Log in

Comparison of the effect of treatment with NSAIDs added to anti-TNF therapy versus anti-TNF therapy alone on the progression of structural damage in the spine over 2 years in patients with radiographic axial spondyloarthritis from the randomised-controlled CONSUL trial

Ann Rheum Dis 2024 Feb. 13, 2024, 6 a.m.

Presented By

Dr Walter Maksymowych MB ChB, FRCP(C), FACP
Chief Medical Officer, CARE Arthritis; Professor of Medicine, University of Alberta

Guest Commentary By:

Dr Louis Bessette MD, MSc, FRCPC
Centre de l’ostéoporose et de rhumatologie de Québec (CORQ) Groupe de recherche en rhumatologie et maladies osseuses (GRMO) Centre de recherche du CHU de Québec, Université Laval Associate professor at Laval University

Synopsis

Introduction
  • Predictors of structural damage in the spine in axSpA include pre-existing ankylosis/syndesmophytes, high disease activity, and spinal inflammation on MRI.
  • A previous study suggested that celecoxib may prevent progression of structural damage, especially in patients with elevated ESR/CRP (Wanders et al), but a similar study evaluating diclofenac (Sieper et al) did not support this effect of NSAID therapy.
  • Observational studies have reported that TNF inhibitors may prevent structural progression (Sepriano et al).
  • It is unclear if adding NSAID therapy to TNF inhibition retards radiographic progression in radiographic axSpA patients at high risk (either presence of syndesmophytes and/or elevated CRP).
Methods
  • Study design included a 12-week ‘run-in phase’ of golimumab 50 mg for all subjects and then responders (BASDAI decrease of ≥2 (0-10 scale)) were randomized 1:1 to golimumab plus celecoxib (GOL/CEL) versus golimumab (GOL) alone for 96 weeks.
  • Primary endpoint was change in mSASSS at 2 years.
  • A difference of 1.5mSASSS units between groups was used for sample size calculations (worsening of 1.7 and 0.2 mSASSS units in GOL and GOL/CEL groups, respectively). 38 were deemed necessary for 80% power.
Results
  • Of 109 patients that were randomized, 97 (89%) completed the study at week 108.
  • Prior treatment with bDMARDs in 31.5% in the GOL/CEL group vs 16.4% in the GOL group.
  • In the GOL group, 23 (42.6%) patients received NSAIDs at least once.
  • Mean mSASSS change over 2 years was numerically higher in the GOL group (1.7; 95% CI 0.8 to 2.6) vs the GOL/CEL group (1.1; 95% CI 0.4 to 1.8) (p=0.79).
  • New syndesmophytes in 25% GOL vs 11% GOL/CEL (p=0.12)
  • All the sensitivity analyses numerically favored the GOL/CEL group
  • No significant difference in disease activity or adverse events between the groups

Commentary

This study was essentially underpowered to demonstrate a treatment group difference. Achieving change as little as 0.2 mSASSS units in the GOL group as proposed in the sample size calculations was probably unrealistic in view of the measurement error inherent to the assessment of change in mSASSS even with expert readers. A previous study had suggested a smallest detectable change cut-off of 2.9 for the mSASSS (Ramiro et al). Moreover, there were twice as many patients previously exposed to bDMARDs in the GLO-CEL group, a factor associated with a worse prognosis.

Special Guest Commentary

This study presents certain methodological limitations which restrict the interpretation of the data: error linked to the main measurement (mSASSS), relatively short follow-up at 2 years to assess the progression of structural damage, groups presenting different prognostic factors at baseline, and the intervention is not carried out blindly. Furthermore, this is a highly selected population, which limits the generalizability of the results to the entire axSpA population. We must also ask ourselves the question of whether this small numerical difference observed in the progression of the mSASSS score between the 2 groups is clinically relevant. However, the numerical difference observed in favor of the combined treatment could be relevant for some patients at high risk of radiographic progression and presenting symptoms with residual inflammation despite treatment with an anti-TNF. In summary, this study does not allow definitive conclusions to be drawn on the role of NSAIDs in the prevention of structural damage in axSpA.

Your Comments