Predictive factors in localized prostate cancer: implications for radiotherapy and clinical trial design.

The anatomic extent of prostate cancer has long served the role of providing prognostic information to assist in therapeutic decision-making, evaluating treatment outcomes, facilitating information exchange between medical centers, and promoting cancer research. However, nonanatomic factors are also associated with important pathological features of this condition and may be used to estimate therapeutic outcome. At present, tumor grade (eg, Gleason score) ascertained from the diagnostic biopsy specimen and the pretherapy serum prostate-specific antigen level are readily available in clinical practice. This information may be used along with clinical tumor stage to construct predictive models. These models may provide reliable estimates for the likelihood of extraprostatic tumor extension, seminal vesicle invasion, or pelvic lymph-node involvement. Consideration of this information may play a vital role in the selection of radiotherapeutic modality and in the definition of external beam radiotherapy treatment volumes. These same factors are also associated with disease relapse and may be combined in a fashion to estimate the prospects for cancer control in the individual patient and in homogeneous patient groups. Grouping patients according to the risk for and site of disease recurrence may be instrumental in the development of clinical trials that assess therapeutic approaches in appropriate subsets of patients.