Analysis of diagnostic therapeutic strategies in prostate cancer. I. When should a pelvic lymphadenectomy for staging be performed?

A decision-tree analysis is presented to compare three strategies for patients with clinical stage-B prostate cancer: strategy I, curative treatment without performing a pelvic lymphadenectomy; strategy II, performing a pelvic lymphadenectomy before deciding on curative treatment; strategy III, noncurative treatment without performing a lymphadenectomy. Ten-year survival rates are used as a measure of comparison. Probability threshold analysis showed that patients with clinical stage-B tumours are candidates for local curative treatment. The net expected value of treatment without testing (expressed in 10-year survival) varied from 9.5% to 19.4%. Sensitivity analysis of the net expected value of lymphadenectomy revealed that the overall survival of these patients is not, or only very slightly, affected by performing a pelvic lymphadenectomy. Therefore, the decision to perform a lymphadenectomy depends on the burden (including the morbidity) to patients resulting from curative treatment, indicated by Cc, and the burden associated with pelvic lymphadenectomy, indicated by Cl. Performing a lymphadenectomy should be recommended if the value of the ratio Cc/Cl is estimated to be higher than 'the percentage of patients submitted to lymphadenectomy' divided by 'the percentage of patients for whom a curative treatment is avoided due to the information provided by lymphadenectomy'.