We conducted this study to clarify a hypothesis that had been generated from preclinical work and small clinical datasets. One additional background point that should be made is that although not all previously reported small datasets had demonstrated a relationship between PTEN expression and outcome, most summaries of these data mentioned just the positive associations.

So, our study has provided definite data that will be helpful to the overall field. It was very important that we had a large set of tumors, long follow-up, blinded analysis, and various cutoffs for negativity that yielded consistent information. ■