Difficult-to-treat rheumatoid arthritis (D2T RA): clinical issues at early stages of disease

Abstract

Objectives

Most studies on difficult-to-treat rheumatoid arthritis (D2T RA) have focused on established RA. Here, we analyse whether disease activity in the early stages of RA could influence progression to a D2T RA under real-life conditions. Other clinical and treatment-related factors were also analysed.

Methods

A longitudinal multicentre study of patients with RA was conducted from 2009 to 2018. Patients were followed up until January 2021. D2T RA was defined based on EULAR criteria (treatment failure, signs suggestive of currently active/progressive disease and management being perceived as problematic by the rheumatologist and/or patient). The main variable was disease activity in the early stages. The covariates were sociodemographic, clinical and treatment-related factors. We ran a multivariable logistic regression analysis to investigate risk factors associated with progression to D2T RA.

Results

The study population comprised 631 patients and 35 (5.87%) developed D2T RA. At the time of diagnosis, the D2T RA group were younger, with a higher disability, 28-joint Disease Activity Score (DAS28) score, tender joint count and pain scores. In our final model, DAS28 was not statistically significantly associated with D2T RA. No differences were found between groups for therapy. Disability was independently associated with D2T RA (OR: 1.89; p=0.01).

Conclusions

In this cohort of patients newly diagnosed with RA, our results do not allow us to prove the influence of active disease according to DAS28. However, we did find that younger patients and those with elevated initial disability scores are more likely to develop D2T RA regardless of other factors.

Correspondence to Dr Leticia Leon; [email protected]

WHAT IS ALREADY KNOWN ON THIS TOPIC

• Despite ‘treat to target’ and the availability of a range of advanced therapies, difficult-to-treat rheumatoid arthritis (D2T RA) remains a relevant clinical problem. Evidence for the D2T RA population has focused on established RA. The aim of our study was to analyse whether disease activity at diagnosis could influence progression to D2T RA under real-life conditions.

WHAT THIS STUDY ADDS

• We did find that patients with elevated initial disability scores are more likely to develop D2T RA regardless of other factors.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• The implementation of more effective strategies in the early stages of the disease and focused on the most influential factors, including severe disability, may change disease course and prevent D2T RA.

Introduction

The main therapeutic objective in rheumatoid arthritis (RA) is to prevent joint damage and disability. Treatment seeks to achieve and maintain ‘remission’ or, at least, minimal disease activity. Recent changes in the management of RA have improved the clinical course of the disease.¹ The ‘window of opportunity’ concept^2–4 underscores the need to initiate aggressive treatment early in the disease process and thus improve prognosis. Intensification of treatment, dose optimisation and the ‘treat-to-target’(T2T) strategy^5–8 are also key goals in the management of affected patients. Remarkable progress in treatment has been made with both biological therapies and JAK-STAT pathway inhibitors.

Recommendations on the management of RA have come to include continuous supervision and monitoring of disease course and associated comorbidity,^{6 9} thus enabling remission 28-joint Disease Activity Score (DAS28) (<2.6) and low disease activity DAS28 (<3.2) in clinical practice.¹ The prevalence of remission in patients with RA is reported to be around 30%.^{10 11} Prognosis has improved in recent decades and may reflect early diagnosis and more accurate treatment rather than a change in the characteristics of the process itself.¹²

Despite these advances, there are patients with poor disease evolution, reflecting the fact that RA is a heterogeneous disease in terms of severity, clinical course and response to treatment, requiring a complex management. A European Alliance of Associations for Rheumatology (EULAR) Task Force recently defined the concept of ‘difficult-to-treat’ (D2T) RA as persistent symptoms and/or signs after failure of at least two biological or targeted synthetic disease-modifying antirheumatic drugs (b/tsDMARDs), each of which has a different mechanism of action.¹³ D2T RA encompasses not only uncontrolled inflammatory disease, but also wider contextual factors such as chronic pain and fatigue, as well as comorbidities, recurrent infections and treatment-limiting adverse events.^14–16

Although the mechanisms leading to D2T RA are varied and not fully understood, they can be classified into two types, namely, multidrug resistance and difficulties in intensifying treatment (including comorbidities, poor adherence, financial problems or the patient’s reluctance to intensify treatment).^{14 17}

In a recent study of more than 1700 patients, 10% of all patients with RA did not achieve remission or low disease activity despite intensive treatment in real clinical practice.¹⁸ The prevalence of DT2 RA is assumed to range between 3% and 17% according to the series reviewed.^{14 19} The main characteristics of patients with D2T RA are younger age, with high disease activity, comorbidity with fibromyalgia, marked pain and disability, poorer quality of life, higher doses of glucocorticoids and more DMARDs, patient desire to intensify treatment and greater consumption of healthcare resources.¹⁷ Evidence for the EULAR-defined D2T RA population is limited and based mainly on definitions that do not take into account features of active disease and the perception of challenging RA by the clinician and/or patient,²⁰ focusing mainly on the first of them (failure of at least two b/tsDMARDs). In addition, almost all patients included have long-standing RA. A recent study found that prognosis was difficult in patients with D2T RA who receive MTX ≤ 8.6 mg/week or who receive glucocorticoids. Therefore, focusing on the most effective treatments from the onset of the disease may prevent increases in its frequency.²¹

Knowledge about why some patients develop D2T RA could be improved. The few studies performed are cross-sectional, regardless of the duration of RA. Consequently, to identify patients at risk, it would be interesting to identify modifiable associated factors at early stages. The aim of our study was to analyse whether disease activity at diagnosis could influence progression to D2T RA under real-life conditions. Other clinical and therapeutic factors were also analysed.

Methods

Setting

The study was performed at three tertiary hospitals of the National Health System of the Community of Madrid, Spain, namely, Hospital Clínico San Carlos (HCSC), Hospital Universitario de La Paz (HULP) and Hospital Universitario de la Princesa (HUP). Each institution has a catchment population of approximately 400 000. The Rheumatology Service at each institution provides care to its catchment population.

Study design

We performed a multicentre observational longitudinal prospective study. Since we have been treating RA according to the EULAR/ACR classification criteria under a T2T strategy since 2010,²² patients were included from June 2009 to June 2018 and followed up for 5 years or less (lost to follow-up or end of the study (January 2021)).

Population, patients and data sources

All patients fulfilled the diagnostic criteria for RA and were included at diagnosis. Data were registered prospectively at baseline and every 6–12 months in the local databases and collected in the IntegrAR cohort.

Briefly, the IntegrAR cohort is the product of integrating local prospective inception RA cohorts with standardised data. IntegrAR was initiated in 2010 and contains a relational database whose records are fed from the databases of the three hospitals involved through web services (HCSC, HULP and HUP). It is an active and dynamic cohort that currently contains more than 210 variables, including sociodemographic and clinical characteristics, diagnostic procedures and therapeutic options.

To be included, patients had to be in the IntegrAR cohort during the study period, age ≥18 years, and diagnosed with RA according to the EULAR 2010 classification criteria.²² They also had to be taking bDMARDs or tsDMARDs.

Patient data were obtained during routine clinical practice, and patients from IntegrAR provided their signed informed consent to participate.

Outcome measures

The primary outcome was D2T RA as defined by the EULAR criteria,¹³ which are set out below. In order to collect the variables, we have had to quantify some of the criteria, based on the Task Force agreements¹³ and the clinical criteria of the research team.

• Treatment according to EULAR recommendations and failure of ≥2 bDMARDs/tsDMARDs (with different mechanisms of action) after failure of conventional synthetic DMARDs (unless contraindicated).

• Presence of at least one of the following: moderate or more severe disease activity (DAS28 ≥3.2), signs and/or symptoms suggestive of active disease (erythrocyte sedimentation rate (ESR) ≥60, C reactive protein ≥0.6), inability to taper glucocorticoids, rapid radiographic progression and RA symptoms that worsen quality of life (Health Assessment Questionnaire, HAQ ≥1.2).

• The management of signs and/or symptoms is perceived as problematic by the rheumatologist and/or the patient (physician global assessment ≥50; patient global assessment ≥50; text recorded by the rheumatologist in the clinical history).

The second and the third criteria had to be fulfilled at least 6 months after diagnosis.

The main variable measure in this study was low disease activity, understood as a DAS28 <3.2.

The covariables initially considered were as follows:

• Baseline sociodemographic characteristics including centre, sex, date of birth and educational level.

• Associated baseline comorbidity (hypertension, hypercholesterolaemia, diabetes mellitus, cardiovascular and cerebrovascular events, congestive heart failure, depression, history of cancer, gastroduodenal ulcer, chronic obstructive pulmonary disease, interstitial lung disease, liver disease, osteoporosis, kidney failure).

• Characteristics of RA: date of onset of symptoms, date of diagnosis, date of first visit, rheumatoid factor (RF), anticitrullinated protein antibody (ACPA).

• Assessment of RA at baseline: number of painful and swollen joints, assessments of the disease by doctor and patient (patient global health assessment, physician/evaluator global health assessment), pain scale, DAS28 score, disability (HAQ), ESR.

• First therapeutic regimen received: DMARDs scheduled with start and end date and glucocorticoids at baseline. The list of DMARDs included conventional synthetic DMARDs (azathioprine, gold, methotrexate, leflunomide, sulfasalazine and antimalarials) and biological agents (ts/bDMARDs, TNF inhibitors (anti-TNF), abatacept, rituximab, tocilizumab and Janus kinasa inhibitors). Patients taking any DMARD for more than 3 months were considered exposed. In the case of glucocorticoids, patients were considered exposed if on medication for at least 2 months.

First, we carried out feature selection, where we combined comorbidity data into two variables: (1) presence of hypertension, hypercholesterolaemia and/or diabetes mellitus; and (2) total number of comorbidities, excluding the previous ones. Second, we selected covariables with a prevalence higher than >5% in the low disease activity group. Finally, only the covariables suspected of playing a relevant role in developing D2T RA were considered to balance the distribution between patients with low disease activity and patients with activity (age, sex, combined therapy, HAQ, RF-positive, ACPA-positive, depression, year of diagnosis, comorbidity, intake of glucocorticoids and methotrexate and centre).

Statistical analyses

Descriptive statistics of patients’ sociodemographic and clinical characteristics, as well as their disease activity and treatment, are presented as frequency distributions for qualitative variables and as the mean and SD or median and percentiles for quantitative variables. The t-test was used for the analysis of normally distributed continuous variables. Continuous variables with a non-normal distribution were analysed using the Mann-Whitney U test or the Kruskal-Wallis test if there were more than two categories. The categorical variables were analysed by using the χ² or Fisher’s exact test. Analyses were performed to assess the differences between D2T RA and non-D2T RA. Logistic multivariate regression models were fitted to examine the possible influence of baseline disease activity (low activity) on the main outcome, D2T RA. These models were adjusted for the independent variables of disease activity (DAS28, HAQ) and therapy (intake methotrexate).

We limited the number of variables in the multivariate model following the rule of Freeman and the value of 10 events per variable to ensure that it was reliable.^23–25

Analyses were performed using STATA V.13 software (Stata Corp) and R statistical software V.4.2.1.²⁶ Missing data (HAQ n=27, DAS28 n=17) were managed using the classification and regression trees algorithm from the mice R package, with the default setting and 5 imputations.²⁷

Results

Sociodemographic, clinical and therapy-related characteristics

A total of 631 patients were included: 178 (28%) from HCSC, 319 (51%) from HULP and 134 (21%) from HUP. All of them were taking DMARDs during the study period, and the median follow-up was 5 (p25–p75: 2.14–5) years. Most patients were women (77.5%) in their fifties, with no statistically significant differences between the centres. Lag time from symptoms to diagnosis of RA was 0.7 (2.2) years. Interestingly, 30% of the patients had at least 1 comorbid condition at baseline and 28% had a cardiovascular risk factor, with arterial hypertension being the most prevalent. Most of the patients were overweight, with obesity in 10% (median body mass index, 25.8 (p25–p75: 22.7–28.9)). Most were RF-positive or ACPA-positive, with a median value of 106 (p25–p75 42–285) and 445 (p25–p75: 202–1055), respectively. Regarding disease activity, most patients had at least a moderate DAS28 (median, 4.7 (p25–p75: 3.6–5.8)) and some degree of disability (median HAQ, 1 (0.375–1.625)).

Glucocorticoids were the first drugs prescribed in 66% of the patients. Most patients were prescribed csDMARDs at diagnosis (71.54%), 97% in the first 6 months and 99% in the first 12 months, with a mean start of 30.41 (65) days after diagnosis. The drugs were prescribed in monotherapy in 87% of the cases, and, as expected, the most prevalent was methotrexate (85%) either in combination or in monotherapy. Ten patients (1.58%) received ts/bDMARDs as their initial treatment, with etanercept (n=3) and abatacept (n=3) the most frequently prescribed, followed by tocilizumab (n=2), adalimumab (n=1) and golimumab (n=1). The remaining patients were prescribed methotrexate, leflunomide or hydroxychloroquine.

Interestingly, 17% of patients used at least one ts/bDMARDs during follow-up, with a median first prescription lag time of 1.4 (0.79–2.5) years after diagnosis. Of these, 42 patients required at least two different biologics during the study period. Concerning the other D2T RA criteria, 63% and 39% of patients met the second and third criteria, respectively (see figure 1). Finally, 35 cases were defined as D2T RA and 596 cases as non-D2T RA. Clinical characteristics and therapy are summarised in tables 1 and 2.

Figure 1. Number of patients who met each of the three D2T RA criteria. b/tsDMARDs, biological or targeted synthetic disease-modifying antirheumatic drugs; D2T RA, difficult-to-treat rheumatoid arthritis; csDMARD, conventional synthetic disease-modifying antirheumatic drugs.

Table 1

Baseline characteristics of patients in the D2T RA and the non–D2T RA groups

Variable	RA n=631	D2T RA n=35	Non-D2T RA n=596	P value
Women, n (%)	489 (77.5)	29 (82.8)	460 (77.1)	0.4
Age, mean (SD), years	55.1 (15.5)	45.7 (12.8)	55.6 (15)	≤0.001
BMI, mean (SD)	26.0 (4.7)	26.9 (4)	26 (4.7)	0.37
Positive RF at baseline, n (%)	526 (83.3)	24 (68.5)	502 (84.2)	0.01
Positive ACPA at baseline, n (%)	499 (79.0)	26 (74.2)	473 (79.3)	0.47
DAS28 at RA diagnosis, mean (SD)	4.6 (1.5)	5 (1.4)	4.6 (1.5)	0.11
CRP at RA diagnosis, mean (SD)	1.9 (5.9)	1.5 (2.1)	2 (6)	0.65
ESR at RA diagnosis, mean (SD)	30 (22.2)	28.3 (22.2)	30.1 (22.3)	0.65
TJC (0–28) at baseline, mean (SD)	6.7 (6.5)	8.9 (7.9)	6.6 (6.4)	0.04
SJC (0–28) at baseline, mean (SD)	6.6 (5.8)	7.7 (6.5)	6.5 (5.7)	0.24
PGHA (mm) at baseline, mean (SD)	46.9 (26.5)	59.2 (26.9)	46.1 (26.3)	≤0.01
PhGHA (mm) at baseline, mean (SD)	38.7 (22.1)	43.8 (20.2)	38.4 (22.2)	0.29
Pain VAS (mm) at baseline, mean (SD)	45.2 (28)	56.5 (27.7)	44.1 (1.6)	0.02
HAQ at RA diagnosis, mean (SD)	1 (0.7)	1.3 (0.6)	1 (0.7)	0.03
Comorbidities, n (%)
Hypertension	124 (19.6)	5 (14.2)	119 (19.9)	0.41
Dyslipidaemia	93 (14.7)	10 (28.5)	83 (13.9)	0.01
Depression	36 (5.7)	4 (11.4)	32 (5.3)	0.13
Diabetes mellitus	34 (5.3)	2 (5.7)	32 (5.3)	0.96
Heart disease	14 (2.2)	0	14 (2.3)	0.35
Vascular disease	33 (5.2)	1 (0.1)	32 (5)	0.72
Liver disease	8 (1.2)	2 (5.7)	6 (1)	≤0.001
Kidney disease	4 (0.6)	0	4 (0.6)	0.62
Lung disease (ILD/COPD)	18 (2.8)	1 (2.8)	17 (2.8)	0.99
Cancer	8 (4.4)	0	8 (4.9)	0.34

ACPA, anticitrullinated-protein antibody; BMI, body mass index; COPD, chronic obstructive pulmonary disease; CRP, C reactive protein; DAS28, 28-joint Disease Activity Score; D2T RA, difficult-to-treat rheumatoid arthritis; ESR, erythrocyte sedimentation rate; HAQ, Health Assessment Questionnaire; ILD, interstitial lung disease; PGA, Patient Global Health Assessment; PhGA, Physician Global Health Assessment; RF, rheumatoid factor; SJC, swollen joint count; TJC, tender joint count; VAS, Visual Analogue Scale.

Table 2

Baseline treatment strategies in the D2TRA and the non-D2TRA groups

Variable	RA n=631	D2TRA n=35	Non-D2TRA n=596	P value
Diagnosis to DMARD treatment initiation, days, mean (SD)	30.4 (65)	13.6 (25.4)	31.4 (66.5)	0.11
Patients under biological therapies at follow-up, n (%)	106 (16.8)	35(100)	71 (11.9)	≤0.001
Diagnosis to biological DMARD treatment initiation, years, mean (SD)	1.6 (1.2)	1.6 (1.1)	1.7 (1.2)	0.67
Glucocorticoid use at baseline (more than 3 months), n (%)	418 (66.2)	23 (65.7)	395 (66.2)	0.94
csDMARDs prescription at baseline, n (%):
DMARD monotherapy	556 (88.1)	26 (74.2)	530 (88.9)	≤0.001
Combination of DMARDs (2 or more)	75 (11.8)	9 (25.7)	66 (11)
csDMARDs drug at baseline, n (%):
Methotrexate	540 (85.5)	507 (85)	33 (94.2)	0.13
Leflunomide	44 (6.9)	43 (7.2)	1 (2.8)	0.32
Antimalarial	60 (9.5)	59 (9.9)	1 (2.8)	0.16

csDMARD, Conventional synthetic disease-modifying antirheumatic drugs; DMARDs, disease-modifying antirheumatic drugs; D2TRA, difficult-to-treat rheumatoid arthritis.

Regarding the biologics in D2T RA patients, 88.5% of them used TNF inhibitors as first biological agent, followed by abatacept (9%). The most prevalent biological drug as second or third choice was rituximab (37%) or tocilizumab (31%), and JAK inhibitors were used in three patients.

At diagnosis, patients in the D2T RA group were younger, with a higher DAS28 tender joint count, higher pain scores and a higher level of disability than the non-D2T RA group. Regarding comorbidities, the main differences were observed for dyslipidaemia (28% vs 13%) and liver disease (5% vs 1%), with a greater prevalence in D2T RA patients, in whom a trend towards depression was observed (11% vs 5%). Other characteristics are shown in table 1.

Table 2 shows differences in therapy between the two groups. No statistically significant differences were recorded between the groups for prevalence of the different DMARDs or glucocorticoids. As expected, D2T RA patients took more bDMARDs during follow-up. Regarding the biologics used as the initial regimen, all except one taking abatacept were in the non-D2T RA group.

Influence of baseline disease activity on D2T RA

In the multivariable logistic regression analysis (table 3), we did not find statistically significant differences for disease activity (OR 1.18 (95% CI 0.37 to 3.20), p=0.76) neither methotrexate intake (OR 2.82 (95% CI 0.82 to 17.70), p=0.16), whereas at least moderate disability (OR 1.89 (95% CI 1.14 to 3.16), p=0.01) and a younger age at baseline (OR 0.95 (95% CI 0.93 to 0.98), p=0.01) increased the risk of D2T RA, regardless of other factors. Glucocorticoids (p=0.95) and sex (p=0.90) dropped from the final model.

Table 3

Multivariate regression analysis

	OR	95% CI	P value
DAS28	1.18	0.37 to 3.20	0.76
Age at baseline	0.95	0.93 to 0.98	0.01
Methotrexate	2.82	0.82 to 17.70	0.16
Health Assessment Questionnaire	1.89	1.14 to 3.16	0.01

DAS28, 28-joint Disease Activity Score.

Discussion

In this study, we investigated whether early disease activity plays a role in the detection of D2T RA under real-world conditions. Our results do not allow us to prove the influence of active disease state according to DAS28, although we did find that younger patients and those with elevated initial disability scores are more likely to develop D2T RA, regardless of other factors. Our evaluation of whether the first therapeutic strategy might result in better outcomes depending on the drugs selected did not yield conclusive results.

As in other studies, recent-onset patients with RA were middle-aged women whose most common clinical and serological manifestations were positive RF and ACPA titers.^{28 29}

We found that 5.87% of patients had D2T RA, consistent with findings reported elsewhere.^{14 19} Previous studies of D2T RA patients were based only on treatment failure, without taking into account other criteria (ie, signs suggestive of presently active/progressive disease and management being perceived as problematic by the rheumatologist and/or patient). Although several studies have described D2T RA-related factors in patients with RA^{17 30 31} with long-standing disease, the effect of different variables on this outcome has not been studied in the early stages of disease, and comparisons between them would not be accurate. In addition, consideration should be given to the possibility of other lifestyle, economics and non-modifiable sociodemographic factors that may contribute to the development of D2T RA.^{14 32–35}

Although an evidence base is not available, identifying affected patients in the early stages of disease could, arguably, limit progression to D2T RA in those experiencing multidrug failure. In our study, we systematically analysed the effect of several variables, mainly those related to disease; however, we also examined the response to treatment (DMARDs and glucocorticoids).

Disease activity (moderate or severe, measured early) based on the DAS 28 was not associated with D2T RA, although statistically significant differences were observed after separate analysis of DAS28 components, painful joints and patient global health assessment, suggesting that rheumatologists act well towards the most objective part of the disease, with little emphasis on the more subjective aspects.³⁶ Interestingly and similarly, a statistically significant higher early disease burden was found in D2T RA than in non D2T RA patients, as seen in lower physical functioning and worse levels of pain; therefore, it was not possible to distinguish disease activity arising from pain or inflammation per se. While temporary, this finding has been reported in other studies of patients with much longer-standing disease who eventually develop D2T RA.¹⁷ Nevertheless, structural and irreversible disability may develop until inflammation is controlled, with the result that early control of function should remain a mandatory objective.

As for the ‘dual-target’ strategy suggested in response to the risks of overtreatment with immunosuppression in pursuit of a treatment target that may not be achievable with drug therapy alone, Ferreira et al³⁷ recommend a personalised approach that also addresses the impact of the disease on the patient, including disability, which can be assessed using patient-reported outcome measures and requires a multidisciplinary approach.

Several non-pharmacological therapeutic strategies have proven helpful in patients with RA. Educational, psychological and self-management interventions have diminished non-inflammatory complaints (pain, fatigue and functional disability),^38–40 and while high-quality evidence of the beneficial effect of these non-pharmacological strategies is still lacking in D2T RA patients, they could prevent a large number of cases of D2T RA. In our setting, we provide educational and self-management interventions by their rheumatologists and specialist nurses of the rheumatology participating centres.

An important aspect of this study was our evaluation of the role of different treatment strategies in the development of D2T RA. We were unable to find differences between RA groups with respect to therapeutic management. The use of glucocorticoids does not seem to have any effect, as reported in a sample of patients with long-standing D2T RA.²¹ Similarly, the choice of methotrexate as the first DMARD does not seem to affect the future development of D2T RA. In a recent study, D2T RA patients receiving methotrexate ≤8.6 mg at baseline are considered to be at high risk of recurrence of D2T RA.²¹ The authors reported that their doses were lower than in Western countries. The mean dose we prescribe after 3 months is around 15–20 mg/week (common practice is to start with a lower dose,⁴¹); however, data were not collected systematically, thus precluding comparisons.

Methotrexate is the cornerstone of treatment of early RA, although it is likely that higher doses are more poorly tolerated. Despite common use of folic acid, intolerance to methotrexate is still reported in 30%–60% of patients.^{42 43} This may be one of the reasons why adherence is only 50%–94% at 1 year and 25%–79% at 5 years.⁴⁴ Intolerance to methotrexate was associated with younger age, as reported elsewhere.^{43 45 46} In our study, fewer patients with D2T RA took methotrexate than in the non-D2T RA group. The age difference between the groups may have affected the use and retention of methotrexate.

Early introduction of aggressive immunosuppressive treatment at appropriate dosages and with combination therapy when needed is recommended for achieving remission.⁴⁷ We found no differences between the lag time to the first DMARDs used, although most were prescribed early, suggesting that the ‘T2T’ strategy and the current EULAR recommendations are being implemented in clinical practice, thus leading to good control of the disease. Consequently, given that disease control has improved, disease activity is probably not an associated factor.

Previous studies questioned whether the sequence of DMARDs and/or resistance to specific drugs affects progression to D2T RA.²⁰ Our analysis of drug choice and combinations showed no trend towards D2T RA. One reason why the first regimen is not a determinant in this condition could be that rheumatologists are applying the T2T strategy in clinical practice and that modifications (adding, switching) are made so quickly that the effect of the initial regimen is diminished. Another reason might be that the most frequently prescribed drug is methotrexate, making it difficult to find differences between treatments. Lastly, we cannot forget that the D2T RA group is too small and heterogeneous in terms of therapy to found differences.

Our study is subject to a series of limitations. First, its observational nature meant that treatment was not allocated randomly, and prescription bias cannot be completely ruled out. Second, as data were collected from clinical practice, some were missing; however, these account for no more than 5%, and imputation methods were applied to correct this issue. Third, given the extended follow-up, some of the data may reflect clinical practice that has changed over time as more therapeutic options have become available. We included cases treated several years ago, when biological agents were not available. Once they did become available, combinations were not usual until experience and evidence started to support them. This is also reflected in our limited experience with other targeted therapies, such as JAK inhibitors. In addition, for the definition of D2T RA, researchers delimited the ranges of the criteria, which are not specifically included in the definition, and it is possible that the cut-off points chosen might be biased by their clinical experience. Finally, despite the 10-year inclusion period, the number of events was small, thus diminishing the strength of the analysis in terms of, for example, therapeutic regimen used. Nevertheless, the strengths of our study include its prospective design and the inclusion of non-selected patients from three hospitals with a follow-up limited to 5 years (more than 600 patients were analysed), which provided us with an overall vision of real-world evidence and thus reflected clinical practice in our setting. Our results add to current knowledge on the real-world management of D2T RA in the window of opportunity.

Our findings should be interpreted as an initial approach to the potential role that early detection of clinical issues could play in the prevention of D2T RA. A shift in current clinical practice is needed to encourage clinicians to consider comorbidities and predisposing conditions early in the disease, before D2T RA develops, rather than after it manifests.³³ Other observational studies, starting at disease onset, will be needed to resolve these issues.

To conclude, the implementation of more effective strategies in the early stages of the disease and focused on the most influential factors, including severe disability, may change disease course and prevent D2T RA. Nevertheless, the presence of refractory disease highlights the need for continued target discovery, new therapies and a multidisciplinary approach.

Data availability statement

Data are available on reasonable request. The datasets used and/or analySed during the current study are available from the corresponding author on reasonable request.

Ethics statements

Patient consent for publication

Not applicable.

Ethics approval

This study involves human participants and was approved by Hospital Clinico San Carlos Ethics Committee approved this study (18/321-E_BS). Participants gave informed consent to participate in the study before taking part.

Footnote

Twitter @LeticiaLeonM

Contributors LL, LA and AMG contributed to the design of the research, to the analysis of the results and to the writing of the manuscript. PL-V, IG-A, MN-N, DFN, ZR and BF-G contributed to the design of the research and data curation. LA is responsible for the overall content as guarantor and accepts full responsibilty for the work.

Funding This work was supported by the Instituto de Salud Carlos III (ISCIII), Ministry of Health, Spain (Fondo de Investigaciones Sanitarias: PI18/01188, PI18/00371; Red de Investigación en Inflamación y Enfermedades Reumáticas (RIER) RD16/0012/0014 and Red de enfermedades Inflamatorias (REI) RD21/0002/0001); and co-funded by el Fondo Europeo de Desarrollo Regional (FEDER).

Competing interests None declared.

Provenance and peer review Not commissioned; externally peer reviewed.

References

¹ Naranjo A, Cáceres L, Hernández-Beriaín JÁ, et al. Ajuste en La práctica clínica de las directrices treat-to-target para La artritis reumatoide: resultados del estudio toarcan. Reumatología Clínica 2016; 12: 34–8. doi:10.1016/j.reuma.2015.03.003

² van Nies JAB, Tsonaka R, Gaujoux-Viala C, et al. Evaluating relationships between symptom duration and persistence of rheumatoid arthritis: does a window of opportunity exist? Results on the Leiden early arthritis clinic and ESPOIR cohorts. Ann Rheum Dis 2015; 74: 806–12. doi:10.1136/annrheumdis-2014-206047

³ van Nies JAB, Krabben A, Schoones JW, et al. What is the evidence for the presence of a therapeutic window of opportunity in rheumatoid arthritis? A systematic literature review. Ann Rheum Dis 2014; 73: 861–70. doi:10.1136/annrheumdis-2012-203130

⁴ Contreras-Yáñez I, Pascual-Ramos V. Window of opportunity to achieve major outcomes in early rheumatoid arthritis patients: how persistence with therapy matters. Arthritis Res Ther 2015; 17: 177. doi:10.1186/s13075-015-0697-z

⁵ Smolen JS, Aletaha D, Bijlsma JWJ, et al. Treating rheumatoid arthritis to target: recommendations of an international task force. Ann Rheum Dis 2010; 69: 631–7. doi:10.1136/ard.2009.123919

⁶ Smolen JS, Landewé RBM, Bijlsma JWJ, et al. EULAR recommendations for the management of rheumatoid arthritis with synthetic and biological disease-modifying antirheumatic drugs: 2019 update. Ann Rheum Dis 2020; 79: 685–99. doi:10.1136/annrheumdis-2019-216655

⁷ Grigor C, Capell H, Stirling A, et al. Effect of a treatment strategy of tight control for rheumatoid arthritis (the TICORA study): a single-blind randomised controlled trial. Lancet 2004; 364: 263–9. doi:10.1016/S0140-6736(04)16676-2

⁸ Toledano E, Ortiz AM, Ivorra-Cortes J, et al. Are rheumatologists adhering to the concepts window of opportunity and treat-to-target? earlier and more intense disease-modifying anti-rheumatic drug treatment over time in patients with early arthritis in the pearl study. Clin Exp Rheumatol 2018; 36: 382–8.

⁹ Saunders SA, Capell HA, Stirling A, et al. Triple therapy in early active rheumatoid arthritis: a randomized, single-blind, controlled trial comparing step-up and parallel treatment strategies. Arthritis Rheum 2008; 58: 1310–7. doi:10.1002/art.23449

¹⁰ Vázquez I, Graell E, Gratacós J, et al. Prognostic markers of clinical remission in early rheumatoid arthritis after two years of dmards in a clinical setting. Clin Exp Rheumatol 2007; 25: 231–8.

¹¹ Sfriso P, Salaffi F, Montecucco CM, et al. MonitorNet: the italian multi-centre observational study aimed at estimating the risk / benefit profile of biologic agents in real-world rheumatology practice monitornet: studio italiano osservazionale multicentrico per la valutazione del. Reumatismo 2009; 61: 132–9. doi:10.4081/reumatismo.2009.132

¹² Welsing PMJ, Fransen J, van Riel PLCM. Is the disease course of rheumatoid arthritis becoming milder? Time trends since 1985 in an inception cohort of early rheumatoid arthritis. Arthritis Rheum 2005; 52: 2616–24. doi:10.1002/art.21259

¹³ Nagy G, Roodenrijs NM, Welsing PM, et al. EULAR definition of difficult-to-treat rheumatoid arthritis. Ann Rheum Dis 2021; 80: 31–5. doi:10.1136/annrheumdis-2020-217344

¹⁴ de Hair MJH, Jacobs JWG, Schoneveld JLM, et al. Difficult-to-treat rheumatoid arthritis: an area of unmet clinical need. Rheumatology (Oxford) 2018; 57: 1135–44. doi:10.1093/rheumatology/kex349

¹⁵ Buch MH. Defining refractory rheumatoid arthritis. Ann Rheum Dis 2018; 77: 966–9. doi:10.1136/annrheumdis-2017-212862

¹⁶ Kearsley-Fleet L, Davies R, De Cock D, et al. Biologic refractory disease in rheumatoid arthritis: results from the British Society for rheumatology biologics register for rheumatoid arthritis. Ann Rheum Dis 2018; 77: 1405–12. doi:10.1136/annrheumdis-2018-213378

¹⁷ Roodenrijs NMT, van der Goes MC, Welsing PMJ, et al. Difficult-to-treat rheumatoid arthritis: contributing factors and burden of disease. Rheumatology (Oxford) 2021; 60: 3778–88. doi:10.1093/rheumatology/keaa860

¹⁸ Takanashi S, Kaneko Y, Takeuchi T. Characteristics of patients with difficult-to-treat rheumatoid arthritis in clinical practice. Rheumatology (Oxford) 2021; 60: 5247–56. doi:10.1093/rheumatology/keab209

¹⁹ Bécède M, Alasti F, Gessl I, et al. Risk profiling for a refractory course of rheumatoid arthritis. Semin Arthritis Rheum 2019; 49: 211–7. doi:10.1016/j.semarthrit.2019.02.004

²⁰ Tan Y, Buch MH. “Difficult to treat” rheumatoid arthritis: current position and considerations for next steps. RMD Open 2022; 8: 1–7. doi:10.1136/rmdopen-2022-002387

²¹ Yoshii I, Sawada N, Chijiwa T. Clinical characteristics and variants that predict prognosis of difficult-to-treat rheumatoid arthritis. Rheumatol Int 2022; 42: 1947–54. doi:10.1007/s00296-022-05124-1

²² Aletaha D, Neogi T, Silman AJ, et al. 2010 rheumatoid arthritis classification criteria: an American college of rheumatology/European League against rheumatism collaborative initiative. Ann Rheum Dis 2010; 69: 1580–8. doi:10.1136/ard.2010.138461

²³ Freeman DH. Applied categorical data analysis. New York: Marcel Dekker, 1987.

²⁴ Concato J, Peduzzi P, Holford TR, et al. Importance of events per independent variable in proportional hazards analysis. I. background, goals, and general strategy. J Clin Epidemiol 1995; 48: 1495–501. doi:10.1016/0895-4356(95)00510-2

²⁵ Doerken S, Avalos M, Lagarde E, et al. Penalized logistic regression with low prevalence exposures beyond high dimensional settings. PLoS One 2019; 14: e0217057. doi:10.1371/journal.pone.0217057

²⁶ Team RDC. A language and environment for statistical computing. R Found Stat Comput; 2020. Available: http://www.r-project.org/

²⁷ van S, Groothuis-Oudshoorn K. Mice: multivariate imputation by chained equations in R. J Stat Softw 2011; 45: 1–67. doi:10.18637/jss.v045.i03

²⁸ Emery P, Solem C, Majer I, et al. A European chart review study on early rheumatoid arthritis treatment patterns, clinical outcomes, and healthcare utilization. Rheumatol Int 2015; 35: 1837–49. doi:10.1007/s00296-015-3312-3

²⁹ Combe B, Rincheval N, Berenbaum F, et al. Current favourable 10-year outcome of patients with early rheumatoid arthritis: data from the ESPOIR cohort. Rheumatology (Oxford) 2021; 60: 5073–9. doi:10.1093/rheumatology/keab398

³⁰ Roodenrijs NMT, Welsing PMJ, van Roon J, et al. Mechanisms underlying DMARD inefficacy in difficult-to-treat rheumatoid arthritis: a narrative review with systematic literature search. Rheumatology (Oxford) 2022; 61: 3552–66. doi:10.1093/rheumatology/keac114

³¹ Messelink MA, Roodenrijs NMT, van Es B, et al. Identification and prediction of difficult-to-treat rheumatoid arthritis patients in structured and unstructured routine care data: results from a hackathon. Arthritis Res Ther 2021; 23: 184. doi:10.1186/s13075-021-02560-5

³² Nagy G, Roodenrijs NMT, Welsing PMJ, et al. EULAR points to consider for the management of difficult-to-treat rheumatoid arthritis. Ann Rheum Dis 2022; 81: 20–33. doi:10.1136/annrheumdis-2021-220973

³³ Dey M, Nagy G, Nikiphorou E. Comorbidities and extra-articular manifestations in difficult-to-treat rheumatoid arthritis: different sides of the same coin? Rheumatology (Oxford) 2022; 7: 1–18. doi:10.1093/rheumatology/keac584

³⁴ Dey M, Busby A, Elwell H, et al. Association between social deprivation and disease activity in rheumatoid arthritis: a systematic literature review. RMD Open 2022; 8: 1–11. doi:10.1136/rmdopen-2021-002058

³⁵ Busby AD, Wason J, Pratt AG, et al. Predictors of poor function in RA based on two prospective UK inception cohorts. do comorbidities matter? Rheumatology (Oxford) 2022; 61: 1563–9. doi:10.1093/rheumatology/keab598

³⁶ Jurgens MS, Overman CL, Jacobs JWG, et al. Contribution of the subjective components of the disease activity score to the response to biologic treatment in rheumatoid arthritis. Arthritis Care Res (Hoboken) 2015; 67: 923–8. doi:10.1002/acr.22532

³⁷ Ferreira RJO, Duarte C, Ndosi M, et al. Suppressing inflammation in rheumatoid arthritis: does patient global assessment blur the target? A practice-based call for a paradigm change. Arthritis Care Res (Hoboken) 2018; 70: 369–78. doi:10.1002/acr.23284

³⁸ Roodenrijs NMT, Hamar A, Kedves M, et al. Strategies regarding goal setting and self-management in difficult-to-treat rheumatoid arthritis: preliminary results of a systematic literature review informing the 2020 Eular recommendations for the management of difficult-to-treat rheumatoid arthritis. Ann Rheum Dis 2020; 79: 595.

³⁹ Roodenrijs NMT, Hamar A, Kedves M, et al. SAT0052 therapeutic strategies in difficult-to-treat rheumatoid arthritis: preliminary results of a systematic literature review informing the 2020 Eular recommendations for the management of difficult-to-treat rheumatoid arthritis. Ann Rheum Dis 2020; 79: 957. doi:10.1136/annrheumdis-2020-eular.4378

⁴⁰ Gossec L, Molto A, Romand X, et al. Recommendations for the assessment and optimization of adherence to disease-modifying drugs in chronic inflammatory rheumatic diseases: a process based on literature reviews and expert consensus. Joint Bone Spine 2019; 86: 13–9. doi:10.1016/j.jbspin.2018.08.006

⁴¹ Bujor AM, Janjua S, LaValley MP, et al. Comparison of oral versus parenteral methotrexate in the treatment of rheumatoid arthritis: a meta-analysis. PLoS One 2019; 14: e0221823. doi:10.1371/journal.pone.0221823

⁴² Dhir V, Aggarwal A. Methotrexate-related minor adverse effects in rheumatoid arthritis: more than a nuisance. J Clin Rheumatol 2012; 18: 44–6. doi:10.1097/RHU.0b013e31823ee540

⁴³ Ćalasan MB, van den Bosch OFC, Creemers MCW, et al. Prevalence of methotrexate intolerance in rheumatoid arthritis and psoriatic arthritis. Arthritis Res Ther 2013; 15: R217. doi:10.1186/ar4413

⁴⁴ Curtis JR, Bykerk VP, Aassi M, et al. Adherence and persistence with methotrexate in rheumatoid arthritis: a systematic review. J Rheumatol 2016; 43: 1997–2009. doi:10.3899/jrheum.151212

⁴⁵ Patil P, Parker RA, Rawcliffe C, et al. Methotrexate-induced nausea and vomiting in adolescent and young adult patients. Clin Rheumatol 2014; 33: 403–7. doi:10.1007/s10067-013-2389-x

⁴⁶ Vijaykumar D, Dhir V, Jain S, et al. Assessing methotrexate intolerance and its prevalence in rheumatoid arthritis: development and validation of the MISA questionnaire. Int J Rheum Dis 2021; 24: 1294–301. doi:10.1111/1756-185X.14207

⁴⁷ Combe B, Landewe R, Daien CI, et al. 2016 update of the EULAR recommendations for the management of early arthritis. Ann Rheum Dis 2017; 76: 948–59. doi:10.1136/annrheumdis-2016-210602