The phrase borderline melanocytic tumor was employed repeatedly by Zembowicz, Carney, and Mihm. In fact, in the very first line of their abstract, they told of having studied "borderline melanocytic tumors," and they went on to write about "borderline melanocytic lesions such as pigmented epithelioid melanocytoma," advising in the last line of their work that "we propose pigmented epithelioid melanocytoma, as a provisional histologic entity, a borderline melanocytic tumor or low-grade melanoma, encompassing animal-type melanoma and epithelioid blue nevus." Nowhere, however, do they define explicitly what they mean by "borderline melanocytic tumor" (or "provisional histologic entity"). The closest they come to that is the statement that follows: "An important implication of this paper [sic] is that it supports a recent paradigm shift occurring in melanocytic pathology, which departs from dichotomous classification of all pigmented lesions as either benign nevus or malignant melanoma. Follow-up data of pigmented epithelioid melanocytoma, its histologic similarity to Carney complex-associated epithelioid blue nevus, and the bland histologic features of some cases all suggest that pigmented epithelioid melanocytoma is a unique melanocytic neoplasm with low probability of systemic spread despite its ability to metastasize to regional lymph nodes."

In fact, this "paper" does not support the "new paradigm"; the authors, unwittingly give the lie to it. Moreover this "paradigm" is incompatible completely with one principle basic of logic, to wit, that of the "excluded third," it stating that of two opposing qualities (in the present occurrence benign/non-malignant versus non-benign/malignant), an object can have only one or the other. In very simple terms, a rock can be either black or non-black, it cannot be anything else concerning blackness although, of course, the matter of blackness could be undecided, that is, unknown.

The coworkers proceeded to comment thus: "pigmented epithelioid melanocytoma is not the first example of metastasizing or recurring melanocytic tumor with favorable clinical outcome. Cases of 'benign' metastasizing Spitz nevi and spitzoid 'nevoid' melanomas have been described. However, histologic features of Spitz nevus and pigmented epithelioid melanocytoma are quite different. Therefore, it is possible that additional borderline melanocytic entities exist within a not uncommon group of atypical spindle and epithelioid cell tumors, which can not be definitively classified as Spitz nevi or melanoma according to present histologic criteria."

Despite the verbiage, a reader has no idea whatever of the meaning of borderline melanocytic tumor, except perhaps that it means that a borderline lesion is of a nature unknown and, therefore, that no diagnosis of use clinically can be rendered with assurance. It would then mean that it is important to ameliorate either the definition or the application of criteria diagnostic. In no way should it justify the introduction of more and more terms incomprehensible.

Several problems of definition are found in the two sentences that follow: "Therefore, to avoid arbitrarily designating controversial lesions as melanoma and hoping to refine histologic criteria separating metastasizing from benign lesions, we have instead more recently categorized suspected examples of animal-type melanoma as 'borderline melanocytic tumors' under a descriptive rubric of 'pigment synthesizing melanocytic tumor,' which was later replaced by a more appropriate term pigmented epithelioid melanocytoma."

If there is no agreed on definition of animal-type melanoma, pigmented epithelioid melanocytoma, and borderline melanocytic tumor, so, too, it is for pigmented synthesizing melanocytic tumor. That phrase, which applies equally to the pigmented spindle-cell tumor of Reed, pigmented examples of Spitz nevus, and cellular blue nevus, contributes to making the matters under consideration here even more incomprehensible. We see no way by which introduction of the term pigmented epithelioid melanocytoma can avoid arbitrariness or help define "criteria separating metastasizing from benign lesions." A metastasizing lesion can only be distinguished from a benign and from a malignant non-metastasizing lesion by the very presence of metastases, whereas the only way to develop new criteria that enable distinguishing a benign from a malignant lesion is to acknowledge, unflinchingly, that, at a particular moment in time, those criteria are not known.

Reference is made by the authors to common nevus, such as in the context of "Three additional cases of combined nevus with a dermal component of a common nevus were present" and to common melanoma, such as in the sentence that "Its [pigmented epithelioid melanocytoma] frequent occurrence in races less susceptible to common melanoma . . ." Nowhere, however, is any attempt made to define common nevus and common melanoma. Any effort in that regard would be fruitless anyway; there is no such thing as a common nevus and a common melanoma, just as there is no agreed on definition of uncommon nevus and uncommon melanoma.

Mention is made of atypical Spitz tumors and of 'benign' metastasizing Spitz nevi but no effort is made at characterization of either of those notions abstruse.

Nuclei are described as being "regular" and having "irregular contours," infiltrates of lymphocytes are referred to as "brisk" and "non-brisk," and some neoplasms are categorized as "benign or low-grade," each of those terms being unfathomable unless one is a member of the fraternity that spawns them, those members communicating with one another in language they deem to be comprehensible, but that is incomprehensible to those who do not belong (and, ironically, to those who do).***

In the abstract, the authors wrote that none of the patients died of disease but in the discussion, they added the following:

Which is it? Did "no patient with pigmented epithelioid melanocytoma die of disease" or did a "patient died 3 years after diagnosis with generalized metastatic" pigmented epithelioid melanocytoma? The answer may come from the "new paradigm": it is neither true nor false or true and false that patients die of pigmented epithelioid melanocytoma.

In brief, pigmented epithelioid melanocytoma, along with "animal-type" melanoma and epithelioid blue nevus, are the source of a quandary that can be summarized as follows: Carney and Ferreiro, in 1996, presented "epithelioid blue nevus" as a nevus and never once mentioned the possibility of it being melanoma. Crowson, Magro and Mihm, in 1999, proposed "animal-type melanoma" as a "low-grade" melanoma but a melanoma (that is, a malignant neoplasm) nonetheless. Zembowicz, Carney and Mihm, in 2004, set forth "pigmented epithelioid melanocytoma" as a melanoma and, in the same breath, averred that it "encompasses" both animal-type melanoma and epithelioid blue nevus. So, in the end, which is it for these three entities? Nevus or melanoma? In our view, what Carney and Ferreiro described and pictured as "epithelioid blue nevus" is a nevus and what Zembowicz, Carney and Mihm described and pictured as "pigmented epithelioid melanocytoma" is a melanoma, whereas the neoplasms described "initially" by Crowson et al. were a mix of benign and malignant. Of course a melanoma can develop in a nevus of any kind, including so-called epithelioid blue nevus, and that possibility cannot be excluded for one or more of the examples in the series of Crowson, Magro and Mihm, of Carney and Ferreiro and of Zembowicz, Carney, and Mihm. What is inconceivable is that a nevus histopathologically can be identical to a melanoma. What surely is conceivable though is that a histopathologist can overlook the differences between them (Fig. 3).

View Figure

Fig. 3A–F  Zembowicz et al.'s Figure 1, original legend: FIGURE 1. Low magnification views (left column) with corresponding illustrations of cytologic detail (right column) of 3 PEMs with sentinel lymph node metastases (A and B, case no. 17; C and D, case no. 26; E and F, case no. 24). Note epidermal hyperplasia (A, C), blue nevus-like architecture (E) and mixture of epithelioid spindled cells (B, D, F). Cytologic detail of large epithelioid cells is best visualized in F. Our comment: Some tumors described as pigmented epithelioid melanocytomas are indubitable melanomas: the tumors presented in Fig. 1 show asymmetry, a satellite nodule metastatic is present in Fig. 1E, and there is no maturation of cells deep. Lymph node—even liver—metastases were present in some patients. Not all tumors having been illustrated satisfactorily, however, and no criteria compelling having been established for the diagnostic of pigmented epithelioid melanocytomas, the reader cannot exclude that some cases were benign, which we think was the case.