Skip to main content
NCBI home page
RMD Open logoLink to RMD Open
. 2023 Dec 1;9(4):e003455. doi: 10.1136/rmdopen-2023-003455

Early axial spondyloarthritis according to the ASAS consensus definition: characterisation of patients and effectiveness of a first TNF inhibitor in a large observational registry

Adrian Ciurea 1,✉, Andrea Götschi 2, René Bräm 3, Kristina Bürki 1, Pascale Exer 4, Michael Andor 5, Michael J Nissen 6, Burkhard Möller 7, Thomas Hügle 8, Andrea Rubbert-Roth 9, Diego Kyburz 10, Oliver Distler 1, Almut Scherer 2, Raphael Micheroli 1
PMCID: PMC10693870  PMID: 38053462

Abstract

Objective

To characterise the population fulfilling the Assessment of SpondyloArthritis international Society (ASAS) consensus definition of early axial spondyloarthritis (axSpA) and to determine the effectiveness of a first tumour necrosis factor inhibitor (TNFi) in early versus established axSpA in a large observational registry.

Methods

A total of 3064 patients with axSpA in the Swiss Clinical Quality Management registry with data on duration of axial symptoms were included (≤2 years=early axSpA, N=658; >2 years=established axSpA, N=2406). Drug retention was analysed in patients starting a first TNFi in early axSpA (N=250) versus established axSpA (N=874) with multiple-adjusted Cox proportional hazards models. Adjusted logistic regression analyses were used to determine the achievement of the ASAS criteria for 40% improvement (ASAS40) at 1 year.

Results

Sex distribution, disease activity, impairments of function and health-related quality of life were comparable between patients with early and established axSpA. Patients with established disease were older, had more prevalent axial radiographical damage and had a higher impairment of mobility. A comparable TNFi retention was found in early versus established disease after adjustment for age, sex, human leucocyte antigen-B27 status, education, body mass index, smoking, elevated C reactive protein and sacroiliac inflammation on MRI (HR 1.05, 95% CI 0.78 to 1.42). The adjusted ASAS40 response was similar in the two groups (OR 1.09, 95% CI 0.67 to 1.78). Results were confirmed in the population fulfilling the ASAS classification criteria.

Conclusion

Considering the recent ASAS definition of early axSpA, TNFi effectiveness seems comparable in early versus established disease.

Keywords: Epidemiology; Spondylitis, Ankylosing; Tumor Necrosis Factor Inhibitors

WHAT IS ALREADY KNOWN ON THIS TOPIC.

  • The Assessment in SpondyloArthritis international Society (ASAS) has recently developed a consensus definition for early axSpA, with up to 2 years of axial symptom duration as a mainstay for use in research settings.

  • In previous analyses, different definitions of early axial spondyloarthritis (axSpA) were used, and earlier treatment was not associated with a better response to biological disease-modifying antirheumatic drugs.

WHAT THIS STUDY ADDS

  • This investigation in a large real-life cohort characterises the population with axSpA fulfilling the ASAS definition of early axSpA.

  • Initiating a first tumour necrosis factor inhibitor during the early disease phase, as newly defined by the ASAS, seems not to lead to better drug retention or better response rates compared with a later start of treatment.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

  • If the results can be replicated in other cohorts, analyses using shorter cut-offs of axial symptom duration are warranted.

Introduction

The concept of a ‘window of opportunity’ is well accepted for rheumatoid arthritis (RA).1 It summarises the findings that treatment with disease-modifying antirheumatic drugs (DMARDs) has an increased potential of preventing functional impairment and radiographical damage in RA if it is initiated within 1–2 years after diagnosis.2 Whether a similar concept holds true for axial spondyloarthritis (axSpA) remains open,3 as early treatment with biological (b)DMARDs was not consistently associated with better outcomes.4 Whether the failure to demonstrate a better response to bDMARDs in early treatment in axSpA might be due to the large heterogeneity of the definition of early disease used in the respective analyses remains unclear. Early axSpA was identified in a recent review of the literature by short symptom duration or short disease duration, with cut-offs ranging from 5 to 10 years, or based on the absence of radiographical sacroiliitis.5 The Assessment in SpondyloArthritis international Society (ASAS) has, therefore, recognised the need for a standardised definition of ‘early axSpA’ and has recently published a consensus definition for research purposes: ‘early axSpA’ is defined as duration of ≤2 years of axial symptoms in patients already diagnosed as having axSpA, irrespective of the presence of radiographical axial damage.6 The definition is complemented by a statement that axial symptoms should include spinal/buttock pain or morning stiffness and that their presence should be considered by a rheumatologist as related to axSpA.6 The proposed consensus definition of early axSpA and its cut-off of 2 years are recognised as ambitious,6 as the diagnostic delay in axSpA in clinical practice is still very long.7 Moreover, there is currently no scientific evidence to support either the chosen cut-off or the start of axial symptoms as mainstays of the definition.6 Therefore, we aimed at characterising the population fulfilling the new definition of early axSpA in a large cohort with available information on the start of first symptoms and also on the start of axial symptoms, as registered by the treating rheumatologist, to better evaluate the feasibility of its use for current research. In a second step, we used the new definition to analyse the effectiveness of treatment with a first tumour necrosis factor inhibitor (TNFi) in early axSpA compared with established axSpA.

Methods

Characterisation of patients with early axSpA at inclusion in the Swiss Clinical Quality Management axSpA cohort

We took advantage of a large ongoing cohort of patients diagnosed with axSpA and recruited between 1 January 2004 and 1 June 2023 in the Swiss Clinical Quality Management (SCQM) registry.8 The treating rheumatologists in private practices, non-academic hospitals and academic institutions9 enter the date of first symptoms and are then asked to enter the presence of back pain of ≥3 months’ duration (yes or no) after interpretation of the patient’s history in the online database. If the answer is yes, the rheumatologist is prompted to specify several items on a list to establish the presence of an inflammatory back pain (IBP) character according to the criteria of ASAS,10 followed by the starting date of axial symptoms.

In comparison with the ASAS definition of early axSpA, which stipulates that axial symptoms should include spinal/buttock pain or morning stiffness, patients with complete absence of axial pain and only presenting with morning stiffness were not included here. Moreover, only patients with ≥3 months of axial symptom duration are considered, which is not an absolute requirement of the ASAS definition of early axSpA. For both reasons, we might have slightly underestimated the actual proportion of early axSpA in SCQM.

Rheumatologists enter data on peripheral and axial manifestations, with information stratified by clinical, radiographical and magnetic resonance-tomographical assessments, only if the respective manifestation is present. The number of imaging procedures performed that yielded negative results and the number of procedures with unknown findings are therefore not known.

Clinical assessments are performed according to the recommendations of ASAS.11 Data on bDMARDs and conventional synthetic DMARDs are entered by the rheumatologist with start and end dates. Data on the use of non-steroidal antirheumatic drugs (NSAIDs) are available as yes or no at the visit level. Laboratory examinations include C reactive protein (CRP) levels and human leucocyte antigen-B27 (HLA-B27) status.

Patients diagnosed as having axSpA were included in the early axSpA characterisation part of the study if information on the starting time point of axSpA-related back pain of at least 3 months’ duration was registered in SCQM by the treating rheumatologist. Patients fulfilling the 2009 ASAS classification criteria for axSpA12 were included in sensitivity analyses. As 1 January and 1 June were indicated as starting dates of axial symptoms in a higher number of instances than expected, it was considered a proxy for not exactly knowing the date within the respective year. We interval-censored all affected dates by assuming the onset date to fall into the reported onset year, and patients were only included if a differentiation between early and non-early axSpA was still possible (2.7% of patients excluded). In instances not affecting differentiation between early and non-early disease, inexact back pain onset dates (knowledge of month and year or only year of start) were mid-imputed. With regard to patient characteristics, missing variables were replaced by values from the closest visit within a range of 150 days before and 100 days after the considered time point (90 days before and 10 days after for disease activity variables). HLA-B27 status and data on family history were mapped from any other visit.

Effectiveness of TNFi treatment in early versus established axSpA

The effectiveness of TNFi treatment was assessed in patients starting a first TNFi after inclusion in SCQM. Drug retention was considered the primary outcome, and we estimated the time individual patients with axSpA maintained their first TNFi treatment when started in early versus established axSpA. Observations were censored at the last visit recorded in SCQM, the last change in medication dosage registered or the patient’s last confirmation of TNFi use via the web-based mySCQM application,13 whatever occurred last. Treatment response, defined as either the proportion of patients achieving the ASAS criteria for 40% improvement (ASAS40) or reaching a 50% reduction in the Bath Ankylosing Spondylitis Disease Activity Index (BASDAI50), was assessed in exploratory analyses in patients with available disease activity measurement at 1 year (±6 months). Patients who had discontinued their first TNFi in the meantime were considered non-responders (response/tolerance analysis).14

Statistical analyses

Baseline characteristics between patients with early versus established axSpA were compared using the Kruskal-Wallis rank-sum test for continuous variables and the Χ2 test for categorical variables. Tests were two sided, with a significance level set at 0.05. Log-rank tests are provided to compare the crude time to treatment discontinuation in early versus established axSpA. The following variables were introduced in multiple-adjusted Cox proportional hazards models to estimate differences in drug retention between early and established disease: age (continuous), sex (female vs male), HLA-B27 status (negative vs positive) and education (vocational and academic vs compulsory, respectively). These variables were considered confounders of our analysis as they may affect not only the outcome but also the exposure (the fact that a patient might be diagnosed earlier and TNFi initiated within 2 years following the start of axial symptoms). The number of missing values per covariate was N=63 for HLA-B27 and N=140 for education. Baseline characteristics are also shown for the population without missing covariates. Ankylosing Spondylitis Disease Activity Score (ASDAS) was added to the main model in a sensitivity analysis. In a further analysis, we added body mass index (BMI), current smoking status, the presence of elevated CRP and the presence of inflammatory MRI changes to the main model. We also checked for the presence of an interaction term between sex and early versus established disease. Event dates were interval-censored to account for the uncertainty of incomplete medication stop dates.

The significance of the unadjusted differences in BASDAI50 and ASAS40 responses was assessed using the Fisher’s exact test. Logistic regression analysis was used to estimate an adjusted ratio for ASAS40 and BASDAI50, with adjustments for age, sex, HLA-B27, and education in the main model and additional adjustments for BMI, current smoking status, the presence of elevated CRP, and the presence of inflammatory MRI changes in a sensitivity analysis. All drug retention and treatment response analyses were performed in all patients diagnosed with axSpA and in the population fulfilling the ASAS axSpA classification criteria.

R statistical software was used for the statistical analyses. Artificial intelligence was only used to check the English grammar in the manuscript (QuillBot).

Results

Characterisation of early axSpA at inclusion in SCQM

Patient disposition in the SCQM axSpA cohort is displayed in figure 1. Out of 3604 patients diagnosed with axSpA by their treating rheumatologist and having information on the start of axial symptoms, 658 patients had axial symptom duration ≤2 years and were classified as having early axSpA6 at the time point of inclusion in SCQM (21.5%). In the population fulfilling the ASAS classification criteria,12 the proportion of patients with early axSpA was 20.3% (N=447 for early disease and N=1753 for established disease). The percentage of early axSpA inclusions per year was constant at around 20%, irrespective of the number of patients fulfilling the ASAS classification criteria included per year since the initiation of the registry (figure 2). The median (IQR) duration since the start of axial symptoms was 0.8 (0.5–1.2) years in early axSpA and 10.1 (5.4–19.0) years in established axSpA (p<0.001; table 1). The median duration since the start of axial symptoms was comparable in the subgroup fulfilling the ASAS classification criteria (table 1). Time from onset of back pain to inclusion is shown for individual patients fulfilling the ASAS classification criteria in figure 3, stratified by early versus established disease. When the start of first symptoms rather than axial symptoms was considered, the median (IQR) symptom duration was only slightly longer: 1.3 (0.8–2.0) years in early axSpA and 10.5 (5.6–19.3) years in established axSpA (p<0.001). While axial symptoms corresponded to the first symptoms in the majority of patients with early axSpA, the time from the first symptom to axial symptoms varied in the remaining population, as shown for individual patients in the online supplemental figure 1. In the population fulfilling the ASAS classification criteria, the disease started with non-axial symptoms in a total of 170 patients with early axSpA (38.0%) and 209 patients with established axSpA (11.9%). The exact nature of these non-axial symptoms is not recorded in SCQM to allow a more detailed analysis.

Figure 1.

Figure 1

Open in a new tab

Patient disposition at inclusion in the SCQM cohort (A) and at start of a first TNF inhibitor (TNFi) (B). ASAS, Assessment of SpondyloArthritis international Society; axSpA, axial spondyloarthritis; SCQM, Swiss Clinical Quality Management; TNF, tumour necrosis factor.

Figure 2.

Figure 2

Open in a new tab

Number of patients with axSpA with early disease fulfilling the ASAS classification criteria included every year in SCQM (A). Proportion of patients with early axSpA for each inclusion year (B). ASAS, Assessment of SpondyloArthritis international Society; axSpA, axial spondyloarthritis; SCQM, Swiss Clinical Quality Management.

Table 1.

Characteristics of patients with axSpA with early disease and established disease at inclusion in SCQM in the population fulfilling the ASAS axSpA classification criteria, as well as in all patients diagnosed as having axSpA

Parameter All patients diagnosed as having axSpA
(main analysis; N=3064)		 Patients fulfilling the ASAS classification criteria
(sensitivity analysis; N=2200)
N Early disease
(≤2 years)
N=658		 N Established disease
(>2 years)
N=2406		 P value N Early disease
(≤2 years)
N=447		 N Established disease
(>2 years)
N=1753		 P value
Male sex, N (%) 658 334 (50.8) 2406 1242 (51.6) 0.73 447 242 (54.1) 1753 994 (56.7) 0.36
Age, years 658 38.0 (12.5) 2406 43.4 (12.2) <0.001 447 33.2 (9.6) 1753 41.2 (11.4) <0.001
Symptom duration, years, median (IQR) 651 1.3 (0.8–2.0) 2380 10.5 (5.6–19.3) <0.001 440 1.3 (0.8–2.5) 1734 11.1 (5.9–20.4) <0.001
Axial symptom duration, years, median (IQR) 658 0.8 (0.5–1.2) 2406 10.1 (5.4–19.0) <0.001 447 0.8 (0.5–1.2) 1753 10.6 (5.6–20.0) <0.001
Time since diagnosis, years, median (IQR) 652 0.3 (0.1–0.9) 2374 2.9 (0.6–8.6) <0.001 441 0.3 (0.1–0.9) 1732 2.9 (0.5–9.5) <0.001
HLA-B27 positive, N (%) 593 353 (59.5) 2179 1396 (64.1) 0.05 411 302 (73.5) 1637 1264 (77.2) 0.13
Family history axSpA, N (%) 578 110 (19.0) 2127 488 (22.9) 0.05 402 90 (22.3) 1567 401 (25.5) 0.21
Body mass index 580 25.1 (4.5) 2090 25.8 (4.7) 0.001 390 24.9 (4.6) 1555 25.7 (4.6) <0.001
Education 516 1848 0.03 346 1374 0.17

  • Compulsory

 69 (13.4) 337 (18.2) 47 (13.6) 245 (17.8)

  • Vocational

 292 (56.6) 963 (52.1) 191 (55.2) 719 (52.3)

  • Academic

 155 (30.0) 548 (29.7) 108 (31.2) 410 (29.8)
Recruiting rheumatologist 658 2406 <0.001 447 1753 <0.001

  • Private practice

 359 (54.6) 1303 (54.2) 217 (48.5) 918 (52.4)

  • Non-academic hospital

 182 (27.7) 505 (21.0) 142 (31.8) 394 (22.5)

  • Academic hospital

 117 (17.8) 598 (24.9) 88 (19.7) 441 (25.2)
Back pain due to axSpA* ≥3 months, N (%) 658 2406 (100.0) 2406 2406 (100.0) N/A 447 447 (100.0) 1753 1753 (100.0) N/A
Inflammatory back pain, N (%) 649 454 (70.0) 2358 1900 (80.6) <0.001 444 337 (75.9) 1719 1457 (84.8) <0.001
Sacroiliitis ever*

  • Clinical assessment, N (%)

 638 422 (66.1) 2304 1409 (61.2) 0.02 436 313 (71.8) 1697 1138 (67.1) 0.07

  • Radiographical assessment, N (%)

 640 121 (18.9) 2316 551 (23.8) 0.01 437 89 (20.4) 1700 481 (28.3) 0.001

  • MRI assessment (inflammation), N (%)

 638 368 (57.7) 2304 979 (42.5) <0.001 436 305 (70.0) 1697 884 (52.1) <0.001
Spine involvement ever*

  • Clinical opinion, N (%)

 638 385 (60.3) 2307 1467 (63.6) 0.15 436 249 (57.1) 1699 1099 (64.7) 0.004

  • Radiographical assessment, N (%)

 638 41 (6.4) 2307 267 (11.6) <0.001 436 25 (5.7) 1699 218 (12.8) <0.001

  • MRI assessment (inflammation), N (%)

 638 196 (30.7) 2307 592 (25.7) 0.01 436 130 (29.8) 1699 478 (28.1) 0.53
ASAS classification criteria, N (%) 585 447 (76.4) 2000 1753 (87.6) <0.001 447 1753 0.002

  • Only clinical arm

 N/A N/A 90 (20.1) 485 (27.7)

  • Only imaging arm

 N/A N/A 164 (36.7) 531 (30.3)

  • Clinical+imaging arm

 N/A N/A 193 (43.2) 737 (42.0)
Ever peripheral arthritis, N (%) 654 251 (38.4) 2366 1023 (43.2) 0.03 446 149 (33.4) 1730 671 (38.8) 0.04
Ever enthesitis, N (%) 656 421 (64.2) 2389 1591 (66.6) 0.27 446 275 (61.7) 1744 1155 (66.2) 0.08
Ever uveitis, N (%) 525 46 (8.8) 1904 337 (17.7) <0.001 355 38 (10.7) 1381 278 (20.1) <0.001
Ever psoriasis, N (%) 502 54 (10.8) 1828 201 (11.0) 0.94 338 30 (8.9) 1328 134 (10.1) 0.57
Ever inflammatory bowel disease, N (%) 482 46 (9.5) 1783 170 (9.5) 1.00 330 26 (7.9) 1299 105 (8.1) 0.99
Physician global disease activity 633 3.9 (2.3) 2286 3.4 (2.2) <0.001 428 3.9 (2.3) 1674 3.4 (2.2) <0.001
Patient global disease activity 519 5.0 (2.7) 1872 4.9 (2.8) 0.32 342 5.0 (2.7) 1381 4.9 (2.8) 0.71
BASDAI 471 4.6 (2.2) 1713 4.6 (2.3) 0.73 313 4.5 (2.2) 1274 4.5 (2.3) 0.99
ASDAS 451 2.8 (1.0) 1576 2.8 (1.0) 0.13 304 2.8 (1.1) 1171 2.8 (1.1) 0.52
Elevated CRP, N (%) 604 209 (34.6) 2134 636 (29.8) 0.03 413 143 (34.6) 1564 515 (32.9) 0.55
CRP (mg/L), median (IQR) 609 5.0 (1.5–10) 2136 3.6 (1.0–8.0) 0.01 415 4.6 (1.5–9.0) 1565 4.0 (1.2–9.0) 0.30
BASFI 465 2.9 (2.5) 1679 3.0 (2.5) 0.64 307 2.8 (2.5) 1244 3.0 (2.5) 0.33
BASMI 585 1.4 (1.4) 2116 1.9 (1.8) <0.001 402 1.3 (1.4) 1589 1.9 (1.9) <0.001
EQ-5D 458 62.5 (20.9) 1647 63.9 (21.7) 0.11 301 64.2 (20.6) 1223 64.1 (22.1) 0.68
SF-12, physical component summary score 439 38.6 (10.5) 1586 39.2 (10.3) 0.27 287 39.6 (10.6) 1181 39.7 (10.4) 0.93
SF-12, mental component summary score 439 43.9 (11.1) 1586 43.8 (11.4) 0.99 287 43.5 (11.3) 1181 43.7 (11.5) 0.79
Non-steroidal antirheumatic drugs, N (%) 589 532 (90.3) 2099 1800 (85.8) 0.01 408 366 (89.7) 1574 1350 (85.8) 0.05
Conventional synthetic DMARDs, N (%) 658 74 (11.2) 2403 289 (12.0) 0.63 447 40 (8.9) 1751 176 (10.1) 0.54
Tumour necrosis factor inhibitors, N (%) 656 163 (24.8) 2401 855 (35.6) <0.001 447 106 (23.7) 1749 567 (32.4) <0.001
Interleukin-17 inhibitors, N (%) 658 3 (0.5) 2405 43 (1.8) 0.02 447 2 (0.4) 1752 19 (1.1) 0.34
Open in a new tab

Except where indicated otherwise, values represent the mean and SD.

*Information provided by the local rheumatologist with unknown total number of patients with imaging performed.

ASAS, Assessment in SpondyloArthritis international Society; ASDAS, Ankylosing Spondylitis Disease Activity Score; axSpA, axial spondyloarthritis; BASDAI, Bath Ankylosing Spondylitis Disease Activity Index; BASFI, Bath Ankylosing Spondylitis Functional Index; BASMI, Bath Ankylosing Spondylitis Metrology Index; CRP, C reactive protein; DMARDs, disease-modifying antirheumatic drugs; EQ-5D, European Quality of Life 5-domains Questionnaire; HLA-B27, human leucocyte antigen-B27; N/A, not applicable; SCQM, Swiss Clinical Quality Management; SF-12, Short Form Questionnaire with 12 questions.

Figure 3.

Figure 3

Open in a new tab

Years from start of axSpA-related axial symptoms to inclusion in SCQM shown for individual patients with axSpA fulfilling the ASAS classification criteria. (A) Early axSpA (≤2 years by definition); (B) established axSpA (>2 years by definition). Time point of inclusion in SCQM=0. ASAS, Assessment in SpondyloArthritis international Society; axSpA, axial spondyloarthritis; SCQM, Swiss Clinical Quality Management.

Supplementary data

rmdopen-2023-003455supp001.pdf (93KB, pdf)

Further, demographic, clinical and imaging characteristics of patients with axSpA with early disease are compared with those of established disease at inclusion for all patients with axSpA and those fulfilling the ASAS classification criteria in table 1. No significant differences could be observed with regard to sex, family history of axSpA or educational level. A trend for a lower proportion of patients being HLA-B27 positive was found in early axSpA, not reaching statistical significance. A lower proportion of patients with early axSpA compared with patients with established axSpA was recruited in academic hospitals, corresponding to the referral bias observed in tertiary institutions. Around 50% of patients with early and established axSpA were included in private rheumatology practices. Peripheral arthritis and uveitis were more often recorded in established disease, while enthesitis, psoriasis and inflammatory bowel disease were comparably distributed between the groups. Regarding axial disease, the majority of patients presented with IBP, although their proportion was higher in established disease (80.6% vs 70.0% in early disease, p<0.001). Inflammatory sacroiliac MRI involvement was more often reported in early disease, while axial radiographical involvement was more frequently registered in established disease. Rheumatologists indicated the presence of sacroiliac radiographical involvement in around 20% of patients with early disease. Mirroring higher radiographical involvement in later disease, spinal mobility as assessed by the Bath Ankylosing Spondylitis Mobility Index (BASMI) was lower in early versus established axSpA. In contrast, disease activity (BASDAI, ASDAS), as well as impairments in function and in health-related quality of life, were comparable in early versus established axSpA (table 1). Ninety per cent of patients with early axSpA were treated with NSAIDs at inclusion, while 25% of patients were already on TNFi treatment. The proportion of patients with established axSpA on current TNFi use was 36%. Only a few patients were on interleukin-17 inhibitors at inclusion in SCQM.

Drug retention analyses

The selection of patients for the retention analyses of a first TNFi is depicted in figure 1B. A total of 1124 patients with axSpA started their first TNFi after inclusion in SCQM and had available data on the start of axial symptoms (250 patients with axial symptom duration ≤2 years at TNFi start (early axSpA; 22%) and 874 patients with longer back pain duration at treatment start (established axSpA)). The baseline characteristics of these patients are shown in table 2. The differences between the two groups were comparable with those at inclusion in SCQM. As expected, patients initiating TNFi treatment had, in comparison with the population at inclusion, higher disease activity levels and a more severe impairment of function and quality of life. Median TNFi retention was slightly shorter in early axSpA (2.0 years; 95% CI 1.4 to 2.8) compared with established axSpA (2.3 years; 95% CI 2.1 to 2.8, log-rank test p=0.04). The reasons for discontinuation included adverse events (15.1%), insufficient response (40.7%), remission (4.0%) and other reasons (15.5) with no differences found between early and established axSpA (online supplemental table 1). The baseline characteristics of patients starting a first TNFi in early versus later disease were confirmed in the population fulfilling the ASAS classification criteria for axSpA (table 2).

Table 2.

Characteristics of patients with axSpA with early disease and established disease at start of first TNF inhibitor

Parameter All patients diagnosed as having axSpA
(main analysis; N=1124)		 Patients fulfilling the ASAS classification criteria
(sensitivity analysis; N=816)
N Early disease
(≤2 years)
N=250		 N Established disease
(>2 years)
N=874		 P value N Early disease
(≤2 years)
N=160		 N Established disease
(>2 years)
N=656		 P value
Male sex, N (%) 250 127 (50.8) 874 447 (51.1) 0.98 160 86 (53.8) 656 362 (55.2) 0.81
Age, years 250 38.4 (12.9) 874 43.6 (12.3) <0.001 160 33.3 (9.7) 656 40.9 (11.5) <0.001
Symptom duration, years, median (IQR) 248 1.4 (0.8–2.4) 867 10.8 (5.9–19.7) <0.001 159 1.4 (0.8–2.6) 651 11.2 (6.1–20.3) <0.001
Axial symptom duration, years, median (IQR) 250 0.8 (0.5–1.3) 874 10.2 (5.3–19.0) <0.001 160 0.8 (0.5–1.3) 656 10.4 (5.5–19.3) <0.001
Time since diagnosis, years, median (IQR) 249 0.4 (0.2–0.9) 864 2.6 (0.5–8.9) <0.001 160 0.3 (0.2–0.8) 649 2.8 (0.5–9.3) <0.001
HLA-B27 positive, N (%) 223 123 (55.2) 804 533 (66.3) 0.003 140 97 (69.3) 613 467 (76.2) 0.11
Family history axSpA, N (%) 231 42 (18.2) 812 185 (22.8) 0.16 151 30 (19.9) 616 159 (25.8) 0.16
Body mass index 224 25.1 (4.5) 807 25.8 (4.6) 0.02 142 24.8 (4.7) 614 25.6 (4.6) 0.02
Current smoking 230 77 (33.5) 804 266 (33.1) 0.32 146 52 (35.6) 602 210 (34.9) 0.04
Education 203 721 0.39 133 0.83

  • Compulsory

 31 (15.8) 124 (17.2) 22 (16.5) 87 (16.0)

  • Vocational

 124 (61.1) 402 (55.8) 77 (57.9) 303 (55.8)

  • Academic

 47 (23.2) 195 (27.0) 34 (25.6) 153 (28.2)
Back pain due to axSpA* ≥3 months, N (%) 250 250 (100.0) 874 874 (100.0) N/A 160 160 (100.0) 656 656 (100.0) N/A
Inflammatory back pain, N (%) 236 173 (73.3) 816 689 (84.4) <0.001 160 122 (76.2) 652 572 (87.7) <0.001
Sacroiliitis ever*

  • Clinical assessment, N (%)

 226 149 (65.9) 766 522 (68.1) 0.59 151 112 (74.2) 615 435 (70.7) 0.46

  • Radiographical assessment, N (%)

 229 40 (17.5) 773 203 (26.3) 0.01 154 31 (20.1) 621 181 (29.1) 0.03

  • MRI assessment (inflammation), N (%)

 226 132 (58.4) 766 345 (45.0) 0.001 151 108 (71.5) 615 317 (51.5) <0.001
Spine involvement ever*

  • Clinical opinion, N (%)

 227 148 (65.2) 770 533 (69.2) 0.29 152 99 (65.1) 619 434 (70.1) 0.27

  • Radiographical assessment, N (%)

 227 13 (5.7) 770 104 (13.5) 0.002 152 8 (5.3) 619 89 (14.4) 0.004

  • MRI assessment, N (%)

 227 80 (35.2) 770 222 (28.8) 0.08 152 48 (31.6) 619 186 (30.0) 0.79
ASAS classification criteria, N (%) 212 160 (75.5) 743 656 (88.3) <0.001 160 656 0.01

  • Only clinical arm

 N/A N/A 27 (16.9) 145 (22.1)

  • Only imaging arm

 N/A N/A 68 (42.5) 200 (30.5)

  • Clinical+imaging arm

 N/A N/A 65 (40.6) 311 (47.4)
Ever peripheral arthritis, N (%) 238 103 (43.3) 815 352 (43.2) 1.00 160 67 (41.9) 652 259 (39.7) 0.68
Ever enthesitis, N (%) 236 163 (69.1) 804 577 (71.8) 0.47 158 109 (69.0) 644 453 (70.3) 0.81
Ever uveitis, N (%) 190 18 (9.5) 642 101 (15.7) 0.04 125 14 (11.2) 517 87 (16.8) 0.16
Ever psoriasis, N (%) 189 23 (12.2) 624 73 (11.7) 0.96 124 13 (10.5) 502 52 (10.4) 1.00
Ever inflammatory bowel disease, N (%) 181 17 (9.4) 614 55 (9.0) 0.98 121 12 (9.9) 497 38 (7.6) 0.53
Physician global disease activity 214 5.0 (1.9) 730 4.7 (1.8) 0.06 145 5.1 (2.0) 581 4.7 (1.9) 0.08
Patient global disease activity 180 6.1 (2.3) 627 6.1 (2.4) 0.48 122 6.0 (2.3) 498 6.1 (2.5) 0.31
BASDAI 172 5.6 (2.0) 612 5.4 (2.0) 0.28 118 5.6 (2.0) 483 5.3 (2.0) 0.11
ASDAS 157 3.3 (0.9) 549 3.3 (0.9) 0.48 109 3.3 (0.9) 440 3.3 (0.9) 0.25
ASDAS ≥2.1 157 143 (91.1) 549 498 (90.7) 1.00 109 99 (90.8) 440 399 (90.7) 1.00
Elevated CRP, N (%) 208 113 (54.3) 703 391 (55.6) 0.80 141 69 (48.9) 561 264 (47.1) 0.76
CRP (mg/L), median (IQR) 209 5.9 (2.0–14.0) 704 6.0 (2.0–13.0) 0.87 141 6.0 (2.0–14.0) 562 6.5 (2.0–14.0) 0.94
BASFI 170 3.7 (2.4) 598 3.7 (2.4) 0.88 118 3.8 (2.6) 474 3.6 (2.4) 0.57
BASMI 186 1.4 (1.3) 657 2.1 (1.9) <0.001 128 1.3 (1.3) 524 2.0 (1.9) <0.001
EQ-5D 167 56.2 (20.6) 586 58.6 (20.7) 0.13 116 56.7 (20.8) 465 59.5 (20.8) 0.16
SF-12, physical component summary score 153 35.4 (9.4) 543 36.1 (9.1) 0.41 107 36.0 (9.7) 434 36.8 (9.2) 0.37
SF-12, mental component summary score 153 40.9 (10.0) 543 42.3 (11.1) 0.19 107 40.2 (9.7) 434 42.2 (11.2) 0.10
Non-steroidal antirheumatic drugs, N (%) 150 144 (96.0) 515 488 (94.8) 0.69 98 96 (98.0) 411 391 (95.1) 0.34
Conventional synthetic DMARDs, N (%) 250 34 (13.6) 873 108 (12.4) 0.68 160 21 (13.1) 655 67 (10.2) 0.36
Open in a new tab

Except where indicated otherwise, values represent the mean and SD.

*Information provided by the local rheumatologist with unknown total number of patients with imaging performed.

ASAS, Assessment of SpondyloArthritis international Society; ASDAS, Ankylosing Spondylitis Disease Activity Score; axSpA, axial spondyloarthritis; BASDAI, Bath Ankylosing Spondylitis Disease Activity Index; BASFI, Bath Ankylosing Spondylitis Functional Index; BASMI, Bath Ankylosing Spondylitis Metrology Index; CRP, C reactive protein; DMARDs, disease-modifying antirheumatic drugs; EQ-5D, European Quality of Life 5-domains Questionnaire; HLA-B27, human leucocyte antigen-B27; N/A, not applicable; SF-12, Short Form Questionnaire with 12 questions; TNF, tumour necrosis factor.

The estimated unadjusted HR to discontinue a first TNFi was slightly higher in early versus later axSpA (1.22, 95% CI 1.03 to 1.44). Baseline characteristics of patients with complete data in the adjusted analyses were comparable to all patients included in the retention analysis (table 3). The difference in retention between early and established disease lost significance in the adjusted model 1 (HR 1.07, 95% CI 0.87 to 1.31, table 4). TNFi retention in the two groups further aligned with each other after additional adjustment for ASDAS or, alternatively, for BMI, current smoking, elevated CRP and inflammatory sacroiliac changes on MRI (adjusted models 2 and 3 in table 4). Female sex and HLA-B27 negativity were associated with a higher risk of TNFi discontinuation in these models, while a higher ASDAS or an elevated CRP was associated with longer drug retention (table 4). We found no interaction between sex and early versus established disease (online supplemental table 2), indicating that the differences in TNFi retention between women and men were comparable in early and later axSpA.

Table 3.

Characteristics of patients with axSpA with early disease and established disease at start of first TNF inhibitor in the main adjusted retention analysis (model 1; patients without missing covariate data)

Parameter All patients diagnosed as having axSpA
(main analysis; N=843)		 Patients fulfilling the ASAS classification criteria
(sensitivity analysis; N=816)
N Early disease
(≤2 years)
N=178		 N Established disease
(>2 years)
N=665		 P value N Early disease
(≤2 years)
N=114		 N Established disease
(>2 years)
N=509		 P value
Male sex, N (%) 178 95 (53.4) 665 333 (50.1) 0.49 114 64 (56.1) 509 275 (54.0) 0.76
Age, years 178 37.6 (13.0) 665 43.7 (12.3) <0.001 114 32.4 (9.4) 509 41.5 (11.5) <0.001
Symptom duration, years, median (IQR) 176 1.4 (0.8–2.6) 661 10.9 (5.9–19.8) <0.001 113 1.5 (0.8–2.6) 506 11.2 (6.1–20.7) <0.001
Axial symptom duration, years, median (IQR) 178 0.8 (0.5–1.2) 665 10.2 (5.2–19.2) <0.001 114 0.8 (0.5–1.2) 509 10.7 (5.6–19.9) <0.001
Time since diagnosis, years, median (IQR) 177 0.4 (0.2–1.0) 658 2.5 (0.5–8.9) <0.001 114 0.3 (0.2–0.9) 503 2.6 (0.5–9.4) <0.001
HLA-B27 positive, N (%) 178 104 (58.4) 665 437 (65.7) 0.09 114 82 (71.9) 509 384 (75.4) 0.51
Family history axSpA, N (%) 168 33 (19.6) 623 148 (23.8) 0.31 110 24 (21.8) 480 130 (27.1) 0.31
Body mass index 133 24.9 (4.4) 630 25.7 (4.5) 0.02 109 24.3 (4.4) 488 25.5 (4.4) 0.003
Current smoking, N (%) 177 63 (35.6) 655 227 (34.7) 0.79 113 43 (38.1) 502 184 (36.7) 0.36
Education 178 665 0.36 114 509 0.88

  • Compulsory

 26 (14.6) 112 (16.8 17 (14.9) 80 (15.7)

  • Vocational

 110 (61.8) 371 (55.8) 67 (58.8) 286 (56.2)

  • Academic

 42 (23.6) 182 (27.4) 30 (26.3) 143 (28.1)
Back pain due to axSpA* ≥3 months, N (%) 178 178 (100.0) 665 665 (100.0) N/A 114 114 (100.0) 509 509 (100.0) N/A
Inflammatory back pain, N (%) 164 122 (74.4) 630 535 (84.9) 0.002 114 87 (76.3) 507 449 (88.6) 0.001
Sacroiliitis ever*

  • Clinical assessment, N (%)

 159 109 (68.6) 593 417 (70.3) 0.74 109 83 (76.1) 482 351 (72.8) 0.56

  • Radiographical assessment, N (%)

 161 31 (19.3) 597 163 (27.3) 0.048 111 25 (22.5) 485 145 (29.9) 0.15

  • MRI assessment (inflammation), N (%)

 159 91 (57.2) 593 266 (44.9) 0.01 109 63 (67.0) 482 244 (50.6) 0.003
Spine involvement ever*

  • Clinical opinion, N (%)

 159 109 (68.6) 596 432 (72.5) 0.38 109 76 (69.7) 485 353 (72.8) 0.60

  • Radiographical assessment, N (%)

 159 10 (6.3) 596 86 (14.4) 0.01 109 6 (5.5) 485 75 (15.5) 0.01

  • MRI assessment, N (%)

 159 47 (29.6) 596 163 (27.3) 0.65 109 28 (25.7) 485 136 (28.0) 0.71
Ever peripheral arthritis, N (%) 166 75 (45.2) 627 280 (44.7) 0.97 114 52 (45.6) 506 208 (41.1) 0.48
Ever enthesitis, N (%) 165 110 (66.7) 619 446 (72.1) 0.21 113 76 (67.3) 500 357 (71.4) 0.45
Ever uveitis, N (%) 129 16 (12.4) 480 74 (15.4) 0.47 86 12 (14.0) 388 66 (17.0) 0.60
Ever psoriasis, N (%) 130 19 (14.6) 471 56 (11.9) 0.50 87 11 (12.6) 379 43 (11.3) 0.88
Ever inflammatory bowel disease, N (%) 123 12 (9.8) 466 42 (9.0) 0.94 84 9 (10.7) 375 31 (8.3) 0.61
Physician global disease activity 152 5.0 (1.9) 556 4.6 (1.8) 0.01 105 5.1 (2.0) 446 4.6 (1.8) 0.04
Patient global disease activity 142 6.0 (2.3) 535 6.1 (2.4) 0.35 97 5.9 (2.3) 427 6.1 (2.4) 0.31
BASDAI 138 5.5 (2.0) 529 5.4 (2.0) 0.87 95 5.5 (2.0) 420 5.3 (2.0) 0.41
ASDAS 130 3.3 (0.9) 484 3.3 (0.9) 0.70 90 3.3 (0.9) 392 3.2 (0.9) 0.39
ASDAS ≥2.1 130 119 (91.5) 484 440 (90.9) 0.96 90 82 (91.1) 392 355 (90.6) 1.00
Elevated CRP, N (%) 145 69 (47.6) 541 236 (43.6) 0.45 101 51 (50.5) 437 202 (46.2) 0.51
CRP (mg/L), median (IQR) 146 5.9 (2.9–16.0) 541 6.0 (2.0–13.0) 0.78 101 6.0 (2.0–18.0) 437 6.5 (2.0–14.0) 0.69
BASFI 138 3.7 (2.4) 521 3.7 (2.4) 0.97 95 3.7 (2.5) 416 3.6 (2.4) 0.73
BASMI 136 1.3 (1.3) 509 2.1 (1.9) <0.001 96 1.2 (1.3) 410 2.0 (1.9) <0.001
EQ-5D 138 56.7 (20.3) 511 59.1 (20.5) 0.10 95 57.2 (20.6) 408 60.4 (20.5) 0.09
SF-12, physical component summary score 131 35.6 (9.4) 477 36.2 (9.2) 0.49 91 36.3 (9.6) 386 37.0 (9.3) 0.47
SF-12, mental component summary score 131 40.7 (9.8) 477 42.3 (11.2) 0.15 91 39.8 (9.7) 386 42.5 (11.1) 0.03
Non-steroidal antirheumatic drugs, N (%) 114 110 (96.5) 396 377 (95.2) 0.74 75 74 (98.7) 321 306 (95.3) 0.32
Conventional synthetic DMARDs, N (%) 178 26 (14.6) 664 83 (12.5) 0.54 114 16 (14.0) 509 53 (10.4) 0.35
Open in a new tab

Except where indicated otherwise, values represent the mean and SD.

*Information provided by the local rheumatologist with unknown total number of patients with imaging performed.

ASAS, Assessment of SpondyloArthritis international Society; ASDAS, Ankylosing Spondylitis Disease Activity Score; axSpA, axial spondyloarthritis; BASDAI, Bath Ankylosing Spondylitis Disease Activity Index; BASFI, Bath Ankylosing Spondylitis Functional Index; BASMI, Bath Ankylosing Spondylitis Metrology Index; CRP, C reactive protein; DMARDs, disease-modifying antirheumatic drugs; EQ-5D, European Quality of Life 5-domains Questionnaire; HLA-B27, human leucocyte antigen-B27; N/A, not applicable; SF-12, Short Form Questionnaire with 12 questions; TNF, tumour necrosis factor.

Table 4.

Multiple adjusted Cox proportional hazards model for analysis of drug discontinuation of a first TNF inhibitor in early versus established axSpA

Population Variable Unadjusted analysis Adjusted model 1 Adjusted model 2 Adjusted model 3
HR 95% CI P value HR 95% CI P value HR 95% CI P value HR 95% CI P value
Main analysis:
all patients diagnosed
as having axSpA		 Early vs established disease 1.22 1.03; 1.44 0.02 1.07 0.87; 1.31 0.51 1.00 0.79; 1.28 0.98 1.01 0.79; 1.29 0.91
Age 1.00 0.99; 1.00 0.46 1.00 0.99; 1.00 0.41 0.99 0.98; 1.00 0.06
Female sex 1.56 1.31; 1.85 <0.001 1.51 1.24; 1.85 <0.001 1.51 1.22; 1.85 <0.001
HLA-B27 negativity 1.39 1.16; 1.67 <0.001 1.40 1.13; 1.73 0.002 1.25 1.01; 1.54 0.04
Education vocational 0.83 0.66; 1.04 0.11 0.77 0.58; 1.01 0.06 0.83 0.64; 1.08 0.17
Education academic 0.77 0.60; 1.00 0.048 0.82 0.61; 1.12 0.21 0.97 0.72; 1.30 0.81
ASDAS 0.82 0.74; 0.92 <0.001
Body mass index 1.03 1.00; 1.05 0.03
Current smoking 1.11 0.91; 1.37 0.30
Elevated CRP 0.58 0.47; 0.71 <0.001
Sacroiliitis on MRI 0.85 0.70; 1.03 0.10
(Number of patients/events) (1124/735) (843/573) (614/424) (619/436)
Sensitivity analysis:
patients fulfilling the
ASAS classification
criteria		 Early vs established disease 1.31 1.06; 1.61 0.01 1.23 0.95; 1.59 0.12 1.12 0.83; 1.52 0.44 1.05 0.78; 1.42 0.73
Age 1.00 0.99; 1.01 0.61 1.00 0.99; 1.01 0.41 0.99 0.98; 1.00 0.04
Female sex 1.53 1.25; 1.87 <0.001 1.49 1.18; 1.87 <0.001 1.45 1.15; 1.83 0.002
HLA-B27 negativity 1.40 1.12; 1.75 0.004 1.33 1.03; 1.72 0.03 1.23 0.94; 1.60 0.13
Education vocational 0.85 0.64; 1.11 0.23 0.76 0.55; 1.05 0.10 0.79 0.58; 1.08 0.14
Education academic 0.93 0.69; 1.25 0.62 0.91 0.64; 1.30 0.61 1.02 0.72; 1.43 0.93
ASDAS 0.82 0.73; 0.93 0.002
Body mass index 1.02 0.99; 1.05 0.12
Current smoking 1.06 0.84; 1.33 0.62
Elevated CRP 0.54 0.43; 0.68 <0.001
Sacroiliitis on MRI 0.89 0.71; 1.12 0.33
(Number of patients/events) (816/526) (623/419) (482/328) (488/340)
Open in a new tab

Statistically significant results are shown in bold. The number of patients assessed and the number of treatment discontinuations are indicated for each statistical model at the bottom.

ASAS, Assessment of SpondyloArthritis international Society; ASDAS, Ankylosing Spondylitis Disease Activity Score; axSpA, axial spondyloarthritis; CRP, C reactive protein; HLA-B27, human leucocyte antigen-B27; TNF, tumour necrosis factor.

Treatment response analyses

Response rates at 1 year of treatment with a first TNFi were available in 577 patients for ASAS40 and 570 patients for BASDAI50. We found numerically slightly lower response rates in early versus established disease in unadjusted analyses: 34.4% vs 36.1% for ASAS40 (OR 0.93, 95% CI 0.61 to 1.40) and 32.5% vs 35.4% for BASDAI50 (OR 0.88, 95% CI 0.58 to 1.34). Statistical models adjusted for potential confounders and additional explanatory variables are presented in table 5. No significant differences could be detected between patients with early and established disease for ASAS40 and BASDAI50 responses in these analyses. Male sex, HLA-B27 positivity, higher education and elevated CRP were consistently associated with significantly better response rates in these models (table 5). There was only a trend for inflammatory sacroiliac changes being associated with a better BASDAI50 response, which did not reach statistical significance and was not found for the ASAS40 response.

Table 5.

Multiple-adjusted response rate analyses at 1 year of treatment with a first TNFi for different outcomes in early versus established axSpA

Outcome Variable All patients diagnosed as having axSpA Patients fulfilling the ASAS classification criteria
Adjusted model 1 Adjusted model 2 Adjusted model 1 Adjusted model 2
OR 95% CI P value OR 95% CI P value OR 95% CI P value OR 95% CI P value
ASAS40 Early vs established disease 1.09 0.67; 1.78 0.72 1.04 0.60; 1.77 0.89 0.80 0.45; 1.42 0.45 0.74 0.39; 1.36 0.33
Age 0.99 0.97; 1.00 0.16 0.99 0.98; 1.01 0.44 0.98 0.96; 1.00 0.054 0.99 0.97; 1.01 0.32
Female sex 0.66 0.44; 0.98 0.04 0.60 0.38; 0.93 0.02 0.73 0.47; 1.14 0.17 0.72 0.44; 1.17 0.19
HLA-B27 negativity 0.55 0.35; 0.84 0.01 0.60 0.37; 0.97 0.04 0.41 0.23; 0.70 0.002 0.49 0.27; 0.88 0.02
Education vocational 2.00 1.08; 3.87 0.03 2.24 1.15; 4.62 0.02 2.22 1.09; 4.82 0.03 2.18 1.04; 4.87 0.047
Education academic 2.35 1.22; 4.73 0.01 2.75 1.35; 5.92 0.01 2.35 1.11; 5.31 0.03 2.29 1.03; 5.35 0.047
Body mass index 0.95 0.90; 1.00 0.06 0.94 0.89; 1.00 0.046
Current smoking 0.91 0.57; 1.43 0.67 1.04 0.63; 1.69 0.88
Elevated CRP 2.10 1.37; 3.24 <0.001 2.26 1.41; 3.64 <0.001
Sacroiliitis on MRI 1.23 0.80; 1.89 0.36 1.20 0.73; 1.96 0.47
(Number of patients/events) (489/172) (433/153) (388/141) (348/130)
BASDAI50 Early vs established disease 0.89 0.53; 1.47 0.73 0.73 0.41; 1.28 0.28 0.71 0.39; 1.26 0.25 0.60 0.31; 1.14 0.13
Age 0.98 0.53; 1.00 0.058 0.99 0.97; 1.01 0.33 0.98 0.96; 1.00 0.03 0.99 0.97; 1.01 0.40
Female sex 0.47 0.31; 0.71 <0.001 0.45 0.28; 0.72 <0.001 0.54 0.34; 0.84 0.01 0.50 0.29; 0.82 0.01
HLA-B27 negativity 0.40 0.25; 0.64 <0.001 0.52 0.31; 0.87 0.02 0.40 0.22; 0.71 0.002 0.48 0.25; 0.89 0.02
Education vocational 2.27 1.18; 4.62 0.02 2.59 1.28; 5.59 0.01 2.70 1.30; 6.08 0.01 2.86 1.31; 6.71 0.01
Education academic 2.82 1.41; 5.95 0.01 3.23 1.52; 7.25 0.003 3.17 1.46; 7.40 0.01 3.30 1.44; 8.07 0.01
Body mass index 0.95 0.90; 1.00 0.06 0.92 0.87; 0.98 0.02
Current smoking 0.82 0.51; 1.33 0.42 0.96 0.57; 1.58 0.86
Elevated CRP 3.22 2.06; 5.10 <0.001 3.04 1.87; 4.99 <0.001
Sacroiliitis on MRI 1.46 0.93; 2.31 0.11 1.44 0.87; 2.41 0.16
(Number of patients/events) (485/164) (429/149) (386/144) (347/132)
Open in a new tab

Response rates in patients with available outcome at 1 year (±6 months) and patients having discontinued the biologic in the meantime being considered non-responders. Significant results are shown in bold.

ASAS, Assessment in SpondyloArthritis international Society; ASAS40, 40% improvement according to the ASAS criteria; axSpA, axial spondyloarthritis; BASDAI50, 50% improvement in the Bath Ankylosing Spondylitis Disease Activity Index; CRP, C reactive protein; HLA-B27, human leucocyte antigen-B27; TNFi, tumour necrosis factor inhibitor.

The results found in the whole population diagnosed with axSpA were confirmed in patients fulfilling the ASAS classification criteria (table 5). A higher BMI was associated with a significantly reduced treatment response according to BASDAI50 and ASAS40 criteria in this population.

Discussion

This first analysis of patients with early axSpA according to the ASAS definition6 in a large observational cohort revealed several important aspects. First, up to 20% of patients recruited over the past two decades had axial symptom duration of ≤2 years and fulfilled the new definition for early axSpA. Second, patients with early axSpA were very comparable with patients with longer axial symptom duration regarding important disease characteristics, with the exception of factors affected by time (such as age, radiographical damage and impairment of spinal mobility). Finally, the effectiveness of TNFi, assessed through the evaluation of their retention as well as ASAS40 and BASDAI50 response rates, was comparable in early and established axSpA.

The fact that 20% of patients recruited to SCQM fulfilled the definition of early axSpA is, on the one hand, reassuring. Although the median diagnostic delay in axSpA is still long,7 a relevant proportion of patients in real-life clinical practice are diagnosed in this early disease stage, guaranteeing the feasibility of future studies using this definition, preconditioned that our results are confirmed in other healthcare systems. On the other hand, we could not identify a trend for earlier recognition of axSpA over the years, at least with regard to the proportion of patients identified within 2 years after the onset of axial symptoms. This highlights the need to intensify the already considerable international and national efforts to improve disease recognition.15 The fact that the first symptoms were not axial symptoms in a significant number of patients fulfilling the ASAS consensus definition of ‘early axSpA’ indicates that the presence of peripheral or extramusculoskeletal manifestations might lead to an earlier recognition of axial disease. Indeed, a recent study found a high prevalence of both overall and previously undiagnosed SpA in patients with acute anterior uveitis.16

Important disease characteristics were comparable between the early and later stages of the disease: sex distribution, proportions of HLA-B27 positivity and of a positive family history of axSpA, markers of disease activity, function and health-related quality of life. Radiographical axial involvement and impairment of spinal mobility as assessed by the BASMI were, as expected, more prominent in established axSpA. Only a few earlier cohorts provide some comparison for the characteristics of our 658 patients with early axSpA at inclusion in SCQM. Inclusion criteria for a small cohort (N=68) from the Netherlands (ESpAC) comprised the presence of IBP of ≤2 years’ duration with onset of back pain before the age of 40 years and persistence for at least 3 months.17 18 Importantly, diagnosis of axSpA before inclusion was not strictly required, and the presence of additional spondyloarthritis features was preferred but not mandatory. Presumably as a consequence of these specific inclusion criteria, only 38% of patients were of male sex, only 46% were HLA-B27 positive and only 35% had inflammatory sacroiliac MRI changes. These features stand in some contrast with our findings, showing a well-balanced sex distribution and considerably higher proportions of patients with HLA-B27 positivity and MRI sacroiliac inflammation (60% and 58%, respectively). Our patients’ characteristics at inclusion in SCQM are more in line with findings from the French DESIR cohort,19 which included patients with IBP of ≤3 years’ duration in the context of overall symptoms suggestive of spondyloarthritis: male sex 46%, HLA-B27 positivity 57%. The proportion of patients with radiographical axSpA found in our patients with early axSpA (20%) was comparable with findings in cohorts of short IBP duration (20% in ESpAC and 26% in DESIR).17 19 Inflammatory axial MRI changes were more prevalent in patients with early axSpA from the SCQM registry (58% for the sacroiliac joints and 31% for the spine) than in patients with IBP of short duration from the DESIR cohort (32% and 20%, respectively).19 The higher inflammation load in patients in the SCQM registry might be explained by the fact that patients considered for bDMARD treatment are preferably recruited. The reason for this observation is the circumstance that, according to regulatory authorities, rheumatologists can deduct the costs of bDMARDs from their global treatment expenditures if the patients are followed in the clinical quality management programme that is at the core of SCQM.20

While the number of research questions to be evaluated in an early axSpA disease stage is substantial, the issue of a potential ‘window of opportunity’ for early treatment to allow for better outcomes remains at the forefront.3 21

Our multiple-adjusted retention and response analyses did not demonstrate better TNFi effectiveness in early versus established axSpA. The ASAS40 response rate found here in early axSpA (34%) is lower than in two prospective trials of infliximab in patients with symptom duration ≤3 years (60% and 75%, respectively).22 23 These studies used, however, additional criteria to select their patients. The presence of inflammatory sacroiliac MRI changes was an absolute requirement for both studies, and all patients were additionally HLA-B27 positive in the first study,22 while all patients were cotreated with an NSAID in the second study.23 As both studies did not compare the treatment response in early versus established disease, the issue of additional criteria for adequate patient selection in early disease to be able to provide evidence for a window of opportunity in axSpA is still open. Comparison with these studies is further hampered by the observational nature of our analysis, which constrained us to measure the outcome at a rather late time point (1 year) and to consider patients who had discontinued the TNFi in the meantime as non-responders, regardless of the reason for discontinuation (response/tolerance analysis). Interestingly, our adjusted analyses did not identify the presence of sacroiliac joint inflammation as a predictor of better drug retention or treatment response. In contrast, HLA-B27 positivity was associated with treatment effectiveness here and is known to be an independent predictor of sacroiliac inflammation on MRI.18 We have previously demonstrated that CRP and male sex seem to better describe the variability of treatment responses than HLA-B27 in individual patients.24 An elevated CRP or a higher ASDAS was consistently associated with significantly better TNFi effectiveness in the analyses presented here. Indeed, CRP was shown to be the best predictor of good response in numerous prospective trials in both radiographical and non-radiographical disease and seems better than sacroiliac inflammation detected by MRI, as demonstrated in subgroup analyses.25–29 Moreover, the amount and intensity of MRI inflammation might better predict response than the mere presence of sacroiliac bone marrow oedema.30 Extensive sacroiliac bone marrow oedema is also a strong predictor of the development of structural lesions, in contrast to limited or intermediate inflammatory lesions.31 Male sex was shown to be a predictor of future sacroiliac inflammation detected by MRI in patients with IBP of short duration.18 It is also known to be a predictor of treatment response in both radiographical32 and non-radiographical33 axSpA and to be associated with accelerated radiographical progression at the levels of the sacroiliac joints34 and the spine.35 The impact of sex on treatment response seemed not to be different in early versus established axSpA, as demonstrated by the interaction analyses shown here.

Future analyses of treatment response and radiographical progression in early versus established disease are therefore warranted. Whether additional requirements on the amount of axial or systemic inflammation might help solve the conundrum of a window of opportunity in axSpA will potentially have to be investigated. As spinal radiographical progression is only minimal in the non-radiographical disease state,36 an adequate length of the investigations might be crucial.

The prospective study design in one of the largest national axSpA cohorts treated under real-life conditions using validated assessments and the systematic collection of the start of axial symptoms in addition to the start of first symptoms represent its major strengths. The main analyses were performed on the whole population diagnosed as having axSpA. However, as the definition of early axSpA is intended to be used for research purposes only, we have presented data for the subgroup fulfilling the ASAS classification criteria12 in parallel to further enhance the homogeneity of the study population.

As a major limitation of our analyses, MRIs were not collected systematically in SCQM to allow for the validation of the sacroiliac and spinal involvement indicated by the rheumatologist.37 38 Additional limitations are related to the observational nature of our investigation and the fact that we might not have been able to adjust for unknown remaining confounders. Recall bias with regard to the start of symptoms is a limitation inherent to the consensually chosen definition of early axSpA.6 It is supposed to be more limited within a range of 2 years than with longer symptom duration.

In conclusion, 20% of patients with axSpA in this contemporary real-life axSpA cohort were included in an early disease stage according to the new consensus definition of early axSpA. While important patient characteristics are comparable in early and established axSpA, our results do not suggest better TNFi retention and better response rates in early axSpA in the context of a cut-off of 2 years of axial symptom duration as defined by ASAS. Comparable analyses in patients with shorter symptom duration might represent a next step for future analyses of early axSpA in suitable observational cohorts.

Acknowledgments

We thank all rheumatologists and their patients for participation in SCQM. The entire SCQM staff were instrumental for data management and support. A list of rheumatology practices and hospitals that are contributing to the SCQM registries can be found on: http://www.scqm.ch/institutions.

Footnotes

Twitter: @ramicheroli

Authors’ contributions: AC, AG and RM conceptualised and designed the study. AC, AR-R, BM, KB, MA, MJN, OD, PE, RB, RM and TH substantially contributed to the acquisition of clinical data. AG and AS processed the data and performed the statistical analyses. All authors contributed to the interpretation of the data. AC wrote the article, and all coauthors critically revised the manuscript for important intellectual content. AC had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. All authors agreed on the final content of the submitted manuscript.

Funding: The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

Competing interests: The SCQM Foundation is supported by the Swiss Society of Rheumatology and by AbbVie, AstraZeneca, Eli Lilly, iQone Healthcare, Janssen, Merck Sharp & Dohme, Novartis, Pfizer, Samsung Bioepis and Sandoz. AC received honoraria for lectures from AbbVie and Novartis. AR-R received consulting fees from AbbVie, Janssen and Pfizer; honoraria for lectures from AbbVie, Janssen, Novartis and Pfizer; as well as support for attending meetings from Janssen and Pfizer. AS received consulting fees from Pfizer and support for attending meetings from Gilead. BM received speaking fees from AbbVie, Janssen, Novartis and Pfizer, as well as support for attending meetings from Janssen. DK received a research grant from AbbVie, honoraria for presentations from AbbVie and Eli Lilly, support for attending meetings from Janssen and Eli Lilly, as well as payments for participation in advisory boards from AbbVie, Eli Lilly, Janssen, Novartis, Pfizer and Roche. MJN received consulting and/or speaking fees from AbbVie, Eli Lilly, Janssen, Novartis and Pfizer, as well as a research grant from Novartis. OD received consulting fees from AbbVie. PE received financial support from UCB to attend a meeting. RM received honoraria for lectures or presentations from AbbVie, Eli Lilly, Janssen, Gilead and Pfizer. TH received royalties from Curmed, payments for lectures and presentations from Pfizer, Fresenius Kabi, AbbVie, Merck Sharp & Dohme, Galapagos, Eli Lilly and Novartis. He participated in advisory boards for DETECTRA and holds stock or stock options of Atreon SA and Vtuls. AG, KB, MA and RB declare they have no conflicts of interest.

Patient and public involvement statement: Patients were involved in the reporting and dissemination plans of this research.

Provenance and peer review: Not commissioned; externally peer reviewed.

Supplemental material: This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.

Data availability statement

Data may be obtained from a third party and are not publicly available. Restrictions apply to the availability of data. The data are owned by a third party, the Swiss Clinical Quality Management in Rheumatic Diseases (SCQM) Foundation. Data may be obtained after approval and permission from the license holder (SCQM). Contact information for data requests: scqm@hin.ch.

Ethics statements

Patient consent for publication

Not required.

Ethics approval

The study was approved by the Ethics Committee of the Canton of Zurich (BASEC 2022-0272) and written informed consent was obtained from all patients.

References

  • 1.van Nies JA, Krabben A, Schoones A, et al. What is the evidence for the presence of a therapeutic window of opportunity in rheumatoid arthritis? A systematic literature review. Ann Rheum Dis 2014;73:861–70. 10.1136/annrheumdis-2012-203130 [DOI] [PubMed] [Google Scholar]
  • 2.Burgers LE, Raza K, van der Helm-van Mil AH. Window of opportunity in rheumatoid arthritis - definitions and supporting evidence: from old to new perspectives. RMD Open 2019;5. 10.1136/rmdopen-2018-000870 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Benavent D, Capelusnik D, Ramiro S, et al. Looking through the window of opportunity in axial spondyloarthritis? Joint Bone Spine 2023;90:105603. 10.1016/j.jbspin.2023.105603 [DOI] [PubMed] [Google Scholar]
  • 4.Capelusnik D, Benavent D, van der Heijde D, et al. Treating spondyloarthritis early: does it matter? Results from a systematic literature review. Rheumatology (Oxford) 2023;62:1398–409. 10.1093/rheumatology/keac532 [DOI] [PubMed] [Google Scholar]
  • 5.Benavent D, Capelusnik D, van der Heijde D, et al. How is early spondyloarthritis defined in the literature? Rsults from a systematic review. Semin Arthritis Rheum 2022;55:152032. 10.1016/j.semarthrit.2022.152032 [DOI] [PubMed] [Google Scholar]
  • 6.Navarro-Compán V, Benavent D, Capelusnik D, et al. ASAS consensus definition of early axial spondyloarthritis. Ann Rheum Dis 2023. 10.1136/ard-2023-224232 [DOI] [PubMed] [Google Scholar]
  • 7.Hay CA, Packham J, Ryan S, et al. Diagnostic delay in axial spondyloarthritis: a systematic review. Clin Rheumatol 2022;41:1939–50. 10.1007/s10067-022-06100-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Ciurea A, Kissling S, Bürki K, et al. Current differentiation between radiographic and non-radiographic axial spondyloarthritis is of limited benefit for prediction of important clinical outcomes: data from a large, prospective, observational cohort. RMD Open 2022;8:e002067. 10.1136/rmdopen-2021-002067 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Ciurea A, Weber U, Stekhoven D, et al. Treatment with tumor necrosis factor inhibitors in axial spondyloarthritis: comparison between private rheumatology practices and academic centers in a large observational cohort. J Rheumatol 2015;42:101–5. 10.3899/jrheum.140229 [DOI] [PubMed] [Google Scholar]
  • 10.Sieper J, van der Heijde D, Landewé R, et al. New criteria for inflammatory back pain in patients with chronic back pain: a real patient exercise by experts from the assessment of spondyloarthritis international society (ASAS). Ann Rheum Dis 2009;68:784–8. 10.1136/ard.2008.101501 [DOI] [PubMed] [Google Scholar]
  • 11.Sieper J, Rudwaleit M, Baraliakos X, et al. The assessment of spondyloarthritis international society (ASAS) handbook: a guide to assess spondyloarthritis. Ann Rheum Dis 2009;68 Suppl 2:ii1–44. 10.1136/ard.2008.104018 [DOI] [PubMed] [Google Scholar]
  • 12.Rudwaleit M, van der Heijde D, Landewé R, et al. The development of assessment of spondyloarthritis international society classification criteria for axial spondyloarthritis (part II): validation and final selection. Ann Rheum Dis 2009;68:777–83. 10.1136/ard.2009.108233 [DOI] [PubMed] [Google Scholar]
  • 13.Shaw Y, Courvoisier D, Scherer A, et al. Do mobile apps improve shared decision making and disease management in the rheumatic diseases? An evaluation of apps in a swiss rheumatology registry. Ann Rheum Dis 2019:125–6. 10.1136/annrheumdis-2019-eular.382 [DOI] [Google Scholar]
  • 14.Ciurea A, Exer P, Weber U, et al. Does the reason for discontinuation of a first TNF inhibitor influence the effectiveness of a second TNF inhibitor in axial spondyloarthritis. Arthritis Res Ther 2016;18:71. 10.1186/s13075-016-0969-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Navarro-Compán V, Ermann J, Poddubnyy D. A glance into the future of diagnosis and treatment of spondyloarthritis. Ther Adv Musculoskelet Dis 2022;14:1759720X221111611. 10.1177/1759720X221111611 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Rademacher J, Müllner H, Diekhoff T, et al. Keep an eye on the back: spondyloarthritis in patients with acute anterior uveitis. Arthritis Rheumatol 2023;75:210–9. 10.1002/art.42315 [DOI] [PubMed] [Google Scholar]
  • 17.Heuft-Dorenbosch L, Landewé R, Weijers R, et al. Performance of various criteria SETS in patients with inflammatory back pain of short duration; the maastricht early spondyloarthritis clinic. Ann Rheum Dis 2007;66:92–8. 10.1136/ard.2006.053918 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.van Onna M, Jurik AG, van der Heijde D, et al. HLA-B27 and gender independently determine the likelihood of a positive MRI of the sacroiliac joints in patients with early inflammatory back pain: a 2-year MRI follow-up study. Ann Rheum Dis 2011;70:1981–5. 10.1136/annrheumdis-2011-200025 [DOI] [PubMed] [Google Scholar]
  • 19.Dougados M, d’Agostino M-A, Benessiano J, et al. The DESIR cohort: a 10-year follow-up of early inflammatory back pain in France: study design and baseline characteristics of the 708 recruited patients. Joint Bone Spine 2011;78:598–603. 10.1016/j.jbspin.2011.01.013 [DOI] [PubMed] [Google Scholar]
  • 20.Ciurea A, Scherer A, Exer P, et al. Tumor necrosis factor α inhibition in radiographic and nonradiographic axial spondyloarthritis: results from a large observational cohort. Arthritis Rheum 2013;65:3096–106. 10.1002/art.38140 [DOI] [PubMed] [Google Scholar]
  • 21.Robinson PC, Brown MA. The window of opportunity: a relevant concept for axial spondyloarthritis. Arthritis Res Ther 2014;16:109. 10.1186/ar4561 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Barkham N, Keen HI, Coates LC, et al. Clinical and imaging efficacy of Infliximab in HLA-B27-positive patients with magnetic resonance imaging-determined early sacroiliitis. Arthritis Rheum 2009;60:946–54. 10.1002/art.24408 [DOI] [PubMed] [Google Scholar]
  • 23.Sieper J, Lenaerts J, Wollenhaupt J, et al. Efficacy and safety of Infliximab plus naproxen versus naproxen alone in patients with early, active axial spondyloarthritis: results from the double-blind, placebo-controlled INFAST study, part 1. Ann Rheum Dis 2014;73:101–7. 10.1136/annrheumdis-2012-203201 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Fröhlich F, Micheroli R, Hebeisen M, et al. HLA-B27 as a predictor of effectiveness of treatment with TNF inhibitors in axial spondyloarthritis: data from the swiss clinical quality management registry. Clin Rheumatol 2023;42:1267–74. 10.1007/s10067-022-06490-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Glintborg B, Ostergaard M, Krogh NS, et al. Predictors of treatment response and drug continuation in 842 patients with ankylosing spondylitis treated with anti-tumour necrosis factor: results from 8 years' surveillance in the danish nationwide danbio registry. Ann Rheum Dis 2010;69:2002–8. 10.1136/ard.2009.124446 [DOI] [PubMed] [Google Scholar]
  • 26.Arends S, Brouwer E, van der Veer E, et al. Baseline predictors of response and discontinuation of tumor necrosis factor-alpha blocking therapy in ankylosing spondylitis: a prospective longitudinal observational cohort study. Arthritis Res Ther 2011;13:R94. 10.1186/ar3369 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Sieper J, Heijde D, et al. Efficacy and safety of adalimumab in patients with non-radiographic axial spondyloarthritis. Ann Rheum Dis 2013;72:815–22. 10.1136/annrheumdis-2012-201766 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Dougados M, van der Heijde D, Sieper J, et al. Symptomatic efficacy of etanercept and its effects on objective signs of inflammation in early nonradiographic axial spondyloarthritis: a multicenter, randomized, double-blind, placebo-controlled trial. Arthritis Rheumatol 2014;66:2091–102. 10.1002/art.38721 [DOI] [PubMed] [Google Scholar]
  • 29.Sieper J, van der Heijde D, Dougados M, et al. A randomized, double-blind, placebo-controlled, sixteen-week study of subcutaneous Golimumab in patients with active Nonradiographic axial Spondyloarthritis. Arthritis Rheumatol 2015;67:2702–12. 10.1002/art.39257 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Braun J, Blanco R, Marzo-Ortega H, et al. Secukinumab in non-radiographic axial Spondyloarthritis: subgroup analysis based on key baseline characteristics from a randomized phase III study, PREVENT. Arthritis Res Ther 2021;23:231. 10.1186/s13075-021-02613-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Arnbak B, Jensen TS, Schiøttz-Christensen B, et al. What level of inflammation leads to structural damage in the sacroiliac joints? A four-year magnetic resonance imaging follow-up study of low back pain patients. Arthritis Rheumatol 2019;71:2027–33. 10.1002/art.41040 [DOI] [PubMed] [Google Scholar]
  • 32.Hebeisen M, Neuenschwander R, Scherer A, et al. Response to tumor necrosis factor inhibition in male and female patients with ankylosing spondylitis: data from a swiss cohort. J Rheumatol 2018;45:506–12. 10.3899/jrheum.170166 [DOI] [PubMed] [Google Scholar]
  • 33.Neuenschwander R, Hebeisen M, Micheroli R, et al. Differences between men and women with nonradiographic axial spondyloarthritis: clinical characteristics and treatment effectiveness in a real-life prospective cohort. Arthritis Res Ther 2020;22:233. 10.1186/s13075-020-02337-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Micheroli R, Kissling S, Bürki K, et al. Sacroiliac joint radiographic progression in axial spondyloarthritis is retarded by the therapeutic use of TNF inhibitors: 12-year data from the SCQM registry. RMD Open 2022;8:e002551. 10.1136/rmdopen-2022-002551 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Ensslin C, Micheroli R, Kissling S, et al. Impact of sex on spinal radiographic progression in axial spondyloarthritis: a longitudinal swiss cohort analysis over a period of 10 years. RMD Open 2023;9:e003340. 10.1136/rmdopen-2023-003340 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Hebeisen M, Micheroli R, Scherer A, et al. Spinal radiographic progression in axial spondyloarthritis and the impact of classification as nonradiographic versus radiographic disease: data from the swiss clinical quality management cohort. PLoS One 2020;15:e0230268. 10.1371/journal.pone.0230268 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.van den Berg R, Lenczner G, Thévenin F, et al. Classification of axial SpA based on positive imaging (radiographs and/or MRI of the sacroiliac joints) by local rheumatologists or radiologists versus central trained readers in the DESIR cohort. Ann Rheum Dis 2015;74:2016–21. 10.1136/annrheumdis-2014-205432 [DOI] [PubMed] [Google Scholar]
  • 38.Jacquemin C, Rubio Vargas R, van den Berg R, et al. What is the reliability of non-trained investigators in recognising structural MRI lesions of sacroiliac joints in patients with recent inflammatory back pain? Results of the DESIR cohort. RMD Open 2016;2:e000303. 10.1136/rmdopen-2016-000303 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary data

rmdopen-2023-003455supp001.pdf (93KB, pdf)

Data Availability Statement

Data may be obtained from a third party and are not publicly available. Restrictions apply to the availability of data. The data are owned by a third party, the Swiss Clinical Quality Management in Rheumatic Diseases (SCQM) Foundation. Data may be obtained after approval and permission from the license holder (SCQM). Contact information for data requests: scqm@hin.ch.

Articles from RMD Open are provided here courtesy of BMJ Publishing Group

RESOURCES