Hi,
This is a good question, and multiple factors could play a role here:
1) Copy-number calling is imperfect, and some of the deep deletion calls are not actually homozygous. This could be further complicated in tumors that have undergone whole genome duplication. That is why we call these events “deep deletions” and not “homozygous deletions”.
2) Clonality could play a role as well, where the deletion and the mutation could occur in different cells.
3) Lastly, a partial homozygous deletion could co-occur with a mutation in a different part of the gene.
That said, scenario #1 is most likely. There tend to be quite a few false positives in the automated copy-number calling pipeline from TCGA. For the TCGA manuscript on prostate cancer, published in Cell, we tried to curate all deep deletion calls in canonical tumor suppressor genes, which is why you will find a lower deletion rate in these two genes in that study:
Niki.