Enrico Martin

38 Chapter 2 most authors believe that survival is better in these patients. All but intracranial tumors had a worse overall survival. In literature, 5-year overall survival range from 39% to 72%. 19,32 Although limb salvage treatment is possible, amputations are still not uncommon for large and deep tumors. 1 Pediatric MPNST Malignant peripheral nerve sheath tumors in pediatric patients have been described previously. 32,41–45 5-year overall survival in children varies between 34.6% and 51%. 32,41,42,44,45 No institutional study has yet been able to find a difference in survival between pediatric and adult tumors. However, in two studies including only pediatric cases, a prolonged survival was seen in younger children compared to adults. 32,44 The SEER data suggests that pediatric patients tend to have a better survival in general, possibly by controlling for many risk factors previously shown to influence survival. Strengths and limitations This study has several, registry associated, limitations. Many data of interest were missing for instance data about tumor grade, tumor size, extent of resection, and site of origin. All missing groups were examined as separate entities and were associated with significantly worse outcomes. This may have resulted in over- or underrepresentation of certain variables. Also, it was not possible to conduct separate analyses for patients with NF1. While many studies found that NF1 patients show poorer survival, 4,10,22,26 more recent studies did not find this difference. 9,11,14 In a meta-analysis by Kolberg et al., NF1 negatively affected survival in studies published before 2000, but significance was lost in data after 2000. 15 Not only better surveillance may have had its impact on this difference, NF1 patients tend to present with larger tumors more frequently originating from trunk sites, both factors associated with worse survival. 10 11.2% of all patients did not receive cancer-directed surgery, which mainly includes patients that were diagnosed at autopsy, but possibly a small heterogeneous group including clinical diagnoses as well. The latter may impede the interpretation of this group of patients. The SEER tumor grading system is also not completely comparable to WHO grading, which may make comparisons to other studies more difficult. Unfortunately, the registry does not contain any information on recurrence, progression-free survival; mode and dosage of radiotherapy are not registered either, nor is the indication of its use. This makes the interpretation of the impact that radiotherapy has, adjacent to surgery, difficult. It is possible that most patients receiving radiotherapy had positive margins, another variable that is not available in the SEER registry, which could skew data on survival. Furthermore, the use and regimen of chemotherapy cannot be extracted either. Nevertheless, the effect of chemotherapy is still subject of debate. 9,14,22,32,46,47 Despite these limitations, the SEER database allows for investigation of small subpopulations of rare tumors as MPNSTs, such as pediatric populations and rare tumor sites. Tumors that arise in different sites may be etiologically different from one another as location seems to be of great influence. Thoroughly examining clinical differences between different