Background Tumor mutational burden (TMB) has been associated with immune therapy response, and the Food and Drug Administration (FDA) has granted accelerated approval to pembrolizumab for the treatment of TMB-H solid tumors. However, there is a lack of data of TMB-H prevalence in children with solid tumors tested by FDA-approved companion diagnostics (FoundationOne®CDx (F1CDx®)).

Methods This was a retrospective cross-sectional study performed in the real-world FoundationCore dataset, which is a US based pan-tumor and multi-center derived genomics dataset. Patients <18 years with malignant solid tumors with available profiling by F1CDx or FoundationOne® (F1®, the prior version of F1CDx) between 2013-2023 were included. TMB was calculated as the number of nondriver somatic coding mutations per megabase (mut/Mb) of genome sequenced; TMB-H was defined as ≥10 mut/Mb. Microsatellite instability (MSI) status was determined using an algorithm which calculates the fraction of microsatellite loci determined to be altered or unstable based on a genome wide analysis across >2000 loci. Admixture-derived genomic ancestry fractions were inferred using a single nucleotide polymorphism-based approach. Subgroup differences were examined using Chi-square tests with two-sided alpha = 0.05.

Results A total of 3559 children were included, 48% were female; 37% were 12 or older; about 60% had predominant European ancestry. Most common tumor groups included glioma (37%), CNS non-glioma (18%), and peripheral nervous system (PNS) tumors (11%). For disease stage at testing, most patients (77%) had missing information, most testing tissues (63%) were from primary tumor site. After excluding 26 MSI-H patients (19 of which were TMB-H), in 3533 non MSI-H patients, the overall TMB-H prevalence was 1.4% (95% CI: 1.1-1.9%). The prevalence was consistent among tests, F1: 1.46%, F1CDx:1.37%. There were no significant differences in TMB-H prevalence between sex, ancestry group, tissue of origin, or disease stage. Older children (12-17 years) had higher prevalence than younger groups (2.2% vs 1.0%, p=0.006). TMB-H prevalence varied by tumor types but was <3% across all examined subgroups (glioma: 1.62%, CNS non-glioma: 0.64%, peripheral nervous system tumors: 2.56%). See table 1 for details.

Among the 50 TMB-H patients, 56% were 12 years or older; 52% were boys; 66% had predominant European ancestry. The most common tumors were gliomas (n=21), PNS tumors (n=10). Other tumors (n<5) included cancers from CNS non-glioma, germ cell, etc.

Conclusions The observed overall real-world pediatric prevalence of TMB-H assessed by F1CDx/F1 (≥10 mut/Mb) is low (~1%). Prevalence was low across all subgroups examined (< 3%)).

Acknowledgements We would like to extend our gratitude to Steven Senglaub for his exceptional study management throughout the course of this project, which ensures the quality and the success of this study. We are deeply appreciative.

Ethics Approval Institutional IRB/EC review board approval was obtained by Foundation Medicine prior to study conduct. A waiver was obtained from the Western Institutional Review Board (Protocol #20152817).

Consent Approval for this study, including a waiver of informed consent and Health Insurance Portability and Accountability Act waiver of authorization, was granted under 45 CFR § 46.116 based on review and determination that this research meets the following requirements: (i) the research involves no more than minimal risk to the subjects; (ii) the research could not practicably be carried out without the requested waiver; (iii) the waiver will not adversely affect the rights and welfare of the subjects.