Dysplastic Nevus Syndrome: A Phenotypic Association of Sporadic Cutaneous Melanoma. Elder DE, Goldman LI, Goldman SC, Greene MH, and Clark WH Jr. Cancer. 1980;46:1787–94.

"Clinical photographs of 79 prospectively studied cases of non-familial cutaneous malignant melanoma were reviewed; special attention was directed to the distribution pattern of coexistent melanocytic lesions. A group of 15 patients had moles on the covered buttock area. Seven of these patients had large clinically atypical nevi, and biopsies of these nevi showed severe melanocytic dysplasia. Residual elements of melanocytic dysplasia were identified in five of the primary melanomas in this group of patients. It is suggested that these patients represent a distinctive syndrome, the Dysplastic Nevus Syndrome (DNS) and that they are at increased risk for development of primary cutaneous malignant melanoma. The clinically and histologically distinctive dysplastic nevi of these patients are identical to the precursor lesion for melanoma that we have previously described in a familial context, the B-K mole syndrome. This paper represents the first description of this form of dysplasia in non-familial melanoma."

In this article, Elder and co-authors claimed to have recognized "melanocytic dysplasia" for the first time in atypical nevi of persons with "non-familial melanoma," and, therefore, to have identified the "dysplastic nevus syndrome (DNS)" in patients with "sporadic melanoma" (Fig. 7). According to them, the only difference in the syndrome as it appeared in those two settings pertained to histopathologic findings, namely, a tendency to a pattern of "lentiginous melanocytic dysplasia" in DN** of those with sporadic melanoma and to "epithelioid melanocytic dysplasia" in those with familial melanoma (Fig. 8, their Table 3).

Elder et al.: "The microscopic features of these dysplastic nevi appear to fall into two classes. The first is described and illustrated in our original reports. The intraepidermal component of these nevi is arranged in forms of nests, usually with horizontally oriented epithelioid melanocytes. More recently, we have identified a second class of dysplastic nevus, which we currently classify histologically as lentiginous melanocytic dysplasia . . . In these nevi, which often co-exist in a given patient with the more classic form of B-K mole (epithelioid-cell melanocytic dysplasia), there is irregular basal proliferation of pleomorphic hyperchromatic melanocytes that show prominent cytoplasmic retraction artifact. The hyperplastic melanocytes are commonly disposed at the margins of elongated rete ridges."

Critique: Clark and co-workers discuss "lentiginous melanocytic dysplasia" for the first time here, and consider it to be the most common histopathologic pattern in DN (Fig. 9, their Fig. 6). But in 1978, only the epithelioid pattern was described in B-K moles by Clark and co-workers. By 1981, Elder et al. had proclaimed that the lentiginous pattern, not the epithelioid one, was the more common in DN.⁹ The shift from "epithelioid" to "lentiginous" is in keeping with other sudden changes by these co-workers, among them the change in name from B-K mole to DN, the change from hereditable to sporadic types of B-K moles, and the change, again and again, in definition of "melanocytic dysplasia." In the article under discussion here, the authors equate "melanocytic dysplasia" with atypia, in contrast to their article in 1978 in which they employed "melanocytic dysplasia" as a synonym for atypical melanocytic hyperplasia. If "dysplasia" is truly synonymous with atypia, why is is necessary to invoke "dysplasia" at all?

Elder et al.: "Though the B-K mole syndrome was originally described in familial matrix, we have seen apparently similar cases in sporadic (non-familial) melanoma. Consequently, the present study was undertaken to determine whether or not special melanocytic nevi, either of the B-K mole type or otherwise, are commonly associated with non-familial melanoma."

Critique: B-K moles are distinctive but not special; they are the most common type of melanocytic nevus, occurring as they do in the vast majority of Caucasians, very few of whom are members of families in which melanomas develop on a genetic basis. For at least 150 years it had been known that some melanomas develop in pre-existing nevi. In fact, Unna, in the German language edition of his textbook in 1892, stated incorrectly that all melanomas arise in pre-existing nevi.

Elder et al.: "Group III was composed of 59 patients who had essentially normal nevus patterns. Their lesions were small, completely benign in appearance, not present on the usually covered areas of the skin, and normal in number. Biopsy, other than of melanoma, was not indicated at their initial clinical evaluation, nor was it performed."

Critique: There are no normal nevus patterns, normal number of nevi, and normal nevi. Melanocytic nevi are all pathologic and there are many patterns of distribution of them. When Clark and co-workers did at last biopsy nevi that were "small, completely benign in appearance," they found the very same histopathologic findings as they encountered in B-K moles. That is why the size of B-K moles shrunk from 10 mm in publications of Clark et al. in 1978 to 5 mm (as DN) in publications of the same group in 1980. In brief, their "normal," "banal," "ordinary," and "common" nevi are merely small versions of B-K moles.

Elder et al.: "Microscopic atypia in these lesions was characterized by two common features: intraepidermal melanocytic dysplasia and a dermal inflammatory host response consisting of a focal or diffuse lymphocytic infiltrate with delicate fibroplasia and new vessel formation."

Critique: "Dermal inflammatory host responses" are not signs of microscopic atypia; only nuclear abnormalities are. Futhermore, atypia of melanocytes is not seen in DN, but is observed commonly in Spitz's nevi and in melanomas.

Elder et al.: "The cytological appearances that were common to all dysplastic nevi were nuclear pleomorphism and hyperchromatism, seen in a few or many cells of each lesion. Mitotic figures were not frequent."

Critique: The authors stress nuclear atypia as the common denominator in DN, yet scrutiny of their photmicrographs reveals no nuclear atypia (Fig. 10, their Fig. 7). The insignificance of nuclear characteristics of melanocytes in regard to diagnosis of these nevi is apparent from the fact that DN can be identified easily by assessment of architectural pattern alone at scanning power (1.0–2.53) of a conventional microscope, a magnification at which nuclear detail cannot be appreciated.^10,11 Criteria for diagnosis of a DN at scanning magnification are silhouette of a benign neoplasm in lesion that is only slightly elevated and confinement of nests of melanocytes to the dermo-epidermal junction and papillary dermis.12 Nuclear atypia and lymphocytic infiltration are irrelevant to diagnosis of DN. In short, the criteria for DN proposed by Elder et al. do not work.

Elder et al.: "The most common sub-type, lentiginous melanocytic dysplasia, was seen in all the dysplastic lesions."

Critique: It is curious that every "dysplastic nevus" studied by the authors in preparation for publication of their manuscript exhibited this new lentiginous pattern, yet "lentiginous melanocytic dysplasia" was not even mentioned in any of their previous articles about the same nevus. The word dysplasia, as used by them in the sentence quoted, has two different meanings, to wit, "lentiginous melanocytic dysplasia" (imprecise histopathologic description) and "dysplastic lesions" (imprecise generic classification).

Elder et al.: "This study was designed to evaluate phenotypic associations of sporadic (non-familial) cutaneous melanoma, with particular emphasis on the number and distribution of melanocytic nevi. We have identified two phenotypes, one of which (Group I) is characterized by an absence of nevi. This group may include patients whose melanoma developed de novo or from a single dysplastic nevus, residual elements of which were present in 2 cases. The second phenotype (Group IIB) is characterized by a prominent nevus pattern. This group was identified by the criterion of presence of nevi on the buttuck area as visualized in clinical photographs. Nevi in these patients had displayed significant clinical atypia at the time of presentation with melanoma, and biopsy had been recommended in each case. Melanocytic dysplasia was seen in one or more lesions from each patient in whom biopsy was performed."

Critique: Classification into 2 groups, i.e., "an absence of nevi" or "a single DN" vs. "a prominent nevus pattern" is arbitary. Any judgment about "clinical atypia" must be predicated on criteria already established for clinical typia, and there are none. "Clinical aytpia" is a fiction. Furthermore, there is no correlation between so-called clinical atypia or dysplasia on one hand and histopathologic findings on the other. Even those nevi misdiagnosed clinically as thin melanomas show stereotypical findings of the commonest of all nevi, namely, silhouette of a benign neoplasm, slight elevation, and nests of small oval monomorphous melanocytes confined to the dermo-epidermal junction and papillary dermis.

Elder et al.: "On the basis of observations in Group IIB and in previously reported patients with familial melanoma, we believe that the dysplastic lesions described in this report may behave as formal histogenetic precursors of melanoma."

Critique: Elder and co-workers verify the well-established fact that some melanomas develop in pre-existing nevi, but they do not define what they mean by "formal histogenetic precursors of melanoma." In fact, no coherent meaning for "precursors" is offered by them in any of their papers about DN. Melanomas usually develop novo (more than 90% of melanomas in persons of all races taken together do not begin in association with a nevus), but melanomas may develop in continuity with a nevus of several different kinds, e.g., acquired ones such as Clark's and Unna's, and congenital ones like Miescher's and giant hairy type. In Caucasians, the majority of nevi in continuity with melanomas are Clark's nevi.

Elder et al.: "We believe that dysplastic melanocytic nevi are important stages in a process aptly defined by Foulds as 'the material continuity of suspected precursor lesions' in the developmental biology of neoplasia. Three lines of evidence support this hypothesis. First is the finding here reported of dysplastic nevi in at least 6 of 79 patients with sporadic melanoma. Our method of case indentification was necessarily crude. When clinical criteria developed since this study began (and summarized below) are applied prospectively, it is likely that the incidence of detection of dysplastic nevi in sporadic melanoma will be signficantly greater than the 8% observed in this series."

Critique: Foulds, the person whose thinking about developmental biology so greatly influenced Clark and students of him, is quoted often by advocates of the concept of "melanocytic dysplasia," but the content of the quotations is difficult to fathom. What does "the material continuity of suspected precursor lesions" mean? Elder et al. found that 8% of patients with sporadic (nonfamilial) melanoma possess these nevi, but that number lacks authenticity because the authors do not present convincing, repeatable, and reliable criteria for either clinical or histopathologic diagnosis of DN. Most Caucasians sport DN whether or not they ever develop melanoma. In brief, the number of persons with melanoma who also bear DN is closer to 80% than 8%.

Elder et al.: "The second line of evidence is our observation in Group IIB of residual elements of intraepidermal dysplasia adjacent to each of the five primary malignant melanomas of the superficial spreading type. This adjacent dysplasia is the lesion that we had tentatively classified as a variant of SSM. The microscopic appearances are, however, clearly distinguishable from classical pagetoid SSM and are identical to those of the growth pattern that we now recognize as lentiginous melanocytic dysplasia."

Critique: Elder et al. acknowledge that originally they had classified the changes at the shoulders of what they called CAN (in reality, small versions of B-K moles) as those of melanoma, but they now regard the changes as those of "dysplasia," i.e., a formal histogenetic precursor of melanoma. But those changes are present at the shoulders of virtually all compound DN and less than 0.01% of those nevi ever eventuate in melanoma. How formal a precursor can those changes be?

Elder et al.: "The third and most direct evidence for the precursor role of dysplastic nevi is the relatively sudden development of malignancy in a continuously observed, previously stable, clinically dysplastic lesion."

Critique: There is no disagreement about the observation that some melanomas arise in association with DN. Nonetheless, 99.99% of these nevi never are a nidus for development of melanoma. How can they be considered formal precursors when the overwhelming majority of them are never linked with melanoma? On the basis of morphologic findings in an individual lesion alone, no clinician or histopathologist can predict which DN will become a site for development of melanoma.

Elder et al.: "The clinical and histologic appearances of the DNS are similar whether the syndrome occurs in a familial melanoma kindred (B-K mole syndrome), in patients with sporadic melanoma (as reported here), or sporadically in patients without melanoma (unpublished observations). For this reason, we now feel that the syndrome should be classified as DNS, sporadic type. The familial type may also be called the B-K mole syndrome. Patients who currently seem to have the sporadic DNS may eventually develop a familial transmission pattern. We have already seen at least two examples of the "sporadic" DNS (not reported here) in whose families melanoma later developed."

Critique: It was necessary for Elder and co-workers to create a new classification of DN because Clark and collaborators had insisted in their original communications about B-K moles that those reputedly uncommon nevi were a marker of persons with familial melanoma at increased risk for developing melanoma. Months after this concept had been stated in print, it became obvious to dermatologists and pathologists, Clark and Elder among them, that B-K moles were extremely common and that the vast majority of them occurred in persons who had no family history of melanoma. In order to sustain the notion of a nevus that represented a "formal histogenetic precursor of malignant melanoma," it was necessary to create a category of sporadic melanoma in addition to that of familial melanoma.

Elder et al.: "The prototypic dysplastic nevus is larger than CAN and is irregular in outline and color. Tan and brown hues predominate, while areas of erythema and slight scaling may develop in lesions with a brisk host response."

Critique: Elder and associates continue to contrast DN with CAN yet what they describe and picture as CAN are simply smaller versions of DN. The authors refer to common acquired nevus as if it were a single, well-defined type of nevus, when, in actuality, there are several well-defined CAN, e.g., those of Unna, Spitz, and, most common of all, Clark.¹² Most DN are not larger than other types of acquired melanocytic nevi, and all DN begin tiny, the majority of them never exceeding a few millimeters in diameter. Most DN are distinguished with ease from melanoma clinically, and nearly all of them are differentiated readily from melanoma histopathologically. Nonetheless, these co-workers persist in describing them as being indistinguishable clinically from melanoma and in sustaining the fiction that they are similar histopathologically to melanoma (at first, by their own admission, they interpreted the changes at the shoulders of B-K moles to be those of intraepidermal melanoma).

Elder: "Though the syndrome was first recognized in patients with 200 or more dysplastic nevi, presentation with ten or fewer or even a single lesion is not uncommon."

Critique: This is the first time that Clark and co-workers stated explicitly that a single DN constitutes the DNS. If this were true, most Caucasians would have the syndrome.¹¹

Elder et al.: "In the Pigmented Lesion of the University of Pennsylvania, all patients who clinically exhibit a dysplastic nevus syndrome have at least two selected nevi removed for histologic study. If the lesions histologically show melanocytic dysplasia, the patients are assigned to the dysplastic nevus syndrome. They are then followed carefully with photographic documentation of changes in any melanocytic lesion. Follow-up studies of this nature will result in the identification and removal of any lesions that undergo incipient malignant change at a stage when cure should routinely be achieved by simple surgical procedures."

Critique: The first sentence of this quotation illustrates how tenuous are criteria for DNS. The authors state that the diagnosis is made clinically, yet they proceed to insist upon confirmation histopathologically in sections from two biopsy specimens of purported DN. Is the diagnosis of DNS based on clinical features alone, histopathologic findings alone, or combination of those features and findings? The answer is not given in this article and has never been given until now. The number "two" (selected nevi) is just as arbitrary as everything else that has been written about this syndrome.

On the basis of an expected histopathologic finding in an exceedingly common nevus, i.e., extension of the junctional component beyond the intradermal one ("shoulders"), a finding that Clark designated "melanocytic dysplasia" in order to adapt it to a preconceived series of steps in tumor progression claimed to have been demonstrated in lesions in organs such as the breast, cervix, and colon, these collaborators created a syndrome. If the thesis of Clark and co-workers is accepted, then there are billions of DN in the skin of persons throughout the world, Caucasians especially, and at least hundreds of millions of persons with the DNS.

Last, it should be stated that most melanocytic proliferations that appear in persons with DN are not melanomas, but newly arisen DN. In these persons, new DN continue to appear throughout life, even into old age.