Seventeen-year reassessment of diagnostic transitions, biologic therapy initiation and mortality in spondyloarthritis: results from the REGISPON-3 study

Abstract

Objectives: To evaluate disease evolution, diagnostic transitions, time to biologic treatment initiation and mortality in patients with spondyloarthritis (SpA) after 17 years from the original Spanish Registry of Spondyloarthritis (REGISPONSER).

Methods: Spondyloarthritis Registry 3 (REGISPON-3) is a two-timepoint longitudinal study with patients fulfilling the European Spondyloarthropathy Study Group criteria participating in the REGISPONSER study and re-evaluated after 17 years. Clinical, laboratory, radiological and treatment data were collected and compared with baseline. Diagnostic changes according to the rheumatologist's judgement and their associated factors were analysed using multivariable logistic regression. Time to initiation of biologic therapy was evaluated using Kaplan-Meier survival analysis.

Results: A total of 536 patients from the REGISPONSER study conducted in 2004 were contacted in 2021 for the REGISPON-3 visit. Of these, 411 were physically re-evaluated, while 125 were confirmed deceased. Among the 411 patients, 31.6% experienced a change in diagnosis in the REGISPON-3 visit, mainly from undifferentiated SpA to axial SpA or psoriatic arthritis. In multivariable analysis, dactylitis, younger age and lower Bath Ankylosing Spondylitis Radiology Index scores were associated with diagnostic change. The use of biologic disease-modifying antirheumatic drugs (bDMARDs) increased from 13.1% to 52.1%. However, the median time to first bDMARD from symptom onset was 32 years (95% CI 30 to 35) and 28 years from diagnosis (95% CI 26 to 32). Among the 125 patients who died, the leading causes were infections (21.6%), cardiovascular (CV) events (20.0%) and cancer (19.2%).

Conclusions: This long-term reassessment reveals significant diagnostic changes in SpA and a mortality burden mainly attributed to infections and CV diseases.

Keywords: arthritis, psoriatic; biological therapy; mortality; outcome assessment, health care; spondyloarthritis.

Figure 1. Sankey plot representing changes in the diagnosis in the overall population (A), among undifferentiated SpA (B) and among undifferentiated SpA stratified by sex (C). (A) Sankey plot illustrating diagnostic transitions of the overall cohort from baseline to follow-up; most patients with AS (n=270) evolved to r-axSpA (n=223, 82.6%). Other changes included AS to axPsA (n=15, 5.6%), IBD-SpA (n=12, 4.4%), PsA (n=9, 3.3%), nr-axSpA (n=5, 1.9%), Juv-SpA (n=2, 0.7%) and pSpA (n=1, 0.4%). (B) From the overall patients with u-SpA (n=61), 36 (60%) progressed towards r-axSpA, 10 (16.67%) towards nr-axSpA, five (8.33%) towards pSpA, four (6.67%) towards other diagnoses outside the SpA spectrum, two (3.33%) towards axPsA and one (1.67%) towatds PsA. Two patients (3.3%) retained the diagnosis of u-SpA, and for one patient (1.6%), diagnostic evolution could not be determined due to missing data. (C) Sankey diagrams illustrating sex-stratified diagnostic transitions for u-SpA: males (top, blue) mainly to r-axSpA (n=21, 72.4%), nr-axSpA (n=3, 10.3%), pSpA (n=2, 6.9%), axPsA (n=1, 3.4%), with two (6.9%) retaining u-SpA; females (bottom, red) to r-axSpA (n=15, 46.9%), nr-axSpA (n=7, 21.9%), pSpA (n=3, 9.4%), axPsA (n=1, 3.1%), PsA (n=1, 3.1%) and other diagnoses (n=4, 12.5%). AS, ankylosing spondylitis; axPsA, axial psoriatic arthritis; IBD-Arthritis, inflammatory bowel disease-associated arthritis; nr-axSpA, non-radiographic axial spondyloarthritis; p-SpA, peripheral spondyloarthritis; PsA, psoriatic arthritis; r-axSpA, radiographic axial spondyloarthritis; ReA, reactive arthritis; uSpA, undifferentiated spondyloarthritis.

Figure 2. Time to bDMARD from first sign/symptom in the overall population (A) and stratified by disease (B). (A) Kaplan-Meier survival curve for the overall population (n=402; 208 events, 194 censored); median time: 32 years (95% CI 30 to 35). (B) Stratified by diagnosis: PsA (n=58; median 24 years, 95% CI 18 to 30) and axSpA (n=258; median 33 years, 95% CI 31 to 36); log-rank test p=0.00013. axSpA, axial spondyloarthritis; bDMARD, biologic disease-modifying antirheumatic drug; PsA, psoriatic arthritis.

Figure 3. Time to bDMARD from diagnosis in the overall population (A) and stratified by disease (B). (A) Kaplan-Meier survival curve for the overall population (n=402; 208 events, 194 censored); median time: 28 years (95% CI 26 to 32). (B) Stratified by diagnosis: PsA (n=58; median 22 years, 95% CI 17 to 27) and axSpA (n=258; median 29 years, 95% CI 25 to 33); log-rank test p=0.02. axSpA, axial spondyloarthritis; bDMARD, biologic disease-modifying antirheumatic drug; PsA, psoriatic arthritis.