Enrico Martin

78 Chapter 4 higher in children diagnosed after 2005 (87.6% SE: 6.7% vs. 53.9% SE: 8.0%, p<0.05). On multivariate analysis of localized pediatric MPNST, NF1 status was the only patient- and tumor-specific variable independently associated with survival (HR: 2.98, 95% CI: 1.17-7.60, p<0.05, Figure 2 and 3 ). Additionally, patients presenting after 2005 were significantly associated with increased survival (HR: 0.20, 95% CI: 0.06-0.69, p<0.05), without demographical or overall treatment differences between these time periods. Surgical margins, the use of chemotherapy, radiotherapy, and any sequence of multimodal treatment were not significantly associated with survival in localized pediatric MPNST (all p>0.05). NF Margin Chemotherapy Period No NF1 NF1 R0 R1 R2/Biopsy NA No Yes 1989−2005 2006−2017 38 26 32 15 7 10 42 22 39 25 Reference 2.98 (1.17, 7.60) Reference 1.40 (0.45, 4.41) 2.07 (0.61, 6.97) 0.83 (0.25, 2.72) Reference 1.96 (0.82, 4.70) Reference 0.20 (0.06, 0.69) 0.02 0.56 0.24 0.75 0.13 0.01 Variable N Hazard ratio p 0.10.2 0.5 1 2 5 Figure 2 Cox proportional hazard model in localized pediatric MPNST Discussion In this large, nationwide, unselected group of MPNST, pediatric patients presented similarly compared to adult MPNST. In children, NF1 patients more commonly had large tumors, but were treated similarly compared to non-NF1 patients. In localized pediatric MPNST, only NF1 status was independently associated with poor survival. No treatment related factors were independently associated with survival. Also, patients presenting after 2005 were independently associated with increased survival.