Predictors of surgical outcome in Wilms' tumor: a single-institution comparative experience.

Background: The merits of primary nephrectomy (PN) vs preoperative chemotherapy (PC) for patients with Wilms' tumor (WT) are much debated. Early data from the International Society of Pediatric Oncology suggested decreased intraoperative spillage but increased risk of local recurrence after PC. Patients with WT at our institution were managed with PC until 1996; subsequently, they underwent PN. This study compares these approaches as they affect tumor spillage, local recurrence, and survival.
Methods: Patients with WT diagnosed at the Hospital for Sick Children from 1985 to 2003 were reviewed.
Results: One hundred sixty patients were identified (114 PC and 46 PN). Tumor spill occurred in 6 (5.3%) of 114 PC and 2 (4.3%) of 46 PN patients. Tumor inhomogeneity, tumor size, and inferior vena cava compression/clot at diagnosis did not affect incidence of spill. Of 6 PC patients with surgical spill, 1 (17%) had significant tumor shrinkage, compared with 87 (81%) of 108 without spill (P < .001). Preoperative chemotherapy and PN had equal rates of surgical complications. Preoperative chemotherapy resulted in 12 (10.5%) of 114 local recurrences vs 5 (10.8%) of 46 with PN. Event-free survival and overall survival were 80% and 92% for PC at 129 months vs 85% and 96% for PN at 61 months.
Conclusions: Preoperative chemotherapy and PN are equally effective in the treatment of WT with no difference in tumor spillage. Failure of the tumor to shrink in size with PC was significantly associated with an increase in tumor spillage and would suggest that a more cautious surgical approach be undertaken in these cases.