Excess of multifocal tumors in nephroblastoma: implications for mechanisms of tumor development and genetic counseling.

Referring to the mutational theory of carcinogenesis in embryonal tumors, it is commonly accepted that patients with multifocal tumors are hereditary cases. This is based on the implicit assumption that each tumor results from a single mutational event occurring in a cell that has already inherited a mutation, and that these tumors are independent. We studied the distribution of tumors in 1,868 cases where the focality was known (SIOP 1, 2, 5 and 6). Using all the supposed gene carriers (bilateral and unilateral multifocal cases), and assuming a Poisson distribution of tumors, we estimated the mean number m of tumors in each kidney to be 0.37. Comparing the observed distribution of cases to the expected one, we found a very bad fit to this hypothesis (P < 10(-9)). This is due to an excess of multifocal tumors, particularly in unilateral cases. These findings have important implications in genetic counseling, since the usual practice of considering multifocal tumor patients as hereditary cases may result in a large overestimate of the recurrence risk in such cases. The implications for the mechanisms of tumor development are also discussed.