The significance of vascular invasion in primary melanomas of thin or intermediate thickness.

Abstract

I n a recent letter entitled “An Inaccuracy in an Otherwise Solid Study,” Dr A. Bernard Ackerman criticizes the statement by Kashani-Sabet et al that “the role of [vascular invasion] in the prognostic assessment of malignant melanoma has remained controversial.” It is the view of Dr Ackerman and his associates that vascular invasion by any melanoma is “virtually equivalent to metastasis.” He says that, except for a single statement in the literature saying that vascular invasion does “not influence prognosis significantly,” he is unaware of any “statement to that effect anywhere.” Perhaps I can help. Clark et al, in their classic article on prognostic factors in melanoma, stated that “[t]wenty-three attributes [including vascular invasion] were tested for entry into [their] model. Six had independent predictive prognostic information.” Vascular invasion was not among these; it dropped out from the multivariate analysis. More recently, using logistic regression, Barnhill et al examined 14 pathologic parameters, including vascular invasion, in melanomas from 548 patients. In their univariate analysis, vascular invasion showed prognostic significance, but in their multivariate analysis it again dropped out. Vollmer reviewed 54 reports of multivariate analyses of prognostic factors in melanoma; he found that only 3 of the studies showed vascular invasion to be significant, whereas 10 of them did not and 41 of them ignored it altogether. It follows that the statement by Kashani-Sabet et al is correct and that Dr Ackerman’s criticism is not justified. The explanation for the confusion is actually quite simple: the presence of vascular invasion in primary melanomas is closely associated with tumor thickness—thick melanomas are very much more likely to show vascular invasion than thin melanomas. Therefore, in univariant analysis, vascular invasion is associated with a poor prognosis, simply because the overwhelming majority of melanomas that show it are thick, and thickness is an extremely strong predictor of a poor prognosis. However, in multivariate analysis, in which the association with tumor thickness has been accounted for, vascular invasion has either been found to be a relatively weak predictor (of a poor prognosis), or it has not been found to be a significant prognosticator at all. Those few studies that have shown a predictive value for vascular invasion have generally been examinations of thick melanomas. Kashani-Sabet et al have made a major contribution to this area by their finding that vascular involvement may be a more important prognosticator than previously thought. Especially important is their observation that melanoma cells need not be found within vessels for vascular involvement to be significant: epiluminal neoplastic cells abutting onto, but outside of or “cuffing,” vessels that are separate from the tumor may be as significant, or even more so. This observation corroborates the findings of Barnhill et al, who have termed the latter phenomenon angiotrophism. For thick melanomas, these recent studies may influence the way that prognostic factors are evaluated. However, for those few thin melanomas (ie, 0.75 mm in thickness [Breslow]) or those of intermediate thickness (ie, 0.75-2 mm in thickness) that show vascular involvement, it is not at all clear whether, or to what extent, this portends a poor prognosis. Even in the study of Kashani-Sabet et al, tumor thickness was found to be a more important prognosticator than vascular involvement. Also, vascular invasion, when it does occur in thinner melanomas, is more likely to involve more superficial vessels than when it occurs in thicker lesions. Additional studies will be needed to resolve this issue and, given the relative infrequency of both vascular invasion into and angiotropism around vessels in melanomas of thin or intermediate thickness, they are not likely to be completed quickly. Until such studies are performed, it is imperative that patients with melanomas of thin or intermediate thickness be counseled, and treated, according to established prognostic indicators, irrespective of vascular invasion, at the very least until many more data are obtained and analyzed.