Project Summary
Osteoarthritis (OA) affects more than 50 million people in the US, eventually leading to chronic pain and
disability. Currently, no FDA-approved biomarkers for OA exist, limiting a clinician’s ability to detect early-
stage disease. Early detection of OA is a rate-limiting step toward improved clinical care because recent
studies show that early intervention can limit or reverse OA symptoms. However, early detection of OA is not
currently possible.
Metabolomic profiling is an innovative approach to characterize biological systems like synovial joints.
While OA is classically described by degeneration of the articular cartilage, the pathophysiology of the disease
involves cell stress, inflammatory activity, and abnormal tissue metabolism. Since the synovial fluid contains
many of the molecules produced by the articular joint’s multiple cell types (eg chondrocytes, synoviocytes,
and osteoblasts), metabolic profiling of the synovial fluid could provide a unique window into active disease
processes in the OA-affected joint, and metabolomic profiles from synovial fluid could facilitate early detection
of OA.
While synovial fluid is “the scene of the crime” for OA, plasma is easier to obtain clinically. Because
metabolites are smaller than 1000 Daltons, plasma metabolite profiles may also reflect OA pathophysiology,
and the goal of this proposal is to advance global metabolite profiles as clinical biomarkers of OA grade. Our
vision is to identify a panel of metabolite biomarkers that aid in the detection OA before the onset of symptoms.
In Aim 1, metabolomic profiles of synovial fluid will be used with statistical learning to develop metabolite
biomarkers that predict both radiographic (e.g. KL-score) and symptomatic (e.g. pain) OA. The largest
available clinical cohort of human synovial fluid and plasma (n=1850, University of Oxford) will be used for
metabolite biomarker development, with model training, testing, and validation performed on a random
independent subsets. Because there is substantial heterogeneity in OA, molecular endotypes (i.e. molecular
OA profiles) will be developed from these metabolomic profiles that may yield important information on OA
pathophysiology. The cohort also contains paired plasma, and studies of Aim 2 will assess the correlations of
each metabolite between synovial fluid and plasma. Because plasma is easier to obtain, these correlations
will define which joint metabolites can be assessed in the circulatory compartment.
The expected outcomes of this project are (1) a validated set of synovial fluid metabolite biomarkers that
predict OA grade and pain levels (2) identification of metabolomic endotypes of OA, and (3) assessment of
correlations between metabolite levels in the synovial fluid and the plasma. This will provide both clinicians
and basic scientists with improved information for diagnosing and treating debilitating osteoarthritis.
