FSTL5 is a marker of poor prognosis in non-WNT/non-SHH medulloblastoma.

PURPOSE: Integrated genomics approaches have revealed at least four distinct biologic variants of medulloblastoma: WNT (wingless), SHH (sonic hedgehog), group C, and group D. Because of the remarkable clinical heterogeneity of group D tumors and the dismal prognosis of group C patients, it is vital to identify molecular biomarkers that will allow early and effective treatment stratification in these non-WNT/non-SHH tumors. PATIENTS AND METHODS: We combined transcriptome and DNA copy-number analyses for 64 primary medulloblastomas. Bioinformatic tools were used to discover marker genes of molecular variants. Differentially expressed transcripts were evaluated for prognostic value in the screening cohort. The prognostic power of follistatin-like 5 (FSTL5) immunopositivity was tested for 235 nonoverlapping medulloblastoma samples on two independent tissue microarrays. RESULTS: Comprehensive analyses of transcriptomic and genetic alterations delineate four distinct variants of medulloblastoma. Stable subgroup separation was achieved by using the 300 transcripts that varied the most. Distinct expression patterns of FSTL5 in each molecular subgroup were confirmed by quantitative real-time polymerase chain reaction. Immunopositivity of FSTL5 identified a large cohort of patients (84 of 235 patients; 36%) at high risk for relapse and death. Importantly, more than 50% of non-WNT/non-SHH tumors displayed FSTL5 negativity, delineating a large patient cohort with a good prognosis who would otherwise be considered intermediate or high-risk on the basis of current molecular subgrouping. CONCLUSION: FSTL5 expression denoted a dismal prognosis both within and across medulloblastoma subgroups. The addition of FSTL5 immunohistochemistry to existing molecular stratification schemes constitutes a reliable and cost-effective tool for prognostication in future clinical trials of medulloblastoma.

TP53 mutation is frequently associated with CTNNB1 mutation or MYCN amplification and is compatible with long-term survival in medulloblastoma.

PURPOSE: The role of TP53 mutations in the tumorigenesis of sporadic medulloblastoma (MB) and the value of TP53 mutation status as a prognostic marker are not yet definitely elucidated. A recent report identified TP53 mutations in MB as an adverse prognostic marker. Hence, the current study was conducted to validate the prognostic role of TP53 mutation in MB and to understand its contribution to tumorigenesis. METHODS: A comprehensive genetic analysis of 310 MB samples was performed by screening for TP53 mutations and further relating the TP53 mutation status to p53 immunostaining, cytogenetic aberrations, and clinical variables. RESULTS: Mutation analysis of TP53 revealed mutations in 21 (6.8%) of 310 samples. Germline TP53 mutations were found in two patients with a history suggestive of a hereditary cancer syndrome. TP53 mutation status was not associated with unfavorable prognosis (P = .63) and was not linked to 17p allelic loss but was over-represented in the prognostically favorable WNT subgroup of MB as defined by CTNNB1 mutation (seven of 35 TP53-mutated tumors v 14 of 271 TP53 wild-type tumors; P = .005) and in tumors carrying high-level MYCN amplification (seven of 21 TP53-mutated tumors v 14 of 282 TP53 wild-type tumors; P = .001). CONCLUSION: The contradictory results in the recent literature concerning the prognostic value of TP53 mutation might be explained by different frequencies of WNT MBs, different frequencies of patients with Li-Fraumeni syndrome, and different cumulative doses of alkylating drugs applied in these studies.

Molecular staging of intracranial ependymoma in children and adults.

PURPOSE: The biologic behavior of intracranial ependymoma is unpredictable on the basis of current staging approaches. We aimed at the identification of recurrent genetic aberrations in ependymoma and evaluated their prognostic significance to develop a molecular staging system that could complement current classification criteria. PATIENTS AND METHODS: As a screening cohort, we studied a cohort of 122 patients with ependymoma before standardized therapy by using array-based comparative genomic hybridization. DNA copy-number aberrations identified as possible prognostic markers were validated in an independent cohort of 170 patients with ependymoma by fluorescence in situ hybridization analysis. Copy-number aberrations were correlated with clinical, histopathologic, and survival data. RESULTS: In the screening cohort, age at diagnosis, gain of 1q, and homozygous deletion of CDKN2A comprised the most powerful independent indicators of unfavorable prognosis. In contrast, gains of chromosomes 9, 15q, and 18 and loss of chromosome 6 were associated with excellent survival. On the basis of these findings, we developed a molecular staging system comprised of three genetic risk groups, which was then confirmed in the validation cohort. Likelihood ratio tests and multivariate Cox regression also demonstrated the clear improvement in predictive accuracy after the addition of these novel genetic markers. CONCLUSION: Genomic aberrations in ependymomas are powerful independent markers of disease progression and survival. By adding genetic markers to established clinical and histopathologic variables, outcome prediction can potentially be improved. Because the analyses can be conducted on routine paraffin-embedded material, it will now be possible to prospectively validate these markers in multicenter clinical trials on population-based cohorts.

Adult and pediatric medulloblastomas are genetically distinct and require different algorithms for molecular risk stratification.

PURPOSE: Medulloblastoma (MB) is the most common malignant brain tumor in children, whereas it rarely presents in adults. We aimed to identify genetic aberrations in 146 adult MBs to evaluate age-dependent differences in tumor biology and adapt age-specific risk stratification models. METHODS: As a screening set, we studied a cohort of 34 adult MBs by using array-based comparative genomic hybridization comparing molecular results with clinical data. DNA copy number aberrations identified as possible prognostic markers were validated in an independent cohort of 112 adult patients with MB by fluorescent in situ hybridization analysis. Results were compared with the data obtained from 404 pediatric patients with MB. RESULTS: CDK6 amplification, 10q loss, and 17q gain are the most powerful prognostic markers in adult MB. Whereas MYC/MYCN oncogene amplifications had a high prognostic value in pediatric MB, these aberrations were rarely observed in adult tumors. Surprisingly, adult MBs with 6q deletion and nuclear beta-catenin activation did not share the excellent prognosis with their pediatric counterparts. CONCLUSION: Adult MB is distinct from pediatric MB in terms of genomic aberrations and their impact on clinical outcomes. Therefore, adult MBs require age-specific risk stratification models. We propose a molecular staging system involving three distinct risk groups based on DNA copy number status of 10q and 17q.