IMMU-13. Dual CTLA4/ PD-1 blockade improves survival for replication-repair deficient high-grade gliomas failing single agent PD-1 inhibition: An IRRDC study

BACKGROUND: High-grade gliomas (HGG) with replication-repair deficiency (RRD) harbour high mutation burden (TMB) and are rapidly fatal following chemo-radiation approaches. Although hypermutation results in objective responses and prolonged survival in >30% of patients undergoing PD1-blockade, salvage following failure of PD1-inhibition remains a challenge. METHODS: We performed a real-world study of Ipilimumab (anti-CTLA4) in combination with Nivolumab/Pembrolizumab for patients failing single-agent PD1-inhibition. RESULTS: Among 68 consortium patients with relapsed HGG treated with single-agent PD1-inhibitors, progression was observed in 43 (63%). Ipilimumab was added to 20/43 (46.5%), 14 (32.5%) received best supportive care (BSC), and 9 (21%) received miscellaneous therapies. For patients receiving CTLA4/PD1-inhibition, median age at progression was 12.3-years (IQR: 9; 15.6). Time from anti-PD1 initiation to progression was 8-months (IQR: 3.8; 18.5). Germline predisposition was observed in all patients (CMMRD: 70%, Lynch: 25%, polymerase-proofreading deficiency: 5%). All HGG were hypermutant (median TMB: 182 mutations/Mb; IQR: 15.6; 369.4). Centralized radiology review revealed objective responses in 3/20 (15%, all ultra-hypermutant: 320, 496, 834 mutations/Mb), stable disease in 5 (25%), and 12 (60%) eventually progressed (iRANO). Following failure of PD1-blockade, estimated progression-free and overall survival at 18-months for patients receiving CTLA4/PD1-inhibition were 11% and 25%, respectively. Importantly, survival was superior to patients receiving BSC (median OS <1-month versus 12-months on CTLA4/PD1-inhibition; p<0.001). All patients receiving BSC died within 3.5-months, while 4/8 survivors were alive for >1-year on the anti-CTLA4/PD1combination (range:1-48 months). The combinational immunotherapy resulted in significant autoimmune toxicity in 11/20 (55%), warranting immunosuppressive therapy in all, and treatment abandonment in 6 patients. CONCLUSION: Combined CTLA4/PD1-blockade after failure of single-agent PD1-inhibition revealed objective responses and prolonged survival in an otherwise rapidly-fatal disease. This needs to be assessed in the context of significant autoimmunity, supporting the need for the current prospective trial (NCT04500548), and novel strategies to limit treatment-related toxicity.

Genomic predictors of response to PD-1 inhibition in children with germline DNA replication repair deficiency

Cancers arising from germline DNA mismatch repair deficiency or polymerase proofreading deficiency (MMRD and PPD) in children harbour the highest mutational and microsatellite insertion–deletion (MS-indel) burden in humans. MMRD and PPD cancers are commonly lethal due to the inherent resistance to chemo-irradiation. Although immune checkpoint inhibitors (ICIs) have failed to benefit children in previous studies, we hypothesized that hypermutation caused by MMRD and PPD will improve outcomes following ICI treatment in these patients. Using an international consortium registry study, we report on the ICI treatment of 45 progressive or recurrent tumors from 38 patients. Durable objective responses were observed in most patients, culminating in a 3 year survival of 41.4%. High mutation burden predicted response for ultra-hypermutant cancers (>100 mutations per Mb) enriched for combined MMRD + PPD, while MS-indels predicted response in MMRD tumors with lower mutation burden (10–100 mutations per Mb). Furthermore, both mechanisms were associated with increased immune infiltration even in ‘immunologically cold’ tumors such as gliomas, contributing to the favorable response. Pseudo-progression (flare) was common and was associated with immune activation in the tumor microenvironment and systemically. Furthermore, patients with flare who continued ICI treatment achieved durable responses. This study demonstrates improved survival for patients with tumors not previously known to respond to ICI treatment, including central nervous system and synchronous cancers, and identifies the dual roles of mutation burden and MS-indels in predicting sustained response to immunotherapy.

Hypermutation and microsatellite burden determine responses and long-term survival following PD-1 blockade in children and young adults with refractory cancers resulting from germline DNA replication repair deficiency.

IMMU-18. FAVORABLE OUTCOME IN REPLICATION REPAIR DEFICIENT HYPERMUTANT BRAIN TUMORS TO IMMUNE CHECKPOINT INHIBITION: AN INTERNATIONAL RRD CONSORTIUM REGISTRY STUDY

Pediatric brain tumors with replication repair deficiency (RRD) are hypermutant and may respond to immune checkpoint inhibition (ICI). We performed a consortium registry study of ICI in recurrent RRD cancers. Clinical and companion biomarkers were collected longitudinally on all patients. Biomarkers included tumor mutational burden (TMB), neoantigens and genetic signatures obtained from whole genome and exome sequencing. Immune inference was obtained by RNAseq and T cell rearrangement was collected in the tumor and in blood throughout treatment. Of the 46 tumors on the study, 32 were brain tumors with glioblastoma in 96%. Rapid, objective responses (>50%) were observed in 50% of glioblastomas. Three year overall survival for the whole cohort was 48+/-8% which compares favorably with historical controls. Brain tumors fared worse with OS of 39+/-10% and late recurrences observed even after 2 years of therapy (p=0.02). Tumor size and acute “flare” constitute poor outcome throughout all cancers. While all tumors are hypermutant, TMB and predicted neoantigens correlated with response to ICI (p=0.02). Specific signatures extracted from SNVs and total mutations predicted response to ICI and favorable outcome (p=0.005). RNA inference and TCR reveal that the FLARE phenotype is mostly acute nonspecific immune response and not true progression. Finally, glioblastomas (n=8) which failed single agent ICI had favorable responses to combinational immunotherapies with prolonged survival of 65%+/-8% at one year after failure vs 0 for other patients (p=0.01). RRD glioblastomas exhibit favorable outcome and responses to ICI. Combinational therapies based on tumor and immune signatures of these cancers are necessary.

HGG-20. DIAGNOSTIC AND BIOLOGICAL ROLE OF METHYLATION PATTERNS IN REPLICATION REPAIR DEFICIENT HIGH GRADE GLIOMAS

Replication repair deficiency (RRD) is an important driving mechanism of pediatric high grade glioma (pHGG) occurring predominantly in the context of germline mutations in RRD-associated genes. Although pHGG present specific patterns of DNA methylation corresponding to driving oncogenic processes, methylation patterns have not been well studied in RRD tumors. We analyzed 52 RRD pHGG using either 450k or 850k methylation arrays. These arrays were compared with 234 PHGG driven by other genetic or epigenetic mechanisms and 10 additional pHGG samples known to be hypermutant. RRD pHGG displayed a methylation pattern corresponding to specific secondary mutations such as IDH1 and H3K27M. Strikingly, RRD pHGG lacking these known secondary mutations largely clustered together with a poorly described group previously labelled Wild type-C. Most of the hypermutant tumors clustered in a similar location suggesting undiagnosed RRD may be a driving force for tumors clustering in this location. Analysis of methylation patterns revealed that RRD pHGG displayed a unique CpG Island Demethylator Phenotype in contrast to the Methylator Phenotype described in other cancers. This effect was most concentrated at gene promotors. Prominent demethylation was observed in genes and pathways critical to cellular survival including cell cycle, gene expression, cellular metabolism and cellular organization. These data suggest that methylation profiles may provide diagnostic information for the detection of RRD pHGG. Furthermore, our findings highlight the unique natural selection pressures in these highly dysregulated, hypermutant cancers and provide novel impact of hypermutation and RRD on the cancer epigenome.

Functional Repair Assay for the Diagnosis of Constitutional Mismatch Repair Deficiency From Non-Neoplastic Tissue

PURPOSE

Constitutional mismatch repair deficiency (CMMRD) is a highly penetrant cancer predisposition syndrome caused by biallelic mutations in mismatch repair (MMR) genes. As several cancer syndromes are clinically similar, accurate diagnosis is critical to cancer screening and treatment. As genetic diagnosis is confounded by 15 or more pseudogenes and variants of uncertain significance, a robust diagnostic assay is urgently needed. We sought to determine whether an assay that directly measures MMR activity could accurately diagnose CMMRD.

PATIENTS AND METHODS

In vitro MMR activity was quantified using a 3'-nicked G-T mismatched DNA substrate, which requires MSH2-MSH6 and MLH1-PMS2 for repair. We quantified MMR activity from 20 Epstein-Barr virus-transformed lymphoblastoid cell lines from patients with confirmed CMMRD. We also tested 20 lymphoblastoid cell lines from patients who were suspected for CMMRD. We also characterized MMR activity from patients with neurofibromatosis type 1, Li-Fraumeni syndrome, polymerase proofreading-associated cancer syndrome, and Lynch syndrome.

RESULTS

All CMMRD cell lines had low MMR activity (n = 20; mean, 4.14 ± 1.56%) relative to controls (n = 6; mean, 44.00 ± 8.65%; P < .001). Repair was restored by complementation with the missing protein, which confirmed MMR deficiency. All cases of patients with suspected CMMRD were accurately diagnosed. Individuals with Lynch syndrome (n = 28), neurofibromatosis type 1 (n = 5), Li-Fraumeni syndrome (n = 5), and polymerase proofreading-associated cancer syndrome (n = 3) had MMR activity that was comparable to controls. To accelerate testing, we measured MMR activity directly from fresh lymphocytes, which yielded results in 8 days.

CONCLUSION

On the basis of the current data set, the in vitro G-T repair assay was able to diagnose CMMRD with 100% specificity and sensitivity. Rapid diagnosis before surgery in non-neoplastic tissues could speed proper therapeutic management.

Immune Checkpoint Inhibition for Hypermutant Glioblastoma Multiforme Resulting From Germline Biallelic Mismatch Repair Deficiency

PURPOSE

Recurrent glioblastoma multiforme (GBM) is incurable with current therapies. Biallelic mismatch repair deficiency (bMMRD) is a highly penetrant childhood cancer syndrome often resulting in GBM characterized by a high mutational burden. Evidence suggests that high mutation and neoantigen loads are associated with response to immune checkpoint inhibition.

PATIENTS AND METHODS

We performed exome sequencing and neoantigen prediction on 37 bMMRD cancers and compared them with childhood and adult brain neoplasms. Neoantigen prediction bMMRD GBM was compared with responsive adult cancers from multiple tissues. Two siblings with recurrent multifocal bMMRD GBM were treated with the immune checkpoint inhibitor nivolumab.

RESULTS

All malignant tumors (n = 32) were hypermutant. Although bMMRD brain tumors had the highest mutational load because of secondary polymerase mutations (mean, 17,740 ± standard deviation, 7,703), all other high-grade tumors were hypermutant (mean, 1,589 ± standard deviation, 1,043), similar to other cancers that responded favorably to immune checkpoint inhibitors. bMMRD GBM had a significantly higher mutational load than sporadic pediatric and adult gliomas and all other brain tumors (P < .001). bMMRD GBM harbored mean neoantigen loads seven to 16 times higher than those in immunoresponsive melanomas, lung cancers, or microsatellite-unstable GI cancers (P < .001). On the basis of these preclinical data, we treated two bMMRD siblings with recurrent multifocal GBM with the anti-programmed death-1 inhibitor nivolumab, which resulted in clinically significant responses and a profound radiologic response.

CONCLUSION

This report of initial and durable responses of recurrent GBM to immune checkpoint inhibition may have implications for GBM in general and other hypermutant cancers arising from primary (genetic predisposition) or secondary MMRD.