Development and validation of a simplified biological clock (GeroAge) to estimate accelerated aging in patients with cancer and its influence on clinical outcomes: a cohort study

Background: Beyond chronological age (ie, how long a person has lived), it is hypothesized that biological age (BA, ie, the extent of cellular aging) may help further refine clinical prognosis of patients with cancer. PhenoAge is a published BA derived from nine blood markers. One challenge to apply PhenoAge is that existing databases, including the Chinese Survey of Investigation on Nutrition Status and Clinical Outcome of Common Cancers (INSCOC), did not have all the required nine biomarkers. Methods: We sought to develop a new BA called GeroAge, which only requires six blood markers (albumin, natural logarithm of C-reactive protein [lnCRP], lymphocyte percentage, creatinine kinase, blood glucose, white blood cell) plus chronological age. We compared the agreement between GeroAge and PhenoAge in the National Health and Nutrition Examination Survey (NHANES) databases which have all the nine blood markers. We also compared GeroAge’s predictive ability of overall survival (OS) compared to PhenoAge in the NHANES. Cox proportional risk models were used to assess the relationship between GeroAge (or PhenoAge) and OS. We further estimated GeroAge in the INSCOC and assessed its influence on clinical outcomes of patients with cancer using the logistic regression, including prolonged hospitalization, malnutrition, risk of intensive care unit (ICU) admission, and high medical costs. Results: Using the NHANES databases, we found a significant linear relationship between GeroAge and PhenoAge ( R = 0.99, P < 0.001); and a similar ability of GeroAge and PhenoAge in predicting OS. In INSCOC database, we found a significant positive association between GeroAge acceleration (GeroAge-chronological age) and adverse clinical outcomes. For every 5-year GeroAge acceleration, the risks of long-term hospitalization, malnutrition, ICU admission, and high healthcare cost increased by 17% to 35%. Moreover, patients with GeroAge acceleration showed a 60% to 232% higher risk of experiencing adverse clinical outcomes than patients without GeroAge acceleration. Conclusions: These findings suggest that BA is a relevant feature of patients with cancer and should be included in the baseline profiling and continuous follow up. When estimation of BA is not feasible using more established biological clock, GeroAge appears a reliable tool to predict the clinical risk of patients with cancer.