Authentic primary cutaneous melanomas*, as proved by metastasis do arise de novo in children despite assertions to the contrary.^1–83 A review of prepubescent melanoma that deals only with patients who fulfilled our criteria for prepubescence, that is, a primary cutaneous de novo melanoma in a child 10 years of age or younger at the time of diagnosis, would not consist of even a single sentence, the reason being that, to the best of our knowledge, no series has ever been published of children 10 years of age or younger with proven primary cutaneous melanoma. In order to elucidate and, thereby, to instruct, we have analyzed all articles pertaining to melanoma that developed de novo in young persons, focusing especially on those that concern children and in which histopathologic findings were presented. Although for purposes of our study we elected to define "prepubescent" as 10 years of age or younger, we have incorporated in our review of the subject all articles that dealt with the matter in general of melanomas in young persons, irrespective of what age was chosen by authors as the upper limit for defining a child.

Our own series, however, includes only primary cutaneous melanomas that developed de novo in children who seemed to be prepubescent. None of those melanomas arose in association with a nevus of any kind, including a congenital one,^84–114 in a fetus acquired in utero from a mother who herself had metastases of melanoma,^115–119 in children with xeroderma pigmentosa,¹²⁰ in conjunction with other malignant neoplasms in children who were immunocompromised,¹²¹ or in a zone of previous irradiation.¹⁰²

At the outset of an undertaking such as this, it is essential to define puberty. Lee, in 1980,¹²² wrote about "Normal ages of pubertal events among American males and females" and offered this definition: "Puberty may begin from age 8.0 to 14.9 years for females and from age 9.7 to 14.1 years for males and is complete by age 12.4 to 16.8 years for females and by age 13.7 to 17.9 years for males." Based on this assessment, we made an arbitrary decision, for the purpose of this work, to regard "prepubertal" in both sexes as 10 years of age or younger, a time when it may be inferred that in the vast majority of instances puberty has not yet been reached. To our knowledge, ours is the largest series reported on of de novo melanomas, proven by metastasis, that began in children who presumably were prepubescent. In 1993, Pensler and coworkers, in an article titled, "Melanoma in pre-pubertal children,"⁵² defined "pre-pubertal" as 13 years of age or younger and one of their patients was 13 years old. Only four of the nine children in their series had sure signs of metastasis of melanoma; in one of the four the melanoma originated in the central nervous system, in one the site of the primary melanoma was not known, and in two the primary melanoma was in the skin. All 11 prepubescent children who are the subjects of our study had primary cutaneous melanoma that metastasized.

By virtue of what is presented here we hope that fellow histopathologists will be able to recognize better an authentic primary cutaneous melanoma in a child, to make the diagnosis with confidence on the basis of repeatability of the appearance of it as viewed especially at scanning magnification of a conventional microscope, and to differentiate it with assurance from a Spitz's nevus.