[Purpose]
Neuroblastoma (NB) is known to exhibit a wide range of clinical behavior and genome-based molecular signature of tumor specimens has been considered as one of the useful risk markers to define high-risk NB. In this study, we collected follow-up data for more than 2,200 NB cases diagnosed in 122 hospitals from 1995 to 2014 according to the common INRG queries as a cooperative effort of the Japan Neuroblastoma Study Group (JNBSG). Array CGH (Agilent microarray system) was conducted with 537 cases (INSS stage 1:73; stage 2:46; stage 4s:41; stage 3:96; stage 4:281) and clinical relevance of each genome signature was assessed with the updated follow-up data. Detailed genome analysis of the 50 patients with high-risk NB enrolled in the JNBSG phase II study (JN-H-07) was also included.
[Results]
Tumors were sub-grouped into three genomic groups (GGs): silent (S), partial (P) and whole (W), which were further segregated by MYCN amplification (a;MYCN-amp), 1p-loss, 11q-loss and 17q-gain. Concordant with our previous results, Ws (s:MYCN single copy) and Ss showed favorable prognosis with 8 year survival rates (8y-SR) of 90% and 87%, respectively, whereas Pa and Wa exhibited poor prognosis (8y-SR:38% and 25%, respectively). As compared with the previous follow-up data in 2008 with same GGs, survival rates have been progressed especially in those with MYCN-amplification (cf. Wa from 0% to 45%). On the other hand, genome analysis of unfavorable cases (death within 2 years) in JN-H-07 suggested that additional chromosomal aberrations such as 12q amplification and 2p gain as well as mutations including ATRX, ARID1 and MYCN seems to affect the prognosis of these patients.
[Conclusion]
Our results suggested that array CGH-based GG subgroups can be useful for the tumor risk classification. Correlation with other clinical markers and additional genomic markers such as gene mutations and expression signature will be also discussed.