Background There is still controversy about the association of TNF-alpha antagonists with malignancies. Although recent meta-analyses suggest that the risk of malignancy with these agents is not increased, the suspicion still exists, at least as reflected by the current labeling information.
Objectives To survey the incidence of malignancy in patients treated with TNF-alpha antagonists for up to 13 years in a single center and to compare the results with the incidence of malignancy in the general population.
Methods The charts of the first consecutive 376 rheumatoid arthritis (RA) patients (307 women; 69 men; mean age 55.5±12.7 SD years) who were prescribed TNF-alpha antagonists between 2001 and 2009 were reviewed retrospectively. Information regarding the demographic and clinical features, duration of use each TNF-alpha antagonist, development time and type of malignancy and the outcomes of patients were recovered from patient charts. Patients were invited to the clinic if their current status was not known. Age and sex standardized cancer incidence rates in the general population used to calculate standardized cancer rates (SIRs) were obtained from the Turkish Ministry of Health, Public Health Institution, Department of Cancer.
Results The mean duration of follow-up was 67.6±17.7 SD months from the start of the first TNF-alpha antagonist to the end of 2013. During this time, 30 (7.9%) patients had died, 132 (35.1%) were still on TNF-alpha antagonists, 67 (17.8%) were using another biologic, 138 (36.7%) were no longer using any biologics and 10 (2.6%) were lost to follow-up. The main reasons of death were cardiovascular causes in 7 patients, sepsis in 5 and malignancies in 4. In 10 (2.6%) patients cause of death was unknown. A total of 8 patients had developed malignancies (2.12% [95% CI 0.66-3.58]). These were solid cancers in 6 (breast =1, lung =1, gastric =1, rectum =1, colon =1, unknown primary=1), hematologic cancers in 2 (acute leukemia=2). Since there were patients who were lost to follow-up and those who died of unknown causes, we calculated SIRs according to three different scenarios. If we assume that none of the patients who were lost to follow-up or whose cause of death was unknown, had a malignancy, the SIR would be 0.78 (95% CI 0.34-1.42). If we assume that all of the patients who were lost to follow-up (n=10) had a malignancy, but none of the patients whose cause of death was unknown had a malignancy, the SIR would be 1.7 (95% CI 1.04-2.0). Finally if we assume all patients who were lost to follow-up or whose reason of death was unknown (n=20) had a malignancy the SIR would be 2.7 (95% CI 1.8-3.8).
Conclusions Our data indicate that an increased incidence of cancer, as compared to the general population, cannot be ruled out among RA patients enrolled in a TNF – alpha antagonist registry. This uncertainty is compounded by the fact that the comparisons we make here are between the incidence in a cohort of practically cancer free patients at the time of cohort entry and a general population incidence in which no such restriction is present.
Disclosure of Interest None declared