Historical Perspective: Lartigau et al., 1995

 
Lartigau and coworkers, in 1995,56 recorded their experience with 17 patients younger than 17 years of age who had what they believed to be de novo melanoma. The neoplasms, culled from a registry of 90 patients seen between January 1956 and December 1990, formed the basis of in an article titled, "Melanoma arising de novo in childhood: Experience of the Gustave-Roussy Institute" in Villejuif, France. The patients had been followed for as long as 35 years. Excluded from the study were children with giant congenital nevi, a family history of the dysplastic nevus syndrome, and xeroderma pigmentosa. The findings, in their own words, were these:
 
"The median age was 9 years and 9 months (range 2 years and 3 months-16 years and 9 months). The primary disease was located in the lower extremities in 10 cases, the trunk in five cases, and the upper extremities or head and neck in one case. The disease was localized for 10 patients at presentation (stage I), six had proven nodal metastasis (stage II) and one patient had nodal and breast metastases. The thickness of the primary lesions ranged from 0.64 mm to 10 mm (median 2.89). Five cases were at Clark's level III, eight at level IV and four cases were at level V. Six cases were classified as superficial spreading melanoma (pagetoid melanoma). In two cases (cases 9 and 16), the radial growth phase could not be readily placed into one or other of the subcategories, and was considered as unclassified (unclassified radial growth phase). For the nine remaining cases, three were nodular and three were classified as melanoma with spitzoid cells and three melanomas were unclassified because of the poor quality of the samples, which did not allow identification of the radial growth phase. There were no cases of lentigo maligna or acral lentiginous melanoma."
 
Based on those observations, Lartigau and coworkers came to consider differentiation histopathologically of Spitz's nevus from melanoma as vexing, and this is how they expressed it:
 
"The pathological features of melanoma in childhood are particular in relation to several features. One of these is the absence, as in this study, of acral lentiginous and lentigo melanoma. Another is the difficulty of the differential diagnosis between Spitz's naevi and melanoma. Even if the criteria of differential diagnosis between Spitz's naevi and melanoma are known, the pathological diagnosis of melanoma in children remains a challenge. The criteria in favour of melanoma include a pagetoid spreading of single tumoural cells into the overlying epidermis, cellular pleomorphism and high mitotic activity. Melanoma with Spitzoid [sic] cells constituted three of our cases."
 
One such melanoma with cells like those of Spitz's nevus was illustrated by Lartigau et al. in a neoplasm that came from a child 5 years of age in whom metastasis developed months later. The melanoma was shown at intermediate and high magnifications (Fig. 14), but was described in the legend as being a "broad melanocytic lesion with effaced rete ridges and striking pagetoid pattern in the epidermis. A diffuse lymphocytic infiltrate is observed in the dermis. The dermis is widely infiltrated by large cells with significant atypia and frequent mitoses."

View Figure  		Fig. 14  Comment: The photomicrographs shown at intermediate and high magnification by Lartigau et al. communicate well how difficult it can be to assess accurately, at those magnifications alone, the changes in the upper part of some authentic melanomas that occur in children. The findings could be misinterpreted as those of Spitz's nevus by virtue of both architectural pattern and cytologic features. In reality, both the architectural pattern and the cytologic features seen here are those of melanoma for these reasons: marked variation in size and shape of aggregations, bizarre shape of some aggregations, and startling pleomorphism of nuclei, some nuclei being very large, but some being very small, the latter finding being incompatible with nuclear features of Spitz's nevus.
 
Serious limitations puncuated the work of Lartigau et al., not the least of them being their adherence to concepts that should have been passe because they are so overtly wrong (for example, so-called Clark's histogenetic classification of melanoma [which is based entirely on anatomic site] as lentigo maligna, superficial spreading, acrrolentiginous, and nodular, and the illogical, and therefore unusable, system of "radial gowth phase versus vertical growth phase" of melanoma [which is not contrasting because vertical is a component of radial]. Perhaps the most eggregious limitation, however, was the failure of those coworkers to set forth workable criteria for distinguishing Spitz's nevus from melanoma.