Novel molecular subgroups for clinical classification and outcome prediction in childhood medulloblastoma: a cohort study

Abstract

Background: International consensus recognises four medulloblastoma molecular subgroups: WNT (MB_WNT), SHH (MB_SHH), group 3 (MB_Grp3), and group 4 (MB_Grp4), each defined by their characteristic genome-wide transcriptomic and DNA methylomic profiles. These subgroups have distinct clinicopathological and molecular features, and underpin current disease subclassification and initial subgroup-directed therapies that are underway in clinical trials. However, substantial biological heterogeneity and differences in survival are apparent within each subgroup, which remain to be resolved. We aimed to investigate whether additional molecular subgroups exist within childhood medulloblastoma and whether these could be used to improve disease subclassification and prognosis predictions.

Methods: In this retrospective cohort study, we assessed 428 primary medulloblastoma samples collected from UK Children's Cancer and Leukaemia Group (CCLG) treatment centres (UK), collaborating European institutions, and the UKCCSG-SIOP-PNET3 European clinical trial. An independent validation cohort (n=276) of archival tumour samples was also analysed. We analysed samples from patients with childhood medulloblastoma who were aged 0-16 years at diagnosis, and had central review of pathology and comprehensive clinical data. We did comprehensive molecular profiling, including DNA methylation microarray analysis, and did unsupervised class discovery of test and validation cohorts to identify consensus primary molecular subgroups and characterise their clinical and biological significance. We modelled survival of patients aged 3-16 years in patients (n=215) who had craniospinal irradiation and had been treated with a curative intent.

Findings: Seven robust and reproducible primary molecular subgroups of childhood medulloblastoma were identified. MB_WNT remained unchanged and each remaining consensus subgroup was split in two. MB_SHH was split into age-dependent subgroups corresponding to infant (<4·3 years; MB_SHH-Infant; n=65) and childhood patients (≥4·3 years; MB_SHH-Child; n=38). MB_Grp3 and MB_Grp4 were each split into high-risk (MB_Grp3-HR [n=65] and MB_Grp4-HR [n=85]) and low-risk (MB_Grp3-LR [n=50] and MB_Grp4-LR [n=73]) subgroups. These biological subgroups were validated in the independent cohort. We identified features of the seven subgroups that were predictive of outcome. Cross-validated subgroup-dependent survival models, incorporating these novel subgroups along with secondary clinicopathological and molecular features and established disease risk-factors, outperformed existing disease risk-stratification schemes. These subgroup-dependent models stratified patients into four clinical risk groups for 5-year progression-free survival: favourable risk (54 [25%] of 215 patients; 91% survival [95% CI 82-100]); standard risk (50 [23%] patients; 81% survival [70-94]); high-risk (82 [38%] patients; 42% survival [31-56]); and very high-risk (29 [13%] patients; 28% survival [14-56]).

Interpretation: The discovery of seven novel, clinically significant subgroups improves disease risk-stratification and could inform treatment decisions. These data provide a new foundation for future research and clinical investigations.

Funding: Cancer Research UK, The Tom Grahame Trust, Star for Harris, Action Medical Research, SPARKS, The JGW Patterson Foundation, The INSTINCT network (co-funded by The Brain Tumour Charity, Great Ormond Street Children's Charity, and Children with Cancer UK).

Figure 1

Novel clinically significant subgroups within the established medulloblastoma subgroups (A) Non-negative matrix factorisation consensus clustering of methylome data from 428 primary medulloblastomas. Each column represents one patient. Missing data are shown in grey. Residuals from χ² tests indicate where subgroup-enrichment has occurred (darker shades of grey indicate stronger relationships), p values are from χ² tests of enrichment; scale bar for residuals (−2 to 2) is shown. Methylation-derived metagene levels (V1–V6), which define subgroup membership, are also shown (red indicates high metagene levels, blue indicates low levels). (B) Overall survival of patients in the seven identified subgroups. All discovery cohort patients with available overall survival information are shown (n=367). (C) Progression-free survival of patients in the consensus four subgroups of medulloblastoma in discovery cohort patients receiving craniospinal irradiation and aged 3–16 years at diagnosis (n=250). (D) Progression-free survival of patients in the seven identified subgroups of medulloblastoma in patients receiving craniospinal irradiation and aged 3–16 years at diagnosis (n=239). Discrepancy in the numbers of patients in (C) and (D) is due to consensus clustering; certain samples could not be confidently classified for the seven subgroup model or the four subgroup model, and were omitted from the figures. DN/MBEN=desmoplastic or nodular medulloblastoma with extensive nodularity. HR=hazard ratio. LCA=large-cell anaplastic. M+=metastatic disease. R+=residual disease.

Figure 2

MB_SHH disease comprises two age-dependent molecular subgroups (A) Log-normal age distributions of MB_SHH-Infant (red) and MB_SHH-Child disease (dark red). Patient ages at diagnosis are shown as ticks along the x-axis and are coloured by subgroup. (B) Clinicopathological and molecular disease features of MB_SHH-Infant and MB_SHH-Child subgroups. Residuals from χ² tests indicate where subgroup-enrichment has occurred (darker shades of grey indicate stronger relationships); scale bar for residuals (−4 to 4) is shown. p values from χ² tests are shown. Differentially methylated probes: Illumina probe identifiers for the top 20 most differentially methylated probes, alongside methylation status of 18 normal cerebella (pink). Each column represents one patient. (C) SHH genome-sequencing data was classified into methylation subgroups on the basis of age. Each column represents one patient. Amp=amplification. DN/MBEN=desmoplastic or nodular medulloblastoma with extensive nodularity. LCA=large-cell anaplastic. M+=metastatic disease. R+=residual disease.

Figure 3

Characterisation of MB_Grp3 and MB_Grp4 subgroups (A) Clinicopathological and molecular disease features. Residuals from χ² tests indicate where subgroup-enrichment has occurred (darker shades of grey indicate stronger relationships); scale bar for residuals (−6 to 6) is shown. p values from χ² tests are shown. (B) Heat map shows the top 20 differentially methylated probes for these subgroups. Methylation data of 18 normal cerebella are shown alongside and magnitude of MB_Grp3 and MB_Grp4 metagenes is shown below. (C) Identification of MB_Grp3 and MB_Grp4 medulloblastoma cytogenetic determinants. Markers with p<0·05 and present in at least 10% of one subgroup are ordered by their subgroup association and then by chromosomal order. Residuals from χ² tests indicate where subgroup enrichment has occurred (darker shades of grey indicate stronger relationships), across all subgroups and within MB_Grp3 and MB_Grp4 individually. p values from χ² tests are shown. i17q=isochromosome 17q. LCA=large-cell anaplastic. M+=metastatic disease. R+=residual disease.

Figure 4

Novel risk stratification scheme for MB_Grp3 and MB_Grp4 medulloblastoma (A) Progression-free survival plots for identified risk subgroups (n=156) defined in table 3 and the appendix (p 20). (B) Time-dependent ROC curves at 5 years are shown for this novel risk stratification alongside a published cytogenetic stratification scheme (MB_Grp4 with chromosome 11 loss or chromosome 17 gain, low risk; MB_Grp4 with M– disease, standard risk; MB_Grp4 with M+ disease, high risk; MB_Grp3 with MYC amplification, i17q, or M+ disease, high risk; MB_Grp3 without MYC amplification, i17q, or M+ disease, standard risk), and the PNET5 risk stratification (patients positive for one or more of LCA pathology, M+ disease, R+ disease, MYC(N) amplification are high risk; patients absent for all high-risk features, standard risk), as well as the stratification derived from considering MB_Grp3 and MB_Grp4 as separate entities (appendix p 22). AUC=area under curve. LCA=large-cell anaplastic. M+=metastatic disease. M–=non-metastatic disease. ROC=receiver operating characteristic.

Figure 5

Summary of the seven primary childhood medulloblastoma subgroups Demographic, clinicopathological, and molecular features are summarised. *Comparisons of cytogenetic, gene expression, and DNA methylation changes are made with respect to their counterpart subgroup, except for MB_WNT cases, which were compared with normal cerebella if data were available. For probe-level comparisons, Kyoto Encyclopedia of Genes and Genomes pathway enrichment of demethylated loci was investigated, after correcting for multiple probes mapping to the same gene (data summarised in appendix pp 27–31). CB=normal cerebella. CLAS=classic histological subtype. DN=desmoplastic nodular. LCA=large-cell anaplastic.

Figure 6

Summary of survival modelling of novel medulloblastoma subgroups (A) Summary of a novel risk-stratification scheme for childhood medulloblastoma in a cohort of patients aged 3–16 years receiving craniospinal irradiation (n=215). The potential to further stratify MB_Grp4-LR patients into favourable and high-risk groups by their metastatic stage is shown (dashed arrows). (B) Kaplan-Meier plot of childhood medulloblastoma risk stratification. (C) Performance of novel stratification scheme in comparison with time-dependent ROC curves of existing schemes of progression-free survival at 5 years. MB_Grp3/4: MB_Grp3 and MB_Grp4 considered as a single entity; MB_Grp3/4 plus M+: MB_Grp3 and MB_Grp4 considered as a single entity with MB_Grp4-LR and non-MYC amplified MB_Grp3-LR further stratified by M+ disease status; MB_Grp3 and MB_Grp4: MB_Grp3 and MB_Grp4 stratified separately; cytogenetic: cytogenetically defined scheme; PNET5: scheme employed by HIT-SIOP-PNET5-MB clinical trial. LCA=large-cell anaplastic. M+=metastatic disease. M–=non-metastatic disease. R+=residual disease.