WASHINGTON – Anti–tumor necrosis factor therapy does not increase the risk of lymphoma in patients with rheumatoid arthritis.
That’s the conclusion of a study presented in a plenary session here by Dr. Kimme L. Hyrich on Monday, Nov. 12, at the annual meeting of the American College of Rheumatology.
"The challenge in studying whether therapies for rheumatoid arthritis (RA) patients can increase the risk of lymphoma is the knowledge that the disease itself has been associated with this outcome," said Dr. Hyrich.
"It’s possible that with immunosuppression, we may see an increased risk of lymphoma, but equally, if we can control the disease activity that has been associated with this outcome, it’s possible that we actually can see a decrease in lymphoma risk over time."
Dr. Hyrich, of the arthritis research UK epidemiology unit at the University of Manchester (England), and colleagues looked at patients from the British Society for Rheumatology Rheumatoid Arthritis Register (BSRBR-RA), a prospective registry of RA patients taking biologic drugs, recruited between 2001 and 2009.
The treatment group comprised 11,987 patients who started therapy with an anti–tumor necrosis factor (TNF) including infliximab, etanercept, or adalimumab; they were compared to 3,465 patients who were treated solely with nonbiologic disease modifying antirheumatic drugs (DMARDs).
Incident cancers were detected either by flagging in a national cancer database (95% of cancers were identified by this method, according to Dr. Hyrich) in one of six monthly patient and physician questionnaires, or in annual physician questionnaires thereafter.
Subjects were followed until Sept. 30, 2010, or their first lymphoma, or death, whichever came first.
The cutoff date of the study was 2010 because there can be up to a 2-year lag between incident cancer and reporting of those data in the registry, due to rigorous confirmatory processes, said Dr. Hyrich.
The investigators found 84 incident lymphomas over the study period: 20 in the nonbiologic DMARD patients and 64 in the anti-TNF cohort, for a rate of 152 per 100,000 person years versus 96 per 100,000 person-years, respectively.
The tally included five cases of Hodgkin’s lymphomas in the nonbiologic DMARD group and nine in the anti-TNF patients. Among the non-Hodgkin’s lymphoma cases, diffuse large B cell lymphomas accounted for the greatest proportion of cancers.
After adjustment for baseline age, gender, disease activity score, health assessment questionnaire results, disease duration, smoking, and current or previous cyclophosphamide use, the hazard ratio for all lymphomas for anti-TNF users was calculated to be a nonsignificant 1.13 (95% confidence interval, 0.55-2.31).
Similarly, looking at non-Hodgkin’s lymphoma only, Dr. Hyrich determined an adjusted hazard ratio among the anti-TNF patients of 1.26, also nonsignificant (95% CI, 0.58-2.72).
In her presentation, Dr. Hyrich commented that the study was unable to assess risk associated with a particular anti-TNF agent, given that the majority of patients in this real-world cohort had a history of treatment with multiple drugs in the class; indeed, 15% of the cohort had a history of treatment with three or more anti-TNF agents.
"How do you attribute risk in a situation where patients have been exposed to all of these agents?" she said.
"I think further analysis of these cohorts is needed to know whether any exposure to anti-TNFs, or any particular order of anti-TNFs, or any particular duration of anti-TNFs" increases risk.
The researchers stated that they had no disclosures relative to this presentation.