2. Critique of Positions of Clark et al. Changed Little or Not at All Since 1978

 
1) Elder et al.: "We will review evidence that nevi, especially atypical or dysplastic nevi, are important simulants of melanoma, potential precursors of melanoma, and markers of increased risk for mela-noma."
 
Critique: DN are not simulators of melanomas histopathologically5–6 (they are characterized by findings polar to those of melanoma) and only rarely are they confused clinically for melanoma (just as is the situation uncommonly for processes as dissimilar as seborrheic keratosis, pigmented basal-cell carcinoma, hemangioma, dermatofibroma, and small congenital nevus). Simulation clinically of sets of conditions, e.g., papules of urticaria pigmentosa and juvenile xanthogranuloma, lichen planus-like keratosis and basal-cell carcinoma, and pyogenic granuloma and eccrine poroma does not imply a relationship between them in terms of either cause or pathogenesis. All types of congenital and acquired melanocytic nevi may be forerunners, rarely, of melanoma, but the mere presence of a single acquired melanocytic nevus or a few of them, no matter the type, puts no person at undue risk for melanoma.
 
2) Elder et al.: "Melanocytic nevi may be regarded as benign neoplasms of melanocytes."
 
Critique: Although acquired melanocytic nevi may be regarded as benign neoplasms, congenital nevi, giant hairy nevi being the quintessential example of them, are hamartomas.
 
3) Elder et al.: "At any stage of their evolution, common acquired nevi are usually smaller than 5 mm in diameter and are rarely larger than 10 mm. They are symmetrical, with regular well-defined borders. There may be slight nuclear size and shape variation, but there is no high-grade atypia."
 
Critique: This description of CAN applies equally to the overwhelming majority of DN, which is not surprising when it is remembered that DN are simply one type of CAN. Parenthetically, what is "high-grade atypia"?
 
4) Elder et al.: "Because the term architectural disorder has not been rigorously defined and lesions identified using this classification scheme have not been subjected to clinical or epidemiological study, we shall use the term melanocytic (histological) dysplasia for the constellation of histological features under discussion. We are also concerned about the lack of an agreed definition for the term atypical mole, and we will use the term clinically atypical (melanocytic) nevi for lesions that meet the 1983 Consensus clinical criteria for dysplastic nevi."
 
Critique: Clark and associates consistently and conveniently ignored views contrary to theirs about the subjects of "melanocytic dysplasia," "DN," and the "DNS," in this instance, the recommendations of the Panel of the NIH Consensus Conference of 1992, to wit, that the term "nevus with architectural disorder" be the histopathologic diagnosis and "atypical mole" be the clinical diagnosis for what the Panel of the Consensus Conference of 1983 had designated "dysplastic nevus."7–8 It is curious that Elder et al. suggested discontinuing use of the term "architectural disorder" in favor of "dysplastic nevus" on the grounds that the former "has not been rigorously defined," when, in actuality, the latter has never been rigorously defined by them, a fact that prompted the Panel of the NIH in 1992 to propose that the designation DN be avoided. Parenthetically, it was Clark and co-workers who popularized the terms "atypical mole" and "funny looking mole," neither of which they ever defined in intelligible fashion.
 
5) Elder et al.: "Clinical features of atypical (dysplastic) nevi discussed in early reports summarized in the 1983 Consensus Statement include consideration of size, profile, border, and color variegation. . . . Dysplastic nevi were defined as lesions greater than 5 mm in diameter, with a macular component always present. . . . Although the border is usually fairly regular, it is characteristically ill-defined or "fuzzy" around all or a part of its circumference. . . . The lesions are usually brown or tan, with some variation of hue and saturation, but without black in most cases."
 
Critique: The stereotypical clinical description of DN given in virtually every article by Clark and fellow workers in the last two decades does not permit differentiation of those nevi from melanomas. Furthermore, when DN were called B-K moles by Clark et al., they were described invariably as being large, i.e., considerably greater than 5.0 mm in diameter. In fact, for a short time between the demise of the designation B-K mole and the ascent of the title DN, Clark and followers dubbed those nevi "large atypical moles."9 It is not to the credit of the Panels of the NIH Consensus Conferences of both 1983 and 1992 that they gave their imprimatur to criteria for clinical diagnosis of DN that were identical to those employed by physicians universally for diagnosis of melanoma clinically. Patients were not served well by such lack of critical acumen.
 
6) Elder et al.: "In summary, the three salient histological features of melanocytic dysplasia are the presence of an immature or disordered lentiginous pattern of growth, of mild-to- moderate focal (random) cytological atypia in melanocytes, and of a lymphocytic host response."
 
Critique: Clark and colleagues consistently made their own rules and then proceeded to ignore or violate them; their criteria for melanocytic dysplasia are no exception. During the course of only a few years, the same group of researchers defined melanocytic dysplasia as synonymous with atypical melanocytic hyperplasia, melanocytic nuclear atypia, and a constellation of a particular pattern (lentiginous) of growth, nuclear atypia of melanocytes, and infiltration of lymphocytes, i.e., DN, to mention but some of their definitions. The last mentioned definition is the one that has been employed by them for nearly a decade. It should be noted, however, that "lentiginous," as it is used in dermatology and pathology, is descriptive of very different findings in very different conditions as unalike as simple lentigo, solar lentigo, labial lentigo, 'ink spot' lentigo, and lentigo maligna. Acrolentigi-nous melanoma and lentigo maligna are melanomas, simple lentigo and labial lentigo are unrelated benign proliferations of melanocytes, and solar lentigo is a benign pigmented proliferation of keratinocytes. In short, the term "lentiginous" is imprecise. "Immature" and "disordered" are inappropriate terms for a pattern of a particular nevus, i.e., DN, which, like benign neoplasms in general, is symmetric and well circumscribed. No definition is provided for the mystical phrases "immature or disordered lentiginous pattern of growth" and "random mild-to-moderate cytological atypia." The presence of a "lymphocytic host response" is not a feature peculiar to DN, being found as it is in such dissimilar mela-nocytic neoplasms as Spitz's nevus and melanoma.
 
7) Elder et al.: "Our criteria to distinguish dysplastic and common nevi are summarized in Table 1 [our Fig. 2] and illustrated in Figs 1 and 2" [Our Figs. 3–4].
 
Critique: The criteria for histopathologic diagnosis of CAN and DN set forth by Elder et al. in the early 1980's and those advocated by them more than a decade later are basically the same, but they lack legitimacy because DN is simply one, albeit the most common by far, of several types of CAN. [Fig. 2–4]

View Figure  		Fig. 2  (orig. Table 1) The Table improperly compares "Melanocytic Dysplasia," which is a descriptive term, with "Common Nevus," which is a histopathologic diagnosis.

View Figure  		Fig. 3  (orig. Fig. 2A) The melanocytic lesion depicted in this Figure is a junctional Clark's nevus. The authors arbitrarily identify these changes as those of "moderate melanocytic dysplasia." However, they neither provide criteria for dysplasia nor for the "grade" of it.

View Figure  		Fig. 4  (orig. Fig. 2B) "Cytological atypia" is not apparent to our eyes and what "forms a bridge" is actually two nests of melanocytes that seem to be fused only because the section has been cut at a tangent.
 
8) Elder et al.: "In our experience, melanoma arising in a precursor nevus has generally been perceived as a clinically and histologically focal event in an otherwise clinically stable lesion. Melanoma may present as a small focus of darker pigment that expands within the precursor lesion, and ultimately replaces it completely. In contrast, melanomas that develop without a precursor lesion typically demonstrate changes in color, topography, and border that are global throughout the lesions as they grow from smaller to larger macules, patches, or nodules."
 
Critique: All melanomas, both those that arise de novo and those that develop in a pre-existing nevus, begin as a tiny flat lesion that, when pigmented, usually can be identified clinically, in time, for what they are by being asymmetric, having scalloped borders, and showing chromatic variation in shades of brown. It is not always possible, though, to judge clinically whether a melanoma arose de novo or developed in a pre-existing nevus because the neoplastic melanocytes of melanoma may overrun the melanocytes of the nevus, sometimes destroying it entirely. In nearly all instances, a melanoma that begins in association with a pre-existing nevus does so at the very periphery of the nevus and not as "a small focus of darker pigment that expands within the precursor lesion. . . ."
 
9) Elder et al.: "Large studies of the relationship between nevi and melanomas have been published (Table 2) [our Fig. 5]. Between 5% and 39% (median 22%) of melanomas exhibit a contiguous melanocytic dysplasia, and banal (dermal) nevic cells are present in an additional 10% to 21% (median 15%) of cases."
 
Critique: Elder et al. make a distinction between "melanocytic dysplasia," a term they apply to abnormal melanocytes situated in the epidermis, and "banal nevic cells," a phrase they invoke for abnormal melanocytes of a nevus positioned in the dermis, but they do not make clear whether the melanocytes of the so-called dysplasia are the equivalent of a junctional type of DN, "intermediate" between a nevus and a melanoma, or elements entirely of a melan-oma, and they do not make plain whether the "banal nevic cells" in the dermis are those of what they call a CAN or a DN. In this context it is not surprising that the percentages given by them for the "relationship between nevi and melanomas" vary so greatly. The considerable discrepancy in these estimates is a consequence of different authors having employed different histopathologic criteria (or no criteria) for these findings. Whatever the reasons, the conclusion is obvious: repeatable criteria for diagnosis of DN by conventional microscopy are not being used.

View Figure  		Fig. 5  (orig. Table 2) The wide variety of numbers about "contiguity" of nevi and melanoma reported on by different authors is due to the lack of consistent, reliable criteria employed in the diagnosis of these lesions. Note that Ackerman et al. (1993) are claimed to have identified "melanocytic dysplasia" in contiguity with 20% of melanomas, but Ackerman has never, ever employed the term "dysplasia."
 
10) Elder et al.: "Considering that banal nevi are much more common than dysplastic nevi, the latter appear overrepresented as precursor lesions."
 
Critique: The authors persist in contrasting "banal nevi" and DN when DN is the most common (banal) of all types of melanocytic nevi. As a result, inferences like "overrepresented" lack validity.
 
11) Elder et al.: "Table 3 summarizes our major histological criteria for distinguishing between nevi with melanocytic dysplasia and RGP melanomas illustrated in Figs 2 and 3" [our Figs. 3–4 and 6].
 
Critique: In Table 3, despite its many inaccuracies, Elder et al. acknowledge, by implication of having constructed two columns of contrasting morphologic findings, how very different DN (presumably "nevus with melanocytic dysplasia" is synonymous with DN) are histopathologically from melanomas, a not unexpected conclusion when one considers that the former are benign neoplasms and the latter are malignant ones. In brief, Table 3 demonstrates that DN are opposite melanoma in terms of histopathologic findings, not "intermediate" between CAN and melanoma.

View Figure  		Fig. 6  (orig. Fig. 3A) In 1984 and thereafter, Clark and co-workers (Human Pathology, 1984) listed the following criteria for specific diagnosis of "dysplastic nevus": 1) Lentiginous melanocytic hyperplasia 2) Random cytological atypia 3) Lamellar fibroplasia 4) Concentric eosinophilic fibroplasia 5) Patchy lymphocytic infiltrates. Note that the melanoma in situ they pictured in this figure fulfills those criteria.
 
12) Elder et al.: "Further, the existence of lesions with properties that are intermediate between wholly benign lesions and established cancers is a test of the tumor progression hypothesis."
 
Critique: Clark and co-workers began with a speculation (tumor progression is a recognizable pathway expressed in 6 distinctive steps) and proceeded to force clinical and histopathological findings of DN into that artificial, hypothetical construct. That the exercise failed is apparent in Table 3 which gives the lie to the hypothesis; DN are not intermediate between CAN and melanoma; they are a particular, exceedingly common type of CAN. Furthermore, the vast majority of melanomas, worldwide, arise de novo, and not in a nevus of any kind. Last, as we view melanomas in continuity with a "DN," through a conventional microscope, both components are identifiable readily for what they are; no component of the proliferation of melanocytes qualifies morphologically as being "intermediate."