Prognostic factors for children with recurrent Wilms' tumor: results from the Second and Third National Wilms' Tumor Study.

The characteristics of 367 stage I-IV National Wilms' Tumor Study (NWTS) children who relapsed after initial treatment for unilateral disease in the second and third NWTS trials (NWTS-2 and -3) were analyzed to identify features predictive of survival. Although modifications in initial therapy resulted in a lower rate of first relapse in these two studies compared with NWTS-1, all previously identified prognostic factors after relapse remained statistically significant predictors of survival. Tumor histology, length of initial remission, initial therapy with two v three drugs, and site of relapse each were independently predictive of postrelapse survival. The 3-year postrelapse survival for all 367 patients was 30% +/- 3%; however, subgroups classified by these prognostic factors were identified that had 3-year postrelapse survival rates of greater than 40%. These subgroups included patients who had tumors of favorable histology (FH) that recurred (1) only in the lungs, (2) in the abdomen when radiotherapy (RT) was not initially given, or (3) that were originally stage I, (4) that were originally treated with only two drugs, or (5) that recurred 12 months or more after diagnosis. These results were achieved with the use of standard treatments, ie, surgery, RT, and chemotherapy using dactinomycin (AMD), vincristine (VCR), and Adriamycin [( ADR] doxorubicin; Adria Laboratories, Columbus, OH). It is suggested that patients in these groups might be managed with aggressive use of conventional therapies until newer chemotherapeutic agents and drug combinations are shown to be more effective. Patients with adverse prognostic features at relapse have a poor survival expectancy with standard measures. Salvage attempts for these patients are better based on experimental approaches.