GO-KIDS Imaging Substudy: MRI treatment response to 30 mg/m² 4-weekly subcutaneous injections of Golimumab in children with polyarticular JIA– preliminary results of the open label portion

Background/Purpose: The GO-KIDS trial is a double blind placebo controlled Phase III trial, where patients 2-17 yrs of age with pJIA receive open label SC GLM 30 mg/m² every 4 weeks until week 16, followed by 1:1 randomization to placebo or GLM from Weeks 16ˆ48 for children achieving ACR Pedi 30 response at Week 16. Primary endpoint is % of children who experience flares of JIA between Weeks 16 and 48.

173 patients were enrolled.

The aim of the MRI based GO-KIDS Imaging  substudy  is to evaluate the efficacy of GLM on MRI measurements within 16 weeks of drug initiation.

Methods: Patients participating in the GO-KIDS trial at a single center could voluntarily participate in the GO-KIDS Imaging substudy after obtaining separate informed consent. Gadolinium enhanced MRI of a reference peripheral joint region and the temporomandibular joints (TMJ) were obtained at week 0 (baseline) and week 16 (before randomization). MRI was scored (grade 0 – 3) for peripheral joints/TMJ for synovitis (S), synovial hypertrophy (SH) and for peripheral joints for tenosynovitis (TS). Statistical analysis: Shapiro- Wilk-W-test (for normality of distribution); non-parametric Wilcoxon rank-sum test (changes of the MRI scoring).

The clinical response data (ACR Pedi Score) were extracted from the GO-KIDS trial and compared to the MRI response.

Results: 19 children (10 female) with a median age of 9 (7-17) years at study inclusion participated in the GO-KIDS Imaging substudy. At week 16 the number of ACR Pedi 100 responders was 11/19, ACRPedi 90 – 15/19, ACR Pedi 70 – 17/19 and 2 children were non-responders.

At week 16 the MRI was significantly improved compared to week 0 in all components of the MRI scoring: peripheral joints (p<0.001 for S, p= 0.001 for SH and p= 0.002 for TS) and TMJ (p<0.001 for S and SH each). Eight patients showed no inflammatory activity on MRI at week 16 and 9 patients significantly (p<0.002) improved. Two patients did not demonstrate significant change on MRI (same as ACR Pedi non-responders).

MRI results did not completely correlate with improvements in ACR Pedi Score: 5 children showed inactivity on MRI but did not achieve an ACR Pedi Score of 100.

Conclusion: During the open label treatment with GLM of the GO-KIDS trial, children with pJIA demonstrated significant improvement on MRI measurements between Weeks 0 and 16 with a large proportion of patients achieving MRI inactivity.  MRI appeared to be more sensitive in detecting inflammatory inactivity than ACR Pedi scoring systems.