Note: To return to the first part, click on PART 1
WHITHER, Part 2
(RELATIVITIES AND DIMENSIONALITIES)
Richard J. Reed, M.D.
In part, the philosophy of pathology is concerned with the manipulation of relativities; variables (real and virtual), and their relationships, one to the other, are to be evaluated and sorted into primalities and extraneities. A structuring of sets of primalities is required before an entity can be defined. In these efforts, virtual images are imposed on real images. Virtual images, some new, some old, some borrowed, some firm, some flimsy, and some most fanciful, are all subject to the vagaries of perspective. In the subjections, individuality is certain to be introduced. A change in perspective may alter the availability of contextual and conceptional options and may require shifts and rearrangements of boundaries along the spectrum of relativities (Fig. 1 of the Desmoplastic Melanoma site, including Fig. 2); in the process, new primalities become requisites and old primalities may become inconsequential extraneities. The subjective nature of these relativities, as encountered along the axis of neoplastic progressions, becomes obvious when parcels of virtual images are shuffled about to accommodate changes in perspective. It may be most evident to old pathologists who have experienced the turmoil accompanying many changes of perspective. The primalities of today often are the extaneities of tomorrow.
The utility of words is based on intensions; intensions are sets of assigned attributes; attributes are arbitrary selections from sets of real and virtual images. Variability in the attributes of intensions is a reflection of the vitality of a language. To impose restrictions on the availability of variables would make language simpler, inflexible, and less compelling; communication would suffer; there would be no Shakepeares. Such impositions would empower the imposer as a censor; prejudice would be the rule. Under such conditions, the identity of the attributes of melanoma in situ and marked melanocytic dysplasia might be lost in the pontifications of an empowered censor.
The designation, melanoma, is laden with intensions, including both native and man- made attributes, and emotional elicitations. The native attributes include locally progressive disease, satellitosis, nodal metastases, disseminated disease, and death. At a histologic level, the assigned native attributes might include some or all of the following: degrees of premalignant melanocytic dysplasia, melanoma in situ, minimal deviation (nevoid) melanomas, and common variants of melanoma. An empowered censor could negate any, or all, of these. Although the native attributes are derivatives of observations, the very act of imbuing a word with such attributes introduces man-made prejudices that are individual, arbitrary and ambiguous; thus all words and even numbers are compromised.
For melanocytic neoplasia, the assigned, man-made attributes, as embodied in intensions, have to do with the relativities among physical parameters and biologic behavior. The emotional elicitations include anxiety, insecurity, and uncertainty. The threats, issuing from all the attributes, evoke the emotional cohorts.
Assigned, man-made attributes are relativities; the likelihood of the threats, imbued in such attributes, becoming realities is relative to size and age of neoplasm. In regard to size, it is not the spread of a lesion on the surface of the skin that relates to prognosis; it is the dimension of a particular component, namely vertical growth. If threats inherent in all these attributes are evaluated, it may be proper to speak of melanocytic neoplasia as melanoma only for those examples in which the prognostic parameters would indicate that there is a sufficient likelihood for progressive disease to also justify special recommendations for treatment beyond conservative local excision. By current criteria, and in a diagram of an axially oriented neoplastic system (Figs. 1 & 2), this boundary is generally characterized as the one beyond which all the representative lesions measure 1.5 mm or greater along the vertical axis of the vertical growth component ("real melanoma"). Our designations assign risk. For thin lesions the risk of metastasis is so unpredictable that a conservative approach is warranted and is the standard. If we can accommodate the distinctive relativities of thin lesions with a conservative approach, might it also be appropriate to accommodate those same relativities with a distinctive designation, something other than "melanoma"? The domain of a common, minimal risk category is diagrammed in Figs. 1 & 2.
In our diagrams (Figs. 1 & 2), both the cone and the expanded portion of the "plane" (the white wedge) which intersects with the cone in Fig. 2 define the domain of transitional melanocytic neoplasia. Even the grade 3 lesions of the plane might be included in the transitional component. This structuring of the "plane" at its extremity as a wedge gives recognition to the ambiguous patterns in the dermis in advanced degrees of melanocytic dysplasia. The papillary dermis in lesions at this segment of the axis of neoplasia is characteristically widened and may contain stratified nests of atypical cells. The strata may be sufficiently developed to introduce concern regarding the emergence of early vertical growth. The white wedge is a transition zone in which ambiguous phenomena are expressed in ambiguous patterns. The cone is representative of a progression in neoplasia, but is also a transition zone in which the patterns qualify as some variant of vertical growth, but not all the vertical growth patterns can be equated with "taking" root. The cone in Fig. 2 consists of two segments, the first on the left being the domain of borderline neoplasia, and the second on the right (a darker gray), abutting on the black sphere (typical vertical growth), defines the domain of intermediate neoplasia. The sphere of Fig. 2 is representative of lesions in typical vertical growth; it is symbolic of the observation that most lesions which have attained a vertical dimension of 1.5 mm in vertical dimension will also be in typical vertical growth, if they have not evolved into lesions in pattern IV (beyond the red boundary in Fig.2). In Fig. 2, all the domains in white, grays, and black are representative of the altered native stroma of the skin in progressive melanocytic neoplasia, including that of the epidermis and the adventitial dermis (i.e., the papillary dermis and the perifollicular connective sheaths); the mucinous avenues of the epidermis and the adventitial dermis are stroma. The red boundary of the diagram defines the interface between the altered stroma and the reticular dermis. Cells beyond the red boundary have attained at least level IV (and are pictured as black dots in Fig. 2). The number of black dots along the axis of neoplasia is relative to a particular stage of neoplastic progression. The higher the stage of neoplasia, the greater the number of dots. A neoplasm in typical vertical growth is more likely to be associated with a component at level IV than is a lesion early on along the axis of neoplasia. On the other hand, neoplastic cells may be found at level IV even in early stages of neoplasia but need not be found in patterns of vertical growth. In this approach, level (pattern) of growth is independent of imbued attributes of the original concept of vertical growth. Boundaries are pertinent to the evaluation of patterns of vertical growth but to identify vertical growth requires something more than the simple demonstration of the violation of a boundary by neoplastic cells.
THE FUNCTIONAL SUPERFICIAL UNIT OF THE SKIN
Epidermis and papillary dermis form a functioning unit (Fig. 2). In Figs.1 & 2, the two are combined in the plane. A change in the epidermis is generally accommodated by compensatory changes in the papillary dermis, and vice versa. Conceptually, the separateness of these two components is a morphologic compromise. Although properly handled as separates, there is utility in combining the two as a single functioning (superficial reactive) unit, particularly if a third cell line has intruded and, by its presence, has altered normal relationships; neoplastic melanocytes of common dysplasias constitute just such an intrusion.
DIAGRAMMATIC REPRESENTATION OF A
MODEL NEOPLASTIC SYSTEM
The epidermis and the papillary dermis can be manipulated as if the two form a morphologic unit. In the process, the domain of the epidermis and its stroma can be treated conceptually as a plane (Fig.1 & 2). If the dimensions of the plane are distorted by contributions from a third population of cells, tools are required to manipulate the new patterns of the plane, and to separate them from those of the cone. A useful approach would be to "dimensionalize" the histologic patterns of both the plane and the cone (Fig. 1). The white wedge (domain of marked melanocytic dysplasia) which blends with the cone at the "1 mm" boundary gives recognition to a dimensionalization, being in turn representative of an expansion of the papillary dermis in response to newly delivered nests of neoplastic cells from the epidermis into the dermis. It gives recognition to the difficulties in making distinctions between "thick" dysplasia patterns, and early "thin" vertical growth patterns. For melanocytic neoplasia, the responses of a "dimensionalized" neoplastic system that need to be evaluated include: 1) induced hyperplasia of the epidermis (primal soil), and 2) modification of the dimensions of the dermal stroma (an adopted soil) by the proliferation and movement of populations of foreign cells, or by expansion (hyperplasia) of the stroma (papillary dermis). To address melanocytic neoplasia as a problem in dimensionalities requires the introduction of new intensions (attributes embodied in words), new labels, and, in turn, new perspectives. Qualitative changes in the stroma (changes other than those of physical dimensions) may also require new tools with which virtual images can be structured and then manipulated.
EPIDERMIS AND ITS RESPONSES
(ANATOMY, FUNCTION, AND THE REACTIONS OF NORMAL EPIDERMIS)
The epidermis and the stroma of melanocytic neoplasia have seldom been given detailed separate attention. The normal epidermis is composed of a basal unit and a superficial unit (Fig. 3). The basal unit is normally composed of only 2 or 3 cell layers. The individual cells have scanty basophilic cytoplasm; they are oriented in a direction that is perpendicular to the basal membrane (Fig. 3). The cells of the basal unit not only replicate, they also maintain a mucinous interstitium. They, in their divisions, compensate for physiologic apoptosis in the basal unit. They also replenish cells of the superficial unit as they are lost along the surface in the form of keratinized debris (a process of so-called terminal differentiation). The superficial unit is composed of polygonal cells with acidophilic cytoplasm (Fig. 3). The cells of the superficial unit have their long axes parallel to the surface of the skin. They, by inlaying the interstitial spaces with lipid rich deposits, and by persisting at the surface as condensed lamellae of keratinized debris, form an impervious barrier (Fig. 4). In their upward migrations (i.e., migrations that are passive and an inevitable consequence of terminal differentiation), keratinocytes desiccate and die.
EPIDERMIS AND ITS INTERACTIONS WITH NEOPLASTIC MELANOCYTES
Many of the virtual images which have utility in the interpretation of the reactions of the skin in inflammatory disorders have, with only limited modifications, application in the interpretation of the melanocytic dysplasias.
RESPONSE OF EPIDERMIS TO STRESS
Under conditions of physiologic, or pathologic, stress, the cells of the basal unit undergo hyperplasia (i.e., an increase in number). If the stimulus is sufficient, the ensuing histologic patterns are likely to be psoriasiform; rete patterns are accentuated (Fig.3). Psoriasiform patterns are prototypic of basal unit hyperplasia (Fig. 3). In moderate to moderately severe melanocytic dysplasias, the response of the epidermis is usually basal unit hyperplasia. Upper migration of neoplastic cells into the epidermis is a feature of moderately severe, and marked, dysplasias. Hyperplasia of the superficial unit is a feature of an established lichenoid reaction (Fig. 5). In melanocytic neoplasia, hyperplasia of the superficial unit might be characterized as a reaction to a foreign population of neoplastic cells. In marked melanocytic dysplasias (i.e., the common final pathway as defined in the concept of minimal deviation melanoma), the superficial unit is hyperplastic; its cells individually are hypertrophied; and rete ridges tend to be effaced (i.e., they become incorporated in the hyperplastic superficial unit). Marked dysplasias with upper migration of neoplastic cells into the epidermis are usually associated with a distinctive form of epidermal hyperplasia. With our model of the lichenoid reaction as a guide, all of the keratinocytes of the affected area of a marked dysplasia, with the exception of the single layer of basal keratinocytes, show features of having committed to terminal differentiation. If this epidermal pattern is associated with scattered, randomly spaced nests of neoplastic melanocytes in the papillary dermis, the respective lesion is, in common parlance, a "superficial spreading melanoma at level II." An hypertrophy of individual cells in the superficial unit is associated with an increased prominence of cytoplasmic expanse (Fig. 3); increased acidophilia of the epidermis is the result. Hyperplasia of the superficial unit of the epidermis, somewhat similar to that seen in lichen planus, is prototypic of patterns encountered in "superficial spreading malignant melanoma in radial growth." It is associated with hypergranulosis and compact orthokeratosis. Exaggerated psoriasiform patterns are common in the radial growth of acral lentiginous melanoma but hyperplasia of the superficial unit is also a feature. In moderately severe to marked dysplasias of acral type, upward migration of neoplastic cells is a prominent feature; the migrants are dendritic; they lie within irregular lacunae. In contrast of all the above, the epidermis of actinic lentiginous dysplasias (lentigo maligna as defined in the concept of minimal deviation melanoma) often is remarkably refractory to the presence of a population of neoplastic melanocytes, even to those in migration into the epidermis. The epidermis is often thin with inconspicuous rete patterns, particularly in the categories of mild to moderately severe actinic melanocytic dysplasias. The kinetics are so distinctive that it is tempting to propose that the kinins which promote the epidermal responses in both SSM and ALM are not equally operative in the actinic lentiginous variants of melanocytic dysplasia. The differences in patterns as related to site may be an indication of the influence of the soil on the nature of the “plant.” Lichenoid reactions have a life history and the late stage qualifies as senescent lichenoid reaction (Fig. 3, - 4,- 5). Lichenoid reactions are common in the setting of the melanocytic dysplasias; they may be encountered in the radial growth components of melanomas. The effects of such a host immune response may lead to focal, or total, regression of the respective neoplasm. The patterns then become those of a senescent lichenoid reaction with spotty lymphoid infiltrates and collections of melanophages in a widened, fibrotic papillary dermis.
In toto, the changes which characterize epidermal hyperplasia of either the basal or superficial unit qualify as a form of "immunostimulation." Varying degrees of cytoplasmic pallor may be seen in immunostimulative disorders affecting the epidermis; they may even be associated with invasive growth of the affected squamous epithelium; nests of pale squamous cells may extend into the dermis in infiltrative patterns. The prototype for the invasive forms is keratoacanthoma. Not all examples of immunostimulation of keratinocytes are represented in patterns as extreme as those of keratoacanthoma, nor are all invasive.
If immunostimulation of terminally differentiating keratinocytes is encountered locally in the setting of a young melanocytic neoplasm, it is likely to be associated with a marked degree of cytologic atypia and with upper migration of neoplastic cells, both individually and in nests. In this association, the growth of cells of one type apparently affects the characteristics of a second population of cells. Such changes are basic to the precursor stages of an evolving superficial spreading melanoma. In the concept of minimal deviation melanoma, this mutual compensation, involving atypical cells and a distinctive hyperplasia of the superficial unit of the epidermis, was characterized as the common final pathway. In all these interactions of an immunostimulative nature, the mucinous avenues of the epidermis are altered. Some of the alterations may favor the upper migration of cells. On the other hand, the epidermal changes in the superficial unit may be a defense mechanism as an attempt to limit the intraepithelial migrations of neoplastic cells; it may be an attempt to close the epidermal interstitial spaces.
If common inflammatory diseases are evaluated, immunostimulation is closely associated with patterns of immunolysis. Lysis affecting the basal unit of the epidermis is a feature commonly encountered in "lichenoid" processes (Figs. 3 & 5). Hyperplasia of the superficial unit of the epidermis is a common accompaniment of lytic phenomena in the basal unit of the epidermis. As a model, lichenoid disorders are immunolytic (by nature and definition). They commonly are associated with a distinctive hyperplasia of the superficial unit of the epidermis (i.e., immunolysis of the basal unit and immunostimulation of the superficial unit) (Fig. 3). Lichenoid patterns are a common feature of the reaction at the dermal-epidermal interface in common melanocytic dysplasias. If extensive and well-developed, they tend to mask the primary melanocytic dysplasia. The reaction in halo nevus-like phenomena can be equated with a lichenoid reaction with the stipulations that the target cell is a melanocytic cell, and that both the epidermal and the dermal components are altered by the effects of the cell mediated reaction. In halo nevi, the initial change may be a genetic alteration in the nature of some, or all, the melanocytic cells; the lymphoid infiltrates probably are a response. From this perspective, some of the lesions which are easily dismissed as common melanocytic dysplasias may represent halo nevi ab lymphoid response.
THE SOIL OF MELANOCYTES
The soils of the common melanocytic neoplasias are both the epidermis and the dermis, both of which interact with neoplastic cells. In the epidermis, the degree of epidermal response is congruent with the degree of dysplasia (Fig. 6). In general, changes in the epidermis are an accompaniment of the progress of the neoplasm. Little alteration of the epidermis is seen in mild dysplasias. In moderate dysplasias, there is usually some degree of acanthosis and a slight accentuation of the rete patterns. In moderately severe dysplasias, the rete usually are accentuated and elongated. In the severe dysplasias, the superficial unit of the epidermis is hyperplastic in a distinctive pattern. The hyperplasia tends to incorporate the rete ridges.
EPIDERMAL STROMA (SOIL)
The interstitial avenues of the epidermis are a type of stroma (Fig. 7). Under physiologic conditions, the stroma of the epidermis loses its watery attributes in the superficial unit near the granular layer (sites in which keratinosomes have been deposited in the epidermal interstitium). By habit, the stroma of the epidermis has been ignored. It is a scanty, watery, or mucinous, matrix with the capability of responding to epidermal stresses. It may expand and become more prominently mucinous or watery. In psoriasiform hyperplasias, it expands and often is mucinous (Fig. 6). In spongiotic disorders, it is expanded and watery.
In the lentiginous actinic dysplasias, progressions along the axis of neoplasia, and progressions as represented in degrees of dysplasia (atypia) are not as closely associated with predictable patterns of epidermal hyperplasia. In uncommon variants of lentigo maligna, the degree of atypia and the epidermal changes come to resemble those of a lesion of SSM; they acquire the characteristics of the common final pathway. Such lesions are likely to be dismissed as SSM regardless of clinical setting; they are dismissed as such even in the face of other markers for patterns that are more characteristic of classic LMM.
MIGRATIONS OF NEOPLASTIC CELLS (RADIAL GROWTH)
AND THE SEEDING OF THE PAPILLARY DERMIS
If competence for metastasis is a primal attribute of "melanoma", the active migration of neoplastic cells among keratinocytes of the stratum malpighii certainly is not a defining feature. Melanocytic neoplasms in which cells are confined to the epidermal domain do not metastasize. A few scattered, widely, and irregularly spaced nests of atypical cells in the papillary dermis might be characterized as a "seeding of the soil (stroma)". If the migrants are to adapt to the new environment of the stroma, something more than the mere physicality of the movement of cells from one domain to another is required. Growth, independent of phenomena at the dermal- epidermal interface, is a vital process; seeds must take root. Vertical growth is the marker for the emergence of the vital process; it is evidence that the tumor has taken root.
MIGRATIONS OF MELANOCYTES;
TYPES OF SOIL;
SOWING OF SEEDS ("DROPPING OFF")
In the evolution of the premalignant melanocytic dysplasias, lentiginous patterns are a marker for the movement of individual neoplastic cells in the epidermal interstitium among keratinocytes (a seeding of the soil of the epidermis). The clustering of cells in nests at the dermal-epidermal interface is in part an economy; if there are too many cells per unit area of epidermal domain, an expansion of the intercellular matrix allows the excess to cluster in a newly formed matrix, and to do so to the exclusion of keratinocytes. In addition, the close clustering in the nests may afford the component cells the opportunity to parcel economies. Depending on their location in a nest and proximity to neighboring structures (i.e., vessels, fibrous stroma, etc.), packets of cells in each nest may be dedicated to different functions. At the interface between a nest of cells and the overlying epidermis, the neoplastic cells may acquire desmoplastic properties. By inlaying new fibrous tissue at the interface, the nest of neoplastic cells may appear to have "dropped off" into the dermis. The newly formed matrix tends to be dense, and acidophilic with random fiber patterns. In this proposition, the "dropping off" of nests of cells acquires the properties of a sequestration.
"DROPPING OFF" AND THE MICROENVIRONMENT OF NESTS OF NEOPLASTIC CELLS
Close clustering of neoplastic melanocytes to form nests in the epidermis may favor a migration of the clusters out of the epidermis into the papillary dermis. The microenvironment in the interior of each nest may evoke a sequence leading to a sequestration of the nest from the epidermis. On the other hand in the "dropping" of nests into the dermis, perhaps the epidermis is the prime mover in the transfer of a nest of foreigners out of the epidermal domain. Perhaps, circumferential fibrosis is instrumental in the sequestration of the nests and the "dropping of nests" into the dermis. In the melanocytic dysplasias, the process may involve the formation of a cleft at the interface between the nests and the epidermis. Eventually, the cleft may be inlaid with fibrous tissue. In this approach, the process remains one of sequestration. In either case, nests of cells, by “dropping” into the dermis, seed the soil.
MELANOCYTIC NEOPLASIA AND THE SOIL OF THE DERMIS
The fibrous stroma of the papillary dermis is the second component of the soil of the melanocytic neoplasias. In the papillary dermis, the neoplastic cells of a dysplasia generally form nests but, early on, the nests mostly are widely and randomly spaced; with only occasional exceptions, they do not contribute bulk in 3 dimensions. The soil of the papillary dermis can accommodate an increase in the number of nests of neoplastic cells; for as long as the neoplasm continues as a dysplasia, neoplastic cells of the dysplasia can survive in it, and still not take root.
MEASURES OF PROGRESSIONS
The progress of a dysplasia can be evaluated by: 1) the density of cells in the lentiginous portions, 2) the density of junctional nests and the upward migration of the nests, 3) the degree of host immune response and the intimacy of that response among neoplastic melanocytes, and 4) the degree of cytologic atypia. Three of these measures are best evaluated in the epidermal component. The lymphoid response is best measured by an evaluation of the density of the dermal infiltrates, and the intimacy in which lymphoid cells and neoplastic cells are intermingled in the dermal component. In some melanocytic dysplasias, lymphoid cells infiltrate the epidermis and partially obscure the epidermal component of neoplastic cells; such patterns are lichenoid; they are associated with lytic defects in the epidermal domain.
THE RESPONSE OF THE DERMAL SOIL TO THE SOWING OF SEEDS
In the dysplasias, the soil (stroma), in response to newly deposited seeds and in anticipation of the seeds taking root, is modified, particularly in the mild to moderate categories. The most characteristic alteration is a laminated pattern of fibrosis. Often the junctional nests of a dysplasia are situated at the extremities of rete ridge. The lamellar fibrosis conforms to the contours of the portion of the nests that bulges out of the epidermal domain and presses on the dermal domain. The greater the amount of the nest that bulges, the closer the lamellar fibrosis comes to be circumferential. The fibrous lamellae often are spaced in a clear matrix and flattened spindle cells (the "sheath cells" of Reed [Fig.8]) may be found among the lamellae. The fibrosis is appositional with one lamella superimposed upon its neighbor. Presumably each new lamella is interposed between an older lamella and the basement membrane of the epidermis (and the junctional nests); the process is accretive at the dermal-epidermal interface.
Eventually, by "dropping off," sequestration, or interstitial fibrosis in the domain of the epidermis (a process which, if established as valid, might anticipate the growth of occasional melanomas in desmoplastic patterns), some of the junctional nests come to lie in the altered soil of the papillary dermis; they lose their attachment to the epidermis. The mechanism of this detachment remains speculative. With increasing numbers of detached nests, the individual, irregularly spaced nests in the dermis may be prominently outlined by concentric fibrous lamellae but, in addition, the neighboring papillary dermis becomes more densely fibrous and acidophilic. This altered fibrous tissue and the entrapped nests form a condensed plaque abutting superficially on the basement membrane zone of the epidermis. In some examples, this diffuse fibrosis encloses the individual nests of cells in the dermis and there may not be fibrous lamellae at the interface with individual dermal nests. If a component of preexisting nevus cells is represented in such a lesion, it generally is beyond the zone of fibrosis and is to be found at the interface between the papillary dermis and the reticular dermis. In addition, if there are perivascular infiltrates of lymphoid cells, they generally are confined outside the zone of fibrosis, particularly in the mild to moderate dysplasias.
THE FUNCTIONALITY OF THE ALTERED (i.e., FIBROTIC), SEEDED SOIL
The altered, fibrotic soil (stroma) of the papillary dermis is a barrier limiting the enlargement, and the number of the dermal nests of neoplastic melanocytic cells. It also limits the contact between native inflammatory cells and the foreign neoplastic cells. Its influence on the dermal component is two-pronged; it limits growth of the neoplastic cells (expansion of the entrapped nest is limited) and also denies the cellular component of the host immune response direct access to the interstitium of the neoplastic cells. It becomes less significant with progression in the degree of dysplasia. In higher grades of dysplasia, the cells of the nests may focally stimulate the cells of the stroma and induce collagenolytic changes; circumferentially around some of the nests, particularly the nests that have most recently descended into the stroma, the matrix becomes loose and clear. Infiltrates of lymphocytes can then intermingle among the nests of neoplastic cells. The likelihood for regression will be greater in high grade than low grade lesions; the neoplastic cells are less sequestered in a condensed stroma.
CORRESPONDENCE OF THE STATE OF THE SOIL AND THE DEGREE OF
"HYBRIDNESS" (i.e., MELANOCYTIC DYSPLASIA)
The patterns, as elaborated in the preceding paragraph, have application to the mild to moderate, common melanocytic dysplasias. In the higher grades of dysplasias, not only is the soil of the epidermis modified but the character of the papillary dermis focally may be altered. With progressions in older dysplasia, markers for succeeding generations of neoplastic cells become evident in the papillary dermis (manifested in stratified nests of cells and variations in degrees of atypia in the nests with the least atypical population of cells at the deepest level in the soil). The newest and most atypical generation in the papillary dermis will be superficial. In these strata, the greatest degree of atypia is near the dermal-epidermal interface; in the latter stratum, the nests of atypical cells in high grade dysplasias are likely to be associated with a significant change in the character of the fibrous matrix. The stroma is likely to be delicately fibrous with delicate, individual fibers spaced widely in an optically clear, or mucinous matrix.
THE CORRESPONDENCE OF BOTH DENSITY OF LYMPHOID INFILTRATES
AND MODIFIED LICHENOID REACTIONS (TUMOR INFILTRATING LYMPHOCYTES),
AND DEGREE OF DYSPLASIA
Lymphoid infiltrates tend to be more intense in proportion to the degree of dysplasia (cytologic atypia) (with the exception of halo nevus-like phenomena). Not only are the infiltrates progressively prominent in higher stages of dysplasias, but they also come to be intimately associated with the dermal nests of cells; they may invade both the nests and the overlying epidermis to intermingle among the neoplastic cells. The end-result is piece-meal necrosis and lysis of not only individual cells but even nests of cells. The patterns at the dermal-epidermal interface may come to resemble those of a lichenoid reaction. In these modified lichenoid reactions, the target cell is primarily neoplastic melanocytes but keratinocytes are also afflicted. Skip areas, devoid of neoplastic cells in the epidermal and dermal domains, may be the end-result (focal regression). In these interactions, the degree of dysplasia is evaluated not only by the cytologic features and the number (and close spacing) of nests (sure primalities), but also by stromal response, and density, and distribution of lymphoid infiltrates.
DYSPLASIA AND PATTERNS OF DIFFERENTIATION
Generally in common premalignant melanocytic dysplasias, the nests of neoplastic cells which are delivered to the dermis are rounded in outline. If strata are represented in the dermal component, the nesting patterns are identical at all levels, although the degree of cytologic atypia may be variable at various levels. The variable patterns of a common nevus are rarely manifested in the evolution of the common dysplasias; type A,B, and C patterns are not a feature of most dysplastic components. The patterns of differentiation as seen in a common nevus are not manifested in the evolution of a common dysplasia. On the other hand, a variant of nevoid melanoma evolves in atypical patterns which resemble those of a common nevus.
THE SEEDS AND THEIR SOWING - THE SOIL AND TAKING ROOT
The process up to this point is, in its nature, a planting. If the seeds are allowed to sprout, the progress of the growth is evidenced in a local change; all subsequent progressive stages then become three dimensional.
In the diagrammatic structuring of a neoplastic system (Figs. 1 & 2), neoplastic progressions may be defined as segments or levels along an axis. In such an approach, time is measured along the axis, and it is possible to give recognition not only to cytologic progression but also to changes in bulk, as well as relationships between tumor and stroma. Vertical growth is a more developed segment (i.e., farther along the pathway of neoplastic progressions) than radial growth. Bulk at some stage then becomes a primality and might be cited as the distinguishing feature of a neoplasm in which the neoplastic cells have made an accommodation for residence in a new domain (i.e., a new stroma) and have acquired some competence for metastasis. The age of a vertical growth component is evaluated by its bulk without attention to any associated radial component.
THE SELECTION OF GUIDES WHICH ALLOW FOR THE SEGMENTATION
OF NEOPLASTIC PROGRESSIONS, THE IMPACT OF THOSE SELECTIONS ON
TERMINOLOGY, AND UNEXPLORED ASPECTS OF VERTICAL GROWTH
In melanocytic neoplasia, initial cellular migrations and the emergence of new histologic patterns are correspondent. A transition from radial to vertical growth cannot be defined finitely by an evaluation of degrees of cytologic atypia. The distinctions between advanced patterns in old radial growth, and thin, new vertical growth, components are always ill-defined.
Several different options have been utilized to diagrammatically segment progressions along the axis of a neoplastic system. With the concept of melanoma in situ as a guide, there are no true segments along the axis of neoplasia; vertical growth must then be characterized as nothing more than progression from the basic pattern of melanoma in situ by expansion in bulk; there are no intermediate stages; the cells of melanoma in situ differ from those of a vertical growth component only in number. The presence of nests of neoplastic cells in the papillary dermis is not, in itself, the defining feature of melanoma.
In the concept of minimal deviation melanoma, a major division between dysplasia and melanoma is defined and the boundary corresponds to the interface between radial and vertical growth. There are no melanomas at levels (or in patterns) "I" or II. From this perspective, a degree of nuclear atypia is secondary in importance to patterns in which nests of cells are arranged in the dermis, particularly the papillary dermis. Some patterns in thin, problematic lesions, which might be characterized as vertical growth, are associated with rather bland cytologic features; such lesions have minimal deviation qualities in regard to both dimensions and degree of cytologic atypia.
STRATIFIED PATTERNS IN VARIANT VERTICAL GROWTH
Some of the processes which eventuate in the progressive delivery of nests of cells into, and in the expansion of, the dermis may not equate with taking root. Stratified patterns are related to the manner in which cells and nests of cells are delivered from the epidermal interstitium into the interstitium of the dermis. Accretive grow of the papillary dermis is an accompaniment of the "dropping off" of nests of cells into the dermis; the result is a random and spotty delivery of nests of neoplastic cells into an altered papillary dermis. The alterations in the papillary dermis include diffuse acidophilic fibrosis and a variable component of concentric lamellar fibrosis. The nests are entrapped in the altered papillary dermis and, if there is to be an increase in the number of nests in the dermis, new nests must "drop" from the epidermis into the dermis as a repeat of the initial process - they do so in sequences. With progressive and sequential delivery of nests of cells into the dermis, patterns may develop in which the distribution of the nests appears to be fairly regular. In this process, each nest often is outlined by concentric fibrous lamellae, and all the loosely clustered nests are entrapped in the distinctive eosinophilic fibrous matrix. The process is one of accretive growth with new generations laid upon older generations, and each generation entrapped in the progressively expanded fibrous matrix. In some examples, the accretive growth is remarkably stratified in a vertical orientation; as a result, the stratified nests form spaced columns; in the columns, the stratified nests are arranged as to degree of cytologic atypia; the new generations are the most atypical and the most superficial. The oldest generations are the deepest and the least atypical. This pattern, in toto, qualifies as arrested variant vertical growth. If it then is to be cited as the basis for a histologic diagnosis of minimal deviation melanoma, the process must be additionally qualified as borderline. The patterns do not have the import of those of typical vertical growth.
RELEVANCE OF TYPICAL AND VARIANT PATTERNS OF VERTICAL GROWTH
The category of vertical growth embraces variable patterns. The most significant of these patterns is "typical" vertical growth in which nests of cells are closely clustered in the dermis and have induced a delicate, vascularized stroma (if stroma is sufficient in its representation to be evaluated as to its character). In Figs. 1 & 2, the variable patterns in vertical growth are not addressed. On the other hand, most of the variant patterns of vertical growth are encountered in thin, new lesions. In the patterns of vertical growth, the character and quantity of stroma is distinctive; the problems related to patterns of vertical growth are most often encountered in lesions measuring less than 1 mm in vertical dimension - the light gray portion of the cone in Fig. 2). Regardless of vertical dimensions, the lesions showing typical vertical growth (the black sphere of Fig. 2) have the greatest correspondence with a competence for metastasis.
DISCRIMINATION IN PROGNOSTICATIONS:
Breslow’s criteria have become the mainstay in the formulation of guidelines upon which the treatment of melanoma can be planned, but really become discriminative only for lesions greater than 1.5 mm in vertical dimensions; they have no relevance to degrees of atypia; they are based on physical dimensionalities and can be manipulated as virtual images to only a limited degree; the manipulations are numerations. Despite the universal acceptance of Breslow’s criteria, and the relevance of the vertical dimensions of certain vertical growth components (i.e., those measuring greater than 1.5 mm in dimensions) to prognosis, there are characteristics of vertical growth that are little understood and hardly have been explored.
THE RELATIVITY OF BOTH DIMENSIONALITIES
AND PATTERNS IN THE DEFINITION OF VERTICAL GROWTH
AND THE IMPORTANCE OF THESE TWO ATTRIBUTES
CHARACTERIZING VERTICAL GROWTH AS A PRIMALITY
Both physical dimensionalities and patterns are relativities to be evaluated in the definition of vertical growth. These two attributes have importance in characterizing vertical growth as a primality: A vertical growth component becomes a primality only in certain ranges of sizes in which size clearly comes to have relevance to a predictable competence for metastasis; in all other settings it is only a designation for a variety of patterns (dimensionalities expressed as patterns) rather than mere measurements. The virtual images associated with these dimensionalities provide insights into dynamic processes involved in structuring the progressive steps along the axis of neoplasia.
RELATIVITY - SIZE AND PROGNOSIS
To characterize a melanocytic neoplasm as "melanoma" in the absence of a convincing vertical growth component is presumptuous, if a demonstrated competence for metastasis is also a basic requisite. On the other hand, relativity between size and competence for metastasis requires caution in the labeling of thin lesions as melanomas, even in the presence of a typical vertical growth component. In this approach in the prognostications that have relevance for thin lesions, a measurement of a physical dimension of a vertical growth component becomes nothing more than an approximate prognostic equivalent. It is to be correlated with patterns and then qualified to indicate that the size of thin, new lesions (i.e., less than 1.5 mm in vertical dimensions) generally has little relevance as an indicator of competence for metastasis; both size and histologic patterns modulate the significance of Breslow’s criteria. In the range of sizes from <0.31 to 1.50 mm, too few melanocytic neoplasms metastasize to assume that size is of sufficient prognostic relevance to be cited as a primality, unqualifiedly legitimizing the designation of melanoma. To label a melanocytic lesion as melanoma, if the vertical growth component is less than 1 mm in vertical dimensions, places an onus on a large population which is at little risk for progressive disease; under duress, these patients easily can be enticed, and then may readily commit to adjuvant therapy. Lesions confined by a boundary at 1 mm. (Fig. 1) do not metastasize with a sufficiently predictable frequency for a pathologist to have confidence in the proposition that such lesions are properly labeled as melanomas. Such lesions are borderline neoplasms of uncertain malignant potential.
THE DEFINITIVE PRIMALITY (i.e., VERTICAL GROWTH) MADE ARBITRARY
(AMBIGUITY OF PATTERNS AS RELATED TO THE LIMITED VERTICAL
DIMENSIONS OF YOUNG MELANOCYTIC NEOPLASMS)
For a melanocytic neoplasm, evidence of actual growth in the new environment of the papillary dermis - rather a mere dropping of nests from the epidermis into the dermis - is to be found in areas in which nests of cells are closely clustered in distinct patterns (i.e., vertical growth) and are not simply distributed as randomly and loosely, spaced nests of neoplastic cells. The most reliable pattern that can be cited as demonstrative of actual growth in the dermis, is a clustering of new, closely spaced nests about the original dermal nest (nidus); the requisite patterns, as embodied in virtual images, is a small expansile nodule. Such a pattern is evidence that a melanocytic neoplasm has taken root. In the process of taking root, the lesion acquires the ability to activate stroma and promote a tumor-responsive plexus of vessels. In theory, these two properties are uncomplicated ideals (primalities). In practice, the stromal responses of evolving melanocytic neoplasias are complex, and some patterns, which might appear as variations of vertical growth, may be indicative of transitional phenomena, of little more significance than that of closely spaced, dormant seeds in unturned soil.
PATTERNS OF INVASION
In addition to depth of invasion as evaluated by Breslow’s criteria, there is utility in modified Clark’s levels of invasion. In the concept of minimal deviation melanoma, patterns of vertical growth were correlated with levels of invasion to define melanoma. In the concept of MDM, it has been proposed that, of all the emerging patterns encountered during neoplastic progressions, melanoma most certainly becomes an appropriate designation if the pattern is that of typical vertical growth, and if the lesion has advanced to at least level III (i.e., pattern III growth as modified in the concept of MDM). These two stipulations clearly indicate the interdependence of two sets of virtual images. Recently, the concept of MDM has been modified and patterns of invasion have been substituted for levels of invasion; the differences in patterns at various levels have biologic significance but patterns are not as readily equated with a diagnosis of melanoma as are levels of invasion. We can speak of patterns without introducing biologic import. Also, as a recent modification of the concept of MDM, it has been proposed that there is relativity in the diagnosis of melanoma, particularly in regard to thin, new lesions. In the thin categories, two domains are defined, one with a boundary at 1 mm and the other with a boundary at 1.5 mm (Figs 1 & 2). In these borderline and intermediate domains, it may be inappropriate to simply assign a lesion in "vertical growth" to the category of melanoma. On the other hand, there seems to be utility in studying patterns of vertical growth, independent of the relationship between such patterns and a competence for metastasis.
THE ADVENTITIAL DERMIS AND THE RETINACULAR DERMIS (SOIL)
The papillary dermis (a component of the adventitial dermis) is true stroma. In its pliancy, it accommodates minor shearing forces at the surface of the skin. It is adapted to provide nourishment to the epidermis. Its associations with the epidermis are so intimate that a change in one component is accommodated by reciprocal changes in the other. It is inappropriate to characterize the reticular dermis as stroma (prepared soil). For most melanomas, neoplastic cells invade and survive in the reticular dermis only by having acquired the capacity to migrate and, in the process, to carry thin sheaths of their own stroma with them. Neoplastic melanocytic cells must have acquired special qualities, if they are to invade the reticular dermis. Such cells give up the security of a community for the life of a migrant.
In the most common sequence, if neoplastic melanocytes have evolved along the neoplastic axis leading to the final common pathway, they tend to be rounded or polygonal and tend to be confined for a period in the pattern of rounded nests in the papillary dermis. The patterns are those of pattern III for a variable period of time. This set of virtual images has relevance for the communal period of neoplastic progressions (the sphere of Fig. 2). In this anatomic confinement in true stroma or prepared soil, there is provision for the eventual emergence of a population of c,ells with the capacity to induce a migrant’s stroma (the migratory phase of neoplastic progression). Level IV invasion (pattern IV) is a property of neoplastic migrants (cells which abandon the security of a community): Fig. 2). In the system evolving from the common final pathway along the neoplastic axis, level IV invasion is generally an acquired property. On the other hand in spindle cell, fasciculated melanomas, that have evolved as such from their inception, the ability to invade the reticular dermis is inherent; the spindle cells of such lesions are potential migrants even when confined to fascicles; fascicular patterns may first facilitate migrations into the reticular dermis, and then facilitate the discharge of individual migrants into the interstitium of the reticular dermis. This is even a property of benign spindle cell neoplasms such as the Spitz melanocytoma, and the pigmented spindle cell melanocytoma.
MIGRANTS - THE EXTENSION OF ROOTS
In the concept of minimal deviation melanoma, the emergence of a vertical growth component signified a neoplastic progression to a higher leve,l or more developed segment (Figs. 1 & 2). The emergence of a typical vertical growth component might be characterized as a demonstration of the capability of a tumor to "take root;" in the process, cells and stroma reciprocally complement each other. By nature, such a tumor enlarges in bulk as a plaque, or an expansile nodule (i.e., the cone of Fig.1 and the cone and sphere of Fig. 2). By progressions, either in sequence from the stage of a confined, expansile nodule, or as phenomena concomitant with the expansile phase, melanocytic neoplasias acquire new capabilities; they come to grow in invasive patterns without regard for either native structures or anatomic boundaries of the dermis; having acquired the new capabilities, such a tumor will freely violate anatomic boundaries, particularly the boundary defining the interface between the papillary dermis and the reticular dermis (Fig. 2), but also those of vessels. Such a lesion not only has taken root but grows without regard for surrounding structures. Cells forming the nodules of typical vertical growth components are communal; they are dependent on their neighbors and on a specialized stroma. Cells that have acquired the capacity to invade the reticular dermis have made a transition from residency in a community to become universal migrants. They are independent of the specialized stroma of the community, and of their neighbors. Spindle cells are more adapted as universal migrants at early stages of neoplasia than are rounded cells.
SOWN SEEDS (VARIANT VERTICAL GROWTH)
In the setting of the common dysplasias, patterns in which nests of cells are regularly, but loosely, spaced in the papillary dermis; show at least moderate cytologic atypia; and are associated with markers for host immune response, qualify as variant vertical growth patterns. If in addition, the nests in variant vertical growth patterns are entrapped in a condensed acidophilic stroma or enclosed by concentrically arranged acidophilic lamellae, the patterns qualify as arrested variant vertical growth. If they are supported by a loose, pale, mucinous or watery matrix, and associated with lymphoid infiltrates, the patterns are those of variant vertical growth without additional qualifications. Variant vertical growth is an accretive phenomenon in which cells are delivered from the epidermis into the dermis in sequences, but expansile growth from a dermal nidus is limited. The stromal response may play a role in limiting expansile growth in the dermis, although the sequestered components from the epidermal domain may themselves contribute limited bulk. Basically, new nests are sown into the dermis in sequential fashion but remain dormant; in the process, the domain of the papillary dermis may be expanded in bulk, but the seeds may not have truly take root.
TAKING ROOT (TYPICAL VERTICAL GROWTH)
In typical vertical growth, the seeds take root. As the population of cells in the soil expands, the transfer of seeds from the epidermal domain to the dermis becomes incidental to phenomena in the dermis; the cells of nests in the dermis proliferate and independently contribute new cells and nests of cells to the dermal population. The new nests do not move far from their site of origin. They cluster tightly and typically the clusters form an expansile nodule, although in some examples the newly formed nests tend to form a plaque in which the nests are closely spaced. In some examples, it may be difficult to make a sharp distinction between typical and variant vertical growth. For either variant, the character of the stroma is an important feature. If the stroma is loose and delicate, then the odds are great that the cells of the dermal component have taken root. If it is condensed, and acidophilic, or concentrically laminated, then the potential is constrained; in the latter setting, the prognostic parameters may not be as predictable of the likelihood for metastasis as they are in the non-arrested variants.
NEOPLASTIC PROGRESSIONS AND INTRALESIONAL POLYMORPHISM
Although typical vertical growth, in its progressions, is independent of phenomena at the dermal-epidermal interface, there is provision for neoplastic progressions to a higher degree of cytologic atypia in the population of cells at the dermal-epidermal interface. These new populations may be transferred to the dermis and may contribute to the dermal components. This may be one explanation for the polymorphism seen in some typical vertical growth components. If typical vertical growth ensues from a population of dermal cells showing less than marked cytologic atypia, there is a potential for neoplastic progressions to a higher of cytologic atypia in the dermal component, independent of the phenomena at the dermal-epidermal interface. This is another potential pathway leading to polymorphism in a vertical growth component.
The manner in which the cells, having taken root, interact with the soil determines the manner in which the expanding neoplasm interacts with anatomic boundaries. The next visit will be to that part of the garden concerned with the mechanics of proper gardening.
1. Reed RJ: Atlas of Dermatopathology: Melanocytic Nevi and Related Tumors of the Skin. Amer Soc Clin Path, CCE, Chicago, Ill, 1975, pp 7-8, figs. 15-19.
End of Part 2:
Click to go to Part 3 or to return to Part 1