Genomic Sequencing Identifies Pediatric Therapy-Related Myeloid Neoplasm Defining Variants Prior to Clinical Presentation

Xiaotu Ma, PhD

Somatic KMT2A rearrangements, RAS/MAPK mutations, RUNX1 and TP53 alterations, and MECOM overexpression were all found to be common oncogenic drivers of pediatric therapy-related myeloid neoplasms, arising predominantly after exposure to cytotoxic therapy and identifiable at least 1 year prior to morphologic evidence of neoplasm, according to findings from a study published in Nature Communications.¹

Additionally, cases with atypical MECOM expression secondary to enhancer hijacking were identified. However, unlike adults with therapy-related myeloid neoplasms, no evidence of pre-existing minor therapy-related myeloid neoplasm clones were identified, including those with TP53 mutations. Instead, the majority of cases harbored unrelated clones as a result of cytotoxic therapy exposure. Also identified were rare cases of lineage switch disease vs true secondary neoplasms.

“This work indicates that we can detect this type of malignancy early, to study if preventative therapies could benefit patients,” said Xiaotu Ma, PhD, senior study author and assistant member in the Departments of Computational Biology and the Comprehensive Cancer Center at St. Jude Children’s Research Hospital.

Prior exposure to cytotoxic therapy predisposes children to pediatric therapy-related myeloid neoplasms and is associated with a poor prognosis. However, the germline and somatic alterations that are responsible for the growth of these myeloid neoplasms in children and the reason for their onset had not previously been described in detail.

“One thing that we’ve known for a long time is once kids develop this secondary tumor, the outcome is really poor,” said senior study author, Jeffery Klco, MD, PhD, associate member, director of the Division of Hematopathology and Molecular Pathology, and medical director of Hematopathology and Immunopathology at St. Jude Children’s Research Hospital. “The alterations that drive these tumors are different in children than they are in adults, underscoring the need to study these tumors specifically in pediatrics.”

To that end, investigators used whole exome, whole genome, and/or RNA sequencing to classify the genomic profile of 84 cases of pediatric therapy-related myeloid neoplasms (therapy-related myelodysplastic syndrome: n = 28; therapy-related acute myeloid leukemia [AML]: n = 56). Samples included tumor and non-tumor tissue for 62 cases and non-tumor material for 22 cases.

Initial diagnoses included hematologic (70%), solid (27%), and brain (3%) neoplasms. The median age at therapy-related myeloid neoplasm diagnosis was 13.6 years (range, 1.2-24.6), and the time to therapy-related myeloid neoplasm after initial diagnosis varied from 0.7 years to 16.2 years (median, 2.9). Somatic variants identified from whole-genome sequencing (median coverage, 50x) or whole-exome sequencing (112x) were validated by targeted resequencing (641x).

A mean of 28 (range, 1-188) somatic mutations per patient were identified, far surpassing the mutational burden found in pediatric primary myelodysplastic syndrome (5 mutations/patient; P < .001) and pediatric de novo core-binding factor AML (13 mutations/patient; P < .001).

Investigators also used targeted capture resequencing and a bioinformatic error suppression approach to understand the timing of acquisition and evolution of the somatic mutations for 37 cases using samples from interval time points prior to the development of therapy-related myeloid neoplasms.

The results indicated that somatic variants most commonly presented after exposure to cytotoxic therapy (n = 23/26; 88%). Moreover, these acquired mutations were able to be detected up to 748 days (mean: 405 days; range, 118-748) prior to morphologic evidence of therapy-related myeloid neoplasms.

A total of 3 cases were found to be clonally related to the primary malignancy; these included a therapy-related myelodysplastic syndrome that developed 8 months after AML and 2 cases where the initial lymphoid malignancy––either acute lymphoblastic leukemia or non-Hodgkin lymphoma––and therapy-related myeloid neoplasm developed from a common clone that subsequently underwent a lineage switch.

Unlike adult therapy-related myeloid neoplasms, the somatic TP53 variants could not be detected with ultra-deep amplicon sequencing (72,000x) and bioinformatic error suppression in pre-treatment samples.

References

1. Study reveals mutations that drive therapy-related myeloid neoplasms in children. News release. St. Jude Children’s Research Hospital. Published February 12, 2021. Accessed March 29, 2021. https://bit.ly/39rPnk1.
2. Schwartz JR, Ma J, Klco JM, et al. The acquisition of molecular drivers in pediatric therapy-related myeloid neoplasms. Nat Com. 2021;12:985. doi:10.1038/s41467-021-21255-8