Identifying biomarkers of accelerated ageing in cancer patients from routine clinical data
Abstract
INTRODUCTION: Cancer and ageing have a bidirectional relationship: age is the strongest risk factor for cancer, and cancer and treatments can accelerate ageing. Therefore, biological age can differ from chronological age; biomarkers are needed to stratify interventions to minimise accelerated ageing. METHODS: PhenoAge was calculated from routine blood test results of patients attending a Geriatric Oncology clinic. PhenoAgeAccel was the residual from a regression of PhenoAge against age. RESULTS: Data were available for 173 patients (62% male). Mean PhenoAge was higher than age (84.3 (12.6) vs 76.2 (7.24), p < 0.001), though the two were correlated (r = 0.579, p < 0.001). Unlike age, PhenoAge and PhenoAgeAccel were associated with one-year mortality (PhenoAge OR=1.083, 95% CI: 1.038-1.136; PhenoAgeAccel OR=1.096, 95% CI: 1.047-1.155). PhenoAge correlated with Clinical Frailty Score and Timed Up and Go (CFS: Rs=0.31, p < 0.001; TUG: Rs=0.25, p < 0.005); there were no correlations with age. PhenoAgeAccel correlated with the number of CGA interventions made (Rs=0.17, p < 0.05), unlike age and PhenoAge. Patients with diabetes mellitus had a higher PhenoAgeAccel compared to those without (3.40 vs -1.71, p = 0.002). In patients receiving systemic anti-cancer treatment, patients with PhenoAgeAccel calculated pre-treatment had less age acceleration than those with PhenoAgeAccel calculated post-treatment, both overall (2.18 vs -2.87; p = 0.048) and in matched samples (n = 21, 7.76 vs -2.87, p < 0.001). CONCLUSIONS: PhenoAgeAccel is a greater predictor of risk than chronological age in older people with cancer. This makes it a promising biomarker to stratify patients for holistic geriatric assessment, dose reductions, or future geroprotective measures which could be integrated within electronic healthcare record systems.
Related
- Routine serum biomarkers and lightgbm modeling for the diagnosis of abomasal ulcers in culled dairy cows.
- An Interpretable Framework for Identifying EEG Spectral Biomarkers of Pathological Brain Aging.
- Copy number signatures in targeted gene panels associate with patient outcomes in routine clinical data
- Identifying Neurophysiological Subtypes in Clinical High Risk for Psychosis Using Functional Connectome Profiling
- Serum Cystatin 4 combined with routine clinical laboratory indicators: A support vector machine diagnostic model for early screening of gastrointestinal tumors
BibTeX
@article{bottomley2026identifying,
title = {{Identifying biomarkers of accelerated ageing in cancer patients from routine clinical data}},
author = {Clodagh Bottomley and Evelyne Liuu and Beatriz Echarte Archanco and Danielle Harari and Tania Kalsi and Carlos de Pablo Gallega and Carly Welch},
journal = {Mechanisms of Ageing and Development},
year = {2026},
month = oct,
doi = {10.1016/j.mad.2026.112234},
url = {https://hal.science/hal-05773712v1},
note = {CC BY},
}
RIS
TY - JOUR
TI - Identifying biomarkers of accelerated ageing in cancer patients from routine clinical data
AU - Clodagh Bottomley
AU - Evelyne Liuu
AU - Beatriz Echarte Archanco
AU - Danielle Harari
AU - Tania Kalsi
AU - Carlos de Pablo Gallega
AU - Carly Welch
PY - 2026
DA - 2026/10/01
JO - Mechanisms of Ageing and Development
DO - 10.1016/j.mad.2026.112234
UR - https://hal.science/hal-05773712v1
AB - INTRODUCTION: Cancer and ageing have a bidirectional relationship: age is the strongest risk factor for cancer, and cancer and treatments can accelerate ageing. Therefore, biological age can differ from chronological age; biomarkers are needed to stratify interventions to minimise accelerated ageing. METHODS: PhenoAge was calculated from routine blood test results of patients attending a Geriatric Oncology clinic. PhenoAgeAccel was the residual from a regression of PhenoAge against age. RESULTS: Data were available for 173 patients (62% male). Mean PhenoAge was higher than age (84.3 (12.6) vs 76.2 (7.24), p < 0.001), though the two were correlated (r = 0.579, p < 0.001). Unlike age, PhenoAge and PhenoAgeAccel were associated with one-year mortality (PhenoAge OR=1.083, 95% CI: 1.038-1.136; PhenoAgeAccel OR=1.096, 95% CI: 1.047-1.155). PhenoAge correlated with Clinical Frailty Score and Timed Up and Go (CFS: Rs=0.31, p < 0.001; TUG: Rs=0.25, p < 0.005); there were no correlations with age. PhenoAgeAccel correlated with the number of CGA interventions made (Rs=0.17, p < 0.05), unlike age and PhenoAge. Patients with diabetes mellitus had a higher PhenoAgeAccel compared to those without (3.40 vs -1.71, p = 0.002). In patients receiving systemic anti-cancer treatment, patients with PhenoAgeAccel calculated pre-treatment had less age acceleration than those with PhenoAgeAccel calculated post-treatment, both overall (2.18 vs -2.87; p = 0.048) and in matched samples (n = 21, 7.76 vs -2.87, p < 0.001). CONCLUSIONS: PhenoAgeAccel is a greater predictor of risk than chronological age in older people with cancer. This makes it a promising biomarker to stratify patients for holistic geriatric assessment, dose reductions, or future geroprotective measures which could be integrated within electronic healthcare record systems.
SN - 0047-6374
SN - 1872-6216
ER -