"De novo replication repair deficient glioblastoma, IDH-wildtype" is a distinct glioblastoma subtype in adults that may benefit from immune checkpoint blockade

Abstract

Glioblastoma is a clinically and molecularly heterogeneous disease, and new predictive biomarkers are needed to identify those patients most likely to respond to specific treatments. Through prospective genomic profiling of 459 consecutive primary treatment-naïve IDH-wildtype glioblastomas in adults, we identified a unique subgroup (2%, 9/459) defined by somatic hypermutation and DNA replication repair deficiency due to biallelic inactivation of a canonical mismatch repair gene. The deleterious mutations in mismatch repair genes were often present in the germline in the heterozygous state with somatic inactivation of the remaining allele, consistent with glioblastomas arising due to underlying Lynch syndrome. A subset of tumors had accompanying proofreading domain mutations in the DNA polymerase POLE and resultant "ultrahypermutation". The median age at diagnosis was 50 years (range 27-78), compared with 63 years for the other 450 patients with conventional glioblastoma (p < 0.01). All tumors had histologic features of the giant cell variant of glioblastoma. They lacked EGFR amplification, lacked combined trisomy of chromosome 7 plus monosomy of chromosome 10, and only rarely had TERT promoter mutation or CDKN2A homozygous deletion, which are hallmarks of conventional IDH-wildtype glioblastoma. Instead, they harbored frequent inactivating mutations in TP53, NF1, PTEN, ATRX, and SETD2 and recurrent activating mutations in PDGFRA. DNA methylation profiling revealed they did not align with known reference adult glioblastoma methylation classes, but instead had unique globally hypomethylated epigenomes and mostly classified as "Diffuse pediatric-type high grade glioma, RTK1 subtype, subclass A". Five patients were treated with immune checkpoint blockade, four of whom survived greater than 3 years. The median overall survival was 36.8 months, compared to 15.5 months for the other 450 patients (p < 0.001). We conclude that "De novo replication repair deficient glioblastoma, IDH-wildtype" represents a biologically distinct subtype in the adult population that may benefit from prospective identification and treatment with immune checkpoint blockade.

Keywords: Giant cell glioblastoma; Hypermutation; Immune checkpoint blockade; Immunotherapy; Lynch syndrome; Mismatch repair deficiency; Molecular neuro-oncology; Molecular neuropathology; POLE; Ultrahypermutation.

Fig. 1

Imaging and histopathologic features of “De novo replication repair deficient glioblastoma, IDH-wildtype”. a, b Imaging features at initial presentation were indistinguishable from conventional glioblastoma, IDH-wildtype. All patients had mass lesions in the cerebral hemispheres demonstrating substantial mass effect, peripheral ring enhancement on post-contrast sequences, and extensive T2 FLAIR hyperintensity extending into the surrounding parenchyma reflective of the infiltrative growth patterns. The tumors were histologically composed of diffuse astrocytic gliomas with high cellularity, brisk mitotic activity, marked nuclear pleomorphism including frequent giant cells, microvascular proliferation, and palisading necrosis. All tumors were IDH-wildtype and negative for IDH1 p.R132H mutant protein expression by immunohistochemistry (a). Many tumors (5/9, 56%) contained inactivating ATRX mutations and had somatic loss of ATRX protein expression in tumor cells (a). Inactivating mutations in TP53 were frequent in the tumor cohort (8/9, 89%), and those tumors with deleterious missense mutations demonstrated aberrant nuclear accumulation of p53 protein (b). All tumors demonstrated biallelic inactivation of a mismatch repair gene, which was either exclusively somatic or with one of the two events being present in the germline in the heterozygous state (Lynch syndrome) accompanied by somatic inactivation of the remaining allele. Immunohistochemistry demonstrated loss of expression of the affected mismatch repair protein in tumor cell nuclei (and concomitant MSH6 loss for those with MSH2 mutational inactivation given the protein dimerization pattern of the mismatch repair complex), with retained/intact expression in endothelial cells and other admixed non-neoplastic cells (b). This is in contrast to children with constitutional mismatch repair deficiency (CMMRD) syndrome who have biallelic germline mutation in an MMR gene and whose tumors demonstrate loss of the affected MMR protein in both tumor cells and non-neoplastic cells (not shown, see references 5 and 9)

Fig. 2

The genomic landscape of “De novo replication repair deficient glioblastoma, IDH-wildtype”. a Oncoprint summary plot of the clinical, histologic, genomic, and epigenomic features of the de novo RRD glioblastoma patient cohort. DNA methylation subclass assignment is based on random forest classification using version 12.7 of the DKFZ MolecularNeuropathology.org online classifier. See Supplementary Tables S3, S4, and S8 for source data. b Comparison of oncogenic alteration frequency between de novo RRD glioblastoma (n = 9) and conventional glioblastoma (n = 450) using the identical genomic testing platform and informatics pipeline. Significant differences in genetic alteration frequency are denoted with an asterisk (p < 0.05). See Table 2 and Supplementary Tables S3 and S4 for source data

Fig. 3

“De novo replication repair deficient glioblastoma, IDH-wildtype” lacks many of the recurrent chromosomal copy number alterations present in conventional glioblastoma, IDH-wildtype including trisomy/gain of chromosome 7 and monosomy/loss of chromosome 10. Shown are copy number summary plots illustrating the fraction of tumors with gain or loss along each chromosome for 7 de novo RRD glioblastomas and 98 conventional glioblastomas. See Supplementary Table S4 for source data and Supplementary Fig. S2 for representative copy number plots of individual tumors

Fig. 4

“De novo replication repair deficient glioblastoma, IDH-wildtype” has a distinct global hypomethylation epigenetic signature compared to conventional glioblastoma, IDH-wildtype. a Violin plot of DNA methylation data showing the mean beta-value for each of approximately 850,000 CpG sites across 7 de novo RRD glioblastomas and 98 conventional glioblastomas. b Heatmap of DNA methylation profiles for 7 de novo RRD glioblastoma alongside 98 conventional glioblastoma. Shown are the 5000 most differentially methylated probes amongst the 105 glioblastomas revealing extensive hypomethylation of CpG sites in the de novo RRD glioblastoma compared to the cohort of conventional glioblastomas. See Supplementary Table S5 for detailed annotations of these 5000 most differentially methylated CpG sites

Fig. 5

“De novo replication repair deficient glioblastoma, IDH-wildtype” has a unique epigenetic signature distinct from all established reference methylation classes of IDH-wildtype glioblastomas in adults. a tSNE dimensionality reduction plot of genome-wide DNA methylation profiles for 7 de novo RRD glioblastomas alongside 1143 reference samples spanning 25 CNS tumor methylation groups and 3 control tissue methylation groups. See Supplementary Table S7 for sample manifest. b Results of DNA methylation-based classification for the 7 de novo RRD glioblastoma samples using version 12.7 of the DKFZ Molecular Neuropathology classifier. See Supplementary Table S8 for further details

Fig. 6

Deconvolution analysis reveals a unique cellular composition of “De novo replication repair deficient glioblastoma, IDH-wildtype” compared to conventional glioblastoma, IDH-wildtype. a Cellular composition of 7 de novo RRD glioblastoma and 98 conventional glioblastoma was estimated by methyCIBERSORT deconvolution of Infinium EPIC DNA methylation profiles at time of initial diagnostic surgery before any adjuvant therapy or immune checkpoint blockade. The de novo RRD glioblastomas demonstrated a greater proportion of microglia and CD8 + T-cells, and lower proportion of regulatory T-cells (Tregs) compared to conventional glioblastoma. See Supplementary Table S9 for source data. b Representative photomicrographs of immunohistochemistry for CD8 and CD163 on a de novo RRD glioblastoma and a conventional glioblastoma for comparison

Fig. 7

Patients with “de novo replication repair deficient glioblastoma, IDH-wildtype” have prolonged survival compared to conventional glioblastoma, IDH-wildtype and may benefit from immune checkpoint blockade. a Swimmer’s plot showing timing of initial surgical intervention, radiation, chemotherapy, immune checkpoint blockade with either pembrolizumab or nivolumab, and clinical outcomes for the 9 patients with de novo replication repair deficient glioblastoma, IDH-wildtype. See Supplementary Table S1 for further clinical data including extent of resection and treatment regimen. b Kaplan–Meier curves showing overall survival for the 459 consecutive adult patients with IDH-wildtype glioblastoma in the cerebral hemispheres, stratified by those with “de novo replication repair deficient glioblastoma, IDH-wildtype” (n = 9, median survival 36.8 months) versus those with conventional glioblastoma (n = 450, median survival 15.5 months)