Article Text

Concise report
The degree of spinal inflammation is similar in patients with axial spondyloarthritis who report high or low levels of disease activity: a cohort study
  1. U Kiltz1,
  2. X Baraliakos1,
  3. P Karakostas1,
  4. M Igelmann2,
  5. L Kalthoff3,
  6. C Klink4,
  7. D Krause4,
  8. E Schmitz-Bortz5,
  9. M Flörecke1,
  10. M Bollow6,
  11. J Braun1
  1. 1Department of Rheumatology Rheumazentrum Ruhrgebiet, Herne, Germany
  2. 2Rheumatology Private Office, Bochum, Germany
  3. 3Rheumatology Private Office, Herne, Germany
  4. 4Rheumatology Private Office, Gladbeck, Germany
  5. 5Rheumatology Private Office, Hattingen, Germany
  6. 6Augusta Krankenanstalten, Department of Radiology, Bochum, Germany
  1. Correspondence to Dr Uta Klitz, Rheumazentrum Ruhrgebiet, St Josefs-Krankenhaus, Landgrafenstr, 15 44652 Herne, Germany; kiltz{at}


Background The threshold for disease activity required to start antitumour necrosis factor (TNF) therapy has been arbitrarily set in patients with axial spondyloarthritis (axSpA) at Bath Ankylosing Spondylitis Disease Activity Index (BASDAI) ≥4. How this relates to spinal inflammation is unknown.

Objective To systematically compare the clinical, laboratory and imaging data of patients with axSpA with respect to their BASDAI level.

Methods A total of 100 consecutive patients with axSpA who had never been treated with TNF blockers were included. Laboratory parameters, spinal MRI and x-rays were quantified. Data were stratified according to BASDAI ≥4.

Results 44 patients were diagnosed as non-radiographic axSpA (nraxSpA) and 56 patients as ankylosing spondylitis (AS): median age 40.3±10.4 years; 57% male, mean disease duration since diagnosis 6.4±8.4 years, 88% HLA-B27+, mean modified Stokes Ankylosing Spondylitis Spinal Score 8.3±16.4. 60% of patients had spinal inflammation by MRI. The stratification based on BASDAI ≥4 disclosed significant differences in most clinical parameters but not for inflammation: patients with nraxSpA and BASDAI <4 versus ≥4 had 0.9±1.4 and 0.5±0.6 inflammatory lesions/patient, respectively (p=0.6), while patients with AS had 3.6±3.7 and 2.7±3.0 inflammatory lesions/patient, respectively (p=0.4).

Conclusion The burden of inflammation is quite comparable in patients with axSpA—regardless of disease activity. These data clearly challenge the concept of the recommended cut-off point of BASDAI ≥4.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Ankylosing spondylitis (AS) as the main subgroup of spondyloarthritis (SpA) is characterised by established radiographic changes in the sacroiliac joints (SIJs).1 Because this reflects chronic changes these criteria have considerable limitations in early disease stages. Therefore, new classification criteria for axial (axSpA) have been recently developed.2

The natural history of axSpA is characterised by a variable disease course. The natural course of patients with non-radiographic axSpA (nraxSpA), in particular, is incompletely understood. Treatment with tumour necrosis factor (TNF)α inhibitors is recommended in patients with SpA with persistent high disease activity according to the Assessment of SpondyloArthritis International Society (ASAS) recommendations.3 The cut-off point for the minimal clinical disease activity required to fulfil criteria for anti-TNF therapy has been set at a Bath Ankylosing Spondylitis Disease Activity Index (BASDAI) level of ≥4. However, this value has been proposed arbitrarily and its validity has not been formally established.4 Nevertheless, a BASDAI of 3.9 discriminated between patients with well-controlled and poorly controlled Sp.5 However, the level of inflammatory activity as demonstrated by MRI or raised C-reactive protein (CRP) levels in patients with SpA is unknown.

The aim of this study was to compare systematically the clinical, laboratory and imaging data of patients with SpA, stratified by the level of disease activity.


Consecutive patients with a diagnosis of axSpA who had never been treated with TNF blockers were included. The disease was diagnosed according the ASAS classification criteria for axSpA,2 (see online supplement 1) including patients with established AS.1 The local ethics committee had approved the study protocol, and all patients gave informed consent. The study was sponsored by an unrestricted educational grant from Pfizer, Germany.

Clinical assessment

All patients reported their demographic data, underwent a physical examination and measurement of laboratory parameters (CRP, HLA-B27). In addition, the Ankylosing Spondylitis Disease Activity Score (ASDAS) was calculated using the ASDAS-CRP formula: 0.12×back pain+0.06×duration of morning stiffness+0.11×patient global+0.07×peripheral pain/swelling+0.58×ln(CRP+1).6 7 A standardised assessment tool was completed by the patients (global pain, physician and patient's global assessment of disease activity, BASDAI,8 Bath AS Functional Index,9 Short Form-36 Health Survey10 11 and the AS Quality of Life questionnaire.12


Conventional radiographs of the SIJ, the cervical and the lumbar spine were collected for each patient. MRI examinations were performed using the routine imaging protocol of our department, based on the T1-weighted and the short-τ inversion-recovery sequences. Since MRIs were performed for those spinal areas which were clinically affected, we could obtain either images of the upper half or the lower half of the spine. All images were scored by an expert (XB) who was blinded for the clinical status of the patients. Radiographs were scored with the modified Stokes Ankylosing Spondylitis Spinal Score (mSASSS).13 Inflammation on MRI was scored based on the grading suggested in the Berlin score.14 Briefly, scoring was performed on the basis of a vertebral unit (VU) defined as the region between two virtual lines drawn through the middle of each vertebral body) and ranged between 0 (no inflammation) and 3 (inflammation in more than 50% of the VUs).

Statistical analysis

Descriptive data are presented as mean±SD when referring to quantitative variables and as absolute frequencies and percentages when referring to the qualitative variables. The Mann–Whitney U test was used to compare the data between subgroups at single time points. Since MRI examinations for individual patients included only half of the spine, we decided to base our MRI analyses on the mean number of inflamed spinal lesions per patient. We analysed the data based on a linear regression analysis. The dependent variables were MRI inflammation or CRP, depending on the analysis. The independent variables were gender and mSASSS baseline status and CRP or MRI inflammation, accordingly. Correlation coefficients were calculated by Spearman rank. A value of p<0.05 was considered statistically significant.


Demographic and clinical characteristics

Data are given in table 1. HLA-B27 was positive in 87 out of 99 (88%) patients; we found no significant differences in HLA-B27 according to gender (81.4% female vs 92.8% male, p=0.16) and in the subgroups studied (p=0.8). Almost all patients took conventional non-steroidal anti-inflammatory drugs or cyclo-oxygenase inhibitors (94%), 51% of them continuously. All 19 patients who took glucocorticoids had concomitant peripheral arthritis or SpA associated with inflammatory bowel disease.

Table 1

Demographic and clinical data of the patients

Disease activity, physical function and quality of life

Data for disease activity, physical function and quality of life are given in table 2.

Table 2

Characteristics of disease activity, physical function and quality of life


The radiographic criteria for AS were fulfilled by 56 patients, while 44 patients were diagnosed as nraxSpA. The mean mSASSS of all patients was 8.3±16.4, indicating relatively low levels of structural spinal changes.

MRI scans of the upper half of the spine (92% of which covered the region from C2/3 to L2/L3), were available from 25 patients, while 75 patients had MR images of the lower half of the spine (of which 97% covered the region from Th4/5 to L5/S1) available. Overall, a majority (60%) of all patients had active spinal inflammation as assessed by MRI. The mean number of VUs per patient was 12.8±3.0, the mean number of inflamed lesions per patient was 2.1±2.8 VUs/patient. There was no correlation between BASDAI and ASDAS and the extent of active inflammatory lesions and CRP (correlation coefficients were 0.08 and 0.06). Significant correlation was found only between CRP and the number of inflamed lesions per patient (r=0.22, p=0.03).

Stratification based on disease activity levels

The stratification according to BASDAI ≥4 showed statistically significant differences in most clinical parameters—with the exception of inflammatory activity, measured either by CRP or by MRI: patients with nraxSpA with BASDAI <4 versus ≥4 had 0.9±1.4 and 0.5±0.6 inflammatory lesions/patient, respectively (p=0.6), while patients with AS had 3.6±3.7 and 2.7±3.0 inflammatory lesions/patient, respectively (p=0.4) (table 3). These results were confirmed by linear regression analysis. In the within-group comparison of all patients, no difference was found between patients with AS and those with nraxSpA (figure 1). The analysis based on ASDAS cut-off point ≥2.1 showed no significant differences in the amount of inflamed MRI lesions between groups, but a comparison of CRP levels showed a significant difference.

Figure 1

Extension of inflammation between patients with nr-axSpA and AS

Table 3

Differences between patients with nraxSpA and AS stratified according to BASDAI level.


Our study shows that the degree of spinal inflammation was largely similar in patients with axSpA, irrespective of the corresponding BASDAI level. Of note, there was a difference in most clinical parameters when stratified according to the BASDAI.

The BASDAI cut-off point ≥4 is widely used to select patients for anti-TNF therapy. Indeed, the BASDAI cut-off point is robust as it discriminates across every-day aspects of patients' experience. There is some evidence that patients with AS may respond to TNF inhibitors despite not fulfilling the ASAS criteria for initiating anti-TNF therapy.15

Lack of correlation between clinical measurements of disease activity and MRI scores has been described for patients with SpA in several studies.16,,18 Most of these surveys have been undertaken in patients with AS receiving anti-TNF therapy. We report in our cohort that 60% of the patients had signs of active inflammation. Although this is lower than the 70–80% reported previously in patients with AS, one has to consider that this percentage have been described in patients with AS who have been screened to receive TNF blockers.

Since it has also been shown that both CRP and MRI may predict the outcome of anti-TNF therapy,19 the question arises whether it might be more meaningful to use these parameters as a guide to starting biological treatment in axial SpA—potentially, even independently of current BASDAI levels.

The ASAS group has recently developed ASDAS, a new tool to assess disease activity.6 It is a composite index that includes back pain, duration of morning stiffness, patient global assessment, peripheral joint complaints and CRP, and it was shown to be highly discriminatory in differentiating patients with different levels of disease activity and patients with different levels of change. The differentiation between nraxSpA and AS based on ASDAS cut-off points for objective signs of inflammation (CRP, inflamed MRI lesions) showed a significant difference only in the analysis of CRP levels. We explain this as due to the CRP-driven formula in ASDAS. For further confirmation, analysis of larger patient groups is needed.

A remarkable finding of our study is that the patient-reported outcomes are the only difference between patients with high and low disease activity. This is clearer for pain and patient global assessment than for physician global assessment. The physician global assessment failed to show differences in patients with nraxSpA.

A limitation of our study is that we did not assess the SIJs and the whole spine by MRI because we had limited personal and financial resources.. This study design does certainly lead to a systematic underestimation of MRI findings in clinically asymptomatic parts of the axial skeleton. However, inflammation has been found more commonly in the thoracic spine, which is the part of the spine which is present in each patient in our study.20

This is the first study to show that inflammatory activity as assessed by MRI and CRP did not differ in patients with axSpA, including patients with AS and BASDAI levels ≥4. These data clearly challenge the concept of the arbitrarily set clinical cut-off point of BASDAI ≥4.


Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

    Files in this Data Supplement:

    • Web Only Data - This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.


  • Funding Unrestricted educational grant, Pfizer, Germany.

  • Competing interests None.

  • Patient consent Obtained.

  • Ethics approval Ethics committee, University of Muenster, Germany.

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