POS1345 Apremilast Reduces Mri Inflammation in Individual Joints and Clinical Dactylitis in Patients with Psoriatic Arthritis: Dynamic Contrast-enhanced Mri (dce-mri) Results From the Phase 4 Mosaic Study

Background:
Psoriatic arthritis (PsA) is a chronic, inflammatory arthritis that often presents with asymmetric manifestations like painful swelling of the joints and dactylitis. Primary results from the phase 4 MOSAIC trial demonstrated that apremilast (APR), an oral immunomodulating phosphodiesterase-4 inhibitor approved for treatment of PsA, improved measures of inflammation in patients with PsA as assessed by Magnetic Resonance Imaging (MRI) using Psoriatic Arthritis Magnetic Resonance Imaging Scoring (PsAMRIS) of the hand and MRI Whole-Body Scoring System for Inflammation in Peripheral Joints and Entheses [1]. In MOSAIC, optional dynamic contrast-enhanced MRI (DCE-MRI) also was performed in the majority of cases. DCE-MRI is a highly sensitive technique that allows for evaluation of perfusion through automated quantification of gadolinium contrast agent uptake within the synovial membrane and bone marrow, thereby estimating the degree of joint inflammation [2].
Objectives:
To assess the impact of APR 30 mg twice daily on joint inflammation and dactylitis in patients with PsA based on DCE-MRI and Leeds Dactylitis Index (LDI) [3] outcomes.
Methods:
MOSAIC (NCT03783026) was a phase 4, multicenter, single-arm, open-label study evaluating APR as monotherapy or in combination with stable methotrexate in patients with diagnosed PsA (3 months to 5 years, meeting CASPAR inclusion criteria). Patients (n=122) received APR for up to 48 weeks. DCE-MRI was performed at baseline, Week 24, and Week 48. Two independent readers blinded to clinical outcomes and time points reviewed DCE-MRI with the Dynamic Contrast-Enhanced MRI Quantification (DEMRIQ) method, which automatically calculates the volume of enhancement (DEMRIQ-Vol), initial rate of enhancement (DEMRIQ-IRE), and degree of maximum enhancement (DEMRIQ-ME) within a region of interest [2]. All finger joints were assessed. Here, we report changes from baseline in these 3 DCE-MRI outcomes from a protocol-specified analysis of all joints of the hand and from a post hoc analysis focusing on the most-affected (according to the highest DEMRIQ value at baseline for each parameter) distal interphalangeal (DIP) joint, proximal interphalangeal (PIP) joint, and metacarpophalangeal joint (MCP) at Weeks 24 and 48; these outcomes were further analyzed based on baseline disease activity per Clinical Disease Activity Score in Psoriatic Arthritis (cDAPSA): moderate (ModDA; cDAPSA >13 to ≤27) or high (HDA; cDAPSA >27). We also report clinical outcomes of dactylitis that were not blinded, including change from baseline in LDI and the proportion of patients with dactylitis count improved to 0, assessed at baseline, Week 24, and Week 48. Statistical significance is defined at a 5% significance level.
Results:
At baseline, DCE-MRI was performed and analyzed in 110 patients. Demographics and other baseline characteristics have been previously reported [1]. The predefined analysis of DCE-MRI data showed no statistically significant changes in mean results of all-joint data (data not shown), possibly due to a diluting effect from the non-affected joints. Focus on the most-affected joints based on DEMRIQ values at baseline revealed a significant decrease from baseline in mean DEMRIQ-VOL and DEMRIQ-ME scores, but not DEMRIQ-IRE, for the most-affected DIP, PIP, and MCP joints at Weeks 24 and 48 (Table 1). DCE-MRI images with superimposed maps of maximum enhancement are presented in Figure 1, showing a reduction from baseline in inflammation of joints (Figure 1A) and dactylitis (Figure 1B) over time in two patients treated with APR. In most cases, both ModDA (n=53) and HDA (n=53) populations showed statistically significant decreases in DEMRIQ-Vol and DEMRIQ-ME for all the most-affected joints at Weeks 24 and 48 (P-values ranging from <0.001 to ≤0.05). DEMRIQ-IRE quantification showed more modest non-significant improvements across all joints for both ModDA and HDA populations, except for the most-affected DIP for the ModDA population, for which the change was negligible at both Weeks 24 and 48. Dactylitis was present in 45 patients at baseline, as per LDI. There were statistically significant decreases from baseline in the least-squares (LS) mean (95% CI) in LDI at Week 24 (-34.38 [-41.49, -27.27]) and Week 48 (-38.71 [-46.85, -30.57]) among patients with pre-existing dactylitis. At Week 48, 93.3% of patients (28/30) had their dactylitis count improve to 0.
Conclusion:
Individual joint inflammation assessed by DCE-MRI in patients with PsA treated with APR significantly decreased from baseline at Weeks 24 and 48 in all of the most-affected joints (DIP, PIP, and MCP) whereas the mean sum of the total joint DEMRIQ indices did not. These data demonstrate a value of DCE-MRI technology for targeted analysis and indicate a treatment-induced decrease in inflammation in the joints most severely affected by inflammation in PsA.