Advertisement

Genomic analysis of paired IDHwt glioblastomas reveals recurrent alterations of MPDZ at relapse after radiotherapy and chemotherapy

Published:February 23, 2022DOI:https://doi.org/10.1016/j.jns.2022.120207

      Highlights

      • MPDZ alterations are increasing at relapse after radio-chemotherapy in IDHwt glioblastoma.
      • MPDZ alterations could lead glioblastoma aggressiveness and therapeutic resistance.
      • MPDZ alterations are associated with pejorative outcome.

      Abstract

      Purpose

      We aimed to identify genomic drivers of glioblastoma inevitable recurrence.

      Methods

      Ten pairs of initial and recurrent frozen IDHwt glioblastoma samples were screened by CGH Array. Next Generation Sequencing (NGS) was then performed on an enriched cohort of 19 pairs. MPDZ alterations were analyzed using TCGA dataset.

      Results

      Nineteen IDHwt glioblastoma patients were included. Median age was 54.5 y/o (37.2–72.8). Using CGH array, unsupervised analysis aggregated the cohort by paired initial and recurrent tumors. Only 44% of CGH Array alterations were conserved at recurrence (amplifications: 55%; deletions: 30%). Two regions (including FPR1, 2 and 3) were lost at relapse: 19q13.33 and 19q13.41. MPDZ and 25 other genes were altered in ≥20% of recurrent tumors. NGS analysis of 29 candidate genes revealed 4 genes with pathogenic mutations: (FPR2, REL, TYRP1 and MPDZ). MPDZ (Multiple PDZ Domain Crumbs Cell Polarity Complex Component) was altered by two pathogenic mutations occurring at relapse. Using TCGA dataset we observed that a lower MPDZ mRNA expression was associated with IDHwt (p < 0.001) and grade IV (p < 0.001) gliomas. Finally, a low mRNA MPDZ expression was significantly correlated to poor overall survival in both IDHwt and IDH mutated gliomas, reinforcing the potential pejorative impact of MPDZ loss.

      Conclusion

      Our results suggest that MPDZ is more frequently altered at relapse after radio-chemotherapy in glioblastoma IDHwt patients, suggesting that MPDZ impairment could contribute to the systematic resistance of these tumors opening new therapeutic perspectives.

      Keywords

      1. Introduction

      IDHwt glioblastoma (GB) is the most frequently occurring primary brain tumor among adults and is one of the most lethal tumors. To date, recurrence is inevitable in a median delay of 7 to 10 months [
      • Lapointe S.
      • Perry A.
      • Butowski N.A.
      Primary brain tumours in adults.
      ]. Standard of care (SOC) in the first-line setting is based on the association of radiotherapy with concomitant and adjuvant temozolomide [
      • Stupp R.
      • Mason W.P.
      • van den Bent M.J.
      • Weller M.
      • Fisher B.
      • Taphoorn M.J.B.
      • Belanger K.
      • Brandes A.A.
      • Marosi C.
      • Bogdahn U.
      • Curschmann J.
      • Janzer R.C.
      • Ludwin S.K.
      • Gorlia T.
      • Allgeier A.
      • Lacombe D.
      • Cairncross J.G.
      • Eisenhauer E.
      • Mirimanoff R.O.
      European Organisation for Research and Treatment of Cancer Brain Tumor and Radiotherapy Groups, National Cancer Institute of Canada Clinical Trials Group, Radiotherapy plus concomitant and adjuvant temozolomide for glioblastoma.
      ]. Biologically, GBs are characterized by extensive angiogenesis, leading to the evaluation of antiangiogenic therapy in this setting. However, the addition of bevacizumab as first line treatment or at recurrence improved progression-free survival and quality of life but failed to significantly improve overall survival (OS). At the same time, no innovative targeting therapies such as ABT-414 or immunotherapies were able to improve OS of GB patients and the identification of relevant targets remains crucial. Molecular profile of GB was recently explored in different patient cohorts aiming to identify early molecular alterations leading to glioma occurrence and development. IDHwt GB are characterized by the loss of chromosome 10q, the gain of chromosome 7p, EGFR amplification, TERT promoter, PTEN mutation and CDKN2A deletion. However, the specific actors of aggressiveness and treatment resistance, particularly after radio-chemotherapy, remain debated despite their identification would open new therapeutic opportunities to prevent the systematic relapse. Some previous studies performed genomic analyses [
      • Kim J.
      • Lee I.-H.
      • Cho H.J.
      • Park C.-K.
      • Jung Y.-S.
      • Kim Y.
      • Nam S.H.
      • Kim B.S.
      • Johnson M.D.
      • Kong D.-S.
      • Seol H.J.
      • Lee J.-I.
      • Joo K.M.
      • Yoon Y.
      • Park W.-Y.
      • Lee J.
      • Park P.J.
      • Nam D.-H.
      Spatiotemporal evolution of the primary glioblastoma genome.
      ,
      • Fp B.
      • Kc J.
      • Fs V.
      • Ad M.
      • T. G, K. E, A. Kj, A. O, A. K, A. Kd, A. D, A. Sb, A. Dm, B. P, B.-S. Js, B. R, B. C, B. Pk, B. Dj, B. Ar, B. Af, B. Kr, C. A, C. A, C. Jh, C. Eb, C. Ej, C. J, C. Jf, F. G, F. Mn, F. Pj, G. Hk, G. Mr, G. Pv, G. Mr, I. A, I. A, J. Md, K. M, K. H, K. Mcm, L. Ps, L. M, L. P, L. Kl, L. Ak, M. Tm, M. T, M. Kl, M. Am, N. Dh, N. N, N. Hk, N. Cy, N. Sp, N. Jm, N. H, N. J, O. Dr, P. Ck, P. Lm, R. R, R. B, R. G, R. G, S. Jk, S. M, S. Bl, S. Sc, S. Pas, S. Ae, S. M, S. H, T. G, V.M. Eg, W. C, W. M, W. P, W. Ba, W. G, W. A, Y. Wka, Z. G, H. Jt, D.G. Jf, S. Lf, V. Rgw
      Longitudinal molecular trajectories of diffuse glioma in adults.
      ,
      • Draaisma K.
      • Chatzipli A.
      • Taphoorn M.
      • Kerkhof M.
      • Weyerbrock A.
      • Sanson M.
      • Hoeben A.
      • Lukacova S.
      • Lombardi G.
      • Leenstra S.
      • Hanse M.
      • Fleischeuer R.
      • Watts C.
      • McAbee J.
      • Angelopoulos N.
      • Gorlia T.
      • Golfinopoulos V.
      • Kros J.M.
      • Verhaak R.G.W.
      • Bours V.
      • van den Bent M.J.
      • McDermott U.
      • Robe P.A.
      • French P.J.
      Molecular evolution of IDH wild-type glioblastomas treated with standard of care affects survival and design of precision medicine trials: a report from the EORTC 1542 study.
      ], transcriptomic analyses [
      • Wang Q.
      • Hu B.
      • Hu X.
      • Kim H.
      • Squatrito M.
      • Scarpace L.
      • deCarvalho A.C.
      • Lyu S.
      • Li P.
      • Li Y.
      • Barthel F.
      • Cho H.J.
      • Lin Y.-H.
      • Satani N.
      • Martinez-Ledesma E.
      • Zheng S.
      • Chang E.
      • Sauvé C.-E.G.
      • Olar A.
      • Lan Z.D.
      • Finocchiaro G.
      • Phillips J.J.
      • Berger M.S.
      • Gabrusiewicz K.R.
      • Wang G.
      • Eskilsson E.
      • Hu J.
      • Mikkelsen T.
      • DePinho R.A.
      • Muller F.
      • Heimberger A.B.
      • Sulman E.P.
      • Nam D.-H.
      • Verhaak R.G.W.
      Tumor evolution of glioma intrinsic gene expression subtype associates with immunological changes in the microenvironment.
      ] or combined both analyses [
      • Wang J.
      • Cazzato E.
      • Ladewig E.
      • Frattini V.
      • Rosenbloom D.I.S.
      • Zairis S.
      • Abate F.
      • Liu Z.
      • Elliott O.
      • Shin Y.-J.
      • Lee J.-K.
      • Lee I.-H.
      • Park W.-Y.
      • Eoli M.
      • Blumberg A.J.
      • Lasorella A.
      • Nam D.-H.
      • Finocchiaro G.
      • Iavarone A.
      • Rabadan R.
      Clonal evolution of glioblastoma under therapy.
      ] on paired samples. However, they failed to identify specific and reproducible recurrence-related signature. Different issues should be considered to improve identification of this signature including the molecular heterogeneity of the previous cohorts in regard to IDH status; the treatment variability between the first and second surgery and finally the timing variability of the samples collection (collected at baseline, at the first relapse or after several treatment lines).
      Multiple PDZ Domain Crumbs Cell Polarity Complex Component (MPDZ) is implicated in G protein-coupled receptors and cell-cell interactions and cell polarity. It has recently been identified as a tumor suppressor gene in cancer while it is also reported to be involved in hydrocephalus or neurological dysfunctions [
      • Liu W.
      • Huang Y.
      • Wang D.
      • Han F.
      • Chen H.
      • Chen J.
      • Jiang X.
      • Cao J.
      • Liu J.
      MPDZ as a novel epigenetic silenced tumor suppressor inhibits growth and progression of lung cancer through the Hippo-YAP pathway.
      ]. In this context, our objective was to analyze the evolutive profile of GB genomic alterations between the initial and the recurrent settings in paired samples of IDHwt GB patients treated by radio-chemotherapy (Stupp protocol) as first-line treatment. Using CGH array (CGHa) and Next Generation Sequencing (NGS) we showed that MPDZ alterations increased at relapse, both through deletion and pathogenic mutations occurring or increasing on recurrent samples. Then, we analyzed MPDZ alteration profile using TCGA dataset and showed that a low MPDZ expression was associated with pejorative patient characteristics and prognosis, reinforcing the interest for this putative target.

      2. Materials and methods

      2.1 Patients and tumor samples (Suppl. Fig. 1)

      Patients ≥18 years old with a diagnosis of IDHwt GB, who underwent a minimum of two surgical resections (at initial diagnosis and at first recurrence) were retrospectively identified from the authors' tumor bank (Assistance Publique-Hôpitaux de Marseille [APHM], Timone, Marseille, France, AC-2018-31,053; CRB BB-0033-00097). Diagnoses were established according to the 2016 World Health Organization classifications [
      • Louis D.N.
      • Perry A.
      • Reifenberger G.
      • von Deimling A.
      • Figarella-Branger D.
      • Cavenee W.K.
      • Ohgaki H.
      • Wiestler O.D.
      • Kleihues P.
      • Ellison D.W.
      The 2016 World Health Organization classification of tumors of the central nervous system: a summary.
      ]. All frozen samples were stored in the APHM Center of Biological Resources. Histological review of the frozen samples (DFB) confirmed the neoplastic nature of the tissue. Methylation of the MGMT promoter and IDH1/2 mutations were evaluated as previously described [
      • Figarella-Branger D.
      • Mokhtari K.
      • Dehais C.
      • Jouvet A.
      • Uro-Coste E.
      • Colin C.
      • Carpentier C.
      • Forest F.
      • Maurage C.-A.
      • Vignaud J.-M.
      • Polivka M.
      • Lechapt-Zalcman E.
      • Eimer S.
      • Viennet G.
      • Quintin-Roué I.
      • Aubriot-Lorton M.-H.
      • Diebold M.-D.
      • Loussouarn D.
      • Lacroix C.
      • Rigau V.
      • Laquerrière A.
      • Vandenbos F.
      • Michalak S.
      • Sevestre H.
      • Peoch M.
      • Labrousse F.
      • Christov C.
      • Kemeny J.-L.
      • Chenard M.-P.
      • Chiforeanu D.
      • Ducray F.
      • Idbaih A.
      • Network P.O.L.A.
      Mitotic index, microvascular proliferation, and necrosis define 3 groups of 1p/19q codeleted anaplastic oligodendrogliomas associated with different genomic alterations.
      ,
      • Metellus P.
      • Coulibaly B.
      • Nanni I.
      • Fina F.
      • Eudes N.
      • Giorgi R.
      • Barrie M.
      • Chinot O.
      • Fuentes S.
      • Dufour H.
      • Ouafik L.
      • Figarella-Branger D.
      Prognostic impact of O6-methylguanine-DNA methyltransferase silencing in patients with recurrent glioblastoma multiforme who undergo surgery and carmustine wafer implantation: a prospective patient cohort.
      ]. All patients have provided informed consent for sample and data analyses. Tumor specimens were obtained after written consent and according to a protocol approved by the local institutional review board and ethics committee. The present study was conducted in accordance with the declaration of Helsinki.

      2.2 Treatment and clinical follow-up

      We retrospectively reviewed the patients included in the APHM Center of Biological Resources to identify all candidates with available frozen tumor sample, treated frontline for GB by maximal surgery then concomitant radio-chemotherapy and whose underwent new surgery at relapse. Patients with early progression within three months after radio-chemotherapy were excluded to avoid potential pseudo-progression. Clinical follow-up was performed every four weeks and magnetic resonance imaging every eight weeks by a senior physician. Disease evaluation was performed according to the RANO (Response assessment in neuro-oncology) [
      • Wen P.Y.
      • Macdonald D.R.
      • Reardon D.A.
      • Cloughesy T.F.
      • Sorensen A.G.
      • Galanis E.
      • Degroot J.
      • Wick W.
      • Gilbert M.R.
      • Lassman A.B.
      • Tsien C.
      • Mikkelsen T.
      • Wong E.T.
      • Chamberlain M.C.
      • Stupp R.
      • Lamborn K.R.
      • Vogelbaum M.A.
      • van den Bent M.J.
      • Chang S.M.
      Updated response assessment criteria for high-grade gliomas: response assessment in neuro-oncology working group.
      ] criteria. Since all but one of our patients presented with local relapse, we were not able to identify different pattern of alterations between local versus distant recurrences. Based on these criterias 19 patients (and so 38 samples) were considered eligible for this study (suppl Fig. 1).

      2.3 DNA extraction

      Total DNA was extracted using TriPrep NucleoSpin® (Macherey-Nagel, Germany), according to the manufacturer's instructions. DNA was analyzed on the Nanodrop spectrophotometer and Agilent 2100 bioanalyzer (Agilent Technologies, Massy, France) for quantitative and qualitative controls.

      2.4 CGHarray

      A wide screening for genomic alteration was done on ten patients (20 samples) of matched samples followed by non-supervised analysis clustered samples by patients (initial and relapse). DNA copy number alterations (CNA) were determined by using high-resolution CGH microarrays (SurePrint G3 Human 4 × 180, Agilent, France) as previously described [
      • Adélaïde J.
      • Finetti P.
      • Bekhouche I.
      • Repellini L.
      • Geneix J.
      • Sircoulomb F.
      • Charafe-Jauffret E.
      • Cervera N.
      • Desplans J.
      • Parzy D.
      • Schoenmakers E.
      • Viens P.
      • Jacquemier J.
      • Birnbaum D.
      • Bertucci F.
      • Chaffanet M.
      Integrated profiling of basal and luminal breast cancers.
      ]. Tumor DNA was cohybridized with a pool of 13 normal male DNA as reference. Scanning was done with Agilent Autofocus Dynamic Scanner (G2565BA, Agilent). Data analysis and visualization were done with CGH Analytics 3.4 software (Agilent). Data extraction (log2 ratio) was done from CGH analytics, while normalized and filtered log2 ratio were obtained from “Feature extraction” software (Agilent). We eliminated data generated by probes mapped to X and Y chromosomes. The final dataset included 161,068 unique probes covering 16,684 genes and intergenic regions according to the hg19/NCBI human genome mapping database (build 37). Data were analyzed using circular binary segmentation as implemented in the DNA copy R/Bioconductor package [
      • Olshen A.B.
      • Venkatraman E.S.
      • Lucito R.
      • Wigler M.
      Circular binary segmentation for the analysis of array-based DNA copy number data.
      ] with default parameters to translate intensity measurements in regions of equal copy number, each region being defined by at least five consecutive probes. Thus, each probe was assigned a segment value referred to as its “smoothed” value. We used two different threshold values (log2 ratio > |0.5|, and |1|) to distinguish low (gain/loss) from high (amplification/deletion) level CNAs respectively [
      • Adélaïde J.
      • Finetti P.
      • Bekhouche I.
      • Repellini L.
      • Geneix J.
      • Sircoulomb F.
      • Charafe-Jauffret E.
      • Cervera N.
      • Desplans J.
      • Parzy D.
      • Schoenmakers E.
      • Viens P.
      • Jacquemier J.
      • Birnbaum D.
      • Bertucci F.
      • Chaffanet M.
      Integrated profiling of basal and luminal breast cancers.
      ]. To identify altered regions, we used the GISTIC 2.0 (v2.0.21) algorithm [
      • Mermel C.H.
      • Schumacher S.E.
      • Hill B.
      • Meyerson M.L.
      • Beroukhim R.
      • Getz G.
      GISTIC2.0 facilitates sensitive and confident localization of the targets of focal somatic copy-number alteration in human cancers.
      ], which computes for each segment through the genome a score based on the frequency of CNA combined with its amplitude, with bootstrapping to calculate the significance level (q < 0.25). The Database for Annotation, Visualization and Integrated Discovery (DAVID) [
      • Huang D.W.
      • Sherman B.T.
      • Lempicki R.A.
      Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources.
      ,
      • Huang D.W.
      • Sherman B.T.
      • Lempicki R.A.
      Bioinformatics enrichment tools: paths toward the comprehensive functional analysis of large gene lists.
      ] was used to identify relevant altered genes and to determine pathways implicated in recurrence after radio-chemotherapy.

      2.5 Library preparation and sequencing using NGS

      Twenty-nine relevant genes identified by DAVID or with recurrent CGHa at relapse were analyzed by NGS in 38 samples (19 patients). Libraries, including the target regions corresponding to the genes of interest (Suppl. Table 1), were performed with reagents from a custom design lon AmpliSeq Library Kit 2.0 (Thermo Fisher Scientific ®, USA). The designed primers for the total target covered more than 98,3% of the coding regions, as well as 20 base-pair regions flanking exon-intron boundaries, corresponding to 608 amplicons and 133,36 kilobase pairs. Libraries were quantified and qualified using the Qubit Fluorometer (Thermo Fisher Scientific ®, USA) and the Agilent 2100 Bioanalyzer instrument (High Sensitivity DNA Kit, Agilent®, USA) to enable equi-molar pooling of barcoded samples. Template preparation, emulsion PCR, and Ion Sphere Particles (ISP) enrichment were carried out using the Ion PGM Template OT2 200 Kit on the Ion OneTouchTM 2 System (Thermo Fisher Scientific ®, USA). The quality of the ISPs was assessed using a Qubit 2.0 Fluorometer, and the ISPs were loaded and sequenced on an Ion 318 Chip Kit v2 using Ion PGM Sequencing 200 Kit v2 on the Ion PGM Sequencer (Thermo Fisher Scientific ®, USA). Raw data were first aligned with the provided software suite included with the Ion PGM system to generate BAM files (hg19). The coverage and sequencing depth analysis were computed using the BEDtools suite v2.17 [
      • Quinlan A.R.
      • Hall I.M.
      BEDTools: a flexible suite of utilities for comparing genomic features.
      ] and in-house scripts. With a mean coverage of depth at 1000×, 95% of target sequences were successfully covered with a depth > 100×, and 70% > 500×. Variants were identified using the Torrent Browser Variant caller (version 4.0.2), annotated and prioritized with the VarAFT system (http://varaft.eu) that includes Annovar [
      • Wang K.
      • Li M.
      • Hakonarson H.
      ANNOVAR: functional annotation of genetic variants from high-throughput sequencing data.
      ] Prediction of functional effects of of non-synonymous SNPs was done with UMD predictor [
      • Salgado D.
      • Desvignes J.-P.
      • Rai G.
      • Blanchard A.
      • Miltgen M.
      • Pinard A.
      • Lévy N.
      • Collod-Béroud G.
      • Béroud C.
      UMD-predictor: a high-throughput sequencing compliant system for pathogenicity prediction of any human cDNA substitution.
      ], MutationTaster and PolyPhen. Variants of interest were verified using the Integrative Genomics Viewer.

      2.6 Immunohistochemistry (IHC)

      Five micrometers formalin fixed paraffin embedded (FFPE) slides of initial and recurrent tumor samples from 17 patients were labeled with anti-MPDZ antibody (rabbit polyclonal antibody used at 1/50 dilution, HPA020255, Sigma Aldrich). Staining was performed on a Benchmark XT (Ventana Medical systems, Illkirch, France) according to manufacturer's instructions. A semi-quantitative analysis was done by a neuropathologist (RA) to define the expression location and level (from 0: absent to 3: high) without knowledge of clinical data.

      2.7 TCGA and GEPIA analyses

      TCGA data of GB and low grade glioma cohort were analyzed using the GlioVis dataset [
      • Bowman R.L.
      • Wang Q.
      • Carro A.
      • Verhaak R.G.W.
      • Squatrito M.
      GlioVis data portal for visualization and analysis of brain tumor expression datasets.
      ] including 669 samples and the GlioVis recurrence dataset [
      • Wang Q.
      • Hu B.
      • Hu X.
      • Kim H.
      • Squatrito M.
      • Scarpace L.
      • deCarvalho A.C.
      • Lyu S.
      • Li P.
      • Li Y.
      • Barthel F.
      • Cho H.J.
      • Lin Y.-H.
      • Satani N.
      • Martinez-Ledesma E.
      • Zheng S.
      • Chang E.
      • Sauvé C.-E.G.
      • Olar A.
      • Lan Z.D.
      • Finocchiaro G.
      • Phillips J.J.
      • Berger M.S.
      • Gabrusiewicz K.R.
      • Wang G.
      • Eskilsson E.
      • Hu J.
      • Mikkelsen T.
      • DePinho R.A.
      • Muller F.
      • Heimberger A.B.
      • Sulman E.P.
      • Nam D.-H.
      • Verhaak R.G.W.
      Tumor evolution of glioma intrinsic gene expression subtype associates with immunological changes in the microenvironment.
      ]. TCGA dataset patient characteristics are described in Supp Table 3. Regarding mRNA analyses, the normalized count reads from the pre-processed data (sequence alignment and transcript abundance estimation) were log2 transformed after adding a 0.5 pseudocount (to avoid infinite value upon log transformation).
      Cell subgroup analyses were performed using the GEPIA interactive web tool [
      • Li C.
      • Tang Z.
      • Zhang W.
      • Ye Z.
      • Liu F.
      GEPIA2021: integrating multiple deconvolution-based analysis into GEPIA.
      ].

      2.8 Statistical analyses

      Data are expressed as mean ± standard error. Statistical analyses were performed using Student's t-test and the Wilcoxon test. The Mann-Whitney U test was used to compare quantitative variables. Correlations were analyzed using the Spearman correlation. Overall survival from the initial diagnosis (OS) was defined to be the time from initial diagnosis to death from any cause, censored at the date of last contact. Progression-free survival (PFS) was defined to be the time from initial diagnosis to progression or death, censored at the date of last contact. The Kaplan-Meier method was used to estimate survival distribution. Log-rank tests were used for univariate comparisons and Cox proportional hazard regression models were used to estimate the hazard ratio (HR) in multivariate analyses. All reported p values are two-sided, and p < 0.05 was considered to be statistically significant. Statistical analyses were performed using SPSS PASW statistics 22.0.

      3. Results

      3.1 Patient characteristics

      Nineteen patients with recurrent IDHwt GB for whom both primary and recurrent tumor samples available were included (Table 1). Initial surgeries were performed between 2003 and 2009. For all of them, first-line treatment consisted of maximal primary tumor resection followed by radiotherapy (60 Gy in 30 fractions) with concomitant temozolomide (75 mg/m2 daily) followed by adjuvant temozolomide (150 mg/m2 up to 200 mg/m2 for five days every 28 days) for 18 of 19 patients. One patient received BCNU (150 mg/m2 on day 1 every 6 weeks) instead of temozolomide. All patient underwent the second surgery at the first recurrence. Median age was 54.5 years [37.2–72.8], 42% were female, 72% had corticosteroid at baseline and Karnofsky performance status was 80% or higher in 51% (Table 1) . All patients experienced progression and only one patient was still alive at the last contact, with a follow-up of 60 months. Median PFS and OS were 9.2 months (95%CI: 9.2–9.3) and 23.4 months (95%CI: 11.6–35.1), respectively.
      Table 1Patient characteristics of paired cohort.
      CharacteristicsN%
      Age (median, range)54.5 (37.2–72.8)
      Gender (Female/Male)8 / 1142 / 58
      Initial KPS
       6016
       70633
       80844
       90317
      Steroids at baseline1372
      Molecular characteristics
      IDHwt19100
       Unmethylated MGMT promoter16/1794
      Gross total resection19100
      First line treatment
       Stupp protocol1895
       Radiotherapy + BCNU15

      3.2 CGH array (CGHa) highlights candidate genes alterations emerging at GB relapse

      First, 10 patients with initial and recurrent samples of GB were analyzed by CGHarray (Supplementary Fig. 1, Supplementary Fig. 2). Unsupervised analysis classified samples by paired tumors (Fig. 1A) . No significant genomic profile was associated with recurrent GB samples versus baseline samples. By gistic analyses, two regions (19q13.33 and 19q13.41) were more frequently lost at relapse compared with initial samples (p = 0.035). These both regions corresponded to 171 genes. The analysis of these genes by DAVID [
      • Huang D.W.
      • Sherman B.T.
      • Lempicki R.A.
      Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources.
      ,
      • Huang D.W.
      • Sherman B.T.
      • Lempicki R.A.
      Bioinformatics enrichment tools: paths toward the comprehensive functional analysis of large gene lists.
      ] highlights one pathway of neutrophils chemotactic factors including three genes: FPR-1, FPR-2 and FPR-3.
      Fig. 1
      Fig. 1Unsupervised analysis clustered the samples by paired of initial and recurrent glioblastoma samples. A- Heat map of genomic alterations over the 20 samples from paired tumors (N = 10 patients) obtained with CGH array and plotted as side to side in each column though each row representing a single gene. The ratio of abundance of each gene is represented: by blue square when the ratio is below, black when is equal and red when is above the median. B- Descriptive analysis of CGH array displaying the distribution of conserved alterations between initial and recurrent samples, alterations observed in the initial samples only or alterations observed in the recurrent samples only: N = 10 patients; 20 samples. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
      Then, global comparison of initial and recurrent alterations revealed that 44% of them were conserved between the initial and the recurrent samples. Fifty-five percents of amplifications were conserved while 30% of deletions were conserved (Fig. 1B). A fifth of gene amplification and half of the gene deletions observed on the initial samples were not observed on the paired recurrent samples. The proportion of new alterations occurring at relapse was comparable between amplifications and deletions (25% and 23% respectively). Among them, 2 genes amplification (PIGU and PLXNA1) occurred at relapse in at least 2 patients (≥ 20%) and 24 genes deletion occurred at relapse in at least 2 patients (≥ 20%, Suppl. Table 2). Taken together, these analyses allowed us to select 29 genes of interest (Suppl. Table 1) for further analyzes using NGS on the whole cohort.

      3.3 Next generation sequencing reveals recurrent alterations in MPDZ at GB relapse

      We performed the NGS analysis of the 29 candidates' genes (top genes) selected previously on the whole cohort of 19 pairs to identify recurrent alterations emerging at relapse, which could lead GB to inevitable progression and treatment resistance (Suppl. Fig. 1). We observed pathogenic mutations for 4 genes: FPR2, REL, TYRP1 and MPDZ (Table 2). For FPR2, REL and TYRP1, the alterations were missense mutations and were exonic. Proportion of mutated variants was stable between initial and recurrent samples (Table 2). For MPDZ, we observed missense mutations in two patients. In the first patient, mutated variant (c.5368G > A) was stable between initial and recurrent settings (47%). For the second patient, mutated variant became the majority variant at relapse (c.25C > T; 86%). In this same patient, another mutation affected MPDZ in the splice site at relapse only. Totally, in our cohort, MPDZ was altered by mutation or loss for 8/19 (42%) patients at initial diagnosis and 10/19 (53%) patients at relapse, including one patient for who the initial alteration was more severe at relapse (SuppTable 2). To note, MPDZ mutations were recorded for only 1% of newly diagnosed GB patients in the TCGA database while they affected 10% of our patients. This variation could be related to the size of our cohort.
      Table 2Next generation sequencing (NGS) results. Four genes (FPR2, REL, TYRP1 and MPDZ) are mutated. MPDZ is more frequently altered at relapse.
      Patients# locusInitialRecurrenceGeneLocationFunctionExonCodingUMD predictor14
      RefVariantRefVariant
      9chr19:

      52,272,632
      CTCTFPR2exonicmissense2c.721C > TProbably pathogenic
      70%30%69%31%
      9chr2:

      61,145,380
      GAGARELexonicmissense6c.590G > APathogenic
      53%47%49%51%
      9chr9:

      12,709,024
      GAGATYRP1exonicmissense8c.1456G > APathogenic
      50%50%51%49%
      11chr9:

      13,247,632
      AGAGMPDZSplice siteDonor site lost3
      100%0%87%13%
      11chr9:

      13,247,792
      GAGAMPDZexonicmissense3c.25C > TPathogenic
      52%48%14%86%
      15chr9:

      13,119,512
      CTCTMPDZexonicmissense39c.5368G > APathogenic
      52%48%53%47%

      3.4 Protein expression of MPDZ by immunohistochemistry

      Using immunohistochemistry, we analyzed the protein expression of MPDZ in initial and recurrent samples from 17 patients with IDHwt GB (12/17 were included in the previous CGHarray and NGS analyses, Suppl. Fig. 1). MPDZ was variably expressed by tumor cells but also by endothelial and microvascular proliferation cells (Fig. 2A). High MPDZ protein expression in tumor cells was significantly less frequent at relapse (p = 0.013) (Fig. 2B,C) while no difference was observed regarding its vascular expression.
      Fig. 2
      Fig. 2MPDZ protein expression is less intense at relapse. A- MPDZ protein expression characteristics. B- Intense MPDZ immunostaining sample fraction at initial diagnosis and recurrence. C- Illustrative case of initial and recurrent samples from patient #4.

      3.5 MPDZ mRNA expression in TCGA database and GEPIA web tool

      To explore the mRNA expression of MPDZ, we analyzed the open source database of the TCGA using the GB and low grade glioma GlioVis dataset [
      • Bowman R.L.
      • Wang Q.
      • Carro A.
      • Verhaak R.G.W.
      • Squatrito M.
      GlioVis data portal for visualization and analysis of brain tumor expression datasets.
      ] (Suppl. Table 3). At diagnosis, mRNA expression of MPDZ was significantly lower in grade IV glioma (p > 0.001) compared to grades II or III (Fig. 3A ). In IDHwt GB, the older patients (> 60 years) presented with a lower MPDZ expression (p = 0.028, Fig. 3B) Moreover, mRNA expression of MPDZ was different between GB molecular subtypes (mesenchymal, classic and proneural) (Fig. 3C). In contrast, mRNA expression of MPDZ was independent from patient gender and MGMT promotor methylation status. Regarding the prognostic value of MPDZ expression, in univariate analyses, a lower expression of MPDZ was associated with a worse prognosis (p = 0.049, Fig. 3D). Finally, MPDZ expression was significanlty lower in IDHwt GB (Suppl. Fig. 4A) than in IDH mutated (IDHm) GB (p < 0.001). In IDHm GB, older age was associated with a lower MPDZ expression (p = 0.022), while a lower MPDZ expression was also associated with a worse outcome (p = 0.040; Suppl. Fig. 4B-C). Regarding the microenvironment cell expression at diagnosis, we observed that MPDZ was more expressed in vascular cells than in immune cells or fibroblasts where the transcript was almost absent (Suppl. Fig. 5).
      Fig. 3
      Fig. 3mRNA expression of MPDZ in IDHwt glioblastoma (TCGA dataset). A- mRNA expression of MPDZ according to sample grade. B- mRNA expression of MPDZ according to patient age. C- mRNA expression of MPDZ according to GB molecular subtypes. D- Overall survival of patients according to MPDZ mRNA expression. * < 0.05; *** < 0.001.

      4. Discussion

      In the present study, we analyzed the genomic profile of paired samples of IDHwt GB at initial diagnosis and at first relapse after radio-chemotherapy. Using CGHa we identified 29 genes that could be preferentially involved in GB recurrence and using NGS we observed complementary pathogenic mutations affecting MPDZ.
      Due to the deep tumor infiltration in the central nervous system and despite maximal surgery, GB relapses systematically leading to fatal outcome [
      • Lapointe S.
      • Perry A.
      • Butowski N.A.
      Primary brain tumours in adults.
      ]. Then, our aim was to describe the molecular alterations associated with IDHwt GB at relapse using comparative genomic analyses with their initial tumor sample counterpart to identify innovant and relevant therapeutic targets.
      First, we observed that only 44% of genomic alterations were conserved at recurrence. Around 25% of alterations occurred at relapse arguing for a clonal evolution and resistance acquisition, under the pressure of radiation and alkylating treatments. In contrast, the 7p/10q alterations remained stable in majority during the disease evolution reinforcing the importance of these alterations and their potential early occurrence in GB development. Among putative oncogenic drivers located on these chromosome arms, we can interestingly find EGFR gain (7p) or PTEN loss (10q). Unsupervised analysis classified all the samples by paired tissue, suggesting that despite molecular evolution between initial and recurrent setting, inter-GB heterogeneity remained significant. This result is in line with previous publications highlighting cellular and molecular heterogeneity of GB, that enhances therapeutic complexity [
      • Wang Z.
      • Sun D.
      • Chen Y.-J.
      • Xie X.
      • Shi Y.
      • Tabar V.
      • Brennan C.W.
      • Bale T.A.
      • Jayewickreme C.D.
      • Laks D.R.
      • Alcantara Llaguno S.
      • Parada L.F.
      Cell lineage-based stratification for glioblastoma.
      ].
      Using CGHa, we identified 29 genes potentially involved in aggressiveness and treatment resistance of GB. Among them, three were implicated in neutrophil chemotactic pathway and one of them, MPDZ (involved in tight junctions and angiogenesis), appeared to be also mutated more severely at relapse than at baseline.
      Comparative molecular analyses of recurrent GB samples are crucial to improve our patient management. We know that cancers present with molecular evolution during their expansion and after treatment, especially when it is composed by DNA-damaging agents such as chemotherapy and radiotherapy. Because the vast majority of new therapies are first evaluated at recurrence, the precise characterization of the targets at this time is required to be able to interpret clinical trials results. However, such analyses are extremely challenging in neuro-oncology when few patients are candidate for a second surgery. Then, a limited number of publications explored previously this major question. Among them, three focused on GB [
      • Draaisma K.
      • Chatzipli A.
      • Taphoorn M.
      • Kerkhof M.
      • Weyerbrock A.
      • Sanson M.
      • Hoeben A.
      • Lukacova S.
      • Lombardi G.
      • Leenstra S.
      • Hanse M.
      • Fleischeuer R.
      • Watts C.
      • McAbee J.
      • Angelopoulos N.
      • Gorlia T.
      • Golfinopoulos V.
      • Kros J.M.
      • Verhaak R.G.W.
      • Bours V.
      • van den Bent M.J.
      • McDermott U.
      • Robe P.A.
      • French P.J.
      Molecular evolution of IDH wild-type glioblastomas treated with standard of care affects survival and design of precision medicine trials: a report from the EORTC 1542 study.
      ,
      • Wang Q.
      • Hu B.
      • Hu X.
      • Kim H.
      • Squatrito M.
      • Scarpace L.
      • deCarvalho A.C.
      • Lyu S.
      • Li P.
      • Li Y.
      • Barthel F.
      • Cho H.J.
      • Lin Y.-H.
      • Satani N.
      • Martinez-Ledesma E.
      • Zheng S.
      • Chang E.
      • Sauvé C.-E.G.
      • Olar A.
      • Lan Z.D.
      • Finocchiaro G.
      • Phillips J.J.
      • Berger M.S.
      • Gabrusiewicz K.R.
      • Wang G.
      • Eskilsson E.
      • Hu J.
      • Mikkelsen T.
      • DePinho R.A.
      • Muller F.
      • Heimberger A.B.
      • Sulman E.P.
      • Nam D.-H.
      • Verhaak R.G.W.
      Tumor evolution of glioma intrinsic gene expression subtype associates with immunological changes in the microenvironment.
      ,
      • Wang J.
      • Cazzato E.
      • Ladewig E.
      • Frattini V.
      • Rosenbloom D.I.S.
      • Zairis S.
      • Abate F.
      • Liu Z.
      • Elliott O.
      • Shin Y.-J.
      • Lee J.-K.
      • Lee I.-H.
      • Park W.-Y.
      • Eoli M.
      • Blumberg A.J.
      • Lasorella A.
      • Nam D.-H.
      • Finocchiaro G.
      • Iavarone A.
      • Rabadan R.
      Clonal evolution of glioblastoma under therapy.
      ]. It should be noticed that these patients' cohorts contained both IDHm and IDHwt GB, with heterogeneous treatments, at the exception of the EORTC study [
      • Draaisma K.
      • Chatzipli A.
      • Taphoorn M.
      • Kerkhof M.
      • Weyerbrock A.
      • Sanson M.
      • Hoeben A.
      • Lukacova S.
      • Lombardi G.
      • Leenstra S.
      • Hanse M.
      • Fleischeuer R.
      • Watts C.
      • McAbee J.
      • Angelopoulos N.
      • Gorlia T.
      • Golfinopoulos V.
      • Kros J.M.
      • Verhaak R.G.W.
      • Bours V.
      • van den Bent M.J.
      • McDermott U.
      • Robe P.A.
      • French P.J.
      Molecular evolution of IDH wild-type glioblastomas treated with standard of care affects survival and design of precision medicine trials: a report from the EORTC 1542 study.
      ]. In the first study, Wang et al., observed a highly branched evolutionary pattern with frequent expression-based subtype changes. They reported that 11% of recurrent tumors harbored mutations in LTBP4, suggesting that TGF-β pathway could be a potential therapeutic target in a sub-group of GB. Finally, they observed some hypermutated tumors at relapse with frequent alterations of MSH6 or other DNA reparation genes. In another study, Wang et al., analyzed the transcriptomic profile of 78 glioma pairs and showed that expression subtype was retained in 55% of cases. Interestingly, they also found significant modifications in the microenvironment of recurrent tumors, identifying a decrease in monocyte gene signature concomitant of an upregulation in M2 macrophage infiltration in mesenchymal phenotype at relapse [
      • Wang Q.
      • Hu B.
      • Hu X.
      • Kim H.
      • Squatrito M.
      • Scarpace L.
      • deCarvalho A.C.
      • Lyu S.
      • Li P.
      • Li Y.
      • Barthel F.
      • Cho H.J.
      • Lin Y.-H.
      • Satani N.
      • Martinez-Ledesma E.
      • Zheng S.
      • Chang E.
      • Sauvé C.-E.G.
      • Olar A.
      • Lan Z.D.
      • Finocchiaro G.
      • Phillips J.J.
      • Berger M.S.
      • Gabrusiewicz K.R.
      • Wang G.
      • Eskilsson E.
      • Hu J.
      • Mikkelsen T.
      • DePinho R.A.
      • Muller F.
      • Heimberger A.B.
      • Sulman E.P.
      • Nam D.-H.
      • Verhaak R.G.W.
      Tumor evolution of glioma intrinsic gene expression subtype associates with immunological changes in the microenvironment.
      ]. They did not observe a decreasing neutrophil recrutment at relapse. In our study, we observed that three neutrophil chemotactic factor genes were more frequently lost at relapse but neutrophil recrutment is certainly lead by complementary pathways [
      • SenGupta S.
      • Hein L.E.
      • Parent C.A.
      The recruitment of neutrophils to the tumor microenvironment is regulated by multiple mediators.
      ]. Finally, the EORTC study sequenced 287 genes and observed that mutation status changed in approximatively 20% of events with changes observed for all examined genes. They underlined that these changes could strongly affect targeted trial results and then size and design. A last study generated through the GLASS consortium analyzed the genomic profile of more than 200 gliomas including half of IDHwt GBs [
      • Fp B.
      • Kc J.
      • Fs V.
      • Ad M.
      • T. G, K. E, A. Kj, A. O, A. K, A. Kd, A. D, A. Sb, A. Dm, B. P, B.-S. Js, B. R, B. C, B. Pk, B. Dj, B. Ar, B. Af, B. Kr, C. A, C. A, C. Jh, C. Eb, C. Ej, C. J, C. Jf, F. G, F. Mn, F. Pj, G. Hk, G. Mr, G. Pv, G. Mr, I. A, I. A, J. Md, K. M, K. H, K. Mcm, L. Ps, L. M, L. P, L. Kl, L. Ak, M. Tm, M. T, M. Kl, M. Am, N. Dh, N. N, N. Hk, N. Cy, N. Sp, N. Jm, N. H, N. J, O. Dr, P. Ck, P. Lm, R. R, R. B, R. G, R. G, S. Jk, S. M, S. Bl, S. Sc, S. Pas, S. Ae, S. M, S. H, T. G, V.M. Eg, W. C, W. M, W. P, W. Ba, W. G, W. A, Y. Wka, Z. G, H. Jt, D.G. Jf, S. Lf, V. Rgw
      Longitudinal molecular trajectories of diffuse glioma in adults.
      ]. They reported that the driver genes detected at initial disease, like the alterations of chromosome 7 and 10 were mainly retained at recurrence, as we also observed, and that alkylating-agents resulted in hypermutator phenotype. However, they did not report recurrence-specific gene alterations. Compared to these previous studies, our cohort is smaller but presents with the advantage to include only IDHwt GB patients all treated with radio-chemotherapy regimen. Then our results are based on a homogenous cohort and the potential extrapolation of our results would be easier. Moreover, we performed a double complementary genomic approach by starting with a CGHa screening then completed by NGS. Finally, our results were reinforced by our analyses of the large TCGA datasets, including the transcriptomic dataset of GB pairs analyzed by Wang et al. [
      • Wang Q.
      • Hu B.
      • Hu X.
      • Kim H.
      • Squatrito M.
      • Scarpace L.
      • deCarvalho A.C.
      • Lyu S.
      • Li P.
      • Li Y.
      • Barthel F.
      • Cho H.J.
      • Lin Y.-H.
      • Satani N.
      • Martinez-Ledesma E.
      • Zheng S.
      • Chang E.
      • Sauvé C.-E.G.
      • Olar A.
      • Lan Z.D.
      • Finocchiaro G.
      • Phillips J.J.
      • Berger M.S.
      • Gabrusiewicz K.R.
      • Wang G.
      • Eskilsson E.
      • Hu J.
      • Mikkelsen T.
      • DePinho R.A.
      • Muller F.
      • Heimberger A.B.
      • Sulman E.P.
      • Nam D.-H.
      • Verhaak R.G.W.
      Tumor evolution of glioma intrinsic gene expression subtype associates with immunological changes in the microenvironment.
      ], reinforcing the relevance of our results.
      MPDZ for “Multiple PDZ domain protein” is a scaffolds membrane protein with multiple PDZ interacting domains, highly expressed in central nervous system, participating to the tight junction, and enriched at the synaptic zones. When constitutively mutated, it causes congenital hydrocephalus. Few data suggested a potential role of MPDZ in cancer: a lower MPDZ expression was associated with a higher risk of breast cancer recurrence and a poor survival in renal cell carcinoma [
      • Huang Y.-S.
      • Liu W.-B.
      • Han F.
      • Yang J.-T.
      • Hao X.-L.
      • Chen H.-Q.
      • Jiang X.
      • Yin L.
      • Ao L.
      • Cui Z.-H.
      • Cao J.
      • Liu J.-Y.
      Copy number variations and expression of MPDZ are prognostic biomarkers for clear cell renal cell carcinoma.
      ,
      • Martin T.A.
      • Watkins G.
      • Mansel R.E.
      • Jiang W.G.
      Loss of tight junction plaque molecules in breast cancer tissues is associated with a poor prognosis in patients with breast cancer.
      ]. Mutation or deletion impairing the stabilizing role of MPDZ could increase the migration and invasive capacities of the tumor cells. Recent data have associated MPDZ with angiogenesis in cellular and embryonic mouse hindbrain models, via the stabilization of DLL1 and DDL4 proteins and by increasing Notch-signaling pathway [
      • Tetzlaff F.
      • Adam M.G.
      • Feldner A.
      • Moll I.
      • Menuchin A.
      • Rodriguez-Vita J.
      • Sprinzak D.
      • Fischer A.
      MPDZ promotes DLL4-induced notch signaling during angiogenesis.
      ]. This pathway activation, known to promote tumor growth and endothelial neoangiogenesis [
      • Kopan R.
      • Ilagan Ma.X.G.
      The canonical notch signaling pathway: unfolding the activation mechanism.
      ], could be associated with resistance to anti-VEGF therapy, opening new therapeutic opportunities [
      • Noguera-Troise I.
      • Daly C.
      • Papadopoulos N.J.
      • Coetzee S.
      • Boland P.
      • Gale N.W.
      • Chieh Lin H.
      • Yancopoulos G.D.
      • Thurston G.
      Blockade of Dll4 inhibits tumour growth by promoting non-productive angiogenesis.
      ]. More recently, MPDZ knockdown was reported to promote cell proliferation, migration and invasion in vitro and in vivo in preclinical lung cancer models. MPDZ deficiency promoted also tumor metastases and reduced mice survival. Interestingly, the authors showed that MPDZ acted as a tumor suppressor through the Hippo-YAP pathway [
      • Liu W.
      • Huang Y.
      • Wang D.
      • Han F.
      • Chen H.
      • Chen J.
      • Jiang X.
      • Cao J.
      • Liu J.
      MPDZ as a novel epigenetic silenced tumor suppressor inhibits growth and progression of lung cancer through the Hippo-YAP pathway.
      ].
      Our study has some limitation including the limited size of our cohort. This size could explain the variations we observed between our results and the TCGA database. However, our cohort is highly homogenous and composed by frozen samples from paired GB.

      5. Conclusion

      To conclude, we have identified by paired genomic analysis of initial and recurrent IDHwt GB heterogeneous molecular profiles at relapse but with recurrent MPDZ alterations, highlighting its potential role in GB recurrence. Further investigations are needed to confirm these results and develop potential therapeutic strategies.
      Supplementary Fig. 1
      Supplementary Fig. 2
      Supplementary Fig. 2Genome analysis between initial and relapse samples. Tumor genomic profile established by CGH array and plotted as whole genomic profile with various copy number alterations observed for each chromosome, with superposition between initial profile (orange) and relapse profile (blue) is shown (one illustrative patient).
      Supplementary Fig. 3
      Supplementary Fig. 3Representative images of semi-quantitative immunostaining scale: 0: absent; 1: low; 2: intermediate; 3: intense.
      Supplementary Fig. 4
      Supplementary Fig. 4mRNA expression of MPDZ in IDH mutated gliomas. A- mRNA expression of MPDZ between IDH mutated and IDH wild-type samples. B- mRNA expression of MPDZ according to patient age. C- Overall survival of patients according to MPDZ mRNA expression. * < 0.005; *** < 0.001.
      Supplementary Fig. 5
      Supplementary Fig. 5MPDZ transcript expression was quantitatively compared in different cell subsets (B cells, T cells, macrophages, endothelial cells and cancer associated fibroblasts) across 166 samples of GB (TCGA expression data set) by using GEPIA2021 interactive web tool. The F-test (one-way ANOVA) identifies the degree of difference, along with the p-value indicating the significance. TPM: Transcripts Per Million.
      • Supplementary material

        Supplementary Table 1 – List of genes of interest (top genes) after CGH array analysis.

        Supplementary table 2 – MPDZ status in the complete cohort. Del.: deletion.

        Supplementary table 3– TCGA dataset patient characteristics (N=669).

      Author contributions

      Conceptualization, E.T.; methodology, C.J.J., D.F.B., O.C., E.T.; software, A.G., J.A., E.D.; validation, E.T., D.F.B., O.C.; formal analysis, BC., R.A., A.G., J.A., C·C, E.D. C.J.J., D.F.B., O.C., E.T.; investigation, R.A., C.J.J., D.F.B., C.B., I.N.M., S.B., T.G., H.D., E.T.; resources, E.T. writing—original draft preparation, B.C., E.T., D.F.B.; writing—review and editing, All Authors.; supervision, E.T; funding acquisition, E.T. All authors have read and agreed to the published version of the manuscript.” +.

      Funding

      This research was funded by ARTC-Sud patient association.

      Institutional review board statement

      The study was conducted according to the guidelines of the Declaration of Helsinki and approved by the Institutional Review Board of Assistance Publique.

      Informed consent statement

      Informed consent was obtained from all patients involved in the study.

      Availability of data and material

      CGH data have been deposited in the ArrayExpress database at EMBL-EBI (www.ebi.ac.uk/arrayexpress) under accession number E-MTAB-10079.

      Declaration of competing interest

      The authors declare no conflict of interest.

      Acknowledgments

      We thank the AP-HM Center of Biological Resources (authorization number: AC2018-31053 ; CRB BB-0033-00097 ) for providing tissue samples, the Cancéropôle PACA and the ARTC-Sud patient association (Association pour le Recherche sur les Tumeurs Cérébrales).

      References

        • Lapointe S.
        • Perry A.
        • Butowski N.A.
        Primary brain tumours in adults.
        Lancet. 2018; 392: 432-446https://doi.org/10.1016/S0140-6736(18)30990-5
        • Stupp R.
        • Mason W.P.
        • van den Bent M.J.
        • Weller M.
        • Fisher B.
        • Taphoorn M.J.B.
        • Belanger K.
        • Brandes A.A.
        • Marosi C.
        • Bogdahn U.
        • Curschmann J.
        • Janzer R.C.
        • Ludwin S.K.
        • Gorlia T.
        • Allgeier A.
        • Lacombe D.
        • Cairncross J.G.
        • Eisenhauer E.
        • Mirimanoff R.O.
        European Organisation for Research and Treatment of Cancer Brain Tumor and Radiotherapy Groups, National Cancer Institute of Canada Clinical Trials Group, Radiotherapy plus concomitant and adjuvant temozolomide for glioblastoma.
        N. Engl. J. Med. 2005; 352: 987-996https://doi.org/10.1056/NEJMoa043330
        • Kim J.
        • Lee I.-H.
        • Cho H.J.
        • Park C.-K.
        • Jung Y.-S.
        • Kim Y.
        • Nam S.H.
        • Kim B.S.
        • Johnson M.D.
        • Kong D.-S.
        • Seol H.J.
        • Lee J.-I.
        • Joo K.M.
        • Yoon Y.
        • Park W.-Y.
        • Lee J.
        • Park P.J.
        • Nam D.-H.
        Spatiotemporal evolution of the primary glioblastoma genome.
        Cancer Cell. 2015; 28: 318-328https://doi.org/10.1016/j.ccell.2015.07.013
        • Fp B.
        • Kc J.
        • Fs V.
        • Ad M.
        • T. G, K. E, A. Kj, A. O, A. K, A. Kd, A. D, A. Sb, A. Dm, B. P, B.-S. Js, B. R, B. C, B. Pk, B. Dj, B. Ar, B. Af, B. Kr, C. A, C. A, C. Jh, C. Eb, C. Ej, C. J, C. Jf, F. G, F. Mn, F. Pj, G. Hk, G. Mr, G. Pv, G. Mr, I. A, I. A, J. Md, K. M, K. H, K. Mcm, L. Ps, L. M, L. P, L. Kl, L. Ak, M. Tm, M. T, M. Kl, M. Am, N. Dh, N. N, N. Hk, N. Cy, N. Sp, N. Jm, N. H, N. J, O. Dr, P. Ck, P. Lm, R. R, R. B, R. G, R. G, S. Jk, S. M, S. Bl, S. Sc, S. Pas, S. Ae, S. M, S. H, T. G, V.M. Eg, W. C, W. M, W. P, W. Ba, W. G, W. A, Y. Wka, Z. G, H. Jt, D.G. Jf, S. Lf, V. Rgw
        Longitudinal molecular trajectories of diffuse glioma in adults.
        Nature. 2019; 576https://doi.org/10.1038/s41586-019-1775-1
        • Draaisma K.
        • Chatzipli A.
        • Taphoorn M.
        • Kerkhof M.
        • Weyerbrock A.
        • Sanson M.
        • Hoeben A.
        • Lukacova S.
        • Lombardi G.
        • Leenstra S.
        • Hanse M.
        • Fleischeuer R.
        • Watts C.
        • McAbee J.
        • Angelopoulos N.
        • Gorlia T.
        • Golfinopoulos V.
        • Kros J.M.
        • Verhaak R.G.W.
        • Bours V.
        • van den Bent M.J.
        • McDermott U.
        • Robe P.A.
        • French P.J.
        Molecular evolution of IDH wild-type glioblastomas treated with standard of care affects survival and design of precision medicine trials: a report from the EORTC 1542 study.
        J. Clin. Oncol. 2020; 38: 81-99https://doi.org/10.1200/JCO.19.00367
        • Wang Q.
        • Hu B.
        • Hu X.
        • Kim H.
        • Squatrito M.
        • Scarpace L.
        • deCarvalho A.C.
        • Lyu S.
        • Li P.
        • Li Y.
        • Barthel F.
        • Cho H.J.
        • Lin Y.-H.
        • Satani N.
        • Martinez-Ledesma E.
        • Zheng S.
        • Chang E.
        • Sauvé C.-E.G.
        • Olar A.
        • Lan Z.D.
        • Finocchiaro G.
        • Phillips J.J.
        • Berger M.S.
        • Gabrusiewicz K.R.
        • Wang G.
        • Eskilsson E.
        • Hu J.
        • Mikkelsen T.
        • DePinho R.A.
        • Muller F.
        • Heimberger A.B.
        • Sulman E.P.
        • Nam D.-H.
        • Verhaak R.G.W.
        Tumor evolution of glioma intrinsic gene expression subtype associates with immunological changes in the microenvironment.
        Cancer Cell. 2017; 32: 42-56.e6https://doi.org/10.1016/j.ccell.2017.06.003
        • Wang J.
        • Cazzato E.
        • Ladewig E.
        • Frattini V.
        • Rosenbloom D.I.S.
        • Zairis S.
        • Abate F.
        • Liu Z.
        • Elliott O.
        • Shin Y.-J.
        • Lee J.-K.
        • Lee I.-H.
        • Park W.-Y.
        • Eoli M.
        • Blumberg A.J.
        • Lasorella A.
        • Nam D.-H.
        • Finocchiaro G.
        • Iavarone A.
        • Rabadan R.
        Clonal evolution of glioblastoma under therapy.
        Nat. Genet. 2016; 48: 768-776https://doi.org/10.1038/ng.3590
        • Liu W.
        • Huang Y.
        • Wang D.
        • Han F.
        • Chen H.
        • Chen J.
        • Jiang X.
        • Cao J.
        • Liu J.
        MPDZ as a novel epigenetic silenced tumor suppressor inhibits growth and progression of lung cancer through the Hippo-YAP pathway.
        Oncogene. 2021; 40: 4468-4485https://doi.org/10.1038/s41388-021-01857-8
        • Louis D.N.
        • Perry A.
        • Reifenberger G.
        • von Deimling A.
        • Figarella-Branger D.
        • Cavenee W.K.
        • Ohgaki H.
        • Wiestler O.D.
        • Kleihues P.
        • Ellison D.W.
        The 2016 World Health Organization classification of tumors of the central nervous system: a summary.
        Acta Neuropathol. 2016; 131: 803-820https://doi.org/10.1007/s00401-016-1545-1
        • Figarella-Branger D.
        • Mokhtari K.
        • Dehais C.
        • Jouvet A.
        • Uro-Coste E.
        • Colin C.
        • Carpentier C.
        • Forest F.
        • Maurage C.-A.
        • Vignaud J.-M.
        • Polivka M.
        • Lechapt-Zalcman E.
        • Eimer S.
        • Viennet G.
        • Quintin-Roué I.
        • Aubriot-Lorton M.-H.
        • Diebold M.-D.
        • Loussouarn D.
        • Lacroix C.
        • Rigau V.
        • Laquerrière A.
        • Vandenbos F.
        • Michalak S.
        • Sevestre H.
        • Peoch M.
        • Labrousse F.
        • Christov C.
        • Kemeny J.-L.
        • Chenard M.-P.
        • Chiforeanu D.
        • Ducray F.
        • Idbaih A.
        • Network P.O.L.A.
        Mitotic index, microvascular proliferation, and necrosis define 3 groups of 1p/19q codeleted anaplastic oligodendrogliomas associated with different genomic alterations.
        Neuro-Oncology. 2014; 16: 1244-1254https://doi.org/10.1093/neuonc/nou047
        • Metellus P.
        • Coulibaly B.
        • Nanni I.
        • Fina F.
        • Eudes N.
        • Giorgi R.
        • Barrie M.
        • Chinot O.
        • Fuentes S.
        • Dufour H.
        • Ouafik L.
        • Figarella-Branger D.
        Prognostic impact of O6-methylguanine-DNA methyltransferase silencing in patients with recurrent glioblastoma multiforme who undergo surgery and carmustine wafer implantation: a prospective patient cohort.
        Cancer. 2009; 115: 4783-4794https://doi.org/10.1002/cncr.24546
        • Wen P.Y.
        • Macdonald D.R.
        • Reardon D.A.
        • Cloughesy T.F.
        • Sorensen A.G.
        • Galanis E.
        • Degroot J.
        • Wick W.
        • Gilbert M.R.
        • Lassman A.B.
        • Tsien C.
        • Mikkelsen T.
        • Wong E.T.
        • Chamberlain M.C.
        • Stupp R.
        • Lamborn K.R.
        • Vogelbaum M.A.
        • van den Bent M.J.
        • Chang S.M.
        Updated response assessment criteria for high-grade gliomas: response assessment in neuro-oncology working group.
        J. Clin. Oncol. 2010; 28: 1963-1972https://doi.org/10.1200/JCO.2009.26.3541
        • Adélaïde J.
        • Finetti P.
        • Bekhouche I.
        • Repellini L.
        • Geneix J.
        • Sircoulomb F.
        • Charafe-Jauffret E.
        • Cervera N.
        • Desplans J.
        • Parzy D.
        • Schoenmakers E.
        • Viens P.
        • Jacquemier J.
        • Birnbaum D.
        • Bertucci F.
        • Chaffanet M.
        Integrated profiling of basal and luminal breast cancers.
        Cancer Res. 2007; 67: 11565-11575https://doi.org/10.1158/0008-5472.CAN-07-2536
        • Olshen A.B.
        • Venkatraman E.S.
        • Lucito R.
        • Wigler M.
        Circular binary segmentation for the analysis of array-based DNA copy number data.
        Biostatistics. 2004; 5: 557-572https://doi.org/10.1093/biostatistics/kxh008
        • Mermel C.H.
        • Schumacher S.E.
        • Hill B.
        • Meyerson M.L.
        • Beroukhim R.
        • Getz G.
        GISTIC2.0 facilitates sensitive and confident localization of the targets of focal somatic copy-number alteration in human cancers.
        Genome Biol. 2011; 12: R41https://doi.org/10.1186/gb-2011-12-4-r41
        • Huang D.W.
        • Sherman B.T.
        • Lempicki R.A.
        Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources.
        Nat. Protoc. 2009; 4: 44-57https://doi.org/10.1038/nprot.2008.211
        • Huang D.W.
        • Sherman B.T.
        • Lempicki R.A.
        Bioinformatics enrichment tools: paths toward the comprehensive functional analysis of large gene lists.
        Nucleic Acids Res. 2009; 37: 1-13https://doi.org/10.1093/nar/gkn923
        • Quinlan A.R.
        • Hall I.M.
        BEDTools: a flexible suite of utilities for comparing genomic features.
        Bioinformatics. 2010; 26: 841-842https://doi.org/10.1093/bioinformatics/btq033
        • Wang K.
        • Li M.
        • Hakonarson H.
        ANNOVAR: functional annotation of genetic variants from high-throughput sequencing data.
        Nucleic Acids Res. 2010; 38: e164https://doi.org/10.1093/nar/gkq603
        • Salgado D.
        • Desvignes J.-P.
        • Rai G.
        • Blanchard A.
        • Miltgen M.
        • Pinard A.
        • Lévy N.
        • Collod-Béroud G.
        • Béroud C.
        UMD-predictor: a high-throughput sequencing compliant system for pathogenicity prediction of any human cDNA substitution.
        Hum. Mutat. 2016; 37: 439-446https://doi.org/10.1002/humu.22965
        • Bowman R.L.
        • Wang Q.
        • Carro A.
        • Verhaak R.G.W.
        • Squatrito M.
        GlioVis data portal for visualization and analysis of brain tumor expression datasets.
        Neuro-Oncology. 2017; 19: 139-141https://doi.org/10.1093/neuonc/now247
        • Li C.
        • Tang Z.
        • Zhang W.
        • Ye Z.
        • Liu F.
        GEPIA2021: integrating multiple deconvolution-based analysis into GEPIA.
        Nucleic Acids Res. 2021; 49: W242-W246https://doi.org/10.1093/nar/gkab418
        • Wang Z.
        • Sun D.
        • Chen Y.-J.
        • Xie X.
        • Shi Y.
        • Tabar V.
        • Brennan C.W.
        • Bale T.A.
        • Jayewickreme C.D.
        • Laks D.R.
        • Alcantara Llaguno S.
        • Parada L.F.
        Cell lineage-based stratification for glioblastoma.
        Cancer Cell. 2020; 38: 366-379.e8https://doi.org/10.1016/j.ccell.2020.06.003
        • SenGupta S.
        • Hein L.E.
        • Parent C.A.
        The recruitment of neutrophils to the tumor microenvironment is regulated by multiple mediators.
        Front. Immunol. 2021; 12734188https://doi.org/10.3389/fimmu.2021.734188
        • Huang Y.-S.
        • Liu W.-B.
        • Han F.
        • Yang J.-T.
        • Hao X.-L.
        • Chen H.-Q.
        • Jiang X.
        • Yin L.
        • Ao L.
        • Cui Z.-H.
        • Cao J.
        • Liu J.-Y.
        Copy number variations and expression of MPDZ are prognostic biomarkers for clear cell renal cell carcinoma.
        Oncotarget. 2017; 8: 78713-78725https://doi.org/10.18632/oncotarget.20220
        • Martin T.A.
        • Watkins G.
        • Mansel R.E.
        • Jiang W.G.
        Loss of tight junction plaque molecules in breast cancer tissues is associated with a poor prognosis in patients with breast cancer.
        Eur. J. Cancer. 2004; 40: 2717-2725https://doi.org/10.1016/j.ejca.2004.08.008
        • Tetzlaff F.
        • Adam M.G.
        • Feldner A.
        • Moll I.
        • Menuchin A.
        • Rodriguez-Vita J.
        • Sprinzak D.
        • Fischer A.
        MPDZ promotes DLL4-induced notch signaling during angiogenesis.
        Elife. 2018; 7https://doi.org/10.7554/eLife.32860
        • Kopan R.
        • Ilagan Ma.X.G.
        The canonical notch signaling pathway: unfolding the activation mechanism.
        Cell. 2009; 137: 216-233https://doi.org/10.1016/j.cell.2009.03.045
        • Noguera-Troise I.
        • Daly C.
        • Papadopoulos N.J.
        • Coetzee S.
        • Boland P.
        • Gale N.W.
        • Chieh Lin H.
        • Yancopoulos G.D.
        • Thurston G.
        Blockade of Dll4 inhibits tumour growth by promoting non-productive angiogenesis.
        Nature. 2006; 444: 1032-1037https://doi.org/10.1038/nature05355