8 Hidden Structural Variants in Childhood Cancer: Landmark St. Jude Study Reveals DNA Rearrangement Patterns That Drive Pediatric Leukemia and Solid Tumors

Structural variants—genomic "cut-and-paste" errors where DNA breaks and rejoins incorrectly—account for over 60% of driver mutations in childhood cancers, according to groundbreaking research published today in Cancer Cell by scientists at St. Jude Children's Research Hospital and the National Cancer Institute. The comprehensive analysis of 1,616 pediatric cancer genomes reveals previously unknown mechanisms driving pediatric malignancies and provides a critical resource for developing targeted therapies.

Understanding Structural Variants: The Hidden Drivers of Childhood Cancer

Genomic structural variants (SVs) represent large-scale rearrangements of DNA segments that occur when chromosomes break in one location and incorrectly rejoin elsewhere. Unlike point mutations—small, single-letter changes that accumulate gradually with age and environmental exposure—structural variants can dramatically alter gene function through deletions, duplications, inversions, or translocations between chromosomes.

"For years, we assumed pediatric cancer had lower mutation burdens than adult cancers," said corresponding author Jinghui Zhang, PhD, St. Jude Department of Computational Biology. "But our analysis reveals that structural variant burden in pediatric blood cancer is actually higher than adult counterparts."

This counterintuitive finding makes biological sense: children simply haven't had time to accumulate the point mutations that drive adult cancers through decades of aging and environmental damage. Instead, childhood cancers rely heavily on structural variants—large-scale genomic disruptions that occur during early development and fundamentally reshape the genome.

The Landmark Dataset: 1,616 Pediatric Cancer Genomes Analyzed

The research team curated structural variants from publicly available whole genome sequencing data spanning 16 major pediatric cancer types, including hematological malignancies (908 patients), brain tumors (183 patients), and solid tumors (525 patients). This represented the first and largest dataset specifically focused on childhood cancer genomic structure variations.

The investigators then compared their pediatric cohort to 2,203 adult cancer genomes from the Pan-Cancer Analysis of Whole Genomes (PCAWG) consortium, identifying what uniquely drives childhood cancers versus adult malignancies.

Key Findings: SV Burden Varies Dramatically Across Cancer Types

The median structural variant burden ranged from just 1 to 249 SVs per pediatric cancer genome—a 100-fold variation across different cancer types that points to fundamentally distinct mutational processes:

• Highest SV burden: Osteosarcoma (median 249 SVs), adrenocortical cancer (median 54.5 SVs), and high-grade glioma (median 45 SVs)
• Lowest SV burden: Acute myeloid leukemia, low-grade glioma, ependymoma, Ewing sarcoma, and Wilms tumor (median ≤5 SVs each)

Notably, the highest SV burden cancers correlated strongly with TP53 mutation rates—91.1% in osteosarcoma, 80.0% in adrenocortical cancer, and 58.1% in high-grade glioma. However, rhabdomyosarcoma proved an exception: despite a median burden of 52 SVs, it rarely harbored TP53 mutations (only 11.1% prevalence).

RAG-Mediated Recombination: The Primary Mechanism Driving Pediatric Leukemia

The most significant discovery centers on errors in a specific DNA rearrangement process called RAG-mediated recombination—a normally beneficial mechanism that immune cells use to create diverse antibody and T-cell receptor repertoires.

During normal immune development, RAG1 and RAG2 proteins deliberately cut DNA at specialized sites called recombination signal sequences (RSS) to reshuffle genetic segments and generate the vast protein diversity needed to recognize countless pathogens. However, when this process goes awry, it creates structural variants that drive cancer.

The SV7 Signature: A Smoking Gun for RAG Activity

Through analysis of 10 prevalent mutational signatures across cancer types, researchers identified SV7 as particularly abundant in both B-cell acute lymphoblastic leukemia (B-ALL) and T-cell acute lymphoblastic leukemia (T-ALL)—together representing the most common childhood cancers.

Key evidence implicating RAG-mediated recombination in SV7:

1. Elevated RAG1/2 expression: Samples positive for SV7 showed significantly elevated expression of both RAG genes
2. RSS heptamer enrichment: The conserved heptamer sequence (5′–CACAGTG–3′) appeared within 20 base pairs of 85.2% of recurrent SV hotspots
3. Hotspot patterns: Recurrent deletions clustered near RSS sites in immune loci and driver genes
4. Adult cancer parallels: SV7's propensity in adult lymphoid cancers—also suspected to involve RAG-mediated recombination—provided corroborating evidence

"This understanding effectively widens the spectrum of possible leukemic drivers and gives new weight to the significance of structural variants in pediatric blood cancers," Zhang explained.

Recurrent Hotspots: A Genome-Wide Pattern

The investigation revealed 332 SV hotspots across the pediatric genome, predominantly contributed by B-ALL (91.5%) and T-ALL (3.3%). These hotspots showed striking patterns:

• 88% were deletions (292 of 332), with only 7 inversions or translocations causing gene fusions
• 63% occurred in immune-related regions, consistent with normal RAG activity at antigen receptor loci
• 123 hotspots affected non-immune genes, including known drivers like RB1, CRLF2, KDM6A, and LEF1
• 65% of non-immune hotspots contained predicted RSS sites, suggesting aberrant RAG activity

The RB1 gene deletion hotspot in intron 17 exemplifies this phenomenon: three recurrent breakpoints within 20 bp of RSS heptamers connected to neighboring genes or intergenic regions, creating two common deletions observed across 16 ALL cases.

Complex Structural Variants and Chromothripsis: Catastrophic Genomic Events

Beyond localized rearrangements, the study identified complex structural variants involving multiple genomic locations in single catastrophic events—a phenomenon called chromothripsis (Greek for "chromosome shattering").

Chromothripsis occurred in:

• 1.5% of hematological malignancies
• 13.1% of brain tumors
• 14.9% of solid tumors

Samples with chromothripsis showed a 4.3-fold median increase in structural variant burden compared to cases without this catastrophic event. In osteosarcoma patients, researchers observed templated insertions forming cycles or chains of inter-chromosomal rearrangements—patterns previously described only in adult soft tissue liposarcomas.

Temporal Evolution: How Structural Variants Drive Cancer Progression

Analyzing 13 pediatric cancer patients with multiple spatiotemporally distinct samples (diagnosis, relapse, metastases), the team discovered that structural variants continue forming throughout tumor evolution:

Osteosarcoma: Continuous Translocation Through Metastasis

Three osteosarcoma patients harbored somatic TP53 inactivation SVs and numerous translocations occurring continuously from primary tumor through metastasis. In one patient with six metastatic samples from both lungs, truncal translocations affected chromosomes 4, 11, 16, and 17, while later shared translocations involved chromosomes 3, 6, and 11.

Relapse-Specific Evolution in B-ALL

Six ETV6::RUNX1 B-ALL patients with matched diagnosis and relapse samples revealed that RAG-associated deletions (SV7) occurred in both truncal and relapse-specific variants at similar proportions—suggesting RAG-mediated recombination remains active throughout B-ALL evolution, fueling leukemia heterogeneity and treatment resistance.

Extrachromosomal DNA: The MDM2 Amplification Story

In rhabdoid tumor patient SJRHB012, clustered SVs created circular extrachromosomal DNA (ecDNA) amplicons with ~100 copies of the MDM2 oncogene at diagnosis. Relapse samples retained this ancestral amplicon but acquired a second, distinct amplicon—demonstrating how complex SVs evolve continuously during therapy.

Structural Variants vs. Point Mutations: A Paradigm Shift

Perhaps most striking was the comparison between structural variant prevalence and point mutation rates in driver genes. Across pediatric cancers:

• SV prevalence exceeded point mutations in nearly all driver genes
• Most genes showed 10-fold higher SV prevalence than point mutations
• Only TP53 and NF1 showed comparable rates of both variant types

This finding reinforces that structural variants—not point mutations—are the dominant force driving pediatric tumorigenesis, fundamentally reshaping how researchers approach childhood cancer genomics.

Genes Disrupted by Structural Variants: 109 Frequently Affected Loci

Across all pediatric cancers, 10,672 genes harbored at least one somatic SV breakpoint. Among the 109 most frequently disrupted genes (≥1% prevalence), researchers identified six categories:

1. Known driver genes (n = 47): ETV6, PAX5, RUNX1, KMT2A in leukemia; EWSR1::FLI1 fusion in Ewing sarcoma; TP53 in osteosarcoma
2. Fusion partners (n = 15): Genes partnered with drivers in translocation events
3. Immune genes (n = 8): T-cell receptor (TR) and immunoglobulin (IG) loci
4. Fragile sites (n = 23): Genomic regions prone to breakage
5. Neuroblastoma breakpoint family (NBPF) genes (n = 4)
6. Genes of unknown status (n = 12): Often the largest genes on their chromosomes

Notably, only a subset of SV-disrupted driver genes (ERG, SHANK2, CDKN2A, NF1, RB1, and RUNX1) were shared with adult cancers—reflecting fundamentally different genomic landscapes between pediatric and adult malignancies.

Intergenic Regulatory Regions: Enhancers Hijacked by Structural Variants

Beyond coding regions, structural variants frequently disrupted regulatory elements, juxtaposing active enhancers or promoters to oncogenes through translocation, deletion, or inversion:

• B-ALL: DUX4, CRLF2, and FLT3 most frequently affected
• Neuroblastoma: TERT promoter disruptions
• T-ALL: TAL1 and LMO2 regulatory region impacts
• Multiple cancers: MYC regulatory region SVs in AML, B-ALL, high-grade glioma, neuroblastoma, and T-ALL

These "enhancer hijacking" events drive aberrant oncogene overexpression without requiring coding mutations—a mechanism increasingly recognized in childhood cancer pathogenesis.

Clinical Implications: Toward Targeted Pediatric Cancer Therapies

The comprehensive dataset, now accessible through the St. Jude Cloud GenomePaint portal, provides unprecedented opportunities for researchers to identify therapeutic targets and investigate structural variation mechanisms.

Potential Therapeutic Applications

1. RAG inhibitors: Compounds blocking aberrant RAG-mediated recombination could prevent SV7-driven leukemias
2. HRD biomarkers: Homologous recombination deficiency (HRD) predicted in 2.4% of pediatric tumors, suggesting potential PARP inhibitor sensitivity
3. SV-based risk stratification: Complex structural variants and chromothripsis patterns may inform prognosis and treatment intensity
4. ecDNA-targeted therapies: Extrachromosomal DNA amplicons represent novel therapeutic vulnerabilities

"This comprehensive dataset will provide more opportunities to identify targets and investigate the mechanism behind the structural variations," Zhang stated. "The data is well curated, so there's a lot of datamining scientists can do to gain insight into the relevance of a target they are interested in studying, making it an incredible resource for the cancer community."

Methodology: Pan-Cancer Analysis Reveals Unique Pediatric Patterns

The research employed sophisticated computational analysis combining:

• Whole genome sequencing of 1,616 pediatric cancer patients (all under 18 years)
• Orthogonal experimental verification and manual review yielding 45,853 curated somatic SVs
• Bi-modal distance distribution analysis identifying the <100 bp peak unique to pediatric cancers
• Recombination Information Content (RIC) scoring evaluating RSS site functionality
• Temporal clustering of spatiotemporally distinct samples tracking evolutionary trajectories

Funding and Collaboration

The study was supported by grants from:

• National Institutes of Health, National Cancer Institute (R01CA216391, contract HHSN261200800001E)
• American Lebanese Syrian Associated Charities (ALSAC), St. Jude's fundraising organization

Lead authors Robert Greenhalgh, Samuel Brady, and Wentao Yang from St. Jude collaborated with Diane Flasch, Michael Edmonson, Nadezhda Terekhanova, Yanling Liu, Jian Wang, Karol Szlachta, Liqing Tian, Daniel Putnam, Delaram Rahbarinia, Pandurang Kolekar, Xin Zhou, Xiaotu Ma (all St. Jude), and Daniela Gerhard (National Cancer Institute).

The Broader Impact: Six Decades of Pediatric Cancer Progress

St. Jude Children's Research Hospital has helped push overall childhood cancer survival rates from 20% to more than 80% since opening over 60 years ago—making it the only NCI-designated Comprehensive Cancer Center devoted solely to children. This landmark structural variant analysis represents another critical milestone in understanding pediatric cancer biology and accelerating progress toward cures for all childhood catastrophic diseases.

The database is publicly accessible through genomepaint.stjude.cloud, enabling researchers worldwide to leverage this comprehensive resource for functional studies and clinical genomic testing design.

---

Disclaimer: This article is for informational purposes only and does not constitute medical advice. Always consult with qualified healthcare professionals for diagnosis and treatment of medical conditions.

References

1. Greenhalgh, R., Brady, S. W., Yang, W., Flasch, D. A., Edmonson, M. N., Terekhanova, N. V., Liu, Y., Wang, J., Szlachta, K. A., Tian, L., Putnam, D. K., Rahbarinia, D., Kolekar, P., Zhou, X., Gerhard, D. S., Ma, X., & Zhang, J. (2026). The landscape of structural variation in pediatric cancer. Cancer Cell, 45(3), S1535-6108. https://doi.org/10.1016/j.ccell.2026.02.012

2. Greenhalgh, R., Brady, S. W., Yang, W., et al. (2026). First-of-its-kind analysis reveals the structural variant landscape driving pediatric cancer development. St. Jude Children's Research Hospital Newsroom. https://www.stjude.org/media-resources/news-releases/2026-medicine-science-news/first-of-its-kind-analysis-reveals-the-structural-variant-landscape-driving-pediatric-cancer-development.html

3. St. Jude Children's Research Hospital. (2026, March 12). First-of-its-kind analysis reveals the structural variant landscape driving pediatric cancer development. EurekAlert!. https://www.eurekalert.org/news-releases/1119813

4. Greenhalgh, R., Brady, S. W., Yang, W., et al. (2026). The landscape of structural variation in pediatric cancer [PMC12991425]. PubMed Central. https://pmc.ncbi.nlm.nih.gov/articles/PMC12991425/

5. Brady, S. W., & Zhang, J. (2026). St. Jude Cloud GenomePaint portal. https://genomepaint.stjude.cloud/