Poster Presentation Advances in Neuroblastoma Research Congress 2016

Survival tree analysis of an independent cohort reveals risk factors as proposed in the INRG system (#162)

Ruth Volland , Thorsten Simon , Frank Berthold , Barbara Hero

Background:

In 2009 Cohn et al. proposed the International Neuroblastoma Risk Group (INRG) classification system based on a risk factor analysis for event free survival (EFS) of a large international cohort. The aim of our work was to evaluate this classification system in an independent representative national cohort with respect to event free survival and overall survival (OS).

Methods:

Risk factors of 1,063 neuroblastoma patients (age 0-21 years) registered in two consecutive national neuroblastoma trials between 01.01.2003 and 31.12.2010, were compared to the published INRG data. Finally, the impact on outcome was analysed by survival tree regression analyses for the endpoints EFS and OS.

Results:

Compared to the INRG cohort, patients of the national cohort were more often diagnosed in infancy (p=0.003) and with localized stages (p=0.004), showed less often elevated LDH (p<0.001), normal serum ferritin (p<0.001), differentiated tumors (p<0.001) and tumors with low/intermediate MKI (p<0.001). Regarding genetic factors, fewer patients had 1p (p<0.001) and 17q aberrations (p= 0.026). In contrast no differences were seen for INPC diagnostic category, 11q and MYCN status.

As previously published for the INRG cohort, survival tree analysis revealed INSS stage as the most important prognostic risk factor in the national cohort for EFS but also for OS. MYCN, age and serum ferritin were further prognostic risk factors for EFS, whereas for OS MYCN, age and LDH were of prognostic relevance.

Conclusion:

Application of survival tree analysis on an independent and differently composed representative neuroblastoma cohort resulted in similar selection and ranking of risk factor as in the published INRG survival tree.