Arthritis & Rheumatism, Volume 60,
October 2009 Abstract Supplement

The 2009 ACR/ARHP Annual Scientific Meeting
Philadelphia October 16-21, 2009.

Gastrointestinal Safety in Patients with Rheumatoid Arthritis Treated with Tocilizumab: Data From Roche Clinical Trials

van Vollenhoven1,  Ronald F., Keystone2,  Edward C., Furie3,  R., Blesch4,  A., Wang4,  C., Curtis5,  J. R.

Karolinska Univ Hosp, Stockholm, Sweden
Professor of Medicine/University of Toronto, Toronto, ON
NS-LIJHS, Lake Success, NY
Roche, Nutley, NJ
University of Alabama at Birmingham, Birmingham, AL


To characterize upper and lower gastrointestinal (GI) perforation events in the tocilizumab (TCZ) worldwide rheumatoid arthritis (RA) Roche clinical trials database and to compare with GI perforation event rates among RA patients enrolled in a managed care database.


Cases of GI perforations have been systematically identified on a quarterly basis since 1Q2006 using the TCZ worldwide RA clinical trials database. Cumulative safety data through February 6, 2009, for adverse events were integrated for patients with RA from the Roche clinical trials database. Rates of GI perforations that occurred in patients with RA in clinical practice were analyzed using administrative claims data from the United Health Care database.


In worldwide Roche clinical trials, 4009 patients with RA have received at least one dose of TCZ (4 mg/kg or 8 mg/kg), with a total cumulative exposure of 9414.3 patient-years (PY). There were 26 cases of GI perforations reported for patients with RA from Roche TCZ clinical trials and long-term extension studies. The rate for lower GI perforation was 1.9 per 1000 PY (Table). The mean age of the 26 individuals experiencing a perforation was 63 years at the initiation of the trials. The majority of patients who perforated were taking chronic corticosteroids (up to 10 mg/day prednisone), one or more NSAIDs, and methotrexate. Three deaths were secondary to GI perforations or associated complications, one of which was an iatrogenic esophageal perforation. Sixteen of the 18 (89%) patients with colonic perforations had diverticulitis. The majority were recognized only after the perforation had occurred. The GI perforation rate in patients with RA who were exposed to corticosteroids in the United Health Care database was 3.9 per 1000 PY (95% CI, 3.1–4.8), and in those exposed to anti-TNFs, it was 1.3 per 1000 PY (95% CI, 0.8–1.9).

Anatomic Distribution of GI Perforation Cases in RA Patients Exposed to TCZ in Clinical Trials or Open Label Extensions

Anatomic LocationNumberRate per 1000 Patient-Yearsa
Upper GI  
    Stomach and duodenum1 
    Jejunum and ileum3 
    Abdominal abscess1 
Lower GI  
      Diverticular (including abscess, fistula)16b1.7
a Total exposure = 9414.3 patient-years.
b One patient had a diverticular perforation in the jejunum and was not included as a colon diverticular perforation case.


Most patients receiving TCZ who experienced GI perforation were also receiving corticosteroids, NSAIDS, and methotrexate. The rate of GI perforations in patients treated with TCZ in the Roche clinical trials was within the range of the rates observed in populations of RA patients treated with corticosteroid and/or anti-TNF therapy in the United Health Care database.

To cite this abstract, please use the following information:
van Vollenhoven, Ronald F., Keystone, Edward C., Furie, R., Blesch, A., Wang, C., Curtis, J. R.; Gastrointestinal Safety in Patients with Rheumatoid Arthritis Treated with Tocilizumab: Data From Roche Clinical Trials [abstract]. Arthritis Rheum 2009;60 Suppl 10 :1613
DOI: 10.1002/art.26687

Abstract Supplement

Meeting Menu