Article Text

Concise report
Long-term effects of secukinumab on MRI findings in relation to clinical efficacy in subjects with active ankylosing spondylitis: an observational study
  1. Xenofon Baraliakos1,
  2. Babul Borah2,
  3. Juergen Braun1,
  4. Dominique Baeten3,
  5. Didier Laurent4,
  6. Joachim Sieper5,
  7. Paul Emery6,
  8. Iain B McInnes7,
  9. Jacob M van Laar8,
  10. Paul Wordsworth9,
  11. Juergen Wollenhaupt10,
  12. Herbert Kellner11,
  13. Laurence Colin4,
  14. Francois Vandenhende12,
  15. Kath Radford4,
  16. Wolfgang Hueber4
  1. 1Rheumazentrum Ruhrgebiet, Herne, Germany
  2. 2Novartis Institutes for BioMedical Research, Cambridge, Massachusetts, USA
  3. 3University of Amsterdam, Amsterdam, The Netherlands
  4. 4Novartis Institutes for BioMedical Research, Basel, Switzerland
  5. 5Charite Campus Benjamin Franklin, Berlin, Germany
  6. 6Leeds Institute of Rheumatic and Musculoskeletal Medicine, University of Leeds and NIHR Leeds Musculoskeletal Biomedical Research Unit, Leeds Teaching Hospitals NHS Trust, Leeds, UK
  7. 7University of Glasgow, Glasgow, UK
  8. 8Newcastle University, Newcastle, UK
  9. 9NIHR Oxford Comprehensive Biomedical Research Centre, Nuffield Orthopaedic Centre, Oxford, UK
  10. 10Schoen Klinik Hamburg, Eilbeck, Hamburg, Germany
  11. 11Centre for Inflammatory Joint Diseases, Munich, Germany
  12. 12ClinBAY, Genappe, Belgium
  1. Correspondence to Professor Juergen Braun, Rheumazentrum Ruhrgebiet, Claudiusstr 45, Herne 44649, Germany; juergen.braun{at}


Introduction A 28-week study suggested efficacy of the anti-interleukin-17A monoclonal antibody secukinumab in active ankylosing spondylitis (AS). MRI-assessed inflammation was reduced at weeks 6, 28.

Objective To analyse the longer-term effects of secukinumab on MRI inflammatory and non-inflammatory spinal lesions in relation to its clinical efficacy in subjects with active AS.

Methods Spinal MRI results (baseline, week 94) for 13 subjects with AS initially treated with secukinumab 2×10 mg/kg intravenously (n=10) or placebo (n=3) and receiving a secukinumab maintenance dose of 3 mg/kg IV every 4 weeks up to week 94 were evaluated by the Berlin score; inflammatory/non-inflammatory (fatty) changes were assessed at vertebral edges (VEs). Results were compared with clinical outcomes.

Results Most of the 13 subjects assessed at week 94 had sustained clinical responses: 8 (62%) achieved Assessment of SpondyloArthritis international Society 20% (ASAS20), including 6 (46%) achieving ASAS40 responses, corresponding to 75% and 83% reductions in the Berlin score, respectively. In the 10 subjects treated with secukinumab throughout the study period, 79/91 (87%) inflammatory VEs at baseline resolved by week 94; new fatty lesions occurred in 39/796 (4.9%) of VEs; 87/124 (70%) VEs with fatty lesions at baseline remained unchanged; 30% were no longer visible.

Conclusions In this pilot study, secukinumab treatment up to 2 years yielded sustained clinical improvement accompanied by regression of spinal inflammation. The impact of secukinumab on the development of fatty changes and bone formation in AS will be assessed in larger trials.

Trial registration number This study is registered with, number NCT00809159.

  • Ankylosing Spondylitis
  • Magnetic Resonance Imaging
  • Inflammation

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.


Early evidence suggests that interleukin-17A may be of pathogenetic relevance in ankylosing spondylitis (AS).1 ,2 Secukinumab, a fully human IgG1κ monoclonal antibody, represents a new approach to treating AS by selectively targeting interleukin-17A. In a randomised, placebo-controlled, proof-of-concept trial involving 30 subjects with active AS (24 randomised to secukinumab 2×10 mg/kg intravenously (IV), 6 to placebo), secukinumab induced significant improvements over 28 weeks and reduced MRI-assessed axial inflammation at weeks 6 and 28 (as measured by Berlin scores3 vs baseline).1 Fat infiltration at the vertebral edges (VEs)—that is, non-inflammatory spinal lesions, has been reported in a large majority of subjects with established AS.4 ,5 There is a growing interest in studying MRI fatty lesions to understand the potential relationship between inflammation, fatty degeneration and new bone formation in AS.6 ,7 In this observational, open-label extension of the proof-of-concept (‘core’) study, we analysed the longer-term effects of secukinumab on MRI inflammatory and non-inflammatory spinal lesions in relation to its clinical efficacy in subjects with active AS.


Subjects diagnosed with AS (according to the modified New York criteria)8 and active disease as defined by a Bath Ankylosing Spondylitis Disease Activity Index (BASDAI) score ≥4 and a total back pain and nocturnal pain score ≥4 by Visual Analogue Scale (0–100 mm) despite maximal tolerated doses of non-steroidal anti-inflammatory drugs were included in the core study. Key exclusion criteria were total ankylosis of the spine, psoriatic arthritis, presence of severe acute or subacute anterior uveitis or clinical contraindications for MRI. Full methods for the core study have been previously published1 and the study protocol is available from the sponsor. There were three contributing study centres for the extension study—two in Germany and one in the Netherlands. General inclusion/exclusion criteria and ethical considerations for the extension study are provided in the online supplementary material.

Complete MRI examinations from baseline and week 94 (end of the open-label study extension) were available from 13 subjects who received secukinumab 2×10 mg/kg IV (n=10) or placebo (n=3) administered 3 weeks apart for 28 weeks in the core study, followed by secukinumab 3 mg/kg IV every 4 weeks (q4w) up to week 94 in the open-label extension (a total of 14 injections, regardless of prior dose or assignment to placebo, for the extension) (figure 1). Clinical efficacy measurements in these subjects included the Assessment of SpondyloArthritis international Society 20% (ASAS20) and 40% (ASAS40) responses, BASDAI50%9 responses and C-reactive protein (CRP) values. All these measurements were performed at baseline of the core study, on day 1 and day 29 of the extension study and every 4 months thereafter. Only assessments performed at core study baseline and at the end of the open-label extension (week 94) are presented here.

Figure 1

Subject disposition flow chart. *Regimens to left of right-facing arrows were those received during the core study; regimens to the right of the arrows were those received during the extension study. PBO, placebo; SEC, secukinumab; VE, vertebral edge.

Spinal inflammation for each of the 23 vertebral units (VUs), including the cervical, thoracic and lumbar spine, was assessed using the Berlin modification of the ASspiMRI-a scoring system.10 Inflammatory and fatty lesions in the four VEs of each VU were scored as 1=present or 0=absent7 ,11 in the 10 MRI subjects treated with secukinumab during the core and extension studies; owing to the small sample size, the three subjects who were receiving placebo in the core study and switched to secukinumab during the extension study were excluded. Maximum score for the VE-level lesions was 4×23=92. MRI results were analysed by an experienced reader blinded to subjects’ characteristics and MRI timepoints.

Owing to the small sample size, no statistical comparisons were made. For MRI and clinical data, subject-level comparisons for changes between core study baseline and week 94 (when all subjects received open-label secukinumab) are presented.


Baseline characteristics for the MRI cohort (n=13), VE/fatty lesions cohort (n=10) and core study cohort (N=30) are shown in online supplementary table S1. Although no statistical comparisons were made, the characteristics for the subset cohorts (MRI and VE/fatty lesions cohorts) did not differ appreciably from those for the larger core study.

MRI findings and comparisons with clinical assessments

The Berlin score was reduced in 8/13 (61.5%) subjects, remained unchanged in three subjects and increased in two subjects at week 94 versus the core study baseline. A total of 8/13 subjects (61.5%) had an ASAS20 response and six of these eight (75%) showed a reduction of their Berlin score. Similarly, 6/13 subjects (46.2%) had an ASAS40 response and five of these six (83%) showed a Berlin score reduction. Finally, a BASDAI response was seen in 13/13 subjects (100%), with 5/13 (38.5%) demonstrating a BASDAI50% response; all five of these (100%) had a Berlin score reduction at week 94 (table 1). CIs for the ASAS20, ASAS40 and BASDAI data are presented in online supplementary table S2.

Table 1

ASAS20, ASAS40, BASDAI and Berlin score data at week 94*

A detailed characterisation of the numbers/types of lesions seen at VEs at baseline and week 94 for the 10 MRI subjects who were treated with secukinumab during the core and extension studies is provided in figure 2. A total of 920 VEs were analysed by MRI at baseline and week 94 (10 subjects×23 VEs×4 VEs/unit). Of the 91 (9.9%) inflammatory lesions seen at baseline, 79 (86.8%) were resolved at week 94. By contrast, of the 124 (13.5%) VEs showing fatty lesions at baseline, 87 (70.2%) did not change during the 94 weeks of secukinumab treatment (figure 2), while the rest were no longer visible at the end of the study. New fatty lesions appeared in 39/796 (4.9%) VEs. In detail, 58 (6.3%) VEs with inflammation at baseline appeared to have no inflammation and no fatty lesions at week 94; 8 (0.9%) VEs with inflammation but no fatty lesions at baseline showed resolution of inflammation but new fatty lesions at week 94; and 2 (0.2%) VEs had fatty lesions and inflammation at baseline and week 94. New inflammatory lesions appeared in 21 (2.5%) VEs at week 94 that were inflammation-free at baseline (figure 2).

Figure 2

Distribution of MRI inflammatory and fatty lesions in VEs in subjects (N=10) treated with secukinumab from core study baseline to week 94. The total number of evaluable VEs was 920. (A and B) Number of vertebral edges (VEs) with inflammatory (Inf+) (A) and fatty (FD+) (B) lesions at core study baseline and at week 94. Data presented over bars are absolute numbers of VEs (percentage of the total number of evaluable VEs (N=920)). FD+/FD−, fatty lesions present/absent; Inf+/Inf−, inflammatory lesions present/absent; VE, vertebral edge.

The observed within-subject changes for inflammatory lesions and fatty lesions, respectively, were not connected in a majority of subjects, either in their magnitude or direction of change over time (see online supplementary table S3). Although these results were derived from a small cohort, it can be speculated that the discrepancy is related to differences in underlying biology for the inflammatory lesions versus fatty lesions. Safety findings are presented in the online supplementary materials.


Our exploratory MRI analysis showed reduced inflammatory activity after 2 years for patients receiving secukinumab, as quantified by Berlin score and VE-level findings (ie, number of VEs showing bone marrow oedema). As previously reported, MRI-assessed axial inflammation was reduced as early as 6 weeks after initiation of secukinumab treatment (loading regimen of 2×10 mg/kg IV) in the core study.1 The current analysis showed that reduction of inflammation was sustained for up to 2 years, with 87% of initially inflamed lesions resolving during this time, consistent with the overall decrease in inflammation as quantified using the Berlin score. These results are consistent with the findings of previous, larger AS trials with tumour necrosis factor blockers (infliximab,12 etanercept,13 adalimumab14 and golimumab15), which showed similar decreases in MRI spinal inflammation. Most of the fatty lesions seen at VEs at core study baseline remained unchanged at week 94. Intriguingly, only a small fraction of VEs that were inflammation positive (Inf+) at baseline converted to fatty lesions (FD+) at week 94.

The significance of the disappearance of about 30% of fatty lesions observed before secukinumab treatment is unclear. Nevertheless, we believe this finding is worth reporting, especially given the much lower rates of fatty lesion disappearance (<6%) reported in tumour necrosis factor-blocker studies.16

We observed a sustained clinical response in most subjects over 94 weeks. Notably, a preliminary dosing regimen was used in this phase 2 study and ongoing phase 3 studies will confirm optimal dosing for secukinumab in AS. We explored the relationship of clinical improvements with MRI score changes up to week 94. The results suggested good correspondence between regression of spinal inflammation at week 94 and clinical response in the same subjects. Moreover, three subjects who received placebo during the core study followed by secukinumab in the current extension and were included in the n=13 MRI cohort, experienced benefits as shown by reduced MRI inflammation and clinical outcomes (table 1). Considering that there was good correspondence between clinical improvement and reduction of Berlin score in our study, the increase of a small fraction (2.5%) of new VU-level inflammatory lesions seen at week 94 may not be of clinical significance.

MRI fatty lesion analyses were conducted to detect correspondence between single VEs. We did not attempt to correlate these changes with subject-level clinical outcomes, as we did not have radiographic data to perform a meaningful analysis.

This observational study had several limitations. First, because we included only subjects with complete clinical and MRI data both at baseline and after 2 years of secukinumab treatment, the number of eligible subjects was small (n=13), precluding group-level statistical comparisons. Second, only 10 of the 13 MRI subjects received secukinumab during the entire study period. Lastly, the MRI analysis was based on evaluation of the spine but excluded sacroiliac joints, a major spondyloarthritis inflammation site.17

Our small observational study of subjects receiving secukinumab suggested that regression of spinal MRI inflammation was sustained up to week 94 and was paralleled by clinical improvements. Comprehensive assessment of the impact of secukinumab treatment on axial MRI changes—and ultimately on radiographic progression of AS—will require larger, long-term studies.


We thank Venugopal Madhusudhana, MBBS, PGDMLE, of Novartis for assistance with writing this manuscript. We also acknowledge the technical assistance of VirtualScopics (Rochester, New York, USA) in obtaining the imaging data. We thank Sheila Gardner of Quintiles for critical review of the manuscript. We also thank Dr Irina Koroleva, Biomarker Development, Novartis Institutes for BioMedical Research, for helpful scientific discussion. BioScience Communications (New York, New York, USA) provided editorial assistance, supported by Novartis Pharma AG.


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.


  • Handling editor Tore K Kvien

  • Contributors XB: Study concept, reading and interpretation of images, critical review of the manuscript for important intellectual content, approval of manuscript for submission. BB: Study concept, data analysis and interpretation, writing of first draft, critical review of the manuscript for important intellectual content, approval of manuscript for submission. JB: Study concept, study design, critical review of the manuscript for important intellectual content, approval of manuscript for submission. LC: Design of statistical analysis, critical review of the manuscript for important intellectual content, approval of manuscript for submission. FV: Statistical analysis. KR: Data collection, study coordination and management, approval of manuscript for submission. WH: Study concept, study design, data interpretation, writing of first draft, critical review of the manuscript for important intellectual content, approval of manuscript for submission. DB, DL, JS, PE, IBM, JMvL, PW, JW, HK: Critical review of the manuscript for important intellectual content, approval of manuscript for submission.

  • Funding The study and publication were funded by Novartis Pharma AG, Basel, Switzerland. Study authors who were employees of Novartis were involved in the design of the study and the collection, analysis and interpretation of the data. The first draft of this report was written by a professional medical writer employed by Novartis (see Acknowledgements above). All authors provided critical review of this report and agreed to submit it to the Annals of the Rheumatic Diseases for publication.

  • Competing interests XB has served as a consultant or paid speaker for or participated in clinical trials sponsored by, AbbVie, Boehringer Ingelheim, Celgene, Centocor, Chugai, MSD, Novartis, Pfizer and UCB. BB is employed by Novartis. JB has received honoraria for talks, advisory boards, paid consultancies and grants for studies from Abbott/AbbVie, Amgen, BMS, Boehringer Ingelheim, Celgene, Celltrion, Centocor, Chugai, EBEWE Pharma, Janssen, Medac, MSD (Schering-Plough), Mundipharma, Novartis, Pfizer (Wyeth), Roche, Sanofi-Aventis and UCB. DB has served as a consultant or paid speaker for or participated in clinical trials sponsored by AbbVie, BMS, Boehringer Ingelheim, Glenmark, Janssen, Lilly, MSD, Novartis, Novo Nordisk, Pfizer, Roche and UCB. DL is employed by Novartis and owns shares of that company. JS has served as a consultant or paid speaker for or participated in clinical trials sponsored by, AbbVie, Janssen, Lilly, Merck, Novartis, Pfizer and UCB. PE has participated in clinical trials for and provided expert advice to, Abbott/AbbVie, BMS, Lilly, MSD, Novartis, Pfizer, Roche, Samsung, Takeda and UCB. IBM has served as a consultant or paid speaker for or participated in clinical trials and research supported by, AbbVie, Amgen, Celgene, Janssen, Novartis, Pfizer and UCB. JMvL has served as a consultant for BMS, Lilly, MSD, Roche and TiGenix and received speaker fees from AbbVie, Miltenyi and Pfizer. JW has served as a consultant or paid speaker for or participated in clinical trials sponsored by, AbbVie, AstraZeneca, BMS, Celgene, Chugai, Johnson & Johnson, Lilly, MSD, Novartis, Pfizer and UCB. HK has served as a consultant or paid speaker for or participated in clinical trials sponsored by, AbbVie, BMS, MSD, Novartis, Pfizer, Roche and UCB. FV works for a company that provides paid services to Novartis for the analysis and reporting of clinical trials and owns shares of Novartis. KR is employed by Novartis. WH is employed by Novartis and owns shares of that company.

  • Patient consent Obtained.

  • Ethics approval All centres received approval from independent ethics committees or institutional review boards, and the study was done in accordance with the principles of the Declaration of Helsinki.

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