Real-world persistence of initial targeted therapy strategy in monotherapy versus combination therapy in patients with chronic inflammatory arthritis

Eur J Clin Invest. 2024;54:e14095. | 1 of 13 https://doi.org/10.1111/eci.14095 wileyonlinelibrary.com/journal/eci Received: 9 May 2023 | Accepted: 7 September 2023 DOI: 10.1111/eci.14095 ORIGINAL ARTICLE Realworld persistence of initial targeted therapy strategy in monotherapy versus combination therapy in patients with chronic inflammatory arthritis LorenaExposito1 | CarlosSánchez- Piedra2 | PalomaVela- Casasempere3 | Manuel JoséMoreno- Ramos4 | CristinaCampos5 | CristinaBohorquez6 | JavierManero7 | JerusalemCalvo- Gutiérrez8 | CarlosRodríguez- Lozano9 | DoloresRuiz- Montesino10 | NoemíBusquets11 | JavierGarcía- González12 | IsabelCastrejón13 | FernandoAlonso14 | SagrarioBustabad1 | FedericoDíaz- González1,15,16 1Rheumatology Unit, Hospital Universitario de Canarias, Santa Cruz de Tenerife, Spain 2Health Technology Assessment Agency (AETS), Instituto de Salud Carlos III, Madrid, Spain 3Rheumatology Service, Hospital General Universitario Dr. Balmis, ISABIAL, Alicante, Spain 4Rheumatology Unit, Hospital Universitario Virgen de la Arrixaca, Murcia, Spain 5Rheumatology Unit, Hospital General Universitario de Valencia, Valencia, Spain 6Rheumatology Unit, Hospital Universitario Príncipe de Asturias, Alcalá de Henares, Madrid, Spain 7Rheumatology Unit, Hospital Universitario Miguel Servet, Zaragoza, Spain 8Rheumatology Unit, Hospital Universitario Reina Sofía, Córdoba, Spain 9Rheumatology Unit, Hospital Universitario Doctor Negrín, Las Palmas de Gran Canaria, Spain 10Rheumatology Unit, Hospital Universitario Virgen Macarena, Sevilla, Spain 11Rheumatology Unit, Hospital de Granollers, Barcelona, Spain 12Rheumatology Unit, Hospital Universitario Doce de Octubre, Madrid, Spain 13Rheumatology Unit, Hospital General Universitario Gregorio Marañón, Spain 14Research Unit, Spanish Society of Rheumatology, Madrid, Spain 15Department of Internal Medicine, Dermatology and Psychiatry, Universidad de La Laguna, Santa Cruz de Tenerife, Spain 16Instituto Universitario de Tecnologías Biomédicas (ITB), Universidad de La Laguna, Santa Cruz de Tenerife, Spain This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made. © 2023 The Authors. European Journal of Clinical Investigation published by John Wiley & Sons Ltd on behalf of Stichting European Society for Clinical Investigation Journal Foundation. Lorena Exposito and Carlos Sanchez- Piedra shared cofirst authorship. They both have worked together on a publication and contributed equally. Abbreviations: AS, ankylosing spondylitis; B, biologics; CI, confidence interval; Cs, conventional synthetic; DMARDs, diseasemodifying antirheumatic drugs; HR, hazard ratio; IRR, incidence rate ratio; MTX, methotrexate; PsA, psoriatic arthritis; RA, rheumatoid arthritis; TNFi, TNF- inhibitor; Ts, targeted synthetic. Correspondence Federico Díaz- González, Rheumatology Unit, Hospital Universitario de Canarias, C/Ofra s/n 38320, La Laguna, Spain. Email: federico.diaz.gonzalez@gmail. com Abstract Objective: The persistence of biologic (b) and targeted synthetic (ts) diseasemodifying antirheumatic drugs(DMARDs) in monotherapy versus in combination with conventional synthetic (cs) DMARDs is still a controversial topic in 2 of 13 | EXPOSITO et al. 1 | INTRODUCTION Clinical practice guidelines for the management of rheumatoid arthritis (RA) recommend that when a biologic (b) diseasemodifying antirheumatic drug (DMARD) is indicated, it should be used in combination with a conventional synthetic (cs) DMARD due to the improved efficacy and reduced immunogenicity of the combination.1 A recent systematic review of 44 randomized clinical trials concluded that the combination of a bDMARD or targeted synthetic (ts) DMARD with a csDMARD was generally more effective than b/tsDMARD monotherapy, both in naïve and in an inadequate response to cs- DMARD in RA patients.2 However, in clinical practice, up to 30% of patients with RA are treated with bDMARD in monotherapy,3– 6 possibly because of csDMARD's intolerance, contraindications, or patient preference. Assessing the persistence of treatments in the real world is particularly useful, as it is considered an indirect marker not only of their safety and efficacy, but also of patient satisfaction.7,8 Several studies using registries and other realworld sources of information have consistently shown that, overall, concomitant use of bDMARD and csDMARD is associated with a higher likelihood of Funding information Agencia Española de Medicamentos y Productos Sanitarios; Biogen; Bristol- Myers Squibb; Celltrion Healthcare; Eli Lilly and Company; Instituto Científico Pfizer; Janssen Biotech; Merck; Novartis; Regeneron Pharmaceuticals; Samsung Bioepis rheumatic diseases. To clarify this issue, the retention of the initial treatment strategy of b/tsDMARD in combination with csDMARD versus monotherapy in rheumatoid arthritis (RA), psoriatic arthritis (PsA) and ankylosing spondylitis (AS) patients under reallife conditions was evaluated. Factors associated with maintenance of the initial strategy were analysed. Methods: Nested cohort study within the Spanish BIOBADASER III registry. Bivariate comparisons and multivariate Cox proportional hazards models were used for the analyses. Results: A total of 2521 patients were included in the study. In the multivariate model, the initial strategy of combination therapy was associated with shorter persistence in patients with RA (hazard ratio [HR] 1.58;95% confidence interval [CI] 1.00– 2.50; p = .049), PsA (HR 2.48; 95% CI 1.65– 3.72) and AS (HR 16.77; 95% CI 7.37– 38.16; p < .001), regardless of sex, time of disease progression, baseline disease activity, glucocorticoid use or type of b/tsDMARD. Overall, the combination strategy was associated with an increased incidence of adverse events (incidence rate ratio [IRR] 1.13; 95% CI 1.05– 1.21). Conclusions: In this reallife study, the strategy of combining a b/tsDMARD with a csDMARD is associated with lower persistence and worse safety profile compared to monotherapy in RA and especially in PsA and AS, suggesting that combination therapy should be rethought as first choice in RA patients, but especially in PsA and AS patients. KEYWORDS ankylosing spondylitis, antirheumatic agents, biologic therapy, psoriatic arthritis, rheumatoid arthritis Key Messages • This analysis compared the persistence of the initial targeted therapy strategy in monotherapy versus combined with csDMARDs. • Combination strategy is associated with shorter persistence and worse safety profile than monotherapy in chronic inflammatory arthritis. • Monotherapy should be considered as first choice in RA, but especially in PsA and AS patients. 13652362, 2024, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/eci.14095 by Readcube (Labtiva Inc.), Wiley Online Library on [01/06/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License | 3 of 13 EXPOSITO et al. bDMARD persistence.3,5,9– 12 Moreover, in a systematic review of 98 studies that used information from registries or health care databases and included over 200,000 patients with RA, the concomitant use of a csDMARD was a significant predictor of a lower likelihood of bD- MARD discontinuation.13 However, there may be some differences among bDMARDs; thus, a study using tocilizumab registries showed no differences in terms of drug persistence between tocilizumab in monotherapy and in combination with a csDMARD, while the combination of a TNF- inhibitor (TNFi) with a csDMARD was associated with a lower likelihood of treatment discontinuation compared to TNF- inhibitor monotherapy.10 In patients with psoriatic arthritis (PsA), when a bD- MARD is considered indicated, there is no conclusive recommendation on the concomitant use of a csDMARD, with this being considered part of the research agenda for the management of PsA.14 Although treatment patterns seem to vary over time, the use of biologic monotherapy is also common among patients with PsA.6,15 A systematic review of six randomized controlled trials found that concomitant use of methotrexate (MTX) with a TNFi is not associated with a clear beneficial effect compared to TNFi monotherapy, except for some data suggesting that the combination may reduce the progression of structural damage.16 The results from registries and other observational sources regarding the impact of combination therapy on drug persistence show inconsistent results. Some studies have shown no difference between combination therapy and monotherapy17– 20; others suggest that bDMARD persistence is better when adding a csDMARD.9,21– 24 According to clinical practice guidelines, in adult patients with ankylosing spondylitis (AS), despite treatment with NSAIDs, treatment with a TNFi over no treatment with a TNFi is strongly recommended, but, although with a low level of evidence, the guidelines recommend against the coadministration of MTX.25 Information from registries or observational studies on biologic persistence in patients with AS is very limited and shows somewhat inconsistent results, but most studies show no impact of adding MTX to the TNFi,22,26,27 while only one study indicates a higher persistence of the TNFi in those patients receiving concomitant MTX.9 To our knowledge, very few studies have compared the persistence of b/tsDMARD using the same source of patients and the same methodology, and none of them have evaluated the survival of the initial target therapy strategy: monotherapy or combined in patients with RA, PsA or AS under reallife conditions. The primary objective of this study was to determine factors associated with potential differences in the persistence of the initial targeted therapy strategy with (combination strategy) versus without (monotherapy strategy) csDMARDs in patients with these three clinical conditions in a realworld clinical setting. 2 | PATIENTS AND METHODS 2.1 | Study design and patient population This is a nested cohort study within a registry based on routine clinical data. The original study is BIOBADASER III, a prospective Spanish registry with 28 participating centres that was established at the end of 2015 to evaluate the longterm safety and effectiveness of patients with rheumatic diseases diagnosed based on clinical grounds who are undergoing treatment with targeted therapy: bDMARD, including biosimilars and tsDMARD.28 The BIOBADASER III registry is supported by the Spanish Agency of Medicines (https://www.aemps.gob.es/en/ home.htm), the Spanish Society of Rheumatology (https:// www.ser.es/) and by most of the pharmaceutical companies with b/tsDMARD in the Spanish market. Adult patients with a diagnosis of RA, PsA (peripheral and mixed forms) or AS (patients with nonradiographic axial SpA were not included in this study) on active followup and who initiated and maintained for at least 3 months treatment with a targeted therapy using a combination strategy or monotherapy were included in this study. The inclusion period was from December 2015 to October 15, 2020. The information in BIOBADASER is collected during the patient's visits to the rheumatology clinics. The information is included in the database at baseline (defined as the time of initiating treatment with a targeted therapy), when an adverse effect or a change in treatment with the biologic occurs, and for effectiveness assessment, at least once per year (followup visit). Data are also recorded in the case of death for any reason or b/tsDMARD discontinuation for any cause. The project was approved by the Ethics Review Committee of the Hospital Universitario Clinic Barcelona, which acted as the reference committee (approval code FERADA- 2015- 01). Written informed consent was obtained from all participants. 2.2 | Exposure, outcomes and covariates The combination therapy strategy was defined as initiating a b/tsDMARD and a concomitant csDMARD simultaneously and monotherapy strategy was defined as initiating b/tsDMARD therapy without the addition of a concomitant csDMARD. The main variables were treatment strategy start date (the date when the patient started b/tsDMARD therapy, either in monotherapy or 13652362, 2024, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/eci.14095 by Readcube (Labtiva Inc.), Wiley Online Library on [01/06/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 4 of 13 | EXPOSITO et al. in combination) and treatment discontinuation date (the date when the patient changed treatment strategy from combination to monotherapy or viceversa) or if the patient stop the target therapy. The primary outcome was persistence of treatment strategy, defined as the time elapsed from the date of initiation of b/tsDMARD in monotherapy or in combination until discontinuation of the initial strategy. After therapy initiation, patients who were switched to a different drug from the same class (b/tsDMARD or csDMARD) but maintained the initial strategy (combination or monotherapy) were considered to persist on the strategy. The covariates included age at the time of initiating the first b/tsDMARD, sex, cotreatment with a csDMARD (yes/ no), cotreatment with glucocorticoids (yes/no) diagnosis, use of TNFi as initial biologic (yes/no), disease duration to initiating treatment with a b/tsDMARD and disease activity as assessed by DAS28 or BASDAI. 2.3 | Statistical analysis Quantitative variables are described with the mean and standard deviation or the median and interquartile range, and qualitative variables are described with the absolute and relative frequencies. For baseline comparisons, quantitative variables were compared using Student's ttest, nonparametric variables were compared with the Mann– Whitney U test or the Kruskal– Wallis H test, and qualitative variables were compared using the chisquared test or Fisher's exact test. The persistence of b/tsDMARD was analysed separately in RA, PsA and AS patients using the Kaplan– Meier method and compared between monotherapy and combination groups with the logrank test. Multivariate Cox proportional hazard models were constructed with b/ tsDMARD persistence as the outcome. Age, disease duration and disease activity at the start of the targeted therapy, as continuous variables, and sex, initial use of TNFi, glucocorticoids and combination therapy, as categorized variables were considered covariates. In the multivariate model were included the covariates which showed a p value <.20 in a bivariate Cox regression analysis, together with those considered clinically relevant (sex). Four multivariate models were constructed: a global model with all patients and including diagnosis as a covariate, and a model for RA, PsA and AS diagnoses. All statistical analyses were performed using STATA software, version 16.0 SE (Stata Corporation). The results were considered significant at p < .05. 3 | RESULTS 3.1 | Patient disposition and characteristics Of the 4051 patients included in the BIOBADASER registry, 2521 met the selection criteria and were included in this study; 47.2% (n = 1192) of them were diagnosed with RA, and the remaining were almost evenly distributed between PsA (27.6%, n = 697) and AS (25.1%, n = 632) (see flowchart in Figure1). b/tsDMARD monotherapy was used as the initial strategy in 20.5% (n = 245) of the patients with RA, 35.6% (n = 248) of patients with PsA and in 76.7% (n = 485) of the patients with AS. The median (interquartile range) duration of the initial treatment strategy (either monotherapy or combination) was 20 (12– 32) months for patients with RA, 22 (12– 40) months for patients with PsA, and 24 (14– 43) months for patients with AS. The characteristics of the patients are shown in Table1. The monotherapy and combination therapy groups were FIGURE 1 Flowchart. Patients included in the analysis. bDMARD, biologic diseasemodifying antirheumatic drug; combo, combination therapy; mono, monotherapy. 13652362, 2024, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/eci.14095 by Readcube (Labtiva Inc.), Wiley Online Library on [01/06/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License | 5 of 13 EXPOSITO et al. TABLE 1 Baseline characteristics of the patients included in the analysis. Rheumatoid arthirtis Psoriatic arthritis Ankylosing spondylitis Total N = 1192 N = 697 N = 632 N = 2521 Variable (%) Mono N = 245 (20.5) Combo N = 947 (79.5) p value Mono N = 248 (35.6) Combo N = 449 (64.4) p value Mono N = 485 (76.7) Combo N = 147 (23.3) p value Mono N = 978 (38.8) Combo N = 1543 (61.2) p value Age at b/ tsDMARD initiation (years), mean ± SD 58.31 ± 2.6 55.3 ± 12.3 <.001 51.3 ± 12.2 49.9 ± 11.3 .122 45.6 ± 12.9 48.0 ± 14.3 .053 50.2 ± 13.7 53.1 ± 12.6 <.001 Sex (females), n (%) 204 (83.3) 712 (75.2) .008 130 (52.4) 236 (52.6) .971 146 (30.1) 49 (33.3) .458 480 (49.1) 997 (64.6) <.001 DAS28 4.7 (1.2) 4.8 (1.3) .375 4.9 (2.6) 4.9 (2.7) .945 4.2 (1.3) 4.6 (1.3) <.001 BASDAI 4.1 (1.2) 4.2 (1.2) .490 5.6 (2.1) 5.4 (2.3) .400 5.5 (2.2) 5.3 (2.4) .200 Disease duration (years), median [IR] 5.3 [2.3– 9.7] 4.4 [1.8– 9.5] .203 3.1 [1.1– 8.0] 3.1 [1.1– 6.9] .4543 2.5 [0.7– 10.0] 3.5 [1.0– 9.8] .2625 3.5 [1.0– 9.5] 3.9 [1.5– 8.5] .0926 First b/tsDMARD, n (%)a TNFi 94 (13.1) 621 (86.9) <.001 132 (27.6) 346 (72.4) <.001 416 (75.2) 137 (24.8) .04 642 (36.8) 1104 (63.2) <.001 Anti- IL6 52 (42.3) 71 (57.7) 52 (42.3) 71 (57.7) Anti- B- cell (CD20) 13 (18.8) 56 (81.2) 13 (18.8) 56 (81.2) JAKi 56 (35.0) 104 (65.0) 4 (57.1) 3 (42.9) 60 (35.9) 107 (64.1) Anti- IL1 1 (50.0) 1 (50.0) 1 (50.0) 1 (50.0) Anti- IL17A 40 (46.5) 46 (53.5) 67 (88.2) 9 (11.8) 107 (66.1) 55 (34.0) Anti- IL12/23 17 (47.2) 19 (52.8) 2 (66.7) 1 (33.3) 19 (47.5) 20 (52.5) PDE4i 55 (61.1) 35 (38.9) 55 (61.1) 35 (38.9) Abatacept 29 (23.6) 94 (76.4) 29 (23.6) 94 (76.4) Glucocorticoids (yes), n (%) 147 (60.0) 612 (64.6) .845 63 (69.2) 149 (70.3) .855 22 (4.5) 44 (29.9) <.001 232 (23.7) 805 (52.2) <.001 aPercentages were calculated by row. Abbreviations: bDMARD, biologic diseasemodifying antirheumatic drug; IL, interleukin; JAKi, Janus kinase inhibitor; PDE4i, phosphodiesterase 4 inhibitor; SD, standard deviation; TNFi, tumour necrosis factor inhibitor. 13652362, 2024, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/eci.14095 by Readcube (Labtiva Inc.), Wiley Online Library on [01/06/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 6 of 13 | EXPOSITO et al. generally wellbalanced regarding age and sex for PsA and AS; however, patients with RA in the monotherapy group were older and more likely to be female than patients in the combination group. Disease duration was similar between the monotherapy and combination groups in patients with RA, PsA and AS, showing PsA patients a trend toward longer disease duration in monotherapy with respect to the combination strategy. With respect to disease activity at the start of biologic therapy, DAS28 in patients with RA and PsA and BASDAI in AS were similar in both types of strategies (Table1). The most frequent initial targeted therapy was a TNFi (n = 1745, 69.2%); when a TNFi was prescribed, it was more frequently used in combination with a csDMARD in RA patients (87%, n = 621 vs. 13%, n = 93 for the combination and monotherapy groups, respectively) and PsA patients (72.4%, n = 346 vs. 27.6%, n = 132 for the combination and monotherapy groups, respectively) but more frequently prescribed as monotherapy in patients with AS (24.8%, n = 137 vs. 75.2%, n = 416 for the combination and monotherapy groups, respectively). The proportion of patients who were receiving glucocorticoids was higher in the combination group than in the monotherapy group (52.2% vs. 23.7%), especially among patients with AS (29.9% vs. 4.5%). 3.2 | Strategy persistence Unadjusted survival analysis showed that the persistence of b/tsDMARD in monotherapy or combination therapy was not significantly different in patients with RA (Figure2A). However, in patients with PsA (Figure2B) and those with AS (Figure2C), the strategy of using b/ts- DMARD in monotherapy showed significantly longer survival than the combination strategy (logrank test p < .001 for both comparisons). Overall, in a global multivariate model, there were no differences in the persistence of the initial treatment strategy depending on the clinical entity (TableS1). Thus, compared to RA patients, the likelihood of treatment strategy persistence was not significantly different in patients with PsA (HR 1.06, 95% CI 0.80– 1.40, p = .677) or patients with AS (HR 1.48, 95% CI 0.99– 2.20, p = .053). Regarding treatments, the use of TNFi compared to other bDMARD was associated with a longer persistence (HR 0.74, 95% CI 0.58– 0,94, p = .016), while glucocorticoids were associated with a shorter persistence (1.36 95% CI 1.02– 1.82, p = .037). In this global model, the b/tsDMARD combination therapy strategy was associated with a shorter persistence than the monotherapy strategy (HR 3.04, 95% CI 2.14– 4.32, p < .001). When analysed by clinical entities in the multivariate model, compared to monotherapy, the combination therapy was associated with shorter b/tsDMARD persistence in patients with RA (HR 1.58, 95% CI 1.00– 2.50, p = .049), PsA (HR 2.48, 95% CI 1.65– 3.72, p < .001) and AS (HR 16.77, 95% CI 7.37– 38.16, p < .001) (Table2). Other factor significantly associated with longer b/tsDMARD persistence was the age at treatment initiation among patients with RA (HR 0.98, 95% CI 0.97– 0.99, p = .004) (Table2). Interestingly, neither disease activity nor TNFi or glucocorticoid use at the initiation of targeted therapy was associated with persistence of the initial therapeutic strategy. 3.3 | Safety profile The frequency of adverse events for the whole group of combination therapy compared to the group who received b/tsDMARD monotherapy is presented in Table3. Overall, the incidence of any adverse event in the combination group significantly increased by 13% (the incidence rates [IR] were 281 and 250 cases per 1000 patientyears in the combination and monotherapy groups, respectively; p = .001), but the rate of serious adverse events did not differ between the study groups (IR 35 and 34 cases per 1000 patientyears in the combination and monotherapy groups, respectively; p = .857). Most groups of adverse events occurred more frequently among patients who received a targeted therapy in combination, although the differences were statistically significant only for ‘infections and infestations’ (incidence rate ratio [IRR] 1.21, 95% CI 1.05– 1.40) and ‘musculoskeletal and connective tissue disorders’ (IRR 1.67, 95% CI 1.10– 2.53). The incidence rates of ‘immune system disorders’ and ‘neoplasms benign, malignant and unspecified’ were 24% and 19%, respectively, higher in the combination group than in the monotherapy group, but the differences were not statistically significant. 4 | DISCUSSION The most important findings of this manuscript can be summarized as follows: (1) Under realworld conditions, the strategy of combining a targeted therapy with a csD- MARD was associated with shorter treatment persistence in RA and especially in PsA and AS compared to a monotherapy strategy; (2) in the multivariable analysis, none of the factors studied including age, sex, time of disease evolution, initial disease activity, use of glucocorticoids, or type of targeted therapy, were associated with worse persistence of the initial strategy; and (3) overall, the combination strategy showed a worse safety profile than the monotherapy strategy. 13652362, 2024, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/eci.14095 by Readcube (Labtiva Inc.), Wiley Online Library on [01/06/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License | 7 of 13 EXPOSITO et al. FIGURE 2 Biologic diseasemodifying antirheumatic drug treatment strategy persistence (Kaplan– Meier plots). 13652362, 2024, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/eci.14095 by Readcube (Labtiva Inc.), Wiley Online Library on [01/06/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 8 of 13 | EXPOSITO et al. The main objective of this work was to analyse patient and disease characteristics associated with persistence on initial targeted therapy strategy, either as monotherapy or in combination with csDMARD, in RA, PsA and AS patients using a national registry database under realworld conditions. The pattern of b/tsDMARD use in routine clinical practice shown in our study is somewhat consistent with clinical practice guideline recommendations. For example, b/tsDMARD monotherapy is more commonly used in patients with AS, where guidelines discourage the use of combination therapy,25 and less commonly used in patients with RA, where guidelines recommend combination therapy.1 However, despite the guideline recommendations, in our series 20% of RA patients (245 of 1192) initiated b/tsDMARD as monotherapy in realworld conditions. To the best of our knowledge, few studies have analysed the impact of the combination with a csDMARD on b/tsDMARD persistence in these three clinical entities using the same source of realworld data.9,22,29,30 George et al., in a retrospective study, showed that the use of MTX in combination with a TNFi increased the persistence of the TNFi in patients with RA (HR 0.85, 95% CI 0.80– 0.89), patients with PsA (HR 0.81, 95% CI 0.74– 0.89) and patients with AS (HR 0.79, 95% CI 0.67– 0.93).9 The inconsistency between their results and ours could be explained in part by differences in methodology. George et al. used a United States administrative claims database from nationally representative private health care insurance.9 Administrative databases were not originally conceived for research purposes, and although they have some advantages, they also have TABLE 2 Factors associated with targeted treatment initial strategy discontinuation in the bivariate and multivariate analyses by clinical entity. Rheumatoid arthritis Bivariate Multivariate HR 95% CI p value HR 95% CI p value Age at b/tsDMARD initiation (years) 0.98 (0.97– 0.99) <.001 0.98 (0.97– 0.99) .004 Sex (male) 1.09 (0.82– 1.46) .553 1.03 (0.73– 1.45) .859 Disease duration (years) 0.98 (0.96– 1.00) .018 – – – DAS28 1.01 (0.91– 1.11) .892 – – – TNFi (other b/tsDMARD) 0.95 (0.75– 1.21) .685 – – – Glucocorticoids (No) 1.46 (0.90– 2.39) .128 1.47 (0.90– 2.40) .128 Combination (Monotherapy) 1.34 (0.95– 1.90) .095 1.58 (1.00– 2.50) .049 Psoriatic arthritis Age at b/tsDMARD initiation (years) 0.99 (0.97– 1.01) .027 0.99 (0.98– 1.0) .097 Sex (male) 0.87 (0.65– 1.17) .351 0.94 (0.70– 1.27) .694 Disease duration (years) 0.98 (0.95– 1.00) .290 – – – DAS28 1.01 (0.89– 1.16) .840 – – – TNFi (other bDMARD) 1.08 (0.78– 1.51) .638 – – – Glucocorticoids (no) 1.73 (1.09– 2.73) .019 1.55 (0.98– 2.46) .059 Combination (monotherapy) 2.53 (1.69– 3.79) <.001 2.48 (1.65– 3.72) <.001 Ankylosing spondylitis Age at bDMARD initiation (years) 1 (0.98– 1.02) .911 0.98 (0.96– 1.00) .114 Sex (male) 0.74 (0.45– 1.21) .225 0.48 (0.21– 1.10) .082 Disease duration (years) 0.99 (0.97– 1.01) .432 – – – BASDAI 0.99 (0.90– 1.10) .909 – – – TNFi (other bDMARD) 1.72 (0.75– 3.95) .202 – – – Glucocorticoids (no) 2.05 (1.15– 3.68) .015 0.62 (0.33– 1.15) .132 Combination (monotherapy) 13.58 (8.28– 22.28) <.001 16.77 (7.37– 38.16) <.001 Abbreviations: b/tsDMARD, biologic/targeted synthetic diseasemodifying antirheumatic drug; CI, confidence interval; HR, hazard ratio; TNFi, tumour necrosis factor inhibitors. 13652362, 2024, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/eci.14095 by Readcube (Labtiva Inc.), Wiley Online Library on [01/06/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License | 9 of 13 EXPOSITO et al. TABLE 3 Adverse events overall and by System Organ Class (MedDRA). Adverse event Incidence rate (95% CI) (cases per 1000 patientyears) Incidence rate ratio 95% CI p valueCombination Monotherapy Overall Any adverse event 280.70 (268.30– 293.70) 250.20 (236.70– 264.50) 1.13 (1.05– 1.21) .001 Serious adverse events 34.60 (30.40– 39.30) 34.00 (29.30– 39.60) 1.02 (0.84– 1.24) .857 System Organ Class Infections and infestations 76.30 (70.00– 83.20) 63.10 (56.40– 70.40) 1.21 (1.05– 1.40) .007 Gastrointestinal disorders 38.60 (34.20– 43.60) 36.30 (31.30– 42.00) 1.07 (0.88– 1.29) .501 General disorders and administration site condition 16.50 (13.70– 19.90) 13.70 (10.80– 17.40) 1.21 (0.90– 1.64) .214 Skin and subcutaneous tissue disorders 5.50 (4.00– 7.60) 4.60 (3.10– 7.00) 1.11 (0.79– 1.55) .555 Injury, poisoning and procedural complications 12.20 (9.80– 15.20) 11.90 (9.20– 15.30) 1.03 (0.74– 1.44) .858 Nervous system disorders 10.10 (8.00– 12.90) 10.70 (8.20– 14.00) 0.95 (0.66– 1.36) .787 Complementary examinations 4.50 (3.10– 6.40) 4.60 (3.10– 7.00) 0.97 (0.56– 1.67) .905 Cardiac disorders 7.90 (6.00– 10.30) 9.30 (6.90– 12.40) 0.85 (0.58– 1.27) .436 Renal and urinary disorders 5.20 (3.70– 7.30) 3.80 (2.40– 6.00) 1.37 (0.78– 2.39) .273 Surgical and medical procedures 10.10 (8.00– 12.90) 8.50 (6.30– 11.40) 1.20 (0.82– 1.76) .351 Blood and lymphatic system disorders 7.90 (6.00– 10.30) 5.40 (3.70– 7.90) 1.46 (0.92– 2.31) .112 Metabolism and nutrition disorders 5.20 (3.70– 7.30) 4.00 (2.60– 6.20) 1.30 (0.75– 2.25) .352 Musculoskeletal and connective tissue disorders 10.70 (8.50– 13.50) 6.40 (4.60– 9.10) 1.67 (1.10– 2.53) .016 Immune system disorders 9.20 (7.20– 11.80) 7.50 (5.40– 10.30) 1.24 (0.83– 1.87) .295 Eye disorders 8.60 (6.70– 11.20) 12.30 (9.60– 15.80) 0.71 (0.49– 1.01) .057 Neoplasms benign, malignant and unspecified (incl cysts and polyps) 5.50 (4.00– 7.60) 4.60 (3.10– 7.00) 1.19 (0.71– 2.01) .506 Respiratory, thoracic and mediastinal disorders 14.00 (11.40– 17.10) 10.90 (8.30– 14.20) 1.29 (0.92– 1.80) .134 Vascular disorders 4.00 (2.80– 5.90) 3.20 (2.00– 5.30) 1.25 (0.67– 2.32) .476 Hepatobiliary disorders 5.10 (3.60– 7.10) 4.20 (2.80– 6.50) 1.20 (0.70– 2.07) .509 Ear and labyrinth disorders 4.60 (3.20– 6.60) 5.80 (4.10– 8.40) 0.79 (0.48– 1.32) .369 Endocrine disorders 2.80 (1.80– 4.40) 2.60 (1.50– 4.50) 1.08 (0.54– 2.20) .822 Psychiatric disorders 3.40 (2.30– 5.20) 3.60 (2.30– 5.80) 0.95 (0.51– 1.76) .865 (Continues) 13652362, 2024, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/eci.14095 by Readcube (Labtiva Inc.), Wiley Online Library on [01/06/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License