Key points are not available for this paper at this time.
BACKGROUND: Accurate detection of macroscopic residual disease after neuroblastoma resection is essential for postoperative treatment decisions, including radiotherapy. Standard imaging follow-up is often performed at a time point when postoperative and therapy-related changes may hamper reliable differentiation between residual disease and reactive findings. OBJECTIVE: To assess inter-reader agreement and the diagnostic accuracy of early postoperative magnetic resonance imaging (MRI) for detecting residual disease after neuroblastoma surgery. METHODS: This retrospective single-center study included patients with histologically confirmed neuroblastic tumors who underwent surgical resection at a reference center and received standardized early postoperative MRI with adequate preoperative imaging. Two independent pediatric radiologists, blinded to all clinical data, assessed all examinations. A hierarchical multimodal reference standard was established based on surgical reports including expert consensus between a senior pediatric radiologist and a pediatric surgeon, and follow-up imaging (median follow-up 33 months). RESULTS: Thirty-nine patients (median age 46 months), predominantly with International Neuroblastoma Staging System (INSS) stage 4 neuroblastoma, were included; all patients had at least one image-defined risk factor. MRI was performed at a mean of 8±5 days after surgery. Residual disease was identified in 14 patients and confirmed by the reference standard. Five residual diseases were expected by the surgeons (median volume 8 ml), whereas nine were unexpected and small (median volume 1 ml). Diagnostic accuracy were 95% and 90% for the two readers, respectively, with substantial inter-reader agreement (Cohen's κ=0.76). Early postoperative MRI findings led to clinically relevant management adaptations in selected cases. CONCLUSION: Early postoperative MRI demonstrates high diagnostic accuracy and substantial inter-reader agreement in detecting macroscopic residual disease, providing a robust baseline for further assessment.
Chaika et al. (Thu,) studied this question.