Article Text

Download PDFPDF
A randomised, placebo controlled, comparative trial of the gastrointestinal safety and efficacy of AZD3582 versus naproxen in osteoarthritis
  1. L S Lohmander1,
  2. D McKeith2,
  3. O Svensson3,
  4. M Malmenäs3,
  5. L Bolin3,
  6. A Kalla4,
  7. G Genti5,
  8. J Szechinski6,
  9. C Ramos-Remus7,
  10. for the STAR Multinational Study Group
  1. 1Department of Orthopaedics, Medical Faculty, Lund University, Lund, Sweden
  2. 2Castlemilk Group Practice, Castlemilk Health Centre, Glasgow, UK
  3. 3AstraZeneca R&D, Södertälje, Sweden
  4. 4Groote Schuur Hospital, Cape Town, South Africa
  5. 5Department of Rheumatology and Physiotherapy, Flor Ferenc Kórház, Rheumatológia, Kistarcsa, Hungary
  6. 6Department of Rheumatology, Medical University of Wroclaw, Wroclaw, Poland
  7. 7Department of Rheumatology, Unidad de Investigacion en Enfermedades Cronico Degenerativas SC, Guadalajara, Jalisco, Mexico
  1. Correspondence to:
    Professor Stefan Lohmander
    Department of Orthopaedics, Lund University, Lund University Hospital, SE-221 85 Lund, Sweden;


Objective: To evaluate the gastrointestinal safety and efficacy of the COX inhibiting nitric oxide donator AZD3582 in patients with hip or knee osteoarthritis.

Methods: 970 patients were randomised (7:7:2) to AZD3582 750 mg twice daily, naproxen 500 mg twice daily, or placebo twice daily in a double blind study. The primary end point was the six week incidence of endoscopic gastroduodenal ulcers (diameter ⩾3 mm). Overall damage measured on the Lanza scale was a secondary end point. Safety and tolerability assessments included endoscopic upper gastrointestinal erosions and the gastrointestinal symptom rating scale (GSRS). Efficacy was primarily assessed by WOMAC.

Results: The incidence of ulcers with AZD3582 was 9.7% and with naproxen 13.7% (p = 0.07, NS), v 0% on placebo. The incidence of Lanza scores >2 was higher with naproxen (43.7%) than with AZD3582 (32.2%) (p<0.001). Compared with baseline, significantly fewer ulcers and erosions developed in stomach and stomach/duodenum combined, and fewer erosions developed in stomach, duodenum, and both combined on AZD3582 than on naproxen. GSRS reflux and abdominal pain subscale scores were lower for AZD3582 than for naproxen but there was no difference for indigestion, constipation, and diarrhoea. AZD3582 was as effective as naproxen at improving WOMAC scores. Both agents were well tolerated, with no significant effects on blood pressure.

Conclusions: At doses with similar efficacy in relieving osteoarthritis symptoms, the primary end point of six week endoscopic gastroduodenal ulcer incidence was not significantly different between AZD3582 and naproxen. Most secondary endoscopic gastrointestinal end points favoured AZD3582.

  • ACR, American College of Rheumatology
  • COX, cyclo-oxygenase
  • CINOD, COX-inhibiting nitric oxide donator
  • GSRS, gastrointestinal symptom rating scale
  • ITT, intention to treat
  • NSAID, non-steroidal anti-inflammatory drug
  • SF-36, 36 item short form health survey
  • VAS, visual analogue scale
  • WOMAC, Western Ontario and McMasters University osteoarthritis index
  • AZD3582
  • naproxen
  • gastrointestinal safety
  • osteoarthritis

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.


  • The members of the STAR Multinational Study Group are listed in the appendix