Detecting Treatment Failure in Rheumatoid Arthritis with Time-Domain Diffuse Optical Methods
Abstract
Rheumatoid arthritis (RA) is the most common type of inflammatory arthritis, and affects approximately 1% of the population in Canada. While the disease has no cure, early treatment within the first 3-6 months of onset is known to substantially reduce disease progression and improve patient prognosis. Nevertheless, identifying which therapy will elicit the appropriate treatment response depends on a time-consuming, trial-and-error approach. Thus, there is a strong clinical motivation to develop treatment monitoring methods which signal the need for treatment adaptation as early as possible; this helps ensure that patients reap the benefits of effective early treatment, and mitigates the risk of irreversible joint damage.
The limitations of current monitoring methods include subjectivity, low sensitivity, high cost, and operator dependence. Diffuse optical methods offer an objective, sensitive, low-cost, and operator-independent solution for monitoring RA disease activity. Previous diffuse optical methods used continuous-wave and frequency-domain techniques to identify joint inflammation; however, little work has explored time-domain (TD) methods. TD techniques typically provide richer information content, which may be leveraged to increase sensitivity to subtle changes in RA disease activity.
This dissertation investigates the prospects of two TD diffuse optical methods for RA treatment monitoring. First, a contrast-enhanced near-infrared spectroscopy technique was used to monitor joint blood flow (BF) changes in a longitudinal study of a rat model of inflammatory arthritis. However, the study found no significant difference in joint BF between controls and rats with induced arthritis. Second, a novel TD diffuse optical imaging (DOI) method for monitoring RA disease activity was developed and assessed in silico; this method was then implemented experimentally and tested on disease-mimicking phantoms. Spatiotemporal Fourier components extracted from simulated TD-DOI images were strongly correlated with a measure of virtual RA disease activity, and components acquired by the experimental TD-DOI system could clearly distinguish between phantoms that mimicked different RA disease activities. These findings suggest that TD-DOI has the potential to be a sensitive treatment monitoring tool for RA, and future work should test its efficacy in RA patients.