Mutation theory of carcinogenesis in retinoblastoma.

To test Knudson's hypothesis that two successive mutations are involved in retinoblastoma, we studied the data on 899 cases. Some of the findings appeared to differ from those that might be expected if Knudson's hypothesis were correct. Certain criticisms of Knudson's methodology and model were suggested. Alternative explanations proposed were 1) the role of the sequence in which mutations occur, and 2) the possibility of three mutational events.