Brain-based and clinical phenotyping of pain pharmacotherapy in knee OA

ABSTRACT This proposal is in response to the funding opportunity Research Project Grant (Parent R01 Clinical Trial Required), FOA number PAR-20-183. This proposal aims to identify specific biomarkers in individual people with osteoarthritis (OA) pain that will allow definition of responder phenotypes distinct for different therapeutic interventions. A mechanistic, prospective randomized trial will be undertaken, treating people having moderate to severe OA pain with either naproxen (a non-steroidal anti-inflammatory agent (NSAID)), duloxetine (a selective serotonin-norepinephrine reuptake inhibitor) or placebo in a 1:1:1 ratio. Randomization will be stratified by sex and prior opioid use. Naproxen, duloxetine and placebo are known to have different mechanisms of action and work at different sites in the pain pathway. Hence, it would be expected that these distinctions would be reflected in differences in individuals who would respond to each of these interventions. To this end, during the study we will collect a wide variety of biomarkers including demographics (sex, age), clinical outcome measures, questionnaires of patient-reported outcomes, neurosensory status (quantitative sensory testing indices), serum- based biomarkers, and joint and brain imaging. The treatment for each participant and collection of biomarkers will occur during an initial 6-week period and then be repeated after a 4-week washout to account for the known within-patient variability. The results obtained will permit the identification of responders (defined by 30% or greater improvement in pain from baseline with other thresholds also evaluated), and the correlation of biomarker status at baseline to response. We will then build a model to define the responder phenotype for each intervention, first using only clinical data and secondly using both clinical and MRI-based brain data. The latter will permit a further understanding of the mechanisms involved in modulation of the pain pathways by each of the agents. Particular attention will be given to treatment by sex interactions. The characterization of responder phenotypes to NSAID, duloxetine and placebo will allow for the practice of personalized medicine, providing the right drug to the right patient, enhancing therapeutic success, and reducing the risks involved with being treated with ineffective drugs having serious potential side effects. In addition, this approach will allow for more targeted and more efficient development of potential new therapeutic agents to treat OA pain.