OP0141 EFFECTS OF FILGOTINIB ON SPINAL LESIONS IN ANKYLOSING SPONDYLITIS: MAGNETIC RESONANCE IMAGING DATA FROM THE TORTUGA TRIAL

Abstract

Background:The oral Janus kinase 1 preferential inhibitor filgotinib (FIL) significantly improved Spondyloarthritis Research Consortium of Canada (SPARCC) magnetic resonance imaging (MRI) inflammation scores (bone marrow oedema) in the spine and sacroiliac joints vs placebo (PBO) in the Phase 2 TORTUGA trial (NCT03117270) in patients with active ankylosing spondylitis (AS).1Objectives:This post-hoc analysis evaluated the effects of FIL on Canada-Denmark (CANDEN) MRI measures of spinal inflammation and structural lesions in patients from the TORTUGA trial.Methods:TORTUGA was a PBO-controlled, multicentre, double-blind, randomised trial. Patients with active AS (as per modified New York classification criteria, with sacroiliitis confirmed by central reading) were treated with FIL 200 mg (n=58) or PBO (n=58) once daily for 12 weeks. MRI of the total spine was conducted at baseline and at treatment end. Scans were re-evaluated post-hoc by 2 independent experts (blinded to time point and assigned treatment) using the CANDEN method;2 inter-reader discrepancies were resolved by an independent adjudicator. Observed changes from baseline were evaluated using analysis of covariance, with factors for treatment, baseline value, and randomisation stratification by prior tumour necrosis factor inhibitor use. Least-squares (LS) mean changes from baseline and between-group differences with 95% confidence intervals (CI) were calculated; P values are nominal.Results:MRI scans from 88 patients (47 FIL, 41 PBO) with an evaluable scan at baseline and Week 12 (or early termination) were re-evaluated. Baseline characteristics were generally similar between patients with/without an MRI scan. Of those with MRI scans, mean total spine inflammation score (which ranges from 0–614) was higher, and mean ankylosis score (which ranges from 0–460) was lower, in the FIL vs PBO group at baseline. Total spine inflammation scores decreased from baseline with FIL but not with PBO (Figure and Table; P=0.0003 for between-group difference). Cumulative probability plots favoured FIL over PBO for change from baseline in subregion inflammation scores, including posterolateral elements (i.e. sum of lesions in ribs, transverse processes, spinous processes, soft tissue inflammation, and postero-lateral vertebral body), facet joint, and vertebral body. Total spine fat lesion scores numerically increased from baseline in the FIL but not PBO group (P=0.0878 for between-group difference; Table). There were no significant differences between groups for changes in erosion (P=0.1956) or ankylosis (P=0.3888) scores (Table).Table 1.Change from baseline at Week 12 in CANDEN total spine inflammation, total spine fat, total spine bone erosion, and ankylosis scoresTreatment groupnSample mean (SE)LS mean (SE)95% CI of treatment meanLS mean of group difference (SE)95% CI of group differenceBetween-group P valueTotal spine inflammation scoreFilgotinib47–4.98 (0.96)–4.40 (1.13)–6.65, –2.15–4.49 (1.21)–6.85, –2.120.0003Placebo410.29 (0.78)0.09 (1.13)–2.17, 2.34Total spine fat scoreFilgotinib471.01 (0.62)1.09 (0.66)–0.22, 2.401.18 (0.69)–0.18, 2.550.0878Placebo41–0.25 (0.19)–0.09 (0.66)–1.40, 1.21Total spine bone erosion scoreFilgotinib470.01 (0.02)0.07 (0.03)0.00, 0.140.05 (0.04)–0.02, 0.120.1956Placebo41–0.02 (0.03)0.02 (0.03)–0.04, 0.09Total ankylosis scoreFilgotinib470.30 (0.29)0.23 (0.31)–0.40, 0.850.28 (0.34)–0.37, 0.940.3888Placebo41–0.01 (0.08)–0.06 (0.31)–0.68, 0.56SE, standard errorConclusion:This is the first PBO-controlled trial to demonstrate a decrease in inflammatory activity with FIL, not only in the spinal vertebrae but also in the postero-lateral elements of the spine and facet joints. As expected in a 12-week study period, no changes in erosion or ankylosis were seen, while fat lesions showed a tendency to increase with FIL. Larger trials are needed to confirm these results.