Retinoblastoma: mutational mosaicism or host resistance?

Carlson and Desnick's [1979] proposals of multiple allelism and mutational mosaicism to account for the observed variability in penetrance and expressivity in retinoblastoma do not seem to give a clear-cut explanation for the facts that 1) a bilaterally affected patient could have an unaffected carrier child who in turn would produce bilaterally affected offspring, and 2) carriers of the retinoblastoma gene are prone to nonradiogenic osteosarcoma irrespective of their phenotypes. However, these observations are readily explained by assuming that inherited host resistance plays an important role in the manifestation of a single major gene for retinoblastoma. Based on their model, Carlson and Desnick offer recurrence risk figures for genetic counseling, but their figures concerning the transmissible types of retinoblastoma do not seem more accurate, as assumed by them, than those previously estimated from the family data. We present evidence that host resistance may be regarded as a multifactorial threshold character with high heritability, and we propose that tissue-specific genes, which we presume to be involved in normal differentiation of retinoblasts and osteoblasts, play a crucial role in malignant transformation. Two lines of suggestions are made for future studies: one, to search for polymorphic genes associated with retinoblastoma, and the other, to see whether the principle of synchronous appearance of multiple tumors in persons at high risk applies to other cancers as well.