Registries are an established method for monitoring the long-term safety of new approved antirheumatic drugs.1 Pivotal studies demonstrate the effectiveness and safety of drugs during a set observation period in selected groups of patients. As a complementary approach, registries allow limitless observation of relevant pathologic events in individuals on treatment with new drugs, without restriction in the number of medications or comorbidities. In fact, registries are determinant in defining the real safety profile of medications in actual clinical situations. Because of the great regional variability in the epidemiology of diseases, especially infectious ones, local registries are extremely important to detect risk of specific adverse events in particular populations. The relationship between tumor necrosis factor inhibitors (TNFi's) biologics and risk of tuberculosis was first established by a registry study in a country with a relatively high incidence of the disease.2 Furthermore, the effectiveness of a patient screening strategy to reduce that specific risk was confirmed by the same registry.3 In 2009, the Brazilian Society of Rheumatology (BSR) implemented its own registry for active monitoring of biologic disease-modifying antirheumatic drugs (bDMARDs) in rheumatic diseases.4 The strong points of this initiative were as follows: (1) since the beginning, it has been an institutional project, with the BSR as sponsor and owner of the data, ensuring scientific independence; (2) a solid methodology, with defined protocol and procedures manual5; (3) a 3-level data quality control5; (4) a platform in common with the established Spanish registry BIOBADASER, which was also available for all other Latin America countries (BIOBADAMERICA project)6; and (5) through the BSR, the project was opened to all Brazilian rheumatology units to achieve national representability, including patients from all the country's states. Both controlled and registry studies pointed out that infectious adverse events are the most frequent in patients exposed to bDMARDs.7,8 This study aimed to define the incidence of serious infectious adverse events (SIAEs) in patients with rheumatoid arthritis (RA) and spondyloarthritis (SpA) monitored in the Brazilian Registry for Biologic Drugs (BIOBADABRASIL).
MATERIAL AND METHODS
BIOBADABRASIL is an observational, longitudinal, multicenter project with no time limitation. Patients have been included by 32 rheumatology units from almost all Brazilian states. A 3-domain online platform was used for data entry: (1) demographics, disease characteristics, comorbidities, and infectious diseases screening; (2) treatment; and (3) adverse events, with outcome information. Disease diagnosis, drug indication, and inclusion in the registry were decisions of the principal investigator of each center, but always according to predetermined guidelines. Patients could be included if they met the following criteria: (1) diagnosis of any rheumatic disease initiating the first bDMARD within the previous 3 months; (2) RA or SpA diagnosis starting a new synthetic disease-modifying antirheumatic drug (sDMARD), up to 3 months, with no previous exposure to bDMARDs (internal control group); and (3) signed informed consent. After inclusion, patients were on continuous follow-up. Data were compulsorily updated in case of an adverse event or treatment modification, registering the cause and the relationship with the drug in use. Adverse events were classified according to the Medical Dictionary for Regulatory Activities (MedDRA).9 Definitions of severity and outcome of adverse events were stated in the BIOBADABRASIL protocol.5 A serious adverse event (SAE) required notification and was defined as a condition that causes death or is life-threatening, implies inpatient hospitalization or prolongation of an existing one, and involves persistent or significant disability or a congenital abnormality/birth defect. Pregnancy was included among SAEs. Serious infectious adverse events were considered all SAEs with clinical characteristics of infection, ideally with the identification of the causative agent. Outcome of adverse events was categorized as follows: unknown, recovered with sequelae, recovered without sequelae, not recovered, death related to the event, death possibly related to the drug in use, and death with no relationship to the drug. Serious adverse events were assigned to a drug if they occurred during drug therapy or within a 90-day period after the last dose. If an event could be associated with several drugs, it was associated with all of them according to international recommendations.1 A specifically trained monitor of the BSR maintained a constant 3-level data quality control program: (1) digital, using the platform resources; (2) by phone, contacting patients every 6 months; and (3) in loco, yearly, comparing registry data and medical files of 20% of patients randomly selected in each center. The local ethical committee of each center approved the study, and all patients signed the informed consent. BIOBADABRASIL is sponsored by the BSR, with funds from the different pharmaceutical companies marketing biological compounds in Brazil. In this study, we limited the analysis to patients with diagnosis of RA or SpA (ankylosing spondylitis and psoriatic arthritis) collected from January 2009 to June 2015. For a general perspective of bDMARDs SIAEs in our population, data of patients with RA and SpA were analyzed jointly and compared with the control group. To compare SIAE incidence in RA and SpA, we focused only on TNFi bDMARDs, because, until 2015, only adalimumab, etanercept, and infliximab were provided for SpA in our country. Subsequent TNFi was defined as the second or further biologic of this class used by the same patient. Afterward, RA data were analyzed independently to obtain a more consistent comparison of SIAE incidence between the bDMARDs and sDMARDs groups, because 92% of the latter is composed of RA patients, mainly on methotrexate and/or leflunomide. In RA, the non-TNFi's (abatacept, rituximab, tocilizumab) were analyzed as a group, due to their relatively small numbers.
Time of exposure was set from initiation of the drug to the date of last administration plus twice the half-life or censorship. If there was overlap of multiple treatments for 1 patient, then a lag-window of 3 months was accounted for each biologic treatment.
Continuous variables were expressed as mean (SD), and categorical variables were expressed in absolute and percentage values. Student t and χ2 tests were used to compare variables between groups with and without serious infections. The SIAE incidence rate (IR) was calculated per 1000 patient/years with 95% confidence interval (CI) and IR ratio (IRR) estimated between groups. The significance level was set at 0.05. The Poisson regression multivariable model was used to estimate adjusted IRR using age, sex, disease duration, corticosteroids, diabetes, and smoking status as confounding factors.
The general characteristics of patients in the BIOBADABRASIL registry, as of June 2015, are presented in Table 1. The total included 1698 subjects with RA (1121, 66%) and SpA (577, 34%) on biologic drugs, with 7119 patient/years and follow-up time of 2.8 (SD 2.2) years. Controls were 572 (RA, 528 [92%]; SpA, 44 [8%]), with 2093 patient/years and follow-up time of 3.6 (SD 2.2). In the biologic group, 1601 (94%) received a TNFi and 97 (6%) a non-TNFi as the first biologic. Controls were mainly on methotrexate (85%), leflunomide (40%), or both drugs (35%). Data comparing characteristics of RA and SpA patients on TNFi to the control group are shown in Table 2. Patients on sDMARDs had, in general, a shorter disease duration. In RA, the mean Disease Activity Score (DAS28) was similar in bDMARDs and controls while 11% on TNFi had no background sDMARD at baseline.
The overall IR of SIAEs for bDMARDs was 36 per 1000 patient/years (95% CI, 31–40; 253 infections) and for TNFi was 35 per 1000 patient/years (95% CI, 30–40; 218 infections) versus 15 per 1000 patient/years (95% CI, 10–21; 31 infections) for controls (IRR, 2.4 [95% CI, 1.65–3.49]; p < 0.001; IRR, 2.34 [95% CI, 1.6–3.5]; p < 0.001, respectively; adjusted IRR, 2.85 [95% CI, 1.94–4.17]; p < 0.001, and adjusted IRR, 2.96 [95% CI, 2.01–4.36]; p < 0.001, respectively). The IR on TNFi was higher in RA, 43 per 1000 patient/years (95% CI, 37–50), than in SpA 21 per 1000 patient/years (95% CI, 16–28; IRR, 0.5 [95% CI, 0.36–0.69]; p < 0.001), but statistical differences disappear after adjusted IRR of 0.96 (95% CI, 0.64–1.44; p = 0.837). An increased SIAE frequency was found when the subsequent TNFi treatment was compared with the first, 31 per 1000 patient/years (95% CI, 26–36) versus 50 per 1000 patient/years (95% CI, 39–64; IRR, 1.6 [95% CI, 1.17–2.17]; p = 0.0013, and adjusted IRR, 1.55 [95% CI, 1.15–2.08]; p = 0.004). There were limited >statistically significant differences in SIAEs between the most prescribed TNFi: adalimumab versus infliximab (adjusted IRR, 0.71 [95% CI, 0.52–0.99]; p = 0.044) or no differences between etanercept versus infliximab (IRR, 1.12 [95% CI, 0.81–1.55]; p = 0.481).
Considering only RA patients, the SIAE incidence for TNFi versus controls was 43 per 1000 patient/years (95% CI, 37–50) versus 14 per 1000 patient/years (95% CI, 9–20; adjusted IRR, 3.06; p < 0.001). Adalimumab showed lower SIAE IR compared with infliximab, 29 per 1000 patient/years (95% CI, 22–39) versus 55 per 1000 patient/years (95% CI, 43–70; adjusted IRR, 0.52 [95% CI, 0.35–0.76]; p = 0.001). There was no statistically significant difference between etanercept and infliximab (adjusted IRR, 0.84 [95% CI, 0.59–1.21]; p = 0.353). Subsequent TNFi treatment confirmed a tendency toward a higher SIAE rate when compared with the first, 50 per 1000 patient/years (95% CI, 37–68) versus 41 per 1000 patient/years (95% CI, 34–48), but without statistical significance (adjusted IRR, 1.19 [95% CI, 0.84–1.70]; p = 0.333). The SIAEs IR for non-TNFi bDMARDs prescribed as the first biologic was 23 per 1000 patient/years (95% CI, 11–48). Serious infectious adverse events, in all groups, were statistically associated with baseline age, sex, corticosteroid use, or smoking status, but surprisingly not with disease duration or diabetes.
All RA subgroups (TNFi, non-TNFi, and controls) had the highest frequency of serious infections in the respiratory and urinary tracts, as well as SpA patients on TNFi. Spondyloarthritis controls registered only serious infections of the urinary tract. No central nervous system infection was reported in any group. Detailed data are shown in Table 3.
As to the outcome of the 154 serious infections events registered during the first biologic treatment, 124 were classified as recovered without sequelae, 13 as recovered with sequelae, 16 as not recovered, and one as fatal.
There were only 3 fatal infectious events registered, all of which were in RA patients exposed to non-TNFi bDMARDs, and were possibly related to the treatment, corresponding to an IR of 3 per 1000 patient/years (95% CI, 0–23). Of note, 2 of 3 occurred in patients exposed to a non-TNFi biologic but with a previous exposure to a TNFi.
The higher risk of SIAEs in patients using biologic agents is well recognized in all rheumatic conditions. However, most of the registry studies data are from European and North American countries.7,8 Furthermore, SIAE data in SpA are scarce. This is the first study focusing on SIAEs in RA and SpA on bDMARDs in a Latin American country, demonstrating that the safety profile is similar to that defined by registries in other parts of the world.
Registries reflect the timeline of bDMARDs availability in a particular country. The Brazilian Public Health System covers 100% of the population and made available adalimumab, etanercept, and infliximab first, for RA and SpA.10,11 For that reason, the large majority of our patients on biologics were using 1 of these 3 medications.
When patients with RA and SpA exposed to TNFi were considered, a higher incidence of serious infections was found, with an adjusted IRR of 2.96 versus sDMARDs. This increased risk is consistent with Spanish (BIOBADASER), Argentinian (BIOBADASAR), and Mexican (BIOBADAMEX) registry studies.12–14 BIOBADASAR, compared with their internal control group on sDMARDs, found an SIAE IRR of 1.66 (95% CI, 1.38–2.0; p < 0.05) (unpublished observations). BIOBADAMEX also found a higher frequency of adverse events in the TNFi group, with infections being the most prevalent cause.14
We found a statistically significant increased risk of SIAEs with a subsequent versus first TNFi (adjusted IRR, 1.55). Given that we could not find similar published data to verify our findings, more research is needed on this point. However, some indirect data from the literature make our findings plausible. Although a study from BIOBADASER did not focus on infections, it showed not only a lower drug survival rate for a second TNFi treatment but also that adverse events were an important reason for drug discontinuation.15 It is important to emphasize that, in this mixed RA-SpA group, no relevant differences in SIAEs were found between adalimumab and infliximab (IRR, 0.71) or etanercept and infliximab (IRR, 1.12).
Our internal cohort of patients treated with sDMARDs comprises 92% of RA patients. Therefore, we recognize that our most consistent data came from analyzing treatments in RA. In these patients, a 3.06-fold greater risk of SIAEs was found in TNFi group (43 per 1000 patient/years vs 14 per 1000 patient/years for controls). This statistically significant increase of frequency of serious infections in the TNFi group was also seen in other longitudinal studies.16–20 The German registry found 6.42 per 100 patient/years for etanercept, 6.15 per 100 patient/years for infliximab, and 2.28 per 100 patient/years for controls.16 The REAL Japanese registry showed 6.42 and 2.64 per 100 patient/years for TNFi and controls, respectively.17 The British registry demonstrated a very similar IR in the TNFi group (42 per 1000 patient/years) but a higher IR of SIAEs in controls (32 per 1000 patient/years).18
Our study showed that a first non-TNFi biologic had a lower frequency of SIAEs than the first TNFi (23 vs 41 per 1000 patient/years). At the moment, few registry data about non-TNFi and infections are available. We could find one study with tocilizumab and one with rituximab versus TNFi that showed a higher IR of serious infections with the non-TNFi (10 vs 3.03 and 11 vs 3.1 per 100 patient/years, respectively).21,22 It is important to consider that our data include abatacept, which could be, at least partially, responsible for the finding that the first non-TNFi biologic had a lower SIAE frequency than the first TNFi in our study population. Noticeably, a recent meta-analysis of randomized controlled trials and long-term extension studies with biologics in RA, reported an IR of SIAEs per 100 patient/years of 4.9 for TNFi, 5.45 for tocilizumab, 3.72 for rituximab, and 3.04 for abatacept.23 In RA patients, adalimumab was associated with a lower SIAE IR compared with infliximab (IRR, 0.52), whereas no significant difference was found between etanercept and infliximab. In contrast, Lampropoulos et al19 demonstrated a trend toward higher risk of serious infection with adalimumab, although not statistically significant, Galloway et al18 showed no difference in SIAEs between adalimumab, infliximab, and etanercept, whereas van Dartel et al24 found a statistically significant lower risk of serious infections among patients on etanercept therapy. These heterogeneous findings may be due to patients' individualized drug prescription at each study site.
Spondyloarthritis patients on TNFi had half the SIAE risk in relation to RA patients receiving the same treatment (IRR, 0.5) but statistical differences disappear after adjusting the data for age, sex, disease duration, corticosteroids intake, diabetes, and smoking (IRR, 0.96). Few data comparing infections in SpA and RA patients on biologic drugs are available. A lower SIAE incidence in SpA versus RA had been found in a meta-analysis of randomized controlled trials and in a cohort study, compared with literature data.25,26
In our study, the most frequent site of infection in RA patients, for both TNFi and controls, was the respiratory tract, followed by the urinary tract and skin/soft tissue. The respiratory tract was the predominant SIAE location in the majority of the studies.16,17,20,24,27 Skin/soft tissue infections were more frequent in several other registry studies than they were in BIOBADABRASIL.16,17,19,24,27,28 Spondyloarthritis patients on TNFi therapy had more SIAEs in the respiratory tract, as did the Wallis et al26 cohort.
In RA, baseline disease duration and diabetes were not associated with serious infections risk in BIOBADABRASIL. Otherwise, Curtis et al27 has found that diabetes was an independent risk factor to develop infections requiring hospitalization. Age was associated with SIAEs as well, in agreement with data from the United Kingdom, Japan, and the United States.17,18,27 Corticosteroid intake was associated with SIAEs. We could not further explore this finding because no information on dosage is available in our dataset. Two American studies showed an association with prednisone use greater than 10 mg daily.20,27
Fatal infectious adverse events were rare and occurred only in RA patients on non-TNFi, although this group had a lower SIAE IR. In patients on tocilizumab, Sakai et al21 found a nonstatistically significant higher risk of overall fatal events, when compared with TNFi.
Our study has relevant strengths. Our data are derived from a registry within an accurate methodological framework and a rigid quality control. An internal parallel control group is another strong point of our database. Moreover, the leadership of the project by the BSR constitutes a guarantee of scientific autonomy. This study shares the same limitations of some others based on registry data. The main one is that the inclusion of patients and events is at principal investigator discretion, being a potential source of bias. In addition, our database covers only a small proportion of patients using biologics in a continental country, namely, Brazil. Another limitation is the lack of information about background corticosteroids and sDMARDs dosage and exposure time for patients in bDMARDs.
In BIOBADABRASIL, biologic drugs, especially the subsequent TNFi, were associated with a higher risk of serious infections compared with synthetic DMARDs. Spondyloarthritis showed a tendency toward a lower risk than RA, but it was not proven in our study. Corticosteroid intake and age represented risk factors for SIAEs, whereas diabetes did not. Even if biologic therapy in Brazil, generally, seems to be as safe as it is in other countries, constant monitoring is required to follow the safety profile of drugs in the clinical setting of rheumatic conditions.
The authors acknowledge Patricia Cabral for the tireless and excellent monitoring work travelling our huge country.
A sincere acknowledgment to other BIOBADABRASIL investigators for their valuable contribution to the database: Luiz Sérgio Guedes Barbosa, MD, Universidade Federal do Mato Grosso; Inês Guimarães Silveira, MD, Pontifícia Universidade do Rio Grande do Sul; André Luiz Hayata, MD, Clínica de Reumatologia de Osasco; Adriana Maria Kakehasi, MD, Universidade Federal de Minas Gerais; Marcelo Medeiros Pinheiro, MD, Universidade Federal Escola Paulista de Medicina; Aline Ranzolin, MD, IMIP Fundação Martiniano Fernandes–Recife; Morton A. Scheinberg, MD, Hospital Abreu Sodré–São Paulo; Flavio Sztajnbok, MD, Universidade Estadual do Rio de Janeiro.
Finally, the authors acknowledge Mr James R. Welsh for the valuable grammatical review.
1. Dixon WG, Carmona L, Finckh A, et al. EULAR points to consider when establishing, analysing and reporting safety data of biologics registries
in rheumatology. Ann Rheum Dis
2. Gómez-Reino JJ, Carmona L, Rodriguez-Valverde V, et al. Treatment of rheumatoid arthritis
with tumor necrosis factor inhibitors may predispose to significant increase in tuberculosis risk: a multicenter active-surveillance report. Arthritis Rheum
3. Carmona L, Gomez-Reino JJ, Rodriguez-Valverde V, et al. Effectiveness of recommendations to prevent reactivation of latent tuberculosis infection in patients treated with tumor necrosis factor antagonists. Arthritis Rheum
4. Titton DC, Silveira IG, Louzada-Junior P, et al. Brazilian biologic registry: BiobadaBrasil implementation process and preliminary results. Rev Bras Reumatol
5. SBR. Manual Biobadabrasil. December, 2009. Available at: https://biobadaser.ser.es/biobadamerica/Brasil/cgi-bin/upload/documentacion.aspx
. Accessed September 24, 2016.
6. Carmona L, De La Vega M, Ranza R, et al. BIOBADASER, BIOBADAMERICA, and BIOBADADERM: safety registers sharing commonalities across diseases and countries. Clin Exp Rheumatol
. 2014;32(Suppl. 85):S163–S167.
7. Ramiro S, Gaujoux-Viala C, Nam IL, et al. Safety of synthetic and biological DMARDs: a systematic literature review informing the 2013 update of the EULAR recommendations for the management of rheumatoid arthritis
. Ann Rheum Dis
8. Curtis JR, Jain A, Askling J, et al. A comparison of patients characteristics and outcomes in selected European and US Rheumatoid Arthritis registries
. Semin Arthritis Rheum
9. MedDRA dictionary. March 1, 2016. Available at: http://www.meddra.org/
. Accessed September 25, 2016.
10. Brazil. Lei No. 8080/90. September 19, 1990. Brasília: DF. 1990. Available at: https://www.planalto.gov.br/ccivil_03/LEIS/L8080.htm
. Accessed March 6, 2018.
11. Brazil. Portaria No. 66. November 1, 2006. Brasília: DF. 2006. Available at: http://bvsms.saude.gov.br/bvs/saudelegis/sctie/2006/prt0066_01_11_2006_comp.html
. Accessed March 6, 2018.
12. Perez-Sola MJ, Torre-Cisneros J, Perez-Zafrilla B, et al. Infections in patients treated with tumor necrosis factor antagonists: incidence, etiology and mortality in the BIOBADASER registry. Med Clin (Barc)
13. De La Vega M, Gomez G, Casado C, et al. Tercer reporte de eventos adversos con tratamientos biológicos en Argentina. Informe de registro BIOBADASAR. Rev Arg Reumatol
14. Ventura-Rios L, Bañuelos-Ramirez D, Hernández-Quiroz Mdel C, et al. Patient survival and safety with biologic therapy. Results of the Mexican National Registry Biobadamex 1.0. Reumatol Clin
15. Gomez-Reino JJ, Carmona L. BIOBADASER Group. Switching TNF antagonists in patients with chronic arthritis: an observational study of 488 patients over a four-year period. Arthritis Res Ther
16. Listing J, Strandfeld A, Kary S, et al. Infections in patients with rheumatoid arthritis
treated with biologic agents. Arthritis Rheum
17. Komano Y, Tanaka M, Nanki T, et al. Incidence and risk factors for serious infection in patients with rheumatoid arthritis
treated with tumor necrosis factor inhibitors: a Report from the Registry of Japanese Rheumatoid Arthritis
Patients for longterm safety. J Rheumatol
18. 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 (Oxford)
19. Lampropoulos CE, Orfanos P, Bournia VK, et al. Adverse events and infections in patients with rheumatoid arthritis
treated with conventional drugs or biologic agents: a real world study. Clin Exp Rheumatol
20. Greenberg J, Reed G, Kremer JM, et al. Association of methotrexate and tumour necrosis factos antagonists with risk of infectious outcomes including opportunistic infections in the CORRONA registry. Ann Rheum Dis
21. Sakai I R, Cho SK, Nanki T, et al. Head-to-head comparison of the safety of tocilizumab and tumor necrosis factor inhibitors in rheumatoid arthritis
patients (RA) in clinical practice: results from the registry of Japanese RA patients on biologics for long-term safety (REAL) registry. Arthritis Res Ther
22. Cobo-Ibañez T, Descalzo MA, Loza-Santamaria E, et al. Serious infections in patients with rheumatoid arthritis
and other immune-mediated connective tissue diseases exposed to anti-TNF or rituximab: data from the Spanish registry BIOBADASER 2.0. Rheumatol Int
23. Strand V, Ahadieh S, French J, et al. Systematic review and meta-analysis of serious infections with tofacitinib and biologic disease-modifying antirheumatic drug treatment in rheumatoid arthritis
clinical trials. Arthritis Res Ther
24. Van Dartel SA, Frasen J, Kievit W, et al. Difference in the risk of serious infections in patients with rheumatoid arthritis
treated with adalimumab, infliximab and etanercept: results from the Dutch Rheumatoid Arthritis
Monitoring (DREAM) registry. Ann Rheum Dis
25. 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
26. Wallis D, Thavaneswaran A, Haroon N, et al. Tumour necrosis factor inhibitor therapy and infection risk in axial spondyloarthritis
: results from a longitudinal observational cohort. Rheumatology
27. 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
28. Dixon WG, Watson K, Lunt TM, et al. Rates of serious infection, including site-specific and bacterial intracellular infection, in rheumatoid arthritis
patients receiving anti-tumor necrosis factor therapy: results from the British Society for Rheumatology Biologics Register. Arthritis Rheum