METHODS: MIBG scans from patients with MIBG avid, INSS stage 4 neuroblastoma enrolled on SIOP-NB were evaluated post-induction (n=330), prior to transplant. Scans were evaluated in 10 different anatomic regions by 2 reviewers, including 9 skeletal segments and 1 soft issue region. Each region was scored 0-3 based upon disease extent, and a cumulative consensus score generated. Optimal cut-points from post-induction scans were determined using the Youden index, with EFS and OS determined. Post-induction CS from patients enrolled on COG A3973 (n= 237) were used for comparison. Patients received busulfan-melphalan (BuMel) in SIOP-NB, and carboplatin-etoposide-melpahlan (CEM) in COG A3973, as transplant conditioning.
RESULTS: The optimal CS cut-point post-induction (pre-transplant) was 2 in both SIOP-NB and COG A3973, with a post-induction CS >2 associated with inferior outcomes in each study. In SIOP-NB, 5-year EFS were 39.2±4.7% vs 16.4±4.2% for patients with post-induction CS ≤2 vs >2 (p<0.001). In comparison, 5-year EFS were 42%±5.8% vs 10.5%±10.0% (p<0.001) for patients with post-induction CS ≤2 vs > 2 in COG A3973. A post-induction CS >2 carried prognostic significance for both MYCN amplified and MYCN non-amplified tumors, in both SIOP-NB and COG A3973. In particular, post-induction CS > 2 were associated with extremely poor outcomes for patients with MYCN amplified disease, with a 5-year EFS 13.8%±12% in SIOP-NB. The post-induction CS maintained independent statistical significance in Cox models when compared to standard predictive covariates age and MYCN in both trials.
CONCLUSION: The prognostic significance of post-induction CS has now been validated in an independent cohort of patients, with a post-induction CS >2 associated with inferior outcome in two large cooperative group trials, SIOPEN-HR-NBL1 and COG A3973. MIBG scoring should be incorporated into pre-transplant evaluations in high risk NBL, and patients with CS >2 considered for alternative regimen.