Predictors of Relapse in Polymyalgia Rheumatica Patients Treated with Low-Dose Glucocorticoid Therapy

Abstract

Polymyalgia rheumatica (PMR) is a chronic inflammatory condition characterised by subacute onset shoulder and pelvic girdle pain, and early morning stiffness. Oral glucocorticoids represent the treatment mainstay and whilst cessation of therapy is the goal, a subset of patients with relapsing disease remains poorly delineated.

This thesis aimed to characterise the demographic, clinical, laboratory and imaging phenotype of patients that relapse by undertaking a prospective longitudinal cohort study of newly diagnosed, steroid naive PMR cases (2012 EULAR/ACR classification criteria) treated with a weaning schedule of prednisolone (BSR guideline). Disease relapse, as defined by the PMR activity score (PMR-AS), represented the primary outcome measure and was observed in 25/32 participants (78.1%) during the 46-week follow-up period. Older age at diagnosis (HR 1.07) and female sex (HR 2.38) were subsequently associated with an increased risk of relapse, whilst a volar pattern of 18F-FDG uptake on whole body PET/CT conferred a distinct prognostic advantage (HR 0.25). Half of the study participants also met criteria for the secondary endpoint of glucocorticoid-resistant disease. Baseline predictors for this outcome again included older age and female sex, along with high BMI and high neutrophil to lymphocyte ratio (NLR).

Results from this work additionally informed a proof-of-concept PET/MRI fusion study identifying hamstring peritendonitis as the anatomic correlate of abnormalities observed adjacent to the ischial tuberosities. A nested case-control subsequently achieved a sensitivity of 90.9% and specificity of 92.4% for diagnosing PMR on whole body PET/CT based upon a novel scoring algorithm that combined this finding with other key sites of extra-capsular pathology. Finally, data was utilised for preliminary validation of candidate instruments being considered for inclusion in a PMR core outcome measurement set to be endorsed by the international body, OMERACT.

Overall, the findings presented in this thesis confirm the hypothesis that discrete phenotypic differences exist among PMR patients with relapsing disease and make a substantive novel contribution to the literature concerning the pathology, diagnosis, monitoring and management of this rheumatic disease. The low incidence of sustained clinical remission observed in this study following a standardised wean of glucocorticoid therapy should however serve as a timely reminder of the ongoing need for an alternate treatment paradigm in PMR.