Genomic and molecular features distinguish young adult cancer from later-onset cancer

Abstract

Young adult cancer has increased in incidence worldwide, but its molecular etiologies remain unclear. We systematically characterize genomic profiles of young adult tumors with ages of onset ≤50 years and compare them to later-onset tumors using over 6,000 cases across 14 cancer types. While young adult tumors generally show lower mutation burdens and comparable copy-number variation rates compared to later-onset cases, they are enriched for multiple driver mutations and copy-number alterations in subtype-specific contexts. Characterization of tumor immune microenvironments reveals pan-cancer patterns of elevated TGF-β response/dendritic cells and lower IFN-γ response/macrophages relative to later-onset tumors, corresponding to age-related responses to immunotherapy in several cancer types. Finally, we identify prevalent clinically actionable events that disproportionally affect young adult or later-onset cases. The resulting catalog of age-related molecular drivers can guide precision diagnostics and treatments for young adult cancer.

Keywords: DNA methylation; clinical actionability; copy-number variation; early-onset cancer; gene expression; gene fusion; metabolic dysregulation; somatic mutation; tumor immune microenvironment; young adult cancer.

Figure 1.. Characterizing 5,365 adult cancers of the TCGA PanCanAtlas cohort

(A) Attributes of the 5,365 cases (1,757 young adult and 3,608 later-onset) across 14 cancer types, including TCGA abbreviation of the cancer type, gender ratio, mean age at diagnosis, and young adult ratio. (B) Subtype percentages between young adult and later-onset cases in non-female-specific cancer types. Log10-transformed nonsilent mutation rates for unique individuals in each cancer type are colored by subtype. (C) Subtype percentages between young adult and later-onset cases in female-specific cancer types. Log10-transformed nonsilent mutation rates for unique individuals in each cancer type are colored by subtype. (D) Log10-transformed nonsilent mutation rates for unique individuals in cancer types without subtype information are distinguished by PanCanAtlas colors. See also Figure S1.

Figure 2.. Somatic mutations between young adult and later-onset tumors

(A) Somatic mutations associated with young adult versus later-onset cancer cohorts. For each gene-level mutation, a coefficient >0 represents a higher rate in young adult cases, while a coefficient <0 represents a higher rate in later-onset cases. Significant mutations (FDR <0.05) are labeled. (B) Percentages of young adult versus later-onset BRCA cases presenting mutation-subtype pairs. (C) Percentages of young adult versus later-onset LGG cases presenting mutation-subtype pairs. (D) Percentages of young adult versus later-onset UCEC cases presenting mutation-subtype pairs. See also Figure S2 and Table S1.

Figure 3.. Copy-number variations between young adult and later-onset tumors

(A) Log10-transformed summed copy-number segments per sample across the 14 cancer types in young adult versus later onset tumors; black bars designate median values. *FDR ≥0.10 and <0.15, **FDR ≥0.05 and <0.10, and ***FDR <0.05. (B) CNVs associated with young adult versus later-onset cancer cohorts. For each deletion or amplification, a log10-transformed coefficient >0 represents a higher rate young adult cases, while a log10-transformed coefficients <0 represents a higher rate in later-onset cases. Significant (FDR <0.05) and suggestive (FDR <0.15) CNVs are labeled. (C) Percentages of young adult versus later-onset BRCA cases presenting CNV-subtype pairs. (D) Percentages of young adult versus later-onset LGG cases presenting CNV-subtype pairs. (E) Percentages of young adult versus later-onset UCEC cases presenting CNV-subtype pairs. See also Figure S3 and Table S2.

Figure 4.. Methylations/fusions between young adult and later-onset tumors

(A) Methylations associated with young adult versus later-onset cancer cohorts. For each gene-level methylation, a log10-transformed coefficient >0 represents a higher rate in young adult cases, while a log10-transformed coefficient <0 represents a higher rate in later-onset cases. Significant (FDR <0.05) and suggestive (FDR <0.15) methylations are labeled. (B) Percentages of young adult versus later-onset BRCA cases presenting methylation-subtype pairs. (C) Percentages of young adult versus later-onset LGG cases presenting methylation-subtype pairs. (D) Fusions associated with young adult versus later-onset cancer cohorts. For each fusion event, a log10-transformed coefficient >0 represents a higher rate in young adult cases, while a log10-transformed coefficient <0 represents a higher rate in later-onset cases. Significant (FDR <0.05) and suggestive (FDR <0.15) fusions are labeled. See also Figure S4 and Table S3.

Figure 5.. Differentially expressed genes and pathways between young adult and later-onset tumors

(A) Differentially expressed genes between young adult and later-onset cancer cohorts. Positive and negative logFC values represent higher levels in young adult and later-onset cases, respectively. Color represents a gene differentially expressed in a specific cancer type; gray represents a gene not differentially expressed in any cancer. Differentially expressed genes with BH-corrected p values less than that of the top third gene are labeled. (B) Overview of pathway perturbed by differential expression in different cancers. Positive and negative normalized enrichment scores represent higher levels in young adult and later-onset cases, respectively. (C) The normalized enrichment scores of pathway-cancer associations. Red and blue indicate higher levels in young adult and later-onset cases, respectively. Significant associations are pinpointed by gray boxes. Cancers are ordered by the number of significantly perturbed pathways involved, and pathways are split based on the number of involved cancers. (D) Enrichment of two immune-related pathways (cytokine-cytokine receptor interaction and graft versus host disease) with consistently lower expressions across three cancer types (HNSC, LGG, and SARC). See also Figure S5.

Figure 6.. Tumor immune microenvironment differences between young adult and later-onset cases

(A) Differences in neoantigen loads between young adult and later-onset cases. Red and blue indicate higher levels in young adult and later-onset cases, respectively. *FDR ≥0.10 and <0.15, **FDR ≥0.05 and <0.10, and ***FDR <0.05. (B) Differences in immune gene signatures and infiltrates between young adult and later-onset cases. Red and blue indicate higher levels in young adult and later-onset cases, respectively. *FDR ≥0.10 and <0.15, **FDR ≥0.05 and <0.10, and *** FDR <0.05. (C) Differences in Th cell levels between young adult and later-onset cases. Red and blue indicate higher levels in young adult versus later-onset cases, respectively. *FDR ≥0.10 and <0.15, **FDR ≥0.05 and <0.10, and ***FDR <0.05. (D) Positive logFC (red) indicates higher expression in young adult cases; negative logFC (blue) indicates down-expression in young adult cases. Genes with p < 0.05 are pinpointed by black boxes. See also Figure S6.

Figure 7.. Clinical actionability in young adult versus later-onset tumors

(A) Percentages of unique young adult and later-onset cases with somatic variants druggable at the A or B evidence levels, further subdivided by on versus off-label status. (B) Percentages of unique young adult and later-onset cases expressing each of the top five (ranked by frequency of appearance in the seven selected cancers) clinically druggable somatic variants. Young adult percentages are red; later-onset are black. After correcting for confounding factors, a single significant association (FDR <0.05) was found between BRAF V600E mutations and young adult SKCM. (C) Percentages of unique young adult and later-onset cases with copy-number amplifications druggable at the A or B evidence levels, further subdivided by on versus off-label status. (D) Percentages of unique young adult and later-onset cases expressing each of the top five (ranked by frequency of appearance in the six selected cancers) clinically druggable copy-number amplifications. Young adult percentages are red; later-onset are black. See also Figure S7.