Article Text

AB0523 Disease activity score (asdas) has increased responsiveness but decreased discriminatory capacity in patients with ankylosing spondylitis(as) on continous nsaid therapy(cnsaid)
  1. S. Kumar1,
  2. Q. Zaheer1,
  3. A. N. Malaviya1
  1. 1Department of Rheumatology, ISIC hospital, New Delhi, India


Background Prospective evaluation has demonstrated that ASDAS had high responsiveness in TNFi treated patients. Change in ASDAS correlated best with change in CRP.1However, inflammatory markers don’t correlate with disease activty in AS.2,3 The present study was aimed at comparison of ASDAS and BASDAI during cNSAID treatment.

Objectives Compare the responsiveness and discriminatory capacity of ASDAS to BASDAI in patients on cNSAID therapy.

Methods Prospective, open label single arm observational cohort. After IEC clearance and consent, all patients >18yrs with AS started on cNSAID were included and followed for 6 months. Exclusions included, among others, the use of cNSAID in the preceeding 3 months, use of TNFi, DMARDs, steroids or pamidronate in the preceeding 3 months or at any point during study period, loss from follow up before completion of 3 months.

Results Sixty two patients(48 males,14 females) with a mean age 32.1(±8.37)yrs completed the study. HLA B27 was positive in 56(90.3%). Mean duration of therapy was 200(±81)days. There was a significant change in mean values of all disease measure indices. Among the indices, ASDAScrp had the largest effect size.(table). Change in patient global(ptGlobal) correlated best with change in BASDAI(r=0.638). The correlation coefficient between change in ptGlobal and change in ASDAScrp & ASDASesr was 0.534 & 0.614 respectively. Change in ESR & CRP correlated poorly with change in ptGlobal(r=0.163 & -0.035 respectively). Change in CRP & ESR contributed 32.5% and 54.1% of the variation in change of ASDAScrp & ASDASesr respectively. Correlation between fulfillment of ASAS20, ASAS5/6 & MCID ptGlobal (minimal clinically important difference) was higher for fulfillment of 50% decrease in BASDAI(rho between 0.523 & 0.6) than fulfillment of decrease ≥1.1 in ASDAS indices(rho between 0.347 & 0.549). When ptGlobal≤10 mm was assumed as the definition of remission, at end of follow up, ROC analysis demonstrated the BASDAI was better in predicting remission(AUC=0.813) compared to ASDAScrp(AUC=0.750) and ASDASesr(AUC=0.796). Similar results were found when ASASpr was assumed as the definition of remission.(AUC for BASDAI=0.941, ASDAScrp=0.883 & ASDASesr=0.888). An ASDAS<1.3 was more likely to miss patients in remission compared to BASDAI<1. When nonresponder status(Pt global>60 at the end of treatment) was considered, ASDAScrp(AUC=0.910) and ASDASesr(AUC=0.859) performed much better than the BASDAI(AUC=0.825).

Conclusions Among AS patients on cNSAIDs, the ASDAS indices were slightly more sensitive to change. However, significant change in BASDAI correlated much better with previously validated measures of change. The discriminatory capacity of the BASDAI was better at determining presence of remission. Thus in patients on cNSAID, use of ASDAS indices may falsely classify responders as non responders leading to unneccessary TNFi usage.

  1. Pendred SJ et al. Ann Rheum Dis.2010;69(6):1065-71;

  2. Spoorenberg A, et al J Rheumatol. 1999;26:980-4;

  3. Ozgocmen S et al,Joint Bone Spine. 2007;74:249-53.

Disclosure of Interest S. Kumar: None Declared, Q. Zaheer: None Declared, A. Malaviya Consultant for: Pfizer, Roche, MSD

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.