Grant Application

Howard Gutstein, MD, Gwen Sowa, MD, PhD, and Ajay Wasan, MD, MSc, University of Pittsburgh School of Medicine

Proposed Innovation

Chronic pain afflicts over 100 million Americans with an estimated economic impact of $500 billion annually. The most prevalent pain condition in the country is low back pain, affecting more than 50 million people nationwide.

Despite decades of research, treatment outcomes remain suboptimal and significant side effects to treatments exist. Opioids — for centuries the gold standard for treatment of severe chronic pain — are highly addictive and have led to an epidemic of abuse and overdose-related deaths. In addition, overuse of various invasive treatments has led to increased costs of care without associated improvements in patients’ pain levels.

This project aims to analyze data stored on the UPMC data warehouse of outpatients and inpatients with low back pain to develop a predictive model of response to pain treatment. Ultimately, this data can help doctors determine the most appropriate pain treatment for each individual patient, as well as identifying those patients most susceptible to addiction.

Improvements in Action

Previous research has shown an association between certain serum protein biomakers and genetic polymorphisms among patients receiving spinal injections for low back pain. But the data collected did not explain all the variability of responses that were observed.

Additional research at UPMC evaluating clinical characteristics of patients has identified 10 unique, clinically homogeneous patient subgroups with different responses to spinal injection. Using mathematical modeling, response to the injection can now be predicted with a high level of accuracy for about 22 percent of all patients.

Intended Outcomes

Combining individualized biology with clinical phenotypes will provide even greater predictive power to support treatment decisions for chronic lower back pain. The algorithms developed using this innovative approach will enable identification of patients most likely to respond to particular pain treatments and those who are more susceptible to opioid dependence and addiction.