Study population
Patients from the IVEPsA study [6] on very early PsA (N = 20) and the PSARTROS [5] study on established PsA (N = 20) were included in this analysis. The very early PsA patients of the IVEPsA study were required to fulfill the following criteria: (i) presence of moderate to severe psoriasis (PASI score > 6) or scalp or nail involvement; (ii) absence of present or past joint swelling, active enthesitis, or dactylitis; and (iii) presence of subclinical inflammatory changes in magnetic resonance imaging (MRI) or erosive changes in MRI or high-resolution peripheral quantitative computed tomography (HR-pQCT). Established PsA patients of the PSARTROS study had to fulfill the CASPAR classification criteria of PsA, had to have symptoms of PsA for at least 6 months prior to inclusion into the study, and had to have an active joint disease with at least 3 tender and 3 swollen joints on the 78/76 joint count.
Corticosteroid treatment was allowed if at stable doses of ≤ 10 mg/day prednisone for at least 2 weeks before baseline and had to remain on a stable dose until the end of the study. Concomitant non-biologic treatment was allowed if on a stable dose for at least 4 weeks before baseline and throughout the study. Previous TNF inhibitor therapy was allowed after appropriate washout periods. Patients, who had previously used drugs targeting IL-17 or IL-23p40, were not eligible to participate in either of the studies. Other conditions such as pregnancy, recent live vaccines, history of tuberculosis, use of opioid analgesics, history of alcohol, or drug abuse within the last 6 months or any uncontrolled medical condition were not allowed. All patients provided written informed consent, and the institutional review board of the University Clinic Erlangen approved the study.
Study procedures and data collection
All patients received secukinumab treatment 300 mg sc. once weekly for the first 4 weeks and once monthly thereafter for a total period of 24 weeks. All patients were longitudinally assessed for 78 tender joint counts (TJC) and 76 swollen joint counts (SJC), the severity of pain and the perceived global disease activity using visual analog scales (VAS), composite disease activity measures for joint (Disease Activity Score 28, DAS-28; Disease Activity Score in Psoriatic Arthritis, DAPSA) and skin disease (total Psoriasis Area and Severity Index, PASI; Body Surface Area, BSA), and several common and disease-specific PROs (SF-36, DLQI, PsAID, and HAQ-DI) at baseline and after 4, 12, and 24 weeks.
Statistical analysis
Baseline study characteristics were tabulated using appropriate summary statistics. The course of patient-reported outcomes and quality of life indicators were plotted as mean scores and 95% confidence intervals over study visits. Changes from baseline values were estimated using linear mixed effects models adjusted for baseline values of each scale, gender, age and disease duration, and plotted as model coefficients and respective 95% confidence intervals that represent adjusted mean absolute improvement from baseline. Scale signs were inverted as necessary to ease interpretability, age, disease duration, and baseline values were centered at their mean. Study visit week was included in the model as a categorical variable using the baseline visit as reference and differences of changes over time by study cohort were assessed using time-cohort interaction terms.
We conducted a cluster analysis of study participants by patient-reported outcomes and overall disease assessment. The scores were centered around their mean and scaled by their standard deviations to generate heat maps, and Euclidean distances were used for k-means clustering to generate marginal dendrograms on the patient and measurement axes. Further subgroup analyses were conducted using the Mann-Whitney U test for comparisons of quantitative variables, chi-square test for comparisons of qualitative variables and Spearman’s rank correlation coefficient for relations. Data manipulation and analyses were conducted using R v. 3.5.3 (R Foundation for Statistical Computing, Vienna. Austria). Two-sided p values less than 0.05 were considered significant.