Personalized dose reduction strategies for biologic disease-modifying antirheumatic drugs for treating ankylosing spondylitis: a clinical and economic evaluation with predictive modeling

Personalized dose reduction strategies for biologic disease-modifying antirheumatic drugs for treating ankylosing spondylitis: a clinical and economic evaluation with predictive modeling | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Personalized dose reduction strategies for biologic disease-modifying antirheumatic drugs for treating ankylosing spondylitis: a clinical and economic evaluation with predictive modeling Hai Binh Bui, Thi Thu Phuong Nguyen, Thi Thanh Hang Vu, Thi Thuc Nhan Ngo, and 2 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5917710/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 26 May, 2025 Read the published version in BMC Rheumatology → Version 1 posted 4 You are reading this latest preprint version Abstract Background Ankylosing spondylitis (AS) is a chronic inflammatory disease that significantly affects quality of life and imposes a high economic burden on patients due to the cost of biologic disease-modifying antirheumatic drugs (bDMARDs). Dose reduction strategies for bDMARDs may offer a feasible approach to maintaining clinical efficacy while reducing costs. This study aimed to evaluate the clinical effectiveness and cost-efficiency of bDMARD dose reduction in patients with AS and apply predictive modeling to identify key factors influencing disease control. Methods This 12-month prospective study included 368 patients with AS who were divided into two groups: those who received dose reduction and those with full-dose therapy. Clinical outcomes such as C-reactive protein (CRP) levels, the Bath ankylosing spondylitis disease activity index (BASDAI) and ankylosing spondylitis disease activity score (ASDAS) were assessed, along with cost effectiveness using incremental cost effectiveness ratios (ICER). Random forest models were developed to predict the achievement of inactive disease (ASDAS < 1.3) and to identify key predictors. Results The ICER to achieve an ASDAS < 1.3 was highly favorable (- $ 16,772.62). Patients in the dose reduction group demonstrated significant improvements in CRP levels (-4.65 vs. -1.32 mg/L, p < 0.001), BASDAI (-3.00 vs. -0.42, p < 0.001), and ASDAS (-1.72 vs. -0.15, p < 0.001), compared with the full dose group. Predictive modeling identified baseline CRP level, baseline ASDAS, and dose adjustment as key factors influencing outcomes, with the medium feature model achieving an area under the receiver operating characteristic curve of 81.86%. Conclusions The reduction in bDMARD doses maintained clinical efficacy and achieved significant cost savings, offering a viable strategy for the management of AS. Predictive modeling provided actionable insights to optimize personalized treatment strategies, balancing efficacy and economic sustainability. These findings support the integration of dose reduction strategies into routine practice, particularly in resource-limited settings. ankylosing spondylitis biological disease-modifying antirheumatic drugs dose reduction cost-effectiveness disease activity Figures Figure 1 Figure 2 Background Ankylosing spondylitis (AS) is a chronic inflammatory disease that primarily affects the spine and sacroiliac joints, causing pain, stiffness, and potential fusion of the vertebrae. This progressive condition is a type of spondyloarthritis, which also includes psoriatic and reactive arthritis ( 1 ). The prevalence estimate of AS per 10,000 people is 23.8 globally; 16.7 in Asia; 31.9 in North America; 10.2 in Latin America; and 7.4 in Africa ( 2 ). The underlying mechanisms of AS involve genetic, environmental and immunological factors, and a notable association with HLA-B27. Despite advances in understanding the pathophysiology of AS, the disease remains incurable, and treatments focus on managing symptoms, reducing inflammation, and preventing complications ( 3 , 4 ). AS imposes a significant burden on patients and society. This disease affects quality of life, productivity at work, and healthcare costs ( 5 ). The total annual cost per patient ranges from €5,155 in Hungary to €20,328 in Spain, with indirect costs accounting for a substantial proportion ( 6 , 7 ). In recent years, therapeutic guidelines for AS have evolved, particularly with the 2022 update of the Assessment of SpondyloArthritis International Society and European Alliance of Associations for Rheumatology recommendations ( 8 ). These guidelines emphasize early diagnosis and a comprehensive treatment approach, prioritizing both nonpharmacological interventions such as physical therapy and pharmacological options, including nonsteroidal anti-inflammatory drugs (NSAIDs) and disease-modifying antirheumatic drugs (DMARDs). Biologic DMARDs (bDMARDs) and Janus kinase inhibitors are recommended as second-line treatments for patients who show inadequate response to NSAIDs. This update highlights the role of DMARDs, particularly tumor necrosis factor inhibitors (TNFi) and interleukin-17 inhibitors (IL-17i), in the management of moderate to severe AS. These agents target pathways involved in the inflammatory response, providing substantial relief and improving functional outcomes ( 9 ). DMARDs, including biologics such as etanercept, adalimumab, and infliximab, have transformed AS management. By targeting TNF-α and IL-17, these drugs reduce inflammation, stop disease progression, and improve quality of life ( 10 , 11 ). Despite their efficacy, DMARDs pose a significant financial burden due to their high treatment costs, limiting accessibility for patients with AS ( 12 ). In many healthcare settings, the cost of DMARD therapy is partially or fully borne by the patients, making long-term adherence challenging. With TNF inhibitors and IL-17A monoclonal antibodies proving effective for AS management and becoming standard treatments, the direct expenses related to AS care have increased considerably. In 2012, the annual direct costs of AS in the US were estimated at approximately $ 6,514 for medical care and $ 11,162 for prescription drugs per patient ( 13 ). Consequently, researchers and clinicians have explored strategies to optimize DMARD use, such as dose tapering or extension, to reduce costs without compromising efficacy. Studies have investigated dose reduction strategies for DMARDs in patients with AS and other rheumatic diseases. Dose reduction, which involves lengthening the interval between doses, aims to maintain disease control while reducing medication frequency and associated costs. Research on rheumatoid arthritis, another condition treated with DMARDs, suggests that dose reduction may be feasible in patients in sustained remission or in those with low disease activity. However, the evidence for AS is less robust, with mixed results regarding the long-term outcomes of dose reduction. Some studies have indicated that extending the dose interval in patients with stable AS can maintain low disease activity, whereas others caution against potential flare-ups and diminished control over disease progression. The high cost of bDMARDs has created a demand for more cost-effective treatment options. Economic analyses in rheumatology have highlighted the financial impact of these drugs on healthcare systems and patients. These studies often employ metrics, such as the incremental cost-effectiveness ratio (ICER), to assess whether the health benefits of dose reduction justify the potential cost savings. However, the variability in disease activity assessments and patient responses to DMARDs pose challenges for establishing standardized protocols for dose reduction. Therefore, this study aimed to evaluate the efficacy and cost-effectiveness of DMARD dose reduction in patients over a 12-month period. By analyzing clinical outcomes such as C-reactive protein (CRP) levels, the Bath ankylosing spondylitis disease activity index (BASDAI), and ankylosing spondylitis disease activity score (ASDAS), this study sought to determine whether dose reduction can achieve disease control comparable to standard dosing. In addition, cost-related outcomes, including ICER, were evaluated to determine the financial implications of dose reduction. This investigation contributes to the growing body of literature on personalized medicine approaches for AS by addressing the clinical and economic aspects of optimizing DMARD therapy. Methods Aim, design, and setting of the study This prospective case-control study aimed to evaluate the clinical and cost-effectiveness of personalized dose reduction strategies for DMARDs among patients with AS. The study was conducted at Bach Mai Hospital in Hanoi, Vietnam, from 1 January 2021 to 30 September 2024. Participants Eligible participants were adults (≥ 18 years old) diagnosed with AS according to the modified New York criteria ( 14 ). Additional inclusion criteria included a history of at least 3 months of DMARD therapy prior to enrollment. Patients with coexisting autoimmune conditions, severe organ dysfunction, or any other conditions that could interfere with treatment evaluation were excluded. Interventions Participants were allocated to one of two groups according to their treatment plan: Group 1 (Dose reduction): Patients who achieved sustained low disease activity—defined as an Ankylosing Spondylitis Disease Activity Score (ASDAS) < 2.1 for at least 3 consecutive months—underwent extended dosing intervals. They also had stable clinical and laboratory parameters, including normal C-reactive protein (CRP) levels and no significant disease flares or radiographic progression over the previous evaluation period. Dose reduction strategies were as follows: Etanercept (Enbrel ® ): reduced from 50 mg weekly to 50 mg bi-weekly. Secukinumab (Fraizeron ® ): reduced from 150 mg monthly to 150 mg bi-monthly. Adalimumab (Humira ® ): reduced from 40 mg every two weeks to 40 mg every three weeks. Infliximab (Remicade ® ): reduced from 5 mg/kg every 6–8 weeks to 5 mg/kg every 10–12 weeks. Golimumab (Simponi ® ): reduced from 50 mg monthly to 50 mg every two months. Group 2 (Full dose): Patients who did not meet criteria for dose reduction or had higher disease activity remained on standard, full-dose DMARD regimens as per manufacturer recommendations, without alterations in dose frequency or quantity. Both groups followed the same follow-up schedule, with evaluations every 3 months to monitor clinical status, laboratory markers (e.g., CRP, erythrocyte sedimentation rate), and radiographic changes. Patients were also assessed at each visit for adverse events, treatment adherence, and quality of life using standardized questionnaires. Outcome measures Primary outcome: The proportion of patients achieving low disease activity at 12 months, defined as a Bath Ankylosing Spondylitis Disease Activity Index (BASDAI) < 4 and an ASDAS < 1.3. Secondary outcomes: Changes in CRP levels and cost-effectiveness analysis (CEA) comparing the two groups ( 15 ). Disease activity was measured using ASDAS ( 16 , 17 ) and BASDAI ( 18 ) at baseline and at 12 months. The ASDAS was calculated using the following formula: ASDAS = 0.12 × Back pain + 0.06 × Duration of morning stiffness + 0.11 × Patient global + 0.07 × Peripheral pain/swelling + 0.58 × Ln (CRP + 1) ASDAS values were categorized according to the following thresholds: scores below 1.3 indicated no disease activity, scores from 1.3 to 2.0 represented moderate activity, scores between 2.1 and 3.4 denoted high activity, and scores of 3.5 or above were classified as very high activity ( 19 ). For the CEA, direct medical costs (medication, hospital visits, laboratory tests) were collected from a healthcare provider perspective and converted to US dollars (USD). Effectiveness was defined as the proportion of patients achieving low disease activity (BASDAI < 4 and ASDAS < 1.3). The incremental cost-effectiveness ratio (ICER) was calculated as the difference in cost between groups divided by the difference in effectiveness, with cost-effectiveness interpreted according to thresholds recommended by the World Health Organization. Statistical analysis Descriptive statistics were used to summarize baseline characteristics. Continuous variables were reported as means ± standard deviations or medians with interquartile ranges, as appropriate. Categorical variables were expressed as frequencies and percentages. Between-group comparisons were made using chi-square or Fisher’s exact tests for categorical variables and independent t tests or Mann–Whitney U tests for continuous variables, based on data distribution. A p-value < 0.05 was considered statistically significant. All statistical analyses were performed using R software (version 4.4.2), employing packages such as dplyr for data manipulation and ggplot2 for graphical presentation. Where applicable, p-values were reported to three decimal places. No formal power calculation was conducted, as patient enrollment was determined by feasibility within the specified study timeframe. Model development and validation A random forest model was developed to predict achievement of ASDAS < 1.3 at 12 months, leveraging its ability to account for nonlinear relationships and complex interactions among predictors. Variables directly reflecting the outcome (ASDAS at 12 months) were excluded to prevent overfitting. Categorical variables were one-hot encoded, and continuous variables were scaled when necessary. The dataset was split into training (80%) and test (20%) subsets using stratified sampling to maintain class balance. Model performance was assessed using accuracy, precision, recall, and the area under the receiver operating characteristic curve (AUC). Feature importance was evaluated via the Gini impurity metric. All machine learning analyses were performed in Python v.3.9 with scikit-learn for model building, pandas and numpy for data handling, and matplotlib and seaborn for visualization. Ethical considerations This study was approved by the Institutional Review Board of Bach Mai Hospital (160520/QĐ-BVBM). Written informed consent was obtained from all participants and the study was performed in accordance with the ethical principles of the Declaration of Helsinki. Results In this study, a total of 480 patients were screened for eligibility. The final analysis included 222 patients in the full dose group and 146 in the reduction dose group (Fig. 1 ). Table 1 provides a comparison of key variables between group 1 (dose reduction) and group 2 (full dose). Significant differences in sex distribution were observed, with group 1 having a higher proportion of male patients, compared with group 2 (84.93% vs. 54.50%, p = 0.0001). Income levels also differed significantly, with 86.99% of patients in group 1 in the high-income category, compared with 55.41% in group 2 (p = 0.0001). Table 1 Comparison of baseline characteristics between dose reduction groups (n = 368) Variable Category Group 1 (dose reduction) (n = 146) Group 2 (full dose) (n = 222) Overall (n = 368) p-value Sex, n, (%) Male 124 (84.93%) 121 (54.50%) 245 (66.58%) 0.0001 Female 22 (15.07%) 101 (45.50%) 123 (33.42%) 0.0001 Income, n (%) Low 19 (13.01%) 99 (44.59%) 118 (32.07%) 0.0001 High 127 (86.99%) 123 (55.41%) 250 (67.93%) 0.0001 Age, (years) (mean ± SD) 31.28 ± 11.64 47.61 ± 19.11 41.13 ± 18.37 0.0001 BMI, (mean ± SD) 20.86 ± 2.59 20.87 ± 2.71 20.87 ± 2.65 0.9783 AS duration, (years) (mean ± SD) 7.08 ± 6.14 9.47 ± 5.98 8.52 ± 6.15 0.0001 Baseline BASDAI, (mean ± SD) 4.11 ± 1.35 5.03 ± 3.41 4.67 ± 2.82 0.0558 Baseline BASDAI, n (%) < 1.3 4 (2.74%) 32 (14.41%) 36 (9.78%) 0.001 1.3-4 42 (28.77%) 63 (28.38%) 105 (28.53%) 0.368 ≥ 4 100 (68.49%) 127 (57.21%) 227 (61.68%) 0.008 Baseline ASDAS, (mean ± SD) 2.15 ± 1.09 2.06 ± 1.16 2.10 ± 1.13 0.6857 Baseline ASDAS, n (%) 1.3–2.0 22 (15.07%) 51 (22.97%) 73 (19.84%) 0.074 2.1–3.4 59 (40.41%) 71 (31.98%) 130 (35.33%) 0.003 < 1.3 42 (28.77%) 65 (29.28%) 107 (29.08%) 0.16 ≥ 3.5 23 (15.75%) 35 (15.77%) 58 (15.76%) 0.107 Baseline CRP level, (mean ± SD) 5.92 ± 3.91 9.59 ± 6.24 8.14 ± 5.72 0.0000 SD, standard deviation; BMI, body mass index; CRP, C-reactive protein; BASDAI, Bath ankylosing spondylitis disease activity index; ASDAS ankylosing spondylitis disease activity score Patients in group 2 were older on average (47.61 ± 19.11 years) than those in group 1 (31.28 ± 11.64 years, p = 0.0001). The duration of AS was longer in group 2 (9.47 ± 5.98 years), compared with group 1 (7.08 ± 6.14 years, p = 0.0001). Baseline BASDAI did not show significant differences between the groups (p = 0.0558). However, group 1 had a higher proportion of patients with BASDAI ≥ 4, compared with group 2 (68.49% vs. 57.21%, p = 0.008). Regarding the baseline ASDAS, there was a higher percentage of patients with an ASDAS of 1.3–2.0 in group 1 than in group 2 (40.41% vs. 31.98%, p = 0.003). Baseline CRP levels were significantly lower in group 1 (5.92 ± 3.91) than in group 2 (9.59 ± 6.24, p = 0.0000). Body mass index was similar between the two groups (20.86 ± 2.59 vs. 20.87 ± 2.71, p = 0.9783). Table 2 shows the distribution of various DMARD therapies among patients in group 1 (dose reduction) and group 2 (full dose), together with overall values and statistical comparisons between the groups. For most combination therapies, including secukinumab + adalimumab and secukinumab + infliximab + golimumab, there were no significant differences between the groups (p > 0.05). However, notable differences were observed for certain therapies. Secukinumab was used significantly more frequently in group 1 (24.66%) than in group 2 (14.86%, p = 0.027). Infliximab was administered significantly more frequently in group 1 (26.71%) compared with group 2 (9.01%, p < 0.001). Conversely, adalimumab was used more frequently in group 1 (14.38%) than in group 2 (6.76%, p = 0.026). Furthermore, the secukinumab + golimumab and secukinumab + infliximab combinations were significantly more frequently administered in group 2 than in group 1 (p = 0.016 and p = 0.031, respectively). Table 2 Characteristics of disease-modifying antirheumatic drug therapies used in the study Therapy Overall (n = 368) Group 1 (dose reduction) (n = 146) Group 2 (full dose) (n = 222) p-value Adalimumab + infliximab 31 (8.42%) 13 (8.90%) 18 (8.11%) 0.939 Secukinumab 69 (18.75%) 36 (24.66%) 33 (14.86%) 0.027 Secukinumab + golimumab 21 (5.71%) 4 (2.74%) 17 (7.66%) 0.078 Golimumab 34 (9.24%) 8 (5.48%) 26 (11.71%) 0.066 Infliximab 59 (16.03%) 39 (26.71%) 20 (9.01%) < 0.001 Adalimumab 36 (9.78%) 21 (14.38%) 15 (6.76%) 0.026 Secukinumab + infliximab 27 (7.34%) 14 (9.59%) 13 (5.86%) 0.255 Secukinumab + adalimumab 17 (4.62%) 4 (2.74%) 13 (5.86%) 0.255 Adalimumab + golimumab 15 (4.08%) 1 (0.68%) 14 (6.31%) 0.016 Infliximab + golimumab 13 (3.53%) 2 (1.37%) 11 (4.95%) 0.125 Etanercept + adalimumab + infliximab 15 (4.08%) 1 (0.68%) 14 (6.31%) 0.016 Secukinumab + infliximab + golimumab 17 (4.62%) 2 (1.37%) 15 (6.76%) 0.031 Table 3 presents a comparison of the outcome variables between group 1 (dose reduction) and group 2 (full dose). Significant improvements were observed in several measures, which favored dose reduction. The reduction in CRP level from baseline to month 12 was significantly higher in group 1 (median − 4.65 [4.44]) than in group 2 (median − 1.32 [12.00], p < 0.001), indicating a significant difference in inflammation reduction. Similarly, the decrease in disease activity scores, based on BASDAI and ASDAS assessments from baseline to month 12, showed significantly greater improvements in group 1 (median improvement in BASDAI: -3.00 [1.96]; median improvement in ASDAS: -1.72 [1.50]) compared with group 2 (median improvement in BASDAI: -0.42 [5.59]; median improvement in ASDAS: -0.15 [2.48], p < 0.001 for both). At month 12, group 1 exhibited significantly lower CRP levels (median: 0.27 [1.19] mg/L) than group 2 did (7.15 [9.40] mg/L, p < 0.001), suggesting better inflammation control in the dose reduction group. Regarding the disease activity scores at month 12, 97.26% of the patients in group 1 had a BASDAI < 4, compared with 50.90% in group 2 (p < 0.001). Based on the ASDAS, 93.15% of the patients in group 1 achieved scores of 2.1–3.4, indicative of better disease control, whereas group 2 had a significantly higher proportion of patients with moderate, high, and very high disease activity (p < 0.001). Table 3 Comparison of outcome variables between dose reduction and full dose groups Variable Overall (n = 368) Group 1 (dose reduction) (n = 146) Group 2 (full dose) (n = 222) p-value CRP level improvement, median [IQR] -3.28 [8.03] -4.65 [4.44] -1.32 [12.00] < 0.001 BASDAI improvement, median [IQR] -2.08 [4.50] -3.00 [1.96] -0.42 [5.59] < 0.001 ASDAS improvement, median [IQR] -1.00 [2.33] -1.72 [1.50] -0.15 [2.48] < 0.001 CRP level at month 12, median [IQR] (mg/L) 2.77 [8.78] 0.27 [1.19] 7.15 [9.40] < 0.001 BASDAI at month 12 (n (%) < 4 255 (69.29%) 142 (97.26%) 113 (50.90%) < 0.001 ≥ 4 113 (30.71%) 4 (2.74%) 109 (49.10%) < 0.001 ASDAS at month 12 (n (%) < 1.3 50 (13.59%) 9 (6.16%) 41 (18.47%) < 0.001 1.3–2.0 65 (17.66%) 1 (0.68%) 64 (28.83%) < 0.001 2.1–3.4 217 (58.97%) 136 (93.15%) 81 (36.49%) < 0.001 ≥ 3.5 36 (9.78%) 0 (0.00%) 36 (16.22%) < 0.001 CRP, C-reactive protein; BASDAI, Bath ankylosing spondylitis disease activity index; ASDAS ankylosing spondylitis disease activity score; IQR, interquartile range As shown in Table 4 , the dose reduction demonstrated cost savings and improved effectiveness compared with the full dose across most criteria. For the improvement of the ASDAS, the ICER was - $ 5,870.20, with a 95% confidence interval (CI) of (- $ 7,982.43; - $ 11,194.39), indicating significant cost-effectiveness. Similarly, achieving the ASDAS target (< 1.3) showed an ICER of $ 16,921.46, with a narrow CI (95% CI: $ 7,982.43; $ 11,194.39). CRP improvement had an ICER of - $ 2,409.25, and the BASDAI improvement reported an ICER of - $ 4,085.17, with their respective 95% CIs suggesting robust cost savings for these outcomes. Conversely, achieving the BASDAI target (< 4.0) showed a higher ICER of $ 20,682.78, indicating that dose reduction was less cost-effective in this criterion. The results for incremental effectiveness showed 0.46 and 0.57 for the BASDAI and ASDAS targets, respectively, with their 95% CIs confirming the statistical significance in favor of dose reduction. Table 4 Incremental cost effectiveness ratio results for dose reduction versus full dose (costs in $ ) Effectiveness criterion ICER (cost per unit of effectiveness) ICER 95% CI Incremental effectiveness Incremental effectiveness 95% CI CRP level improvement -2409.25 (7982.43; 11194.39) -3.98 (-5.75; -2.21) BASDAI improvement -4085.17 (7982.43; 11194.39) -2.35 (-3.16; -1.53) ASDAS improvement -5870.20 (7982.43; 11194.39) -1.63 (-2.04; -1.23) BASDAI target (< 4.0) 20682.78 (7982.43; 11194.39) 0.46 (0.37; 0.56) ASDAS target (< 1.3) 16921.46 (7982.43; 11194.39) 0.57 (0.46; 0.67) CRP, C-reactive protein; BASDAI, Bath ankylosing spondylitis disease activity index; ASDAS ankylosing spondylitis disease activity score; ICER, incremental cost effectiveness ratio; CI, confidence interval ICER represents the cost per unit increase in effectiveness for dose reduction compared with full dose; Incremental effectiveness denotes the difference in effectiveness between the dose reduction and full dose groups. Table 5 presents the analysis of the predictors for achieving an ASDAS < 1.3 (inactive disease) in patients with AS undergoing DMARD treatment. Key predictors with significant impact include “reduce dose,” baseline CRP level, and baseline ASDAS, all with p-values below 0.001. The negative estimates for “reduce dose” and the baseline ASDAS suggest that these factors are associated with a higher probability of achieving inactive disease. Conversely, a positive association was observed with baseline CRP levels. In our study, a significant difference was found between male and female patients in the achievement of low disease activity (ASDAS < 1.3); compared with female patients, male patients were more likely to reach this state of low disease activity. Other variables, including age, AS duration, income, and drug therapy, did not show significant predictive value, indicating a limited influence on achieving low disease activity. A similar multivariate analysis was performed to explore the associations between various factors and achieving a BASDAI < 4. However, no significant association was observed. The random forest models demonstrated robust performance in predicting ASDAS < 1.3, as summarized in Table 5 . The medium feature model (20 variables) achieved the best performance, with an AUC of 81.86%, precision of 86.49%, and accuracy of 77.03%, balancing predictive power and model complexity. The full model (50 variables) performed similarly (AUC, 81.21%; precision, 84.62%) but was more complex. The reduced model (10 variables) offered a streamlined approach with an AUC of 81.63% and precision of 76.74%. The key clinical variables model (five variables), focusing on the core clinical factors, delivered reasonable performance (AUC, 81.10%; precision 80.00%) and was ideal for resource-limited settings. These findings highlight the utility of the medium feature model as the optimal choice for balancing interpretability and accuracy, whereas the key clinical variables model offers simplicity and clinical relevance. Table 5 Performance metrics of random forest models for predicting ankylosing spondylitis disease activity score < 1.3 Model Accuracy AUC Precision Recall Number of variables Full model (encoded) 77.03% 81.21% 84.62% 75.00% 50 Reduced model (encoded) 71.62% 81.63% 76.74% 75.00% 10 Medium feature model (top 20) 77.03% 81.86% 86.49% 72.73% 20 Key clinical variables model 72.97% 81.10% 80.00% 72.73% 5 AUC, area under the receiver operating characteristic curve The feature importance analysis (Fig. 2 ) highlights the key predictors of achieving ASDAS < 1.3 using the medium feature model. Total cost (USD), CRP at the month 12 (mg/L), and 12-month total cost (USD) are identified as the most influential variables, reflecting the importance of financial and inflammatory factors. Clinical measures, such as the BASDAI at month 12 and dose adjustment emphasize the importance of disease severity and treatment optimization. Early indicators, such as CRP level at month 1 and CRP at month 3 demonstrate the relevance of intermediate responses to treatment. These results underscore the value of integrating clinical, economic and longitudinal factors into personalized treatment strategies. Discussion This study highlights the effectiveness of dose reduction strategies for bDMARDs for treating AS. Patients who underwent dose reduction achieved superior clinical outcomes compared with those on full dose therapy, with significant improvements in CRP levels (4.65 vs. 0.90 mg/L, p < 0.001), the BASDAI (3.00 vs. 0.42, p < 0.001), and ASDAS (1.72 vs. 0.15, p < 0.001). Interestingly, 93.15% of the dose reduction group achieved inactive disease (ASDAS < 1.3), compared with 47.62% in the full dose group. Economic evaluation through ICER analysis demonstrated the cost-effectiveness of dose reduction, with a favorable ICER of - $ 16,772.62 to achieve ASDAS < 1.3. Predictive modeling using random forest further identified key predictors of successful treatment outcomes, including the baseline CRP level, ASDAS, and early response to treatment. Our findings align with earlier studies supporting the clinical feasibility of bDMARD dose reduction. Recent studies also support the clinical feasibility of bDMARD dose reduction in AS and other inflammatory arthritis conditions. Evidence shows that 50–60% of patients with AS and patients with psoriatic arthritis can successfully maintain reduced TNFi doses for approximately a year ( 20 , 21 ). Disease activity-guided tapering and fixed dose reduction appear to be effective strategies, whereas complete discontinuation is not recommended due to high flare rates ( 22 , 23 ). Patients express concerns about dose reduction but are generally willing to try if given a clear rationale and participate in collaborative decision-making ( 24 ). Successful tapering is associated with lower disease activity prior to dose reduction ( 20 ). Cost-effectiveness analyses of dose adjustment strategies for axial spondyloarthritis treatments have shown promising results. Tapering etanercept by 25% every 3 months is a pragmatic approach for more cost-effective use ( 25 ). Secukinumab is cost-effective in both biologic-naïve and biologic-experienced patients ( 26 ). Adalimumab is considered cost-effective, compared with conventional therapy, over a 30-year period ( 27 ). Infliximab administered every 6 weeks is cost-effective, compared with on-demand regimens ( 28 ). Tocilizumab addition to standard care is estimated to be cost-effective in both methotrexate-tolerant and contraindicated patients ( 29 ). Disease activity-guided dose optimization of TNFis results in significant cost savings without a relevant loss of quality of life ( 30 ). Higher treatment persistence with subcutaneous TNFis is cost-effective from both the payer and societal perspectives ( 31 ). Therapeutic drug monitoring-guided adalimumab therapy has increased quality-adjusted life years at lower costs ( 32 ). Machine learning (ML) techniques have also shown promise in improving the diagnosis and treatment of AS. ML models can predict early AS diagnosis with high accuracy using administrative claims and electronic medical record data ( 33 , 34 ). These models can also identify patients likely to require early TNFi treatment ( 35 , 36 ). Additionally, ML and deep learning approaches facilitate early diagnosis through patient characteristic profiling and biomarker identification ( 37 ). The integration of ICER analysis and ML provides a robust framework for implementing dose reduction strategies in clinical practice. The economic benefits of dose reduction, including significant cost savings, can enhance patient adherence, particularly in resource-constrained settings. Random forest models enable clinicians to identify patients who are more likely to benefit from dose reduction based on clinical and biomarker predictors, such as the baseline CRP level and early improvement in ASDAS. This personalized approach supports the more efficient use of bDMARDs, balancing clinical efficacy and financial sustainability. The strengths of this study include its prospective design, comprehensive evaluation of clinical and economic outcomes, and application of ML for predictive modeling. However, the study also has limitations, including the nonrandomized allocation of patients to the dose reduction and full dose groups, which may have introduced a selection bias. Additionally, the relatively short follow-up period limits insights into the long-term sustainability and safety of dose reduction. Furthermore, the ICER results may vary across healthcare systems, reducing generalizability. Future studies should validate these findings in larger multicenter trials with diverse populations and longer follow-up periods. Investigating additional predictors, including patient-reported outcomes and quality-of-life measures, can refine candidate selection for dose reduction. Further exploration of real-world ICER results across different healthcare systems will strengthen the evidence for the economic feasibility of dose adjustment strategies. Overall, this study supports the feasibility of dose-reduction of bDMARDs in patients with AS, demonstrating that it can maintain clinical efficacy and significantly reduce treatment costs. These findings advocate for future studies supporting dose reduction strategies in clinical practice, with careful patient selection and monitoring to ensure optimal results. Such approaches will contribute to more sustainable healthcare by balancing effective disease management with economic considerations. Conclusions This study provides compelling evidence that the reduction in the dose of bDMARDs for treating ankylosing spondylitis is both clinically effective and economically advantageous. Patients with dose reduction achieved better disease control rates (ASDAS < 1.3) while significantly reducing treatment costs, as demonstrated by a favorable ICER. The integration of random forest modeling further enhanced the study by identifying key predictors of treatment success, enabling a personalized approach to dose adjustment. These findings not only reinforce the feasibility of dose reduction in maintaining long-term disease remission but also highlight its potential to alleviate the financial burden associated with bDMARD therapy. Combining clinical, economic, and predictive insights, this study establishes a strong foundation for optimizing treatment strategies in AS, advocating for their adoption in routine practice with customized patient management. Abbreviations AS, ankylosing spondylitis ASDAS, ankylosing spondylitis disease activity score BASDAI, Bath ankylosing spondylitis disease activity index bDMARDs, biologic disease-modifying antirheumatic drugs ICER, incremental cost effectiveness ratios NSAIDs, nonsteroidal anti-inflammatory drugs TNFi, tumor necrosis factor inhibitors Declarations Acknowledgments We would like to thank all the physicians and nurses at the Rheumatology Center and Laboratory Department of Bach Mai Hospital, Vietnam, for their help with patient recruitment and clinical laboratory testing. Ethics approval and consent to participate This study was reviewed and approved by the Institutional Review Board of Bach Mai Hospital, approval number of 160520/QĐ-BVBM. All methods were carried out in accordance with relevant guidelines and regulations. Written informed consent was obtained from all individual participants included in the study. Consent for publication Not applicable Availability of data and materials All data generated or analyzed during this study are included in this published article and its supplementary information files. Any additional data or materials can be obtained from the corresponding author upon reasonable request. Competing interests The authors declare that they have no competing interests. Funding None of the authors have any financial or non-financial competing interests to disclose. Authors' contributions BHB and HVD conceived and designed the study. NTTP, VTTH, NTTN, and NTNH contributed to data acquisition, analysis, and interpretation. HVD supervised the project and served as the corresponding author. BHB, NTTP, and VTTH drafted the initial version of the manuscript. NTTN, NTNH, and HVD critically revised the manuscript for important intellectual content. Acknowledgements The authors would like to express their sincere gratitude to the clinical and administrative staff at Bach Mai Hospital for their invaluable support and collaboration throughout the study period. Special thanks go to all the patients who participated in this research for their time and cooperation. References McVeigh CM, Cairns AP. Diagnosis and management of ankylosing spondylitis. BMJ (Clinical Res ed). 2006;333(7568):581–5. Dean LE, Jones GT, MacDonald AG, Downham C, Sturrock RD, Macfarlane GJ. Global prevalence of ankylosing spondylitis. Rheumatology (Oxford). 2014;53(4):650–7. Brewerton DA, Hart FD, Nicholls A, Caffrey M, James DC, Sturrock RD. Ankylosing spondylitis and HL-A 27. Lancet (London England). 1973;1(7809):904–7. Wordsworth BP, Cohen CJ, Davidson C, Vecellio M. Perspectives on the Genetic Associations of Ankylosing Spondylitis. Front Immunol. 2021;12:603726. Boonen A, van der Linden SM. The burden of ankylosing spondylitis. J Rheumatol Suppl. 2006;78:4–11. Hegyi R, Nagy B, Koncz A, Huybrechts I, Lavicky J, Ferenczik A. Burden Of Disease Analysis Of Ankylosing Spondylitis In Hungary. Value Health. 2014;17(7):A376–7. Kobelt G, Sobocki P, Mulero J, Gratacos J, Pocovi A, Collantes-Estevez E. The burden of ankylosing spondylitis in Spain. Value Health. 2008;11(3):408–15. Webers C, Ortolan A, Sepriano A, Falzon L, Baraliakos X, Landewe RBM, et al. Efficacy and safety of biological DMARDs: a systematic literature review informing the 2022 update of the ASAS-EULAR recommendations for the management of axial spondyloarthritis. Ann Rheum Dis. 2023;82(1):130–41. Klavdianou K, Tsiami S, Baraliakos X. New developments in ankylosing spondylitis—status in 2021. Rheumatology. 2021;60(Supplement6):vi29–37. Oo WM, Yu SP, Daniel MS, Hunter DJ. Disease-modifying drugs in osteoarthritis: current understanding and future therapeutics. Expert Opin Emerg Drugs. 2018;23(4):331–47. Gregori D, Giacovelli G, Minto C, Barbetta B, Gualtieri F, Azzolina D, et al. Association of Pharmacological Treatments With Long-term Pain Control in Patients With Knee Osteoarthritis: A Systematic Review and Meta-analysis. JAMA. 2018;320(24):2564–79. Franke LC, Ament AJ, van de Laar MA, Boonen A, Severens JL. Cost-of-illness of rheumatoid arthritis and ankylosing spondylitis. Clin Exp Rheumatol. 2009;27(4 Suppl 55):S118–23. Greenberg JD, Palmer JB, Li Y, Herrera V, Tsang Y, Liao M. Healthcare Resource Use and Direct Costs in Patients with Ankylosing Spondylitis and Psoriatic Arthritis in a Large US Cohort. J Rheumatol. 2016;43(1):88–96. van der Linden S, Valkenburg HA, Cats A. Evaluation of diagnostic criteria for ankylosing spondylitis. A proposal for modification of the New York criteria. Arthritis Rheum. 1984;27(4):361–8. Gold MRSJ, Russell LB, Weinstein MC, editors. Cost-Effectiveness in Health and Medicine. ed. n, editor. Oxford University Press; 2016. Goie The HS, Steven MM, van der Linden SM, Cats A. Evaluation of diagnostic criteria for ankylosing spondylitis: a comparison of the Rome, New York and modified New York criteria in patients with a positive clinical history screening test for ankylosing spondylitis. Br J Rheumatol. 1985;24(3):242–9. van der Heijde D, Lie E, Kvien TK, Sieper J, Van den Bosch F, Listing J, et al. ASDAS, a highly discriminatory ASAS-endorsed disease activity score in patients with ankylosing spondylitis. Ann Rheum Dis. 2009;68(12):1811–8. Garrett S, Jenkinson T, Kennedy LG, Whitelock H, Gaisford P, Calin A. A new approach to defining disease status in ankylosing spondylitis: the Bath Ankylosing Spondylitis Disease Activity Index. J Rheumatol. 1994;21(12):2286–91. Lukas C, Landewe R, Sieper J, Dougados M, Davis J, Braun J, et al. Development of an ASAS-endorsed disease activity score (ASDAS) in patients with ankylosing spondylitis. Ann Rheum Dis. 2009;68(1):18–24. Fong W, Holroyd C, Davidson B, Armstrong R, Harvey N, Dennison E, et al. The effectiveness of a real life dose reduction strategy for tumour ne crosis factor inhibitors in ankylosing spondylitis and psoriatic arthr itis. Rheumatology. 2016;55(10):1837–42. vdV SAE. F BK, P H, R B, H B, Patient-tailored dose reduction of TNF-α blocking agents in ankylosing spondylitis patients with stable low disease activity in daily clinic al practice. Clin Exp Rheumatol. 2015. Verhoef LM, Tweehuysen L, Hulscher ME, Fautrel B, den Broeder AA. bDMARD Dose Reduction in Rheumatoid Arthritis: A Narrative Review with Systematic Literature Search. Rheumatol Ther. 2017;4(1):1–24. Chaudhary H, Bittar M, Daoud A, Magrey M. Dose Tapering and Discontinuation of Biologic DMARDs in Axial Spondylo arthritis: A Narrative Review (2023 SPARTAN Annual Meeting Proceedings). Curr Rheumatol Rep. 2024;26(5):155 – 63. Hewlett S, Haig-Ferguson A, Rose‐Parfitt E, Halls S, Freke S, Creamer P. Dose reduction of biologic therapy in inflammatory arthritis: A qualit ative study of patients' perceptions and needs. Musculoskelet Care. 2018;17(1):63–71. Jun FL, Yu Z, Dongying W, Hanshi C. X, L L. Efficiency of dose reduction strategy of etanercept in patients with a xial spondyloarthritis. Clinical and Experimental Rheumatology; 2018. Emery P, Van Keep M, Beard S, Graham C, Miles L, Jugl SM, et al. Cost Effectiveness of Secukinumab for the Treatment of Active Ankylosi ng Spondylitis in the UK. PharmacoEconomics. 2018;36(8):1015–27. Botteman MF, Hay JW, Luo MP, Curry AS, Wong RL, van Hout BA. Cost effectiveness of adalimumab for the treatment of ankylosing spond ylitis in the United Kingdom. Rheumatology. 2007;46(8):1320–8. Fautrel B, Benhamou M, Breban M, Roy C, Lenoir C, Trape G, et al. Cost effectiveness of two therapeutic regimens of infliximab in ankylo sing spondylitis: economic evaluation within a randomised controlled t rial. Ann Rheum Dis. 2010;69(2):424–7. Diamantopoulos A, Finckh A, Huizinga T, Sungher DK, Sawyer L, Neto D, et al. Tocilizumab in the Treatment of Rheumatoid Arthritis: A Cost-Effective ness Analysis in the UK. PharmacoEconomics. 2014;32(8):775–87. Kievit W, van Herwaarden N, van den Hoogen FHJ, van Vollenhoven RF, Bijlsma JWJ, van den Bemt BJF, et al. Disease activity-guided dose optimisation of adalimumab and etanercept is a cost-effective strategy compared with non-tapering tight control rheumatoid arthritis care: analyses of the DRESS study. Ann Rheum Dis. 2016;75(11):1939–44. Ivergård M, Dalén J, Svedbom A, Black CM, Borse RH, Kachroo S. FRI0636 The value of persistence in treatment with subcutaneous tnf-al pha inhibitors for ankylosing spondylitis. Ann Rheum Dis. 2018;77:840–1. Gómez-Arango C, Gorostiza I, Úcar E, García-Vivar ML, Pérez CE, De Dios JR, et al. Cost-Effectiveness of Therapeutic Drug Monitoring-Guided Adalimumab Th erapy in Rheumatic Diseases: A Prospective, Pragmatic Trial. Rheumatol Ther. 2021;8(3):1323–39. Walsh JA, Rozycki M, Yi E, Park Y. Application of machine learning in the diagnosis of axial spondyloarth ritis. Curr Opin Rheumatol. 2019;31(4):362–7. Kennedy J, Kennedy N, Cooksey R, Choy E, Siebert S, Rahman M, et al. Predicting a diagnosis of ankylosing spondylitis using primary care he alth records–A machine learning approach. PLoS ONE. 2023;18(3):e0279076. Lee S, Eun Y, Kim H, Cha H-S, Koh E-M, Lee J. Machine learning to predict early TNF inhibitor users in patients with ankylosing spondylitis. Sci Rep. 2020;10(1). Deodhar A, Rozycki M, Garges C, Shukla O, Arndt T, Grabowsky T, et al. Use of machine learning techniques in the development and refinement o f a predictive model for early diagnosis of ankylosing spondylitis. Clin Rheumatol. 2019;39(4):975–82. Dhall S, Vaish A, Vaishya R. Machine learning and deep learning for the diagnosis and treatment of ankylosing spondylitis- a scoping review. J Clin Orthop Trauma. 2024;52:102421. Additional Declarations No competing interests reported. Supplementary Files finaldata.csv Supplementary File 1: Study data. Cite Share Download PDF Status: Published Journal Publication published 26 May, 2025 Read the published version in BMC Rheumatology → Version 1 posted Editorial decision: Revision requested 03 Feb, 2025 Editor assigned by journal 28 Jan, 2025 Submission checks completed at journal 28 Jan, 2025 First submitted to journal 28 Jan, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-5917710","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":408153416,"identity":"1c3416a3-8353-4fe1-85c9-ab286c938ff1","order_by":0,"name":"Hai Binh Bui","email":"","orcid":"","institution":"Bach Mai Hospital","correspondingAuthor":false,"prefix":"","firstName":"Hai","middleName":"Binh","lastName":"Bui","suffix":""},{"id":408153417,"identity":"98ab8a45-66b7-4ede-aa55-c801169e03da","order_by":1,"name":"Thi Thu Phuong Nguyen","email":"","orcid":"","institution":"Hai Phong University of Medicine and Pharmacy","correspondingAuthor":false,"prefix":"","firstName":"Thi","middleName":"Thu Phuong","lastName":"Nguyen","suffix":""},{"id":408153418,"identity":"f27a326c-7cf5-4d4e-be7c-36ff9f3a2524","order_by":2,"name":"Thi Thanh Hang Vu","email":"","orcid":"","institution":"Hanoi Medical University","correspondingAuthor":false,"prefix":"","firstName":"Thi","middleName":"Thanh Hang","lastName":"Vu","suffix":""},{"id":408153419,"identity":"8ce0ec50-3775-4ba1-b3b5-0e998d4dc81d","order_by":3,"name":"Thi Thuc Nhan Ngo","email":"","orcid":"","institution":"Nam Dinh University Of Nursing","correspondingAuthor":false,"prefix":"","firstName":"Thi","middleName":"Thuc Nhan","lastName":"Ngo","suffix":""},{"id":408153420,"identity":"c4ad8f0c-a275-4246-a4e1-223bebdc732b","order_by":4,"name":"Thi Nhu Hoa Nguyen","email":"","orcid":"","institution":"Bach Mai Hospital","correspondingAuthor":false,"prefix":"","firstName":"Thi","middleName":"Nhu Hoa","lastName":"Nguyen","suffix":""},{"id":408153421,"identity":"531fba2d-561a-4905-a337-d21199eb4370","order_by":5,"name":"Dung Van Hoang","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA70lEQVRIiWNgGAWjYDCCA2BSAoiZDz74AKTY2InUAtTDlmw4A6SFmTgtIGt4zIR5QExCWvhu5B5+XVBhUWc+I8GM2ebXNnk+ZgbGDx9zcGuRvJGXZj3jjISEzI2EtMe5fbcN25gZmCVnbsOtxeBGjpkxb5sEECQcN87tuc0I1MLGzEuclsQ2acue2/bEaDF+DNGSzCbN8ON2IkEtkmfemDHznJGQnMHzjNmwt+F2chszYzNev/AdzzH+zFNRxy/Bnv/xwY8/t23ntzcf/PARjxYgYJOAMxnbwGQDXvVAwPwBwf5DSPEoGAWjYBSMRAAAh2pL7IcqIxMAAAAASUVORK5CYII=","orcid":"","institution":"Hai Phong International Hospital","correspondingAuthor":true,"prefix":"","firstName":"Dung","middleName":"Van","lastName":"Hoang","suffix":""}],"badges":[],"createdAt":"2025-01-28 09:53:36","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5917710/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5917710/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1186/s41927-025-00516-9","type":"published","date":"2025-05-26T15:57:10+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":75087239,"identity":"54c96621-9dea-484d-b4d4-44e5534febdf","added_by":"auto","created_at":"2025-01-30 10:20:02","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":62813,"visible":true,"origin":"","legend":"\u003cp\u003eFlow diagram\u003c/p\u003e","description":"","filename":"Figure11.png","url":"https://assets-eu.researchsquare.com/files/rs-5917710/v1/6f80ad8606be0c2ee931ba79.png"},{"id":75088607,"identity":"49e54706-249a-4c0f-94fa-1cd362227163","added_by":"auto","created_at":"2025-01-30 10:28:02","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":176224,"visible":true,"origin":"","legend":"\u003cp\u003eFeature importance in predicting ankylosing spondylitis disease activity score \u0026lt;1.3 using the medium feature model\u003c/p\u003e","description":"","filename":"Figure2.png","url":"https://assets-eu.researchsquare.com/files/rs-5917710/v1/737a1245275d86a139a3f4f3.png"},{"id":83782813,"identity":"2160c19c-9c17-42a6-8be9-61a1397bb6a9","added_by":"auto","created_at":"2025-06-02 16:06:49","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1199938,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5917710/v1/c85f1f02-85c2-466d-a2fb-2ee4cbef5f36.pdf"},{"id":75088605,"identity":"43e21c99-120e-4127-be5f-b3f4309fedf7","added_by":"auto","created_at":"2025-01-30 10:28:02","extension":"csv","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":99746,"visible":true,"origin":"","legend":"\u003cp\u003eSupplementary File 1: Study data.\u003c/p\u003e","description":"","filename":"finaldata.csv","url":"https://assets-eu.researchsquare.com/files/rs-5917710/v1/69b1dbae71886999eee9e35b.csv"}],"financialInterests":"No competing interests reported.","formattedTitle":"Personalized dose reduction strategies for biologic disease-modifying antirheumatic drugs for treating ankylosing spondylitis: a clinical and economic evaluation with predictive modeling","fulltext":[{"header":"Background","content":"\u003cp\u003eAnkylosing spondylitis (AS) is a chronic inflammatory disease that primarily affects the spine and sacroiliac joints, causing pain, stiffness, and potential fusion of the vertebrae. This progressive condition is a type of spondyloarthritis, which also includes psoriatic and reactive arthritis (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e). The prevalence estimate of AS per 10,000 people is 23.8 globally; 16.7 in Asia; 31.9 in North America; 10.2 in Latin America; and 7.4 in Africa (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe underlying mechanisms of AS involve genetic, environmental and immunological factors, and a notable association with HLA-B27. Despite advances in understanding the pathophysiology of AS, the disease remains incurable, and treatments focus on managing symptoms, reducing inflammation, and preventing complications (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e). AS imposes a significant burden on patients and society. This disease affects quality of life, productivity at work, and healthcare costs (\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e). The total annual cost per patient ranges from \u0026euro;5,155 in Hungary to \u0026euro;20,328 in Spain, with indirect costs accounting for a substantial proportion (\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eIn recent years, therapeutic guidelines for AS have evolved, particularly with the 2022 update of the Assessment of SpondyloArthritis International Society and European Alliance of Associations for Rheumatology recommendations (\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e). These guidelines emphasize early diagnosis and a comprehensive treatment approach, prioritizing both nonpharmacological interventions such as physical therapy and pharmacological options, including nonsteroidal anti-inflammatory drugs (NSAIDs) and disease-modifying antirheumatic drugs (DMARDs). Biologic DMARDs (bDMARDs) and Janus kinase inhibitors are recommended as second-line treatments for patients who show inadequate response to NSAIDs. This update highlights the role of DMARDs, particularly tumor necrosis factor inhibitors (TNFi) and interleukin-17 inhibitors (IL-17i), in the management of moderate to severe AS. These agents target pathways involved in the inflammatory response, providing substantial relief and improving functional outcomes (\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eDMARDs, including biologics such as etanercept, adalimumab, and infliximab, have transformed AS management. By targeting TNF-α and IL-17, these drugs reduce inflammation, stop disease progression, and improve quality of life (\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e). Despite their efficacy, DMARDs pose a significant financial burden due to their high treatment costs, limiting accessibility for patients with AS (\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e). In many healthcare settings, the cost of DMARD therapy is partially or fully borne by the patients, making long-term adherence challenging. With TNF inhibitors and IL-17A monoclonal antibodies proving effective for AS management and becoming standard treatments, the direct expenses related to AS care have increased considerably. In 2012, the annual direct costs of AS in the US were estimated at approximately \u003cspan\u003e$\u003c/span\u003e6,514 for medical care and \u003cspan\u003e$\u003c/span\u003e11,162 for prescription drugs per patient (\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e). Consequently, researchers and clinicians have explored strategies to optimize DMARD use, such as dose tapering or extension, to reduce costs without compromising efficacy.\u003c/p\u003e \u003cp\u003eStudies have investigated dose reduction strategies for DMARDs in patients with AS and other rheumatic diseases. Dose reduction, which involves lengthening the interval between doses, aims to maintain disease control while reducing medication frequency and associated costs. Research on rheumatoid arthritis, another condition treated with DMARDs, suggests that dose reduction may be feasible in patients in sustained remission or in those with low disease activity. However, the evidence for AS is less robust, with mixed results regarding the long-term outcomes of dose reduction. Some studies have indicated that extending the dose interval in patients with stable AS can maintain low disease activity, whereas others caution against potential flare-ups and diminished control over disease progression.\u003c/p\u003e \u003cp\u003eThe high cost of bDMARDs has created a demand for more cost-effective treatment options. Economic analyses in rheumatology have highlighted the financial impact of these drugs on healthcare systems and patients. These studies often employ metrics, such as the incremental cost-effectiveness ratio (ICER), to assess whether the health benefits of dose reduction justify the potential cost savings. However, the variability in disease activity assessments and patient responses to DMARDs pose challenges for establishing standardized protocols for dose reduction.\u003c/p\u003e \u003cp\u003eTherefore, this study aimed to evaluate the efficacy and cost-effectiveness of DMARD dose reduction in patients over a 12-month period. By analyzing clinical outcomes such as C-reactive protein (CRP) levels, the Bath ankylosing spondylitis disease activity index (BASDAI), and ankylosing spondylitis disease activity score (ASDAS), this study sought to determine whether dose reduction can achieve disease control comparable to standard dosing. In addition, cost-related outcomes, including ICER, were evaluated to determine the financial implications of dose reduction. This investigation contributes to the growing body of literature on personalized medicine approaches for AS by addressing the clinical and economic aspects of optimizing DMARD therapy.\u003c/p\u003e"},{"header":"Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eAim, design, and setting of the study\u003c/h2\u003e \u003cp\u003eThis prospective case-control study aimed to evaluate the clinical and cost-effectiveness of personalized dose reduction strategies for DMARDs among patients with AS. The study was conducted at Bach Mai Hospital in Hanoi, Vietnam, from 1 January 2021 to 30 September 2024.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eParticipants\u003c/h3\u003e\n\u003cp\u003eEligible participants were adults (\u0026ge;\u0026thinsp;18 years old) diagnosed with AS according to the modified New York criteria (\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e). Additional inclusion criteria included a history of at least 3 months of DMARD therapy prior to enrollment. Patients with coexisting autoimmune conditions, severe organ dysfunction, or any other conditions that could interfere with treatment evaluation were excluded.\u003c/p\u003e\n\u003ch3\u003eInterventions\u003c/h3\u003e\n\u003cp\u003eParticipants were allocated to one of two groups according to their treatment plan:\u003c/p\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003eGroup 1 (Dose reduction): Patients who achieved sustained low disease activity\u0026mdash;defined as an Ankylosing Spondylitis Disease Activity Score (ASDAS)\u0026thinsp;\u0026lt;\u0026thinsp;2.1 for at least 3 consecutive months\u0026mdash;underwent extended dosing intervals. They also had stable clinical and laboratory parameters, including normal C-reactive protein (CRP) levels and no significant disease flares or radiographic progression over the previous evaluation period. Dose reduction strategies were as follows:\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eEtanercept (Enbrel\u003csup\u003e\u0026reg;\u003c/sup\u003e): reduced from 50 mg weekly to 50 mg bi-weekly.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eSecukinumab (Fraizeron\u003csup\u003e\u0026reg;\u003c/sup\u003e): reduced from 150 mg monthly to 150 mg bi-monthly.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eAdalimumab (Humira\u003csup\u003e\u0026reg;\u003c/sup\u003e): reduced from 40 mg every two weeks to 40 mg every three weeks.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eInfliximab (Remicade\u003csup\u003e\u0026reg;\u003c/sup\u003e): reduced from 5 mg/kg every 6\u0026ndash;8 weeks to 5 mg/kg every 10\u0026ndash;12 weeks.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eGolimumab (Simponi\u003csup\u003e\u0026reg;\u003c/sup\u003e): reduced from 50 mg monthly to 50 mg every two months.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eGroup 2 (Full dose): Patients who did not meet criteria for dose reduction or had higher disease activity remained on standard, full-dose DMARD regimens as per manufacturer recommendations, without alterations in dose frequency or quantity.\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003c/p\u003e \u003cp\u003eBoth groups followed the same follow-up schedule, with evaluations every 3 months to monitor clinical status, laboratory markers (e.g., CRP, erythrocyte sedimentation rate), and radiographic changes. Patients were also assessed at each visit for adverse events, treatment adherence, and quality of life using standardized questionnaires.\u003c/p\u003e\n\u003ch3\u003eOutcome measures\u003c/h3\u003e\n\u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003ePrimary outcome: The proportion of patients achieving low disease activity at 12 months, defined as a Bath Ankylosing Spondylitis Disease Activity Index (BASDAI)\u0026thinsp;\u0026lt;\u0026thinsp;4 and an ASDAS\u0026thinsp;\u0026lt;\u0026thinsp;1.3.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eSecondary outcomes: Changes in CRP levels and cost-effectiveness analysis (CEA) comparing the two groups (\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e).\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003c/p\u003e \u003cp\u003eDisease activity was measured using ASDAS (\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e, \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e) and BASDAI (\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e) at baseline and at 12 months. The ASDAS was calculated using the following formula:\u003c/p\u003e \u003cp\u003eASDAS\u0026thinsp;=\u0026thinsp;0.12 \u0026times; Back pain\u0026thinsp;+\u0026thinsp;0.06 \u0026times; Duration of morning stiffness\u0026thinsp;+\u0026thinsp;0.11 \u0026times; Patient global\u0026thinsp;+\u0026thinsp;0.07 \u0026times; Peripheral pain/swelling\u0026thinsp;+\u0026thinsp;0.58 \u0026times; Ln (CRP\u0026thinsp;+\u0026thinsp;1)\u003c/p\u003e \u003cp\u003eASDAS values were categorized according to the following thresholds: scores below 1.3 indicated no disease activity, scores from 1.3 to 2.0 represented moderate activity, scores between 2.1 and 3.4 denoted high activity, and scores of 3.5 or above were classified as very high activity (\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eFor the CEA, direct medical costs (medication, hospital visits, laboratory tests) were collected from a healthcare provider perspective and converted to US dollars (USD). Effectiveness was defined as the proportion of patients achieving low disease activity (BASDAI\u0026thinsp;\u0026lt;\u0026thinsp;4 and ASDAS\u0026thinsp;\u0026lt;\u0026thinsp;1.3). The incremental cost-effectiveness ratio (ICER) was calculated as the difference in cost between groups divided by the difference in effectiveness, with cost-effectiveness interpreted according to thresholds recommended by the World Health Organization.\u003c/p\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003eStatistical analysis\u003c/h2\u003e \u003cp\u003eDescriptive statistics were used to summarize baseline characteristics. Continuous variables were reported as means\u0026thinsp;\u0026plusmn;\u0026thinsp;standard deviations or medians with interquartile ranges, as appropriate. Categorical variables were expressed as frequencies and percentages. Between-group comparisons were made using chi-square or Fisher\u0026rsquo;s exact tests for categorical variables and independent t tests or Mann\u0026ndash;Whitney U tests for continuous variables, based on data distribution.\u003c/p\u003e \u003cp\u003eA p-value\u0026thinsp;\u0026lt;\u0026thinsp;0.05 was considered statistically significant. All statistical analyses were performed using R software (version 4.4.2), employing packages such as dplyr for data manipulation and ggplot2 for graphical presentation. Where applicable, p-values were reported to three decimal places. No formal power calculation was conducted, as patient enrollment was determined by feasibility within the specified study timeframe.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eModel development and validation\u003c/h2\u003e \u003cp\u003eA random forest model was developed to predict achievement of ASDAS\u0026thinsp;\u0026lt;\u0026thinsp;1.3 at 12 months, leveraging its ability to account for nonlinear relationships and complex interactions among predictors. Variables directly reflecting the outcome (ASDAS at 12 months) were excluded to prevent overfitting. Categorical variables were one-hot encoded, and continuous variables were scaled when necessary. The dataset was split into training (80%) and test (20%) subsets using stratified sampling to maintain class balance.\u003c/p\u003e \u003cp\u003eModel performance was assessed using accuracy, precision, recall, and the area under the receiver operating characteristic curve (AUC). Feature importance was evaluated via the Gini impurity metric. All machine learning analyses were performed in Python v.3.9 with scikit-learn for model building, pandas and numpy for data handling, and matplotlib and seaborn for visualization.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eEthical considerations\u003c/h3\u003e\n\u003cp\u003e This study was approved by the Institutional Review Board of Bach Mai Hospital (160520/QĐ-BVBM). Written informed consent was obtained from all participants and the study was performed in accordance with the ethical principles of the Declaration of Helsinki.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003eIn this study, a total of 480 patients were screened for eligibility. The final analysis included 222 patients in the full dose group and 146 in the reduction dose group (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eTable\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e provides a comparison of key variables between group 1 (dose reduction) and group 2 (full dose). Significant differences in sex distribution were observed, with group 1 having a higher proportion of male patients, compared with group 2 (84.93% vs. 54.50%, p\u0026thinsp;=\u0026thinsp;0.0001). Income levels also differed significantly, with 86.99% of patients in group 1 in the high-income category, compared with 55.41% in group 2 (p\u0026thinsp;=\u0026thinsp;0.0001).\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eComparison of baseline characteristics between dose reduction groups (n\u0026thinsp;=\u0026thinsp;368)\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"6\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVariable\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCategory\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eGroup 1 (dose reduction) (n\u0026thinsp;=\u0026thinsp;146)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eGroup 2 (full dose) (n\u0026thinsp;=\u0026thinsp;222)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eOverall (n\u0026thinsp;=\u0026thinsp;368)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003ep-value\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eSex, n, (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMale\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e124 (84.93%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e121 (54.50%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e245 (66.58%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.0001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eFemale\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e22 (15.07%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e101 (45.50%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e123 (33.42%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.0001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eIncome, n (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eLow\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e19 (13.01%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e99 (44.59%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e118 (32.07%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.0001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eHigh\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e127 (86.99%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e123 (55.41%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e250 (67.93%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.0001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAge, (years) (mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e31.28\u0026thinsp;\u0026plusmn;\u0026thinsp;11.64\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e47.61\u0026thinsp;\u0026plusmn;\u0026thinsp;19.11\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e41.13\u0026thinsp;\u0026plusmn;\u0026thinsp;18.37\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.0001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBMI, (mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e20.86\u0026thinsp;\u0026plusmn;\u0026thinsp;2.59\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e20.87\u0026thinsp;\u0026plusmn;\u0026thinsp;2.71\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e20.87\u0026thinsp;\u0026plusmn;\u0026thinsp;2.65\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.9783\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAS duration, (years) (mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e7.08\u0026thinsp;\u0026plusmn;\u0026thinsp;6.14\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e9.47\u0026thinsp;\u0026plusmn;\u0026thinsp;5.98\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e8.52\u0026thinsp;\u0026plusmn;\u0026thinsp;6.15\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.0001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBaseline BASDAI, (mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e4.11\u0026thinsp;\u0026plusmn;\u0026thinsp;1.35\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e5.03\u0026thinsp;\u0026plusmn;\u0026thinsp;3.41\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e4.67\u0026thinsp;\u0026plusmn;\u0026thinsp;2.82\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.0558\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003eBaseline BASDAI, n (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;1.3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e4 (2.74%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e32 (14.41%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e36 (9.78%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.3-4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e42 (28.77%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e63 (28.38%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e105 (28.53%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.368\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u0026ge;\u0026thinsp;4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e100 (68.49%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e127 (57.21%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e227 (61.68%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.008\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBaseline ASDAS, (mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e2.15\u0026thinsp;\u0026plusmn;\u0026thinsp;1.09\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e2.06\u0026thinsp;\u0026plusmn;\u0026thinsp;1.16\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e2.10\u0026thinsp;\u0026plusmn;\u0026thinsp;1.13\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.6857\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"3\" rowspan=\"4\"\u003e \u003cp\u003eBaseline ASDAS, n (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.3\u0026ndash;2.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e22 (15.07%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e51 (22.97%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e73 (19.84%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.074\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2.1\u0026ndash;3.4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e59 (40.41%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e71 (31.98%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e130 (35.33%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.003\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;1.3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e42 (28.77%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e65 (29.28%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e107 (29.08%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.16\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u0026ge;\u0026thinsp;3.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e23 (15.75%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e35 (15.77%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e58 (15.76%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.107\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBaseline CRP level, (mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e5.92\u0026thinsp;\u0026plusmn;\u0026thinsp;3.91\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e9.59\u0026thinsp;\u0026plusmn;\u0026thinsp;6.24\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e8.14\u0026thinsp;\u0026plusmn;\u0026thinsp;5.72\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.0000\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"6\"\u003eSD, standard deviation; BMI, body mass index; CRP, C-reactive protein; BASDAI, Bath ankylosing spondylitis disease activity index; ASDAS ankylosing spondylitis disease activity score\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003ePatients in group 2 were older on average (47.61\u0026thinsp;\u0026plusmn;\u0026thinsp;19.11 years) than those in group 1 (31.28\u0026thinsp;\u0026plusmn;\u0026thinsp;11.64 years, p\u0026thinsp;=\u0026thinsp;0.0001). The duration of AS was longer in group 2 (9.47\u0026thinsp;\u0026plusmn;\u0026thinsp;5.98 years), compared with group 1 (7.08\u0026thinsp;\u0026plusmn;\u0026thinsp;6.14 years, p\u0026thinsp;=\u0026thinsp;0.0001).\u003c/p\u003e \u003cp\u003eBaseline BASDAI did not show significant differences between the groups (p\u0026thinsp;=\u0026thinsp;0.0558). However, group 1 had a higher proportion of patients with BASDAI\u0026thinsp;\u0026ge;\u0026thinsp;4, compared with group 2 (68.49% vs. 57.21%, p\u0026thinsp;=\u0026thinsp;0.008). Regarding the baseline ASDAS, there was a higher percentage of patients with an ASDAS of 1.3\u0026ndash;2.0 in group 1 than in group 2 (40.41% vs. 31.98%, p\u0026thinsp;=\u0026thinsp;0.003). Baseline CRP levels were significantly lower in group 1 (5.92\u0026thinsp;\u0026plusmn;\u0026thinsp;3.91) than in group 2 (9.59\u0026thinsp;\u0026plusmn;\u0026thinsp;6.24, p\u0026thinsp;=\u0026thinsp;0.0000). Body mass index was similar between the two groups (20.86\u0026thinsp;\u0026plusmn;\u0026thinsp;2.59 vs. 20.87\u0026thinsp;\u0026plusmn;\u0026thinsp;2.71, p\u0026thinsp;=\u0026thinsp;0.9783).\u003c/p\u003e \u003cp\u003eTable\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e shows the distribution of various DMARD therapies among patients in group 1 (dose reduction) and group 2 (full dose), together with overall values and statistical comparisons between the groups. For most combination therapies, including secukinumab\u0026thinsp;+\u0026thinsp;adalimumab and secukinumab\u0026thinsp;+\u0026thinsp;infliximab\u0026thinsp;+\u0026thinsp;golimumab, there were no significant differences between the groups (p\u0026thinsp;\u0026gt;\u0026thinsp;0.05).\u003c/p\u003e \u003cp\u003eHowever, notable differences were observed for certain therapies. Secukinumab was used significantly more frequently in group 1 (24.66%) than in group 2 (14.86%, p\u0026thinsp;=\u0026thinsp;0.027). Infliximab was administered significantly more frequently in group 1 (26.71%) compared with group 2 (9.01%, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001). Conversely, adalimumab was used more frequently in group 1 (14.38%) than in group 2 (6.76%, p\u0026thinsp;=\u0026thinsp;0.026). Furthermore, the secukinumab\u0026thinsp;+\u0026thinsp;golimumab and secukinumab\u0026thinsp;+\u0026thinsp;infliximab combinations were significantly more frequently administered in group 2 than in group 1 (p\u0026thinsp;=\u0026thinsp;0.016 and p\u0026thinsp;=\u0026thinsp;0.031, respectively).\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eCharacteristics of disease-modifying antirheumatic drug therapies used in the study\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"5\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTherapy\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eOverall (n\u0026thinsp;=\u0026thinsp;368)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eGroup 1 (dose reduction) (n\u0026thinsp;=\u0026thinsp;146)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eGroup 2 (full dose) (n\u0026thinsp;=\u0026thinsp;222)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003ep-value\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAdalimumab\u0026thinsp;+\u0026thinsp;infliximab\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e31 (8.42%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e13 (8.90%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e18 (8.11%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.939\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSecukinumab\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e69 (18.75%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e36 (24.66%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e33 (14.86%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.027\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSecukinumab\u0026thinsp;+\u0026thinsp;golimumab\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e21 (5.71%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e4 (2.74%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e17 (7.66%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.078\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGolimumab\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e34 (9.24%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e8 (5.48%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e26 (11.71%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.066\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eInfliximab\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e59 (16.03%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e39 (26.71%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e20 (9.01%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAdalimumab\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e36 (9.78%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e21 (14.38%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e15 (6.76%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.026\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSecukinumab\u0026thinsp;+\u0026thinsp;infliximab\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e27 (7.34%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e14 (9.59%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e13 (5.86%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.255\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSecukinumab\u0026thinsp;+\u0026thinsp;adalimumab\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e17 (4.62%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e4 (2.74%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e13 (5.86%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.255\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAdalimumab\u0026thinsp;+\u0026thinsp;golimumab\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e15 (4.08%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1 (0.68%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e14 (6.31%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.016\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eInfliximab\u0026thinsp;+\u0026thinsp;golimumab\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e13 (3.53%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e2 (1.37%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e11 (4.95%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.125\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEtanercept\u0026thinsp;+\u0026thinsp;adalimumab\u0026thinsp;+\u0026thinsp;infliximab\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e15 (4.08%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1 (0.68%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e14 (6.31%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.016\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSecukinumab\u0026thinsp;+\u0026thinsp;infliximab\u0026thinsp;+\u0026thinsp;golimumab\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e17 (4.62%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e2 (1.37%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e15 (6.76%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.031\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eTable\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e presents a comparison of the outcome variables between group 1 (dose reduction) and group 2 (full dose). Significant improvements were observed in several measures, which favored dose reduction. The reduction in CRP level from baseline to month 12 was significantly higher in group 1 (median \u0026minus;\u0026thinsp;4.65 [4.44]) than in group 2 (median \u0026minus;\u0026thinsp;1.32 [12.00], p\u0026thinsp;\u0026lt;\u0026thinsp;0.001), indicating a significant difference in inflammation reduction. Similarly, the decrease in disease activity scores, based on BASDAI and ASDAS assessments from baseline to month 12, showed significantly greater improvements in group 1 (median improvement in BASDAI: -3.00 [1.96]; median improvement in ASDAS: -1.72 [1.50]) compared with group 2 (median improvement in BASDAI: -0.42 [5.59]; median improvement in ASDAS: -0.15 [2.48], p\u0026thinsp;\u0026lt;\u0026thinsp;0.001 for both).\u003c/p\u003e \u003cp\u003eAt month 12, group 1 exhibited significantly lower CRP levels (median: 0.27 [1.19] mg/L) than group 2 did (7.15 [9.40] mg/L, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001), suggesting better inflammation control in the dose reduction group. Regarding the disease activity scores at month 12, 97.26% of the patients in group 1 had a BASDAI\u0026thinsp;\u0026lt;\u0026thinsp;4, compared with 50.90% in group 2 (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001). Based on the ASDAS, 93.15% of the patients in group 1 achieved scores of 2.1\u0026ndash;3.4, indicative of better disease control, whereas group 2 had a significantly higher proportion of patients with moderate, high, and very high disease activity (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001).\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eComparison of outcome variables between dose reduction and full dose groups\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"5\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVariable\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eOverall (n\u0026thinsp;=\u0026thinsp;368)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eGroup 1 (dose reduction) (n\u0026thinsp;=\u0026thinsp;146)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eGroup 2 (full dose) (n\u0026thinsp;=\u0026thinsp;222)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003ep-value\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCRP level improvement, median [IQR]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e-3.28 [8.03]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e-4.65 [4.44]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e-1.32 [12.00]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBASDAI improvement, median [IQR]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e-2.08 [4.50]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e-3.00 [1.96]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e-0.42 [5.59]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eASDAS improvement, median [IQR]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e-1.00 [2.33]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e-1.72 [1.50]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e-0.15 [2.48]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCRP level at month 12, median [IQR] (mg/L)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e2.77 [8.78]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.27 [1.19]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e7.15 [9.40]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBASDAI at month 12 (n (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e255 (69.29%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e142 (97.26%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e113 (50.90%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u0026ge;\u0026thinsp;4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e113 (30.71%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e4 (2.74%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e109 (49.10%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eASDAS at month 12 (n (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;1.3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e50 (13.59%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e9 (6.16%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e41 (18.47%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e1.3\u0026ndash;2.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e65 (17.66%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1 (0.68%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e64 (28.83%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2.1\u0026ndash;3.4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e217 (58.97%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e136 (93.15%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e81 (36.49%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u0026ge;\u0026thinsp;3.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e36 (9.78%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0 (0.00%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e36 (16.22%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eCRP, C-reactive protein; BASDAI, Bath ankylosing spondylitis disease activity index; ASDAS ankylosing spondylitis disease activity score; IQR, interquartile range\u003c/p\u003e \u003cp\u003eAs shown in Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e, the dose reduction demonstrated cost savings and improved effectiveness compared with the full dose across most criteria. For the improvement of the ASDAS, the ICER was -\u003cspan\u003e$\u003c/span\u003e5,870.20, with a 95% confidence interval (CI) of (-\u003cspan\u003e$\u003c/span\u003e7,982.43; -\u003cspan\u003e$\u003c/span\u003e11,194.39), indicating significant cost-effectiveness. Similarly, achieving the ASDAS target (\u0026lt;\u0026thinsp;1.3) showed an ICER of \u003cspan\u003e$\u003c/span\u003e16,921.46, with a narrow CI (95% CI: \u003cspan\u003e$\u003c/span\u003e7,982.43; \u003cspan\u003e$\u003c/span\u003e11,194.39). CRP improvement had an ICER of -\u003cspan\u003e$\u003c/span\u003e2,409.25, and the BASDAI improvement reported an ICER of -\u003cspan\u003e$\u003c/span\u003e4,085.17, with their respective 95% CIs suggesting robust cost savings for these outcomes.\u003c/p\u003e \u003cp\u003eConversely, achieving the BASDAI target (\u0026lt;\u0026thinsp;4.0) showed a higher ICER of \u003cspan\u003e$\u003c/span\u003e20,682.78, indicating that dose reduction was less cost-effective in this criterion. The results for incremental effectiveness showed 0.46 and 0.57 for the BASDAI and ASDAS targets, respectively, with their 95% CIs confirming the statistical significance in favor of dose reduction.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab4\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 4\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eIncremental cost effectiveness ratio results for dose reduction versus full dose (costs in \u003cspan\u003e$\u003c/span\u003e)\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"5\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEffectiveness criterion\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eICER (cost per unit of effectiveness)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eICER 95% CI\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eIncremental effectiveness\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eIncremental effectiveness 95% CI\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCRP level improvement\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e-2409.25\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e(7982.43; 11194.39)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e-3.98\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(-5.75; -2.21)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBASDAI improvement\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e-4085.17\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e(7982.43; 11194.39)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e-2.35\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(-3.16; -1.53)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eASDAS improvement\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e-5870.20\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e(7982.43; 11194.39)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e-1.63\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(-2.04; -1.23)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBASDAI target (\u0026lt;\u0026thinsp;4.0)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e20682.78\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e(7982.43; 11194.39)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.46\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(0.37; 0.56)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eASDAS target (\u0026lt;\u0026thinsp;1.3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e16921.46\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e(7982.43; 11194.39)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.57\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(0.46; 0.67)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eCRP, C-reactive protein; BASDAI, Bath ankylosing spondylitis disease activity index; ASDAS ankylosing spondylitis disease activity score; ICER, incremental cost effectiveness ratio; CI, confidence interval\u003c/p\u003e \u003cp\u003eICER represents the cost per unit increase in effectiveness for dose reduction compared with full dose; Incremental effectiveness denotes the difference in effectiveness between the dose reduction and full dose groups.\u003c/p\u003e \u003cp\u003eTable\u0026nbsp;\u003cspan refid=\"Tab5\" class=\"InternalRef\"\u003e5\u003c/span\u003e presents the analysis of the predictors for achieving an ASDAS\u0026thinsp;\u0026lt;\u0026thinsp;1.3 (inactive disease) in patients with AS undergoing DMARD treatment. Key predictors with significant impact include \u0026ldquo;reduce dose,\u0026rdquo; baseline CRP level, and baseline ASDAS, all with p-values below 0.001. The negative estimates for \u0026ldquo;reduce dose\u0026rdquo; and the baseline ASDAS suggest that these factors are associated with a higher probability of achieving inactive disease. Conversely, a positive association was observed with baseline CRP levels. In our study, a significant difference was found between male and female patients in the achievement of low disease activity (ASDAS\u0026thinsp;\u0026lt;\u0026thinsp;1.3); compared with female patients, male patients were more likely to reach this state of low disease activity. Other variables, including age, AS duration, income, and drug therapy, did not show significant predictive value, indicating a limited influence on achieving low disease activity.\u003c/p\u003e \u003cp\u003eA similar multivariate analysis was performed to explore the associations between various factors and achieving a BASDAI\u0026thinsp;\u0026lt;\u0026thinsp;4. However, no significant association was observed.\u003c/p\u003e \u003cp\u003eThe random forest models demonstrated robust performance in predicting ASDAS\u0026thinsp;\u0026lt;\u0026thinsp;1.3, as summarized in Table\u0026nbsp;\u003cspan refid=\"Tab5\" class=\"InternalRef\"\u003e5\u003c/span\u003e. The medium feature model (20 variables) achieved the best performance, with an AUC of 81.86%, precision of 86.49%, and accuracy of 77.03%, balancing predictive power and model complexity. The full model (50 variables) performed similarly (AUC, 81.21%; precision, 84.62%) but was more complex. The reduced model (10 variables) offered a streamlined approach with an AUC of 81.63% and precision of 76.74%. The key clinical variables model (five variables), focusing on the core clinical factors, delivered reasonable performance (AUC, 81.10%; precision 80.00%) and was ideal for resource-limited settings. These findings highlight the utility of the medium feature model as the optimal choice for balancing interpretability and accuracy, whereas the key clinical variables model offers simplicity and clinical relevance.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab5\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 5\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003ePerformance metrics of random forest models for predicting ankylosing spondylitis disease activity score\u0026thinsp;\u0026lt;\u0026thinsp;1.3\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"6\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eModel\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAccuracy\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eAUC\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003ePrecision\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eRecall\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eNumber of variables\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFull model (encoded)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e77.03%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e81.21%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e84.62%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e75.00%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e50\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eReduced model (encoded)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e71.62%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e81.63%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e76.74%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e75.00%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e10\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMedium feature model (top 20)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e77.03%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e81.86%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e86.49%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e72.73%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e20\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eKey clinical variables model\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e72.97%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e81.10%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e80.00%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e72.73%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eAUC, area under the receiver operating characteristic curve\u003c/p\u003e \u003cp\u003eThe feature importance analysis (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e) highlights the key predictors of achieving ASDAS\u0026thinsp;\u0026lt;\u0026thinsp;1.3 using the medium feature model. Total cost (USD), CRP at the month 12 (mg/L), and 12-month total cost (USD) are identified as the most influential variables, reflecting the importance of financial and inflammatory factors. Clinical measures, such as the BASDAI at month 12 and dose adjustment emphasize the importance of disease severity and treatment optimization. Early indicators, such as CRP level at month 1 and CRP at month 3 demonstrate the relevance of intermediate responses to treatment. These results underscore the value of integrating clinical, economic and longitudinal factors into personalized treatment strategies.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eThis study highlights the effectiveness of dose reduction strategies for bDMARDs for treating AS. Patients who underwent dose reduction achieved superior clinical outcomes compared with those on full dose therapy, with significant improvements in CRP levels (4.65 vs. 0.90 mg/L, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001), the BASDAI (3.00 vs. 0.42, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001), and ASDAS (1.72 vs. 0.15, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001). Interestingly, 93.15% of the dose reduction group achieved inactive disease (ASDAS\u0026thinsp;\u0026lt;\u0026thinsp;1.3), compared with 47.62% in the full dose group. Economic evaluation through ICER analysis demonstrated the cost-effectiveness of dose reduction, with a favorable ICER of -\u003cspan\u003e$\u003c/span\u003e16,772.62 to achieve ASDAS\u0026thinsp;\u0026lt;\u0026thinsp;1.3. Predictive modeling using random forest further identified key predictors of successful treatment outcomes, including the baseline CRP level, ASDAS, and early response to treatment.\u003c/p\u003e \u003cp\u003eOur findings align with earlier studies supporting the clinical feasibility of bDMARD dose reduction. Recent studies also support the clinical feasibility of bDMARD dose reduction in AS and other inflammatory arthritis conditions. Evidence shows that 50\u0026ndash;60% of patients with AS and patients with psoriatic arthritis can successfully maintain reduced TNFi doses for approximately a year (\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e, \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e). Disease activity-guided tapering and fixed dose reduction appear to be effective strategies, whereas complete discontinuation is not recommended due to high flare rates (\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e, \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e).\u003c/p\u003e \u003cp\u003ePatients express concerns about dose reduction but are generally willing to try if given a clear rationale and participate in collaborative decision-making (\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e). Successful tapering is associated with lower disease activity prior to dose reduction (\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e). Cost-effectiveness analyses of dose adjustment strategies for axial spondyloarthritis treatments have shown promising results. Tapering etanercept by 25% every 3 months is a pragmatic approach for more cost-effective use (\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e). Secukinumab is cost-effective in both biologic-na\u0026iuml;ve and biologic-experienced patients (\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e). Adalimumab is considered cost-effective, compared with conventional therapy, over a 30-year period (\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e). Infliximab administered every 6 weeks is cost-effective, compared with on-demand regimens (\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e). Tocilizumab addition to standard care is estimated to be cost-effective in both methotrexate-tolerant and contraindicated patients (\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e). Disease activity-guided dose optimization of TNFis results in significant cost savings without a relevant loss of quality of life (\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e). Higher treatment persistence with subcutaneous TNFis is cost-effective from both the payer and societal perspectives (\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e). Therapeutic drug monitoring-guided adalimumab therapy has increased quality-adjusted life years at lower costs (\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e). Machine learning (ML) techniques have also shown promise in improving the diagnosis and treatment of AS. ML models can predict early AS diagnosis with high accuracy using administrative claims and electronic medical record data (\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e, \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e). These models can also identify patients likely to require early TNFi treatment (\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e, \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e). Additionally, ML and deep learning approaches facilitate early diagnosis through patient characteristic profiling and biomarker identification (\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe integration of ICER analysis and ML provides a robust framework for implementing dose reduction strategies in clinical practice. The economic benefits of dose reduction, including significant cost savings, can enhance patient adherence, particularly in resource-constrained settings.\u003c/p\u003e \u003cp\u003eRandom forest models enable clinicians to identify patients who are more likely to benefit from dose reduction based on clinical and biomarker predictors, such as the baseline CRP level and early improvement in ASDAS. This personalized approach supports the more efficient use of bDMARDs, balancing clinical efficacy and financial sustainability.\u003c/p\u003e \u003cp\u003eThe strengths of this study include its prospective design, comprehensive evaluation of clinical and economic outcomes, and application of ML for predictive modeling. However, the study also has limitations, including the nonrandomized allocation of patients to the dose reduction and full dose groups, which may have introduced a selection bias. Additionally, the relatively short follow-up period limits insights into the long-term sustainability and safety of dose reduction. Furthermore, the ICER results may vary across healthcare systems, reducing generalizability.\u003c/p\u003e \u003cp\u003eFuture studies should validate these findings in larger multicenter trials with diverse populations and longer follow-up periods. Investigating additional predictors, including patient-reported outcomes and quality-of-life measures, can refine candidate selection for dose reduction. Further exploration of real-world ICER results across different healthcare systems will strengthen the evidence for the economic feasibility of dose adjustment strategies.\u003c/p\u003e \u003cp\u003eOverall, this study supports the feasibility of dose-reduction of bDMARDs in patients with AS, demonstrating that it can maintain clinical efficacy and significantly reduce treatment costs. These findings advocate for future studies supporting dose reduction strategies in clinical practice, with careful patient selection and monitoring to ensure optimal results. Such approaches will contribute to more sustainable healthcare by balancing effective disease management with economic considerations.\u003c/p\u003e"},{"header":"Conclusions","content":"\u003cp\u003eThis study provides compelling evidence that the reduction in the dose of bDMARDs for treating ankylosing spondylitis is both clinically effective and economically advantageous. Patients with dose reduction achieved better disease control rates (ASDAS\u0026thinsp;\u0026lt;\u0026thinsp;1.3) while significantly reducing treatment costs, as demonstrated by a favorable ICER. The integration of random forest modeling further enhanced the study by identifying key predictors of treatment success, enabling a personalized approach to dose adjustment. These findings not only reinforce the feasibility of dose reduction in maintaining long-term disease remission but also highlight its potential to alleviate the financial burden associated with bDMARD therapy. Combining clinical, economic, and predictive insights, this study establishes a strong foundation for optimizing treatment strategies in AS, advocating for their adoption in routine practice with customized patient management.\u003c/p\u003e"},{"header":"Abbreviations","content":" \u003cp\u003eAS, ankylosing spondylitis\u003c/p\u003e \u003cp\u003eASDAS, ankylosing spondylitis disease activity score\u003c/p\u003e \u003cp\u003eBASDAI, Bath ankylosing spondylitis disease activity index\u003c/p\u003e \u003cp\u003ebDMARDs, biologic disease-modifying antirheumatic drugs\u003c/p\u003e \u003cp\u003eICER, incremental cost effectiveness ratios\u003c/p\u003e \u003cp\u003eNSAIDs, nonsteroidal anti-inflammatory drugs\u003c/p\u003e \u003cp\u003eTNFi, tumor necrosis factor inhibitors\u003c/p\u003e "},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgments\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe would like to thank all the physicians and nurses at the Rheumatology Center and Laboratory Department of Bach Mai Hospital, Vietnam, for their help with patient recruitment and clinical laboratory testing.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eEthics approval and consent to participate\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study was reviewed and approved by the Institutional Review Board of Bach Mai Hospital, approval number of 160520/QĐ-BVBM. All methods were carried out in accordance with relevant guidelines and regulations. Written informed consent was obtained from all individual participants included in the study.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eConsent for publication\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eAvailability of data and materials\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll data generated or analyzed during this study are included in this published article and its supplementary information files. Any additional data or materials can be obtained from the corresponding author upon reasonable request.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eCompeting interests\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that they have no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eFunding\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNone of the authors have any financial or non-financial competing interests to disclose.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eAuthors\u0026apos; contributions\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eBHB and HVD conceived and designed the study.\u003c/p\u003e\n\u003cp\u003eNTTP, VTTH, NTTN, and NTNH contributed to data acquisition, analysis, and interpretation.\u003c/p\u003e\n\u003cp\u003eHVD supervised the project and served as the corresponding author.\u003c/p\u003e\n\u003cp\u003eBHB, NTTP, and VTTH drafted the initial version of the manuscript.\u003c/p\u003e\n\u003cp\u003eNTTN, NTNH, and HVD critically revised the manuscript for important intellectual content.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eAcknowledgements\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors would like to express their sincere gratitude to the clinical and administrative staff at Bach Mai Hospital for their invaluable support and collaboration throughout the study period. Special thanks go to all the patients who participated in this research for their time and cooperation.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eMcVeigh CM, Cairns AP. Diagnosis and management of ankylosing spondylitis. BMJ (Clinical Res ed). 2006;333(7568):581\u0026ndash;5.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDean LE, Jones GT, MacDonald AG, Downham C, Sturrock RD, Macfarlane GJ. Global prevalence of ankylosing spondylitis. Rheumatology (Oxford). 2014;53(4):650\u0026ndash;7.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBrewerton DA, Hart FD, Nicholls A, Caffrey M, James DC, Sturrock RD. Ankylosing spondylitis and HL-A 27. Lancet (London England). 1973;1(7809):904\u0026ndash;7.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWordsworth BP, Cohen CJ, Davidson C, Vecellio M. Perspectives on the Genetic Associations of Ankylosing Spondylitis. Front Immunol. 2021;12:603726.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBoonen A, van der Linden SM. The burden of ankylosing spondylitis. J Rheumatol Suppl. 2006;78:4\u0026ndash;11.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHegyi R, Nagy B, Koncz A, Huybrechts I, Lavicky J, Ferenczik A. Burden Of Disease Analysis Of Ankylosing Spondylitis In Hungary. Value Health. 2014;17(7):A376\u0026ndash;7.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKobelt G, Sobocki P, Mulero J, Gratacos J, Pocovi A, Collantes-Estevez E. The burden of ankylosing spondylitis in Spain. Value Health. 2008;11(3):408\u0026ndash;15.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWebers C, Ortolan A, Sepriano A, Falzon L, Baraliakos X, Landewe RBM, et al. Efficacy and safety of biological DMARDs: a systematic literature review informing the 2022 update of the ASAS-EULAR recommendations for the management of axial spondyloarthritis. Ann Rheum Dis. 2023;82(1):130\u0026ndash;41.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKlavdianou K, Tsiami S, Baraliakos X. New developments in ankylosing spondylitis\u0026mdash;status in 2021. Rheumatology. 2021;60(Supplement6):vi29\u0026ndash;37.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOo WM, Yu SP, Daniel MS, Hunter DJ. Disease-modifying drugs in osteoarthritis: current understanding and future therapeutics. Expert Opin Emerg Drugs. 2018;23(4):331\u0026ndash;47.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGregori D, Giacovelli G, Minto C, Barbetta B, Gualtieri F, Azzolina D, et al. Association of Pharmacological Treatments With Long-term Pain Control in Patients With Knee Osteoarthritis: A Systematic Review and Meta-analysis. JAMA. 2018;320(24):2564\u0026ndash;79.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFranke LC, Ament AJ, van de Laar MA, Boonen A, Severens JL. Cost-of-illness of rheumatoid arthritis and ankylosing spondylitis. Clin Exp Rheumatol. 2009;27(4 Suppl 55):S118\u0026ndash;23.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGreenberg JD, Palmer JB, Li Y, Herrera V, Tsang Y, Liao M. Healthcare Resource Use and Direct Costs in Patients with Ankylosing Spondylitis and Psoriatic Arthritis in a Large US Cohort. J Rheumatol. 2016;43(1):88\u0026ndash;96.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003evan der Linden S, Valkenburg HA, Cats A. Evaluation of diagnostic criteria for ankylosing spondylitis. A proposal for modification of the New York criteria. Arthritis Rheum. 1984;27(4):361\u0026ndash;8.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGold MRSJ, Russell LB, Weinstein MC, editors. Cost-Effectiveness in Health and Medicine. ed. n, editor. Oxford University Press; 2016.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGoie The HS, Steven MM, van der Linden SM, Cats A. Evaluation of diagnostic criteria for ankylosing spondylitis: a comparison of the Rome, New York and modified New York criteria in patients with a positive clinical history screening test for ankylosing spondylitis. Br J Rheumatol. 1985;24(3):242\u0026ndash;9.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003evan der Heijde D, Lie E, Kvien TK, Sieper J, Van den Bosch F, Listing J, et al. ASDAS, a highly discriminatory ASAS-endorsed disease activity score in patients with ankylosing spondylitis. Ann Rheum Dis. 2009;68(12):1811\u0026ndash;8.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGarrett S, Jenkinson T, Kennedy LG, Whitelock H, Gaisford P, Calin A. A new approach to defining disease status in ankylosing spondylitis: the Bath Ankylosing Spondylitis Disease Activity Index. J Rheumatol. 1994;21(12):2286\u0026ndash;91.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLukas C, Landewe R, Sieper J, Dougados M, Davis J, Braun J, et al. Development of an ASAS-endorsed disease activity score (ASDAS) in patients with ankylosing spondylitis. Ann Rheum Dis. 2009;68(1):18\u0026ndash;24.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFong W, Holroyd C, Davidson B, Armstrong R, Harvey N, Dennison E, et al. The effectiveness of a real life dose reduction strategy for tumour ne crosis factor inhibitors in ankylosing spondylitis and psoriatic arthr itis. Rheumatology. 2016;55(10):1837\u0026ndash;42.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003evdV SAE. F BK, P H, R B, H B, Patient-tailored dose reduction of TNF-α blocking agents in ankylosing spondylitis patients with stable low disease activity in daily clinic al practice. Clin Exp Rheumatol. 2015.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVerhoef LM, Tweehuysen L, Hulscher ME, Fautrel B, den Broeder AA. bDMARD Dose Reduction in Rheumatoid Arthritis: A Narrative Review with Systematic Literature Search. Rheumatol Ther. 2017;4(1):1\u0026ndash;24.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChaudhary H, Bittar M, Daoud A, Magrey M. Dose Tapering and Discontinuation of Biologic DMARDs in Axial Spondylo arthritis: A Narrative Review (2023 SPARTAN Annual Meeting Proceedings). Curr Rheumatol Rep. 2024;26(5):155\u0026thinsp;\u0026ndash;\u0026thinsp;63.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHewlett S, Haig-Ferguson A, Rose‐Parfitt E, Halls S, Freke S, Creamer P. Dose reduction of biologic therapy in inflammatory arthritis: A qualit ative study of patients' perceptions and needs. Musculoskelet Care. 2018;17(1):63\u0026ndash;71.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJun FL, Yu Z, Dongying W, Hanshi C. X, L L. Efficiency of dose reduction strategy of etanercept in patients with a xial spondyloarthritis. Clinical and Experimental Rheumatology; 2018.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEmery P, Van Keep M, Beard S, Graham C, Miles L, Jugl SM, et al. Cost Effectiveness of Secukinumab for the Treatment of Active Ankylosi ng Spondylitis in the UK. PharmacoEconomics. 2018;36(8):1015\u0026ndash;27.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBotteman MF, Hay JW, Luo MP, Curry AS, Wong RL, van Hout BA. Cost effectiveness of adalimumab for the treatment of ankylosing spond ylitis in the United Kingdom. Rheumatology. 2007;46(8):1320\u0026ndash;8.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFautrel B, Benhamou M, Breban M, Roy C, Lenoir C, Trape G, et al. Cost effectiveness of two therapeutic regimens of infliximab in ankylo sing spondylitis: economic evaluation within a randomised controlled t rial. Ann Rheum Dis. 2010;69(2):424\u0026ndash;7.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDiamantopoulos A, Finckh A, Huizinga T, Sungher DK, Sawyer L, Neto D, et al. Tocilizumab in the Treatment of Rheumatoid Arthritis: A Cost-Effective ness Analysis in the UK. PharmacoEconomics. 2014;32(8):775\u0026ndash;87.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKievit W, van Herwaarden N, van den Hoogen FHJ, van Vollenhoven RF, Bijlsma JWJ, van den Bemt BJF, et al. Disease activity-guided dose optimisation of adalimumab and etanercept is a cost-effective strategy compared with non-tapering tight control rheumatoid arthritis care: analyses of the DRESS study. Ann Rheum Dis. 2016;75(11):1939\u0026ndash;44.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIverg\u0026aring;rd M, Dal\u0026eacute;n J, Svedbom A, Black CM, Borse RH, Kachroo S. FRI0636 The value of persistence in treatment with subcutaneous tnf-al pha inhibitors for ankylosing spondylitis. Ann Rheum Dis. 2018;77:840\u0026ndash;1.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eG\u0026oacute;mez-Arango C, Gorostiza I, \u0026Uacute;car E, Garc\u0026iacute;a-Vivar ML, P\u0026eacute;rez CE, De Dios JR, et al. Cost-Effectiveness of Therapeutic Drug Monitoring-Guided Adalimumab Th erapy in Rheumatic Diseases: A Prospective, Pragmatic Trial. Rheumatol Ther. 2021;8(3):1323\u0026ndash;39.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWalsh JA, Rozycki M, Yi E, Park Y. Application of machine learning in the diagnosis of axial spondyloarth ritis. Curr Opin Rheumatol. 2019;31(4):362\u0026ndash;7.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKennedy J, Kennedy N, Cooksey R, Choy E, Siebert S, Rahman M, et al. Predicting a diagnosis of ankylosing spondylitis using primary care he alth records\u0026ndash;A machine learning approach. PLoS ONE. 2023;18(3):e0279076.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLee S, Eun Y, Kim H, Cha H-S, Koh E-M, Lee J. Machine learning to predict early TNF inhibitor users in patients with ankylosing spondylitis. Sci Rep. 2020;10(1).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDeodhar A, Rozycki M, Garges C, Shukla O, Arndt T, Grabowsky T, et al. Use of machine learning techniques in the development and refinement o f a predictive model for early diagnosis of ankylosing spondylitis. Clin Rheumatol. 2019;39(4):975\u0026ndash;82.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDhall S, Vaish A, Vaishya R. Machine learning and deep learning for the diagnosis and treatment of ankylosing spondylitis- a scoping review. J Clin Orthop Trauma. 2024;52:102421.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"bmc-rheumatology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"brhm","sideBox":"Learn more about [BMC Rheumatology](http://bmcrheumatol.biomedcentral.com)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/brhm/default.aspx","title":"BMC Rheumatology","twitterHandle":"@BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"ankylosing spondylitis, biological disease-modifying antirheumatic drugs, dose reduction, cost-effectiveness, disease activity","lastPublishedDoi":"10.21203/rs.3.rs-5917710/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5917710/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground\u003c/h2\u003e \u003cp\u003eAnkylosing spondylitis (AS) is a chronic inflammatory disease that significantly affects quality of life and imposes a high economic burden on patients due to the cost of biologic disease-modifying antirheumatic drugs (bDMARDs). Dose reduction strategies for bDMARDs may offer a feasible approach to maintaining clinical efficacy while reducing costs. This study aimed to evaluate the clinical effectiveness and cost-efficiency of bDMARD dose reduction in patients with AS and apply predictive modeling to identify key factors influencing disease control.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e \u003cp\u003eThis 12-month prospective study included 368 patients with AS who were divided into two groups: those who received dose reduction and those with full-dose therapy. Clinical outcomes such as C-reactive protein (CRP) levels, the Bath ankylosing spondylitis disease activity index (BASDAI) and ankylosing spondylitis disease activity score (ASDAS) were assessed, along with cost effectiveness using incremental cost effectiveness ratios (ICER). Random forest models were developed to predict the achievement of inactive disease (ASDAS\u0026thinsp;\u0026lt;\u0026thinsp;1.3) and to identify key predictors.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e \u003cp\u003eThe ICER to achieve an ASDAS\u0026thinsp;\u0026lt;\u0026thinsp;1.3 was highly favorable (-\u003cspan\u003e$\u003c/span\u003e16,772.62). Patients in the dose reduction group demonstrated significant improvements in CRP levels (-4.65 vs. -1.32 mg/L, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001), BASDAI (-3.00 vs. -0.42, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001), and ASDAS (-1.72 vs. -0.15, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001), compared with the full dose group. Predictive modeling identified baseline CRP level, baseline ASDAS, and dose adjustment as key factors influencing outcomes, with the medium feature model achieving an area under the receiver operating characteristic curve of 81.86%.\u003c/p\u003e\u003ch2\u003eConclusions\u003c/h2\u003e \u003cp\u003eThe reduction in bDMARD doses maintained clinical efficacy and achieved significant cost savings, offering a viable strategy for the management of AS. Predictive modeling provided actionable insights to optimize personalized treatment strategies, balancing efficacy and economic sustainability. These findings support the integration of dose reduction strategies into routine practice, particularly in resource-limited settings.\u003c/p\u003e","manuscriptTitle":"Personalized dose reduction strategies for biologic disease-modifying antirheumatic drugs for treating ankylosing spondylitis: a clinical and economic evaluation with predictive modeling","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-01-30 10:19:58","doi":"10.21203/rs.3.rs-5917710/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-02-03T21:46:37+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-01-28T11:27:09+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-01-28T11:25:13+00:00","index":"","fulltext":""},{"type":"submitted","content":"BMC Rheumatology","date":"2025-01-28T09:51:57+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"bmc-rheumatology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"brhm","sideBox":"Learn more about [BMC Rheumatology](http://bmcrheumatol.biomedcentral.com)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/brhm/default.aspx","title":"BMC Rheumatology","twitterHandle":"@BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"2240b7bb-92ad-4cff-a0b1-bf8f9c0b4c27","owner":[],"postedDate":"January 30th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2025-06-02T15:59:52+00:00","versionOfRecord":{"articleIdentity":"rs-5917710","link":"https://doi.org/10.1186/s41927-025-00516-9","journal":{"identity":"bmc-rheumatology","isVorOnly":false,"title":"BMC Rheumatology"},"publishedOn":"2025-05-26 15:57:10","publishedOnDateReadable":"May 26th, 2025"},"versionCreatedAt":"2025-01-30 10:19:58","video":"","vorDoi":"10.1186/s41927-025-00516-9","vorDoiUrl":"https://doi.org/10.1186/s41927-025-00516-9","workflowStages":[]},"version":"v1","identity":"rs-5917710","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5917710","identity":"rs-5917710","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}