Background It has previously been shown that the survival of anti-TNF drugs was lower in Rheumatoid arthritis (RA) than in other chronic arthritis.
Objectives The objective of this study was to investigate the drug retention, as surrogate of effectiveness, of Golimumab (GOL) with axial Spondyloarthritis (AxSpA), Psoriatic arthritis (PsA) and RA. Furthermore, predictors of drug retention were evaluated.
Methods This is a multicenter prospective observational study of unselected patients with AxSpA, PsA and RA starting GOL because of their active disease despite prior treatment with conventional disease modifying drugs (DMARDs) or biological DMARDs (bDMARDs) in the settings of standard care from 11 rheumatologic centres of Apulia (Southern Italy). The primary endpoint was GOL drug retention at 2 years. Time to discontinuation was defined as the time between drug initiation and last administration plus one dispensation interval. Kaplan-Meier curve analysis was used to assess drug retention, and a Cox proportional regression model, adjusting for potential confounders, including prior bDMARD exposure, glucocorticoid or DMARD intake, patient demographics and disease characteristics, for the analysis of predictors. Clinical response was evaluated at 3,6,12 and 24 months.
Results To date, 416 patients (PsA 180, RA 89, AxSpA 147) have been included in the study, 171 (41%) were biologic-naïve. Disease duration was 6.9 ±6 yrs for PsA, 8.1 ±8 for RA, and 7.4 ±7 for AxSpA, respectively. RF/ACPA were present in 72.5% of RA patients, and HLA-B27 in 43% of AxSpa. Within PsA 58 patients had also spinal involvement. At 2 years (Fig.1), the drug persistence rate was 74.5% (20.8±0.5 months) for AxSpa, 62.5% (19.5±0.8 months) for RA, and 63.9% (18.8±0.6) months for PsA, being the difference not statistically significant. For AxSpa patients, the chance of stopping GOL was associated with the female gender (hazard ratio (HR) 3.15 (95% CI 1.4–8.5), and the presence of extra-articular manifestation (HR 0.20 (95% CI 0.48–0.88). In RA the only predictor of GOL discontinuation was the achievement good EULAR response at 3 months (HR 0.36 ((95% CI 0.15–0.88), while no baseline factors correlating with drug persistence was detected for PsA. The percentage of AxSpa patients achieving BASDAI<4 was 63.4% at3 months, 69.3% at6 months, 77.5% at12 months, and 70.8% at24 months. In RA, DAS28 based remission was attained in 22.5% at3 months, 26.8% at6 months, 36.2% at12 months, and 43.9% at24 months. Finally for PsA DAS28 based remission was found in 27.5% at3 months, 33.2% at6 months, 42.1% at12 months, and 42.6% at24 months, while BASDAI<4 was 56.4% at 3 months, 73.1% at 6months, 76.2% at12 months, and 82.1% at24 months.
Conclusions Unlike previous studies on TNF inhibitors, GOL seems to have similar drug survival in different chronic arthritides. Male sex and the presence of extra-articular manifestations were independently associated with GOL discontinuation in AxSpa. The achievement of good EULAR response at 3 months was the best predictor of continuing GOL in RA.
Disclosure of Interest None declared