Clinical implementation of integrated whole-genome copy number and mutation proﬁling for glioblastoma

Background. Multidimensional genotyping of formalin-ﬁxed parafﬁn-embedded (FFPE) samples has the potential to improve diagnostics and clinical trials for brain tumors, but prospective use in the clinical setting is not yet routine. We report our experience with implementing a multiplexed copy number and mutation-testing program in a diagnostic laboratory certiﬁed by the Clinical Laboratory Improvement Amendments. Methods. We collected and analyzed clinical testing results from whole-genome array comparative genomic hybridization (Onco-Copy) of 420 brain tumors, including 148 glioblastomas. Mass spectrometry–based mutation genotyping (OncoMap, 471 mutations) was performed on 86 glioblastomas. Results. OncoCopy was successful in 99% of samples for which sufﬁcient DNA was obtained ( n ¼ 415). All clinically relevant loci for glioblastomas were detected, including ampliﬁcations ( EGFR, PDGFRA, MET ) and deletions ( EGFRvIII , PTEN , 1p/19q). Glioblastoma patients ≤ 40 years old had distinct proﬁles compared with patients . 40 years. OncoMap testing reliably identiﬁed mutations in IDH1 , TP53 , and PTEN . Seventy-seven glioblastoma patients enrolled on trials, of whom 51% participated in targeted therapeutic trials where multiplex data informed eligibility or outcomes. Data integration identiﬁed patients with complete tumor suppressor inactivation, albeit rarely (5% of patients) due to lack of whole-gene coverage in OncoMap. Conclusions. Combined use of multiplexed copy number and mutation detection from FFPE samples in the clinical setting can efﬁciently replace singleton tests for clinical diagnosis and prognosis in most settings. Our results support incorporation of these assays into clinical trials as integral biomarkers and their potential to impact interpretation of results. Limited tumor suppressor variant capture by targeted genotyping highlights the need for whole-gene sequencing in glioblastoma.

Molecular biomarkers are increasingly used for many aspects of clinical care based on their potential diagnostic, prognostic, and predictive capacities. Multiplexed genomic data offer the possibility to obtain results for known biomarkers while simultaneously generating information that may be useful for future exploratory analyses or for clinical trial eligibility screening. The Cancer Genome Atlas (TCGA) and others have identified several pathways with recurrent aberrations in glioblastoma (GBM); however, routine incorporation of genomic precision medicine tools into the clinical environment of laboratories certified by the Clinical Laboratory Improvement Amendments (CLIA) still presents significant challenges. While targeted multiplexed assays are increasingly being adopted, integration of results from multiple tests on patients' tumors is not widely implemented in the clinical or clinical trial research setting. 1 -3 For patients diagnosed with malignant glioma, copy number aberrations such as EGFR amplification currently support tumor classification and are frequently identified by fluorescence in situ hybridization (FISH). While whole-genome somatic copy number analysis platforms, including array comparative genomic hybridization (aCGH) and single nucleotide polymorphism (SNP) arrays, have become routine tools to characterize cancer genomes in the research setting, 4 -7 reliable somatic copy number analysis of formalin-fixed paraffin-embedded (FFPE) tissues has been slow to emerge in routine clinical practice, largely due to relatively poor DNA integrity isolated from clinical samples. 8,9 Recent advances, however, now enable reliable and sensitive aCGH testing of FFPE samples. 10 Targeted multiplexed somatic mutation testing has been more rapidly implemented using several platforms. 11,12 Here, we report our results and experience with FFPE-based aCGH (OncoCopy) and mass spectrometry mutational genotyping assays 13,14 (OncoMap) in a CLIA-laboratory setting at the Dana-Farber/Brigham and Women's Cancer Center (DF/ BWCC). We demonstrate that these tests reliably provide critical diagnostic and prognostic data for use in clinical management, including selection of targeted therapies and enrollment into biomarker-based clinical trials.

Patient Selection
Analysis of data generated from tumor specimens and clinical variables was conducted following approval from the DF/BWCC institutional review board. Genotyping data from clinical Onco-Copy reports were obtained from the medical record, while raw aCGH data were obtained from the Brigham and Women's Hospital (BWH) Cytogenetics Laboratory. Somatic mutation profiling was performed with consent for the DF/BWCC PROFILE clinical research study approved by the DF/BWCC institutional review board. Tests were performed within the Cytogenetics (OncoCopy) and Molecular Diagnostics (OncoMap) Divisions of the BWH Center for Advanced Molecular Diagnostics, a CLIA-certified laboratory environment. All samples underwent central histopathologic review by at least 2 board-certified neuropathologists (S.H.R., S.S., R.D.F. or K.L.L) using World Health Organization criteria. Patient trial participation was assessed by retrospective review of medical records. Trial participation was scored as positive for any therapeutic trial at any time point in a patients' care.
Molecularly informed trials were defined as those that had molecular enrollment criteria (immunohistochemistry or genomic assay based) where OncoCopy or OncoMap results could have theoretically contributed to enrollment or other decisions in the trial.

Clinical Array Comparative Genomic Hybridization
Minimum sample requirements were established based on research assay performance and validation studies performed as part of migration to the CLIA lab. Pathologists were instructed to submit 10 × 5 mm FFPE sections from a block with a minimum tumor size of 0.5 cm 2 in cross section, .50% tumor nuclei. Coring or microdissection was performed on circled regions of the tumor when necessary to achieve these requirements. Post-extraction, the minimum amount of DNA required to perform hybridization was set at 1.3 mg DNA by Nanodrop.
A minimum of 1.3 mg DNA (10×5 mm FFPE sections) was obtained as part of clinical care. Due to the impact of necrosis on aCGH signal detection, the total necrosis was required to be ≤30% for GBM samples being submitted for aCGH without macrodissection. The value of ≤30% was selected as a particularly strict lower level limit because it allows for pathologist interobserver variability in scoring tumor necrosis. For GBM with .30% total necrosis, macrodissection (including coring) of viable tumor tissue was recommended to achieve this.
Patient and reference DNA (Promega) were fragmented using previously described fragmentation simulation methods and hybridized to Agilent SurePrint G3 Human 1 Million feature arrays. 10 Clinical analysis was performed using Agilent Workbench software, and log ratios were normalized using the centralization algorithm, with a threshold score of 6.0 and bin size of 10. Somatic copy number analyses of 8 consecutive probes with mean log 2 ratios .0.18 (gains) and ,-0.30 (losses) were called using the ADM2 algorithm (Algorithm Design Manual, 2nd ed). Clinical reports contained only aberrations that affected a previously annotated list of genes/regions with diagnostic, prognostic, and/or therapeutic significance in brain tumors (Supplementary Table S1, termed OncoCopy). The time from tissue submission to data reporting averaged 2 weeks. Raw whole-genome data files were retained and used for exploratory research analyses (eg, Genomic Identification of Significant Targets in Cancer [GISTIC]).

OncoMap
DNA was isolated from five to ten 5-mm FFPE sections containing at least 50% tumor nuclei as previously described. 13 Somatic mutations in tumor DNA were detected using the multiplexed Sequenom-based assay OncoMap in the DF/BWCC CLIA-certified laboratory. The assay OncoMap v4 detects mutations in 471 different loci from 41 cancer genes (Supplementary Table S2). The average time from tissue submission to report of data was 6 weeks, which includes 1 week for tissue acquisition/ DNA preparation, 3 weeks for genotyping analysis and validation, and 1-2 weeks for pathologist review. Data visualization was performed using the Memorial Sloan-Kettering Cancer Center cBioPortal for Cancer Genomics software. 15 Ramkissoon et al.: Integrating aCGH and genotyping for neuro-oncology Fluorescence In situ Hybridization FISH was performed on 5-mm FFPE tissue sections using methods described previously 16

Exploratory Research Analysis of aCGH Data
Circular binary segmentation was used to segment copy number data for research aCGH analysis using parameters (a ¼ 0.01, undo.splits ¼ none, minimum width ¼ 5). 17 Segmented data were analyzed with GISTIC 2.0 18 to determine statistically significant recurrent somatic copy number aberrations (SCNAs), after filtering germline copy number variations, using the following parameters: minimum segment size ¼ 10, lesion amplitude threshold of 0.2 for aCGH cohort; focal versus broad SCNA events were defined with a cutoff of 0.5× chromosome arm length, and gene confidence level ¼ 0.99. Results were compared with well-established data from TCGA. 19 Segmented data were visually presented using Integrative Genomics Viewer 2.3 in heatmap format. 20 A 2-tailed Fisher's exact test was applied to identify peaks that were differentially observed based on the patient's age. For hypothesis testing, the 148 GBM patients were sequentially partitioned into 2 groups based on arbitrarily selected age cutoffs (35,40, or 45 y), and copy number profiles were compared. The most significant differences were detected when the patient cohort was split into a ≤40 year group and a .40 year group. Copy number alterations were called when associated with log 2 copy number changes .0.2. Peaks with a Benjamini -Hochberg false discovery rate q-value of ,0.2 were highlighted between the 2 groups. Wilcoxon rank-sum analysis was performed to compare frequency of broad, focal, and all SCNA events between patients younger versus older than 40 years for aCGH and TCGA cohorts (Supplementary Table S3).
To compare the incidence of focal regions of amplification and deletion between the cohorts of TCGA and DF/BWCC aCGH, genes representative of each peak from the analyses by TCGA were chosen (Supplementary Table S4A and B). The frequencies of amplification or deletion above a threshold of 0.1 for TCGA data and 0.2 for aCGH data were compared for these representative genes using Fisher's 2-tailed exact test with Bonferroni correction for multiple hypotheses. Only subjects .40 years of age were included in this comparison to minimize age variation between younger subjects across TCGA and DF/BWCC aCGH populations.

Clinical Performance Characteristics of OncoCopy Applied to FFPE Brain Tumors
From the period 2012-2013, OncoCopy (whole-genome aCGH) was clinically requested on 469 primary brain tumors for diagnostic, prognostic, or treatment-related indications (Fig. 1A).
OncoCopy is a routine clinical test requested by the patient's oncologist, pathologist, or other physicians, performed through Ramkissoon et al.: Integrating aCGH and genotyping for neuro-oncology the BWH Division of Cytogenetics and paid for by the patient's insurance company. Examples of frequent indications for testing in gliomas were EGFR amplification (diagnostic, trial eligibility for glioblastoma), 1p/19q codeletion (diagnostic, prognostic, treatment for oligodendrogliomas), and BRAF duplication (diagnostic of pilocytic astrocytoma). Following DNA isolation and implementation of quality control measures, 49 samples (10.4%) yielded insufficient DNA. Insufficient DNA yields were highest in GBM (17.7%, 33/186) and lowest in meningiomas (0/122); however, 85% of GBM that yielded insufficient DNA were consult cases from outside hospitals where access to additional FFPE material was challenging. The remaining samples were from biopsies with limited amounts of tumor material. Importantly, although 18% of GBM samples were insufficient for OncoCopy, the extracted DNA was used for OncoMap and/ or MGMT promoter methylation analysis.
We successfully obtained genome-wide SCNA profiles with derivative log ratio spread noise measures of ,0.30 in 99% of cases (415/420). Repeat review of histology for 5 GBM samples with derivative log ratio spread .0.30 revealed .30% necrotic tissue. In total, 415 brain tumors, including 148 GBM, were analyzed and included during this pilot period (2012 -2013). Patient demographics for 148 and 86 GBM cases are reported in this study by OncoCopy and OncoMap, respectively (Fig. 1B).

OncoCopy Reliably Detects Common Clinically Relevant Aberrations in GBM
Data generated from whole-genome copy number analysis by aCGH comprehensively detect all known genomic gains or losses clinically relevant to GBM diagnoses and prognoses in a single assay, but also detect numerous SCNAs of unknown significance which we elected not to report. 21 To capitalize on the comprehensive nature of this test while reducing the complexity for clinical use, we utilized manufacturer recommended software (Agilent Cytogenomics) to restrict clinical reporting to 41 SCNAs (termed OncoCopy) with known diagnostic or prognostic value to brain tumors, including those that are relevant to clinical trial enrollment or targeted therapies 22 (Supplementary Table S1).
We examined the collective OncoCopy genotyping results derived from individual adult GBM patients for 14 common and relevant GBM aberrations and compared incidences for these reported by the 2008 GBM dataset of TCGA. 4 The most significant amplification encompassed EGFR (7p11.2) and was present in 35% (52/148) of samples. The oncogenic EGFRvIII variant (deletion of exons 2 -7) was detected in 11.5% (6/52) of EGFR-amplified GBM, which was comparable to rates previously reported using aCGH and SNP-based platforms but lower than that for RNA sequencing or Nanostring assays. 23 -26 The most significant deletion encompassed CDKN2A/B (9p21.3) and was detected in 39% (58/148) of samples. Differences between our clinical results on FFPE material versus frozen tissue used by TCGA were statistically significant for only CDKN2A homozygous loss, which occurred at a lower incidence in our dataset (P ¼ .004; Fig. 1C). 4 Taken together, our analysis demonstrates that OncoCopy data generated from FFPE GBM samples for clinical purposes provide reliable, reproducible results by identifying diagnostic and clinically relevant GBM copy number aberrations.

OncoCopy Identifies Loci With Clonal Heterogeneity Resulting From Genomically Distinct Tumor Cell Subpopulations
In a subset of patient samples, our clinical analysis found focal (117.3 Kb to 5.3 Mb) low-level gains (log 2 ratio range 0.25 -2.0) in MYCN, PDGFRA, EGFR, MYC, or MET. To determine whether these represent a subpopulation of amplified cells or low-level gain in the majority of cells, we performed FISH with gene-specific probes on 8 representative cases in which OncoCopy showed low-level gains involving MYCN (n ¼ 2), MYC (n ¼ 2), PDGFRA (n ¼ 2), MET (n ¼ 2), or EGFR (n ¼ 1). In all cases, we detected genomic amplifications in subpopulations of tumor cells ranging from 5% to 35% ( Fig. 2A and B). One tumor, GBM08, demonstrated at least 2 distinct genomic subpopulations detected by aCGH, corresponding to PDGFRA or MET amplification in different tumor cell subpopulations, as has been previously described 27,28 ( Fig. 2A, bottom panel).

Distinct Copy Number Profiles in GBM Based on Age at Diagnosis
Adult GBM tumors are commonly thought to be associated with a copy number "signature," typified by polysomy 7 with or without EGFR amplification, CDKN2A/B single copy or homozygous loss, and monosomy 10 that leads to single copy PTEN loss. Based on observations during routine clinical care, we hypothesized that genomic profiles of young GBM patients were different from those of GBM patients in the later decades of life. The SCNA patterns among younger GBM patients looked strikingly different from patterns observed in older GBM patients (Fig. 4A). To quantify these differences, we compared the number of SCNAs per sample for patient cohorts dichotomized at ages 35, 40, and 45. We found significant differences in the number of arm-level SCNAs per GBM between patients ≤40 years (n ¼ 26) compared with .40 years of age (n ¼ 122) (7.4 and 5.3 events per sample, respectively; P ¼ .04 after multiple hypotheses correction); however, we did not see similar differences among focal SCNAs. We also observed fewer arm-level events among TCGA GBM in older patients relative to younger patients, but this was not statistically significant (P ¼ .3). In our cohort of 148 GBM cases, only 8/26 GBM patients ≤40 years had an IDH1/2 mutation compared Ramkissoon et al.: Integrating aCGH and genotyping for neuro-oncology with 2/122 patients .40 years, suggesting that the copy number profiles for the vast majority of younger GBM patients are related to other mechanisms than those proposed for IDH mutant gliomas.
These genome-wide differences were also associated with differences in individual chromosomes (Fig. 4B). Specifically, chromosome 7 gain and chromosome 10 loss (containing EGFR and PTEN, respectively) each occurred in 75% of GBM in patients older than 40 compared with only 25% of patients in the younger cohort. Chromosome 19 gain was also observed almost exclusively among older patients; no patients younger than 40 demonstrated gains in chromosome 19q, and only Distinct sets of recurrent focal SCNAs were also enriched in younger or older patients (Fig. 4C). Older patients were enriched for amplifications of EGFR and deletions of CDKN2A/B. Younger patients were enriched for deletion of a region on chromosome 11 encompassing CDKN1C and CEND1.

Computational Analysis of Clinical OncoCopy Data Compares Favorably With TCGA Analysis
We next compared copy number profiles amongst the DF/ BWCC aCGH cohort with the population in TCGA. Due to the significant differences seen between younger and older patients, we restricted this comparison to adults older than 40 years. Globally, overlay of GISTIC peaks representing significantly amplified and deleted focal chromosome regions demonstrated high levels of similarity between aCGH and TCGA data (Fig. 5A). We also compared frequencies of events at each peak region between the 2 datasets. Only 4 of 22 amplification peaks and 2 of 39 deletion peaks exhibited significant differences; all of these were enriched in the dataset from TCGA (Fig. 5B,  Supplementary Table S4). These included amplifications of MDM4 (1q32.1), CDK4 (12q14.1), 2 regions without known GBM oncogenes: 1p36.21 and 19p13.3, deletions of PTEN (10q23.31), and a region of 10p13 without a known GBM tumor suppressor gene.

OncoMap Somatic Mutation Profiling Reliably Captures Clinically Relevant Mutations in GBM-Associated Oncogenes
OncoMap testing was prospectively performed as an enterprise-level clinical research program at the DF/BWCC with the goal of providing tumor genotyping data to requesting clinicians and consenting patients for clinical trial decision making. OncoMap testing was paid for by the DF/BWCC and provided to patients at no cost. We analyzed OncoMap data from 86 GBM patients and found recurrent mutations in IDH1 (4.7%), PIK3CA (3.5%), PIK3R1 (3.5%), and BRAF (2.3%) at rates similar to those reported in TCGA for the specific mutations targeted by the assay (Fig. 1D). Mutations in TP53, PTEN, and RB1, the most common GBM-associated tumor suppressor genes queried in the OncoMap assay, were detected in 5.8%, 3.5%, and 1.2% of tumors in our study cohort, whereas TCGA reported frequencies of 34%, 31.9%, and 9.9%, respectively, using Sanger-based whole-gene sequencing. Comparison of OncoMap probes, Ramkissoon et al.: Integrating aCGH and genotyping for neuro-oncology designed to capture "hot spots" in tumor suppressor genes, with actual TCGA 2008 mutation data showed that the expected theoretical detection frequencies of OncoMap applied to the cohort from TCGA would be 21% (8/38) for TP53, 20% (6/30) for PTEN, and 0% (0/9) of the RB1 mutations, while 100% (9/9) of PIK3R1 (oncogene) mutations would have been detected.

Integrative OncoCopy and OncoMap Reporting Can Inform Trial Inclusion and Exclusion Criteria Based on Pathway Status of Tumor Suppressors or Oncogenes
In our cohort, comprehensive testing for both OncoCopy and OncoMap was performed in 37 GBM patients and we examined this group for ability of combined testing to provide unique results not achieved with either test alone (Fig. 6). Such integrative analysis identified one tumor with concurrent RB1 mutation and RB1 single copy loss, and one tumor with PTEN mutation and loss of the second PTEN allele. Such events are consistent with complete pathway inactivation. As an example of the potential utility of this integrative information in the clinical trial setting, a patient with combined PTEN alterations was considered "PTEN inactivated" and was molecularly eligible for the pan -phosphatidylinositol-3 kinase inhibitor BKM120 (Novartis), as part of an open phase II clinical trial at DF/BWCC for recurrent GBM (Ivy Early Phase Clinical Trials Consortium, NCT01339052). However, the patient did not subsequently enroll due to failure to meet other clinical eligibility requirements. To further determine the theoretical value of multiplex testing in our patient cohort, we retrospectively determined the number of patients in our study who were involved in clinical therapeutic trials at any point in their care at our institution. Of 198 GBM patients, 77 (39%) were enrolled into 41 distinct clinical trials. Among those 77 cases, 39 (51%) participated in a clinical trial of a targeted agent where eligibility criteria or trial results could be critically informed by data generated by Onco-Copy or OncoMap. As expected due to their longer survival, recurrent GBM patients seen in a tertiary care setting were more likely to be enrolled in a clinical trial over the course of their disease than newly diagnosed GBM patients: 53% of recurrent (17/32) and 36% (60/166) of newly diagnosed GBM patients ultimately participated in one or more clinical trials. Patients with recurrent and newly diagnosed GBM were equal in level of enrollment to trials with molecular eligibility criteria (50% of both cohorts).

Discussion
Implementation of both aCGH and somatic mutation detection technologies in CLIA-certified clinical laboratories at DF/BWCC has enabled us to report both genome-wide SCNAs and mutations from FFPE primary brain tumors. We found that these data streams provide efficient and complementary tumor genotyping data useful for diagnostics/prognostics while also offering genetic profiles that can be used in real time for clinical trial selection and decision making. Ramkissoon et al.: Integrating aCGH and genotyping for neuro-oncology Gliomas, including GBM, are a disease in which copy number alterations appear to be predominant drivers of tumorigenesis and for which whole-genome copy number analysis provides particularly relevant diagnostic, prognostic, and therapeutic information. 4,19,30 In our clinical experience, the most useful prognostic information based on copy number was related to simultaneous detection of 1p/19q codeletion and EGFR amplification. These copy number aberrations are relevant to prognosis in tumors histologically diagnosed as mixed gliomas, and emerging data suggest in fact that they are likely pathognomonic diagnostic features of oligodendroglioma and glioblastoma, soon to be adopted by the field. 31,32 To date, most genome-wide SCNA studies, such as those in TCGA, rely solely on frozen tumor tissue, which essentially excludes this type of testing from the routine clinical diagnostic pipeline, particularly for referral patients, who rarely have frozen tissue available. Our results demonstrate that clinical FFPE aCGH can replace targeted FISH studies that were traditionally performed as standard of care, while providing extensive additional genomic data. The failure to extract sufficient DNA in GBM was higher than expected based on our research experience with the assay; however, this was primarily explained by logistical factors of tissue acquisition from outside sites and failures in the early stages of the implementation where pathologists were not sufficiently trained in the importance of avoiding necrosis in submitted material; these shortcomings have been readily improved as experience with the testing has been gained.
Molecular heterogeneity in the form of subclonal tumor cell populations is well described in GBM and is hypothesized to have clinical importance as a treatment resistance mechanism. 27,33 While there are no specific criteria that can be uniformly applied to definitively identify subclonal populations based on copy number aberrations in any individual tumor, our results do support that specific loci can be reliably identified as highly likely to represent subclonal events based on prior knowledge of consistent patterns of aberration involving specific loci. Our experience suggests that low-level, focal gains affecting specific oncogenes (MYCN, PDGFRA, EGFR, MET, and MYC) using criteria in this study of ,10 Mb and log 2 ratios between 0.25 and 2.0 are sufficiently indicative of intratumoral heterogeneity that clinical reports should formally note this possibility. Furthermore, it seems reasonable to make the suggestion in the report that targeted FISH may need to be performed if formally required for clinical decision making (eg, entry onto clinical trial).
While sensitivity of aCGH is important to consider, our findings demonstrate that aCGH can detect evidence of amplifications occurring in as few as 5%-10% of cells (Fig. 2). However, we note that our EGFRvIII detection rate is lower than some previously reported incidences, such as RNA sequencing and Nanostring assays, but is similar to levels reported in aCGH or SNP assays, including the cohort from TCGA. 26 This is likely due to the intrinsic dynamic range limitations of aCGH, which can be exceeded in the setting of high copy number amplifications seen in GBM. Based on single cell sequencing of GBM, the EGFRvIII mutation typically exists in subclones of tumor cells, which may be masked by the more dominant wild-type epidermal growth factor receptor -amplified cell population. 28 This range problem is not unique to the aCGH platform and is also a limitation of SNP-based copy number assays and SCNAs calculated by quantification of reads in next-generation sequencing. In our experience, these latter methods are currently less sensitive than aCGH for detection of the deleted exons in EGFRvIII. While aCGH offers a viable and rapid-turnaround screen for EGFRvIII and other copy number alterations in the clinical setting, we suggest that in situations where EGFRvIII status is required for clinical trial entry, EGFR-amplified GBM that lacks EGFRvIII should be reflexively submitted for an EGFR-vIII dedicated assay such as immunohistochemistry, reverse transcriptase PCR, or Nanostring.
Our study also revealed data suggestive of biologic underpinnings of the disease. Patients ≤40 years old had a much lower incidence of "classic" GBM copy number aberrations involving EGFR, PTEN, and CDKN2A, and higher frequencies of several other events, including amplifications of AKT3 and CCND2. These findings suggest that GBM arising in younger patients differs from that which develops later in life and may require alternative treatment strategies.
While SCNAs represent key driver events in GBM, mutations in oncogenes and/or tumor suppressor genes represent a complementary level of molecular disruption contributing to the cancer phenotype. 4,19 Incorporating clinical single nucleotide variation analysis for brain tumors is important because it identifies diagnostic and clinically relevant events, including mutations in IDH1/ IDH2 (adult low-grade gliomas), BRAF (gangliogliomas, pleomorphic xanthoastrocytomas), or INI1 (atypical teratoid rhabdoid tumors). 11,13,34 -36 Our results show that mass spectrometrybased methods reliably capture oncogenic mutations and other targeted somatic events in a clinical environment; however, they also highlight the need in GBM for rapid adoption of whole gene/exome sequencing given the low incidence of "hot spots" within the most common tumor suppressor genes involved in GBM. Integration of results from these technologies should greatly increase the completeness of assessing tumor suppressor genes and improve interpretation of responses in clinical trials.
While TCGA required extremely large infrastructure and staff investments, current technologies and methods now allow similar integrative genomics on a scale and timeframe feasible for patients in an academic laboratory setting. Genomically defined clinical trials increasingly require costly screening of large numbers of patients using singleton tests such as FISH. Such approaches are particularly problematic as clinical trials increase in number and complexity. Furthermore, multi-arm genomically stratified trials of targeted agents are currently being designed for GBM and other cancers where routine assessment of multiple biomarkers is essential to the trial design. 37,38 Such trials would be efficiently enabled and their costs reduced by incorporation of multiplexed genotyping approaches described here. Indeed, with a 1-to 2-week turnaround time for OncoCopy, copy number data are reported within a timeframe compatible with the time it takes to obtain final pathologic diagnoses and transfer patient care from neurosurgery to the oncology service. Improvements in the sample-processing pipeline, including reducing DNA preparation and genotyping time or dedicating a pathologist review of molecular data, can further accelerate the turnaround in the future. OncoMap results are complementary and often reveal trial-specific mutations in oncogenes (eg, PIK3CA, PIK3C2B, BRAF), for which patients would be eligible. The 6-week turnaround for large panels, which are run as enterprise-wide batched results, was Ramkissoon et al.: Integrating aCGH and genotyping for neuro-oncology less problematic for patients with newly diagnosed disease given that few trials currently were available. However, this longer time to results did represent a significant challenge for patients presenting at recurrence from outside hospitals, who generally need to make decisions within 4 weeks or less. This suggests that laboratories may need to consider having separate testing for high-throughput, high-dimension, non-timesensitive assays and lower-throughput, low-dimension, more rapid testing for clinical trial incorporation. As the number of genomics-based clinical trials continues to expand, we expect an increase in the percentage of patients with newly diagnosed and recurrent GBM enrolling in molecularly stratified trials.