Evaluation of circulating tumor DNA as a biomarker for early relapse detection and molecular profiling in patients with high-risk neuroblastoma

Neuroblastoma (NBL) is the most common extracranial solid tumor of childhood, and nearly half of patients are diagnosed with high-risk disease. Patients with high-risk NBL often present with metastasis and frequently develop drug resistance; the 5-year survival rate is about 50%. Standard diagnostic and surveillance modalities not only involve radiation exposure and possible procedural sedation in pediatric patients, but also lack the ability to effectively monitor genetic evolution and heterogeneity. Circulating tumor DNA (ctDNA) has emerged as a minimally invasive biomarker capable of detecting microscopic disease, offering potential advantages for earlier relapse identification, reduced radiation exposure, and improved accessibility. We conducted a retrospective chart review of 13 patients with high-risk NBL, capturing salient clinical, molecular, diagnostic, and therapeutic information for each patient. Using this data, clinical evaluations of treatment response were recorded according to the revised International Neuroblastoma Response Criteria (INRC). Plasma samples taken in tandem with these evaluations were sequenced, and the ctDNA results were used to examine recurrence-associated ctDNA positivity and explore shared genomic alterations across patients. Additionally, variant information collected from tumor sequencing - performed as part of routine clinical care - was compared to the variants found in ctDNA sequencing. We found that 11 relapse-associated time points had preceding ctDNA measurements; among these, 72% were ctDNA-positive prior to radiographic relapse. Samples positive for ctDNA, with higher variant allele frequency (VAF), and with higher tumor fraction (TF) were significantly associated with stable or progressive disease, whereas complete response correlated strongly with samples negative for ctDNA. Progressive disease was associated with increasing VAF and TF, while partial response was associated with decreasing VAF. Six patients had concordant sequencing results; however, ctDNA failed to detect certain key alterations, including HRAS K117N and HGF amplification. Conversely, ctDNA detected several variants not observed in tumor sequencing, including alterations in tumor suppressor genes such as TP53 and CHEK2, as well as a rare germline SMARCA4 Variant of Uncertain Significance (VUS). ctDNA demonstrates promising utility as a non-invasive tool for monitoring treatment response and predicting relapse in high-risk NBL while also showing potential to reveal clinically relevant genetic alterations. Although our interpretation is limited by small sample size, heterogeneous sampling, and analytical bias, these findings support further prospective evaluation of ctDNA as a complementary biomarker to current imaging-based surveillance strategies.