AMSTERDAM – , according to an analysis of three large databases presented at the European Congress of Rheumatology.
“When we combined the databases, the incidence of any malignancy excluding nonmelanoma skin cancer was 13.09 per 1,000 patient years in the tocilizumab group and 13.46 in the TNF-inhibitor group,” reported Seoyoung C. Kim, MD, ScD, of the division of pharmacoepidemiology & pharmacoeconomics at Brigham and Women’s Hospital, Boston.
Dr. Seoyoung C. Kim
The study was conducted with data from 10,393 adult RA patients treated with tocilizumab and 26,357 patients treated with TNFi in the Medicare, QuintilesIMS PharMetrics Plus, and Truven Health MarketScan databases. All patients were new starts on tocilizumab or the TNFi on which they were evaluated, but all were required to have been exposed to at least one different biologic prior to starting the treatment. A diagnosis of RA at least 365 days prior to inclusion in this analysis was required to rule out prevalent cancers, which was an exclusion criterion.
More than 60 covariates were employed in the analysis to minimize the risk of confounders. These included demographics, RA characteristics, comorbidities, and other medications.
There also was no difference in the rates of the 12 most common cancer types when those exposed to tocilizumab were compared with those exposed to TNFi in a secondary analysis of these data, according to Dr. Kim. When expressed as hazard ratios, there were some numerical differences in relative risk among these cancers on both as-treated and intention-to-treat analyses, but confidence intervals were large, and none approached significance.
RA itself has been associated with an increased risk of some malignancies, such as lung cancer, but the relationship between the proinflammatory state of RA, its treatments, and the risk of cancer has been unclear, according to Dr. Kim. She said, “There is some concern relative to use of TNFi or other biologics in regard to developing malignancy, but studies have been inconsistent.”
Dr. Kim conceded that a lack of data on patients’ disease duration or activity is one limitation of this analysis. Another is that residual confounding can never be ruled out from a retrospective analysis. However, she said that, because the two biologics were compared for the same indication in patients exposed to at least one previous biologic, the confounding may be less than it would be if tocilizumab was compared with a conventional synthetic disease modifying antirheumatic drug (csDMARD), such as methotrexate. Again, there also was a requirement for exposure to at least one prior biologic, and this also is reassuring for the final conclusion.
“In other words, even among RA patients who were exposed to more than one biologic, the risk of cancer was similar between tocilizumab and TNF-inhibitor initiators,” Dr. Kim reported.
Roche provided funding for the study. Dr. Kim reports financial relationships with Bristol-Myers Squibb, Pfizer, and Roche.