Multivariate Machine Learning to Characterize Opioid-induced Alterations in the Brain in Chronic Pain

点击翻译按钮获取中文摘要

PROJECT SUMMARY/ABSTRACT Prescription opioids are a potent class of drugs for treating pain. However, growing body of research has described iatrogenic consequences of long-term (> 90 days) opioid use in patients with chronic pain including hyperalgesia and impaired executive function. Dopamine is a critical modulator of executive function. While changes in pain and behavior have been noted, little is known about the brain’s morphology, neural and dopaminergic activity that change over time with long-term prescription opioid use. Consistent with the NIDA Strategic Plan objective 1.3, this K25 proposal seeks to “establish the effects of drug use, addiction, and recovery on brain circuits, behavior, and health” using neuroimaging-informed tools. Specifically, the present study combines multiple levels of investigation, including structural and functional Magnetic Resonance Imaging (MRI), Positron Emission Tomography (PET), Quantitative Sensory Testing (QST) and neuropsychological assessments of executive function, and employ machine learning techniques for analysis to identify the effects of long-term prescription opioid use on the brain in chronic pain patients. The applicant will use her advanced quantitative skills in neuroimaging data analysis and modeling to training in QST, and experience in cognitive neuropsychology, epidemiology of chronic pain and addiction to develop an independent research plan in translational pain and successfully compete for future R01 funding. To achieve the training needed to facilitate this investigation, the applicant has consulted with an expert in chronic pain research and opioid therapy, a substance abuse specialist, a neuropsychologist, an epidemiologist, an imaging scientist, and a machine learning leader in neuroimaging field to develop an innovative study and training plan. 40 patients with a diagnostically homogeneous chronic pain condition (i.e., chronic low back pain; CLBP) on long-term opioid therapy, as compared to 40 opioid-naïve CLBP patients, will be studied to achieve the following Aims: 1) Measure pain, cognitive performance, neural and dopaminergic activity during concurrent pain and executive function task fMRI-PET to characterize the effects of opioids on pain processing and executive function in CLBP; 2) measure intrinsic brain activity during resting state fMRI-PET to identify intrinsic brain alterations associated with long-term opioid use in CLBP; and 3) apply high-resolution structural MRI to measure opioid-induced morphological changes in CLBP. This research is innovative in its use of combined QST and neuro- psychological measures with multimodal imaging and sophisticated statistical approaches. It is significant because of its comprehensive approach towards addressing the NIDA Strategic Plan objective. Findings stand to inform medical decision-making regarding pain care and opioid prescription, as well as risk mitigation strategies. The research, training and results obtained will provide a platform for applicant’s long-term scientific research goal of becoming an independent R01-funded, faculty-level principal investigator performing translational pain research aimed at developing neuroimaging tools to have clinical application.