Melanoma Driver Selection Varies With Patient Age

Key Takeaways
- Among melanoma tumors in TCGA and AACR Project GENIE, inferred driver selection showed stronger age-associated changes than neutral mutation rates.
- BRAF V600E prevalence and inferred selection declined with age, while evidence for increasing selection was stronger for NRAS Q61L than for other NRAS Q61 variants.
- TERT promoter mutations became modestly more prevalent with age even as their inferred selection declined.
- UV-associated signatures predominated at every age, while the clock-like SBS1 and SBS5 signatures differed across age groups.
In Liu and colleagues’ Science Advances melanoma analysis, investigators retrospectively examined 5,472 tumor samples: 459 whole-exome Skin Cutaneous Melanoma (SKCM) samples from The Cancer Genome Atlas (TCGA) and 5,013 targeted-panel tumors from American Association for Cancer Research (AACR) Project GENIE v13.1. They excluded samples without recorded age and retained one GENIE sample per patient, favoring a primary specimen when available. The report generally describes age groups as younger than 40, 40–59, and 60 years or older, although one methods passage lists the middle group as 40–50; continuous-age fits used age at tumor resection. Age-specific synonymous variants in the exome cohort supplied sequence-context-adjusted neutral mutation-rate estimates. Panel data used age-matched exome-derived baselines. Selection strength, γ, represented observed variant-fixation flux relative to the estimated neutral mutation rate—not a clinical hazard ratio or treatment-benefit measure. Investigators compared age groups and fitted linear, logistic, and sigmoid age trajectories, selecting among them with Akaike Information Criterion.
Median gene-level neutral mutation rates in tumors from patients aged 60 or older were 20% higher than at ages 40–59 and 30% higher than below age 40. Ultraviolet (UV)-associated mutational signatures predominated across all age groups; modest differences in UV attribution were not statistically significant. The clock-like single-base substitution (SBS) signatures SBS1 and SBS5 differed significantly across age groups, although SBS1 did not rise in every successive group. These patterns describe the background supply of mutations rather than selection on individual variants.
BRAF V600E prevalence and inferred selection declined with age, with age-stratified selection P < 0.001. In contrast, NRAS Q61L showed increasing inferred selection in a continuous-age logistic fit (k = 0.03; probability of an opposite direction ≲0.025). Evidence for NRAS Q61H was less clear, and Q61K was particularly uncertain. TERT promoter C250T and C228T became modestly more frequent despite declining inferred selection. Compound selection estimates for TP53, PTEN, and CDKN2A also declined. SBS- and double-base substitution (DBS)-derived BRAF V600E shared a declining trajectory, whereas DBS-derived BRAF V600K increased.
Patient age correlates with other biological and exposure-related characteristics, so these associations do not isolate age as a cause or establish treatment response. The cohorts contained mostly metastatic tumors; metastatic-only reanalysis preserved the principal trends, but extension to primary melanoma remains uncertain. Loss-of-function–restricted analyses preserved the tumor-suppressor pattern. Exome and targeted-panel sequencing largely confined assessment to coding regions and selected noncoding sites. Exploratory variant-screening P values were nominal and unadjusted for multiple testing, and continuous fits considered only prespecified monotonic forms. The findings do not validate age-based testing or treatment decisions.
According to the researchers, separating neutral mutation supply from inferred selection revealed mutation-specific age associations in melanoma evolution. Potential applications to biomarkers or treatment remain hypotheses rather than demonstrated clinical effects.
Clinician Questions
How were mutational signatures assigned to AACR Project GENIE melanoma tumors?
In the melanoma genomic analysis, investigators directly attributed signatures in whole-exome tumor samples but assigned AACR Project GENIE targeted-panel samples the whole-exome cohort’s average signature proportions. Panel assignments were not directly measured, tumor-specific signature profiles and do not establish individual exposure histories.
Does the age-related BRAF V600E–NRAS Q61 pattern show that a melanoma switches drivers as a patient gets older?
No. Melanoma tumors were compared across patients of different ages, recorded at resection; investigators did not track driver states within individual tumors over time. BRAF V600E and NRAS Q61 variants were nearly mutually exclusive in these samples.
This content is intended for healthcare professionals for educational purposes. The article and its image were developed with AI assistance and reviewed by members of the ReachMD team prior to publication. Please refer to the original source linked for the complete and most up-to-date information.