Rheumatology Advance Access published August 13, 2014
RHEUMATOLOGY
263
Concise report
doi:10.1093/rheumatology/keu255
Tumour necrosis factor inhibitor therapy and infection risk in axial spondyloarthritis: results from a longitudinal observational cohort
Objectives. Long-term data on infection risk in axial SpA (axSpA) are sparse. TNF inhibitors (TNFis) are increasingly being used in axSpA, with infection being the most important adverse event. We aimed to investigate the frequency of infections in axSpA and to identify factors predisposing to infection. Methods. Data were extracted from a longitudinal observational cohort of patients with axSpA. Infection rates were calculated and multivariate analysis was performed to investigate the association of independent variables with infection. Results. Data were analysed for 440 patients followed for a total of 1712 patient-years (pys). A total of 259 infections, of which 23 were serious, were recorded in 185 patients. The overall rate of any infection was 15 (95% CI 13, 17)/100 pys and the serious infection rate was 1.3 (95% CI 0.9, 2.0)/100 pys. There was no significant difference in the rate of any infection or serious infection in patients on TNFis compared with patients never on biologic agents. In the multivariate analysis, DMARD treatment, but not TNFi treatment, was associated with risk of infection. Age, disease duration, smoking status, BASFI, BASDAI, co-morbidity score and hospitalization were not associated with an increased risk of infection. Conclusion. The serious infection rate in axSpA in this observational cohort is low when compared with rates reported in other rheumatic diseases. Biologic use was not a significant risk factor for serious infection. Key words: axial spondyloarthritis, ankylosing spondylitis, infection, biologic.
Introduction Axial SpA (axSpA) is a chronic, inflammatory disease characterized by back pain, spinal ankylosis, peripheral arthritis and extra-articular manifestations. TNF inhibitors (TNFis) are effective in those patients who do not respond to first-line pharmacological management with NSAIDs and have become an important part of the management of AS. In RA, serious infections are an important adverse event of TNFis, with an incidence of 410/100 patient-
1 Division of Rheumatology, Toronto Western Hospital, University of Toronto, Ontario, Canada.
Submitted 15 January 2014; revised version accepted 28 April 2014. Correspondence to: Robert D. Inman, Division of Rheumatology, Toronto Western Hospital, 399 Bathurst Street, Toronto M5T 2S8, ON, Canada. E-mail: [email protected]
years (pys) [13]. The risk of infection is greatest in the first year of treatment [2]. Long-term data on infection risk associated with biologics in axSpA are sparse. There is an impression that the risk of infection appears to be lower than in RA [46]. Infection risk is considered when selecting patients for TNFi therapy, which is increasingly prescribed for axSpA patients, both those meeting modified New York criteria for AS and those with non-radiographic axSpA. Most of the current data on infection risk in AS is drawn from randomized controlled trials (RCTs) with small numbers of patients followed for short periods of time and in which selected patients may not be representative of the general AS population because co-morbidities are excluded. In this study we aimed to investigate the frequency and types of infection in patients with axSpA and to identify factors that predispose patients to infection.
! The Author 2014. Published by Oxford University Press on behalf of the British Society for Rheumatology. All rights reserved. For Permissions, please email: [email protected]
1
CLINICAL SCIENCE
Abstract
Downloaded from http://rheumatology.oxfordjournals.org/ at University of California, San Diego on September 30, 2014
Dinny Wallis1, Arane Thavaneswaran1, Nigil Haroon1, Renise Ayearst1 and Robert D. Inman1
Dinny Wallis et al.
Patients and methods Patients
Statistical analyses Crude infection rates were calculated per 100 pys. Multivariate analysis was performed using a generalized estimating equation model with occurrence of infection as the dependent variable. The reduced model was obtained by stepwise regression. The independent variables tested in the model were age (1 year increase), disease duration (1 year increase), smoking status (ever/never), BASFI (1 U increase), BASDAI (1 U increase), biologic exposure at the time of infection (yes/no), glucocorticoid exposure at the time of infection (yes/no), DMARD exposure at the time of infection (yes/no), presence of co-morbidities (1 U increase in score) and history of hospitalization (yes/no). A co-morbidity score was allocated to each patient scoring one for each of the following: asthma, chronic obstructive lung disease, hypertension, angina, cardiomyopathy, congestive cardiac failure, peptic ulcer, IBD, stroke, psoriasis, liver disease, cancer and diabetes.
Results Data were analysed for 440 patients (413 with AS according to the modified New York criteria and 27 with nonradiographic axSpA) followed for a total of 1712 pys with a mean follow-up of 3.89 years (S.D. 2.26). Characteristics of the study population are shown in Table 1. A total of 259 infections, of which 23 were serious, were recorded in 185 patients. Eighty per cent of infections were bacterial and 15% viral. The most common site of infection was the lung, followed by skin, genitourinary tract, upper respiratory tract, sinus and gastrointestinal tract. Ninety per cent of infections were treated with antibiotics, of which 10% required i.v.
2
Age at baseline, mean (S.D.), years Males, n (%) Duration of AS at baseline, mean (S.D.), years Age at symptom onset, mean (S.D.), years Age at diagnosis, mean (S.D.), years History of smoking, n (%) ESR at baseline, mean (S.D.), mm/h CRP at baseline, mean (S.D.), mg/l Co-morbidity score (013), mean (S.D.) Iritis, n (%) IBD, n (%) Psoriasis, n (%) Asthma, n (%) Hypertension, n (%) Angina, n (%) Cardiomyopathy, n (%) Congestive cardiac failure, n (%) Peptic ulcer, n (%) Stroke, n (%) Liver disease, n (%) Cancer, n (%) Diabetes, n (%) Chronic obstructive lung disease, n (%) Glucocorticoid use, n (%) DMARD use, n (%) MTX, n (%) SSZ, n (%) LEF, n (%) Other, n (%) Biologic drug use, n (%) BASDAI at baseline (00), mean (S.D.) BASFI at baseline (010), mean (S.D.) BASMI at baseline (010), mean (S.D.)
38.3 (12.8) 322 (73.2%) 14.7 (10.7) 24.0 (9.7) 30.1 (11.8) 180 (40.9) 16.3 (16.9) 14.1 (22.5) 0.78 (1.03) 139 (31.6) 61 (13.9) 48 (10.9) 35 (8.0) 77 (17.5) 11 (2.5) 2 (0.5) 0 (0) 49 (11.1) 4 (0.9) 14 (3.2) 20 (4.6) 13 (3.0) 9 (2.1) 42 (9.5) 124 (28.2) 66 (15.0) 48 (10.9) 5 (1.1) 5 (1.1) 264 (60) 4.8 (2.5) 3.9 (2.8) 2.6 (2.4)
antibiotics. The mean co-morbidity score for the study population was low.
Glucocorticoid, DMARD and biologic drug use For patients using glucocorticoids during the study period (n = 42), the mean time on glucocorticoids was 1.7 years and the mean prednisone dose was 14 mg/day. For the 124 (28.2%) patients who used a DMARD (Table 1), 65 (52%) used one DMARD, 25 (20%) used two DMARDs and 3 (2.4%) used three DMARDs during the study period. TNFis were prescribed for 264 patients, of whom 61 (23%) used the drug for a maximum of 3 months during the study period, 14 (5.3%) for 612 months, 41 (15.5%) for 1224 months, 42 (15.9%) for 2436 months and 106 (40.2%) for >36 months.
Infection rates Crude infection rates and 95% CIs were calculated. The overall rate of any infection was 15 (95% CI 13, 17)/100 pys and the serious infection rate was 1.3 (95% CI 0.9, 2.0)/100 pys. No deaths due to infection were recorded. Of the 264 patients prescribed a TNFi during a total
www.rheumatology.oxfordjournals.org
Downloaded from http://rheumatology.oxfordjournals.org/ at University of California, San Diego on September 30, 2014
Patients attending the Toronto Western Hospital AS clinic are invited to be registered in the spondylitis database. All patients provide written consent to participate, in compliance with the Declaration of Helsinki, and the study was approved by the Research Ethics Board of the Toronto University Health Network. Clinical, laboratory and radiological data are collected annually according to a standardized protocol. Details of intercurrent infections (type of infection, target organ, antibiotic usage, hospitalization) are captured on each protocol visit. Cofactors that might contribute to infection risk (diabetes and other co-morbidities, glucocorticoid or DMARD use) are collected in the protocol. Data were extracted for all patients enrolled between 1 January 2003 and September 2012 diagnosed with axSpA according to the Assessment of SpondyloArthritis international Society classification criteria [7, 8] who had at least two assessments completed 1 year apart. Serious infection was defined as infection requiring i.v. antibiotics or hospitalization. All infections occurring in the year prior to study enrolment (for which details are recorded at the first visit) and those recorded at subsequent visits were included.
TABLE 1 Characteristics of the study population (n = 440)
Infection risk in axial spondyloarthritis
TABLE 2 Multivariate analysis of risk factors for infection (n = 440) Multivariate model Univariate model OR
Age Duration of disease Smoking status DMARD use Glucocorticoids Biologics BASFI BASDAI Co-morbidity score Hospitalization
0.99 1.01 1.21 1.69 1.90 1.14 1.03 1.03 1.09 1.21
95% CI (0.98, (0.99, (0.90, (1.19, (1.08, (0.86, (0.98, (0.97, (0.92, (0.69,
1.01) 1.02) 1.63) 2.39) 3.34) 1.51) 1.09) 1.10) 1.29) 2.13)
P-value
OR
0.35 0.37 0.21 0.003 0.03 0.37 0.25 0.28 0.31 0.51
0.99 1.01 1.23 1.76 1.10 1.25 0.96 1.09 1.06 0.85
95% CI (0.97, (0.99, (0.86, (1.12, (0.54, (0.90, (0.87, (0.98, (0.86, (0.41,
1.00) 1.03) 1.75) 2.76) 2.26) 1.73) 1.07) 1.22) 1.31) 1.75)
Reduced model P-value
OR
95% CI
P-value
0.17 0.40 0.26 0.01 0.80 0.19 0.46 0.11 0.58 0.66
— — — 1.73 — — — — — —
— — — (1.21, 2.48) — — — — — —
— — — 0.003 — — — — — —
OR: odds ratio.
follow-up time of 684 pys on biologic therapy with a mean follow-up of 2.59 years (S.D. 2.10), 127 infections (10 serious) were recorded in 101 patients. The rate of any infection while on a TNFi was 19 (95% CI 16, 22)/100 pys and the serious infection rate was 1.5 (95% CI 0.7, 3.0)/100 pys. Of the 186 patients never prescribed a TNFi during a total follow-up time of 651 pys with a mean follow-up of 3.50 years (S.D. 2.12), 91 infections (12 serious) were recorded in 71 patients. The rate of any infection in this group was 14 (95% CI 11, 17)/100 pys and the serious infection rate was 1.8 (95% CI 1.0, 3.0)/100 pys. There was no significant difference in the rate of any infections or serious infections in patients on TNFis compared with patients never on biologic agents.
Risk factors for infection The results of multivariate analysis are shown in Table 2. In the univariate analysis, DMARD treatment and glucocorticoid treatment were associated with an increased risk of infection. In the multivariate analysis, only DMARD treatment remained significant. TNFi treatment was not associated with an increased risk of infection. Age, disease duration, smoking status, BASFI, BASDAI, co-morbidity score and hospitalization were not associated with an increased risk of infection. Specifically, there was no difference in the frequency of IBD between patients who developed infections (29/185, 15.6%) and those without infections (32/255, 12.5%) (P = 0.35).
Discussion In this prospective axSpA cohort, the serious infection rate was low. Long-term data on infection risk in axSpA are limited. A systematic review and meta-analysis of clinical trial experience comparing the risk of infection between TNFi and placebo in AS [4] found the risk of
www.rheumatology.oxfordjournals.org
serious infections in 2202 patients with AS not exposed to immunosuppressive drugs was very low (0.4/100 pys). However, as these were patients selected for clinical trials, the real-world infection risk may be higher. In the meta-analysis of nine RCTs of TNFis, where the trial duration ranged from 12 to 30 weeks, the risk of serious infection was 1.9/100 pys in the treatment group vs 1.0/100 pys in the control group. This difference was not statistically significant. The serious infections reported were both viral and bacterial and no deaths due to infection were reported. The lack of significance may be explained by a lack of power, as the power calculation indicated that the overall sample size was insufficient to detect significance for this difference in risk. In addition, the period of observation was short for agents that are commonly used for long-term management of the disease. Burmester et al. [5] evaluated 71 global clinical trials of adalimumab across six diseases for safety and mortality rates. Four clinical trials (comprising 1684 patients) in AS were included. The rate of serious infection was 1.4/100 pys, compared with 4.6/100 pys in RA and 2.8/100 pys in PsA. No cases of tuberculosis or serious opportunistic infections were reported in AS patients. This analysis did not include control AS patients who were not treated with a TNFi. A retrospective cohort study from four large US databases investigated the risk of hospitalization for infection in patients treated with TNFi [6]. The investigators combined patients with psoriasis, PsA or AS and found the serious infection hospitalization rates to be 5.41/100 pys in the TNFi group vs 5.37/100 pys in the comparator group, with no significant difference between the groups. Baseline use of glucocorticoids in these patients was associated with a significantly increased risk of serious infections in a dose-responsive manner. The databases used in this study comprised predominantly lowincome patients from medically underserviced areas,
3
Downloaded from http://rheumatology.oxfordjournals.org/ at University of California, San Diego on September 30, 2014
Covariate
Full model
Dinny Wallis et al.
Rheumatology key messages The serious infection rate in this axial SpA cohort is low when compared with RA. . TNF inhibitors were not associated with an increased risk of infection in axial SpA. .
4
Funding: This study was partly funded by an Assessment of SpondyloArthritis international Society fellowship to D.W. Disclosure statement: D.W. was partly funded by an ASAS fellowship and received research support from Janssen Pharmaceuticals. N.H. and R.D.I. serve as consultants for Abbvie, Celgene, Pfizer, Amgen, UCB and Janssen Pharmaceuticals. All other authors have declared no conflicts of interest.
References 1 Askling J, Fored CM, Brandt L et al. Time-dependent increase in risk of hospitalisation with infection among Swedish RA patients treated with TNF antagonists. Ann Rheum Dis 2007;66:133944. 2 Galloway JB, Hyrich KL, Mercer LK et al. Anti-TNF therapy is associated with an increased risk of serious infections in patients with rheumatoid arthritis especially in the first 6 months of treatment: updated results from the British Society for Rheumatology Biologics Register with special emphasis on risks in the elderly. Rheumatology 2011;50: 12431. 3 Curtis JR, Patkar N, Xie A et al. Risk of serious bacterial infections among rheumatoid arthritis patients exposed to tumor necrosis factor alpha antagonists. Arthritis Rheum 2007;56:112533. 4 Fouque-Aubert A, Jette-Paulin L, Combescure C et al. Serious infections in patients with ankylosing spondylitis with and without TNF blockers: a systematic review and meta-analysis of randomised placebo-controlled trials. Ann Rheum Dis 2010;69:175661. 5 Burmester GR, Panaccione R, Gordon KB, McIlraith MJ, Lacerda AP. Adalimumab: long-term safety in 23 458 patients from global clinical trials in rheumatoid arthritis, juvenile idiopathic arthritis, ankylosing spondylitis, psoriatic arthritis, psoriasis and Crohn’s disease. Ann Rheum Dis 2013;72:51724. 6 Grijalva CG, Chen L, Delzell E et al. Initiation of tumor necrosis factor-alpha antagonists and the risk of hospitalization for infection in patients with autoimmune diseases. JAMA 2011;306:23319. 7 Rudwaleit M, Landewe R, van der Heijde D et al. The development of Assessment of SpondyloArthritis international Society classification criteria for axial spondyloarthritis (part I): classification of paper patients by expert opinion including uncertainty appraisal. Ann Rheum Dis 2009;68:7706. 8 Rudwaleit M, van der Heijde D, Landewe R et al. The development of Assessment of SpondyloArthritis international Society classification criteria for axial spondyloarthritis (part II): validation and final selection. Ann Rheum Dis 2009;68:77783. 9 van der Veen MJ, van der Heide A, Kruize AA, Bijlsma JW. Infection rate and use of antibiotics in patients with rheumatoid arthritis treated with methotrexate. Ann Rheum Dis 1994;53:2248. 10 Greenberg JD, Reed G, Kremer JM et al. Association of methotrexate and tumour necrosis factor antagonists with
www.rheumatology.oxfordjournals.org
Downloaded from http://rheumatology.oxfordjournals.org/ at University of California, San Diego on September 30, 2014
which may have contributed to the higher absolute rates of infection compared with previously published cohort studies. In this cohort, the serious infection rate in patients treated with TNFi was in keeping with rates reported in RCTs and TNFi use was not associated with an increased risk of infection. DMARD use (which in this population was most often MTX) appeared to be associated with an increased risk of infection. The association of MTX with infection has been most widely studied in RA, and while early work suggested that there was a higher risk of infection in patients on MTX than those not on MTX [9], later studies have not supported these findings [1012]. RA itself is associated with an infection risk greater than in the general population [13] and this may confound the interpretation of differences between treatment groups. Patients on immunosuppressants may be more likely to report infections than other patients, which could contribute to the apparent association of infection with DMARDs. In the AS population, DMARDs are often used in patients with IBD. However, a subanalysis demonstrated that patients who had reported infections did not have a significantly higher frequency of IBD than those who had never reported an infection. Further studies are required to confirm any association of infection with DMARD use in other AS cohorts. There are limitations to interpreting these observational data. We relied on self-reporting of infection by the patient, so under- or overestimation of infection rates is possible due to recall bias. However, patients on biologic therapy are counselled about the risk of infection and might be expected to recall and report infections more often than patients not on biologic therapy. It is therefore unlikely that recall bias resulted in underestimation of the effect of biologic therapy on infection risk. Patients with a high baseline risk of infection may be less likely to receive biologic therapy (selection bias) and therefore the true risk of infection on biologic therapy may be higher in unselected patients. Patients starting biologic therapy are also screened for infections prior to commencing the drug, which may explain the low rate of infection seen in this group. Our findings represent a real-worldsetting that includes AS patients with co-morbidities under long-term followup. The apparent lower rate of infection in patients with AS compared with RA may reflect several factors: the younger age and lower frequency of co-morbidities in AS, differing immunogenetic mechanisms in the respective diseases and differences in treatment regimens, with glucocorticoid and concurrent DMARD therapy being much more common in RA than AS.
Infection risk in axial spondyloarthritis
risk of infectious outcomes including opportunistic infections in the CORRONA registry. Ann Rheum Dis 2010;69: 3806. 11 Doran MF, Crowson CS, Pond GR, O’Fallon WM, Gabriel SE. Predictors of infection in rheumatoid arthritis. Arthritis Rheum 2002;46:2294300. 12 Dixon WG, Kezouh A, Bernatsky S, Suissa S. The influence of systemic glucocorticoid therapy upon the risk of
non-serious infection in older patients with rheumatoid arthritis: a nested case-control study. Ann Rheum Dis 2011;70:95660. 13 Doran MF, Crowson CS, Pond GR, O’Fallon WM, Gabriel SE. Frequency of infection in patients with rheumatoid arthritis compared with controls: a population-based study. Arthritis Rheum 2002;46: 228793.
Downloaded from http://rheumatology.oxfordjournals.org/ at University of California, San Diego on September 30, 2014
www.rheumatology.oxfordjournals.org
5
RHEUMATOLOGY
263
Concise report
doi:10.1093/rheumatology/keu255
Tumour necrosis factor inhibitor therapy and infection risk in axial spondyloarthritis: results from a longitudinal observational cohort
Objectives. Long-term data on infection risk in axial SpA (axSpA) are sparse. TNF inhibitors (TNFis) are increasingly being used in axSpA, with infection being the most important adverse event. We aimed to investigate the frequency of infections in axSpA and to identify factors predisposing to infection. Methods. Data were extracted from a longitudinal observational cohort of patients with axSpA. Infection rates were calculated and multivariate analysis was performed to investigate the association of independent variables with infection. Results. Data were analysed for 440 patients followed for a total of 1712 patient-years (pys). A total of 259 infections, of which 23 were serious, were recorded in 185 patients. The overall rate of any infection was 15 (95% CI 13, 17)/100 pys and the serious infection rate was 1.3 (95% CI 0.9, 2.0)/100 pys. There was no significant difference in the rate of any infection or serious infection in patients on TNFis compared with patients never on biologic agents. In the multivariate analysis, DMARD treatment, but not TNFi treatment, was associated with risk of infection. Age, disease duration, smoking status, BASFI, BASDAI, co-morbidity score and hospitalization were not associated with an increased risk of infection. Conclusion. The serious infection rate in axSpA in this observational cohort is low when compared with rates reported in other rheumatic diseases. Biologic use was not a significant risk factor for serious infection. Key words: axial spondyloarthritis, ankylosing spondylitis, infection, biologic.
Introduction Axial SpA (axSpA) is a chronic, inflammatory disease characterized by back pain, spinal ankylosis, peripheral arthritis and extra-articular manifestations. TNF inhibitors (TNFis) are effective in those patients who do not respond to first-line pharmacological management with NSAIDs and have become an important part of the management of AS. In RA, serious infections are an important adverse event of TNFis, with an incidence of 410/100 patient-
1 Division of Rheumatology, Toronto Western Hospital, University of Toronto, Ontario, Canada.
Submitted 15 January 2014; revised version accepted 28 April 2014. Correspondence to: Robert D. Inman, Division of Rheumatology, Toronto Western Hospital, 399 Bathurst Street, Toronto M5T 2S8, ON, Canada. E-mail: [email protected]
years (pys) [13]. The risk of infection is greatest in the first year of treatment [2]. Long-term data on infection risk associated with biologics in axSpA are sparse. There is an impression that the risk of infection appears to be lower than in RA [46]. Infection risk is considered when selecting patients for TNFi therapy, which is increasingly prescribed for axSpA patients, both those meeting modified New York criteria for AS and those with non-radiographic axSpA. Most of the current data on infection risk in AS is drawn from randomized controlled trials (RCTs) with small numbers of patients followed for short periods of time and in which selected patients may not be representative of the general AS population because co-morbidities are excluded. In this study we aimed to investigate the frequency and types of infection in patients with axSpA and to identify factors that predispose patients to infection.
! The Author 2014. Published by Oxford University Press on behalf of the British Society for Rheumatology. All rights reserved. For Permissions, please email: [email protected]
1
CLINICAL SCIENCE
Abstract
Downloaded from http://rheumatology.oxfordjournals.org/ at University of California, San Diego on September 30, 2014
Dinny Wallis1, Arane Thavaneswaran1, Nigil Haroon1, Renise Ayearst1 and Robert D. Inman1
Dinny Wallis et al.
Patients and methods Patients
Statistical analyses Crude infection rates were calculated per 100 pys. Multivariate analysis was performed using a generalized estimating equation model with occurrence of infection as the dependent variable. The reduced model was obtained by stepwise regression. The independent variables tested in the model were age (1 year increase), disease duration (1 year increase), smoking status (ever/never), BASFI (1 U increase), BASDAI (1 U increase), biologic exposure at the time of infection (yes/no), glucocorticoid exposure at the time of infection (yes/no), DMARD exposure at the time of infection (yes/no), presence of co-morbidities (1 U increase in score) and history of hospitalization (yes/no). A co-morbidity score was allocated to each patient scoring one for each of the following: asthma, chronic obstructive lung disease, hypertension, angina, cardiomyopathy, congestive cardiac failure, peptic ulcer, IBD, stroke, psoriasis, liver disease, cancer and diabetes.
Results Data were analysed for 440 patients (413 with AS according to the modified New York criteria and 27 with nonradiographic axSpA) followed for a total of 1712 pys with a mean follow-up of 3.89 years (S.D. 2.26). Characteristics of the study population are shown in Table 1. A total of 259 infections, of which 23 were serious, were recorded in 185 patients. Eighty per cent of infections were bacterial and 15% viral. The most common site of infection was the lung, followed by skin, genitourinary tract, upper respiratory tract, sinus and gastrointestinal tract. Ninety per cent of infections were treated with antibiotics, of which 10% required i.v.
2
Age at baseline, mean (S.D.), years Males, n (%) Duration of AS at baseline, mean (S.D.), years Age at symptom onset, mean (S.D.), years Age at diagnosis, mean (S.D.), years History of smoking, n (%) ESR at baseline, mean (S.D.), mm/h CRP at baseline, mean (S.D.), mg/l Co-morbidity score (013), mean (S.D.) Iritis, n (%) IBD, n (%) Psoriasis, n (%) Asthma, n (%) Hypertension, n (%) Angina, n (%) Cardiomyopathy, n (%) Congestive cardiac failure, n (%) Peptic ulcer, n (%) Stroke, n (%) Liver disease, n (%) Cancer, n (%) Diabetes, n (%) Chronic obstructive lung disease, n (%) Glucocorticoid use, n (%) DMARD use, n (%) MTX, n (%) SSZ, n (%) LEF, n (%) Other, n (%) Biologic drug use, n (%) BASDAI at baseline (00), mean (S.D.) BASFI at baseline (010), mean (S.D.) BASMI at baseline (010), mean (S.D.)
38.3 (12.8) 322 (73.2%) 14.7 (10.7) 24.0 (9.7) 30.1 (11.8) 180 (40.9) 16.3 (16.9) 14.1 (22.5) 0.78 (1.03) 139 (31.6) 61 (13.9) 48 (10.9) 35 (8.0) 77 (17.5) 11 (2.5) 2 (0.5) 0 (0) 49 (11.1) 4 (0.9) 14 (3.2) 20 (4.6) 13 (3.0) 9 (2.1) 42 (9.5) 124 (28.2) 66 (15.0) 48 (10.9) 5 (1.1) 5 (1.1) 264 (60) 4.8 (2.5) 3.9 (2.8) 2.6 (2.4)
antibiotics. The mean co-morbidity score for the study population was low.
Glucocorticoid, DMARD and biologic drug use For patients using glucocorticoids during the study period (n = 42), the mean time on glucocorticoids was 1.7 years and the mean prednisone dose was 14 mg/day. For the 124 (28.2%) patients who used a DMARD (Table 1), 65 (52%) used one DMARD, 25 (20%) used two DMARDs and 3 (2.4%) used three DMARDs during the study period. TNFis were prescribed for 264 patients, of whom 61 (23%) used the drug for a maximum of 3 months during the study period, 14 (5.3%) for 612 months, 41 (15.5%) for 1224 months, 42 (15.9%) for 2436 months and 106 (40.2%) for >36 months.
Infection rates Crude infection rates and 95% CIs were calculated. The overall rate of any infection was 15 (95% CI 13, 17)/100 pys and the serious infection rate was 1.3 (95% CI 0.9, 2.0)/100 pys. No deaths due to infection were recorded. Of the 264 patients prescribed a TNFi during a total
www.rheumatology.oxfordjournals.org
Downloaded from http://rheumatology.oxfordjournals.org/ at University of California, San Diego on September 30, 2014
Patients attending the Toronto Western Hospital AS clinic are invited to be registered in the spondylitis database. All patients provide written consent to participate, in compliance with the Declaration of Helsinki, and the study was approved by the Research Ethics Board of the Toronto University Health Network. Clinical, laboratory and radiological data are collected annually according to a standardized protocol. Details of intercurrent infections (type of infection, target organ, antibiotic usage, hospitalization) are captured on each protocol visit. Cofactors that might contribute to infection risk (diabetes and other co-morbidities, glucocorticoid or DMARD use) are collected in the protocol. Data were extracted for all patients enrolled between 1 January 2003 and September 2012 diagnosed with axSpA according to the Assessment of SpondyloArthritis international Society classification criteria [7, 8] who had at least two assessments completed 1 year apart. Serious infection was defined as infection requiring i.v. antibiotics or hospitalization. All infections occurring in the year prior to study enrolment (for which details are recorded at the first visit) and those recorded at subsequent visits were included.
TABLE 1 Characteristics of the study population (n = 440)
Infection risk in axial spondyloarthritis
TABLE 2 Multivariate analysis of risk factors for infection (n = 440) Multivariate model Univariate model OR
Age Duration of disease Smoking status DMARD use Glucocorticoids Biologics BASFI BASDAI Co-morbidity score Hospitalization
0.99 1.01 1.21 1.69 1.90 1.14 1.03 1.03 1.09 1.21
95% CI (0.98, (0.99, (0.90, (1.19, (1.08, (0.86, (0.98, (0.97, (0.92, (0.69,
1.01) 1.02) 1.63) 2.39) 3.34) 1.51) 1.09) 1.10) 1.29) 2.13)
P-value
OR
0.35 0.37 0.21 0.003 0.03 0.37 0.25 0.28 0.31 0.51
0.99 1.01 1.23 1.76 1.10 1.25 0.96 1.09 1.06 0.85
95% CI (0.97, (0.99, (0.86, (1.12, (0.54, (0.90, (0.87, (0.98, (0.86, (0.41,
1.00) 1.03) 1.75) 2.76) 2.26) 1.73) 1.07) 1.22) 1.31) 1.75)
Reduced model P-value
OR
95% CI
P-value
0.17 0.40 0.26 0.01 0.80 0.19 0.46 0.11 0.58 0.66
— — — 1.73 — — — — — —
— — — (1.21, 2.48) — — — — — —
— — — 0.003 — — — — — —
OR: odds ratio.
follow-up time of 684 pys on biologic therapy with a mean follow-up of 2.59 years (S.D. 2.10), 127 infections (10 serious) were recorded in 101 patients. The rate of any infection while on a TNFi was 19 (95% CI 16, 22)/100 pys and the serious infection rate was 1.5 (95% CI 0.7, 3.0)/100 pys. Of the 186 patients never prescribed a TNFi during a total follow-up time of 651 pys with a mean follow-up of 3.50 years (S.D. 2.12), 91 infections (12 serious) were recorded in 71 patients. The rate of any infection in this group was 14 (95% CI 11, 17)/100 pys and the serious infection rate was 1.8 (95% CI 1.0, 3.0)/100 pys. There was no significant difference in the rate of any infections or serious infections in patients on TNFis compared with patients never on biologic agents.
Risk factors for infection The results of multivariate analysis are shown in Table 2. In the univariate analysis, DMARD treatment and glucocorticoid treatment were associated with an increased risk of infection. In the multivariate analysis, only DMARD treatment remained significant. TNFi treatment was not associated with an increased risk of infection. Age, disease duration, smoking status, BASFI, BASDAI, co-morbidity score and hospitalization were not associated with an increased risk of infection. Specifically, there was no difference in the frequency of IBD between patients who developed infections (29/185, 15.6%) and those without infections (32/255, 12.5%) (P = 0.35).
Discussion In this prospective axSpA cohort, the serious infection rate was low. Long-term data on infection risk in axSpA are limited. A systematic review and meta-analysis of clinical trial experience comparing the risk of infection between TNFi and placebo in AS [4] found the risk of
www.rheumatology.oxfordjournals.org
serious infections in 2202 patients with AS not exposed to immunosuppressive drugs was very low (0.4/100 pys). However, as these were patients selected for clinical trials, the real-world infection risk may be higher. In the meta-analysis of nine RCTs of TNFis, where the trial duration ranged from 12 to 30 weeks, the risk of serious infection was 1.9/100 pys in the treatment group vs 1.0/100 pys in the control group. This difference was not statistically significant. The serious infections reported were both viral and bacterial and no deaths due to infection were reported. The lack of significance may be explained by a lack of power, as the power calculation indicated that the overall sample size was insufficient to detect significance for this difference in risk. In addition, the period of observation was short for agents that are commonly used for long-term management of the disease. Burmester et al. [5] evaluated 71 global clinical trials of adalimumab across six diseases for safety and mortality rates. Four clinical trials (comprising 1684 patients) in AS were included. The rate of serious infection was 1.4/100 pys, compared with 4.6/100 pys in RA and 2.8/100 pys in PsA. No cases of tuberculosis or serious opportunistic infections were reported in AS patients. This analysis did not include control AS patients who were not treated with a TNFi. A retrospective cohort study from four large US databases investigated the risk of hospitalization for infection in patients treated with TNFi [6]. The investigators combined patients with psoriasis, PsA or AS and found the serious infection hospitalization rates to be 5.41/100 pys in the TNFi group vs 5.37/100 pys in the comparator group, with no significant difference between the groups. Baseline use of glucocorticoids in these patients was associated with a significantly increased risk of serious infections in a dose-responsive manner. The databases used in this study comprised predominantly lowincome patients from medically underserviced areas,
3
Downloaded from http://rheumatology.oxfordjournals.org/ at University of California, San Diego on September 30, 2014
Covariate
Full model
Dinny Wallis et al.
Rheumatology key messages The serious infection rate in this axial SpA cohort is low when compared with RA. . TNF inhibitors were not associated with an increased risk of infection in axial SpA. .
4
Funding: This study was partly funded by an Assessment of SpondyloArthritis international Society fellowship to D.W. Disclosure statement: D.W. was partly funded by an ASAS fellowship and received research support from Janssen Pharmaceuticals. N.H. and R.D.I. serve as consultants for Abbvie, Celgene, Pfizer, Amgen, UCB and Janssen Pharmaceuticals. All other authors have declared no conflicts of interest.
References 1 Askling J, Fored CM, Brandt L et al. Time-dependent increase in risk of hospitalisation with infection among Swedish RA patients treated with TNF antagonists. Ann Rheum Dis 2007;66:133944. 2 Galloway JB, Hyrich KL, Mercer LK et al. Anti-TNF therapy is associated with an increased risk of serious infections in patients with rheumatoid arthritis especially in the first 6 months of treatment: updated results from the British Society for Rheumatology Biologics Register with special emphasis on risks in the elderly. Rheumatology 2011;50: 12431. 3 Curtis JR, Patkar N, Xie A et al. Risk of serious bacterial infections among rheumatoid arthritis patients exposed to tumor necrosis factor alpha antagonists. Arthritis Rheum 2007;56:112533. 4 Fouque-Aubert A, Jette-Paulin L, Combescure C et al. Serious infections in patients with ankylosing spondylitis with and without TNF blockers: a systematic review and meta-analysis of randomised placebo-controlled trials. Ann Rheum Dis 2010;69:175661. 5 Burmester GR, Panaccione R, Gordon KB, McIlraith MJ, Lacerda AP. Adalimumab: long-term safety in 23 458 patients from global clinical trials in rheumatoid arthritis, juvenile idiopathic arthritis, ankylosing spondylitis, psoriatic arthritis, psoriasis and Crohn’s disease. Ann Rheum Dis 2013;72:51724. 6 Grijalva CG, Chen L, Delzell E et al. Initiation of tumor necrosis factor-alpha antagonists and the risk of hospitalization for infection in patients with autoimmune diseases. JAMA 2011;306:23319. 7 Rudwaleit M, Landewe R, van der Heijde D et al. The development of Assessment of SpondyloArthritis international Society classification criteria for axial spondyloarthritis (part I): classification of paper patients by expert opinion including uncertainty appraisal. Ann Rheum Dis 2009;68:7706. 8 Rudwaleit M, van der Heijde D, Landewe R et al. The development of Assessment of SpondyloArthritis international Society classification criteria for axial spondyloarthritis (part II): validation and final selection. Ann Rheum Dis 2009;68:77783. 9 van der Veen MJ, van der Heide A, Kruize AA, Bijlsma JW. Infection rate and use of antibiotics in patients with rheumatoid arthritis treated with methotrexate. Ann Rheum Dis 1994;53:2248. 10 Greenberg JD, Reed G, Kremer JM et al. Association of methotrexate and tumour necrosis factor antagonists with
www.rheumatology.oxfordjournals.org
Downloaded from http://rheumatology.oxfordjournals.org/ at University of California, San Diego on September 30, 2014
which may have contributed to the higher absolute rates of infection compared with previously published cohort studies. In this cohort, the serious infection rate in patients treated with TNFi was in keeping with rates reported in RCTs and TNFi use was not associated with an increased risk of infection. DMARD use (which in this population was most often MTX) appeared to be associated with an increased risk of infection. The association of MTX with infection has been most widely studied in RA, and while early work suggested that there was a higher risk of infection in patients on MTX than those not on MTX [9], later studies have not supported these findings [1012]. RA itself is associated with an infection risk greater than in the general population [13] and this may confound the interpretation of differences between treatment groups. Patients on immunosuppressants may be more likely to report infections than other patients, which could contribute to the apparent association of infection with DMARDs. In the AS population, DMARDs are often used in patients with IBD. However, a subanalysis demonstrated that patients who had reported infections did not have a significantly higher frequency of IBD than those who had never reported an infection. Further studies are required to confirm any association of infection with DMARD use in other AS cohorts. There are limitations to interpreting these observational data. We relied on self-reporting of infection by the patient, so under- or overestimation of infection rates is possible due to recall bias. However, patients on biologic therapy are counselled about the risk of infection and might be expected to recall and report infections more often than patients not on biologic therapy. It is therefore unlikely that recall bias resulted in underestimation of the effect of biologic therapy on infection risk. Patients with a high baseline risk of infection may be less likely to receive biologic therapy (selection bias) and therefore the true risk of infection on biologic therapy may be higher in unselected patients. Patients starting biologic therapy are also screened for infections prior to commencing the drug, which may explain the low rate of infection seen in this group. Our findings represent a real-worldsetting that includes AS patients with co-morbidities under long-term followup. The apparent lower rate of infection in patients with AS compared with RA may reflect several factors: the younger age and lower frequency of co-morbidities in AS, differing immunogenetic mechanisms in the respective diseases and differences in treatment regimens, with glucocorticoid and concurrent DMARD therapy being much more common in RA than AS.
Infection risk in axial spondyloarthritis
risk of infectious outcomes including opportunistic infections in the CORRONA registry. Ann Rheum Dis 2010;69: 3806. 11 Doran MF, Crowson CS, Pond GR, O’Fallon WM, Gabriel SE. Predictors of infection in rheumatoid arthritis. Arthritis Rheum 2002;46:2294300. 12 Dixon WG, Kezouh A, Bernatsky S, Suissa S. The influence of systemic glucocorticoid therapy upon the risk of
non-serious infection in older patients with rheumatoid arthritis: a nested case-control study. Ann Rheum Dis 2011;70:95660. 13 Doran MF, Crowson CS, Pond GR, O’Fallon WM, Gabriel SE. Frequency of infection in patients with rheumatoid arthritis compared with controls: a population-based study. Arthritis Rheum 2002;46: 228793.
Downloaded from http://rheumatology.oxfordjournals.org/ at University of California, San Diego on September 30, 2014
www.rheumatology.oxfordjournals.org
5
