This was a 74-week prospective study on the safety of infliximab in patients that had active RA despite treatment with MTX and at least one other DMARD. This study was longer than most randomized clinical trials that have examined the safety of infliximab (mean 0.8 years) [16
]. In this study, the most common reason for discontinuing treatment was an AE, of which infections and IRs were the most frequent causes. We also found that AEs as well as discontinuations for AEs most often occurred during the first 26 weeks of treatment.
Our study confirmed that infections are the most common type of AE in RA patients receiving the combination of infliximab and MTX [20
]. Approximately half of all infections were considered SAEs, and infections were also the most common type of SAE. The rate of infections considered SAEs in this study (7.4%; (42 of 575 patients)) was similar to that reported using the same dose of infliximab in ATTEST (Abatacept or infliximab vs placebo, a Trial for Tolerability, Efficacy and Safety in Treating rheumatoid arthritis) (8.5%) [26
]. In addition, after IRs, infections were the second leading cause of discontinuation. Also, infections led to five of the nine deaths. The most common type of infection was non-tuberculosis bacteria, although viral infections were also common. This agrees with data from the Swedish practice-based registry ARTIS (AntiRheumatic Therapies In Sweden), which indicate that there is a slight increase in the risk of infection in RA patients treated with anti-TNF-α agents but that it is not driven by any particular type of infection [27
]. A recent meta-analysis of randomized clinical trials by Leombruno and colleagues, however, did not find an increased risk of serious infections in RA patients treated with recommended doses of anti-TNF-α therapies [16
]. Similar to Takeuchi and colleagues [28
], we found that older patients were more likely to have infections, although the odds ratio was close to 1.0. It is also noteworthy that all patients that died from infections were at least 69 years old. Finally, although our study confirms that infections are a reason for concern in refractory RA patients, we cannot determine whether the risk for infection or death due to an infection was increased by treatment with infliximab.
With regard to infections, of particular concern is the increased risk for tuberculosis in patients treated with infliximab, which is generally thought to be due to a lack of compliance with recommendations to prevent reactivation of latent tuberculosis infections [29
]. In the current study, there were four cases of tuberculosis (three confirmed, one suspected). All four were in patients with negative Mantoux tests and chest X-rays at screening. One of the confirmed cases of tuberculosis appeared to be a new case caused by exposure to a family member with overt tuberculosis. The remaining could have been new cases of tuberculosis, but they may have also been due to latent infections that went undetected by the screening tests [30
]. Interestingly, all four cases of tuberculosis were in patients taking corticosteroids at baseline, which could have masked the Mantoux test or caused further suppression of the patient's immune system. Regardless of the reasons for these infections, we concur with the conclusion of Theis and Rhodes [30
] that, despite screening and efforts to treat latent infections, clinicians need to carefully monitor for the emergence of tuberculosis infections in patients receiving anti-TNF-α therapies.
In addition to infection, IRs are common in patients treated with anti-TNF-α therapies and are a frequent reason for discontinuation [31
]. In this study, IRs were the second-most common type of AE. In nearly half of these cases (42.8%), the IRs caused treatment discontinuation, although, in many cases, the IR causing discontinuation was not considered an SAE. In agreement with Kapetanovic and colleagues [32
], age, sex, and HAQ results were not risk factors for IRs. In contrast to their report, however, we did not find an association between IRs and age at diagnosis/onset or longer disease duration.
Some early studies suggest that anti-TNF-α agents may increase the risk of malignancies, especially lymphoma [13
]. However, this is not supported by a more recent meta-analysis of clinical trial data or more recent data from clinical registries [16
]. Nevertheless, we paid close attention to the appearance of malignant tumors. There were four cases of malignant tumors, three of which led to treatment discontinuation. However, there were no cases of lymphoma, all four were different tumor types, and there were no obvious relations between the incidence of tumors and any of the patient characteristics.
We also paid close attention to the incidence of cardiovascular AEs because RA patients are at increased risk [33
]. Cardiovascular events accounted for 4 of 74 treatment discontinuations, and they were the fourth most common AE overall. They also accounted for two of the nine deaths. Despite the importance of cardiovascular events, there is good evidence that anti-TNF-α therapies reduce the risk in patients with RA to the level in the non-RA population [33
AEs overall, SAEs, and IRs were most common during the first 26 weeks of treatment. We found an association with higher age and the appearance of AEs overall. Also, higher age was a predictor of SAEs, infections, and discontinuation due to a SAE, but the odds ratios were all close to 1.0. Otherwise, we did not identify significant risk factors for AEs overall, SAEs, infections, or IRs in this study.
One of the key aims of this study was to identify reasons for discontinuation in RA patients treated with infliximab. In the evaluable population, the continuation rate at 74 weeks was 68.7%. This is comparable with most other studies of daily clinical practice, which have shown one-year continuation rates between 65% and 73% and two-year continuation rates between 67% and 75% [21
]. The continuation rate in the current study was lower than the one-and two-year rates (91% and 81%, respectively) in a previous multicenter study carried out in Belgium [39
]. This difference was partly due to the fact that dose increases were possible in the previous study but not here. In addition, the current study took place after etanercept and adalimumab became available, so that patients had the option of switching to alternative anti-TNF-α therapies. Thus, patients would have been more likely in the current study to discontinue treatment if they or the investigator were uncomfortable with the AEs or the level of efficacy.
Treatment discontinuations were most frequent during the first 26 weeks. The AEs most frequently leading to discontinuation were IRs, followed by infections. Baseline characteristics, including age, did not appear to predispose patients to discontinuation due to an AE. Higher age was a significant predictor of discontinuation due to an SAE, but the odds ratio was close to 1.0. Similarly, Chevillotte-Maillard et al.
reported no difference in discontinuation rates (median one-year follow-up) or drug survival curves between older and younger patients treated with infliximab [40
Infections were also most common during the first 26 weeks of the study. This agrees with the findings of the ARTIS study, where the risk of infection was highest in the first year [27
]. We suspect that this was due to the discontinuation of susceptible patients rather than an adaptation to the treatment. This is supported by the fact that discontinuation for any AE was most common during the first 26 weeks. Moreover, using data from a registry of British patients, Dixon and colleagues showed that the risk of serious infection is highest in the first six months after the initiation of anti-TNF-α therapies and that the reduction in risk thereafter is associated with physicians excluding patients considered at high risk [41
]. Regardless of the reason for the lower risk for infection with time, some risk is always present, so physicians should remain vigilant during the course of treatment with infliximab or any other anti-TNF-α therapy.
Prior to beginning the study, we speculated that the use of corticosteroids would reduce the frequency of IRs and increase the frequency of infections. However, our analysis showed that the use of corticosteroids was not associated with a difference in the likelihood of AEs overall, SAEs, IRs, allergic skin reactions, or infections, nor did it appear to influence the likelihood of discontinuation due to AEs or IRs. These results suggest that patients can continue corticosteroid use during treatment with infliximab, if indicated, without increasing the chance for discontinuation or occurrence of an AE, including infections. The results also suggest that corticosteroids do not prevent infliximab-induced IRs. Notably, all four reported cases of tuberculosis were in patients taking corticosteroids at baseline, and all had negative Mantoux tests at screening. Thus, it is possible that the corticosteroids masked the Mantoux results or increased the risk for tuberculosis infections in these patients by suppressing their immune systems.