Combined analysis of PHOX2B at two time points and its value for further risk stratification in high‐risk neuroblastoma

Abstract

Background Risk stratification of high‐risk neuroblastoma (NB) is crucial for exploring treatments. This study aimed to explore the value of minimal residual disease (MRD) based on PHOX2B levels for further risk stratification in high‐risk NB. Methods The expression of PHOX2B was monitored at two time points (after two and six cycles of induction chemotherapy, TP1 and TP2, respectively) by real‐time polymerase chain reaction (RT‐PCR). The clinical characteristics between groups and survival rates were analyzed. Results The study included 151 high‐risk patients. Positive expression of PHOX2B at diagnosis was seen in 129 cases. PHOX2B was mainly expressed in patients with high lactate dehydrogenase (LDH) and neuron‐specific enolase (NSE) levels (p < .001), bone marrow metastasis (p < .001), more than three metastatic organs (p < .001), 11q23 loss of heterozygosity (LOH) (p = .007), and when more events occurred (p = .012). The 4‐year EFS rate was significantly lower in patients with positive PHOX2B expression compared to the negative group at diagnosis (32.9% ± 6.2% vs. 74.5% ± 10.1%, p = .005). We stratified the 151 patients into three MRD risk groups: low high‐risk (low‐HR), with TP1 less than 10−4 and TP2 less than 10−4; ultra‐HR, with TP1 greater than or equal to 10−2 or TP2 greater than or equal to 10−4, and others classified as intermediate‐HR. Patients in ultra‐HR had the worst survival rate compared with other two groups (p = .02). In a multivariate model, MRD risk stratification based on PHOX2B levels at TP1 and TP2 was an independent prognostic factor for high‐risk patients (p = .001). Patients in ultra‐HR were associated with 11q23 LOH (p < .001), more than three organs of metastasis (p = .005), bone marrow metastasis (p < .001), and occurrence of more events (p = .009). Conclusions MRD risk stratification based on PHOX2B levels at two time points (after two and six cycles of induction chemotherapy) provided a stratification system for high‐risk NB, which successfully predicted treatment outcomes. Our results present an effective method for further stratification of high‐risk NB.

Machine‐learning radiomics to predict bone marrow metastasis of neuroblastoma using magnetic resonance imaging

“…The amplification of the oncogene MYCN is widely considered a deteriorative biomarker in neuroblastoma. The mutations and/or overexpressions of some other genes such as PHOX2B, ALK, LMO1, HACE1, PRAF2, and LIN28B are also representative of poor prognosis [ 34 , 35 , 36 ], while quite the opposite is observed when ADRB2 over‐expresses [ 37 ]. Loh et al [ 38 ] managed to enrich and characterize intact floating cells in peripheral circulation using a label‐free and size‐based separation method.…”