Immune Repertoire Aging Accelerates After 50, Shaped by Sex, CMV and Cancer

The collection of T cell receptors circulating in human blood does not wear down at a steady pace. According to a new analysis published in Aging Cell, the diversity of the T cell receptor (TCR) repertoire declines slowly through early adulthood and then accelerates markedly after around age 50, with sex and latent cytomegalovirus (CMV) infection each leaving additional, measurable imprints on the repertoire. In patients with cancer, the repertoire appears contracted beyond what age alone would predict, and the degree of diversity carries prognostic information across several tumor types.

T cells recognize pathogens and cancerous cells through the TCR, a surface receptor whose antigen-binding specificity is generated by somatic V(D)J recombination. Each T cell clone carries a unique receptor sequence, and an effective immune response depends on a repertoire diverse enough to recognize a wide range of antigens. When a naïve T cell meets its cognate antigen, it expands rapidly into many daughter cells with identical receptors. With age, thymic involution progressively limits the production of new naïve T cells, and immune maintenance shifts toward peripheral expansion of existing clones, narrowing the repertoire’s ability to respond to previously unencountered pathogens and tumor neoantigens.

To map this landscape systematically, the researchers assembled TCR sequencing data targeting the CDR3 region of the TCR β-chain from 14 studies, covering 1.32 billion unique T cell clones from 798 patients with cancer and 5052 individuals without cancer, the latter including healthy donors as well as people with COVID-19 or pre- or at-onset type 1 diabetes. Ten of the cohorts were generated with the same commercial sequencing assay, which reduced platform-related variability; the remaining four were generally excluded from joint modeling to limit assay-specific bias. All repertoires were downsampled to 5000 TCR-β sequences per individual because sequencing depth influenced diversity estimates.

The team quantified diversity using the normalized Shannon index, which captures both the richness of clones and the evenness of their frequencies. Median values ranged from 0.85 to 0.99 across cohorts, yet inter-individual variability was substantial. In 117 patients with paired samples collected seven months apart, variation between individuals significantly exceeded variation within the same person over time (p < 0.001), indicating that TCR diversity is relatively stable on short timescales and that the observed differences are unlikely to stem from temporal fluctuation alone. A cohort of young people spanning a narrow age range showed particularly low variance, consistent with a repertoire shaped by lifelong exposures, health status and age-associated remodeling.

Pooling the platform-matched cohorts of individuals without cancer, the analysis revealed a significant age-associated decline in diversity, and the downward trend steepened in older age groups. A segmented linear regression model with a fixed breakpoint at age 50 estimated a 2.3-fold faster reduction in TCR diversity after that age (p < 0.0001), a result that held when diversity was instead measured with the Gini and Simpson indices. The authors characterize this as a nonlinear trajectory: systemic loss of TCR diversity is not constant throughout life but accelerates around midlife.

Sex left a distinct signature. Males showed lower diversity than females both below age 50 (p = 0.0004) and above it (p < 0.0001), and the gap widened after midlife. On average, males exhibited diversity comparable to that of females roughly ten years older. When samples were ranked by diversity and split into ten equal bins, the lowest-diversity deciles were enriched for males (χ² = 22.1, df = 9, p = 0.0085). The authors note this aligns with prior reports of sexual dimorphism in immune remodeling, including faster thymic involution in males, and underscore the need to treat sex as a biological variable in immune aging studies.

CMV infection emerged as another major determinant. Using a published classifier that detects CMV-associated TCR-β sequences and predicts status from their abundance—a model that achieved an area under the curve of 0.98 on its training set and 0.94 on its validation set—the researchers found that the predicted fraction of CMV-positive individuals rose from under 30% in childhood to more than 65% in older adults, mirroring known age-dependent prevalence. Across age and sex, CMV-positive individuals consistently showed lower repertoire diversity. Strikingly, a CMV-positive person aged 51–60 displayed diversity comparable to a CMV-negative person aged 71–80, suggesting that latent CMV infection may approximate the effect of two decades of aging on the repertoire, with the effect itself intensifying after midlife.

Clonal composition told a parallel story. Grouping clones by their share of the repertoire—non-expanded singletons, low-expanded, intermediate-expanded and hyper-expanded clones—the proportion of non-expanded clones fell from 94.7% to 65.0% across age groups, while hyper-expanded clones increased, with a distinct change in slope between the 41–50 and 51–60 age groups. CMV-positive individuals had more hyper-expanded and fewer non-expanded clones at every age, with the absolute difference growing over the lifespan. Single-cell RNA sequencing data from 981 healthy individuals in the OneK1K cohort helped disentangle the mechanisms: aging was associated with a robust loss of naïve CD8+ T cells independent of CMV status, whereas CMV positivity was marked by an enrichment of CD8+ effector memory cells re-expressing CD45RA (TEMRA cells) and reduced relative proportions of naïve CD4+ T cells. In other words, aging depletes the naïve compartment while CMV selectively inflates cytotoxic effector lineages.

Cancer compounded these patterns. Compared with individuals without cancer of similar age (41–90), patients with cancer showed generally lower TCR diversity, and a linear model adjusted for age, sex and CMV status confirmed that cancer status was significantly associated with reduced diversity among those over 50 (estimate = −0.0225, 95% CI: −0.032 to −0.013, p < 0.0001)—an average effect corresponding to roughly ten years of repertoire aging. Notably, the effect of CMV on diversity was amplified in patients with cancer (interaction p = 0.0001), while among CMV-negative individuals no association between cancer and diversity was observed, suggesting the cancer-associated reduction is driven largely by an amplification of the CMV effect. Cancer was also associated with a shift in clone size distributions after adjustment: a reduced proportion of non-expanded clones (estimate = −7.44, 95% CI: −8.14 to −6.74) and an increased proportion of hyper-expanded clones (estimate = 5.32, 95% CI: 4.75–5.88), both p < 0.0001, with the differences most pronounced in CMV-positive patients.

The prognostic implications were examined in Cox proportional hazards analyses across seven cancer cohorts spanning pancreatic, lung, bladder, breast, melanoma and urothelial cancers. Higher TCR diversity was generally associated with better overall survival, with a hazard ratio of 0.86 per 0.1-unit increase in normalized Shannon diversity (95% CI: 0.77–0.96, p = 0.0078); all cohort-specific hazard ratios fell below 1, with the strongest association in the Snyder cohort (HR = 0.52, 95% CI: 0.29–0.92). CMV status itself was not significantly associated with survival in these cohorts, treated primarily with chemotherapy. When the researchers corrected each patient’s diversity for age, sex and CMV status using a model trained on individuals without cancer, higher-than-expected diversity—potentially reflecting a biologically younger immune profile—remained associated with improved survival. In an independent analysis of The Cancer Genome Atlas, where TCR diversity was estimated from whole-exome sequencing data with a limited 3–57 TCR reads per sample, patients in the bottom diversity quartile had worse overall survival (HR = 1.21, 95% CI: 1.06–1.39, p = 0.0051), as did those with low T cell fractions estimated by T cell ExTRECT (HR = 1.18, 95% CI: 1.05–1.33). Patients with both low diversity and low T cell abundance fared worst, with a meta-analytic hazard ratio of 1.54 (95% CI: 1.22–1.94, p = 0.0003), though effect sizes varied by cancer type. The authors caution that heterogeneity in treatment history, sampling timing and disease stage across the cancer cohorts may confound these survival associations, and that the prognostic relevance of peripheral immune measures appears cancer-type specific. They position TCR diversity, measurable non-invasively from blood, as a candidate biomarker of immune aging and a potential complement to existing prognostic indicators—while emphasizing that chronic viral exposures such as CMV must be accounted for in any study of immunosenescence.

Subject of Research: Nonlinear age-related decline of human T cell receptor repertoire diversity and its modification by sex, CMV infection and cancer

Article Title: Nonlinear Aging of the Human TCR Repertoire Is Influenced by Sex, CMV, and Cancer

Article References: Lange, N., Attermann, A. K., Juul, R. I., Andersen, L., Nordentoft, I., Jessen, L. E., McGranahan, N., Aerts, H. J. W. L., Dyrskjøt, L., & Birkbak, N. J. (2026). Nonlinear Aging of the Human TCR Repertoire Is Influenced by Sex, CMV, and Cancer. Aging Cell, 25(11), Article e70777. https://doi.org/10.1111/acel.70777

Image Credits: AI Generated

DOI: 10.1111/acel.70777

Keywords: TCR repertoire, immunosenescence, T cell diversity, cytomegalovirus, thymic involution, cancer prognosis, TCR sequencing, clonal expansion, OneK1K cohort, TCGA, overall survival, sexual dimorphism