Irrelevance of prognostic parameters: Prognostic parameters such as Breslow's criteria, are based upon statistical data; they primarily are of value in the formulation of therapeutic guidelines (8). In practice, the potential of a particular melanoma for dissemination cannot be reliably predicted by the utilization of prognostic parameters. Only in retrospect can the biologic potential of both thin and thick melanomas be truly discovered. Only those melanomas, that have fulfilled their potential and metasta- sized, are truly manipulatable in statistical analyses. Unfortunately, members of the legal profession habitually abuse these cautions; they assume that the behavior of any given example of a newly diagnosed melanoma can be anticipated by the evaluation of prognostic parameters. In practice, these parameters have little relevance to the inherent potential of newly diagnosed melanomas, particularly thin variants.
Speculations, that a given lesion in the borderland of melanocytic neoplasia (e. g., thin lesions in the common dysplasia-melanoma sequence) had metastasized prior to a local excision of the primary lesion, if mostly based on findings from a histologic examination of the excised primary lesion, will be incorrect more often than not. In this borderland, lesions on both sides of the border might be characterized as borderline melanocytic neoplasia of indeterminate malignant potential. Some lesions of the borderland have the patterns of a dysplasia (patterns I & II) and some have the patterns of a melanoma (thin patterns III or IV). For some examples of vertical growth components, in which as few as 5 or 6 nests of cells are regularly spaced in at least two strata, the patterns themselves become borderline in regard to making a precise separation between dysplasia patterns (i.e., patterns I & II) and melanoma patterns (i.e., patterns III & IV). A definition of borderline melanocytic neoplasia in the premalignant dysplasia-melanoma sequence greatly lessens the threat that our terminology will lend itself to manipulation by the legal industry.
Evidence of progressive malignant disease, having been documented after a interval of observation of a lesion, which initially had been histologically documented only by examination of an incisional biopsy specimen, interpreted as benign, and then followed and observed - but not controlled in the primary site - does not, in retrospect, qualify the primary lesion at the time of the initial histologic examination as a "malignancy.” Significant alterations in the nature and potential of the neoplastic cells may have occurred in the interval between the two encounters: the neoplasm may have progressed and transformed during the interval to something quite different from the neoplasm represented on the initial biopsy specimen. This position would not be defensible in the concept of "one melanorna, biologically and histologically.”
Margins of excision: On histologic sections of biopsy specimens, the margins of excision of premalignant melanocytic dysplasias commonly are involved. A shave is the common technique - the initial biopsy procedure. For low grade (mild to moderate) dysplasias, if the epidermal margins are involved histologically, a re-exicision of the area is not required, if, by an evaluation of clinical criteria (a clinical inspection of the biopsy site shows no evidence of residual lesion), the lesion can be considered to have been completely excised (caveat: if the atypia of a lesion is sufficient clinically to initially provide an indication for a biopsy, if a portion of the lesion remains evident clinically after biopsy, and if the histologic report documents both a dysplasia - regardless of degree of atypia - and histologically involved margins of excision, the original indications remain valid and the lesion should be excised [with an additional shave biopsy being adequate for the low grade lesions] ). For moderately severe to severe dysplasias, if the margins are histologically involved, then a conservative re- excision of the site, with margins sufficient to encompass any clinically evident, residual disease, is recommended.
For melanomas of the common type, if the biopsy has been conservative with only a few millimeters of uninvolved skin, or if the margins are involved by the vertical or radial growth components, a currently popular guideline is a recommendation for a re- excision with a margin of 1 cm for each 1 mm of height in the vertical growth component. To endorse an even more conservative approach may lead to a charge of malpractice, if the subsequent clinical course is complicated by local recurrence, satellitosis, or nodal metastases. There is some evidence that margins less than 1 cm may be associated with an increased incidence of local recurrence. The 5 cm margins, that have been commonly utilized in the past, are currently touted as unnecessary even for "thick" lesions. If faced with the controversy regarding the recommendations for what constitute an adequate margin of excision, respect for the nature of a melanoma may take precedence over recommendations for treatment that are based solely on statistical analyses. Common sense may be an important attribute in providing recommendations regarding the need for wider margins of excision. Even our experts, who promote a conservative approach in the local management of a "real" melanoma, may vary in their approach under special circumstances. A pathologist who finds himself a defendant in a malpractice suit in which the adequacy of margins of excision is a contention may be chagrined to find that a proponent of the more conservative approaches has found fault with the defendant's performance, and has agreed to serve as an expert for the plaintiff (the needs of an expert's ego commonly outweigh his needs to support his recorded recommendations).
Intravascular nests of tumor cells in the immediate vicinity of a vertical growth component are a flag identifying a risk for both local recurrence of tumor and metastasis; they are findings which may be offered in support of recommendations of wider margins of excision.
Local recurrence (a designation embracing both recurrent and locally metastatic lesions) and implications regarding adequacy of margins of excision: Recurrent melanoma often is situated in the lower dermis and subcutis. It tends to be expansile and may even be outlined by a condensation of inflamed fibrous tissue. Many examples are relatively devoid of inflammatory infiltrates. A local recurrence in the excision site (i.e., within 2 cm of the excision site) of a primary melanoma evokes a variety of virtual images. Local recurrence may be in the form of a true local recurrence in which a portion of the original lesion was left behind at the time of the original biopsy. Most "local recurrences” represent local metastases (within 2 cm of the original site of the biopsy), and many actually are immediately adjacent to, or beneath, the original biopsy site.
If the first indication of a mistaken diagnosis following excision of a problematic lesion is a local recurrence within the region of the scar, the histologic character of the local recurrence must be defined. If on review of the initial histologic sections a vertical growth component of a primary melanoma has been cut across at a surgical margin, and if the subsequent recurrence is superficial, is manifested in patterns of vertical growth, and is in continuity with the epidermis (and likely to be associated with remnants of a melanocytic dysplasia, or radial growth component in the overlying or adjacent epidermis), the recurrence would be properly characterized as a surgical persistence of a primary melanoma (i.e., the original biopsy cut across tumor).
If, in a somewhat different scenario, sections of a primary melanoma, on review, show the histologic patterns of only radial growth at the surgical margin, then the nature of the radial growth component must be defined. Much of what currently is accepted as radial growth is best interpreted as a persistence of the population of dysplastic cells from which the vertical growth component ensued (3). If only radial growth had been cut across by an initial incomplete excision, and the lesion subsequently recurred both superficially in the dermis (in vertical growth patterns) and at the dermal-epidermal interface (in radial growth patterns), then conceptually the vertical growth component of the "recurrent" lesion may in fact represent a new primary melanoma; one that has ensued from a remnant of the same radial growth component from which the original vertical growth component had its origin. In this scenario, the same phenomena which gave origin to the first melanoma (herein, melanoma is defined as a lesion in vertical growth), would have been recapitulated in the remnant of the radial growth component. If the recurrence is "primary" in configuration, and within the confines of the scar of the original surgical procedure, the possibility that the "recurrence" is a local, epidermotropic metastasis (satellite) must also be considered (10).
In a somewhat different scenario, surgical margins of excision of the primary lesion will have been found to be histologically free of involvement and a "recurrence," that lacks a primary configuration, will have been found to involve the deep dermis and the subcutaneous fat. For a lesion with these features, the most likely explanation for the recurrence is that it represents a metastasis to the deep dermis and subcutis. In this scenario, the primary lesion may be identified as not only having had the capacity for metastasis prior to the initial biopsy or excision, but also as having locally metastasized prior to the original excision. In such a setting, the "adequacy" of the margins of excision of the initial biopsy specimen likely will be questioned. In assessing the culpability of a pathologist, who may have imposed false, "benign" virtual images in his initial interpretation of what later proved to be a primary melanoma. the demonstration of a capacity for metastasis (as manifested in a deep, local recurrence) must be afforded great significance. In this scenario, any harm is properly attributed to an act of nature, as an expression of nature's way: the decisions and interpretations of the original pathologist are merely incidental. In practice, metastases to regional lymph nodes commonly appear shortly after, or even synchronously with, the appearance of a metastatic subcutaneous nodule in the site of the primary excision. In addition, distant metastases often follow, in short order, the appearance of the nodal metastases. Such events, occurring at close intervals, provide support for the interpretation that all of the metastases are more or less synchronous, and that they occurred prior to the initial excision of a primary lesion (a conclusion that is based on the premise that the original margins of excision were histologically free of involvement). The most reasonable conclusion would be that a patient, manifesting metastases in this manner and in the defined scenario, would not have been harmed by an error in diagnosis of his primary melanoma. A clinical "recurrence” would represent growth of tumor in what was an occult (micro-) metastasis at the time of the original excision. The error would be incidental to the observed phenomena (post hoc, ergo proper hoc) .
Distinctions should be made between local recurrences and local metastases. If the margins of an excisional biopsy of a melanoma are histologically free, then the subsequent appearance of tumor in the lower portion of the dermis and the subcutis qualifies as a local metastasis. If, on the other hand, the margins of the excised specimen were histologically involved by tumor and the biopsy site had not been re- excised with clear margins, then a distinction between local recurrence and metastasis cannot be established with certainty but the burden would be on those who would promote the problematic lesion as a metastasis.
Metastases: Metastases are generally characterized as discontinuous spread of tumor, but this characterization gives recognition to neither the fluid phase of malignancies (e.g., diffuse level IV invasion of melanomas) nor the connectedness of various organ systems by way of the vascular system.
The lymphatic component of the fluid phase of vertical growth gives rise to lymphatic metastases. The blood vascular component of the fluid phase of vertical growth gives rise to hematogenous metastases. The two components are independent (tumor does not selectively go first to regional lymph nodes and then disseminate to the viscera by way of the blood stream) but the resulting metastases, once manifested, may appear synchronously, or be temporally, isolated.
The designation, "in transit metastasis," is used in an inconsistent manner but gives recognition to cutaneous satellite lesions which are distributed between a primary site, and the respective regional lymph nodes. Originally, the concept was promoted as a justification for allowing an interval of time to pass between excision of a primary lesion, and a dissection of regional lymph nodes. As the rationale, the delay would give tumor cells, that might be free in the lymphatics, an opportunity to complete their journey from the primary site to the regional lymph nodes, the supposition being that not all lymphatics are open, and functioning, at all times. It was assumed that lymphatic vessels, i.e., those which are not functionally "open," could contain tumor cells. A short interval of observation presumably would provide the opportunity for the sequestered cells to find their way into a functioning tributary, and to then be delivered to the regional nodes. If the primary excision and the node dissection were to be performed concurrently, the "in transit" component would be more likely to become sequestered in the altered flow of lymph; to locally implant; and to then present in the pattern of local satellites.
The designation, satellitosis, gives recognition to cutaneous metastases in the vicinity of the site of a primary melanoma - an individual example is a "satellite." The metastases are visible on the surface of the skin and, thus, are superficial. Often, satellites are not associated with prominent lymphoid infiltrates. In "satellitosis,” regional cutaneous metastases are evident at the time of excision of the primary melanoma, or appear subsequently, often following excision of histologically positive regional nodes. In all examples of regional cutaneous metastases, if the lesions first become clinically apparent after a regional lymph node dissection, the most likely explanation is that the tumor cells were already implanted in the lymphatic vessels and neighboring dermis at the time of the excision of the primary melanoma; the implants were indolent in growth, and late in producing clinical manifestations (i.e., occult). In the above defined clinical setting, the above scenario is the most likely sequence regardless of the presence, or absence, of a recurrent nodule of tumor in the excision site of the primary melanoma.
If a distant, solitary cutaneous metastasis is followed by the appearance of multiple lesions in its immediate vicinity, then these might also be characterized as satellites.
A demonstration of a dermal and epidermal component (the latter in lentiginous, or lentiginous and junctional, patterns) is required to qualify a satellite as epidermo- tropic. Such a lesion is distinguished from a new primary by the confinement of the lentiginous and junctional component to the epidermis over the dermal component (no significant radial spread), and by a high cytologic grade in monomorphic population of cells in the dermis. Generally, such a lesion is small but often shows level IV invasion, and a population of neoplastic cells that are remarkably uniform. Some of these small, high grade lesion may also show neurotropic spread along small peripheral nerves in the underlying dermis. Lymphoid infiltrates often are inconspicuous.
In some patients, many epidermotropic satellites are purely lentiginous and junctional; they have the qualities of a primary configuration. Some examples of so- called epidermotropic metastases may be primary lesions; they may be representative of a field effect in which neoplasia is multi-focal and widespread in a regional distribution and, in addition, is cytologically high grade.
Micro-metastases: Microscopic foci of melanoma in the dermis in the vicinity of a primary melanoma, but in continuity with neither the epidermis nor the vertical growth component of the primary lesion, have been characterized as micro-metastases. They may be represented as small irregular nests of compactly aggregated cells, or as traceries of cells among collagen bundles of the reticular dermis. It is sometimes possible to identify tumor emboli in small dilated vessels beneath the vertical growth component in such lesions.
The designation, micro-metastasis, also has been used In a more specific manner to give recognition to discontinuous, reticular dermal foci of melanoma which measure more than 0.05 mm in diameter, and are in the immediate vicinity of, but not in contact with, a vertical growth component. In factorinq the significance of these isolated nests into the prognostic evaluations, they, regardless of size, should be characterized by a modification of clinical stage, rather than introduced into the prognostic evaluation by including the micro-metastases in the measurement by Breslow's criteria. Not all micro-metastases are at the deep margin of vertical growth components. Often, they are present in the upper portion of the dermis in linear arrays; they are often situated in patterns that would not lend themselves to manipulation by Breslow's criteria. They are comparable in significance to the finding of invasion of small vessels in the dermis beneath the vertical growth component, but have the added significance of being not only in the lumen of a vessel but of also of having extended beyond the confines of a vessel into the dermis, or adjacent soft tissue; such microscopic foci are incipient satellites. Micro-metastases might be cited as a finding which should influence recommendations as to the optimal width of margins of excision.
Agminate Spitz nevus: Agminate Spitz nevus is a well documented clinical entity, but the nature of the secondary lesions has not been well defined. The phenomena of multiple Spitz nevi in a limited locus may be observed in infancy and childhood. In the primary lesion, the patterns may deviate from those of typical Spitz nevus. Rounded nests of cells may be more common than fascicles of cells. In addition, the cells in the nests tend to have plump, round nuclei. In early examples, the agminate lesions may not be clinically apparent. They are manifested histologically by small clusters of nests of cells in patterns of local satellitosis at the interface of the papillary dermis and the reticular dermis. Agminate nevi of this type may represent local lvrnphatic metastases. This opinion does not imply that the cells of these lesions would have the capacity to colonize regional lymph nodes. The ability to establish an ecological niche in a lymph node may be of a different order from the ability to manifest local satellitosis in the skin. Small clusters of tumor cells within small, dilated lymphatics are occasionally a feature of the dermis adjacent to the border of a typical Spitz
Thin, minimal deviation melanoma metastasize, and in the size ranges of 1 mm or less, rare thin, small, melanomas metastasize (9): In the act of metastasis, metastasizing, thin melanomas will have demonstrated their nature as borderline or incipient melanomas, regardless of how the real and virtual images, and the assigned linguistic forms may have been manipulated prior to the appearance of metastases. Often, such offenders also manifest thin, level IV invasion, and may be relatively free of lymphoid infiltrates; often, they also are fascicular variants showing variant level IV invasion.
Areas of regression are common in thin metastasizing melanomas. A note of caution (i.e., a conservative prognostic evaluation) is required in the prognostic evaluation of thin "melanomas" showing focal areas of reqression. To identify a relationship between regression and metastasis is not the same as an identification of all thin "melanomas” with areas of regression as lesions that have metastasized. For all such lesions, the areas of regression should be viewed as a caution that real images and biologic behavior are not always congruous. Only virtual images can place the deceptive real images in proper perspective.
Several arbitrary segments of various melanocytic neoplastic continua have been characterized as "entities." Some of the “entities” have been sufficiently documented, and generally accepted, to be characterized as types, with SSM, LMM, and ALM as examples. In the common expressions of common rnelanocytic neoplastic continua, a precursor (dysplasia) has been identified as a two dimensional process at its deepest level. Its malignant counterpart (melanoma) has been identified as a three dimensional process at level III, or deeper. Nuclear grades and vertical dimensions have been emphasized; they have been offered as the basis for the delineation of a group of thin lesions in a prognostically favorable borderland, ernbracing both high grade dysplasias (patterns I & II) and thin melanomas (patterns III & IV). The management of an individual lesion in this borderland is little affected regardless of the manner in which portions of the borderland has been segmented, and then labeled. All of the common lesions (arbitrary segments with special designations) in this borderland would be managed by conservative excision and followup. Herein, it is proposed that this portion of the borderland might be characterized as borderline melanocytic neoplasia of indeterminate (for sorne of these lesions the qualifier, insignificant, might even be appropriate) malignant potential. This category could then be qualified as either a dysplasia variant or a melanoma variant depending on whether the lesions are basically two dimensional, or three dimensional.
An appreciation of nuclear grade has a role in the partitioning of this borderland of thin lesions. The dysplasia variants are lesions showing moderately severe to severe cytologic atypia (high grade dysplasias). The category would include not only those segments of the borderland that are commonly characterized as "melanoma" in situ, but also segments that by custom have been assigned to the category of "melanomas” (including SSM) at level II. In regard to the "rnelanoma" variants in this borderland, the category would lnclude lesions that are three dimensional (a definition of vertical growth), and measure 1 mm or less. It would embrace the borderline, thin MDM of Reed, and would include SSM, LMM, and ALM showing thin vertical growth (1 mm or less) and either level III or level IV patterns. The examples in this borderland in which widely, but regularly, spaced nests of cells of a thin, variant vertical growth component are individually isolated by concentrically arranged fibrous lamellae qualify as borderline melanocytic neoplasia of indeterminate malignant potential (arrested variant vertical growth pattern at level III).
With respect to Breslow's criteria, the upper limit of this borderline categorv might be defined as 1 mm. For patients with lesions confined to this range, the survival at 5 years would be greater than 90%. This is not the same as saying that an individual patient with a lesion in this range has an 90% chance of survival, if the lesion is properly diagnosed and excised locally, and the patient then carefully followed. These figures apply only to predictions when groups of patients are studied.
Variants of Melanoma (Lesions not of the Type)
(other nevi, other variants of melanoma, and other borderlands) (with emphasis on relationships with variant nevi)
(Note: for lesions in this section, size alone is not always the
chief criterion for the delineation of a borderline category)
In practice, Breslow's criteria and the related therapeutic guidelines have removed the onus from characterizations of a variety of melanocytic lesions as "melanomas;" in the limited range in which lesions are thin (i.e., 1 mm or less in vertical dimensions), a degree of relativity in the diagnosis of melanoma is unavoidable. Many surgical pathologists, even when confronted with variant histologic patterns, still react to the diagnosis of rnelanoma as if there is no provision for relativity, or for degrees of malignancy. For this group of observers, there is utility in neither the concept that size of a vertical growth component has relevance in estimating the age of the respective neoplasm nor in the proposition that age of a neoplasm and grade of neoplasia commonly are related. Variant histologic features in young lesions would not be appreciated by these observers as a possible marker for an early stage in neoplastic progression.
Many surgical pathologists react to the diagnosis of melanoma as if the segments of neoplastic continua, which they have individually selected and currently are utilizing, are inviolate. They consider modifications of their favored segments, or alterations of existing categories as a desecration of what for them have become sacred principles of neoplasia. Segments such as "melanoma" in situ, "melanoma" at level II, and hiqh grade dysplasias, though stoutly defended by proponents, have not been chiseled in stone and then brought down from the Mount; these segments are so insignificant biologically that they might be offered in support of the concept of relativity in the diagnosis of melanoma.
If we can accept relativity in the category of the common melanomas, then a slight modification of our virtual images might provide accommodation for variants of "melanoma," including a.........
(Continued on NEXT PAGE)