Objective To analyze changes in baseline characteristics of patients with very early rheumatoid arthritis (RA) over 24 years in the Early Undifferentiated Polyarthritis (EUPA) cohort.
Methods Consecutive patients with recent-onset polyarthritis fulfilling RA classification criteria recruited in EUPA were assessed at baseline. Three successive periods were defined: (1) prior to the general availability of biologics (1998-2004; 245 patients), (2) prior to the implantation of the 2010 classification criteria (2005-2010; 266 patients), and (3) the most recent decade (2011-2022; 329 patients).
Results At baseline, demographics, BMI, swollen and tender joint counts, proportion fulfilling 2010 American College of Rheumatology/European Alliance of Associations for Rheumatology criteria, modified Health Assessment Questionnaire, shared epitope status, patient-reported outcomes except pain, and patient global assessment of disease activity remained stable over the 3 periods. Despite a marked decrease in active smoking (22.2% to 12.1%), prevalence of cardiovascular comorbidities and prior cancer increased. Although duration of symptoms increased from a median of 2.9 to 4.1 months, decreases were seen in seropositivity (53.9% to 42.2%) and C-reactive protein beginning in the 2005-2010 period. A large decrease in erosive status (Sharp/van der Heijde erosion score ≥ 5; 18.3% to 9.4%) was only observed after 2011; this decrease occurred mostly in seronegative patients. Use of disease-modifying antirheumatic drugs prior to inclusion remained low and stable (25.7%), but use of oral corticosteroids increased (18% to 33.4%).
Conclusion Baseline characteristics of patients with RA evolved since 2005 toward less seropositivity and lower blood inflammation but with more comorbidities. Milder erosive damage at baseline became evident only since 2011, mostly in seronegative patients. These changes at baseline, before any intervention, suggest ongoing secular trends that may favorably affect outcomes in patients with early RA.
Key Indexing Terms:Rheumatoid arthritis (RA) is an inflammatory joint disease that affects 0.5% to 1% of adults worldwide.1,2 Chronic synovitis in multiple joints, periarticular bone destruction (erosions), and extraarticular manifestations such as rheumatoid nodules and interstitial lung disease characterize RA. The presence of rheumatoid factor (RF) and/or anticitrullinated protein antibodies (ACPA) distinguishes seropositive from seronegative RA. Patients with seropositive RA present with more extraarticular features; they also develop more severe and persistent disease and more erosions.3-5
The clinical manifestations in individual patients with RA are modulated by the interaction of environmental and genetic variables.2,6 Exposure to tobacco smoke or to inorganic dust, in the appropriate genetic context, is associated with the development of ACPA and RA.7 Female sex, tobacco smoke, and obesity all play a role in disease severity.6,7
The outcomes of patients with RA have improved over the past decades,5,8 largely because of the availability of more effective advanced treatments, notably the introduction of biologics in the early 2000s. Use of biologics started in 2001 in Quebec, Canada, but became available for patients with early RA around 2004, after the Régie de l’assurance-maladie du Québec (RAMQ), a provincial governmental agency, agreed to reimburse the drugs. A second positive change is related to earlier diagnosis facilitated by the 2010 American College of Rheumatology (ACR)/European Alliance of Associations for Rheumatology (EULAR) RA classification criteria.9 Early diagnosis must be coupled with more intensive treatment with conventional synthetic (cs) disease-modifying antirheumatic drugs (DMARDs), preferably high-dose methotrexate (MTX). MTX is frequently administered subcutaneously or in combination with other csDMARDs. Around 2011, the treat-to-target (T2T) approach was schematized, although this strategy was already being used informally.10,11 In addition to these landmark changes, we wanted to explore whether part of the improvement in outcomes observed in the more recent years could also be associated with patient changes at the initial presentation. Reports on the evolution of baseline characteristics over extended periods of time of patients with early RA are sparse and concern affluent, mostly White populations (eg, Sweden, Finland, Netherlands, and Olmsted County, Minnesota, USA).12-17 To complement the 2 reports providing baseline characteristics beyond demographics, smoking status, BMI (calculated as weight in kilograms divided by height in meters squared), and seropositivity status,13,16 we analyzed the evolution of a wide range of baseline clinical, immunogenetic, environmental, radiographic, and patient-reported characteristics over 24 years of consecutive patients with incident RA recruited into the Early Undifferentiated Polyarthritis (EUPA) cohort.
METHODSCohort description. The EUPA cohort was previously described.18-21 Starting in July 1998, consecutive adult patients with synovitis affecting ≥ 3 joints for 1 to 12 months and evaluated at the Centre hospitalier universitaire de Sherbrooke (CHUS; now Centre intégré universitaire de santé et de services sociaux [CIUSSS] de l’Estrie - CHUS) were asked to participate in a longitudinal observational study. CIUSSS de l’Estrie - CHUS is the single rheumatology referral center for approximately 500,000 people. Exclusions were presence of bacterial or crystal-induced arthritis and defined connective tissue disease or systemic vasculitis according to the ACR criteria. Enrolled patients were considered by rheumatologists as having chronic immune-mediated inflammatory polyarthritis and were probable patients with RA; therefore, they benefited from prompt management with early introduction of DMARDs, irrespective of their fulfillment of current RA criteria. Only those patients fulfilling ACR criteria for RA over the first 18 months of follow-up (representing 90% of EUPA patients) are reported here. The 2010 RA criteria were thus retrospectively applied to patients recruited before 2010. Since cohort inception, treatment was individualized with the objective of sustained control of synovitis, aiming for 0 swollen joints.
Statement of ethics and consent. We obtained ethical approval from the CIUSSS de l’Estrie – CHUS Ethics Review Board (97-04 Project EUPA). The project was explained to all patients, and all signed a consent form in order to be included in EUPA (ClinicalTrials.gov: NCT00512239).
Disease variables. Data collection was performed at the time of visit (usually the first) with the treating rheumatologist, who completed a physical examination. A trained coordinator performed a structured interview. Time of onset of arthritis was assumed to be the week or month during which the patient indicated that joint symptoms/signs had appeared or, in patients with previous musculoskeletal complaints (eg, osteoarthritis), the time when signs or symptoms of inflammatory arthritis appeared. Evaluations were scheduled relative to symptom onset to ensure homogeneity in disease duration upon subsequent follow-ups at 12, 18, 24, 36, 48, and 60 months, irrespective of symptom duration at baseline.
Variables assessed included demographics, tender joint count of 68 joints (TJC68), swollen joint count of 66 joints (SJC66), physician global assessment of disease activity (PGA), duration of morning stiffness, drug use at and between each visit, and functional status using the modified Health Assessment Questionnaire (mHAQ) with a range of 0-3 (normal function to maximal disablement).22 Psychosocial characteristics were evaluated with patient-reported outcome (PRO) questionnaires, including the Center for Epidemiologic Studies Depression Scale (CES-D)23,24 and visual analog scales (VAS; 0-10, where 0 = none and 10 = maximal) for patient global assessment of disease activity (PtGA), pain, fatigue, and sleep problems. Disease activity was measured with the Simplified Disease Activity Index (SDAI).25 Rheumatic Disease Comorbidity Index (RDCI) was calculated using self-reported comorbidities completed by chart review.26
Biologic variables included serum C-reactive protein (CRP; upper normal limit 8.0 mg/L), erythrocyte sedimentation rate (ESR; upper normal limit 20 mm/h), genomic HLA-DR typing (described below), serum IgM RF (positive when ≥ 40 IU/mL), and second-generation anticyclic citrullinated peptides (anti-CCP2; initially QuantaLite, Inova Diagnostics with a positive test > 20 U/mL, and since 2009, EuroImmun Anti-CCP ELISA with a positive test > 5 U/mL). As both anti-CCP assays use the same antigen plates, titers are convertible using the following formula: anti-CCP2 titer using EuroImmun at a maximum of 300 U = 2.5 × e0.0347 × anti-CCP2 titer using QuantaLite. Standardized hand and foot radiographs were obtained at each visit. Joint space narrowing and erosions were scored according to the Sharp/van der Heijde (SvdH) method, with a maximum score of 448 units.27 Blinded review of all radiographs was performed in known time sequences by a trained reviewer (GB). The smallest detectable difference in EUPA was previously shown to be < 5 units.18,28,29 PCR-based HLA-DR locus genotyping determined the presence of alleles of HLA-DR isotypes encoding a QKRRAA amino acid sequence motif in residues 70-74 of the HLA-DRβ chain (RA-associated shared epitopes [SEs]) and of alleles encoding the protective DERAA motif in the same residues.19
Definition of groups according to date of inclusion. Since coverage of antitumor necrosis factor drugs by the Quebec governmental health agency (and rapidly thereafter by private insurers) became accessible outside special access programs by the end of 2004, we divided the patients with early RA as before and after 2005. As T2T and novel classification criteria for RA became widely publicized in late 2010,9,11 and additional biologics (certolizumab pegol, golimumab, tocilizumab, and easier access to rituximab) became available to our patients in 2010-2011, we divided the later recruited EUPA patients as before and after 2011, yielding 3 subgroups of relatively similar sizes over 3 consecutive periods: 1998-2004, 2005-2010, and 2011-2022.
Statistical analysis. Patients who failed to fulfill 1987/2010 ACR/EULAR criteria for RA between inclusion and 18 months were excluded. Standard descriptive statistics were used for continuous and categorical variables. Baseline characteristics were compared between 3 subgroups based on years of inclusion (1998-2004, 2005-2010, and 2011-2022) using the Kruskal-Wallis test for continuous variables and chi-square test or Fisher exact test for categorical variables. The Benjamini-Hochberg false discovery rate correction was used to adjust P values for multiple comparisons. All analyses used only available data without imputation. Statistical analyses were performed with SAS software version 9.4 (SAS Institute) and GraphPad Software version 9 (GraphPad). Two-sided adjusted P < 0.05 was considered significant.
RESULTSCohort characteristics. As of October 3, 2022, 935 patients were recruited in EUPA, of which 840 fulfilling RA criteria were analyzed. This represents a mean of 34 patients (range 6-54) identified with early RA per year. Of the 95 excluded patients, 33 did not complete the baseline visit or fulfilled EUPA exclusions during follow-up and 62 (19, 12, and 31, in each consecutive period, respectively) never fulfilled RA criteria over the first 18 months of disease (Figure 1).
Flowchart of patient recruitment in EUPA over 24 years (1998-2022). ACR: American College of Rheumatology; EULAR: European Alliance of Associations for Rheumatology; EUPA: Early Undifferentiated Polyarthritis; RA: rheumatoid arthritis.
Baseline demographic and clinical characteristics. The demographic and clinical variables of the 840 patients at baseline distributed into 3 time periods are presented in Table 1. Globally, there were < 5% missing data for all variables, except for fatigue (n = 784). Numbers of patients with data for each variable and each period are presented in Supplementary Table S1 (available with the online version of this article). Age (median 62.4 years), sex (59.9% women), race (99.2% White), and BMI (median 26.6) remained stable over time. Similarly, there were no changes in SJC66 or TJC68, nor in SDAI-assessed disease activity. A stable proportion (89.8%) of EUPA patients were found to fulfill the 2010 criteria for RA, even among those included before 2010, in which those criteria were applied retrospectively.
Table 1.Baseline demographic and patient-related characteristics according to period of inclusion.
The proportion of active smokers decreased from 22.2% to 12.1% (P = 0.04). Nonetheless, the proportion of patients ever exposed to smoking remained stable at > 60%. Education significantly increased by a median of 1 year over each period. Despite constant educational and organizational efforts to facilitate evaluation of patients with early arthritis, the median duration of symptoms at inclusion increased from 2.9 to 4.1 months (P < 0.001).
Significant increases over time were also found in patient comorbidities. More recent patients had more cardiovascular (CV) comorbidities (40% to 51.1%; P = 0.03), mainly dyslipidemia (19.2% to 28.6%; P = 0.003) and hypertension (26.1% to 36.2%; P = 0.01) but not diabetes (stable at 10.7%). Prevalence of prior cancer, excluding nonmelanoma skin cancer, increased numerically (4.1% to 9.4%; P = 0.19) over time.
Although PtGA did not vary significantly over the years, EGA increased in the 2011-2022 period. Contrary to PtGA stability, reported pain levels slightly but significantly decreased from 59/100 to 53/100 (P = 0.049), essentially after 2011. CES-D and VAS for sleep problems were added in 2006 in our cohort; therefore, these data were available for the 2 most recent periods only. Despite the observed decrease in pain, there were no significant changes for variables such as functional impairment (mHAQ), fatigue, depressive symptoms, anxiety, or sleep problems.
DMARDs were used in approximately one-quarter of patients at inclusion, without significant change over time, with a median duration of 1.4 (IQR 0.5-2.8) months. In a sensitivity analysis in which patients (n = 46) with exposure to MTX for > 2 months at inclusion were removed, no significant differences were noted for any baseline characteristic. There was a significant increase over time in the prevalence of oral corticosteroid (CS) use at baseline (18% to 33.4%; P = 0.002). The proportion of patients completely treatment-naïve at inclusion fluctuated from 52.2%, 40.2%, and 50.5% in the 1998-2004, 2005-2010, and 2011-2022 cohorts, respectively.
HLA-DR alleles, antibodies, markers of inflammation, and radiographic damage. Prevalence of positive SEs (47.6%) and of DERAA alleles (27.2%) remained stable over the 3 time periods (Table 2). During that time, mean RF and anti-CCP2 levels at baseline significantly decreased, as did the prevalence of RF positivity (48.2% to 34.9%; P = 0.01) and RF and/or anti-CCP2 seropositivity (53.9% to 42.2%; P = 0.02). The decrease in anti-CCP2 positivity did not reach significance (41.2% to 33.9%; P = 0.19). Despite the known pathogenic link between smoking and SE alleles with RA autoantibodies, the only significant decrease in seropositivity occurred among SE-negative nonsmokers, from 48.7% to 12.5% (P < 0.01; Figure 2).
Table 2.Baseline genetic, serologic, and inflammatory characteristics according to period of inclusion.
Baseline seropositivity prevalence during each period of recruitment according to HLA-DR SE status and to smoking status (nonsmoker vs ever exposed to smoking). * P < 0.01. SE: shared epitope.
Over the years, the baseline prevalence of high CRP and high ESR decreased significantly, as did the levels of both ESR and CRP. As a sensitivity analysis, the changes in variables over time among CS-naïve patients are presented in Supplementary Table S2 (available with the online version of this article). Almost identical decreases in levels of antibodies, CRP, and ESR were found in the CS-naïve subset.
Baseline radiographs showed that the prevalence of elevated baseline erosion scores (ie, ≥ 5) significantly decreased from 18.3% to 9.4% (P = 0.004; Table 3). The decrease in erosive damage was obvious only for the 2011-2022 period. In contrast, elevated joint space narrowing scores (ie, ≥ 5) increased from 18.6% to 29.7% (P = 0.001) over the 2011-2022 period. As a result of these opposite changes, total SvdH scores (erosion + narrowing) did not vary significantly over time.
Table 3.Radiographic damage at baseline assessed according to the Sharp/van der Heijde score for each period of inclusion.
Lack of effect of the coronavirus disease 2019 pandemic on baseline characteristics. As described in the Methods section, 3 consecutive periods were defined a priori. Over the last decade, only the coronavirus disease 2019 (COVID-19) pandemic represented a sufficiently important event to potentially define a new period. From March 2020 to October 2020, COVID-19 restrictions prevented recruitment. Therefore, we compared patients recruited during the COVID-19 pandemic, October 2020 to October 2022 (n = 52), with a similar prepandemic period October 2017 to October 2019 (n = 37; Supplementary Table S3, available with the online version of this article). No baseline characteristic was significantly different between pandemic and prepandemic patients. Combining the patients recruited after and before the COVID-19 pandemic thus appeared appropriate.
Evolution of baseline characteristics depending on RA antibody status. We then verified if changes in baseline characteristics over time differed between seropositive and seronegative patients (Table 4A and Table 4B). Most variables did not vary over time in both subsets; similar increases were seen in education and disease duration in both, and an increase in use of oral CS occurred among seronegative patients only. Notably, the decrease in erosive damage was statistically significant only in seronegative patients. Within the entire cohort, seronegative patients were older (aged 65.9 vs 59.1 yrs; P < 0.001), were less frequently active smokers (13.2% vs 21.6%; P = 0.002), had higher comorbidities (RDCI ≥ 1, 60.1% vs 46.6%; P < 0.001), less frequently bore SEs (37.6% vs 59%; P < 0.001) but more frequently bore DERAA alleles (34.1% vs 19.3%; P < 0.001), and presented lower erosive status (erosion score ≥ 5, 12.4% vs 18.2%; P = 0.02).
Table 4A.Baseline characteristics in the 3 successive periods among seronegative patients.
Table 4B.Baseline characteristics in the 3 successive periods among seropositive patients.
DISCUSSIONOur data show that despite remarkable stability over 24 years of their initial demographic, immunogenetic, and clinical presentation, patients with recent-onset RA in the EUPA cohort became significantly more seronegative after 2005 and less erosive after 2011; the most significant decrease in erosive status was seen among seronegative patients. We also noted a significant decrease after 2005 in ESR and CRP, similar to Matthijssen et al.16 During the 24-year period, we documented environmental changes related to increased levels of education, as in Hallert et al,13 as well as higher prevalence of CV and neoplastic comorbidities and lower exposure to active smoking. Our data strongly support secular trends toward gradual changes in baseline patient characteristics, prior to any treatment. These modified baseline characteristics starting around 2005 but most apparent since 2011 are likely to interact with current therapeutic strategies to modulate outcomes of present and future patients with early RA. Although most observed changes predict better outcomes, more frequent comorbidities are likely to impose therapeutic restrictions.
Our observations are not a result of the introduction of early arthritis classification criteria in 2010. The proportions of patients fulfilling these criteria were similar before (in which the criteria were applied retrospectively) and after 2011. TJC68, SJC66, age, sex, BMI, and prevalence of SE also remained unchanged. This reflects that RA is a clinical diagnosis based on the presence of an immune-mediated polyarthritis, and that rheumatologists in the early 2000s rightly diagnosed RA without strictly adhering to current classification criteria. Similarly, no changes were observed at baseline over time for functional impairment, depressive symptoms, fatigue, and sleep perturbation.
A major decrease in seropositive status, particularly for RF, had already occurred during the 2005-2010 period. Increasing proportions of seronegative RA from the 1980s to early 2010s were previously reported,12,15,17 but not uniformly.13,14,16 Our results suggest that this trend continued into the 2020s. Increasing age of the general population was proposed as a major explanation.17 In our cohort, patients with seronegative RA were indeed older, but median age of the full cohort remained similar over time. Alternative explanations are thus needed. This shift toward seronegative RA status accompanies the decrease in current smoking prevalence among patients with early RA, paralleling similar changes in the adult Canadian population, from 25% in 1999 to 16% in 2017.30,31 Smoking is associated with positive RF and anti-CCP in patients with RA, especially in the presence of SEs.7 Although current smoking decreased, lifetime exposure to tobacco remained positive in almost two-thirds of patients during each period. Prevalence of an SE allele among seropositive patients was very stable (almost 60%) over time and remained higher than in seronegative patients. However, the most significant decrease in seropositivity occurred in never smokers without an SE allele, possibly because of less secondary smoke exposure. In the absence of tobacco exposure, other triggers likely lead to development of clinical RA, but without inducing associated antibodies. One such trigger could be immune exhaustion.32 Immune dysfunction would manifest clinically in our patients as higher rates of CV and neoplastic comorbidities, rather than by autoantibodies. It is interesting to note that inflammatory arthropathies arising as immune-related adverse events of immunotherapy are typically seronegative.33 Increases in prevalence of CV comorbidities may in part be because of improved screening. In contrast, age-adjusted cancer rates in Canada have declined by 1.5% per year in male individuals since 2011 and by 1.2% in female individuals since 2013.34,35 The increase in the number of cancers in our patients with early RA (and more so in seronegative patients) thus points toward other environmental exposures (eg, drugs, toxins) affecting development of both cancer and RA, but not of RA-associated antibodies.
A second important change in our cohort was the decrease in erosive damage at presentation, mostly seen after 2011, and more so in seronegative patients. Certainly, the combination of more seronegative status, lower inflammation, less tobacco exposure, and greater use of CS may all have contributed to the reduction in baseline erosion scores over time. However, disease duration at baseline was quite short, and exposure to treatment was minimal. This suggests that recent RA disease at onset may be intrinsically less erosion-prone than in earlier decades. This also suggests ongoing environmental influences. Although erosions became rarer, more joint space narrowing was observed over time. Joint space narrowing is associated with prior erosions and higher inflammation, whereas obesity may be slightly protective.36 Patients with early RA in EUPA were not significantly older or heavier, but they suffered from more frequent comorbidities. The increase in narrowing scores might thus be linked to the same environmental factors implicated in RA development whose relative importance increased with smoking reduction.
Finally, we also observed a decrease in the burden of inflammation biologically assessed by blood inflammatory markers and clinically reported by patients as pain. Increased use of oral CS does not appear to explain this reduction in inflammation and pain, as patients who were CS-free presented the same characteristics over time. Seronegative patients with RA were previously reported to present higher disease activity and higher inflammation at diagnosis.5,37 Less frequent seropositive status should thus have translated into higher inflammation at baseline, but we observed the contrary. Seronegative status and lower inflammatory markers potentially contributed to recent delays in referral, although the increasing duration of symptoms at baseline over time was similar in seronegative and seropositive patients.
Our study has several strengths. The first is that, over 24 years of recruitment, inclusion criteria remained uniform. Another strength is the thorough evaluation of patients, including PROs, comorbidities, radiographs, and immunogenetics. One reader (GB) of the SvdH score remained the same over the full 24 years. Third, CIUSSS de l’Estrie - CHUS is the single referral site with rheumatologists in a large administrative area, ensuring a real-world representation of de novo patients with early arthritis, relative to multicentric registries in which subsets of patients may be selected.
Our study also has limitations. First, not all consecutive patients with early RA were recruited. Some patients declined or were unable to consent, and these patients may have divergent characteristics. Only those reporting < 1 year of symptoms were included. Recall bias increases with time, potentially leading to inappropriate inclusions or exclusions as a result of duration of symptoms. The numbers and identity of rheumatologists varied markedly over time. Among the 15 rheumatologists and 26 rheumatology fellows practicing at the CIUSSS de l’Estrie - CHUS over these 24 years, some were inconsistent or periodic recruiters. In addition, recruiters varied in their confidence to identify subtle joint swelling as synovitis.38 Most importantly, intrinsic hurdles affect recruitment occurring directly within the clinic. For example, the patient and/or physician may feel too exhausted or busy to go through recruitment processes, research assistants may not be available, or radiographs and blood tests for research may not be done close enough to the visit. We estimate that approximately 50% of the potentially eligible patients evaluated were indeed included in EUPA; however, the patients not included were likely missed at random. Second, our study is unicentric; although this represents a strength to assess temporal trends, it decreases external validity of our results. This is especially true because our local population is primarily White, like previous reports, and mostly of French-Canadian ethnicity. Confirmatory studies from other areas and populations are needed. Finally, the practice conditions of rheumatologists at the hospital and of referring primary physicians have been submitted to multiple administrative reforms, some potentially affecting access of patients with recent-onset RA to appropriate care.
To conclude, over 24 years, patients with recent-onset RA presented to rheumatologists with similar demographics, joint manifestations, immunogenetic background, functional impairment, and PROs. During the same time, there was a significant shift toward more seronegative, less erosive, and less inflammatory presentations, suggesting ongoing environmentally driven secular trends, especially since 2011. Our data underline the need for more pathogenic and therapeutic studies in seronegative RA.17,39 The role of active smoking in RA development is clearly decreasing. Because of the increasing prevalence of comorbidities, particularly those that are CV or neoplastic, we propose that alternative pathogenic pathways, such as immune exhaustion, may be at play. Finally, whether these changes at diagnosis, especially toward more seronegativity and less erosive status, contributed to the improvements in outcomes observed over the last decades in EUPA patients merits further study.
ACKNOWLEDGMENTWe thank the staff rheumatologists Dr. Hugues Allard-Chamard, Dr. Guylaine Arsenault, Dr. Lyne Bissonnette, Dr. Alessandra Bruns, and Dr. Pierre Dagenais for their contribution to the recruitment and follow-up of EUPA patients. We also thank our research assistants Chantal Guillet, Noémie Poirier, and Christine Rosa for their long-term contribution to the EUPA study.
FootnotesThe EUPA cohort was supported by the Canadian Institutes of Health Research (grant no. MOP-110959). We also acknowledge previous support from the Arthritis Society (grants no. 00/201 and RG06/108). AJBF, PL, AM, SR, and GB are part of the Centre de recherche du Centre hospitalier universitaire de Sherbrooke (CRCHUS), which received a team grant from the Fonds de recherche du Québec – Santé (FRQ-S). Since 2007, the Sherbrooke Early Undifferentiated Polyarthritis cohort has also received financial support from the Canadian Arthritis Cohort, a study designed and implemented by investigators. BMS contributed to the funding of the analyses of this manuscript through an unrestricted research grant to GB that was managed by the CRCHUS of the Centre intégré universitaire de santé et de services sociaux de l’Estrie - Centre hospitalier universitaire de Sherbrooke (CIUSSS de l’Estrie - CHUS) in Sherbrooke, Quebec, Canada.
MM was previously an employee of BMS Canada. The remaining authors declare no conflicts of interest relevant to this article.
Accepted for publication September 3, 2024.Copyright © 2025 by the Journal of RheumatologyThis is an Open Access article, which permits use, distribution, and reproduction, without modification, provided the original article is correctly cited and is not used for commercial purposes.
留言 (0)