Summation and analysis of the data of Crotty and her associates presented in their Table 1 of children of an age comparable to those in our series, namely, 10 years of age or younger, of which there were eight, revealed four children with long-term follow-up who died of metastasis melanoma.⁴⁸ Their case 5 was a 6-year-old girl with a melanoma 2.6 mm in thickness which metastasized to a regional lymph node and became apparent there 39 months after diagnosis of the primary neoplasm. She died of liver metastases 158 months after diagnosis of the primary melanoma had first been made; an 8-year-old girl with a melanoma that was 5.5 mm in thickness (case 8) was found to have metastasized to a regional lymph node eight months after diagnosis of the primary neoplasm had been made and two months later caused her death; a 5-year-old girl, their case 11, with a melanoma 1.6 mm in thickness and that metastasized to the ovaries and omentum, resulting in her death 146 months following diagnosis of the primary neoplasm; and an 8-year-old boy in whom the thickness of the melanoma was not given but who died 14 months after diagnosis of the primary neoplasm. In October 2000, Dr. Crotty visited our Academy in New York City, and we had the opportunity to review with her sections of tissue from the melanomas in five of the children reported on by her. The melanomas she showed us came from children who were 4, 10, 10, 12, and 12 years of age. The histopathologic findings in the melanomas in those children were distinctly different from those that comprise our series; those melanomas were thinner and resembled more closely melanomas that develop in postpubescents, for one prime example, being horizontal rather than vertical in orientation to the skin surface. The melanoma in the 12-year-old was entirely in situ.

In the series of Crotty, the two children with "spitzoid melanomas" (their cases 5 and 6) died of the effects of metastases 11 years and one year, respectively, following diagnosis of the primary neoplasm. Case 5 was a 9-year-old girl and case 6 was a 13-year-old girl. Both "spitzoid melanomas" had been misinterpreted histopathologically as Spitz's nevi. The findings were pictured only at high magnification.

In sum and in short, all of the melanomas in the 11 prepubescent children in our series were thick and all metastasized. The speed at which some of them grew may be inferred from the thickness of them at the time of biopsy, for example, by one year of age a melanoma already had entered the subcutaneous fat. That being the case, and knowing that for practical purposes all de novo melanomas begin in situ, it is reasonable to think that, as a rule, melanomas in prepubescent children grow with celerity. Very exceptionally a melanoma in a prepubescent child may be biopsied when it is flat (in situ) or very slightly elevated (thin), and that was true of some youngsters in the series of Crotty et al. Because melanomas in prepubescent children usually have morphologic features that differ from those of melanomas in postpubescents, clinicians and histopathologists must be exquisitely alert to those differences. The challenge for clinicians is to identify a melanoma in a young child at a stage when it is still curable and the challenge for histopathologists is to avoid misdiagnosis of such a melanoma as Spitz's nevus or cellular blue nevus.