Purpose:

Checkpoint inhibitors have limited efficacy for children with unselected solid and brain tumors. We report the first prospective pediatric trial (NCT02992964) using nivolumab exclusively for refractory nonhematologic cancers harboring tumor mutation burden (TMB) ≥5 mutations/megabase (mut/Mb) and/or mismatch repair deficiency (MMRD).

Patients and Methods:

Twenty patients were screened, and 10 were ultimately included in the response cohort of whom nine had TMB >10 mut/Mb (three initially eligible based on MMRD) and one patient had TMB between 5 and 10 mut/Mb.

Results:

Delayed immune responses contributed to best overall response of 50%, improving on initial objective responses (20%) and leading to 2-year overall survival (OS) of 50% [95% confidence interval (CI), 27–93]. Four children, including three with refractory malignant gliomas are in complete remission at a median follow-up of 37 months (range, 32.4–60), culminating in 2-year OS of 43% (95% CI, 18.2–100). Biomarker analyses confirmed benefit in children with germline MMRD, microsatellite instability, higher activated and lower regulatory circulating T cells. Stochastic mutation accumulation driven by underlying germline MMRD impacted the tumor microenvironment, contributing to delayed responses. No benefit was observed in the single patient with an MMR-proficient tumor and TMB 7.4 mut/Mb.

Conclusions:

Nivolumab resulted in durable responses and prolonged survival for the first time in a pediatric trial of refractory hypermutated cancers including malignant gliomas. Novel biomarkers identified here need to be translated rapidly to clinical care to identify children who can benefit from checkpoint inhibitors, including upfront management of cancer.

See related commentary by Mardis, p. 4701

This article is featured in Selected Articles from This Issue, p. 4699

Translational Relevance

This trial represents the first prospective assessment of the utility of immune checkpoint blockade in pediatric cancers with increased mutation burden and/or mismatch repair deficiency. The best overall response rate of 50% and remarkable prolonged overall survival particularly in patients with relapsed high-grade glioma demonstrates a clear role for immune checkpoint inhibition (ICI) in this rare pediatric population and lays a foundation for incorporating ICI in the upfront treatment of these patients.

Immune checkpoint inhibition (ICI) targeting programmed-death 1 (PD1) and its ligand (PD-L1) has shown improved survival in adults with advanced cancers (1–4). The best responses have been observed in cancers exhibiting high tumor mutation burden (TMB), mismatch repair deficiency (MMRD), microsatellite instability (MSI), or PD-L1 expression (3–5). Notably, hypermutant gliomas have failed to respond due to the immune privilege of the central nervous system (CNS), their immunosuppressive microenvironment (6), and the subclonal nature of MMRD in these cancers (7–9).

Previous clinical trials in children using anti-PD1 [nivolumab (10), pembrolizumab (11)], anti–PD-L1 [atezolizumab (12), avelumab (13)], and anti-CTLA4 [ipilimumab (14)] have shown limited efficacy restricted to lymphomas and rare solid tumors. In addition to low TMB and immunogenicity contributing to an “immune-cold” microenvironment (11), this has been attributed to low expression of MHC (15), an immature immune system (16) and gut microbiome (17), and the presence of an immunosuppressive microenvironment enriched for macrophages (18). The failure of ICI in unselected pediatric cancers has led to recommendations of avoiding ICI monotherapy trials in children without robust biological rationale (19, 20).

The improved outcomes using ICI in refractory MMRD cancers in adults (21) and detection of hypermutation in 5% of pediatric cancers (22), the majority of which were driven by MMRD, prompted us to hypothesize that such a molecularly selected cohort of cancers in children may respond to ICI monotherapy. Hence, we developed NCT02992964, an investigator-initiated, multicenter, open-label, single-arm pilot study in which pediatric patients with relapsed/refractory cancers with increased TMB and/or MMRD were treated with nivolumab. The trial included an initial “Part 1: Molecular Profiling,” allowing patients to consent to analyses to confirm eligibility. This required the presence of a mutation in MMR genes (MLH1, MSH2, MSH6, PMS2, EPCAM, MSH3) or a loss of MMR protein expression on IHC, hypermutation, or a corresponding mutational signature (COSMIC; ref. 23) on sequencing, MSI, a history of germline conditions linked to hypermutation (constitutional mismatch repair deficiency syndrome/CMMRD, Lynch syndrome, xeroderma pigmentosum), or prior treatment with temozolomide. Patients were eligible for “Part 2: Treatment and Companion Biomarker Studies” provided they separately consented for therapy, had a measurable or evaluable relapsed/refractory cancer with no alternative treatment, and confirmation of MMRD and/or hypermutation.

Herein, we report the unique antitumor responses and survival benefit using nivolumab observed in NCT02992964 for children with advanced, nonhematologic malignancies with elevated mutation burden and/or MMRD. We also present the results of exploratory analyses that suggest that definite genomic and immune biomarkers can help identify children whose cancers can benefit from ICI monotherapy.

Trial design and treatment

NCT02992964 was an investigator-initiated, multicenter, open-label, single-arm pilot study in which pediatric patients ages ≥12 months and <25 years of age with relapsed/refractory cancers with elevated TMB and/or MMRD were treated with nivolumab 3 mg/kg every 2 weeks until confirmed disease progression, intolerable toxicity, or for a maximum of 24 months. The study was approved by Institutional Ethics Review Boards and competent authorities in each center, and conducted in accordance with International Ethical Guidelines for Biomedical Research Involving Human Subjects (CIOMS). Written informed consent was obtained from all patients or parents/guardians according to local regulations. Nivolumab and funding to support the conduct of the study was provided by Bristol Myers Squibb.

Patients

The protocol included an initial “Part 1: Molecular Profiling” option allowing patients to consent to sequencing to confirm eligibility. This required, either in the recurrent or primary tumor, the presence of a mutation in MMR genes (MLH1, MSH2, MSH6, PMS2, EPCAM, MSH3) or loss of protein expression on IHC, hypermutation or a corresponding mutational signature (COSMIC; ref. 23) on sequencing performed locally, MSI, a history of germline conditions linked to hypermutation (CMMRD, Lynch syndrome, xeroderma pigmentosum), or prior treatment with temozolomide.

Patients were eligible for “Part 2: Treatment and Companion Biomarker Studies” provided they separately consented for therapy, had a relapsed/refractory cancer for which there was no known curative therapy, or therapy proven to prolong survival with an acceptable quality of life, baseline imaging within 14 days of starting treatment with measurable or evaluable disease, and confirmation of MMRD and/or TMB ≥5 mutations/megabase (mut/Mb). The latter was determined by measurement of TMB either via FoundationOne (Foundation Medicine), or an equivalent next-generation sequencing cancer-gene panel. Patients with a TMB ≥5 mut/Mb were eligible, with separate cohorts for TMB 5 to 10 mut/Mb and >10 mut/Mb. Patients with a history of autoimmune diseases, human immunodeficiency virus, chronic hepatitis B/C, prior allogeneic or solid organ transplant or ICI treatment, uncontrolled infection or requiring corticosteroid therapy at greater than physiologic doses were excluded.

Study assessments and endpoints

Disease status was monitored every 2 months while on therapy, and the response was evaluated using different criteria based on the type of tumor: iRECIST (Immunotherapy Response Evaluation Criteria in Solid Tumours) for solid tumors, revised INRC (International Neuroblastoma Response Criteria) for neuroblastoma, and iRANO (Immunotherapy Response Assessment in Neuro-Oncology) for CNS malignancies. The response was labeled as complete remission (CR), partial remission (PR), or progressive disease (PD) based on the assessment. The primary objective was to evaluate objective response rate (ORR = CR + PR). For correctly assigning immune responses following possible pseudoprogression, clinically stable patients could continue therapy beyond initial unconfirmed progression (iUPD), with subsequent scans being used to establish confirmed progression, or delayed “immune” CR or PR, which could then be included in the best overall response (iBOR). iBOR was defined as the best response recorded from the start of the treatment until disease progression/recurrence), even after discontinuation of protocol therapy (refs. 24, 25, 27, 28; details in Supplementary Tables S5–S9 of trial protocol: Supplementary Appendix). It was mandatory for all patients who demonstrated CR, PR, or stable disease (SD) lasting beyond six cycles to have a central review of the radiology at the Hospital for Sick Children. If feasible, histologic evaluation was allowed to resolve ambiguity regarding immune responses.

Secondary objectives included determination of progression-free survival (PFS), overall survival (OS), and evaluating the safety and toxicity of treatment. PFS was defined from time of first study medication dose to first progression that either involved significant clinical deterioration and/or was unambiguously confirmed radiologically as PD. OS was defined as time from the first study medication to death. Adverse events were graded according to the NCI's Common Terminology Criteria for Adverse Events v.4.03.

Biomarker analyses

Exploratory biomarker analyses included whole-exome sequencing (WES) and analyses of total and microsatellite in(sertion)-del(etions) (MSI) burden, plus IHC for CD8 and PD-L1 expression, and peripheral blood immune profiling using flow cytometry, and T-cell receptor (TCR) rearrangement using methods previously published by our group (29). All bioinformatics analyses were performed on the SickKids High Performance Cluster (HPF) and the University Health Network (UHN) High Performance Cluster for Health (H4H). The details are provided below.

For WES, genomic DNA, along with matched germline blood samples for available cases was extracted using the PaxGene Blood DNA Extraction Kit (catalog No./ID: 761133) for blood samples, Qiagen DNeasy Blood & Tissue Kits (catalog No./ID: 69504) for frozen tissue, and MasterPure Complete DNA and RNA Purification Kit (Epicentre #MC85200) for paraffin-embedded tissue. WES was performed at The Centre for Applied Genomics, SickKids, using SureSelect Agilent All Exon v5 kit, followed by sequencing (150X) on Illumina HiSeq 2500. The software bcl2fastq2 v2.17 was used to generate raw fastq files. Alignment to the hg38 reference genome, followed by preprocessing and QC was adapted from the GATK standard pipeline, using BWA-MEM 0.7.12 (alignment), BAMQC, Picard 2.6.0 (QC). Somatic variant calling was done post alignment, using processed bam files from tumor and matched normal samples, to call both single-nucleotide variants (SNV) and insertion deletion (indel) variants. A consensus vcf file of shared variants across two or more of four variant callers (Mutect v1.1.5, GATK v3.6/Mutect2, Strelka v1.0.14, and Varscan2 Somatic v2.4.2) was generated for SNV and indels separately, using VCFtools 0.1.15, and these vcfs were annotated using VEP v83. We used the built-in algorithms in the GATK mutect2 pipeline sequence to correct for context-dependent artifacts related to tissue source and processing by using the read-orientation bias filter and excluding low allele fraction SNP where evidence for the alternate allele consists almost entirely of F1R2 reads or F2R1 reads. The TMB (SNV per megabase) from WES data was calculated by counting the total number of somatic SNV divided by the total number of callable bases in megabases (∼50 Mb). DeconstructSigs (30) was used to determine COSMIC signatures (23, 31) in the mutation spectrum within a trinucleotide context for each sample.

Microsatellite (MS) indels were called on the bam files of tumor and matched normal samples, using an in-house pipeline using MSMuTect. The detailed methods for this algorithm have been reported previously (32, 33). Briefly, repeats of five or more nucleotides were considered to be MS loci, and using the PHOBOS algorithm and the lobSTR approach, tumor and normal BAM files were aligned with their 5′ and 3′ flanking sequences. Each MS-locus allele was estimated using the empirical noise model, which is the probability of observing a read with a MS length k and motif m, where the true length of the allele is j with the motif m. This was used to call the MS alleles with the highest likelihood of being the true allele at each MS-locus. The alleles of microsatellite (MS) markers for each tumor and its matched normal sample were determined separately. These alleles were then compared to identify the mutations present at the MS loci in the tumor samples. The Akaike Information Criterion score was assigned to both the tumor and normal models, and a threshold score that was determined by using simulated data was applied to make the final MS-indel call.

For IHC analysis of the immune microenvironment, 4-μm-thick sections of formalin-fixed paraffin-embedded surgical specimens were stained using an automated stainer (Dako OMNIS) with the following primary antibodies: PD-L1 (clone28-8, Abcam, 1/500) and CD8 (Clone c8/144B, Dako OMINS). Measurements were recorded as the number of positive cells per tissue surface unit in square millimetres. Quantitative evaluation of the IHC stains was performed by examine each section using at least five to seven different high-power fields with the most abundant tumor-infiltrating lymphocyte areas.

For TCR rearrangement repertoire profiling, genomic DNA was extracted (methods as above) from tumors. These were transferred to the Pugh laboratory at the UHN in Toronto, where library preparation and capTCRseq (34) hybrid capture were performed. Following library preparation, the samples were sequenced first on a MiSeq for QC purposes and then 300 ng of each sample, pooled in a ratio of 1:1:1, was processed for a three-step capture using target hybrid capture panel (8). Postcapture QC was performed on a MiSeq, followed by sequencing of up to a depth of approximately 2 million reads on the NextSeq. Postsequencing, the raw data were analyzed using MiXCR version 2.1.12, “iNext,” “immunarch” R packages and Pugh Lab customized functions on R version 3.5 to look at TCR rearrangements in the form of unique clonotypes (VDJ rearranged sequences) for TCRs alpha, beta, gamma, and delta. As the total read depth varied across the cohort, affecting the total successfully aligned reads, all raw fastq reads were downsampled to approximately 1 million reads. QC parameters of percent aligned reads, reads used in clonotypes, final clonotype count, and the total number of clonotypes per 1,000 reads were considered.

For flow cytometry analysis, viable frozen peripheral blood mononuclear cells were incubated with Fc block (BD Biosciences) prior to staining for surface markers (anti-CD3 - clone UCHT1, anti-CD4 – clone RPA-T4, anti-CD8 – clone RPA-T8, anti-4-1BB– clone 4B4-1, anti-TIGIT – clone MBSA43, anti-Ki67 – clone 20Raj1) and viability dye (eBioscience). Cells were fixed and permeabilized for intercellular staining with the Foxp3 transcription factor staining buffer set (BD). Flow cytometry voltages were set using Rainbow beads (Spherotech) with the same setting between experiments. Samples were acquired on a BD LSR Fortessa flow cytometer, and data were analyzed using the FlowJo software.

Statistical analyses

The trial was originally planned as a pilot study for which the initial aim was to accrue 20 pediatric patients with CMMRD. Following encouraging results in adult trials for hypermutant cancers, an amendment was made, and the study inclusion criteria was expanded to include patients with tumors harboring high TMB, as well as those with CMMRD, and consequently the total enrollment was increased to 50 patients. We planned to recruit patients into two cohorts of TMB 5 to 10 (cohort A) and ≥10 mut/Mb (cohort B) using a Simon two-stage optimal design within each cohort. Cohort A was designed with power of 85% and one-sided alpha of 0.05 to test a true response rate of 35% again a null hypothesis ORR of 10%. If there were ≥2 responders within the first 10 patients, the cohort would expand to 21 patients, and the outcome would be positive if there were ≥5 responses. Cohort B was designed with a power of 80% and one-sided alpha of 0.05 to test a true response rate of 30% againt a null hypothesis ORR of 10%. If there were ≥2 responders within the first 10 patients, the cohort would expand to 29 patients, and the outcome would be positive if there were ≥6 responses. Unfortunately, the study was terminated prematurely due to slow recruitment and loss of funding. As per predefined protocol criteria, patients exhibiting objective disease progression prior to the end of cycle 1 were considered evaluable for response. For all other patients, only those who have measurable disease present at baseline, had received at least one cycle of therapy (two doses of nivolumab) and had their disease reevaluated were considered evaluable for response. All enrolled patients who had received at least one dose of nivolumab were evaluable for the safety data. Patients who were removed from protocol therapy continued to have the required follow-up observations and documentation of additional treatments received, with the only exception for this continued documentation being the patient's withdrawal of consent, in which case only data collected prior to the withdrawal of consent could be evaluated.

PFS and OS for treated patients were estimated using the Kaplan–Meier method with patients censored, as necessary, at date of last follow-up. For correlative biomarker analyses, best overall response (iBOR) was used as a manifestation of true immune, including delayed, responses. For the serial immune correlates, landmark analysis was performed at 3 and 6 months. Furthermore, as these correlates changed over time, we performed time-dependent covariate analysis using a modified Cox model as described previously (https://cran.r-project.org/web/packages/survival/vignettes/timedep.pdf). All statistical analyses were performed with R v.4.2.1. All P values were two sided, with a cutoff of 0.05 for significance. Adjustments to the P values for multiple comparisons to control for FDR were performed using methods as previously reported by Benjamini and colleagues (35, 36). Plots were edited for aesthetics using Adobe Illustrator v.23.0.1.

Data availability

All data relevant to this work are available in European Genome Phenome Archive under accession number EGAS00001007393 (https://ega-archive.org/studies/EGAS00001007393) and accessible through communication with the corresponding authors. All codes are publicly available.

Patient demographics

Twenty patients were enrolled in “Part 1” and screened for eligibility. Nine did not meet criteria for enrollment for “Part 2.” Between May 2017 and November 2020, 11 patients received nivolumab on study at a dose of 3 mg/kg every 2 weeks until confirmed disease progression, intolerable toxicity, or for a maximum of 24 months. Data cutoff for outcomes was March 2022. All patients had failed first-line therapies. Radiation was delivered >6 months prior to recurrence and trial enrollment for all patients receiving prior radiotherapy. Although the protocol allowed palliative radiation at least 14 days before trial commencement, none of the patients received (re)-irradiation prior to the initiation of nivolumab. Patients were eligible only after recovery of acute toxic effects of previous anticancer therapy (see trial protocol, Supplementary Appendix).

One individual withdrew consent from all aspects of the trial immediately following the first dose without any evidence of progression or toxicity. As per the trial mandate requiring completion of at least one cycle (two doses) for endpoint assessments, this patient was excluded from further follow-up and outcome analyses (Fig. 1). Median age of the final cohort (n = 10) was 14 years (range, 9–18; Table 1; Supplementary Table S1). Cancer diagnoses included malignant gliomas (n = 7; glioblastoma n = 5), neuroblastoma (n = 1), colorectal carcinoma (n = 1), and adrenocortical carcinoma (n = 1; ACC; Fig. 2A).

Baseline genomic characteristics and cancer predisposition

Three children had been enrolled on the basis of loss of MMR protein expression (all had TMB> 10 mut/Mb), six based on TMB >10 mut/Mb (five of whom also had IHC loss), while a single patient was enrolled for TMB of 7.4 mut/Mb with an MMR-proficient tumor (Table 1; Fig. 2B and C). Mutational signature (COSMIC; ref. 23) analyses demonstrated single-base substitution (SBS) signatures of MMRD in eight of 10 cancers and concomitant DNA polymerase proofreading deficiency signatures along with a pathogenic POLE somatic variant in three of these eight MMRD cancers. Germline cancer predisposition was detected in nine of 10 patients [CMMRD: n = 2; Lynch syndrome: n = 5; one patient each with PALB2 variant and Li-Fraumeni syndrome (LFS); Table 1]. The single patient without germline predisposition (P8) had been previously treated with temozolomide, showed focal loss of staining for MSH6 in subpopulations of tumor cells suggesting subclonal MMR deficiency, and demonstrated signature 11 (23), suggesting that the MMR deficiency was acquired and driven by temozolomide treatment (Fig. 2B). None of the patients with germline MMR deficiency and high-grade gliomas had received temozolomide (37, 38).

Toxicity

Toxicity data were evaluated for all patients who received at least one dose of nivolumab and are summarized in Table 2 (n = 11). Treatment was well tolerated. Seven patients experienced grade ≤2 treatment-related adverse events. The only significant autoimmune adverse event leading to treatment discontinuation was in a patient with glioblastoma (P6), who developed grade 3 pancreatitis (with grade 4 lipase elevation), after having previously experienced other immune-related side effects including grade 2 autoimmune gastritis and grade-1 noninfective cystitis. A second patient (P1) with glioma stopped following symptomatic hydrocephalus due to a rapid progression of a posterior fossa mass that subsequently needed to undergo debulking surgery.

Antitumor activity and patient survival

Disease status was evaluated every 2 months, and patients could continue therapy beyond iUPD, with subsequent scans being used to establish confirmed progression, or delayed “immune” responses, which could then be included in the best overall response (iBOR) even after discontinuation of protocol therapy (24, 25, 27, 28). Among 10 patients, nine had measurable disease, with evaluable-only disease in the patient with neuroblastoma. The initial radiological responses were complete response (CR) in one (P4), partial response (PR) in one (P6), SD in four, and PD in three patients (Table 1; Fig. 2D). Remarkably, delayed responses (iBOR) were observed in multiple patients even when the trial was terminated because of initial progression. This resulted in best overall responses to be reclassified as CR in three (P2, P4, P7), and PR in two patients (P6, P3; Fig. 2D and E). Hence, while the initial objective response rate was 20%, the best overall response was 50%. Recurrent tumors in patients who achieved iBOR exhibited significant to complete radiological regression (Figs. 3 and 4), and all who continued ICI treatment are alive at the time of reporting at a median follow-up of 37 months (range, 32.4–60) from trial enrollment. The best percentage change from baseline in target lesions is shown in Fig. 2D.

The median follow-up for the cohort was 20.8 months [95% confidence interval (CI), 11.2–30.4]. The median OS was 23.7 months (95% CI, 7.2–not reached), culminating in an estimated 2-year OS of 50% (95% CI, 27–93). The median PFS was 3.6 months (95% CI, 0.9–not reached), with an estimated 2-year PFS of 20% (95% CI, 6–69; Fig. 2F). For patients with malignant gliomas, the median OS was 16.2 months (95% CI, 7.2–not reached) with an estimated 2-year OS of 43% (95% CI, 18.2–100), while the median PFS was 2.7 months (95% CI, 0.9–not reached) with an estimated 2-year PFS of 14% (95% CI, 2.3–87.7; Fig. 2G). The delayed immune responses following initial progression contributed to the differences observed between PFS and OS, as only minority received additional nonimmune therapies after progression (Table 1).

Tumor genomic and immune biomarkers and neuroimaging response trajectories

On the basis of previous studies showing impact of tumoral genomic and immune biomarkers in determining outcome to ICI treatment (5, 29, 39–42), we investigated several biomarkers in responders and nonresponders at baseline (Table 1; Fig. 2C) and their spatiotemporal variability (Figs. 3 and 4; Supplementary Fig. S1) in select cases where repeat biopsies were performed. The unique trajectories to best overall responses in multiple patients in our trial in the context of these biomarkers revealed the following interesting observations.

An adolescent with Lynch syndrome and multifocal recurrent glioblastoma (P4), where the baseline tumor had elevated TMB, MSI, CD8 T-cell infiltration and PD-L1 expression (ref. 29; Fig. 3A; Table 1) developed initial tumor “flare” (ref. 29; pseudoprogression or iUPD; Figs. 2E and 3A) at multiple sites after the second dose of nivolumab. Clinical deterioration mandated admission to the intensive care unit. Symptoms subsided with supportive management without any immunosuppressive (steroid) treatment. He continued on protocol treatment, achieved CR, and is alive 3 years following initiation of nivolumab and 1 year after completing the trial.

The second survivor (P6) with CMMRD and recurrent multifocal glioblastoma demonstrated initial radiological progression without worsening symptoms (iUPD; Figs. 2E and 3B) and continued protocol therapy, eventually achieving PR prior to stopping the trial due to severe pancreatitis (Fig. 2E). This tumor also has extreme mutation and MSI burden, along with elevated CD8 and PD-L1 expression. After a local relapse 6 months after stopping protocol treatment due to severe toxicity, he was rechallenged with nivolumab (compassionate access) and remains in CR without toxicity recurrence at data cutoff.

A third survivor (P7) with CMMRD achieved delayed CR after sustained massive disease progression on serial imaging for 3 months on protocol treatment, prompting discontinuation of trial therapy (Figs. 2E and 3C). Notably, the tumor at baseline had highly elevated TMB and MSI, but lacked CD8 and PD-L1 expression. Continued clinical improvement while at home on hospice care prompted an imaging more than 6 months after stopping therapy. Remarkably, even in the absence of any intervening therapy, this demonstrated PR. She was restarted on nivolumab (compassionate access) and is alive in CR after completing 2 years of therapy. All three patients with aggressive CNS tumors also harbored a second, somatic POLE mutation in their tumors contributing to the ultra-hypermutation (TMB >100 mut/Mb; ref. 43) in their tumors.

The fourth survivor had Lynch syndrome and metastatic colorectal carcinoma (P2) with moderately elevated TMB but a very high MSI burden (Table 1). She demonstrated radiological SD, but pathologic CR was demonstrated when the residual lesion was resected and showed only inflammation with no detectable tumor cells (Supplementary Fig. S1). Furthermore, there was elevation of CD8 and PD-L1 expression in this second biopsy as compared with the primary specimen. She remains in remission for 60 months from enrollment (Fig. 2E).

Interestingly, two additional deceased patients with Lynch syndrome had mixed responses attributable to divergent TMB and/or tumor locations that deserve further elaboration. Sustained symptomatic progression in patient P3 following initial disease stability prompted a biopsy (Fig. 4A). An increase in TMB and MSI burden were noted. In addition, 83% of mutations in the second biopsy were novel and distinct from the primary specimen. There was also higher CD8 T-cell infiltration and PD-L1 expression as compared with baseline. The tumor continued to symptomatically progress following rechallenge of immunotherapy postsurgery, leading to midline shift and visual deterioration, and no further ICI treatment was administered thereafter. Remarkably, there was delayed PR 5 months later without further intervention. Ultimately, there was another progression after 12 months, which culminated in fatality approximately 16 months from trial enrollment.

Another patient with metastatic recurrence (P1) demonstrated progression of two of her lesions, including the primary target temporal glioma that had a moderately elevated TMB and high MSI burden (Table 1). On trial, a rapidly enlarging posterior fossa mass caused significant symptomatic hydrocephalus, requiring surgical debulking. This specimen did not have any CD8 or PD-L1 expression in contrast to the primary lesion. Furthermore, this posterior fossa lesion and the primary tumor had only 1% of mutations shared between them, suggesting that the two lesions were genomically disparate, plausibly with distinct neoantigen expression that resulted in different immune microenvironment characteristics. Though treatment was stopped culminating in rapid fatality, strikingly, a third lesion detected pretherapy that was never biopsied showed reduction in size with nivolumab treatment (Fig. 4B).

Overall, within this cohort of cancers with elevated TMB, we observed that children whose tumors harbored higher (≥median) total indel and MS indel burden had improved survival (median OS not reached for both; Fig. 5A). Association of CD8 and PD-L1 expression with response and survival were variable (Fig. 2C).

Immune biomarkers in peripheral blood

We performed exploratory immune cell subset analyses from serial blood samples collected at predetermined timepoints as indicated in the trial protocol that provided additional insights. First, responders, including patients who had delayed responses after sustained initial progression (P7, P3), had higher 4-1BB+ CD8 T cells in blood at baseline, as well as achieved a higher peak of these activated populations of T cells on immunotherapy (Fig. 5B). Nonresponders had higher regulatory T cells (CD4+) in blood at baseline, and their nadir counts in serial measurements stayed higher than responders (Fig. 5C). Furthermore, patients with persistently elevated regulatory T cells after the first 3 months on trial had a higher risk of death (Supplementary Table S2). Finally, pilot analyses of serial peripheral blood TCR-beta (TRB) rearrangement were performed for three responders and three nonresponders over the first 3 months of treatment (44). The responders had a higher total clone count, as well as higher richness (measured by Shannon diversity) and evenness (measured by Gini coefficient) of clonal distribution than nonresponders (ref. 44; Fig. 5D).

In this article, we present the first prospective clinical trial using nivolumab exclusively for children with refractory solid and brain tumors with high TMB. Best overall responses (50%) exceeded initial objective radiological response (20%), translating to ongoing prolonged ongoing survival in four of five responders with rapidly fatal cancers when continued on ICI treatment. Despite a premature termination for slow accrual and loss of funding, our trial, like select others (45), could still provide us with several important clinical and biological insights for this unique cohort of childhood cancers with elevated mutation load.

We noted that the high TMB in our cohort of refractory pediatric cancers had diverse etiologies, which plausibly had an impact on ICI response, as had also been recently proposed by an international multistakeholder group on ICI use in pediatric cancers (20). All five responders in this trial exhibited MMRD, which uniquely for our cohort of pediatric patients was predominantly of germline origin and related to well-characterized cancer predisposition syndromes. Cancers in the two children with constitutional (biallelic) MMRD demonstrated ultra-hypermutation (TMB > 100 mut/Mb) with concomitant DNA polymerase proofreading deficiency (43) and responded following ICI. Remarkably, we also witnessed responses among three of five patients with Lynch syndrome, where the cancers had both hypermutation and evidence of somatic MMRD as was demonstrated using IHC and mutational signature analysis [COSMIC SBS signatures 6, 15, 20 (31)]. It is likely that germline MMRD contributes to improved survival on immunotherapy because of the continuous obligatory mutation accumulation (43) leading to higher neoantigen expression. An additional important contributor could be the concomitant presence of high MS-indel burden (29), as frameshift mutations can generate “higher-quality” neoantigens, especially if the indels are in MS in coding loci (29, 46). It is important to note that for pediatric cancers, MSI panels developed for adults have low sensitivity and appropriate genomic assays should be incorporated for incorporation of MSI-high pediatric cancers in future ICI studies (33, 47).

In this context, it is important to highlight that the patient with GBM where MMRD was acquired following temozolomide treatment (P8) did not respond to nivolumab despite having a high TMB. As observed in adult glioblastomas that historically do not respond to ICI (7), this tumor harbored SBS 11 (7, 31) and had only limited focal areas of loss of MMR staining (Fig. 2B), suggesting subclonality for acquired MMRD, plausibly contributing to the low MSI burden and lack of ICI response (7, 29). Similarly, an elevated TMB arising from other etiologies was not associated with response to ICI. This included a patient with neuroblastoma (P5) with a defect in the homologous recombination repair (HRD) pathway harboring SBS 3 (HRD signature; ref. 31) and a patient with LFS and ACC (P10). Together, these data suggest that in children, hypermutation driven by germline MMRD is likely essential for ICI response.

The dynamic responses that we observed in several patients with germline MMRD-induced hypermutation are worth further discussion. While early tumor “flare” or pseudoprogression has been reported previously (ref. 29; P4; Fig. 3A), we witnessed clear sustained and symptomatic tumor progression on nivolumab prior to delayed responses and prolonged survival in three children (P3, P6, P7; Figs. 3B and C and 4A). Such prolonged survival with continuation of ICI postprogression has been noted in adult cancers like melanoma (48). The patient with ultra-hypermutant MMRD glioblastoma (P7) had low CD8 expression at diagnosis and sustained progression for >3 months resulting in trial termination. Repeat scans 6 months without additional therapy demonstrated response, leading us to hypothesize that MMRD-driven genomic instability can contribute to stochastic mutation accumulation over time, conferring delayed immunogenicity (43). This was objectively demonstrated in another patient (P3), where a biopsy during progression showed that the glioma now harbored a higher TMB, leading to higher immunogenicity, as demonstrated by both higher CD8 and PD-L1 expression (29) in the tumor, plausibly contributing to the delayed response. Similar evidence of radiological progression for several weeks was also noted in patient P6 before delayed and sustained response, contributing to prolonged ongoing survival. Of note, among CNS tumors, responders with lower tumor burden (P4, P6) had a relatively earlier response than those with higher disease burden (P3, P7), where despite favorable genomic biomarkers, response was delayed beyond 6 months (Fig. 2E). There is increasing evidence that a higher tumor burden negatively impacts the microenvironment and can modulate response to checkpoint inhibitors (49).

It is also important to note that we observed different lesions in the same patient, which are conventionally considered as metastatic tumors at relapse, could be biologically heterogeneous and consequently have different responses (P1; Fig. 4B). This could be attributed to random mutation accumulation secondary to MMRD in lesions at different sites. These findings highlight the challenges of conventional radiological assessments in children treated with immunotherapy and combined with the delay in responses observed by us, may account for limited responses seen in some previous studies in children receiving ICI treatment.

Finally, although limited by the small numbers, continuous collection of biomarkers in the blood and in tumor tissue at progression in our trial provided us with additional biological insights. We did not find a reliable association between PD-L1 expression and patient outcome (Fig. 2C) plausibly linked to cancer-driven immune editing (50), as MMRD tumors in our trial exhibited change in PD-L1 expression over time (Fig. 4A; refs. 51, 52). Similarly, while we observed variable and consistent elevation of CD8 (+) T cells in the microenvironment plausibly driven by neoantigen expression caused by the high TMB, this too could change over time (Fig. 4A and B), confounding associations of baseline measurements on diagnostic biopsies with response at recurrence following ICI treatment.

Interestingly, serial peripheral blood analysis during treatment on trial did show important correlation between tumor response and immune activation. Activated 4-1BB+CD8+ T cells, higher levels of which have been previously associated with antitumor immune responses (53, 54, 29), were uniformly and significantly elevated in all five responders in our study. Interestingly, patient P1, who demonstrated divergent responses at three sites including radiological response in one lesion, also had elevated activated 4-1BB+CD8+ T-cell counts (Fig. 5B). This suggested that asymmetric and atypical responses could be potentially identified by peripheral biomarkers. We also found lower regulatory T cells in the peripheral blood of responders, as has been recently linked to more effective antitumor responses (55). In addition, we observed that responders had both higher richness and evenness of their TRB clonal repertoire. While we did observe increase in skewness (Gini coefficient) of the TRB repertoire over time, we did not find enrichment of specific TCR sequences (measured by reduction of Shannon entropy) as has been previously reported in responders following ICI treatment (56–60). This is plausibly related to the high, ongoing mutation accumulation and neoantigen expression in MMRD cancers that leads to a dominance of multiple TRB clones. These intriguing findings may be specific to germline MMRD where cancers arise in the backdrop of constitutional genomic instability that need to be evaluated in more detail in future studies.

In summary, our trial shows benefit of nivolumab treatment in pediatric patients with MMRD cancers with elevated TMB, including lethal malignant gliomas, with an acceptable safety profile. During ICI treatment, clinicians need to be aware of early pseudoprogression, differential responses at distinct sites, and delayed responses due to changing tumor immunogenicity. Future approaches could involve biomarker-driven front-line use in aggressive tumors like malignant ultra-hypermutant gliomas with favorable biomarkers (61, 62) and exploration of combinatorial approaches in patients failing anti-PD1 monotherapy (63). Routine use of MMR IHC, the incorporation of TMB and mutational signatures into cancer sequencing analyses, and screening patients from low and middle-income countries (64, 65) with a higher prevalence of germline MMRD can improve outcomes for these patients worldwide.

L.C. Sambira Nahum reports grants from BioCanRx Grant during the conduct of the study. S. Sudhaman reports other support from Natera Inc outside the submitted work. G. Getz reports personal fees from Scorpion Therapeutics during the conduct of the study; grants from Pharmacyclics outside the submitted work; has a patent for MSMutect pending, a patent for MSMutSig pending, and a patent for MSIDetect pending; and is an inventor on additional patent applications related to POLYSOLVER, SignatureAnalyzer-GPU, and MinimuMM-seq. Y.E. Maruvka reports grants from Israeli Science Foundation and grants from Israel Cancer Association during the conduct of the study and personal fees from ForeseeGenomics outside the submitted work. B. Ertl-Wagner reports personal fees from Bayer Healthcare outside the submitted work; and spouse is an employee of Siemens Healthineers. T. Pugh reports personal fees from AstraZeneca, Chrysalis Biomedical Advisors, Merck, and SAGA Diagnostics, and grants from Roche/Genentech outside the submitted work. C. Hawkins reports grants from CIHR, CCS, The Cure Starts Now, Meagan's Hug, and Genome Canada outside the submitted work; and has a patent for LOGIC pending. E. Bouffet reports grants from BMS during the conduct of the study and personal fees from Novartis and Alexion (AstraZeneca) outside the submitted work. D.A. Morgenstern reports grants from BMS during the conduct of the study; personal fees from ymAbs Therapeutics, Clarity Pharmaceuticals, Regeneron, and Rayzebio Inc, and other support from Abbvie and Oncoheroes Biosciences outside the submitted work. No disclosures were reported by the other authors.

The funders had no role in the collection, analyses, or interpretation of data; in the writing of the article; or in the decision to publish the results.

A. Das: Software, formal analysis, funding acquisition, validation, investigation, visualization, methodology, writing–original draft, writing–review and editing. U. Tabori: Conceptualization, resources, data curation, formal analysis, supervision, funding acquisition, investigation, methodology, writing–review and editing. L.C. Sambira Nahum: Resources, investigation, project administration. N.B. Collins: Conceptualization, resources, investigation, project administration. R. Deyell: Conceptualization, resources, investigation, project administration. R. Dvir: Resources, investigation. C. Faure-Conter: Resources, investigation. T.E. Hassall: Resources, investigation. J.E. Minturn: Resources, investigation. M. Edwards: Resources. E. Brookes: Resources. V. Bianchi: Resources. A. Levine: Investigation. S.C. Stone: Investigation, visualization. S. Sudhaman: Investigation, visualization. S. Sanchez-Ramirez: Investigation, visualization. A.B. Ercan: Investigation. L. Stengs: Resources. J. Chung: Investigation. L. Negm: Investigation. G. Getz: Software, investigation. Y.E. Maruvka: Software, validation. B. Ertl-Wagner: Formal analysis, investigation, visualization. P.S. Ohashi: Formal analysis, investigation, visualization. T. Pugh: Formal analysis, investigation, visualization. C. Hawkins: Formal analysis, investigation. E. Bouffet: Conceptualization, resources, formal analysis, supervision, funding acquisition, investigation, methodology, writing–review and editing. D.A. Morgenstern: Conceptualization, resources, data curation, formal analysis, supervision, funding acquisition, validation, investigation, visualization, methodology, writing–original draft, project administration, writing–review and editing.

The authors thank Aiman Siddiqi, Emily Taylor, Shaherose Nanji, Entela Zaffino, Nicole Sarvaria, and Alana Bougie at Hospital for Sick Children and the team at Ozmosis for assistance with trial management. Nivolumab and funding for the study and correlative biology were provided by BMS. A. Das would like to acknowledge the support of the St. Baldrick's Foundation International Scholar Award (with support from Kai Slockers Pediatric Cancer Research Fund), Stand Up to Cancer Maverick and Hold'em for Life awards. P.S. Ohashi would like to acknowledge the support from We Love You Conni Foundation.

The publication costs of this article were defrayed in part by the payment of publication fees. Therefore, and solely to indicate this fact, this article is hereby marked “advertisement” in accordance with 18 USC section 1734.

Note: Supplementary data for this article are available at Clinical Cancer Research Online (http://clincancerres.aacrjournals.org/).

1.
Liu
D
,
Schilling
B
,
Liu
D
,
Sucker
A
,
Livingstone
E
,
Jerby-Arnon
L
, et al
.
Integrative molecular and clinical modeling of clinical outcomes to PD1 blockade in patients with metastatic melanoma
.
Nat Med
2019
;
25
:
1916
27
.
2.
Garon
EB
,
Rizvi
NA
,
Hui
R
,
Leighl
N
,
Balmanoukian
AS
,
Eder
JP
, et al
.
Pembrolizumab for the treatment of non-small-cell lung cancer
.
N Engl J Med
2015
;
372
:
2018
28
.
3.
Topalian
SL
,
Hodi
FS
,
Brahmer
JR
,
Gettinger
SN
,
Smith
DC
,
McDermott
DF
, et al
.
Five-year survival and correlates among patients with advanced melanoma, renal cell carcinoma, or non-small cell lung cancer treated with nivolumab
.
JAMA Oncol
2019
;
5
:
1411
20
.
4.
Mandal
R
,
Samstein
RM
,
Lee
K-W
,
Havel
JJ
,
Wang
H
,
Krishna
C
, et al
.
Genetic diversity of tumors with mismatch repair deficiency influences anti-PD-1 immunotherapy response
.
Science
2019
;
364
:
485
91
.
5.
Rousseau
B
,
Foote
MB
,
Maron
SB
,
Diplas
BH
,
Lu
S
,
Argilés
G
, et al
.
The spectrum of benefit from checkpoint blockade in hypermutated tumors
.
N Engl J Med
2021
;
384
:
1168
70
.
6.
Notarangelo
G
,
Spinelli
JB
,
Perez
EM
,
Baker
GJ
,
Kurmi
K
,
Elia
I
, et al
.
Oncometabolite d-2HG alters T cell metabolism to impair CD8(+) T cell function
.
Science
2022
;
377
:
1519
29
.
7.
Touat
M
,
Li
YY
,
Boynton
AN
,
Spurr
LF
,
Iorgulescu
JB
,
Bohrson
CL
, et al
.
Mechanisms and therapeutic implications of hypermutation in gliomas
.
Nature
2020
;
580
:
517
23
.
8.
Sampson
JH
,
Gunn
MD
,
Fecci
PE
,
Ashley
DM
.
Brain immunology and immunotherapy in brain tumours
.
Nat Rev Cancer
2020
;
20
:
12
25
.
9.
Khasraw
M
,
Reardon
DA
,
Weller
M
,
Sampson
JH
.
PD-1 inhibitors: do they have a future in the treatment of glioblastoma?
Clin Cancer Res
2020
;
26
:
5287
96
.
10.
Davis
KL
,
Fox
E
,
Merchant
MS
,
Reid
JM
,
Kudgus
RA
,
Liu
X
, et al
.
Nivolumab in children and young adults with relapsed or refractory solid tumours or lymphoma (ADVL1412): a multicentre, open-label, single-arm, phase 1–2 trial
.
Lancet Oncol
2020
;
21
:
541
50
.
11.
Geoerger
B
,
Kang
HJ
,
Yalon-Oren
M
,
Marshall
LV
,
Vezina
C
,
Pappo
A
, et al
.
Pembrolizumab in paediatric patients with advanced melanoma or a PD-L1-positive, advanced, relapsed, or refractory solid tumour or lymphoma (KEYNOTE-051): interim analysis of an open-label, single-arm, phase 1–2 trial
.
Lancet Oncol
2020
;
21
:
121
33
.
12.
Geoerger
B
,
Zwaan
CM
,
Marshall
LV
,
Michon
J
,
Bourdeaut
F
,
Casanova
M
, et al
.
Atezolizumab for children and young adults with previously treated solid tumours, non-Hodgkin lymphoma, and Hodgkin lymphoma (iMATRIX): a multicentre phase 1–2 study
.
Lancet Oncol
2020
;
21
:
134
44
.
13.
Loeb
DM
,
Lee
JW
,
Morgenstern
DA
,
Samson
Y
,
Uyttebroeck
A
,
Lyu
CJ
, et al
.
Avelumab in paediatric patients with refractory or relapsed solid tumours: dose-escalation results from an open-label, single-arm, phase 1/2 trial
.
Cancer Immunol Immunother
2022
;
71
:
2485
95
.
14.
Merchant
MS
,
Wright
M
,
Baird
K
,
Wexler
LH
,
Rodriguez-Galindo
C
,
Bernstein
D
, et al
.
Phase I clinical trial of ipilimumab in pediatric patients with advanced solid tumors
.
Clin Cancer Res
2016
;
22
:
1364
70
.
15.
Haworth
KB
,
Leddon
JL
,
Chen
CY
,
Horwitz
EM
,
Mackall
CL
,
Cripe
TP
.
Going back to class I: MHC and immunotherapies for childhood cancer
.
Pediatr Blood Cancer
2015
;
62
:
571
6
.
16.
Simon
AK
,
Hollander
GA
,
McMichael
A
.
Evolution of the immune system in humans from infancy to old age
.
Proc Biol Sci
2015
;
282
:
20143085
.
17.
Routy
B
,
Le Chatelier
E
,
Derosa
L
,
Duong
CPM
,
Alou
MT
,
Daillère
R
, et al
.
Gut microbiome influences efficacy of PD-1-based immunotherapy against epithelial tumors
.
Science
2018
;
359
:
91
7
.
18.
Majzner
RG
,
Simon
JS
,
Grosso
JF
,
Martinez
D
,
Pawel
BR
,
Santi
M
, et al
.
Assessment of programmed death-ligand 1 expression and tumor-associated immune cells in pediatric cancer tissues
.
Cancer
2017
;
123
:
3807
15
.
19.
Long
AH
,
Morgenstern
DA
,
Leruste
A
,
Bourdeaut
F
,
Davis
KL
.
Checkpoint immunotherapy in pediatrics: here, gone, and back again
.
Am Soc Clin Oncol Educ Book
2022
;
42
:
1
14
.
20.
Pearson
ADJ
,
Rossig
C
,
Lesa
G
,
Diede
SJ
,
Weiner
S
,
Anderson
J
, et al
.
ACCELERATE and European medicines agency paediatric strategy forum for medicinal product development of checkpoint inhibitors for use in combination therapy in paediatric patients
.
Eur J Cancer
2020
;
127
:
52
66
.
21.
Le
DT
,
Uram
JN
,
Wang
H
,
Bartlett
BR
,
Kemberling
H
,
Eyring
AD
, et al
.
PD-1 blockade in tumors with mismatch-repair deficiency
.
N Eng J Med
2015
;
372
:
2509
20
.
22.
Campbell
BB
,
Light
N
,
Fabrizio
D
,
Zatzman
M
,
Fuligni
F
,
de Borja
R
, et al
.
Comprehensive analysis of hypermutation in human cancer
.
Cell
2017
;
171
:
1042
56
.
23.
Alexandrov
LB
,
Nik-Zainal
S
,
Wedge
DC
,
Aparicio
SAJR
,
Behjati
S
,
Biankin
AV
, et al
.
Signatures of mutational processes in human cancer
.
Nature
2013
;
500
:
415
21
.
24.
Seymour
L
,
Bogaerts
J
,
Perrone
A
,
Ford
R
,
Schwartz
LH
,
Mandrekar
S
, et al
.
iRECIST: guidelines for response criteria for use in trials testing immunotherapeutics
.
Lancet Oncol
2017
;
18
:
e143
52
.
25.
Park
JR
,
Bagatell
R
,
Cohn
SL
,
Pearson
AD
,
Villablanca
JG
,
Berthold
F
, et al
.
Revisions to the international neuroblastoma response criteria: a consensus statement from the national cancer institute clinical trials planning meeting
.
J Clin Oncol
2017
;
35
:
2580
7
.
26.
Ellingson
BM
,
Huang
RY-K
,
Villanueva-Meyer
J
,
Lim-Fat
MJ
,
George
E
,
Iorgulescu
B
, et al
.
Estimated clinical efficacy and radiographic response characteristics of PD1 inhibition in newly diagnosed and recurrent glioblastoma in clinical practice: a report from the iRANO working group
.
J Clin Oncol
38
:
15s
,
2020
(suppl; abstr 2521).
27.
Okada
H
,
Weller
M
,
Huang
R
,
Finocchiaro
G
,
Gilbert
MR
,
Wick
W
, et al
.
Immunotherapy response assessment in neuro-oncology: a report of the RANO working group
.
Lancet Oncol
2015
;
16
:
e534
42
.
28.
Ellingson
BM
,
Chung
C
,
Pope
WB
,
Boxerman
JL
,
Kaufmann
TJ
.
Pseudoprogression, radionecrosis, inflammation or true tumor progression? Challenges associated with glioblastoma response assessment in an evolving therapeutic landscape
.
J Neurooncol
2017
;
134
:
495
504
.
29.
Das
A
,
Sudhaman
S
,
Morgenstern
D
,
Coblentz
A
,
Chung
J
,
Stone
SC
, et al
.
Genomic predictors of response to PD-1 inhibition in children with germline DNA replication repair deficiency
.
Nat Med
2022
;
28
:
125
35
.
30.
Rosenthal
R
,
McGranahan
N
,
Herrero
J
,
Taylor
BS
,
Swanton
C
.
DeconstructSigs: delineating mutational processes in single tumors distinguishes DNA repair deficiencies and patterns of carcinoma evolution
.
Genome Biol
2016
;
17
:
31
.
31.
Alexandrov
LB
,
Kim
J
,
Haradhvala
NJ
,
Huang
MN
,
Tian Ng
AW
,
Wu
Y
, et al
.
The repertoire of mutational signatures in human cancer
.
Nature
2020
;
578
:
94
101
.
32.
Maruvka
YE
,
Mouw
KW
,
Karlic
R
,
Parasuraman
P
,
Kamburov
A
,
Polak
P
, et al
.
Analysis of somatic microsatellite indels identifies driver events in human tumors
.
Nat Biotechnol
2017
;
35
:
951
9
.
33.
Chung
J
,
Maruvka
YE
,
Sudhaman
S
,
Kelly
J
,
Haradhvala
NJ
,
Bianchi
V
, et al
.
DNA polymerase and mismatch repair exert distinct microsatellite instability signatures in normal and malignant human cells
.
Cancer Discov
2021
;
11
:
1176
91
.
34.
Mulder
DT
,
Mahé
ER
,
Dowar
M
,
Hanna
Y
,
Li
T
,
Nguyen
LT
, et al
.
CapTCR-seq: hybrid capture for T-cell receptor repertoire profiling
.
Blood Adv
2018
;
2
:
3506
14
.
35.
Benjamini
Y
,
Hochberg
Y
.
Controlling the false discovery rate: a practical and powerful approach to multiple testing
.
J R Statist Soc B
1995
;
57
:
289
300
.
36.
Benjamini
Y
,
Cohen
R
.
Weighted false discovery rate controlling procedures for clinical trials
.
Biostatistics
2017
;
18
:
91
104
.
37.
Tabori
U
,
Hansford
JR
,
Achatz
MI
,
Kratz
CP
,
Plon
SE
,
Frebourg
T
, et al
.
Clinical management and tumor surveillance recommendations of inherited mismatch repair deficiency in childhood
.
Clin Cancer Res
2017
;
23
:
e32
7
.
38.
Liu
L
,
Markowitz
S
,
Gerson
SL
.
Mismatch repair mutations override alkyltransferase in conferring resistance to temozolomide but not to 1,3-bis(2-chloroethyl)nitrosourea
.
Cancer Res
1996
;
56
:
5375
9
.
39.
Marabelle
A
,
Fakih
M
,
Lopez
J
,
Shah
M
,
Shapira-Frommer
R
,
Nakagawa
K
, et al
.
Association of tumour mutational burden with outcomes in patients with advanced solid tumours treated with pembrolizumab: prospective biomarker analysis of the multicohort, open-label, phase 2 KEYNOTE-158 study
.
Lancet Oncol
2020
;
21
:
1353
65
.
40.
Le
DT
,
Kim
TW
,
Van Cutsem
E
,
Geva
R
,
Jäger
D
,
Hara
H
, et al
.
Phase II open-label study of pembrolizumab in treatment-refractory, microsatellite instability-high/mismatch repair-deficient metastatic colorectal cancer: KEYNOTE-164
.
J Clin Oncol
2020
;
38
:
11
9
.
41.
Blank
CU
,
Haanen
JB
,
Ribas
A
,
Schumacher
TN
.
Cancer immunology. The “cancer immunogram.”
Science
2016
;
352
:
658
60
.
42.
Bai
R
,
Lv
Z
,
Xu
D
,
Cui
J
.
Predictive biomarkers for cancer immunotherapy with immune checkpoint inhibitors
.
Biomark Res
2020
;
8
:
34
.
43.
Shlien
A
,
Campbell
BB
,
de Borja
R
,
Alexandrov
LB
,
Merico
D
,
Wedge
D
, et al
.
Combined hereditary and somatic mutations of replication error repair genes result in rapid onset of ultra-hypermutated cancers
.
Nat Genet
2015
;
47
:
257
62
.
44.
Chiffelle
J
,
Genolet
R
,
Perez
MA
,
Coukos
G
,
Zoete
V
,
Harari
A
.
T-cell repertoire analysis and metrics of diversity and clonality
.
Curr Opin Biotechnol
2020
;
65
:
284
95
.
45.
Morgan
CJ
.
Statistical issues associated with terminating a clinical trial due to slow enrollment
.
J Nucl Cardiol
2017
;
24
:
525
6
.
46.
Roudko
V
,
Bozkus
CC
,
Orfanelli
T
,
McClain
CB
,
Carr
C
,
O'Donnell
T
, et al
.
Shared Immunogenic poly-epitope frameshift mutations in microsatellite unstable tumors
.
Cell
2020
;
183
:
1634
49
.
47.
Chung
J
,
Negm
L
,
Bianchi
V
,
Stengs
L
,
Das
A
,
Liu
ZA
, et al
.
Genomic microsatellite signatures identify germline mismatch repair deficiency and risk of cancer onset
.
J Clin Oncol
2023
;
41
:
766
77
.
48.
Beaver
JA
,
Hazarika
M
,
Mulkey
F
,
Mushti
S
,
Chen
H
,
He
K
, et al
.
Patients with melanoma treated with an anti-PD-1 antibody beyond RECIST progression: a US Food and Drug Administration pooled analysis
.
Lancet Oncol
2018
;
19
:
229
39
.
49.
Dall'Olio
FG
,
Marabelle
A
,
Caramella
C
,
Garcia
C
,
Aldea
M
,
Chaput
N
, et al
.
Tumour burden and efficacy of immune-checkpoint inhibitors
.
Nat Rev Clin Oncol
2022
;
19
:
75
90
.
50.
Yi
M
,
Niu
M
,
Xu
L
,
Luo
S
,
Wu
K
.
Regulation of PD-L1 expression in the tumor microenvironment
.
J Hematol Oncol
2021
;
14
:
10
.
51.
Xue
W
,
Li
W
,
Zhang
T
,
Li
Z
,
Wang
Y
,
Qiu
Y
, et al
.
Anti-PD1 up-regulates PD-L1 expression and inhibits T-cell lymphoma progression: possible involvement of an IFN-gamma-associated JAK-STAT pathway
.
Onco Targets Ther
2019
;
12
:
2079
88
.
52.
Takahashi
T
,
Tateishi
A
,
Bychkov
A
,
Fukuoka
J
.
Remarkable alteration of PD-L1 expression after immune checkpoint therapy in patients with non-small-cell lung cancer: two autopsy case reports
.
Int J Mol Sci
2019
;
20
:
2578
.
53.
Ugolini
A
,
Nuti
M
.
CD137(+) T-cells: protagonists of the immunotherapy revolution
.
Cancers
2021
;
13
:
456
.
54.
Clouthier
DL
,
Lien
SC
,
Yang
SYC
,
Nguyen
LT
,
Manem
VSK
,
Gray
D
, et al
.
An interim report on the investigator-initiated phase 2 study of pembrolizumab immunological response evaluation (INSPIRE)
.
J Immunother Cancer
2019
;
7
:
72
.
55.
Yoshida
K
,
Okamoto
M
,
Sasaki
J
,
Kuroda
C
,
Ishida
H
,
Ueda
K
, et al
.
Anti-PD-1 antibody decreases tumour-infiltrating regulatory T cells
.
BMC Cancer
2020
;
20
:
25
.
56.
Poran
A
,
Scherer
J
,
Bushway
ME
,
Besada
R
,
Balogh
KN
,
Wanamaker
A
, et al
.
Combined TCR repertoire profiles and blood cell phenotypes predict melanoma patient response to personalized neoantigen therapy plus anti-PD-1
.
Cell Rep Med
2020
;
1
:
100141
.
57.
Ott
PA
,
Hu-Lieskovan
S
,
Chmielowski
B
,
Govindan
R
,
Naing
A
,
Bhardwaj
N
, et al
.
A phase Ib trial of personalized neoantigen therapy plus anti-PD-1 in patients with advanced melanoma, non-small cell lung cancer, or bladder cancer
.
Cell
2020
;
183
:
347
62
.
58.
Kato
T
,
Kiyotani
K
,
Tomiyama
E
,
Koh
Y
,
Matsushita
M
,
Hayashi
Y
, et al
.
Peripheral T cell receptor repertoire features predict durable responses to anti-PD-1 inhibitor monotherapy in advanced renal cell carcinoma
.
Oncoimmunology
2021
;
10
:
1862948
.
59.
Postow
MA
,
Manuel
M
,
Wong
P
,
Yuan
J
,
Dong
Z
,
Liu
C
, et al
.
Peripheral T cell receptor diversity is associated with clinical outcomes following ipilimumab treatment in metastatic melanoma
.
J Immunother Cancer
2015
;
3
:
23
.
60.
Han
J
,
Duan
J
,
Bai
H
,
Wang
Y
,
Wan
R
,
Wang
X
, et al
.
TCR repertoire diversity of peripheral PD-1(+)CD8(+) T cells predicts clinical outcomes after immunotherapy in patients with non-small cell lung cancer
.
Cancer Immunol Res
2020
;
8
:
146
54
.
61.
Larkin
T
,
Das
A
,
Bianchi
V
,
Sudhaman
S
,
Chung
J
,
Alsafwani
N
, et al
.
Upfront adjuvant immunotherapy of replication repair–deficient pediatric glioblastoma with chemoradiation-sparing approach
.
JCO Precis Oncol
2021
;
5
:
1426
31
.
62.
Rittberg
R
,
Harlos
C
,
Rothenmund
H
,
Das
A
,
Tabori
U
,
Sinha
N
, et al
.
Immune checkpoint inhibition as primary adjuvant therapy for an IDH1-mutant anaplastic astrocytoma in a patient with CMMRD: a case report-usage of immune checkpoint inhibition in CMMRD
.
Curr Oncol
2021
;
28
:
757
66
.
63.
Campbell
BB
,
Galati
MA
,
Stone
SC
,
Riemenschneider
AN
,
Edwards
M
,
Sudhaman
S
, et al
.
Mutations in the RAS/MAPK pathway drive replication repair deficient hypermutated tumors and confer sensitivity to MEK inhibition
.
Cancer Discov
2021
;
11
:
1454
67
.
64.
Amayiri
N
,
Tabori
U
,
Campbell
B
,
Bakry
D
,
Aronson
M
,
Durno
C
, et al
.
High frequency of mismatch repair deficiency among pediatric high grade gliomas in Jordan
.
Int J Cancer
2016
;
138
:
380
5
.
65.
Alphones
S
,
Chatterjee
U
,
Singh
A
,
Das
A
,
Zameer
L
,
Achari
R
, et al
.
Immunohistochemical screening for mismatch repair protein deficiency in paediatric high-grade gliomas - institutional experience and review of literature
.
Childs Nerv Syst
2021
;
37
:
2521
30
.
This open access article is distributed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0) license.