#5454 DESIGN AND OPTIMIZATION OF DRUG COMBINATION THERAPIES FOR OSTEOPOROSIS VIA SIMULATIONS OF BONE TURNOVER

Abstract

Abstract Background and Aims Many chronic kidney disease (CKD) patients suffer from mineral and bone disorders like osteoporosis, which also affects a huge part of the general population, predominantly postmenopausal women. For the treatment of osteoporosis, several drug types with different mechanisms of action are available, such as bisphosphonates, PTH analogs, RANKL inhibitors and sclerostin inhibitors (some of which are contraindicated for CKD patients beyond CKD stage 3). Different drugs can be given individually or combined in various ways, simultaneously or sequentially, resulting in a huge number of theoretically possible drug combinations that are practically impossible to study in clinical studies. We aimed to study using simulations whether there are combination therapies that would work considerably better than those tested and employed in clinical practice. If so, this could be beneficial for the treatment of a large number of osteoporosis patients by leading to a higher increase in bone strength and a higher reduction in bone fracture risk than standard therapies. Method We developed a physiology-based mathematical model that simulates the process of bone renewal in the human body and can predict how different osteoporosis drugs (including bisphosphonates, PTH analogs, RANKL inhibitors, and sclerostin inhibitors) affect this process, alone and in combination [1]. Results The model was validated on over 30 clinical osteoporosis studies with about 90 study arms. It predicted the time courses of bone mineral density and bone turnover markers with a high degree of accuracy based on the knowledge of the used drugs and dosing schemes alone. We used the model to study how treatment results changed when merely reshuffling the order in which different drugs were given. For example, for 3-year drug therapies involving alendronate, denosumab and romosozumab (1 year each), our simulations showed that the order in which these drugs were given can have a considerable effect on short- and long-term therapy success [1]. This is due to different drugs favorably interacting when applied in the right sequence. Conclusion Our findings indicate that some osteoporosis drug administration schemes are superior to others while relying on the exact same types and amounts of drugs. Therefore, there is a large potential to improve pharmacologic therapies of osteoporosis using physiological simulations. For renal patients beyond CKD stage 3, a restricted set of drug types may be considered to account for contraindications related to impaired renal clearance. If translated into clinical practice, findings obtained using our model could give rise to novel treatments using combinations of existing drugs for osteoporosis that lead to a lower bone fracture risk than standard treatments. Moreover, our model could serve as a basis for personalizing osteoporosis therapy to individual patients.