The taxonomy of melanoma is related to selections, from manifold offerings, of virtual images. It is additionally influenced by the fickleness of the selector; the selections are individualized. There is the option of characterizing melanoma as a single disease, biologically and histologically. In this approach, the designation, melanoma, embraces both thin and thick lesions, and low and high grade lesions. Distinction between round and spindle cell lesions are not relevant.
In an alternate approach, the wide range of real images in collections of "melanomas" are matched by diverse virtual images, and by a broad taxonomic spectrum. Observed real images, selected virtual images, and the respective taxonomic selections have an interdependency. The designation, melanoma, with a variety of modifiers acquires a relativity to various real and virtual images. In addition, some of physical parameters in this broad category are insufficient to carry with them a serious threat for the welfare of the affected patients. In other examples, the definition of a variation may be too poorly documented to provide a statistical basis for meaningful prognostications; qualifiers then are required to denote the imprecision of the designation, melanoma. In the final analysis, the only true validation of a diagnosis of melanoma is to be found in documentation of the subsequent clinical course, after the primary lesion has been excised, and hopefully controlled.
Nuclear grade and prognostications: In the general category of melanomas, histologic "grade" and size ("age") seem to be correlated. Most large (thick) melanomas are high grade; they not only tend to be high grade lesions, but commonly are polymorphic in vertical growth (they are old enough to have experienced many intralesional transformations). With size as the chief criterion in prognostications, and also as a measure of the "age" of a neoplasm, few "malignancies" are as unpredictable, or as prone to metastasis at an "early age" as melanoma. Many thin melanomas are low, or intermediate in regard to histologic grade (e.g., nuclear grade). Thin lesions of low nuclear grade have been given recognition in the concept of MDM. Generally, it is convenient to utilize three nuclear grades.
On occasion, the degree of atypia in melanocytic neoplasia is extreme: such lesions have an anaplastic quality; they might be characterized as grade IV. In some anaplastic lesions, polymorphism is also a feature (e.g., melanoma may even masquerade in the generic category of atypical fibrous xanthoma). For the common dysplasia-melanoma syndromes (including those in the setting of the dysplastic nevus-melanoma syndrome, the LM-LMM syndrome, and the acral lentiginous dysplasia-melanoma syndrome), small, thin lesions are less likely to show a high nuclear grade than are thick lesions. In part, deviations in degree of atypia in thin melanomas of the common (typical) categories are given recognition in the concept of minimal deviation melanoma.
Variant vertical growth, a common feature of thin, typical melanomas, might be cited as an additional deviant parameter encountered in the minimal deviation categories of the common dysplasia-melanoma syndromes. In the most common sequence of dysplasia to melanoma, as expressed in SSM, the early stages of vertical growth often are not only variant in patterns, but are additionally characterized by fibrosing reactions of the type characterized as arrested variant growth.
The quality of low to intermediate nuclear grade (grades I & II) as often manifested in thin, common melanomas is not a regular attribute of some of the less common variants of minimal deviation melanoma (e.g., the Spitz variant).
The very fact, that most examples of low qrade lesions are thin (less than 1 mm in height), provides an indication for appending qualifications to, or in place of, the diagnosis, "melanoma" (e.g., borderline melanocytic neoplasia of indeterminate malignant potential) .
In practice, size, as evaluated by a measurement of the dimension of a melanocytic neoplasm along an axis that is perpendicular to the epithelial surface - in this orientation, the axis also should overly the greatest expanse of tumor - is the prime prognostic parameter. In consultative practice in the histologic evaluation of a specific lesion, conflicts will occasionally be discovered when several such measurements, all by different observers but presumably of the same general area, are compared. Often, the conflicts relate to the interpretation of variable patterns, particularly the interpretation of the significance of a population of nevus-like cells at the deep margin of the lesion. In general, the mild atypia often displayed in rernnants of a nevus at the margin of a melanoma should be ignored in defining the parameters of a vertical growth component; the nevus celllike population should not be included in the prognostications. The nevus-like variant of MDM is an exception.
Vertical growth: In defining vertical growth (i.e., patterns III & IV) as the sine qua non for the diagnosis of melanoma, the criteria for the recognition of early vertical growth are based on extrapolations from the patterns in a fully evolved vertical growth component of "real" (i.e., category Ill) melanoma (a lesion measuring 2.5 mm in vertical dimensions would be an example of a "real" melanorna) . Neophytes or the uninformed (e.g., surgeons) will often assume that there are predictable correlations between Clark's criteria and Breslow's criteria. They will assume that there will be a precise cut-off in vertical dimensions at the transitions from level II to level III and from level III to level IV. In practice, the correlations are imprecise with significant overlaps. A level IV spindle cell lesion may be as thin as the thinest level III lesion, and some level II lesions may be "thicker" than a level III or IV lesion. Some thin melanomas, with a minimal representation of the requisite number of nests for the recognition of vertical growth, may measure as little as 0.3 mm in vertical dimensions.
Generally, in the evaluation of melanomas, depth of invasion (Breslow) is likely to be considered as a correlate of bulk - a tumor at level V would be expected to be a bulky lesion. Actually, invasion to Level V and significant tumor bulk, at that same level, are not always correlated. Some melanomas may enter vertical growth in a superficial site, and then extend in situ along and within skin appendages. Subsequently, a second vertical growth component, near or at level V, may issue from the adnexal component without having attained significant overall bulk (i.e., secondary loci of vertical growth arising from neoplastic melanocytes affecting the skin appendages). Such lesions may behave as would a bulky, solid level V lesion: level V may be a significant measure of prognosis without attention to bulk, if the component at this level also shows diffuse vertical growth.
Lentigo maligna melanoma is an example of a melanoma that may enter vertical growth from components involving hair follicles, or the sweat gland apparatus. The deep portions of the respective lesions at level V may be relatively small in crosssectional diameter but may show diffuse vertical growth, with infiltration among collagen bundles of the reticular dermis.
Some examples of SSM may enter vertical growth deep in the dermis, or sub-cutis, from components involving the sweat gland apparatus and, at level V, the vertical growth components of such lesions rnight have small cross - sectional diameters. Such lesions usually enter vertical growth at this level in patterns of diffuse infiltration (diffuse vertical growth). These qualities in combination with a rich vascular plexus may provide an explanation for the aggressive behavior of some of these lesions.
Malpractice - standard of care: In the prosecution of claims of malpractice, lawyers often fuss over the manner in which the dimensions of a melanoma relate to prognosis. Another point of contention is "standard of care " which in part involves a judgment regarding how other pathologists would have performed, if confronted with the material in contention. In particular, the question posed is "would other path- ologists have been more "likely than not” (i.e., >50%) to render a correct diagnosis, or to commit the same “error” as the defendant. To put this approach in proper perspective, we should also define at what histologic stage (Breslow) 51% of patients, all with a similar lesion, would be committed to a fatal course. For purposes of discussion, a figure such as 2.5 mm might be offered as an arbitrary cutoff point. In other words, the likelihood of a fatal course above the cutoff point would become "legally" significant. Only beyond this point would statistics become "valid" (using the biased parlance of lawyers) as the basis for a claim of malpractice. Below the point, the assumptions that the patient is committed to a fatal course is mostly a guess, with accuracy dependent on the vagaries of nature; not the prescience of a pathologist.
Documentation of real and virtual images: A report without a microscopic description is certainly acceptable by current standards, but does not provide evidence that the pathologist's observations, and his evoked virtual images, are relevant to the real images, or that the final diagnosis will be a guide to clinical behavior, or treatment. It may come to pass that such a report will be cited as a deviation from a "standard of care," but such an assessment would not automatically identify the related performance of a pathologist as negligent. On the other hand, careful evaluations of histologic and cytologic features, and documentation of the evaluations, in a report, go a long way in resolving problems, if there is a question of culpability.
Ulceration: Ulceration is common over the vertical growth components of melanomas. Its prevalence is directly proportional to the "thickness" of the respective vertical growth components. Having been often cited as a significant prognostic parameter, efforts have been expended to provide a definition of what constitutes significant ulceration. Objectively, ulceration is nothing more than a covariant of "thickness" (i.e., size of vertical growth component). The significance of the frequency of either typical, or variant, vertical growth components as related to the frequency of ulceration has not been examined.
Density and distribution of lymphoid infiltrates have often been emphasized in prognostic evaluations. Tumor infiltrating lymphocytes have been cited as prognostically important.
Although mostly denigrated, level IV invasion has utility particularly in the evaluation of thin lesions. If, in developing therapeutic options, emphasis is placed on scant infiltrates of lymphocytes, the pathologist should also record whether those examples which are relatively free of lymphoid infiltrates are also at level IV. The two features are correlated. Often, they are also correlated with fascicular, spindle cell patterns.
Regression: An appreciation of areas of regression requires the manipulation of virtual images. In the process, virtual images in which the destructive effects of a cellular immune reaction are integrated with the real images of markers left behind after a population of neoplastic melanocytes has undergone extensive, or complete, lysis. Regression is common in the setting of halo nevus, and may be encountered focally in the dysplasias. The lysis of melanocytic cells in association with dense infiltrates of lymphocytes might be characterized as halo nevus phenomena without attention to the clinical presentation. The diagnosis of halo nevus does not require clinical documentation of a de-pigmented halo. The histologic patterns of the end- stage of halo nevus phenomena, in either halo nevus, or in an area in which the vertical growth component of a melanoma has undergone complete regression, are basically indistinguishable.
Some melanomas, having metastasized , then regress, but do so only in the primary sites. In this sequence, a careful examination of the patient and a history of a lesion, that spontaneously regressed or "fell off," may provide important clues for the identification of the site from which the metastases originated. Some of these lesions, if manifesting partial preservation of a dermal component will, on occasion, also show patterns of the halo nevus phenomena: they probably represent examples of regressing MDM of halo nevus type. From this perspective, some halo nevi serve as the nidus for neoplastic progressions leading to a special variant of melanoma, and some of these lesions may metastasize during the evolution of the melanoma. Such a sequence does not then preclude extensive, or complete, regression of the primary lesion.
Biased manipulation of real images: If, on the histologic examination of a problematic lesion, no melanoma cells are extant or, even if only a few widely spaced nests of atypical cells are represented, a diagnosis of melanoma, that is based solely on the real images, may be difficult to defend. On the other hand, a weak, or biased, expert may interpret the melanophages in such a site as residual melanoma cells; he may even be able to successfully defend such an erroneous position in a court of law.
All of the difficulties a pathologist may experience in the interpretation of the troublesome patterns of a regressed primary lesion become moot in the face of a documented metastasis; a subsequent review of the earlier, problematic material, is likely to seem less problematic to an observer who has the vantage of a historical perspective. To impose, in retrospect, the virtual images of a metastasizing melanoma on the real images of a regressed primary melanoma, as seen in the initial histologic preparations, is presumptuous. Such arrogance often leads to expressions of confidence that may not have been justified, if the roles of respective pathologists in the sequence of events could be reversed. Unbiased manipulations of real and virtual images are required to put the images of the prirnary pathologist in proper perspective.
That a metastasis may flourish and the primary source of the metastasis may simultaneously regress is a biologic paradox. The explanation probably is related to peculiarities of anatomic sites, to dedicated systems of immunity in the respective sites, and to an interplay of tumor cells and distinctive stromas. The process of regression in the site of a primary melanoma may adversely alter relationships between neoplastic cells and stroma. These alterations may favor the dissemination of neoplastic cells. With dissemination to new sites, the cells may find sites that are immunologically favorable. In this sequence in which metastases are identified prior to the identification of a site of regression of a primary lesion in the skin, the metastases would have occurred prior to, or during. the regression of a vertical growth component.
In one collection of fanciful virtual images, the growth of occult (latent) metastatic melanoma in a lymph node may be adversely influenced by ”humors" originating in a site of regression in the skin. In this approach, the metastases would have been occult, but would become overt during the process of regression in the primary site.
Conservative prognostic evaluation: Prognostic evaluations are based primarily on size (Breslow's criteria) with other parameters being of lesser significance. In the presence of regression, a qualifier, promoting the caution that any prognostic evaluation may prove to be conservative, is required. Regression in a melanocytic lesion generally is manifested by local interruption of the cellular patterns of a neoplasm. In the void, both the epidermis and the papillary dermis tend to be free of atypical cells. In addition, the papillary dermis tends to be widened; it often is delicately fibrotic; it will contain an increased number of dilated, thin walled vessels, and variable infiltrates of lymphoid cells. Clusters of melanophages are often a prominent feature. If patterns of regression are represented in a lesion which otherwise qualifies as a dysplasia and, if they extend to a lateral margin, it is advisable to recommend a wider excision of the area, even if the atypical cells themselves are not represented at the margin.
Many specirnens from melanocytic lesions are the product of a punch or shave biopsy; some of the specimens are not even from the most worrisome portion of the primary lesion. Commonly on these incomplete biopsy specimens, worrisome histologic patterns extend to margins of excision. The available patterns on such specimens, even though incomplete, should be evaluated by modified Clark's and Breslow's criteria to provide some indication of what the lesion may represent. Often when the entire lesion becomes available, no identifiable residua of the melanocytic lesion will be found; the provisional guidelines then become the final guidelines.
When providing therapeutic guidelines from material which obviously has been incompletely sampled, the interpretation should be qualified as conservative. It should be noted in the report that the prognostic evaluations and therapeutic guidelines must be considered conservative, and that, when the entire specimen becomes available, the evaluations and prognostications may need to be modified. For example, a biopsy of a lesion showing the pattern of either a high grade dysplasia, or a vertical growth component may also show one or both margins to be involved by one or the other. In the face of a compromised specimen in which margins are involved, the lesion should be characterized on the basis of the most significant component that is available for histologic examination. If a portion of a vertical growth component is represented, the evaluation should include a measurement of its vertical dimensions. The diagnosis should also be qualified as being "conservative" and subject to change when the entire lesion becomes available for examination. The re-excised specimen may not contain residuaI tumor; the conservative evaluation must then be utilized as the definitive evaluation.
The conservative prognostic evaluation is a caution to the clinician for either a wider excision, or for the need to not underestimate the potential of a common melanocytic dysplasia on the basis of what may be an incomplete representation of important components, or parameters.
Documentation of uncertainties: If, by histologic criteria , the nature of a melanocytic lesion is uncertain, the uncertainties should be documented along with cautions against overly aggressive treatment. In these conundrums, a consultation (in writing and from someone actively engaged in consultation work) may lessen the hazards for both patients and physicians. Concern for the manner in which both histologic criteria and predictions of clinical behavior may influence the care and well being of a patient are important in formulating a diagnosis. Unfortunately, such concerns may also color the interpretative process and modify the manner in which virtual images, having once been imposed upon real images, will impact upon the significance of the real images.
In the examination of real images of a histologic section, an inability to muster relevant virtual images places constraints on attempts to formulate a firm diagnosis. If, in this quandary, melanoma is a tentative consideration then, before committing the patient to the impact of extensive surgery, or other forms of aggressive therapy, consultations may put the confounding virtual images in a better perspective.
In the evocation of virtual images for the interpretation of melanomas, individual prejudices find expression in the character of the virtual images. A rather common prejudice, leading to the evocation of fanciful virtual images, holds that cutaneous melanomas first metastasize to regional lymph nodes, and that any subsequent dissemination occurs only from the nodal site. Support for this prejudice has not been documented and clinical observations commonly contradict it. For a lesion with demonstrated capacity for metastasis, the initial rnetastatic site is not predictably limited. Blood-borne and lymphatic metastases may be synchronous, and are a possible eventuality from any primary melanoma once it has entered vertical growth. The factors controlling the selection and expression of metastatic sites are natural for the individual neoplasm; they include genetic alterations, immune responses, stromal peculiarities, and the accidents encountered in the growth of tumors (e.g., vascular or nerve sheath invasion). To ignore all these factors and simply cite delay in treatment as the determinant of a poor result, is to claim an understanding of all of the many biologic attributes. Representatives of the legal industry, as if oblivious to the privileges of nature, commonly behave as if they have such an understanding.
(Continued on NEXT PAGE)