Clinical and Radiological Manifestations of Coxitis in Patients With Ankylosing Spondylitis Treated With Golimumab

Rationale. Coxitis in AS is inflammation of hip(s) affecting significant number of patients. It is associated with worse function and more expressed axial disease requiring hip replacement at end-stage. Number of studies dedicated to coxitis treated with TNF alpha inhibitors is very limited. Primary objective. To evaluate change of functional impairment in AS patients with coxitis from baseline to 12 months of therapy with golimumab by BASFI in daily clinical practice Study design. This study is a non-interventional prospective observational cohort study conducted in multiple centers across Russia. Study population. Patients with ankylosing spondylitis (according to the modified New York criteria) with coxitis newly prescribed golimumab during the course of usual clinical care will be enrolled and followed prospectively for 24 months with data collection at the approximate time points: baseline (pre-treatment) and consequent every 6 months. Patients will receive golimumab as prescribed in regular clinical practice. Statistical Methods. Quantitative variables will be tested for normal distribution using the Shapiro-Wilk test. The hypothesis of equality of variances will be tested using Levene's test. Quantitative variables matching a normal distribution will be described in terms of the mean ± standard deviation, and values outside of the normal distribution as medians, 25% and 75% quartiles. Qualitative variables will be presented in the form of percentages of the absolute value N. Sample Size and Power Calculations. Sample size calculation is based on data from GO-RAISE registration study in which golimumab was evaluated in patients with AS. The baseline BASFI was 5.0 which changed by mean (± SD, standard deviation) -2.5 (± 2.12) at week 52 (approximately 12 months). The analysis of the data shows that the minimum necessary sample size should be 18 patients to show a statistically significant change of BASFI from baseline to 12 months. On this basis and considering that the dropout rate over the two years is expected to be 30%, we decided that the necessary number of patients to be included into protocol should be 39 individuals. It is expected that around 27 patients will be included in the patient set completing the study. Power of the study is 90% with formula evaluation P = 1-β where P is power and β is type 2 error = 10%.