In silico prediction of novel effective combinational treatment of chronic pain in individual patients: A joint white paper of the H2020 QSPainRelief consortium

Opioids are prescribed widely for chronic pain despite well-recognised risks and variable long-term benefit, reflecting the lack of effective alternatives for many patients. Combination therapies offer a promising strategy to enhance efficacy whilst reducing side effects. However, identifying effective drug combinations is notoriously difficult. The H2020 QSPainRelief project developed a model platform for in-silico prediction of efficacy and adverse effects of analgesic drug combinations, focussing on opioid-non-opioid combination strategies. It integrates physiologically-based pharmacokinetic, pharmacodynamic and neural circuit models that capture key aspects of nociceptive processing and central nervous system (CNS) side effects, and enables more advanced personalised pain management by incorporating patient-specific variables, patient-reported outcomes and patient preferences. After discussing the problem of chronic pain treatments and critical determinants of CNS drug effects, we introduce the QSPainRelief platform development and share illustrative results on prediction of morphine-non-opioid combinations effects, and inclusion of patient preferences in dealing with the side effects using a clinical utility index model. Finally, we discuss remaining gaps in data, and directions of future research to strengthen the validation and predictive performance of the platform to further support its application for the development of safer and more effective combination therapies for chronic pain. We conclude that the QSPainRelief model platform can reduce reliance on costly and slow trial-and-error methods in clinical drug development for chronic pain by bridging mechanistic insights and clinical needs, and representing a key enabler for more effective, faster, safer and personalised chronic pain management.