Risk of Lower Gastrointestinal Perforation More than Double With Certain Biologics

Patients with rheumatoid arthritis (RA) who take tocilizumab or tofacitinib may have more than double the risk of lower gastrointestinal perforation (GIP) compared to patients taking tumor necrosis factor inhibitors (TNFis), according to a study reported in Arthritis & Rheumatology.¹

Compared with the general population, patients with RA may have an increased risk for GIP.²  Previously, it was established that such patients were most commonly affected by perforations of the upper GI tract, which have been linked with the use of nonsteroidal anti-inflammatory drugs (NSAIDs).² However, there have been a growing number of lower GIPs in recent years, particularly in association with tocilizumab since its 2010 FDA approval for RA treatment.³

The initial data from the clinical trials pertaining to the drug’s development show 26 cases of GIP, while no cases were observed in the placebo group. Because the trials allowed non-responders to switch to active treatment after 16 weeks, however, there was a minimal amount of placebo exposure. “Thus neither a stable background rate, nor a comparator rate for patients receiving alternative RA agents, was available,” noted the authors of the new study.

Since that time, the newest class of RA drugs–Janus kinase (JAK) inhibitors–have been introduced. “Given the pleitropic nature of this molecule and others select inhibitors of JAK1, JAK2, JAK3, or tyrosine protein kinase 2, with some downstream effect on interleukin(IL)-6 signaling, the potential for an increased risk for GIP exists,” according to the new review.¹ Because there has been limited comparative or real-world evidence regarding the risk of GIP from this new class of medications, researchers at the University of Alabama at Birmingham and McGill University in Montreal compared risk between 167,109 adult RA patients using a variety of RA drugs. 

Patient information was drawn from Medicare and Marketscan databases from 2006 to 2014. Patients were excluded if they had a previous diagnosis of GIP, as well as inflammatory bowel disease or non-melanoma skin cancer, since these could affect RA drug choice and risk of GIP.

The main exposures in the study were the JAK-inhibitor tofacitinib, the IL-6 inhibitor tocilizumab, and other biologics, while several TNF-inhibitors (TNFi) comprised the reference group. The primary outcome was hospitalized GIP as indicated by hospital discharge diagnosis codes.

Results show that GIP incidence per 1,000 person-years was 1.29 for tofacitinib, 1.55 for tocilizumab, 1.10 for abatacept, 0.73 for rituximab, and 0.84 for TNFis. Approximately 62% of these cases were lower-GI perforations. After adjusting for potentially confounding variables including age, race, sex, concurrent medications, and comorbid conditions, the researchers found“a significantly increased risk for lower tract GIP amongst the tocilizumab users (hazard ratio [HR]=2.55, 95% confidence interval [CI] 1.16-3.73) and tofacitinib users (HR=3.24, 95%CI 1.05-10.04).

Predictors of lower tract GIP were older age (HRs=1.16 per 5 years, 95%CI 1.10-1.22), diverticulitis/other gastrointestinal conditions (HR=3.25, 95%CI 1.62-6.51), and prednisone use>7.5mg/day (HR=2.24, 95%CI1.36-3.70. Of the 108 patients who experienced lower GIP, 17 died in the hospital, resulting in a mortality rate of 16%.

While upper GIP has been linked with NSAIDs, experts are less clear on the causes of lower GIP, though “some evidence suggests drug-related toxicity may extend into the lower GI tract,” and mechanisms leading to lower GIP could possibly include breaches in host defenses. 

Summary and Clinical Applicability

Patients taking tocilizumab and tofacitinib for RA had more than twice the risk of lower-GI perforations than patients taking TNFi. Although absolute rates of perforation and differences between treatments were low (1-3 per 1000 patient years), GIP is an expensive and potentially fatal adverse event. 

It “is clearly only one consideration in evaluating the overall risk-benefit profile of choosing a specific therapy,” the researchers concluded. 

Limitations and Disclosures

Due to the types of databases used the researchers could not clinically confirm cases, and they were not able to measure other potentially confounding factors including RA severity and duration and use of over-the-counter NSAIDs.

This project had no external funding.

Author disclosures | JRC: Research grants and/or consulting for unrelated work with Amgen, Abbvie, BMS, CORRONA, Crescendo, Janssen, Pfizer, Roche, UCB | SB: Supported by the Drug Safety and Effectiveness Network, Canadian Institutes of Health Research | HY: research grants from Amgen