Lever

 
"In early malignant melanoma and at the advancing border of older lesions one may find pathologic changes within the epidermis but little or no invasion of the corium by tumor cells. This stage may be referred to as malignant melanoma in situ. In some areas the epidermis, because of the presence of numerous atypical clear cells, appears vacuolated, torn apart and even disintegrated. In other areas the epidermis contains large, irregularly shaped, often deeply pigmented nests composed of nevus cells which appear atypical and may show mitotic figures. In the upper corium, close to the epidermis, one finds a dense, bandlike inflammatory infiltrate intermingled with numerous chromatophores." . . .
 
"In invasive malignant melanoma one finds considerable junction activity. The junction activity is more pronounced and more irregular than that seen in junction nevi. The nevus cells are large, atypical and show occasional mitotic figures in contrast to junction nevi which do not show them. One finds not only dropping off of nevus cells and cell nests downward into the corium, but also migration upward into the rete malpighii. The epidermis may become so fully permeated with tumor cells that it disintegrates and ulceration results . . . "
 
"The size and shape of the tumor cells in the corium show great variation. Nevertheless, two types of cells can be clearly recognized, a cuboidal and a fusiform type . . . Mitotic figures are usually present in malignant melanoma, but often only in small numbers. They should be searched for, because their presence is of great value as evidence that the tumor is malignant, inasmuch as in pigmented nevi mitotic figures are never found." (Fig. 3) Lever WF. Histopathology of the skin. Philadelphia: JB Lippincott, 1949:400–3.

View Figure  		Fig. 3   Our diagnosis and comment: Melanoma. The designations "early" and "late" for melanoma are not accurate. Abnormal melanocytes in a patch of melanoma may remain in situ, that is, confined to the epidermis and adnexal epithelium, for decades, whereas abnormal melanocytes in a papule of melanoma may have come to involve the dermis in a matter of but a few years. The melanoma pictured here cannot be deemed "early;" it measures about 1.0 mm in thickness to the base of the photomicrograph.
 

Brief Critique

 
The criteria utilized by Lever are as follows: (1) atypical clear cells, (2) mitotic figures, (3) pleomorphic nuclei of neoplastic cells, (4) dense, bandlike infiltrate of inflammatory cells, (5) chromatophores, (6) junctional activity, (7) dropping off of individual nevus cells and nests of nevus cells, (8) upward migration of nevus cells in the epidermis, (9) disintegration of the epidermis, and (10) ulceration. Apart from atypical melanocytes, mitotic figures in melanocytes, and melanocytes positioned above the dermo-epidermal junction, none of the criteria advanced by Lever benefits a histopathologist in coming to a specific diagnosis of melanoma. And each of those three findings that have validity are encountered also in Spitz's nevi. In addition, several ideas advocated by Lever are incorrect, among them disintegration of the epidermis, nevus cells, junctional activity, and mitotic figures never being encountered in nevi.
 
Some sense for how different was the use of the term melanoma in 1949 can be learned by glancing at the conditions said to be "Melanomas" in Chapter 19 of Walter Lever's first edition of the textbook. In brief, "pigmented nevus," "lentigo," and "blue nevus" were all considered by him to be melanomas, which in those days meant any melanocytic neoplasm.