Personalised prediction of anomalous bone regeneration in multiple myeloma patients

Abstract

Recent advances in immunotherapies have dramatically extended the lives of patients with multiple myeloma, resulting in a growing population of elderly patients suffering from damage to their vertebrae caused by this blood cancer. Clinicians typically intervene to fill in these lytic lesions in an attempt to reduce the risk of vertebral collapse and spinal cord injury, however observations of rapid bone regeneration post-cancer treatment suggest that surgery may not always be required. Here we develop a series of mechanistic computational models, optimised using cohorts of healthy volunteer and patient data, to show that vertebrae of recovering multiple myeloma patients rapidly strengthen in a manner not entirely explained by known bone remodelling mechanisms. Furthermore, using pathways informed by genomics data and patient-specific bone biomarkers, we demonstrate the potential for these models to provide personalised predictions of bone disease trajectory.