RABBIT Score for Infection Risk Validated in Second Population

A novel risk score for predicting serious infections in individual patients with rheumatoid arthritis who are treated with either tumor necrosis factor–alpha inhibitors or conventional disease-modifying antirheumatic drugs showed high agreement between observed and expected rates of serious infection, said Dr. Angela Zink at the annual European Congress of Rheumatology.

The score, known as the RABBIT risk score, is named for the German biologics register from which patients were enrolled. The risk score was previously evaluated in 5,044 patients enrolled in the RABBIT register between May 2001 and December 2006, and was shown to provide detailed information about the expected absolute risk of serious infection findings (Ann. Rheum. Dis. 2011;70:1914-20).

The new findings demonstrate the value of the score in a new population of patients not included in the previous analysis because they enrolled in the register after January 2009. Taken together, the two sets of data suggest that the score will serve as a useful tool with which rheumatologists can identify patients at increased risk of infection. By knowing the score, rheumatologists can avoid the use of treatment combinations that confer an increased risk for serious infection, said Dr. Zink of the German Rheumatism Research Center, Berlin.

"We showed that the risk of an individual patient at any point in time results from background risks such as age, comorbidity, or treatment history, and from risks that change over time, such as the activity of the disease, functional status, comedication with glucocorticoids, or treatment with TNF inhibitors or conventional DMARDs," Dr. Zink said, noting that multivariate methods were used to calculate the relative contributions of each of these factors. Dr. Zink presented the results for Dr. Anja Strangfeld, a colleague at the German Rheumatism Research Center.

For the analysis the investigators used data from 1,327 RA patients who were treated with TNF-alpha inhibitors and 1,276 RA patients treated with conventional DMARDs. The patients included in this analysis were enrolled in the RABBIT register at the start of their treatment with anti-TNF-alphas or DMARDs. The investigators used the RABBIT score to determine the serious infection risk, and they compared the observed serious infection rate with the expected rate in the total cohort and certain defined subgroups.

Agreement between the observed and expected rates of serious infection was high, as was agreement in subgroups of patients with a higher risk of infections. For example, the observed and expected rates of serious infection in those taking a TNF-alpha inhibitor were 3.2 and 3.0 serious infections per 100 person-years, respectively, and the observed and expected rates of serious infection in those taking a conventional DMARD were 1.3 and 1.5 serious infections per 100 person-years, respectively. Among those at high risk (based on the presence of one or more risk factors) who were taking a TNF-alpha inhibitor, the observed and expected rates were 5.2 and 4.3 serious infections per 100 person-years, respectively, and among those at high risk who were taking a conventional DMARD, the observed and expected rates were 1.7 and 1.9 serious infections per 100 person-years, respectively.

"We think that long-term observational studies of new treatments offer excellent opportunities to understand what happens during the treatment of real-life patients," Dr. Zink said of the value of biologics registers in general.

These data allow us to balance risks against each other, she added.

"For instance, an almost fivefold risk of serious infection was found in patients who had to be treated with dosages of glucocorticoids of 15 mg/day or above, compared to those with less than 7.5mg/day. TNF inhibition, which was associated with a 1.8-fold increased overall risk of infection, can lead to a decrease in risk if the disease activity is successfully suppressed, thereby allowing the glucocorticoid dose to be reduced," she explained.

The researchers did not disclose any conflicts of interest. This research was supported by a joint unconditional grant from Abbott, Amgen/Swedish Orphan Biovitrum, Bristol Myers Squibb, Essex/MSD, Pfizer, Roche, and UCB.

A novel risk score for predicting serious infections in individual patients with rheumatoid arthritis who are treated with either tumor necrosis factor–alpha inhibitors or conventional disease-modifying antirheumatic drugs showed high agreement between observed and expected rates of serious infection, said Dr. Angela Zink at the annual European Congress of Rheumatology.

The score, known as the RABBIT risk score, is named for the German biologics register from which patients were enrolled. The risk score was previously evaluated in 5,044 patients enrolled in the RABBIT register between May 2001 and December 2006, and was shown to provide detailed information about the expected absolute risk of serious infection findings (Ann. Rheum. Dis. 2011;70:1914-20).

The new findings demonstrate the value of the score in a new population of patients not included in the previous analysis because they enrolled in the register after January 2009. Taken together, the two sets of data suggest that the score will serve as a useful tool with which rheumatologists can identify patients at increased risk of infection. By knowing the score, rheumatologists can avoid the use of treatment combinations that confer an increased risk for serious infection, said Dr. Zink of the German Rheumatism Research Center, Berlin.

"We showed that the risk of an individual patient at any point in time results from background risks such as age, comorbidity, or treatment history, and from risks that change over time, such as the activity of the disease, functional status, comedication with glucocorticoids, or treatment with TNF inhibitors or conventional DMARDs," Dr. Zink said, noting that multivariate methods were used to calculate the relative contributions of each of these factors. Dr. Zink presented the results for Dr. Anja Strangfeld, a colleague at the German Rheumatism Research Center.

For the analysis the investigators used data from 1,327 RA patients who were treated with TNF-alpha inhibitors and 1,276 RA patients treated with conventional DMARDs. The patients included in this analysis were enrolled in the RABBIT register at the start of their treatment with anti-TNF-alphas or DMARDs. The investigators used the RABBIT score to determine the serious infection risk, and they compared the observed serious infection rate with the expected rate in the total cohort and certain defined subgroups.

Agreement between the observed and expected rates of serious infection was high, as was agreement in subgroups of patients with a higher risk of infections. For example, the observed and expected rates of serious infection in those taking a TNF-alpha inhibitor were 3.2 and 3.0 serious infections per 100 person-years, respectively, and the observed and expected rates of serious infection in those taking a conventional DMARD were 1.3 and 1.5 serious infections per 100 person-years, respectively. Among those at high risk (based on the presence of one or more risk factors) who were taking a TNF-alpha inhibitor, the observed and expected rates were 5.2 and 4.3 serious infections per 100 person-years, respectively, and among those at high risk who were taking a conventional DMARD, the observed and expected rates were 1.7 and 1.9 serious infections per 100 person-years, respectively.

"We think that long-term observational studies of new treatments offer excellent opportunities to understand what happens during the treatment of real-life patients," Dr. Zink said of the value of biologics registers in general.

These data allow us to balance risks against each other, she added.

"For instance, an almost fivefold risk of serious infection was found in patients who had to be treated with dosages of glucocorticoids of 15 mg/day or above, compared to those with less than 7.5mg/day. TNF inhibition, which was associated with a 1.8-fold increased overall risk of infection, can lead to a decrease in risk if the disease activity is successfully suppressed, thereby allowing the glucocorticoid dose to be reduced," she explained.

The researchers did not disclose any conflicts of interest. This research was supported by a joint unconditional grant from Abbott, Amgen/Swedish Orphan Biovitrum, Bristol Myers Squibb, Essex/MSD, Pfizer, Roche, and UCB.

A novel risk score for predicting serious infections in individual patients with rheumatoid arthritis who are treated with either tumor necrosis factor–alpha inhibitors or conventional disease-modifying antirheumatic drugs showed high agreement between observed and expected rates of serious infection, said Dr. Angela Zink at the annual European Congress of Rheumatology.

The score, known as the RABBIT risk score, is named for the German biologics register from which patients were enrolled. The risk score was previously evaluated in 5,044 patients enrolled in the RABBIT register between May 2001 and December 2006, and was shown to provide detailed information about the expected absolute risk of serious infection findings (Ann. Rheum. Dis. 2011;70:1914-20).

The new findings demonstrate the value of the score in a new population of patients not included in the previous analysis because they enrolled in the register after January 2009. Taken together, the two sets of data suggest that the score will serve as a useful tool with which rheumatologists can identify patients at increased risk of infection. By knowing the score, rheumatologists can avoid the use of treatment combinations that confer an increased risk for serious infection, said Dr. Zink of the German Rheumatism Research Center, Berlin.

"We showed that the risk of an individual patient at any point in time results from background risks such as age, comorbidity, or treatment history, and from risks that change over time, such as the activity of the disease, functional status, comedication with glucocorticoids, or treatment with TNF inhibitors or conventional DMARDs," Dr. Zink said, noting that multivariate methods were used to calculate the relative contributions of each of these factors. Dr. Zink presented the results for Dr. Anja Strangfeld, a colleague at the German Rheumatism Research Center.

For the analysis the investigators used data from 1,327 RA patients who were treated with TNF-alpha inhibitors and 1,276 RA patients treated with conventional DMARDs. The patients included in this analysis were enrolled in the RABBIT register at the start of their treatment with anti-TNF-alphas or DMARDs. The investigators used the RABBIT score to determine the serious infection risk, and they compared the observed serious infection rate with the expected rate in the total cohort and certain defined subgroups.

Agreement between the observed and expected rates of serious infection was high, as was agreement in subgroups of patients with a higher risk of infections. For example, the observed and expected rates of serious infection in those taking a TNF-alpha inhibitor were 3.2 and 3.0 serious infections per 100 person-years, respectively, and the observed and expected rates of serious infection in those taking a conventional DMARD were 1.3 and 1.5 serious infections per 100 person-years, respectively. Among those at high risk (based on the presence of one or more risk factors) who were taking a TNF-alpha inhibitor, the observed and expected rates were 5.2 and 4.3 serious infections per 100 person-years, respectively, and among those at high risk who were taking a conventional DMARD, the observed and expected rates were 1.7 and 1.9 serious infections per 100 person-years, respectively.

"We think that long-term observational studies of new treatments offer excellent opportunities to understand what happens during the treatment of real-life patients," Dr. Zink said of the value of biologics registers in general.

These data allow us to balance risks against each other, she added.

"For instance, an almost fivefold risk of serious infection was found in patients who had to be treated with dosages of glucocorticoids of 15 mg/day or above, compared to those with less than 7.5mg/day. TNF inhibition, which was associated with a 1.8-fold increased overall risk of infection, can lead to a decrease in risk if the disease activity is successfully suppressed, thereby allowing the glucocorticoid dose to be reduced," she explained.

The researchers did not disclose any conflicts of interest. This research was supported by a joint unconditional grant from Abbott, Amgen/Swedish Orphan Biovitrum, Bristol Myers Squibb, Essex/MSD, Pfizer, Roche, and UCB.