ORIGINAL PAPER

Are pediatric brain tumors on the rise in the USA? Significantincidence and survival findings from the SEERdatabase analysis

Swetal Patel & Amit Bhatnagar & Christopher Wear & Stephen Osiro & Abigail Gabriel &David Kimball & Alana John & Paul J. Fields & R. Shane Tubbs & Marios Loukas

Received: 29 September 2013 /Accepted: 9 October 2013 /Published online: 27 October 2013# Springer-Verlag Berlin Heidelberg 2013

AbstractIntroduction Central nervous system tumors are the secondmost common form of cancer in children between the ages of1 and 19 years. We aimed to provide the most recent data onthe incidence and survival of these tumors in the USA and toassess the literature.Methods Frequency, rates, and survival sessions werecalculated using the November 2008 submission for the USSurveillance Epidemiology and End Results Program. Data

were collected and analyzed for children and adolescents aged1 to 19 years with primary brain tumors.Results We found that the incidence rate of all pediatric braintumors has been on a gradual but steady increase from 1973 to2008 (p <0.001). The average annual increase was 1.37 %.Our survival analysis of the individual tumors revealed thatthe 5-year overall survival for children diagnosed between1974 and 1978 with medulloblastoma was 43.7 %. However,this increased to 62.8 % for children diagnosed between 1999and 2003. A similar survival trend was also observed when allthe other pediatric brain cancer histologies were collectivelyanalyzed (p <0.001).Conclusions From our study, we can conclude that contrary toprevious reports indicating a plateau in the incidence rates ofpediatric brain tumors since the mid-1980s, there has been anincrease from 1973 to 2008. Potential causes includeenvironmental carcinogens, but more research is needed toinvestigate the factors behind this sustained rise in incidenceover the years.

Keywords Pediatric . Brain tumors . SEER . Incidence .

Survival

Introduction

Pediatric central nervous system (CNS) tumors are the secondmost common cancers in childhood after leukemia, andaccount for almost 20 % of all malignancies in this age group[9]. They also comprise the majority of pediatric cancer deaths[14]. In 1973, these tumors were diagnosed in the USA at arate of 13.5 per 100,000/year [3]. According to Keirian et al.[13] and Bleyer [3], the incidence rates increased to 16.0 casesper 100,000/year in 1994, and remained stable as of early2001[3, 13]. Per year, worldwide data shows that there are

Novelty and impact There has been a steady increase in the incidence ofbrain tumors in the USA among children of all ages from 1973 to 2008contrary to previous reports indicating that it had plateaued in the mid-1980s. Environmental carcinogens could be a possible cause, but furtherresearch is needed to investigate the factors behind this sustained increaseover the years.

S. Patel :A. Bhatnagar :C. Wear : S. Osiro :A. Gabriel :D. Kimball :A. John :M. Loukas (*)Department of Anatomical Sciences, School of Medicine,St. George’s University, St. George’s, Grenadae-mail: mloukas@sgu.edu

S. OsiroDepartment of Medicine, Woodhull Medical and Mental HealthCenter, Brooklyn, NY 11206, USA

A. JohnDepartment of Internal Medicine, SUNYDownstate Medical Center,Brooklyn, NY, USA

P. J. FieldsWindward Islands Research and Education Foundation, St. George’s,Grenada

R. S. TubbsPediatric Neurosurgery, Children’s Hospital, Birmingham, AL, USA

M. LoukasDepartment of Anatomy, Medical School, University of Varmia andMazuria, Olsztyn, Poland

Childs Nerv Syst (2014) 30:147–154DOI 10.1007/s00381-013-2307-1

approximately 11 new cases of CNS tumors per 100,000persons under the age of 19 [3].

Astrocytomas account for just over half of all diagnosedCNS tumors in children [13]. Medulloblastoma, which is atype of primitive neuroectodermal tumors (PNET), comprises21 % of the cases, while approximately 15 % of these tumorsare classified as assorted gliomas. The remainder constitutesependymomas and miscellaneous tumors, including germ celltumors [3, 13, 28]. It is important to note that pediatric CNStumors receive a great deal of attention due to their highprevalence, the vulnerable population affected, and the factthat most of them are highly responsive to treatments [16].

As described by Bleyer [3], existing research stronglyindicates that spontaneous mutation is the most common causeof pediatric brain tumors. However, patients with geneticconditions such as neurofibromatosis, Gorlin syndrome, Li–Fraumeni syndrome, and others have also been shown to haveup to 70-fold increased risk of developing these tumors [3, 30].Gender is not a risk factor for developing CNS tumors inchildren, except for medulloblastoma, which has a predilectionfor males [20]. Other possible etiologies of these cancers includecellular phone use, gene methylation, ionizing radiation, viruses,and amyriad of potential environmental toxins [4, 10, 11, 16, 22,31]. Apart from ionizing radiation, no conclusive evidence hasbeen found linking the other environmental exposures with thedevelopment of CNS cancers.

Patients with CNS tumors can present with non-specificneurological symptoms such as headache, nausea, vomiting,visual loss, and seizures. Infratentorial tumors, which are morecommon in young children, often present with dizziness,nystagmus, and motor weakness [15]. Diagnoses are usuallyconfirmed with either CT or MRI scan or biopsy [9].

Isolated studies show that the survival for pediatric patientswith CNS tumors has improved considerably over the past35 years, with an overall 5-year survival rate of approximately65 % [9]. This can probably be attributed to improvedtreatment modalities. For instance, advanced radiotherapymethods have allowed not only for a localized radiation field,but also diminished long-term side effects such as hearing lossand neurocognitive deficits [9]. In addition, children nowaccount for about 25 % of all patients enrolled in clinicaltrials, and up to 75 % of pediatric patients receiving cancertreatments in the USA are enrolled in a nationally supportedclinical trial [20].

Although CNS tumors have been extensively studied overthe last two decades, most research has focused either on onetumor histology or treatment modalities. As a result, there is apaucity of recent data on the incidence and survival rates ofthese tumors in children. Therefore, we proposed to examinethese parameters using the Surveillance Epidemiology andEnd Results (SEER) database. Our aim was to analyze thechange in survival among these tumors since 1973, and also toassess whether or not the incidence rates have been on a

plateau since the 1980s, as has previously been reported inthe literature [3, 10, 17, 18, 25, 33].

Methods

Internal Classification of Diseases (ICD)-0-3 topography codesC71.0 to C71.9 and ages from 1 to 19 years were selected toquery the National Cancer Institute's SEER 17 Registry databaseand generate the number of pediatric brain tumors diagnosedfrom 1973 to 2008 [27]. Consistent with the National CancerInstitute's definitions, the brain topography codes were used toselect patients aged 1 to 19 yearswith primary tumors in the brain[1]. Other CNS tumors originating in the meninges, spinal cord,and cranial nerves were excluded [1]. Demographic, age atdiagnosis, year of diagnosis, and histology data were collectedfor this pediatric population [13, 30]. Both malignant and non-malignant cases were selected for inclusion to allow for low-grade gliomas, which are often considered benign [13, 30].Incidence was also obtained from SEER 17 Registry databaseusing the age-adjusted rate statistic and US population data fromthe 2000 census [27].

The Joinpoint Regression program (version 3.5.2) was used toanalyze incidence rates retrieved from the SEER Stat software[12]. Trends in the incidence of pediatric brain cancer for theyears 1973 to 2008 and the annual percentage change werecalculated. A log linear model was used, and 0–3 joinpoints weretried for a plot of the year of diagnosis against age-adjustedincidence rates. Using a 95 % confidence interval, the softwareselected the best fit line(s) with the fewest join points resulting ina statistically significant change in rate.

The many different types of pediatric brain tumors werecategorized by related groups to facilitate analysis. Pilocyticastrocytomas, one of the two types of low-grade gliomas(WHO grade I), and medulloblastomas were consideredseparately because they are the two most frequent braintumors of early childhood [13]. The group “OtherAstrocytomas” include diffuse, anaplastic, unique, and nototherwise specified (NOS) astrocytomas, as well asglioblastoma (WHO grade IV). Oligodendrogliomas, mixed,and NOS gliomas are in the “Mixed Glioma” category. Allependymomas and variants are listed under “Ependymoma.”The “Other Brain” category groups all other primary braintumor histologies and includes choroid plexus tumors, germcell tumors, nerve sheath tumors, craniopharyngiomas,hemangiomas, and chordomas. These groupings areconsistent with prior pediatric brain tumor studies [2, 10, 13].

Mixing tumors with different pathological grades would beproblematic for some analyses, especially survival. However,the groupings are an effort to highlight the most commonlyoccurring pediatric brain tumor types and to track the changesin their relative frequency within different pediatric age sub-groups. Etiology and likelihood, as opposed to clinical

148 Childs Nerv Syst (2014) 30:147–154

outlook, were the determining factors for devising the tumorcategories; the cancers in the “Other Brain” group all haverelatively low incidence rates and were of diverse origins andwere not the focus of this study.

Overall survival percentages were calculated for allpediatric brain tumors as well as for the aforementionedhistology sub-groups using the SEER 17 registry databaseand a 95 % confidence interval [27]. Cases with multipleprimaries, and those without follow-up information wereexcluded. Death from any cause was used as the endpoint ofsurvival. Two-tailed t tests were used to determine statisticalsignificance.

Results

Between 1973 and 2008, 10,668 patients aged 1 to 19 yearswith primary brain tumors were recorded in the SEERdatabase. There were more brain tumors in younger childrenthan adolescents. We found 2,835 cases in the 1–4-year-oldage group; 3,060 in the 5–9-year-old age group; and 2,151 inthe 15–19-year-old group (Fig. 1). The frequency ofindividual tumors also varied with age. Astrocytomasaccounted for 36.5 % of all brain tumors in the 1–4 year oldsand 51.2% of tumors in the 15–19 year age group. In contrast,PNET/medulloblastomas and ependymomas are tumors ofearly childhood. For instance, medulloblastomas represent27.9 % of all brain tumors in 1–4 year olds, but only 9.9 %in 15–19 year olds. The percentage of pilocytic astrocytomasdiagnosed across all age groups remained relatively stableover time.

Until 1983, the overall incidence rate of these tumors wasrelatively constant. However, there was an increase between1984 and 1986 before it plateaued until 2003 (Fig. 2).Nevertheless, when we did a regression of all pediatric braintumors from 1973–2008, we found that there has been asteady increase in the incidence rate during this period, withan annual percentage change of 1.37 % (p <0.001) (Fig. 3).The incidence was 19.07 cases per million in 1973, butincreased to 35.15 cases per million in 2008. A similar trendwas also observed for the majority of histological sub-groups,with pilocytic astrocytomas and mixed gliomas showing thegreatest increase during this period (Table 1).

Our analysis also showed differences in raw incidence ratesamong the pediatric demographics. For instance, the incidencerate of brain tumors in whites (27.73 cases per million) wasgreater than in blacks (22.68 cases per million). There werenear similarities in the incidence rates between blacks and theother races recorded by SEER, which include AmericanIndians, Alaska natives, and Asian/Pacific Islanders (Table 1).It is important to note that when the incidence rate for eachrace was compared against the others using a t test, none of thedifferences were statistically significant.

The survival among brain tumors also varies greatly. Wefound pilocytic astrocytomas to be the histology with the mostfavorable prognosis. Pediatric patients diagnosed withpilocytic astrocytomas between 1974 and 1978 had a 5-yearsurvival rate of 81.1 %. However, this percentage rose to93.0 % for children diagnosed between 1999 and 2003(Fig. 4). The second most common tumor type,medulloblastoma, also showed an increased trend in the 5-year overall survival over the same time period. The mixedgliomas, which includes high-grade histologies, had a 5-yearoverall survival of 53.8 % from 1999–2003. The survivals forother CNS tumors are summarized in (Table 2).

Discussion

Based on our study, brain cancer diagnoses among allpediatric ages have increased at a similar rate since 1973.The annual average percentage change (AAPC) for each agesub-group ranged from 0.99 % in 5–9 year olds, to 1.69 % in10–14 year olds. However, we found a much higher incidencerate (IR) in younger children. Between 1973 and 2008, thesetumors were diagnosed at an average rate of 41.0 cases permillion per year in 1–4 year olds and 29.2 cases per million peryear in 15–19 year olds. This finding confirms previousreports, which showed that the IR of pediatric brain tumorsis inversely proportional to age [13, 20]. It is postulated that asbrains age, their plasticity and cell division decreases. As aresult, fewer cancers arise from these dormant cells [13].

Previous SEER studies have reported that the incidencerates of pediatric brain tumors across all ages was fairlyconstant until 1984 at which time there was a sudden increase[2, 3, 5, 25]. This rise lasted for 2 years. In 1986, the IRplateaued again and was thought to remain stable (Fig. 2). Ouranalysis of the SEER data from 1973–2003 revealed a similarfinding. Most authors have hypothesized that the increasedavailability of MRI as a diagnostic tool and/or changes toneurosurgical practices, including stereotactic biopsies led tothe rise in incidence seen around 1984 [3, 10, 17, 18, 25, 33].However, some have countered that the observed increase inincidence could not be due to improved diagnostic tools alone.They argue that had there been no true increase in pediatricbrain tumors, the rates would have returned toward the pre-1984 levels since the improved diagnostic tools only allowcases to be identified at an earlier juncture [3, 24].

Another theory is that the changes in the histologicalclassification of “benign” and “malignant” tumors, whichwere adopted in 1985, contributed to the observed increase[25]. This is because tumors which would have beenpreviously identified by pathologists as “benign,” wereinstead being diagnosed as malignant neoplasms [25].Although this could be partially true, such changes wouldhave only increased the incidence of high-grade gliomas,

Childs Nerv Syst (2014) 30:147–154 149

since low-grade tumors were unrecorded by SEER at the time.As it occurred, there was a steady increase in pilocyticastrocytomas (low-grade gliomas) during the period, with an

overall AAPC of 13.49 (p value<0.001). Moreover, changesin cancer classification codes have since been controlled forby the SEER*Stat program [27].

Fig. 1 A bar graph showing the percentages of tumor histologies in thefour pediatric age sub-groups. Pilocytic astrocytomas, embryonal/primitive/medulloblastomas, and ependymomas are consideredindividually. The group, “Other Astrocytomas,” includes diffuse,anaplastic, unique, and not otherwise specified (NOS) astrocytomas, aswell as glioblastoma (WHO grade IV glioma). Oligodendrogliomas,

mixed, and NOS gliomas are in the “Mixed Glioma” category. The“Other Brain” category includes all the remaining primary brain tumorhistologies such as choroid plexus tumors, germ cell tumors, nerve sheathtumors, craniopharyngiomas, hemangiomas and chordomas. The count isthe total number of cases in the SEER database for that age group between1973 and 2008

Fig. 2 Joinpoint analysis of the incidence of all pediatric brain tumors diagnosed between 1973 and 2003. A model with two joinpoints was selected asthe best fit and shows an overall increase in incidence in 1984 before the “plateau”

150 Childs Nerv Syst (2014) 30:147–154

Contrary to prior studies showing that the incidence rates ofpediatric brain tumors rose around 1984 and has since beenstable, our joinpoint analysis showed that there has been asignificant increase since 1973 and that the rates have notstabilized (p <0.001) (Fig. 3). We found the overall averageannual percentage increase in the incidence rates of thesetumors to be 1.37 % per year (p <0.001). This increase wasobserved with virtually all tumor types, except formedulloblastomas and astrocytomas NOS. Medulloblastomasonly had a slight decrease in incidence rates over the years (0.4cases per million), while astrocytomas NOS had a largedecrease, ostensibly due to improvements in ICD classification[3, 23].

To the best of our knowledge, there are no reports indicatingthat the incidence rates of pediatric tumors have been on the rise

in the USA over the last two decades. The possible explanationis that by using the most recent SEER database available [27],we were able to observe for the first time a more accuratepicture of the incidence rates over the years. In addition, studiesof the brain tumors in the past decade have almost exclusivelyfocused on specific histologies or treatment modalities [6, 14,16, 19, 21, 32]. As a result, our theory on this increase iscurrently speculative. Propagation of the potential risk factorsfor brain cancers such as ionizing radiations, gene-environmentinteractions, electromagnetic fields, N -nitroso-compounds,parental occupational exposure, pesticides, viruses, hormones,and inherited disorders could be the most likely cause [3, 8, 11,26, 29, 31]. It is important to note that so far, ionizing radiationis the only environmental factor proven to cause brain cancer inhumans [11, 33]. The inherited disorders are rare, however, and

Fig. 3 Joinpoint analysis of the incidence of all pediatric brain tumors diagnosed between 1973 and 2008. A model with 0 joinpoints was selected as thebest fit. This analysis shows that tumor incidence has increased annually by an average of 1.37 % over the period (p <0.001)

Table 1 Average annual incidence rates for pediatric brain tumors from 1973 to 2008

All Race Sex

[Races and Sexes] White Black Other Male Female

IR Change N IR N IR N IR N IR N IR

Pilocytic astrocytoma 4.61 +7.36 906 4.88 115 3.48 80 3.36 573 4.62 545 4.6

PNET/MB 5.01 −0.39 987 5.33 116 3.49 111 4.64 774 6.25 442 3.73

Other astrocytoma 7.62 −1.38 1,522 8.17 216 6.57 102 4.35 1,001 8.08 846 7.14

Mixed glioma 5.5 +7.22 1,037 5.58 194 5.89 91 3.86 733 5.92 599 5.06

Ependymoma 1.88 +0.55 357 1.91 63 1.88 36 1.49 259 2.07 200 1.68

Other Brain 1.84 +2.71 347 1.86 45 1.37 48 2.06 232 1.86 214 1.81

Total 26.46 27.73 22.68 19.76 28.8 24.02

Rates are per 1,000,000 and age-adjusted to the 2000 US Std Population (19 age groups—Census P25-1130) standard. Change is the difference inincidence rates from 1973 to 2008. Other race category includes American Indian, Alaskan Native, Asian and Pacific Islander

IR incidence rates, N total number of tumor cases in a given demographic from 1973 to 2008

Childs Nerv Syst (2014) 30:147–154 151

the majority of the increase could be predominantly due toenvironmental factors [6, 8, 10, 11, 31, 33].

Other potential brain tumor risk factors include race andgender [3]. Nevertheless, in our study, race was not a significantfactor in the incidence rates, and onlymedulloblastoma showeda large disparity with regard to gender (Table 1). This furtherimplicates environmental and molecular risk factors as thepossible etiologies of pediatric brain tumors. It is worth notingthat the SEER demographic data may be slightly skewedtoward minorities, since foreign-born and urban populationstend to be overrepresented due to their proximity to majorhealth research centers [7]. This might have caused the

incidences among the races to appear closer than they are inreality.

Several studies show that pediatric brain cancer prognosishas fairly improved over the last 30 years [3, 9, 15, 20]. OurSEER study confirms these findings. The 5-year overallsurvival rate for all histologies has risen from 19.0 % forpatients diagnosed between 1974 and 1978, to 24.6 %between 1984 and 1988, and finally to 28.6 % for 1999 to2003 (p value <0.01). Our analysis of the individual tumortypes also showed either an improvement in survival rates or aslight decrease from 1974 to 2003. For instance, the 5-yearoverall survival among pediatric patients diagnosed between

Fig. 4 Overall survival (OS) for pediatric patients diagnosed with pilocytic astrocytoma between 1974 and 2008. The survivals were analyzed by 5-yearintervals. Trend lines were added for the 1974–1978, 1984–1988, and 1994–1998 intervals to show the changes in survivals over time

Table 2 Five-year overall survival percentages for different tumor types grouped by year(s) of their diagnosis

1974–1978 1979–1983 1984–1988 1989–1993 1994–1998 1999–2003

OS (%) N OS (%) N OS (%) N OS (%) N OS (%) N OS (%) N

Pilocytic astrocytoma 81.1 37 84.8 66 84.6 130 89.7 263 94.3 445 93.0 964

PNET/MBα 43.7 199 53.7 188 59.9 212 54.8 329 63.1 441 62.8 791

Other astrocytomaβ 14.0 3,480 15.8 4,213 17.9 4,684 16.4 5,543 12.4 6,190 12.3 11,249

Mixed gliomaα 27.4 589 30.8 764 41.3 869 47.5 1,123 53.2 1,650 53.8 3,233

Ependymomaα 36.3 113 46.3 108 60.4 129 62.0 188 68.6 201 70.8 432

Other brainβ 26.6 445 20.1 336 18.9 515 12.4 1,030 19.5 1,373 25.9 1,935

Totalα 19.0 4.863 20.7 5,675 24.6 6,539 24.8 8,476 26.7 10,300 28.6

αPaired, 2-tailed t test for tumor type over time had a p value of <0.025βPaired, 2-tailed t test for tumor type overtime had a p value of >0.5

PNET primitive neuroectodermal tumor; MB medulloblastoma; OS observed 5-year overall survival rate; N total number of tumor cases in a given 5-year period

152 Childs Nerv Syst (2014) 30:147–154

1974 and 1978 with pilocytic astrocytoma was 81.1 %. Thispercentage rose to 93.0 % for children diagnosed between1999 and 2003. While the survival among other astrocytomasdecreased from 14.0 to 12.3 % over the same period, thischange in rates was not significant (p =0.50). We agree withprevious authors that the reason for this increased survivalamong the majority of brain tumors could be due toadvancements in imaging techniques, standardization oftreatments, and novel therapeutic modalities such as intensitymodulated radiotherapy [9, 15, 19, 20]. The decline insurvival among “Other Astrocytomas” and “Other Brain”tumors is probably due to their resistance to these therapies[13, 15, 20]. However, this is subject to further research.

Conclusions

The incidence rate of pediatric brain tumors has been onthe rise in the USA from 1973 to 2008 according to theSEER database, despite previous reports indicating thatit had stabilized in the mid-1980s. Further research isneeded to confirm whether environmental factors havecontributed to this increase. On the other hand, survivalstatistics has shown an overall improvement, a trendwhich can be attributed to the advancement in bothdiagnostic and treatment techniques.

Conflict of Interest None.

References

1. Barnholtz-Sloan J, Sloan A, Schwartz A (2007) Cancer of the brainand other central nervous system. In: Ries LAG, Young JL, Keel GE,Eisner MP, Lin YD, Horner M-J (eds) SEER survival monograph:cancer survival among adults: U.S. SEER Program, 1988–2001,patient and tumor characteristics. National Cancer Institute,Bethesda, pp 203–216

2. Bishop A, McDonald M, Chang A, Esiashvili N (2012) Infant braintumors: incidence, survival, and the role of radiation based onSurveillance, Epidemiology, and End Results (SEER) data. Int JRadiat Oncol Biol Phys 82:341–347

3. Bleyer W (1999) Epidemiologic impact of children with braintumors. Childs Nerv Syst 15:758–763

4. Buttarelli FR,MassiminoM,AntonelliM,Lauriola L,Nozza P,DonofrioV, Arcella A, Oliva MA, Di Rocco C, Giangaspero F (2010) Evaluationstatus and prognostic significance of O6-methylguanine-DNAmethyltransferase (MGMT) promotermethylation in pediatric high gradegliomas. Childs Nerv Syst 26:1051–1056

5. Childhood cancers. National Cancer Institute at the NationalInstitutes of Health, (2008) http://www.cancer.gov/cancertopics/factsheet/Sites-Types/childhood. Accessed 30 Jan 2012

6. Coco S, Valdora F, Bonassi S, Scaruffi P, Stigliani S, Oberthuer A,Berthold F, Andolfo I, Servidei T, Riccardi R, Basso E, Iolascon Aet al (2011) Chromosome 9q and 16q loss identified by genome-wide

pooled-analysis are associated with tumor aggressiveness in patientswith classic medulloblastoma. OMICS 15:273–280

7. Data on health and well-being of American Indians, Alaska Natives,and Other Native Americans, 2007. Surveillance, Epidemiology, andEnd Results (SEER). http://aspe.hhs.gov/hsp/06/catalog-ai-an-na/SEER.htm. Accessed 30 Jan 2012

8. Deorah S, Lynch C, Sibenaller Z, Ryken T (2006) Trends in braincancer incidence and survival in the United States: Surveillance,Epidemiology, and End Results Program, 1973 to 2001. NeurosurgFocus 20:1–7

9. Gibbs I, Tuamokumo N, Yock T (2006) Role of radiation therapy inpediatric cancer. Hematol Oncol Clin N Am 20:455–470

10. Gurney J (1999) Brain cancer incidence in children: time to lookbeyond the trends. Med Pediatr Oncol 33:110–112

11. Inskip P, Linet M, Heineman E (1995) Etiology of brain tumors inadults. Epidemiol Rev 17:382–414

12. Joinpoint Regression Program, Version 3.5.2. October 2011;Statistical Research and Applications Branch, National CancerInstitute

13. Kieran M, Walker D, Frappaz D, Prados M (2010) Brain tumors:from childhood through adolescence into adulthood. J Clin Oncol 28:4783–4789

14. Kilday JP, Rahman R, Dyer S, Ridley L, Lowe J, Coyle B, Grundy R(2009) Pediatric ependymoma: biological perspectives. Mol CancerRes 7:765–786

15. Knab B, Comell P (2007) Radiotherapy for pediatric brain tumors:when and how. Expert Rev Anticancer Ther 7:S69–S77

16. Küpeli S, Yalçin B, Bilginer B, Akalan N, Haksal P, BüyükpamukçuM (2011) Posterior fossa syndrome after posterior fossa surgery inchildren with brain tumors. Pediatr Blood Cancer 56:206–210

17. Legler J, Ries L, Smith M (1999) Brain and other central nervoussystem cancers: recent trends in incidence and mortality. J NatlCancer Inst 91:1382–1390

18. Linet M, Ries L, Smith M, Tarone R, Devesa S (1999) Cancersurveillance series: recent trends in childhood cancer incidence andmortality in the USA. J Natl Cancer Inst 91:1051–1058

19. MacDonald SM, Trofimov A, Safai S, Adams J, Fullerton B, Ebb D,Tarbell NJ, Yock TI (2011) Proton radiotherapy for pediatric centralnervous system germ cell tumors: early clinical outcomes. Int JRadiat Oncol Biol Phys 79:121–129

20. Mueller S, Chang S (2009) Pediatric brain tumors: current treatmentstrategies and future therapeutic approaches. Neurotherapeutics 6:570–586

21. Musial-Bright L, Panteli L, Driever P (2011) Pediatric low-gradeglioma survivors experience high quality of life. Childs Nerv Syst27:1895–1902

22. Rao A, Packer R (2012) Impact of molecular biology studies on theunderstanding of brain tumors in childhood. CurrOncol Rep 14:206–212

23. Rorke L, Gilles F, Davis R, Becker L (1985) Revision of the WorldHealth Organization classification of brain tumors for childhoodbrain tumors. Cancer 56:1869–1886

24. Schechter C (1999) Re: Brain and other central nervous systemcancers: recent trends in incidence and mortality [Letter to theEditor]. J Natl Cancer Inst 92:77–78

25. Smith M, Freidlin B, Ries L, Simon R (1998) Trends in reportedincidence of primary malignant brain tumors in children in the UnitedStates. J Natl Cancer Inst 90:1269–1277

26. Steen RG, Taylor JS, Langston JW, Glass JO, Brewer VR, ReddickWE,Mages R, Pivnick EK (2001) Prospective evaluation of the brainin asymptomatic children with neurofibromatosis Type 1:relationship of macrocephaly to T1 relaxation changes and structuralbrain abnormalities. Am J Neuroradiol 22:810–817

27. Surveillance, Epidemiology, and End Results (SEER) Program (www.seer.cancer.gov) SEER*Stat Database: Incidence - SEER 9 RegsResearch Data, Nov 2010 Sub (1973–2008) (Katrina/Rita populationadjustment) Linked to county attributes - total U.S., 1969–2009

Childs Nerv Syst (2014) 30:147–154 153

Counties, National Cancer Institute, DCCPS, Surveillance ResearchProgram, Cancer Statistics Branch, released April 2011, based on theNovember 2010 submission.

28. Tatter S (2006) The new WHO classification of tumors affecting thecentral nervous system. Massachusetts General Hospital NeurologicalService http://neurosurgery.mgh.harvard.edu/newwhobt.htm. Accessed28 Jan 2012

29. Varley JM (2003) Germline TP53 mutations and Li–Fraumenisyndrome. Hum Mutat 21:313–320

30. Villani A, Malkin D, Tabori U (2011) Syndromes predisposing topediatric central nervous system tumors: lessons learned and newpromises. Curr Neurol Neurosci Rep 12:153–164

31. Werbowetski-Ogilvie T, Morrison L, Fiebig-Comyn A, BhatiaM (2011) In vivo generation of neural tumors from neoplasticpluripotent stem cells models early human pediatric braintumor formation. Stem Cells 30:392–404

32. Wiegering V, Eyrich M, Rutkowski S, Wölfl M, Schlegel P,Winkler B (2011) TH1 predominance is associated withimproved survival in pediatric medulloblastoma patients.Cancer Immunol Immunother 60:693–703

33. Wu XC, Chen VW, Steele B, Roffers S, Klotz JB, Correa CN,Carozza SE (2003) Cancer incidence in adolescents and youngadults in the United States, 1992–1997. J Adolesc Health 32:405–415

154 Childs Nerv Syst (2014) 30:147–154