Decrease in ASDAS score at follow-up compared with ASDAS score at baseline of at least 1.1.

Decrease in ASDAS score at follow-up compared with ASDAS score at baseline of at least 1.1.

Maximum 1 disco-vertebral unit with inflammation present.

At least 50% reduction in Canada-Denmark (CANDEN) MRI spine inflammation score.

Inflammation in 2 or fewer quadrants as assessed slice by slice across MRI slices that depict the cartilagenous compartment of the sacroiliac joints.

A score of maximally 2 of the MRI Peripheral Joints and Entheses Inflammation Index.

Race and Ethnicity were not collected from any participant.

Each sacroiliac joint is assigned a score of 0 to 4 based on the appearance on a radiograph, grade 0 is normal, grade 1 is suspicious changes, grade 2 is minimum abnormality, grade 3 is unequivocal abnormality and grade 4 is severe abnormality. A patient fulfils the radiographic part of the modified New York criteria if there is sacroiliitis of at least grade 2 bilaterally or sacroiliitis grade 3 or 4 unilaterally. Reference: Arthritis Rheum 1984;27:361-8.

C-reactive protein. The assay had a lower limit of detection of 0.3 mg/L.

ASDAS is a composite index to assess disease activity in axSpA. ASDAS parameters: 1) Total back pain 2) Patient global 3) Peripheral pain/swelling 4) Duration of morning stiffness 5) CRP in mg/L: ASDAS calculation: 0.121 × total back pain + 0.110 × patient global + 0.073 × peripheral pain/swelling + 0.058 × duration of morning stiffness + 0.579 × Ln(max(CRP,2)+1). CRP is in milligram/liter (mg/L), the range of other variables is from 0 to 10. Data from five variables combine to yield a score (minimum 0.636 to no defined upper limit), higher scores indicate higher disease activity.

BASDAI is a measure of disease activity which was calculated based on 6 separate questionnaire items that were answered by the patients using visual analogue scale scales with range 0-10. A weighted average was performed, where items 5 and 6 had a weight of 0.5 whereas items 1, 2, 3 and 4 had a weight of 1. Range is from 0 to 10, with 10 indicating more severe disease activity.

BASFI is a measure of physical function which was calculated based on 10 separate questionnaire items that were answered by the patients using visual analogue scales with range 0-10, an average across 10 items was calculated. The range is 0-10 with 10 indicating that the level of physical functioning is more severely affected.

5 components of BASMI were each scored 0, 1 or 2, based on physical examination tests of the range of motion of the vertebral column and hips. The scores were added, and the range was 0-10, where 10 indicates a worse metrology (i.e. more severely affected range of motion of the vertebral column and hips).

Range is 0-35. The entheses assessed were those that are included in either the Maastricht Ankylosing Spondylitis Enthesitis Score (MASES), the Spondyloarthritis Research Consortium of Canada (SPARCC) Enthesitis index or the Leeds Enthesitis Index (LEI).

Questionnaire filled in by the patient, visual analogue scale with range 0-10, 10 indicates more severe pain.

Questionnaire filled in by the patient to capture the global disease activity as assessed by the patient; visual analogue scale with range 0-10, 10 indicates worse global score

Questionnaire filled in by the physician to capture the physician's global assessment of disease activity; visual analogue scale with range 0-10, 10 indicates a worse disease activity.

The mSASSS is an assessment of structural lesions of the spine using lateral radiographs of the cervical and lumbar spine. The mSASSS is the sum of the lumbar and cervical spine score (range 0 [no change] to 72 [severe change]), derived from scoring the anterior sites of the lumbar spine (T12 to S1) and the cervical spine (C2 to T1) as either 0 (normal), 1 (erosion, sclerosis, or squaring), 2 (syndesmophyte), 3 (bridging syndesmophyte), and adding all scores together.

Semiquantitative score of inflammation of the spine as assessed by MRI. A comprehensive assessment system is applied to different anatomical parts of the spine at 23 spinal levels and all scores are summed; range is 0-614 with higher scores indicating more severe inflammation. Reference: RMD Open 2019;5:e001057.

Semiquantitative score of fat metaplasia of the spine as assessed by MRI. A comprehensive assessment system is applied to different anatomical parts of the spine at 23 spinal levels and all scores are summed; range is 0-510 with higher scores indicating more severe fat metaplasia. Reference: RMD Open 2019;5:e001057.

Semiquantitative score of erosion of the spine as assessed by MRI. A comprehensive assessment system is applied to different anatomical parts of the spine at 23 spinal levels and all scores are summed; range is 0-208 with higher scores indicating more severe erosion. Reference: RMD Open 2019;5:e001057.

Semiquantitative score of new bone formation of the spine as assessed by MRI. A comprehensive assessment system is applied to different anatomical parts of the spine at 23 spinal levels and all scores are summed; range is 0-460 with higher scores indicating more severe new bone formation. Reference: RMD Open 2019;5:e001057.

Spondyloarthritis Research Consortium of Canada (SPARCC) Magnetic Resonance Imaging (MRI) Spine Inflammation Index. 23 levels of the spine are assessed for the presence, intensity and depth of inflammation at a maximum of 3 slices per level, and scores are summed together. Range is 0-414 with higher scores indicating more severe inflammation of the spine.

Spondyloarthritis Research Consortium of Canada (SPARCC) Magnetic Resonance Imaging (MRI) Sacroiliac Joint Inflammation Index as based on an assessment of the presence, depth and intensity of inflammation as assessed on 6 MRI slices, and all scores are summed together; range is 0-72, with higher scores indicating more severe inflammation of the sacroiliac joints.

Spondyloarthritis Research Consortium of Canada (SPARCC) Magnetic Resonance Imaging (MRI) Sacroiliac Joint Structural Score - Fat is based on an assessment of MRI images of the sacroilliac joints where the presence of fat is assessed on 5 MRI slices; range is 0-40, with higher scores indicating worse fat metaplasia.

Spondyloarthritis Research Consortium of Canada (SPARCC) Magnetic Resonance Imaging (MRI) Sacroiliac Joint Structural Score - Erosion is based on an assessment of MRI images of the sacroilliac joints where the presence of erosion is assessed on 5 MRI slices; range is 0-40, with higher scores indicating worse erosion.

Spondyloarthritis Research Consortium of Canada (SPARCC) Magnetic Resonance Imaging (MRI) Sacroiliac Joint Structural Score - Backfill is based on an assessment of MRI images of the sacroilliac joints where the presence of backfill is assessed on 5 MRI slices; range is 0-20, with higher scores indicating worse backfill. According to the recently updated terminology by Assessment of Spondyloarthritis International Society MRI Group, backfill is termed "fat metaplasia in an erosion cavity".

Spondyloarthritis Research Consortium of Canada (SPARCC) Magnetic Resonance Imaging (MRI) Sacroiliac Joint Structural Score - Ankylosis is based on an assessment of MRI images of the sacroilliac joints where the presence of ankylosis is assessed on 5 MRI slices; range is 0-20, with higher scores indicating worse ankylosis.

Magnetic Resonance Imaging (MRI) Peripheral Joints and Entheses Inflammation Index was based on an assessment of MRI images of several image stations: anterior chest wall and shoulder, hands, pelvis, knees, ankles and feet; 74 joints were assessed for synovitis and osteitis, 30 entheses were assessed for soft tissue inflammation and osteitis, and all scores were summed together; range was 0-210, with higher scores indicating worse inflammation of joints and entheses throughout the body as assessed by MRI.

IBP according to experts: four out of five of the following parameters present: (1) age at onset less than 40 years, (2) insidious onset, (3) improvement with exercise, (4) no improvement with rest, (5) pain at night (with improvement upon getting up). (Reference: Ann Rheum Dis 2009;68:777-83)

Past or present active synovitis diagnosed by a doctor. (Reference: Ann Rheum Dis 2009;68:777-83)

Heel enthesitis: past or present spontaneous pain or tenderness at examination at the site of the insertion of the Achilles tendon or plantar fascia at the calcaneus. (Reference: Ann Rheum Dis 2009;68:777-83)

Past or present uveitis anterior, confirmed by an ophthalmologist. (Reference: Ann Rheum Dis 2009;68:777-83)

Past or present dactylitis diagnosed by a doctor. (Reference: Ann Rheum Dis 2009;68:777-83)

Past or present psoriasis diagnosed by a doctor. (Reference: Ann Rheum Dis 2009;68:777-83)

Past or present Crohn disease or ulcerative colitis diagnosed by a doctor. (Reference: Ann Rheum Dis 2009;68:777-83)

At 24-48 h after a full dose of NSAID the back pain is not present anymore or much better. (Reference: Ann Rheum Dis 2009;68:777-83)

Presence in first-degree or second-degree relatives of any of the following: (a) ankylosing spondylitis, (b) psoriasis, (c) uveitis, (d) reactive arthritis, (e) inflammatory bowel disease. (Reference: Ann Rheum Dis 2009;68:777-83)

CRP above upper normal limit in the presence of back pain, after exclusion of other causes for elevated CRP concentration. (Reference: Ann Rheum Dis 2009;68:777-83)

Conventional synthetic disease-modifying antirheumatic drugs (csDMARDs) were allowed, but the dose could not be changed 4 weeks prior to baseline.

At least 50% decrease or at least an improvement of 2.0 in BASDAI, when measured on a scale with range 0-10.

At least 50% decrease or at least an improvement of 2.0 in BASDAI, when measured on a scale with range 0-10.

Decrease in ASDAS score at follow-up compared with ASDAS score at baseline of at least 1.1.

Decrease in ASDAS score at follow-up compared with ASDAS score at baseline of at least 1.1.

Maximum 1 disco-vertebral unit with inflammation present.

At least 50% reduction in Canada-Denmark (CANDEN) MRI spine inflammation score.

At least 50% decrease in Spondyloarthritis Research Consortium of Canada (SPARCC) MRI sacroiliac joint inflammation index

At least 50% reduction in MRI Peripheral Joints and Entheses Inflammation Index. MRIs of 74 peripheral joints were evaluated separately for synovitis and osteitis (i.e. bone marrow edema) and 30 peripheral entheses were evaluated separately for soft tissue inflammation and osteitis. All types of lesions (i.e. joint synovitis, joint osteitis, entheseal soft tissue inflammation and entheseal osteitis) were scored on a semi-quantitative scale of 0-2 ('none', 'mild', 'moderate to severe').

Inflammation in 2 or fewer quadrants as assessed slice by slice across MRI slices that depict the cartilagenous compartment of the sacroiliac joints.

A score of maximally 2 of the MRI Peripheral Joints and Entheses Inflammation Index.