Dysplastic melanocytic nevi are common acquired melanocytic proliferations with distinct clinical and histologic features. This category of melanocytic lesions has been one of the most controversial terms introduced into dermatopathology and has resulted in completely divergent entrenched opinions as to its existence, diagnostic criteria, prevalence, and its significance [1–5]. Since dysplastic melanocytic nevi were first reported in 1978 by Wallace Clark and colleagues as histologically defined lesions in melanoma-prone families , there has been extensive debate about the definition, classification, and biologic importance of these lesions [6–8]. And although a number of authors object to the use of the term “dysplastic,” we think that it has been so much popularized that it is probably best to continue using it. The most important point that has been debated is whether dysplastic nevi represent premalignant lesions that will progress to melanoma or not. There are authors who view dysplastic nevi as a discreet entity of clinical significance and others who dismiss the concept entirely [4, 9–11]. It is unknown the probability of a particular dysplastic nevus to become melanoma; however, patients with a diagnosis of dysplastic nevus are more likely to have had or to have in the future a diagnosis of melanoma in some other area of their bodies (see also below).

The precise definition of these melanocytic nevi has been a source of great controversy, as many synonyms have been proposed such as atypical nevus, Clark nevus, melanocytic nevus with architectural disorder, cytologic atypia, etc. First, Wallace H. Clark and colleagues described distinctive nevi in 37 patients from six melanoma families with “B-K mole syndrome” (B and K were the first letters of the names of two patients) [6]. These melanocytic nevi exhibited variable clinical appearance (>5 mm in size with variable color and borders). Second, Henry T. Lynch and colleagues reported a similar nevus phenotype in a melanoma-prone family. These phenotypic “syndromes” were referred to, respectively, as B-K mole syndrome and familial atypical multiple-mole melanoma (FAMMM) syndrome [8]. Then, David Elder and colleagues coined the term “dysplastic nevus syndrome” (DNS) and postulated the existence of sporadic and familial variants [1]. The term “dysplastic nevus” was proposed as these benign melanocytic nevi are characterized histologically by architectural disorder and cytologic atypia, similar to dysplastic lesions in other organs, such as the cervix, uterus, or esophagus [12]. These investigations of melanoma kindreds suggested a close relationship with these clinically atypical and histologically dysplastic nevi. However, it was noted early on that this nevus phenotype was not required for melanoma development in these kindreds, given that Clark and colleagues described two family members who developed melanoma without having clinically dysplastic nevi. An autosomal dominant mode of inheritance of the nevus phenotype has been described in melanoma-prone families, and it has been reported that up to 40 % of these families harbor a mutation in the CDKN2A locus , which encodes the p16 and ARF tumor suppressor proteins [13, 14].

The term “dysplastic nevus” remains the most commonly used name; an alternative term was proposed by the National Institutes of Health (NIH) consensus group as “nevus with architectural disorder and cytologic atypia”; however, it has been used less often as it is not very well accepted by many authorities in the field [15]. Some dermatopathologists have promoted using the term “Clark nevus” instead of dysplastic nevus to honor Dr. Clark as being the researcher that popularized the concept. Despite the NIH consensus conference and multiple studies to assess reproducibility of dysplastic melanocytic nevi, no single definition for these melanocytic nevi has yet been accepted by all [12, 16–21]. Some authors suggest that the term “dysplasia” requires histologic evaluation , and thus the term “atypical” is considered more appropriate for any clinical classification. In our view, the problem with using “atypical” is that “atypical nevi” include a broad range of different types of nevi with unusual morphology. Thus, the lack of clarity could create not only diagnostic confusion but also unnecessary treatment.

The etiology of dysplastic nevi is not well defined. In general, the major environmental risk factor for melanoma is ultraviolet radiation but the presence of an increased number of melanocytic nevi is another major host risk factor. Other host factors include increased freckling, poor tanning ability, fair complexion, light hair and eye color, and family history of melanoma. The major genetic risk factors for melanoma include the high-risk susceptibility genes CDKN2A and CDK4 as well as numerous low-risk susceptibility loci. Similar to melanoma, dysplastic nevi seem to result from genetic, host, and environmental factors. Evidence suggests that sun exposure, in addition to genetic susceptibility, can affect the development of dysplastic nevi. The incidence of dysplastic nevi among patients with fair skin types is higher compared with those with darker skin, which might explain the increased risk of melanoma in these individuals [22]. At the molecular level, common nevi and dysplastic nevi have proven some differences. Dysplastic nevi tend to display mutation-deletion of p16 gene, altered expression of p53, and increased microsatellite instability [23]. Familial melanoma and dysplastic nevi were originally thought to be pleiotropic effects of a single gene, but subsequent studies have shown that the genetics are more complex and that the genetic causes of familial melanoma and dysplastic nevi are not the same. There is not much evidence that dysplastic nevi are caused by mutations in the major melanoma susceptibility genes CDKN2A or CDK4; however, similar to dysplastic nevi in unselected patients with melanoma, dysplastic nevi are risk factors for melanoma in melanoma-prone families independent of CDKN2A mutations [24–28].

Dysplastic nevi are acquired atypical melanocytic lesions that oftentimes share some of the same clinical features as melanomas, including one or more of the ABCDs (asymmetry, border irregularity, color variability, and a diameter which may be >6 mm). Although these criteria are useful for recognizing suspicious lesions, there is often great difficulty in differentiating dysplastic nevi from melanoma.

Clark and colleagues initially defined dysplastic nevi as a nevus >5 mm in diameter with variegation in its pigmentation and irregular borders; however, since this morphologic description, there have been proposed further criteria [6]. Tucker et al. clinically defined dysplastic nevi as being >5 mm in size and having “flatness” (being entirely flat or having a flat component). In addition, two of three characteristics were also necessary: variable pigmentation, irregular and asymmetric outline, and having indistinct borders [30]. Kelly and colleagues classified dysplastic nevi as lesions with a macular component that showed at least three of five criteria: irregularly distributed pigmentation, ill-defined or irregular border, background of erythema, and >5 mm [35]. The Dutch Working Group has used the following criteria for atypical nevi: ≥5 mm in diameter, vague border, asymmetric shape, irregular pigmentation, and red hue [36].

Dysplastic nevi differ from melanomas in situ clinically, as they tend to be symmetrical. The common appearance of a dysplastic nevus is that of a “fried egg,” where there is a central papular component (“the head”) and an adjacent macular component (“shoulders”). Some dysplastic nevi are entirely junctional, and thus they usually lack this fried egg appearance. The borders of dysplastic nevi are less irregular than those of melanomas, and perhaps paradoxically they are often ill defined compared with the sharper border of most melanomas. Lentigo maligna melanoma can have an ill-defined border, similar to that of a dysplastic nevus; however, these lesions are usually much larger and show greater pigmentary variegation and asymmetry. The color of dysplastic nevi is usually less variable (uniform color) than that of melanomas and does not usually include the red, white, and blue colors that may signify partial regression in a melanoma. Dysplastic nevi, by definition, are >5 mm; however, it has been proven that dysplastic nevi <5 mm may be also associated with melanoma risk. Dysplastic nevi are most often seen on the trunk, especially the upper back, although they can appear anywhere on the skin, including sun-protected areas such as the scalp, breasts, and buttocks.

It is also important to remember that small nevi have been reported to cover the same histopathological spectrum as large lesions and to show a high proportion of atypical features (this is not the exception with dysplastic nevi). Most of the studies about dysplastic nevi concentrate on lesions larger than 5 mm in diameter, mostly because it is relatively uncommon for clinicians to biopsy lesions of smaller size. One study focused on small lesions, measuring less than 3 mm, and concluded that small junctional melanocytic lesions may display severe architectural disorder, which might be related to active melanocytic proliferation [37]. In such lesions it is important to know how to interpret the severe architectural disorder when lesions are small in size and with features of dysplastic nevi, in order to avoid overdiagnosis of melanoma.

The current definition of dysplastic nevus syndrome has been a source of controversy due its nomenclature and definition. Dysplastic nevus syndrome has also been referred as atypical mole syndrome, FAMMM syndrome, B-K mole syndrome, etc. Although originally described as a familial condition, there are also sporadic cases. Familial cases were originally called the B-K mole syndrome and the FAMMM syndrome (see above) [6, 8, 38]. In familial cases the melanoma trait is inherited as an autosomal dominant with incomplete penetrance. As explained above, the dysplastic nevus trait has a more complicated inheritance, occurring more commonly than it would be expected for a dominant inheritance. Mutations of the CDKN2A gene on 9p21–22 have been found in patients with familial melanoma (in approximately 40 % of families), but these mutations are uncommon in dysplastic nevi themselves [39–41]. CDKN2A encodes the tumor suppressor gene products p14ARF and p16INK4a. Patients with dysplastic nevus syndrome have also been associated with partial deletion of chromosome 11 and with deletion of 17p13 (p53) [42, 43]. Patients with dysplastic nevus syndrome have increased risk of developing other neoplasms, especially pancreatic cancer [44, 45].

The importance of the dysplastic nevus syndrome is that it identifies an at-risk population group for the development of melanoma. The sporadic dysplastic nevus syndrome is characterized by the presence of 2–5 dysplastic nevi on sun-exposed areas, patients have no personal or family history of melanoma, and the nevi first appear at puberty and then develop new lesions during adult life (only 20 % of these nevi appear after 50). The familial dysplastic nevus syndrome is characterized by multiple nevi (>100 melanocytic), one or more large nevi (>6 mm), and one or more clinically atypical nevi (ABCD rule), and the nevi are distributed all over the body including sun-exposed and sun-protected areas (involvement of face and scalp). The risk of developing melanoma in melanoma-prone families with the dysplastic nevus syndrome varies in the literature, and some studies have suggested that it exceeds 50 %, being highly positively correlated with higher number of nevi [46].

Dysplastic nevi, whether familial or sporadic, show identical histologic features. The changes seen in dysplastic nevi can be divided into architectural, cytological, and host responses. Dysplastic nevi by definition should display melanocytic proliferative changes (intraepidermal, at least focally lentiginous, hyperplasia of melanocytes), cytologic and architectural atypia, and a stromal response [9, 47, 48].

Architecture : Dysplastic nevi are either junctional or compound. The surface of these nevi is flat and mildly papillated. As mentioned above, compound lesions usually tend to show an extension of the epidermal component beyond the dermal component (“shoulder”). The main architectural feature of dysplastic nevi is the irregular intraepidermal melanocytic proliferation. The junctional component generally displays nests as well solitary melanocytes arranged in a lentiginous pattern. In cases in which a lentiginous pattern predominates, the melanocytes are arranged as variable nests and single cells along the basilar aspect of elongated, club-shaped rete. The rete ridges tend to be irregularly elongated. The junctional melanocytic nests are irregularly sized and shaped and often have a horizontal orientation. These junctional melanocytic nests are found near the tips and laterally to the elongated rete ridges and above the dermal papilla. Melanocytic nests connect the bases of the rete and fuse to the adjacent rete (“bridging”). This bridging is more prominent when the nests are not uniform in size and shape. The junctional component is characteristically hypercellular in areas replacing the vast majority of basal keratinocytes, and in some cases, there is marked lentiginous melanocytic hyperplasia in which single melanocytes predominate at the side of the rete ridges. Both nests and lentiginous proliferations can extend somewhat along the epithelium of cutaneous adnexa. Melanocytes within the junctional nests tend to show a dyshesive pattern with shrinkage artifact with scant cytoplasm and a spindle-shaped pattern, but in some lesions, there are epithelioid melanocytes with dusty pigment. In advanced lesions, there can be confluence of nests which may involve three or four rete ridges. In some cases, melanocytes can display retraction spaces, which give a vacuolated appearance to the basal layer. While pagetoid upward migration is not common in dysplastic nevi, there may be focal or limited suprabasal extension of melanocytes (see below).

The intradermal component, if present, often is focal and located in the central part of the lesion with the junctional component extending beyond the dermal component (“shoulders”). In most cases, there are well-defined nests and strands of melanocytes underneath the epidermis, along with sclerotic changes (fibroplasia, see “host response). Deeper areas of the nevus consist of ill-defined aggregates or compact small groups of melanocytes. The dermal component appears cytologically banal with small, round melanocytes. However, some nevi show superficial cytologic atypia comparable to the junctional component. The dermal component routinely lacks mitotic figures; therefore, the presence of a single mitotic figure in the dermal component should be a red flag to further evaluate the lesion to rule out melanoma.

Cytology : Dysplastic nevi show a variation in morphology of melanocytes including epithelioid, spindle cell, vacuolated, small melanocytes, etc. In cases in which a lentiginous pattern predominates, melanocytes tend to have a retracted cytoplasm giving a vacuolated appearance. Those cases with a predominantly nested pattern usually have epithelioid melanocytes (advanced cases may show spindle morphology). A very important criterion in the diagnosis of dysplastic nevi is the characteristic discontinuous cytologic atypia of the intraepidermal melanocytes (the degree of atypia varies from melanocyte to melanocyte).

Some melanocytes show with large, hyperchromatic, and pleomorphic nuclei; some authors considered them a requisite for the diagnosis of dysplastic nevus. In our opinion, cytologic atypia occurs along a continuum characterized by nuclear pleomorphism, chromatin variation, and presence of nucleoli, as previously described [48]. The low-end of spectrum tends to show only very minimal cytologic atypia that is almost imperceptible. Some cases may display a vacuolated cytoplasm with only mild enlargement of the nuclei (similar to the size of the adjacent keratinocyte nuclei). In this category, the chromatin is evenly dispersed with inconspicuous nucleoli. In advanced lesions, cytologic atypia is characterized by the presence of melanocytes with large oval shape with abundant pale cytoplasm with large central vesicular and hyperchromatic nuclei and prominent eosinophilic nucleoli. The nuclei are larger in size than the nucleus of the overlying keratinocytes. In some cases, the cytoplasm is ample with granular appearance or filled with dusty pigment. These atypical melanocytes may be present singly or in small clusters. These atypical melanocytes do not represent the majority, and in any one nevus, there is an admixture of normal and atypical melanocytes (random/discontinuous cytologic atypia). Mitotic figures are very rarely seen, and they should be located in the upper dermis near the epidermis. Although rare melanocytes may be seen in the suprabasal epidermis, any significant degree of pagetoid spread or its presence at the periphery of the lesion should raise the possibility of melanoma.

Host response: Host response is referred to as the presence of eosinophilic lamellar and concentric fibroplasia of the papillary dermis, mononuclear cell infiltrate, and prominent vascularity. In some cases, there is fibrosis in the upper reticular dermis, resulting in widely spaced nests.

The inflammatory lymphocytic infiltrate varies from sparse to dense. When sparse, the lymphocytes are arranged along the perivascular spaces in the superficial dermis, while in cases where there is dense inflammation, there is a band-like distribution that fills the papillary dermis. Melanophages (pigment incontinence) are usually present along with lymphocytes in the papillary dermis. Dermal fibroplasia is observed as lamellar or concentric. Lamellar fibroplasia is characterized by horizontally dispersed collagen subjacent to the epidermal rete ridges. Concentric fibroplasia is characterized by hyalinized collagen that is compactly disposed around a rete ridge. Dysplastic nevi usually show prominent vascularity in the papillary dermis, represented by the presence of ectatic vessels.

There have been many studies that have tried to consolidate histologic criteria; however, great controversy still surrounds the histopathological diagnosis of dysplastic nevi, and the possible usefulness of histological grading have raised interest in assessing the reproducibility of grading criteria.

Differences between the degree of cytologic atypia and architectural disorder have been noted in dysplastic nevi; however, in our opinion significant correlation exists between these two parameters. Several groups have proposed that dysplastic nevi be graded using architectural and cytologic features. To make dysplasia grading more objective and reproducible, a semiquantitative scoring system (the Duke criteria) was proposed and validated by Shea et al. [48]. This scoring system assessed both the architecture and the cytologic atypia of dysplastic nevi. Other authors have recommended the distinction between low-grade and high-grade dysplastic nevi using slightly different histologic criteria [49]. One study used a system for grading melanocytic dysplasia and concluded that severe dysplasia could be reliably distinguished from mild and moderate dysplasia; thus, these authors recommended a two-grade system (low and high grade) for classifying dysplastic nevi [49]. Some authors recommend not to try to grade dysplastic nevi and rather establish a diagnosis of “atypical” or “dysplastic” or “Clark” for these lesions. However, as discussed above, patients with lesions with high degree of dysplastic seem to have a higher risk of developing melanoma.

In our practice , we use the Duke system, to grade dysplastic nevi (see Table 5.2). Using this method the architectural disorder is graded as measuring the junctional component at both edges (both edges nested equal 0; single cells equal 1), the overall symmetry of the lesion (symmetrical equals 0; asymmetrical equals 1), cohesiveness of junctional nests (>50 % of nests equals 1), suprabasal spread (prominent or at edge equals 1), confluence of rete ridges (>50 % equals 1), and single-cell proliferation (>50 % equals 1). Lesions are then graded as mild (scores of 0 and 1), moderate (scores of 2 and 3), and severe (scores of 4–6) architectural disorder (see Table 5.2).

Cytologically, lesions are evaluated in the two most atypical high-powered fields. The points are given if more than 50 % of the melanocytes in those two fields show the studied criterion: 1) shape and chromasia of melanocytes (round/oval, euchromatic equals 0; else equals 1), 2) nuclear size (smaller or equal to basal cell keratinocyte nuclei equals 0; larger equals 1), cellular size (cells smaller or equal to twice the size of basal cell keratinocyte nuclei equals 0; larger equals 1), and nucleolus (small, inconspicuous equals 0; visible equals 1). Lesions are then graded as mild (score of 0), moderate (scores of 1 and 2), and severe (scores of 3 to 4) (see Table 5.2).

Although we do not grade the amount of alteration of the dermal stroma, we consider that dysplastic nevi have some degree of host response against the atypical melanocytes and thus will show different degrees of lymphocytic and macrophage infiltrate , dermal fibrosis, and vascular proliferation (see Table 5.2).