systems is a dramatic reduction of doubtful, suspicious cases [22]. ( Figure 10 ) shows a Gaussian distribution of the scores of common nevi and MM with no overlap. The problematical lesions, the DN, also distribute in a Gaussian manner in the range between TDS 4.75 to 5.45, with little overlap. To us this proves the existence of indeterminate melanocytic tumors which cannot be further resolved. Therefore, regardless of the nature of DN and irrespective of the current semantic confusion, a practical approach could be to remove lesions with a TDS of over 4.75 or, alternatively, to closely follow-up such lesions by computerized dermatoscopy. Tumors with a TDS of higher than 5.45 have already a 90% probability of being a MM and should in any case be removed with an appropriate safety margin [82].

A more widespread use of this practical approach for research purposes would also contribute to a better comparability of studies dealing with melanocytic tumors. Pigmented tumors would then not fall into variable numbers of ill-defined categories, but would be classified by scores, e.g., a study concept could be the analysis of lesions with TDS 4.75 to 5.45. Concurrent studies would then profit in terms of comparability.

Conclusion and practical approach

Acknowledgements

References

2 Elder DE. The dysplastic nevus. Pathology 1985; 17: 291-7.

3 Barnhill RL. Malignant melanoma, dysplastic melanocytic nevi, and Spitz tumors. Histologic classification and characteristics. Clin Plast Surg 2000; 27(3): 331-60; (viii).

4 Clark Jr. WH, Reimer RR, Greene M, Ainsworth AM, Mastrangelo MJ. Origin of familial malignant melanomas from heritable melanocytic lesions. ’The B-K mole syndrome’. Arch Dermatol 1978; 114(5): 732-8.

5 Friedman RJ, Rigel DS, Kopf AW. Early detection of malignant melanoma: the role of physician examination and self-examination of the skin. CA Cancer J Clin 1985; 35(3): 130-51.

6 Hofmann-Wellenhof R, Blum A, Wolf IH, et al. Dermoscopic classification of Clark’s nevi (atypical melanocytic nevi). Clin Dermatol 2002; 20(3): 255-8.

7 Ackerman AB. Dysplastic nevus" syndrome: does a survey make it real? J Am Acad Dermatol 2003; 48(3): 461-3.

8 McBride A, Rivers JK, Kopf AW, et al. Clinical features of dysplastic nevi. Dermatol Clin 1991; 9(4): 717-22.

9 Kim JC, Murphy GF. Dysplastic melanocytic nevi and prognostically indeterminate nevomelanomatoid proliferations. Clin Lab Med 2000; 20(4): 691-712.

10 Tucker MA, Fraser MC, Goldstein AM, et al. A natural history of melanomas and dysplastic nevi: an atlas of lesions in melanoma-prone families. Cancer 2002; 94(12): 3192-209.

11 Soyer HP, Massone C, Ferrara G, Argenziano G. Limitations of histopathologic analysis in the recognition of melanoma: a plea for a combined diagnostic approach of histopathologic and dermoscopic evaluation. Arch Dermatol 2005; 141(2): 209-11.

12 Cho KR, Vogelstein B. Genetic alterations in the adenoma--carcinoma sequence. Cancer 1992; 70(6 Suppl): 1727-31.

13 Rudolph P. The dysplastic nevus. An Ariadne’s thread through a conceptual labyrinth. Pathologe 2002; 23(2): 128-34.

14 Ackerman AB. What naevus is dysplastic, a syndrome and the commonest precursor of malignant melanoma? A riddle and an answer. Histopathology 1988; 13(3): 241-56.

15 Kerl H, Soyer HP, Cerroni L, Wolf IH, Ackerman AB. Ancient melanocytic nevus. Semin Diagn Pathol 1998; 15(3): 210-5.

16 Tronnier M, Garbe C, Brocker EB, et al. Standards in histopathologic diagnosis of malignant melanomas. Recommendations of the Working Group of the Central Malignant Melanoma Registry of the German Dermatologic Society. Hautarzt 1997; 48(10): 720-9.

17 Ackerman AB, Cerroni L, Kerl H. Pitfalls in histopathologic diagnosis of malignant melanoma. Philadelphia, Baltimore, Hong Kong, Munich, Sydney, Tokyo: Lea and Febiger, 1994.

18 Barnhill RL. Tumors of melanocytes. In: Barnhill RL, ed. Textbook of Dermatopathology. 1998: 537-91; (New York).

19 Barnhill RL, Roush GC, Duray PH. Correlation of histologic architectural and cytoplasmic features with nuclear atypia in atypical (dysplastic) nevomelanocytic nevi. Hum Pathol 1990; 21(1): 51-8.

20 Duncan LM, Berwick M, Bruijn JA, Byers HR, Mihm MC, Barnhill RL. Histopathologic recognition and grading of dysplastic melanocytic nevi: an interobserver agreement study. J Invest Dermatol 1993; 100(3): 318S-321S.

21 Cramer S. Interobserver variability in surgical pathology. Adv Pathol 1996; 9: 3.

22 In: Stolz W, Braun-Falco O, Bilek P, Burgdorf WHC, Landthaler M, eds. Color Atlas of Dermatoscopy. 2nd ed. Berlin, Germany: Blackwell, 2002.

23 Stolz W, Schmoeckel C, Landthaler M, Braun-Falco O. Association of early malignant melanoma with nevocytic nevi. Cancer 1989; 63(3): 550-5.

24 Harley S, Walsh N. A new look at nevus-associated melanomas. Am J Dermatopathol 1996; 18(2): 137-41.

25 Nordlund JJ, Kirkwood J, Forget BM, et al. Demographic study of clinically atypical (dysplastic) nevi in patients with melanoma and comparison subjects. Cancer Res 1985; 45(4): 1855-61.

26 Holly EA, Kelly JW, Shpall SN, Chiu SH. Number of melanocytic nevi as a major risk factor for malignant melanoma. J Am Acad Dermatol 1987; 17(3): 459-68.

27 MacKie RM, Freudenberger T, Aitchison TC. Personal risk-factor chart for cutaneous melanoma. Lancet 1989; 2(8661): 487-90.

28 Augustsson A, Stierner U, Rosdahl I, Suurkula M. Common and dysplastic naevi as risk factors for cutaneous malignant melanoma in a Swedish population. Acta Derm Venereol 1991; 71(6): 518-24.

29 Grob JJ, Gouvernet J, Aymar D, et al. Count of benign melanocytic nevi as a major indicator of risk for nonfamilial nodular and superficial spreading melanoma. Cancer 1990; 66(2): 387-95.

30 Halpern AC, Guerry D, Elder DE, et al. Dysplastic nevi as risk markers of sporadic (nonfamilial) melanoma. A case-control study. Arch Dermatol 1991; 127(7): 995-9.

31 Garbe C, Buttner P, Weiss J, et al. Risk factors for developing cutaneous melanoma and criteria for identifying persons at risk: multicenter case-control study of the Central Malignant Melanoma Registry of the German Dermatological Society. J Invest Dermatol 1994; 102(5): 695-9.

32 Carli P, Biggeri A, Giannotti B. Malignant melanoma in Italy: risks associated with common and clinically atypical melanocytic nevi. J Am Acad Dermatol 1995; 32(5 Pt 1): 734-9.

33 Grulich AE, Bataille V, Swerdlow AJ, et al. Naevi and pigmentary characteristics as risk factors for melanoma in a high-risk population: a case-control study in New South Wales, Australia. Int J Cancer 1996; 67(4): 485-91.

34 Bataille V, Bishop JA, Sasieni P, et al. Risk of cutaneous melanoma in relation to the numbers, types and sites of naevi: a case-control study. Br J Cancer 1996; 73(12): 1605-11.

35 Tucker MA, Halpern A, Holly EA, et al. Clinically recognized dysplastic nevi. A central risk factor for cutaneous melanoma. JAMA 1997; 277(18): 1439-44.

36 Bataille V, Grulich A, Sasieni P, et al. The association between naevi and melanoma in populations with different levels of sun exposure: a joint case-control study of melanoma in the UK and Australia. Br J Cancer 1998; 77(3): 505-10.

37 Rieger E, Soyer HP, Garbe C, et al. Overall and site-specific risk of malignant melanoma associated with nevus counts at different body sites: a multicenter case-control study of the German Central Malignant-Melanoma Registry. Int J Cancer 1995; 62(4): 393-7.

38 Weiss J, Bertz J, Jung EG. Malignant melanoma in southern Germany: different predictive value of risk factors for melanoma subtypes. Dermatologica 1991; 183(2): 109-13.

39 Kanzler MH, Mraz-Gernhard S. Primary cutaneous malignant melanoma and its precursor lesions: diagnostic and therapeutic overview. J Am Acad Dermatol 2001; 45(2): 260-76.

40 Tronnier M, Muller C. Relationship between season and diagnoses of melanocytic tumours. Acta Derm Venereol 2001; 81(2): 112-5.

41 Bale SJ, Dracopoli NC, Tucker MA, et al. Mapping the gene for hereditary cutaneous malignant melanoma-dysplastic nevus to chromosome 1p. N Engl J Med 1989; 320(21): 1367-72.

42 Goldstein AM, Goldin LR, Dracopoli NC, Clark Jr. WH, Tucker MA. Two-locus linkage analysis of cutaneous malignant melanoma/dysplastic nevi. Am J Hum Genet 1996; 58(5): 1050-6.

43 Fountain JW, Karayiorgou M, Ernstoff MS, et al. Homozygous deletions within human chromosome band 9p21 in melanoma. Proc Natl Acad Sci USA 1992; 89(21): 10557-61.

44 Cannon-Albright LA, Goldgar DE, Meyer LJ, et al. Assignment of a locus for familial melanoma, MLM, to chromosome 9p13-p22. Science 1992; 258(5085): 1148-52.

45 Lee JY, Dong SM, Shin MS, et al. Genetic alterations of p16INK4a and p53 genes in sporadic dysplastic nevus. Biochem Biophys Res Commun 1997; 237(3): 667-72.

46 Park WS, Vortmeyer AO, Pack S, et al. Allelic deletion at chromosome 9p21(p16) and 17p13(p53) in microdissected sporadic dysplastic nevus. Hum Pathol 1998; 29(2): 127-30.

47 Boni R, Zhuang Z, Albuquerque A, Vortmeyer A, Duray P. Loss of heterozygosity detected on 1p and 9q in microdissected atypical nevi. Arch Dermatol 1998; 134(7): 882-3.

48 Piepkorn M. Melanoma genetics: an update with focus on the CDKN2A(p16)/ARF tumor suppressors. J Am Acad Dermatol 2000; 42(5 Pt 1): 705-22; (quiz 723-6).

49 Healy E, Sikkink S, Rees JL. Infrequent mutation of p16INK4 in sporadic melanoma. J Invest Dermatol 1996; 107(3): 318-21.

50 Papp T, Pemsel H, Zimmermann R, Bastrop R, Weiss DG, Schiffmann D. Mutational analysis of the N-ras, p53, p16INK4a, CDK4, and MC1R genes in human congenital melanocytic naevi. J Med Genet 1999; 36(8): 610-4.

51 Levin DB, Wilson K, Valadares de Amorim G, Webber J, Kenny P, Kusser W. Detection of p53 mutations in benign and dysplastic nevi. Cancer Res 1995; 55(19): 4278-82.

52 Vogt T, Zipperer KH, Vogt A, Holzel D, Landthaler M, Stolz W. p53-protein and Ki-67-antigen expression are both reliable biomarkers of prognosis in thick stage I nodular melanomas of the skin. Histopathology 1997; 30(1): 57-63.

53 Cristofolini M, Boi S, Girlando S, et al. p53 Protein expression in nevi and melanomas. Arch Dermatol 1993; 129(6): 739-43.

54 Hussein MR, Roggero E, Sudilovsky EC, Tuthill RJ, Wood GS, Sudilovsky O. Alterations of mismatch repair protein expression in benign melanocytic nevi, melanocytic dysplastic nevi, and cutaneous malignant melanomas. Am J Dermatopathol 2001; 23(4): 308-14.

55 Hussein MR, Sun M, Tuthill RJ, et al. Comprehensive analysis of 112 melanocytic skin lesions demonstrates microsatellite instability in melanomas and dysplastic nevi, but not in benign nevi. J Cutan Pathol 2001; 28(7): 343-50.

56 Birindelli S, Tragni G, Bartoli C, et al. Detection of microsatellite alterations in the spectrum of melanocytic nevi in patients with or without individual or family history of melanoma. Int J Cancer 2000; 86(2): 255-61.

57 Shukla VK, Hughes DC, Hughes LE, McCormick F, Padua RA. ras mutations in human melanotic lesions: K-ras activation is a frequent and early event in melanoma development. Oncogene Res 1989; 5(2): 121-7.

58 Van Duinen CM, Fleuren GJ, Bruijn JA. The extracellular matrix in pigmented skin lesions: an immunohistochemical study. Histopathology 1994; 24(1): 33-40.

59 Reed JA, McNutt NS, Albino AP. Differential expression of basic fibroblast growth factor (bFGF) in melanocytic lesions demonstrated by in situ hybridization. Implications for tumor progression. Am J Pathol 1994; 144(2): 329-36.

60 Pollock PM, Harper UL, Hansen KS, et al. High frequency of BRAF mutations in nevi. Nat Genet 2003; 33(1): 19-20.

61 Glaessl A, Bosserhoff AK, Buettner R, Hohenleutner U, Landthaler M, Stolz W. Increase in telomerase activity during progression of melanocytic cells from melanocytic naevi to malignant melanomas. Arch Dermatol Res 1999; 291(2-3): 81-7.

62 Miracco C, Pacenti L, Santopietro R, Laurini L, Biagioli M, Luzi P. Evaluation of telomerase activity in cutaneous melanocytic proliferations. Hum Pathol 2000; 31(9): 1018-21.

63 Rudolph P, Schubert C, Tamm S, et al. Telomerase activity in melanocytic lesions: A potential marker of tumor biology. Am J Pathol 2000; 156(4): 1425-32.

64 Hussein MR, Wood GS. Molecular aspects of melanocytic dysplastic nevi. J Mol Diagn 2002; 4(2): 71-80.

65 Herlyn M, Satyamoorthy K. Melanoma. In: DeVita V, Hellman S, Rosenberg SA, eds. Cancer: Principles and practice of oncology. Philadelphia, USA: Lippincott-Raven Publishers, 2000: 2003-12.

66 Wachsmuth RC, Harland M, Bishop JA. The atypical-mole syndrome and predisposition to melanoma. N Engl J Med 1998; 339(5): 348-9.

67 Rhodes AR, Harrist TJ, Day CL, Mihm Jr. MC, Fitzpatrick TB, Sober AJ. Dysplastic melanocytic nevi in histologic association with 234 primary cutaneous melanomas. J Am Acad Dermatol 1983; 9(4): 563-74.

68 Tsao H, Zhang X, Majewski P, Haluska FG. Mutational and expression analysis of the p73 gene in melanoma cell lines. Cancer Res 1999; 59(1): 172-4.

69 Richetta A, Silipo V, Calvieri S, et al. Microsatellite instability in primary and metastatic melanoma. J Invest Dermatol 1997; 109(1): 119-20.

70 Kroiss MM, Vogt TM, Schlegel J, Landthaler M, Stolz W. Microsatellite instability in malignant melanomas. Acta Derm Venereol 2001; 81(4): 242-5.

71 Becker B, Roesch A, Hafner C, et al. Discrimination of melanocytic tumors by cDNA array hybridization of tissues prepared by laser pressure catapulting. J Invest Dermatol 2004; 122(2): 361-8.

72 Roesch A, Vogt T, Stolz W, Dugas M, Landthaler M, Becker B. Discrimination between gene expression patterns in the invasive margin and the tumour core of malignant melanomas. Melanoma Res 2003; 13(5): 503-9.

73 Vogl A, Sartorius U, Vogt T, et al. Gene Expression Profile Changes between Melanoma Metastases and their Daughter Cell Lines: Implication for Vaccination Protocols. J Invest Dermatol 2005; 124(2): 401-4.

74 Bittner M, Meltzer P, Chen Y, et al. Molecular classification of cutaneous malignant melanoma by gene expression profiling. Nature 2000; 406(6795): 536-40.

75 Davis Jr. RL, DuBreuil RM, Reddy SP, Dooley TP. Methods for gene expression profiling in dermatology research using DermArray nylon filter DNA microarrays. Methods Mol Biol 2005; 289: 399-412.

76 Baldi A, Battista T, De Luca A, et al. Identification of genes down-regulated during melanoma progression: a cDNA array study. Exp Dermatol 2003; 12(2): 213-8.

77 Mirmohammadsadegh A, Baer A, Nambiar S, Bardenheuer W, Hengge UR. Rapid identification of dysregulated genes in cutaneous malignant melanoma metastases using cDNA technology. Cells Tissues Organs 2004; 177(3): 119-23.

78 Skvara H, Teban L, Fiebiger M, Binder M, Kittler H. Limitations of dermoscopy in the recognition of melanoma. Arch Dermatol 2005; 141(2): 155-60.

79 Naeyaert JM, Brochez L. Clinical practice. Dysplastic nevi. N Engl J Med 2003; 349(23): 2233-40.

80 Argenziano G, Soyer HP, Chimenti S, et al. Dermoscopy of pigmented skin lesions: results of a consensus meeting via the Internet. J Am Acad Dermatol 2003; 48(5): 679-93.

81 Argenziano G, Fabbrocini G, Carli P, De Giorgi V, Sammarco E, Delfino M. Epiluminescence microscopy for the diagnosis of doubtful melanocytic skin lesions. Comparison of the ABCD rule of dermatoscopy and a new 7-point checklist based on pattern analysis. Arch Dermatol 1998; 134(12): 1563-70.

82 Nachbar F, Stolz W, Merkle T, et al. The ABCD rule of dermatoscopy. High prospective value in the diagnosis of doubtful melanocytic skin lesions. J Am Acad Dermatol 1994; 30(4): 551-9.

83 Stolz W. ABCD rule of dermatoscopy: a new practical method for early recognition of malignant melanoma. Eur J Dermatol 1994; 4: 521-7.

84 Stolz W, Riemann A, Cognetta AB, Pillet L. ABCD rule of dermatoscopy: a new practical method for early recognition of malignant melanoma. Eur J Dermatol 1994; 4: 521-7.

85 Binder M, Schwarz M, Winkler A, et al. Epiluminescence microscopy. A useful tool for the diagnosis of pigmented skin lesions for formally trained dermatologists. Arch Dermatol 1995; 131(3): 286-91.

86 Menzies SW, Crotty KA, McCarthy WH. The morphologic criteria of the pseudopod in surface microscopy. Arch Dermatol 1995; 131(4): 436-40.

87 Kittler H, Pehamberger H, Wolff K, Binder M. Follow-up of melanocytic skin lesions with digital epiluminescence microscopy: patterns of modifications observed in early melanoma, atypical nevi, and common nevi. J Am Acad Dermatol 2000; 43(3): 467-76.

88 Binder M, Kittler H, Dreiseitl S, Ganster H, Wolff K, Pehamberger H. Computer-aided epiluminescence microscopy of pigmented skin lesions: the value of clinical data for the classification process. Melanoma Res 2000; 10(6): 556-61.