Objectives: The aim of this study was to investigate the risk of Tbc among patients with rheumatoid arthritis (RA) receiving anti-TNF therapy and compare it with a control RA group and with a non-RA reference population. In this study we also examined the compliance with the national consensus report published by the Turkish Rheumatology Research and Education Association for the prevention of Tbc infections in patients on anti-TNF therapy.

Methods: The electronic prescription database of the Pension Fund of Turkey which is one of the 3 main state run health insurance systems in Turkey, was searched for the period from January 2000 to June 2005. 103.459.230 prescriptions of 2.652.013 health beneficiaries registered in the database were reviewed to determine the patients who were prescribed TNF inhibitors or anti-Tbc drugs. Patients who were prescribed anti-TNF agents and anti-Tbc drugs compatible with a known treatment protocol were considered to have tuberculosis. Other anti-Tbc drug prescriptions were considered to be given prophylactically. A control group of 2353 patients who were registered in the database to have RA, but had never been on biologic therapy was formed for comparison. Tbc incidence for all the health beneficiaries of the Pension Fund registered in the database was also estimated. Standardized incidence ratio (SIR) was calculated using this population as a reference population. In the general Pension Fund population anyone who had been prescribed isoniazid along with rifampicin at least one time were considered to have Tbc (in Turkey, after a diagnosis of Tbc, patients are provided free anti-Tbc medication by Tbc dispensaries when they referred). Tbc incidence rates were standardized according to the general Turkish population and StatsDirect program was used for the statistical analysis.

Results: Standardized Tbc incidence rate for the whole Pension Fund population was 26/100.000 (crude rate 24/100.000), which is very similar to the World Health Organization's estimate. Two cases of Tbc were detected in the RA control group (crude rate 42/100.000, SIR= 1,29 (95% CI = 0,16- 4.66; p=0.45). Six Tbc cases were identified among 424 RA patients (346,3 patient-years). However, in 3 of them Tbc had occurred before the initiation of biological therapy. Crude Tbc incidence rate during anti-TNF therapy was 866/100.000 (SIR= 28.9 (95% CI = 5,96-84.41; p = 0,0002)). All three patients were on infliximab (crude rate 1.496/100.00, SIR= 49.8 (95% CI = 10.3-145.5; P < 0, 0001). The mean onset of tuberculosis after anti-TNF therapy was 111 ± 25 days. None of the patients had received prophylactic treatment which was found to be given to 92 of the 424 RA patients (22%). Prophylaxis rate was similar for all three agents. For prophylaxis, all patients were prescribed isoniazid, except for three. In 23% of the patients, prophylactic treatment was started after 30 days following the initiation of biologic therapy. Duration of prophylaxis was less than 6 months in the majority of patients (65%).

Conclusion: The increased rate of Tbc in patients taking TNF inhibitors is not unexpected. But, it is noteworthy that there appears to be an increased risk of Tbc in these patients even before the onset of biologic therapy. It is of major concern that in a significant proportion of patients, prophylactic treatment is initiated with a significant delay and prescribed for a shorter duration than recommended by the national consensus report.

---