RESUMEN
Purpose: This retrospective study sought to identify predictors of metastatic site failure (MSF) at new and/or original (present at diagnosis) sites in high-risk neuroblastoma patients. Methods and materials: Seventy-six high-risk neuroblastoma patients treated on four institutional prospective trials from 1997 to 2014 with induction chemotherapy, surgery, myeloablative chemotherapy, stem-cell rescue, and were eligible for consolidative primary and metastatic site (MS) radiotherapy were eligible for study inclusion. Computed-tomography and I-123 MIBG scans were used to assess disease response and Curie scores at diagnosis, post-induction, post-transplant, and treatment failure. Outcomes were described using the Kaplan-Meier estimator. Cox proportional hazards frailty (cphfR) and CPH regression (CPHr) were used to identify covariates predictive of MSF at a site identified either at diagnosis or later. Results: MSF occurred in 42 patients (55%). Consolidative MS RT was applied to 30 MSs in 10 patients. Original-MSF occurred in 146 of 383 (38%) non-irradiated and 18 of 30 (60%) irradiated MSs (p = 0.018). Original- MSF occurred in post-induction MIBG-avid MSs in 68 of 81 (84%) non-irradiated and 12 of 14 (85%) radiated MSs (p = 0.867). The median overall and progression-free survival rates were 61 months (95% CI 42.6-Not Reached) and 24.1 months (95% CI 16.5-38.7), respectively. Multivariate CPHr identified inability to undergo transplant (HR 32.4 95%CI 9.3-96.8, p < 0.001) and/or maintenance chemotherapy (HR 5.2, 95%CI 1.7-16.2, p = 0.005), and the presence of lung metastases at diagnosis (HR 4.4 95%CI 1.7-11.1, p = 0.002) as predictors of new MSF. The new MSF-free survival rate at 3 years was 25% and 87% in patients with and without high-risk factors. Conclusions: Incremental improvements in systemic therapy influence the patterns and type of metastatic site failure in neuroblastoma. Persistence of MIBG-avidity following induction chemotherapy and transplant at MSs increased the hazard for MSF.
