Factors influencing the therapeutic response of methotrexate plus leflunomide combination in Colombian patients with rheumatoid arthritis: Real-life evidence

Abstract

Objective

To assess the therapeutic response, factors influencing this outcome, and adverse reactions of methotrexate and leflunomide combination therapy in patients with rheumatoid arthritis (RA), previously refractory to either methotrexate or leflunomide monotherapy.

Methods

This multicenter historical cohort study included patients with active RA despite previous monotherapy treatment with methotrexate or leflunomide between 2010 and 2021. The variables were recorded from electronic medical records at the beginning of combination therapy and at 3 and 6 months of follow-up. Disease activity and treatment response were assessed using the Disease Activity Score 28 with C-reactive protein (DAS28-CRP) and the European Alliance of Associations for Rheumatology (EULAR) response criteria. A logistic regression model was performed to identify the factors associated with clinical response.

Results

The study included 90 patients, 75% were female, and the mean age was 55 years. Adverse reactions occurred in 24% of patients, the most frequent being elevation of aminotransferase levels (10%). The median DAS28-CRP was 4.0, 2.6, and 2.1 at baseline, three-month, and six-month follow-up, respectively. According to the EULAR response criteria, 83% of patients achieved a moderate or good response at 6 months. The erosive disease was identified as a risk factor for non-response according to EULAR criteria at 3 months of follow-up.

Conclusion

In RA refractory to methotrexate or leflunomide, the combination therapy achieves an adequate clinical response with an acceptable safety profile. Patients with erosive disease represent a subgroup with a lower response to the combination therapy, and therefore could benefit from another therapeutic option.

Resumen

Objetivo

Evaluar la respuesta terapéutica, los factores que influyen en este desenlace y las reacciones adversas de la combinación terapéutica de metotrexato y leflunomida en pacientes con artritis reumatoide (AR) previamente refractarios a la monoterapia con metotrexato o con leflunomida.

Métodos

Estudio de cohorte histórica multicéntrico que incluyó pacientes con AR activa a pesar del tratamiento previo en monoterapia con metotrexato o leflunomida entre 2010 y 2021. Las variables se recolectaron a partir de la historia clínica al inicio de la combinación terapéutica y a los tres y seis meses de seguimiento. La actividad de la enfermedad se evaluó con el DAS28-CRP y los criterios de respuesta de la EULAR. Se elaboró un modelo de regresión logística para identificar los factores asociados con la respuesta clínica.

Resultados

El estudio incluyó 90 pacientes; el 75% eran de género femenino y tenían una edad promedio de 55 años. Los eventos adversos ocurrieron en el 24% de los pacientes, con la elevación de transaminasas como la más frecuente (10%). La mediana de DAS28-CRP fue de 4,0, 2,6 y 2,1 puntos en el momento inicial y en el seguimiento a los tres y seis meses, respectivamente. Según los criterios de respuesta de la EULAR, el 83% de los pacientes presentaron una respuesta moderada o buena a los seis meses. La enfermedad erosiva se identificó como un factor de riesgo para no lograr respuesta según los criterios EULAR a los tres meses de seguimiento.

Conclusiones

En pacientes con AR refractaria al metotrexato o a la leflunomida, la combinación terapéutica logra una respuesta clínica adecuada con un perfil de seguridad aceptable. Los pacientes con enfermedad erosiva representan un subgrupo con menor respuesta a la combinación terapéutica y, por lo tanto, podrían beneficiarse de otra opción terapéutica.

Introduction

Methotrexate (MTX) is recommended as the first-line therapy for patients with rheumatoid arthritis (RA) according to international treatment guidelines [1–5]. However, between 20% and 60% of patients do not have an adequate response [6–9], and up to 25% withdrew MTX due to adverse events [9,10], requiring the addition of biological disease-modifying antirheumatic drugs (bDMARDs) or targeted synthetic DMARDs (tsDMARDs). However, both bDMARDs and tsDMARDs treatments can increase the risk of infections and costs to the healthcare system [11,12].

Treatment with leflunomide (LEF) as monotherapy has been shown to improve disease activity and slow the progression of structural damage [13,14], making it an alternative for patients who do not reach the therapeutic goal or who experience adverse reactions to MTX. Furthermore, some studies have reported that combination therapy of MTX + LEF is another therapeutic option before starting bDMARDs with favorable clinical responses [7,8,10,15–17]. However, there is concern about the risk of potentiating serious adverse effects, especially hepatotoxicity or myelotoxicity.

In Latin America, combination therapy for RA has been reported to result in a significant decrease in the Disease Activity Score 28 (DAS28), with 76% of patients achieving the American College of Rheumatology 20% improvement criteria (ACR20) [17]. Furthermore, in a Brazilian study, MTX + LEF had a safety profile comparable to MTX or LEF monotherapy [18].

Bluett et al. [19] identified rheumatoid factor (RF) positivity, younger age at onset, and higher baseline DAS-28 as factors associated with MTX treatment failure. However, there is limited data on the risk factors that influence the clinical response to MTX + LEF and its benefit in high-risk populations, such as patients with high titers of RF or anti-citrullinated peptide antibodies (ACPA), elevated acute-phase reactants, and erosive disease.

Our study aimed to determine the adverse reactions, therapeutic responses, and factors influencing outcomes in a Colombian cohort of RA patients treated with MTX + LEF using real-world data from routine healthcare delivery.

Materials and methods

Study design

We conducted a historical cohort study at the Hospital San Vicente Fundación (HSVF) in Medellín, Colombia, and at Riesgo de Fractura CAYRE IPS in Bogotá, Colombia. The medical records registered by rheumatologists were identified in the database of both institutions using the International Classification of Diseases (ICD-10) codes for RA (ICD-10: M053, M058, M059, M060, M068, and M069). Subsequently, three of the authors selected patients with MTX + LEF during the period from January 2010 to December 2021.

Patients

The inclusion criteria were as follows: age ≥18 years; diagnosis of RA according to the 2010 ACR/European Alliance of Associations for Rheumatology (EULAR) classification criteria [20]; failure to MTX (orally or subcutaneous administration) or LEF monotherapy; MTX + LEF combination therapy for at least six months after failure of MTX or LEF monotherapy; follow-up visits at the outpatient clinic at three and six months after initiation of combination therapy, the baseline for the study. Failure to monotherapy was defined as the absence of a decrease in DAS28 with C-reactive protein (DAS28-CRP) of at least 0.6 points in patients with moderate or high disease activity, or persistence of DAS28-CRP above 5.1 points despite a dose of MTX between 15 and 25 mg weekly or LEF 20 mg daily for at least three months.

The exclusion criteria were as follows: concomitant treatment with antimalarial, sulfasalazine, bDMARDs, tsDMARDs, or prednisolone at a dose >10 mg daily or equivalent; pregnancy; overlapping with other rheumatic autoimmune diseases; history of cancer or infections diagnosed before the onset of the combination therapy that could interfere with the outcome of the study. Modifications in MTX dose or route of administration at the three-month follow-up due to lack of response or adverse reactions were not exclusion criteria, but the addition of other treatments was. All patients received folic acid supplementation as part of standard care, with a minimum weekly dose of 7 mg.

Variables

The variables collected were the following: age, sex, time from RA diagnosis to the combination therapy onset, RF titers by the latex turbidimetry technique, ACPA by the micro Enzyme-Linked Immunosorbent Assay method, the presence of erosions (any erosion on X-rays of hands or feet recorded by the treating rheumatologist), previous use of MTX or LEF as monotherapy and their doses, route of MTX administration (oral or subcutaneous), adverse reactions to the combination therapy (gastrointestinal intolerance, oral ulcers or mucositis, aminotransferase elevation, alopecia, leukopenia, nodulosis, allergic reaction, and infections), and the need to discontinue the combination therapy (one or both medications) due to an adverse reaction. Blood cell count, serum creatinine, aspartate aminotransferase (AST), alanine aminotransferase (ALT), C-reactive protein (CRP), and erythrocyte sedimentation rate (ESR) at the onset of combination treatment and at follow-up visits were also recorded. The outcomes evaluated were the response to combination therapy, as assessed by DAS28-CRP and its components, and the EULAR response criteria [21]. Participants with missing outcome information at follow-up visits were not excluded; their remaining data were used in the study.

Sample

A sample size of 70 patients was calculated using the formula for the difference in proportions, with α = 0.05 and power = 80%, based on the described proportions of patients with therapeutic failure to MTX in RA of 54.9% [22] and to combination therapy of 24% [17]. Given the possibility of missing outcome data due to incomplete registration in the electronic medical records, we collected an additional 20% of the calculated sample as a contingency.

Data analysis

Categorical variables were described in absolute and relative frequencies. The quantitative variables were described using measures of central tendency and dispersion: the mean and standard deviation (SD) for normally distributed variables, and the median and interquartile range (IQR) for non-normally distributed variables.

The Wilcoxon signed-rank test was used to evaluate change over time for quantitative variables, since they did not have normal distributions, while the difference in proportions with paired data was used for qualitative variables. We developed a logistic regression model to identify risk factors for non-response by EULAR criteria adjusted for age, sex, RF titers, ACPA titers, and erosive disease. In addition, a multiple linear regression was performed to identify risk factors associated with the mean decrease in DAS28-CRP at 3 and 6 months, adjusted for the same variables. The statistical analysis was performed with Stata version 16. A p-value <0.05 was considered significant.

Ethical considerations

This research study was conducted retrospectively from data obtained for clinical purposes, without details that might disclose the identity, and it did not require informed consent. Approval was obtained from the Institutional Review Board and ethics committee of the HSVF and Riesgo de Fractura CAYRE IPS in Bogotá, Colombia. Additionally, this study was performed in line with the principles of the Declaration of Helsinki 2013.

Results

From 2291 medical records reviewed, 90 patients (3.9%) were included in this study, 31 (34%) from the HSVF and 59 (66%) from Riesgo de Fractura CAYRE IPS. The main exclusion criteria were concomitant use of bDMARDs and comorbidities. Table 1 describes the baseline characteristics of the patients, the treatment characteristics, and the adverse reactions during the combination therapy. Adverse reactions to MTX + LEF therapy occurred in approximately one-quarter of patients (Table 1). No allergic reactions, cytopenias, or infections occurred with this combination therapy.

Table 1.

Baseline demographics, treatment, and adverse reactions.

Characteristics	n = 90
Sociodemographic characteristics
Age (years), mean (±SD)	55 (13)
Female, n (%)	75 (83)
Ethnic group
Mestizo, n (%)	88 (97.8)
Afro-Latin American, n (%)	2 (2.2)
Features of the disease
Time since diagnosis (months), median (IQR)	8.5 (4–13)
Positivity for RF, n (%)*	81 (90)
RF titers, median (IQR)	83.5 (36.7–220)
Positivity for ACPA, n (%)^a	72 (80)
ACPA titers, median (IQR)	250 (116–500)
Erosive disease, n (%)	28 (31)
Characteristics of the treatment
Previous use of methotrexate monotherapy, n (%)	49 (54)
Time on methotrexate monotherapy, median in months (IQR)	14 (6–45)
Weekly dose of methotrexate (mg/week), median (IQR)	15 (15–20)
Methotrexate route of administration
Oral, n (%)	63 (70)
Subcutaneous, n (%)	27 (30)
Time with combination therapy MTX + LEF (months), median (IQR)	15 (9–24)
Adverse reaction to combination therapy, n (%)	22 (24)
Elevated aminotransferases, n (%)	9 (10)
>3 times the URL, n (%)	4 (4.4)
>5 times the URL, n (%)	2 (2.2)
>10 times the URL, n (%)	0
Therapy withdrawal because of increased transaminases, n (%)	3 (3.3)
Gastrointestinal intolerance, n (%)	5 (5.6)
Alopecia, n (%)	3 (3.3)
Oral ulcers or mucositis, n (%)	2 (2.2)
Leucopenia, n (%)	2 (2.2)
Nodulosis, n (%)	1 (1.1)
Therapy withdrawal due to any adverse reaction	16 (17.7)

ACPA: anti-citrullinated peptide antibodies; DAS28-CRP: Disease Activity Score 28 with C-reactive protein; IQR: inter-quartile range; RF: rheumatoid factor; SD: standard deviation; URL: upper reference limit.

a

From a total of 83 patients with ACPA measurement.

*

From a total of 88 patients with RF measurement.

Table 2 and Table S1 (supplementary material) summarize the laboratory variables and DAS28-CRP at baseline and the follow-up visits. There was a statistically significant difference in the median values of leukocytes, neutrophils, lymphocytes, and platelets between baseline and follow-up. Nevertheless, these differences were not clinically meaningful, and none of them presented a value of cytopenia (Table S1). There were missing data for calculating the DAS28-CRP score and the EULAR response criteria in 16% and 19% at three months of follow-up, and in 3% and 11% at six months of follow-up, respectively. The DAS28-CRP categories of moderate and high activity progressively decreased during follow-up. In contrast, the remission category and low disease activity showed a progressive increase (Fig. 1 and Table S1). We found a moderate or good response according to EULAR criteria in 73% and 83% of cases at 3 months and 6 months, respectively (Fig. 2 and Table S1).

Table 2.

Baseline and follow-up laboratory characteristics.

Characteristics	Baseline	At 3 months	At 6 months
Laboratories, median (IQR)^a
CRP (mg/L)	4.2 (1.3–9.9)	2.3 (1.1–9.5)	2.5 (1.1–5.9)
ESR (mm/h)	16 (7–27)	15 (7–25)	12 (7–22)
AST (U/L)	19 (16–24)	20 (16–27)	21 (17–28)
ALT (U/L)	20 (15–26)	22 (17–37)	22 (18–32)
Creatinine (mg/dL)	0.73 (0.65–0.80)	0.74 (0.64–0.84)	0.71 (0.68–0.80)
Leukocytes (/μL)	7.550 (6.210–8.910)	6.440 (5.040–7.610)	6.050 (4.860–7.200)
Neutrophils (/μL)	4.690 (3.530–6.490)	4.045 (2.870–5.090)	3.715 (2.780–4.465)
Lymphocytes (/μL)	1.785 (1.430–2.200)	1.560 (1.200–2.100)	1.400 (1.240–1.990)
Hemoglobin (g/dL)	13.9 (13.2–14.6)	14.0 (12.8–14.5)	13.8 (13.0–14.6)
Platelets (/1000 μL)	317 (268–377)	297 (258–365)	286 (246–336)
Disease activity score, median (IQR)^b
DAS28-CRP (points)^c	4.0 (3.6–4.7)	2.6 (1.8–3.8)	2.1 (1.7–3.1)

ALT: alanine transferase; AST: aspartate transferase; CRP: C-reactive protein; DAS28-CRP: Disease Activity Score 28 with C-reactive protein; ESR: erythrocyte sedimentation rate; IQR: inter-quartile range.

a

Variables analyzed using the Wilcoxon signed-rank test.

b

Variables analyzed with the difference in proportions for paired data.

c

DAS28-CRP measure was obtained in 76 (84.4%) participants at three-month follow-up and 87 (97%) participants at six-month follow-up.

Fig. 1.

Clinical response by DAS28-CRP categories. Variables analyzed with the difference in proportions for paired data. *Significant difference between baseline and three-month follow-up values. ^†Significant difference between baseline and six-month follow-up values. ^‡Significant difference between the three-month and six-month follow-up values. DAS28-CRP: Disease Activity Score 28 with C-reactive protein.

Fig. 2.

Clinical response by EULAR criteria. Variables analyzed with the difference in proportions for paired data. ^‡Significant difference between three-month and six-month follow-up values. EULAR: European Alliance of Associations for Rheumatology.

There was no significant difference in demographics or disease characteristics between patients with no response and those with moderate or good responses to MTX + LEF by EULAR criteria (Tables 3 and 4).

Table 3.

Comparison between groups according to the three-month follow-up EULAR response and logistic regression.

Characteristics	No response	Moderate/good response	Unadjusted model			Adjusted model
	n = 20	n = 53	OR	95% CI	p	OR	95% CI	p
Age (years), mean (±SD)	57.0 ± 12.2	55.4 ± 13.6	1.01	0.97–1.05	0.659	1.00	0.96–1.04	0.963
Female, n (%)	15 (75)	45 (85)	1.88	0.53–6.61	0.44	1.61	0.37–6.96	0.522
Erosive disease, n (%)	9 (45)	12 (23)	2.79	0.94–8.32	0.065	3.37	1.01–11.3	0.048
RF titers, median (IQR)	52 (28–307)	117 (37–266)	1.00	0.99–1.01	0.481	1.00	0.96–1.05	0.963
ACPA titers, median (IQR)	408 (148–500)	199 (69–500)	1.00	0.99–1.01	0.403	1.00	0.99–1.01	0.611

ACPA: anti-citrullinated peptide antibodies; CI: confidence interval; EULAR: European Alliance of Associations for Rheumatology; IQR: interquartile range; OR: odds ratio; RF: rheumatoid factor. Values in bold indicate statistical significance (p < 0.05).

Table 4.

Comparison between groups by six-month follow-up EULAR response and logistic regression analyses.

Characteristics	No response	Moderate/good response	Unadjusted model			Adjusted model
	n = 14	n = 66	OR	95% CI	p	OR	95% CI	p
Age (years), mean (±SD)	56.8 ± 10.3	54.3 ± 13.6	1.02	0.97–1.06	0.525	1.01	0.97–1.06	0.582
Female, n (%)	11 (79)	57 (86)	1.72	0.40–7.42	0.462	1.87	0.39–8.84	0.431
Erosive disease, n (%)	3 (21)	22 (33)	0.54	0.14–2.16	0.388	0.53	0.12–2.30	0.395
RF titers, median (IQR)	45 (29–366)	89 (37–255)	1.00	0.99–1.01	0.373	1.00	0.99–1.01	0.488
ACPA titers, median (IQR)	378 (199–452)	343 (108–500)	1.00	0.99–1.01	0.544	1.00	0.99–1.01	0.434

ACPA: anti-citrullinated peptide antibodies; CI: confidence interval; EULAR: European Alliance of Associations for Rheumatology; IQR: interquartile range; OR: odds ratio; RF: rheumatoid factor.

The adjusted logistic regression identified erosive disease as a risk factor for non-response by EULAR criteria at three months, but not at six months of follow-up (Tables 3 and 4). No significant risk factors were identified in the multiple linear regression for mean change in DAS28-CRP (Table S2).

Discussion

The results of this study have shown that MTX + LEF achieved an adequate clinical response in RA patients who had been previously refractory to either MTX or LEF monotherapy. In addition, we identified that patients with erosive disease are three times more likely to have no therapeutic response to the combination therapy at follow-up.

The 2024 Colombian RA high-cost account showed that among patients receiving conventional DMARD treatment, MTX + LEF was used in 17.5% [23]. These data are similar to those described in Brazil, where this combination is used in 13.3% of patients with RA [24]. This therapeutic choice is relevant for Latin American countries with high tuberculosis incidence, such as Brazil, Peru, Mexico, and Colombia [25], given the increased risk of tuberculosis infection with biological therapy, particularly tumor necrosis factor alpha (TNF-α) inhibitors [12]. In addition, there are barriers and delayed access to rheumatology consultation and biological therapy in Latin American RA patients [6]. Therefore, MTX + LEF remains a good and reasonable option for the treatment of RA in these countries.

The clinical response to combination therapy with these two DMARDs has been demonstrated in observational studies and clinical trials [7,8,10,11,15]. Kremer et al. [8] reported in a 24-week randomized, placebo-controlled clinical trial with 263 patients that 46.2% of patients treated with MTX + LEF achieved an ACR20 response, compared with 19.5% of those treated with MTX + placebo. These differences were maintained for up to 48 weeks in the open-label extension of this study [10].

Observational studies with smaller samples have also shown this clinical response using combined therapy. For example, in a study from South Korea [7], 65 RA patients received MTX + LEF for 20 weeks, achieving ACR20, ACR50, and ACR70 responses in 71.6%, 50%, and 32.4%, respectively.

In another study conducted in Panama [26], adding LEF to treatment in 278 patients with MTX-refractory RA led to DAS28 remission in 35.8% and low disease activity in 35.2%. In our cohort, we also found MTX + LEF to be an effective therapeutic option, with a higher percentage of patients achieving remission (67%) using DAS28-CRP and a lower proportion with low disease activity (10%) compared with the Panamanian study conducted in an ethnic population similar to ours.

In another Colombian open-label, multicenter study, Londoño et al. [17] also demonstrated the efficacy and safety of MTX + LEF in patients with active RA despite treatment with MTX, glucocorticoids, and non-steroidal anti-inflammatory drugs. In this study, a significant reduction in DAS28 score was observed, from a baseline value of 5.8 to 3.8 at week 24 (p < 0.001). Similarly, in our study, the MTX + LEF combination therapy resulted in a significant reduction in the DAS28-CRP score from a median of 4 points at baseline to 2.1 at six-month follow-up. In contrast, in a retrospective analysis of a Brazilian cohort of 113 patients with RA that compared monotherapy with LEF versus combined therapy with MTX + LEF, no significant differences in DAS28-CRP were observed between the groups at 6 and 12 months [27].

Regarding risk factors influencing clinical response to MTX + LEF, a retrospective Greek study conducted in 15 patients with early RA who received MTX + LEF after failing to respond to MTX monotherapy, found that 67% of them had at least one poor prognostic factor (positive RF, DAS28 >5.1, or early joint erosions on X-rays of hands or feet). A good EULAR response was achieved in 30% of those with at least one poor prognostic factor, compared with 80% in those without any such factors. However, this difference was not statistically significant (p = 0.28).

In another retrospective study of predominantly Black South African RA patients (n = 194) treated with MTX + LEF, a good or moderate EULAR response was achieved by 86% of these patients after 12 months of combination therapy [15]. These data are similar to ours, as a moderate or good EULAR response was observed in 83% of our patients after 6 months of MTX + LEF therapy. In addition, a baseline DAS28-CRP ≤5.5 was identified as a predictor of remission or low disease activity in the South African RA patients. Unlike our results, erosive disease was not found to be a factor for poor response in South African patients [15].

In the current study, the mean ESR and CRP values were within the normal range at baseline and follow-up, likely due to prior MTX or LEF monotherapy or concomitant low-dose glucocorticoids, highlighting the refractory nature of RA in our patients despite normalization of acute-phase reactants. Interestingly, an erosive disease was a risk factor for lack of response to combination therapy at three months, but not at six months of follow-up. This finding could be due to the fact that patients with erosive disease have a slower response to this therapy or because a longer cumulative treatment is required to achieve the therapeutic effect, as can be seen in Tables 3 and 4, where the percentage of patients with erosive disease without response decreases from 45% at three months to 21% at six months of follow-up.

Regarding adverse events associated with MTX + LEF, the literature reports variable results across studies. The safety of the MTX + LEF combination in RA, compared with other therapeutic regimens involving conventional synthetic DMARDs, bDMARDs, or Janus kinase inhibitors (JAKi), was evaluated in a multicenter, registry-based cohort study in Brazil (BiobadaBrasil) [18]. In this study, MTX + LEF showed a relatively good safety profile compared with the other therapeutic regimens. The risk of serious adverse events (SAEs) for MTX + LEF was not higher than that for MTX of LEF monotherapy (adjusted HR [aHR] 1.00, 95% CI 0.76–1.31, p = 0.98). In addition, MTX + LEF showed a lower risk of SAEs (aHR 0.56, 95% CI 0.36–0.88, p = 0.01) than bDMARDs/JAKi with MTX or LEF, and a lower hazard of SAEs than MTX + sulfasalazine (aHR 0.33, 95% CI 0.16–0.65, p = 0.002). The most frequent adverse reactions were infections, cardiovascular, and gastrointestinal (rates of 10.8, 2.5, and 1.9 per 100 patient-years, respectively).

A systematic review and network meta-analysis [16] showed that MTX + LEF had the highest risk of alanine aminotransferase elevations (RR 4.75, 95% CI 1.16–20.7) among all MTX combination therapies. However, the probability of withdrawal due to adverse events was not different from MTX monotherapy (RR 1.86, 95% CI 0.74–4.68). Regarding this issue, our results showed a low frequency of withdrawals due to elevated aminotransferases (3.3%), whereas the frequency was 17.7% due to any other adverse event (Table 1).

The Safety of Methotrexate in Combination with Leflunomide in Rheumatoid Arthritis (SMILE), a multicenter, cross-sectional, retrospective study [28], assessed the safety of MTX + LEF in clinical practice. The frequency of adverse events (liver function abnormalities or neutropenia) due to the combination of MTX + LEF was comparable to that of MTX monotherapy and LEF monotherapy. Liver function abnormalities were reported in 12%, 16%, and 19% of the MTX monotherapy group, LEF monotherapy group, and the MTX + LEF combination group, respectively. Neutropenia was reported in 2.3%, 5.5%, and 3.9% of the MTX monotherapy, LEF monotherapy, and MTX + LEF combination groups, respectively.

Similar to safety data reported in previous studies [16], the combination of MTX + LEF was well tolerated in our population. However, the hepatotoxic potential of the MTX + LEF combination remains a concern. The most frequent adverse event in our study was elevation of transaminases in 9 patients (10%), which was similar to the safety data reported in another Colombian study, where elevated transaminases were also the most common adverse event, occurring in 10% of patients treated with MTX + LEF [17]. It should be noted that no infections occurred in our cohort, likely due to registration bias and a relatively short follow-up time.

Among the strengths of this study is that it included patients from two rheumatology reference centers in two of the main Colombian cities, providing real-world evidence; therefore, the results can be extrapolated to the general population. In addition, it has an organized analysis and recording of outcomes previously supported in the literature, such as the DAS28-CRP and the EULAR response criteria.

This study has several limitations. Due to its retrospective design, some data may have been missing because of incomplete medical records. Furthermore, the sample size and follow-up time are insufficient to evaluate all possible SAEs that the combination of MTX + LEF could produce. Also, the resulting sample of patients without clinical response could be considered insufficient for risk factor analysis. Finally, factors such as lack of patient acceptance and treatment persistence were not evaluated.

Conclusion

In patients with inadequate response to monotherapy with MTX or LEF, the combination of MTX + LEF improves clinical response and may be an adequate therapeutic option before considering bDMARDs or tsDMARDs, except in patients with erosive disease. Furthermore, this is a relevant therapeutic alternative considering the high incidence of tuberculosis in Latin America and the barriers to access to bDMARDs or tsDMARDs.

Authors’ contributions

Andrés Cardona-Cardona: Conceptualization, Methodology, Resources, Investigation, Formal analysis, Data curation, Visualization, Writing – review & editing; Carlos Agudelo-Cardona: Conceptualization, Methodology, Resources, Investigation, Formal analysis, Data curation, Visualization, Writing – review & editing; Ana Peralta: Investigation, Data curation; Edwin Jauregui: Conceptualization, Methodology, Writing – review & editing; Joaquín Rodelo: Methodology, Resources, Formal analysis, Data curation; Carlos Muñoz-Vahos: Conceptualization, Investigation, Writing – review & editing; Gloria Vásquez: Conceptualization, Writing – review & editing; Luis González Naranjo: Conceptualization, Writing – review & editing; Adriana Vanegas-García: Conceptualization, Methodology, Resources, Investigation, Visualization, Supervision, Project administration.

Ethics approval

Approval was obtained from the Institutional Review Board and the ethics committee of HSVF in Medellín, Colombia, and from the Riesgo de Fractura CAYRE IPS in Bogotá, Colombia. Additionally, this study was performed in line with the principles of the Declaration of Helsinki 2013.

Informed consent

This research study was conducted retrospectively from data obtained for clinical purposes, without details that might disclose the identity; it did not require informed consent.

Declaration of generative AI and AI-assisted technologies in the writing process

Nothing to disclose.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Conflicts of interest

The authors declare that they have no known competing financial interests or personal relationships that could have influenced the work reported in this paper.

Data availability

Data from this study are available upon reasonable request to the corresponding author.

Acknowledgments

The authors thank Alba Luz León for the analysis and interpretation of the data, and Pharmalab for English-language editing.