Giant Cell Arteritis: Diagnostic Prediction Models, Temporal Artery Biopsy and Epidemiology

Giant cell arteritis (GCA) is the most common primary vasculitis in the elderly and can cause irreversible blindness, aortitis, and stroke. Diagnostic confirmation of GCA usually entails temporal artery biopsy (TABx) - a time-consuming and invasive test, or ultrasound. The primary treatment of GCA is with high dose glucocorticoids that have numerous potential side effects. Glucocorticoids are initiated prior to the TABx result, due to the risk of interim blindness. By 2050 the cost of blindness from GCA in the United States is estimated at $76 billion with an additional $6 billion from glucocorticoid-induced fractures. This thesis examines knowledge gaps in the diagnosis and epidemiology of GCA. Needed refinements in the diagnosis of GCA included: i) the optimization of diagnostic prediction models (PMs) and ii) clarification of the contemporary utilization parameters of TABx. With regards to i) previous PMs are usually based on limited sample size, do not leverage sufficient clinical predictors, or include continuous variables, and not compliant with the transparent reporting guidelines for diagnostic PMs (TRIPOD). Using multicentre data of consecutive patients undergoing TABx, the largest (n=1,201) and most comprehensive logistic regression and, neural network PMs for GCA were formulated. Age, platelet level, jaw claudication and vision loss eventuated as the key predictor variables. An online risk calculator was developed from the PM and could decrease both the number of TABx performed on low-risk patients, and the morbidity from unneeded glucocorticoids. Regarding ii) although TABx has long been acknowledged as the gold standard test for GCA the current preference for TABx versus ultrasound amongst neuro-ophthalmologists and the utility rate of TABx are unknown. The thesis survey revealed that 91% of neuro- ophthalmologists preferred TABx over ultrasound as the confirmatory test for GCA. The first systematic review for the utility rate of TABx disclosed a median positive yield of 25% and provides a benchmark for institutions performing this procedure. Knowledge gaps in the epidemiology of GCA important for public health planning included the incidence of GCA in Ontario, Canada, and the controversial role of herpes zoster in the development of GCA given the advent of zoster vaccines. Pathology audit and an assay of billing data revealed the incidence of biopsy-proven GCA in Ontario to be 4.9 per 100,000 individuals 50 years of age or older. On ecologic analysis, the inverse relationship of the incidence rates of herpes zoster versus GCA per country suggested zoster is not a major immunopathogenic trigger for GCA. In summary, this thesis discusses the diagnosis and epidemiology of GCA, most notably in the area of clinical prediction models that aid in the triage of patients with suspected GCA.