Predictors of differential response to induction therapy in high-risk neuroblastoma: A report from the Children's Oncology Group (COG)

Background: Induction chemotherapy plays an important role in the management of patients with high-risk neuroblastoma. Predictors of response to induction therapy are largely lacking. We sought to describe clinical and biological features associated with induction response. Methods: Patients from four consecutive COG high-risk trials were included. Response was evaluated by the 1993 International Neuroblastoma Response Criteria. The primary end-point was end-induction partial response (PR) or better. Univariate analyses were performed to compare response as a function of clinical or biologic predictors. A multivariate logistic regression model using significant predictors from univariate analyses was constructed to model PR or better. Results: The analytic cohort included 1242 patients. End-induction response ≥PR was significantly associated with higher event-free and overall survival. Baseline factors associated with ≥PR included age <18 months (87.4% with ≥PR vs. 78.7% if older; p = 0.0103), International Neuroblastoma Staging System non-stage 4 (89.0% vs. 78.4% if stage 4; p = 0.0016), MYCN amplification (85.5% vs. 77.1% if non-amplified; p = 0.0006), 1p loss of heterozygosity (LOH; 85.6% vs. 76.0% if no LOH; p = 0.0085), no 11q LOH (84.8% vs. 70.9% if 11q LOH; p = 0.0004) and high mitosis-karyorrhexis index (MKI; 84.5% vs. 77.5% if low-intermediate MKI; p = 0.0098). On multivariable analysis (n = 407), the absence of 11q LOH was the only factor that remained significantly associated with ≥PR (odds ratio: 1.962 vs. 11q LOH; 95% confidence interval 1.104–3.487; p = 0.0216). Conclusions: Improved end-induction response in high-risk neuroblastoma is associated with longer survival. Patients with 11q LOH are less likely to respond to induction therapies and should be prioritised for novel approaches in future trials.