In 2004, Kazakov et al. in a piece titled, "Melanoma with prominent pigment synthesis (animal type melanoma). A case report with ultrastructural studies," sought to lend credence and to give legitimacy to the notion of "animal-type melanoma" by adding information ultrastructural and immunohistochemical to the definition of it (Figs. 3A–B). [13]

The patient, a 28-year-old woman, presented herself with a solitary, exophytic, heavily pigmented nodule on the right arm which was said to have been there two years before excision of it. A sentinel lymph node biopsy revealed involvement massive of the parenchyma of the node by abnormal melanocytes virtually identical to those of the primary neoplasm, the latter being made up of large, heavily pigmented cells, structures adnexal having been destroyed partially or completely. The cells constituent formed large sheets, which were divided from one another by thick bands of connective tissue. The nuclei, obscured largely by melanin, displayed little pleomorphism. A rare mitotic figure could be detected after bleaching had been performed. Zones en masse of necrosis were observable, and the epidermis was seen to be unaffected.

Studies ultrastructural were carried out on paraffin-embedded tissue prepared in a manner suboptimal. The cytoplasm of the neoplastic cells was noted to contain melanosomes at various stages of development (mostly stages II and III). Melanosomes in abundance, some of them abnormal, were apparent. Melanophages formed a "minor component," they housing large compound melanosomes and not single melanosomes. No elements schwannian were identified, contrary to what had been the experience of Tuthill et al. [8] in the only other account of a study ultrastructural of heavily pigmented lesions in the skin like those of interest to Kazakov et al. Parenthetically, what Tuthill and coworkers described was called by them "epithelioid neurocristic hamartoma" and not melanoma, which reinforces how great is the extent of perplexity in the mind of authors about the entity under discussion here.

By methods immunohistochemical, Kazakov et al. were able to recognize the following: "The same pattern of staining was seen on the bleached and unbleached slides both in the skin and in the lymph node. The neoplastic cells stained positively with MiTF (nuclei), NSE, NK1/C3, tyrosinase (weak), p53, and CD68, S-100 protein, HMB45, Melan-A, Mac367, and lysozyme reacted negatively. Occasional cells (< 1%) reacted mith MIB-1."[13] Unfortunately, no illustration of positivity to the different markers studied was provided, but the findings immunohistochemical were interpreted by Kazakov and fellow workers thus: "While the expression of CD68 in conventional melanoma ["common" melanoma], as reported in large series, ranges between 70 and 94 %, a positive immunoreaction for Mac387 and lysozyme is a rare finding. CD387 and lysozyme are lysosome-associated glycoprotein. They do not normally present in melanocytes and their occurrence in melanomas has been explained by the acquisition of a phagocytic ability during carcinogenesis. The presence of intralysosomal melanosomes complexes in the cytoplasm of the neoplastic cells as detected by electron microscopy may explain CD68-positivity although the reason for the negative reaction to Mac387 and lysozyme is unknown." [13]

A reader attentive will appreciate how contradictory are these sentences. Were Mac387 and lysozyme positive or negative? Moreover, considering the immunoreactivity present, it is surprising that melanophages were claimed to be only a "minor component" in the assessment ultrastructural.

The authors contended that the cells constituent of this neoplasm had acquired properties histiocytic through a process of transformation. Although such acquisition is conceivable, it is far-fetched and, we think, without foundation in this instance; it is pure speculation. In order to make a case convincing for such a supposition, it would have been important to include images of the immunostains in order to give authenticity to a hypothesis so daring and it would have been useful to picture ultrastructurally cells possessing characteristics of both types of cell. Instead, the authors appeal to faith in these words: "We believe that the results of our immunochemical studies, which were observed independently in our 2 institutions, are reliable." [13] That declaration simply is not sufficient. What a reader wants to know is whether the positivities were compelling everywhere in the proliferation or were only weak in loci and whether there was any correspondence ultrastructurally ? It is not demonstrated convincingly that the very same cells exhibited both attributes melanocytic and histiocytic.

It might seem that electron microscopy provides support for the notion of "animal-type melanoma." but identification of melanoma by way of observations ultrastructural surely is not a subject about which experts agree, as is made clear in the volume by Erlandson devoted to the subject of Ultrastructural Electron Microscopy of Tumors (Raven Press 1994:pp 151158) [21] This is what is stated there concerning the very matter that engages us: "To the electron microscopist the most significant pigment is melanin, as ultrastructural evidence of the synthesis of this inclusion in poorly differentiated neoplastic cells conclusively establishes the diagnosis of MM." [21] Thus, synthesis of melanin is conclusive of character melanocytic only in the setting of "poorly differentiated neoplastic cells." That is confirmed by this fact: "The rarely occurring synthesis or storage of melanin pigment by nonmelanocytic neoplastic cells of neural crest origin or other types of nonneoplastic or neoplastic cells has also been reported . . . [The] "melanophages" often [but not necessarily!] contain compound melanosomes . . . and occasionally premelanosomes within membranous sacs." [21] To us, these lines mean that the melanocytes with phagocytic properties immunohistochemical of Kazakov et al. may have been melanophages that share some properties ultrastructural with melanocytes.

In conclusion, we think that the neoplasm presented by Kazakov et al. is a melanoma that regressed partially and metastasized to a lymph node where the metastasis also regressed. Although no disease residual was apparent in the patient two years later, the prognosis, nevertheless, remains bleak because metastases of melanoma are the equivalent of time bombs.

A melanoma in the process of regression, or having regressed completely, reveals itself through two characteristics histopathologic, one designated "fibrosis" and the other "melanosis." Usually, regression as assessed grossly assumes the aspect of a lesion flat, i.e., a macule or small patch, but, rarely, as in the example under scrutiny here, it can appear as a nodule. Moreover, one can see a "nodule" of regressed melanoma in the deep reticular dermis or the subcutis, it representing regression of a "satellite" metastasis, rather than of a primary melanoma and, in that circumstance, always taking the form of melanosis, i.e., a collection, ranging in size fanciful from a puddle to a lake, of melanophages. Such a finding is indicative that metastasis of melanoma already has occurred, the regression complete being thought to be a consequence of lymphocytes sensitized by cells of metastatic melanoma returning from a regional lymph node to obliterate abnormal melanocytes of the lesion primary in the skin, and that is what we think happened in the case of the lesion reported on by Kazakov et al., which for us is not an example convincing of "animal-type melanoma." Last, even a metastasis more distant can regress, for example in a lymph node, the melanocytes malignant of it being replaced by melanophages.