LONDON – Although a potentially serious complication, perforation of the gastrointestinal tract is rare, judging from the findings of an analysis of more than 140,000 patients with rheumatoid arthritis.
Upper or lower GI tract perforation occurred in 696 (0.5%) of patients, with an overall, unadjusted incidence rate of 1.7 cases per 1,000 person-years, according to a report by Dr. Jeffrey Curtis.
"For [most of] the rheumatoid arthritis patients someone has in their practice, [GI perforation] is going to be very uncommon; I think that should be reassuring," Dr. Curtis said during an interview at the annual European Congress of Rheumatology.
"We also observed that there were cases [of GI perforation] for every biologic group," added Dr. Curtis, a rheumatologist, epidemiologist, and associate professor of medicine at the University of Alabama at Birmingham. This should help dispel any concerns that the adverse event might occur with certain biologic agents used to treat RA, he suggested.
In a retrospective study, Dr. Curtis and his associates analyzed records of 143,433 RA patients from a large U.S.-based administrative claims database for the years 2001-2009. The investigators used a validated algorithm to identify cases of upper and lower GI perforation and to determine predictive factors. The median follow-up was 2.5 years.
Older age was found to be a predictor of GI perforation, with adjusted relative risks of 1.6 and 2.1 for people aged 40-64 years and 65 years, respectively, compared with RA patients younger than 40 years. The mean age of the 142,737 patients who did not have a GI perforation was 57.6 years, and the mean age of the 696 patients who did was 62 years (P less than .01).
Diverticulitis and diverticulosis without diverticulitis were also significantly more common in patients who experienced a GI perforation than in those who did not, although the incidence was still low, with rates of 2.9% vs. 0.3% and 1.4% vs. 0.4%, respectively (both P less than .01). Diverticulitis but not diverticulosis was a significant risk factor for perforation.
The main risk factors among the RA medication groups of most relevance were the use of oral glucocorticosteroids and nonsteroidal anti-inflammatory drugs (NSAIDs), not biologics and not really the disease-modifying antirheumatic drugs (DMARDs), Dr. Curtis said in the interview. Indeed, the incidence of GI perforation was highest in patients who used glucocorticoids in combination with DMARDs other than methotrexate (3.03 per 1,000 patient years).
Steroid monotherapy carried an incidence of 2.86 per 1,000 patient-years. Steroids used in combination with methotrexate and biologics also increased the risk of the GI complications (2.24 and 1.87 per 1,000 patient-years, respectively).
The rates of GI perforations in patients treated with biologics, methotrexate, or other DMARDs without steroids were 1.02, 1.08, and 1.71 per 1,000 patient-years, respectively. NSAID use was associated with an incidence rate of 1.68 per 1,000 patient-years.
"In contrast to about 10 years ago, when there was a lot more GI bleeding with perforation in the upper GI tract, there are population studies now suggesting that we need to be more worried about perforations of the lower GI tract," Dr. Curtis said.
In the study, 80% of the perforations seen were in the lower GI tract, so the use of gastroprotective medications may not be that useful.
"I think that the relative contribution of NSAIDs [to GI perforation] is probably diminished, just because we are seeing more lower now than upper."
Although still important, upper GI bleeding and peptic ulcer disease are perhaps less critical than antecedent diverticulitis and its associated complications.
Minimization of NSAID and steroid use is probably warranted in higher-risk patients, Dr. Curtis advised. "In somebody with a history of diverticulitis, I would be very cautious," he noted.
Dr. Curtis disclosed research and consulting relationships with Abbot, Amgen, BMS, Centocor, CORRONA, Crescendo, Pfizer, Roche, and UCB.