Developing and validating predictive models for clinical outcomes in mild autonomous cortisol secretion 

This research project was awarded the 2024 Birmingham BRC Collaboration Fund

Status: Ongoing

Small growths in the adrenal glands (which sit above the kidneys) are increasingly found by chance during scans done for other reasons. Many of these growths produce slightly too much of the stress hormone cortisol, a condition known as mild autonomous cortisol secretion (MACS). Although MACS does not usually cause obvious hormone symptoms, it is linked to a higher risk of high blood pressure, diabetes, heart disease, and early death. At present, doctors cannot reliably predict which patients with MACS will go on to develop these serious health problems. 

Project aims

This project uses data from the DEX-AI, NAPACA-OUTCOME, and EURINE-ACT studies, the largest long-term studies of adrenal tumours worldwide, which have followed over 6,000 patients for more than seven years. We will use advanced statistical methods to develop and test tools that can predict an individual patient’s future risk of heart and metabolic disease using information collected at diagnosis. We will also examine whether repeating tests over time improves prediction and helps identify the best timing for follow-up. 

The main output will be a validated clinical prediction model that can be used in routine NHS practice to better target treatment and monitoring to those at highest risk. This will improve personalised care, avoid unnecessary follow-up for low-risk patients, and inform future research into biological markers of disease risk.