Long-Term Golimumab Treatment Persistence and Clinical Outcomes in Axial Spondyloarthritis: A Real-World Cohort from Kazakhstan

Background/Objectives: Long-term real-world evidence on golimumab in axial spondyloarthritis (axSpA) from Central Asia is limited. We evaluated clinical outcomes through week 144, treatment persistence through 36 months, factors associated with discontinuation, and reasons for discontinuation in Kazakhstan. Methods: This prospective observational cohort included 113 adults with axSpA enrolled at two rheumatology centers in Almaty in October 2019. All had received golimumab 50 mg subcutaneously every four weeks for at least 12 weeks before enrollment. Time zero for persistence analyses was the enrollment date. Clinical assessments were performed at baseline and weeks 12, 24, 52, 104, and 144. Treatment persistence was estimated using Kaplan–Meier analysis, and factors associated with confirmed discontinuation were explored using Cox regression. Clinical responses were assessed using non-responder imputation (NRI; N = 113). Results: During 36 months, 24 patients had confirmed permanent discontinuation; two patients lost to follow-up without confirmed discontinuation were censored. Kaplan–Meier estimated persistence was 89.3% (95% CI, 83.8–95.2%) at 12 months, 79.4% (72.2–87.3%) at 24 months, and 78.5% (71.2–86.5%) at 36 months. Among 88 patients with paired baseline and week 144 assessments, the Bath Ankylosing Spondylitis Disease Activity Index (BASDAI), Ankylosing Spondylitis Disease Activity Score based on erythrocyte sedimentation rate (ASDAS-ESR), Bath Ankylosing Spondylitis Functional Index (BASFI), Bath Ankylosing Spondylitis Metrology Index (BASMI), and Maastricht Ankylosing Spondylitis Enthesitis Score (MASES) improved significantly (all p < 0.001). With NRI, 64.6% achieved BASDAI50, 59.3% ASDAS inactive disease, and 54.9% ASDAS major improvement at week 144. Higher Rheumatic Disease Comorbidity Index (RDCI) score (adjusted hazard ratio [HR] 1.312, 95% CI, 1.042–1.653) and functional class III–IV versus I–II (adjusted HR 2.412, 95% CI, 1.037–5.611) were associated with higher discontinuation hazards in exploratory analyses. Loss of effectiveness was the most frequent reason for discontinuation (54.2%). Adverse events were recorded only when they led to discontinuation and therefore do not represent overall safety incidence. Conclusions: Golimumab was characterized by substantial 36-month treatment persistence and clinical improvement among patients who remained on treatment and were evaluable at week 144 (n = 88). Full-cohort NRI analyses also demonstrated clinically relevant responses at week 144. Comorbidity burden and poorer functional class were associated with discontinuation in exploratory analyses. These findings add long-term real-world evidence from an under-represented geographic setting.