In spite of the shortcomings of expressed criteria for diagnoses, they do provide an adjunct for teaching pathology. The danger lies in relying on such criteria, particularly a single criterion, in making a diagnosis. Pundits who are quite capable of correctly diagnosing a difficult case often state that they have done so by utilizing a few pet criteria. That this has involved a more complex, though subconscious, process has not been appreciated. This becomes harmful when a resident with insufficient visual experience attempts to diagnose a complex case based upon one or two such criteria. This can be a serious problem with melanocytic lesions in that there is probably not a single criterion, taken in isolation, that allows one to distinguish between an benign and a malignant lesion.

A diagnosis or a narrow differential diagnosis occurs almost immediately to an experienced dermatopathologist when first confronted with a slide. Subsequent manipulation and examination of the slide is for confirmatory purposes when the diagnosis is apparent, or  sequential fields that satisfy a subconscious algorithm are examined  when there is a differential diagnosis to ponder. At this point he/she either knows the diagnosis or, hopefully, is aware that he/she does not. It is also at this point that he/she dictates observations or statements that support the diagnosis.
 

The images in Part One could have been picked by someone who sought supporting criteria after making a benign diagnosis, possibly spindle cell nevus (Spitz), .  Let us examine some of the 'benign' criteria that one could invoke in this case. This examination is not to denigrate these criteria but to put them into perspective.

• SYMMETRY^{Figures 1 & 2}
Scanning power lesional profile: Symmetry is often in the eye of the beholder, and this determination frequently is made after deciding the nature of the lesion based on other factors. I can recall a national meeting where the speaker projected a markedly asymmetrical banal nevus and verbally described it as being symmetrical. Snickers from the audience were ignored. There are symmetrical melanomas and there are asymmetrical nevi though these are in the minority. Neoplastic progression can start anywhere, including equidistant from the edges, in a precursor lesion.

Intralesional  profiles as seen under higher powers:  Symmetry from top to bottom is lacking in many compound nevi, intradermal nevi, and spindle cell nevi. In such cases it is referred to, euphemistically, as maturation.

There can be dissimilar populations of nevus cells side by side resulting in asymmetry. We seldom notice this because the dissimilar cells are 'normal' as determined by previous experience. The 'abnormal' cells in combined nevi stand out and are troublesome to those not familiar with them. The 'abnormal' cells or cell cluster in a combined nevus  that contact an adjacent population of 'abnormal' cells should not have expansile properties (in the easier cases). When several dissimilar melanoma cell populations are found in the dermis, they characteristically are in contact with other melanoma cells, either cytologically malignant or minimal deviation variants.
 

• INTRAEPIDERMAL NESTS IN THE PERIPHERY^{Figures 3 & 4}
So-called fuzzy edges can be seen in dysplastic nevi, nevi having architectural disorder, and in subtle melanoma precursors, particularly lentiginous ones. Melanomas arising in such backgrounds have led to the false corrollary that clusters of normal nevus cells should not be in the periphery of a melanoma. Nests of normal nevus cells can be found in the periphery of a melanoma arising in an ordinary nevus if neoplastic transformation has not occured in the periphery.
• KAMINO BODIES^{Figure 5}
Kamino bodies were once purported to be found exlcusively in spindle cell nevi (Spitz). They can be found in melanomas though they are usually absent or few in number. Dyskeratotic keratinocytes also can be found uncommonly in melanomas and in halo nevi.
• SPACED FASCICLES OF SPINDLE CELLS^{Figure 6}
Fascicles composed of cytologically benign spindle cells that are separated by stroma are characteristic of spindle cell nevi (Spitz). Similarly spaced fascicles of spindle cells can be found in desmoplastic melanomas, and the degree of atypia is minimal in some examples.

• MATURATION^{Figure 7 andFigure 12}
Maturation is best illustrated in those Spitz nevi that have large spindle cells or clusters of large, myoid, epithelioid cells in the superficial part of the lesion with diminution in cell size in the deeper aspects of the lesion. This can be emulated in melanomas where normal nevus cells have been pushed down or displaced by invasive melanoma cells. This is also emulated in some dysplastic nevi and in some melanomas where there is sequential invasion by phenotypically different cell populations. I would interpret the large, atypical epithelioid cells in the most superficial nest in Figure 7 as being fully developed melanoma cells, the next layers as melanoma cells of minimal deviation type, and the very deepest cells as nevus cells.