Introduction

Rheumatoid arthritis (RA) is a common progressive systemic disease characterized by inflammation of the synovium leading to irreversible joint destruction and patient disability [1, 2]. Although cardinal signs of RA are related to the musculoskeletal system, extra-articular involvement, such as by the pulmonary and cardiovascular system, accompanies peripheral joint manifestations [3, 4]. The clinical course of RA is most commonly insidious [5]. The chance to develop a severe disability in RA is reported in nearly 33% of afflicted patients in 20 years, with less than 50% of patients being able to continue working 10 years following diagnosis [5].

The treatment of RA is a comprehensive approach including effective inflammation control by pharmacological agents as well as educational programs and psychological support to prevent quality of life deterioration and to improve everyday activities [5, 6]. The goal of RA treatment is to attain sustained remission or low disease activity based on shared decision-making between patients and healthcare professionals [2]. Biological disease-modifying antirheumatic drugs (bDMARDs) targeting cytokines that regulate chronic inflammation have opened a new treatment era, especially in patients who cannot tolerate conventional synthetic DMARDs (csDMARDs) [7, 8]. Monoclonal antibodies against tumor necrosis factor-alpha (TNF-α) have been widely used in clinical practice, but are not effective in all patients [9]. Approximately one-third of patients receiving a TNF inhibitor (anti-TNF) experience primary treatment failure or inefficacy according to the American College of Rheumatology (ACR) criteria 20 (ACR20), while more than 50% fail to achieve at least an ACR50 response [6]. Moreover, efficacy is lost over time or patients develop adverse events [6]. These limitations have prompted an investigation into new targets and the development of therapies with alternative mechanisms of action.

Interleukin-6 (IL-6) is a glycoprotein that is well-recognized for its role in the acute-phase inflammatory response, bone metabolism and remodeling, reproductive system, cancer progression, autoimmunity, and metabolic alterations in the liver [10,11,12,13,14]. High IL-6 levels have been detected in the synovial fluid of patients with various forms of inflammatory arthritis including RA [15]. After the identification of the signaling pathways and biological activities of IL-6 and its receptor (IL-6R), researchers decided to target IL-6R through novel therapeutic agents. It has been found that the concentrations of Il-6R have less interpersonal variability, a factor that might potentially allow for the simplification of dosing and selection of the regimen [16]. Employing the advances in biotechnology has been developed [16].

Tocilizumab (TCZ) is a humanized monoclonal antibody of the immunoglobulin G1 subclass targeting IL-6R [16]. TCZ has been shown to improve the clinical signs and symptoms, laboratory results, and radiological findings in refractory RA as monotherapy or in combination with conventional synthetic DMARDs (csDMARDs) [16] [2, 13, 16, 17]. Other than intravenously, TCZ can be administered subcutaneously (SC) once monthly [18, 19]. Despite the high treatment efficacy rates of biological DMARDs (bDMARDs), persistence to treatment is lower than anticipated in real-life settings and registries compared to randomized studies [20]. In the Danish DAINBO registry study of RA patients that initiated bDMARD treatment, nearly 70% of patients were persistent to treatment with bDMARDs, with 19% of patients achieving ACR70 after 6 months of treatment [21]. The availability of dDMARDs that are self-administered via SC injection has empowered patients and are expected to improve the overall patient experience and, thereby, persistence to treatment. However, there are challenges to this mode of administration, such as patient phobia due to self-administration, particularly in elderly patients [22].

Persistence to treatment is critical for all chronic diseases and is used as a surrogate for treatment effectiveness [21, 23]. To date, there are limited data on the persistence to TCZ and particularly TCZ-SC under real-life conditions in Greece. The objective of this study was to evaluate treatment persistence among RA patients under conditions of routine clinical practice and to determine the impact of TCZ-SC on patient-reported outcomes (PROs) as well as clinical efficacy parameters in the real-life setting. We also aimed to identify the impact of baseline patient and/or disease characteristics on persistence to treatment (Fig. 1, Table 5).

Fig. 1
figure 1

Kaplan-Meier estimates of TCZ SC persistence per treatment group. Censored observations are marked with “+”

Full size image

Material and methods

Setting, study design, and population

The present observational study was conducted in 23 sites across Greece. Overall, 222 adult patients with moderate or severe RA, according to the revised 1987 ACR or 2010 ACR/EULAR classification criteria, who started receiving treatment with TCZ-SC (monotherapy or combination) up to 4 weeks at the physician’s discretion were enrolled after written consent was provided [24, 25]. Patients who had been receiving tocilizumab for more than 4 weeks prior to enrollment or received tocilizumab in past treatments or had received treatment with any investigational agent within 4 weeks (or 5 half-lives of investigational agent, whichever is longer) of starting treatment with TCZ-SC or had a history of autoimmune disease (other than autoimmune disease of the endocrine system) or any joint inflammatory disease other than RA were excluded from this study. Participants were evaluated at baseline, week 24, and week 52. Ethics approval was obtained by the corresponding Hospital Ethics Committee/Institutional Review Board. Informed consent was obtained from all subjects. This study was conducted according to the Declaration of Helsinki and Good Clinical Practice.

Data collection and outcome variables

Patient and disease characteristics, routine general and inflammatory blood tests, clinical and patient-reported outcomes (disease activity score-28 (DAS-28), EULAR response, Health Assessment Questionnaire (HAQ), Visual Analog Scale (VAS) for pain, and patient global assessment of disease activity (PtGA)), information about attendance in a patient support program (PSP) for TCZ-SC, and information about adverse events were recorded during the study visits.

The primary efficacy endpoint was the percentage of patients still on treatment with TCZ-SC at week 52. Clinical and PROs were analyzed at weeks 24 and 52. Clinical outcomes were evaluated by determining the proportion of patients in remission or with low disease activity (DAS28 < 2.6 or DAS28 ≤ 3.2, respectively) and patients with good/moderate EULAR response criteria [26, 27]. The PRO measures used were the ΗΑQ score (decrease by ≥ 0.3 and ≥ 0.5, converted to the original tool), VAS pain (improvement of at least 10 mm in a 100 mm scale), and PtGA (improvement of at least 15 mm in a 100 mm VAS scale) [28,29,30,31].

Associations between treatment persistence and clinical outcome parameters, PROs, involvement in a PSP, and patient and disease characteristics were assessed. The persistence in monotherapy and combination groups was also investigated.

The main safety parameters were the incidence of adverse events (AEs), serious adverse events (SAEs), non-serious adverse reactions, serious adverse drug reactions (SADRs), injection site reactions (ISRs), and adverse events of special interest (AESIs). The incidence, severity, and type of these safety events were compared between persistent vs. non-persistent patients and between monotherapy vs. combination therapy groups.

Statistical analysis

Statistical analysis was performed via SAS

Descriptive statistics and standard deviation (SD) were used to compare demographic and disease characteristics, laboratory values, medical history, and treatment responses. Exploratory statistical testing and modeling were performed by using paired t-tests and ANOVA. All tests were two-sided and carried out with a 5% α-error rate.

Continuous efficacy parameters and PROs, their changes from baseline, and concomitant glucocorticoids (GCs) and csDMARDs were summarized at each time point by mean, SD, and median. Categorical efficacy parameters including the EULAR response rate and the involvement in a TCZ PSP were summarized by numbers and percentages (n, %).

The proportion of patients experiencing at least one AE was estimated with 95% confidence intervals (CIs). Adverse event rates per 100 patient-years were calculated separately for patient subpopulations. Logistic univariate regression analysis was performed to assess the impact of baseline characteristics and TCZ-SC persistence at the end of the study. The level of significance was set at 5%.

Results

The FAS included 222 patients, of whom 199, 192, and 185 patients completed 24 weeks, 40 weeks, and 52 weeks of treatment with TCZ-SC.

Of the study population (n = 222), 186 were women (83.8%). The mean (SD) age was 62.2 years (± 12.3) with approximately 60% of patients being ≥ 60 years. The mean duration of RA was 6.9 ± 7.3 years. At baseline, the mean DAS28 score was 5.6 ± 1.1 for the overall study population, and 9% of all patients had good EULAR response. The most common extra-articular manifestation was rheumatoid nodules (34.2%). Before enrollment, 74.3% of subjects had received at least one RA treatment excluding TCZ and 49.1% were bDMARD-naïve. The baseline demographics of patients and disease characteristics are presented in Table 1.

Table 1 Baseline characteristics of the study population
Full size table

Efficacy outcomes

During the study, 66.7% of patients received combination treatment with methotrexate (MTX) or other csDMARDs, and 33.3% received monotherapy with TCZ-SC.

Patients persistent to treatment with TCZ-SC comprised 89.6% (95% CI; 79.9–98.2%) of the population at week 24 and 85.1% (95% CI; 74.3–95.9%) at week 52. Overall, 6.3% of patients persistent to treatment were enrolled in the PSP. Persistence to TCZ-SC was analyzed in subgroups and was found to be statistically higher in patients receiving combination treatment at both weeks 24 and 52, though no difference was observed per line of treatment (Table 2; Fig. 1 in Supplementary Material). Disease recurrence was observed in 5% and 11.5% of patients under TCZ-SC at the same study time points (Fig. 2 in Supplementary Material).

Table 2 Drug persistence according to monotherapy vs. combination therapy and line of treatment at study timepoints
Full size table

At baseline, high disease activity (DAS > 5.1) was reported for 72.5% of the overall cohort and was decreased to 5.9% and 12% at weeks 24 and 52, respectively. Compared to baseline, DAS28 values were statistically significantly improved in both monotherapy and combination therapy groups (Table 3).

Table 3 Clinical outcome assessments for overall, monotherapy, and combination therapy groups during the study
Full size table

In the overall study population, the DAS28 score decreased from 5.6 to 3.2 at week 24 and to 2.9 at week 52 with remission rates of 37.2% and 52.9%, respectively (Table 3). Over half of the study population attained a good response according to the EULAR criteria at weeks 24 and 52 (Table 3). Statistically significant improvements in DAS28 values and good EULAR response were detected in both monotherapy and combination groups (Table 3) and in patients who received TCZ-SCas first line (TCZ-SC naïve), as second and other lines of treatment (Supplementary Table 2).

The baseline HAQ score recorded by participants described an initial mild to severe disability. A statistically significant improvement in the HAQ score was recorded at week 24 which continued to improve throughout week 52. HAQ ≥ 0.5 improvement was observed in 64.4% and 66.5% of patients at weeks 24 and 52, respectively (Table 4). The same significant improvement using the HAQ score was also observed both in monotherapy and combination therapy groups and in all treatment lines. In accordance with this, the patient’s assessment of their symptoms, using the PtGA and VAS tools, showed a statistically significant improvement in both scores at weeks 24 and 52. At week 52, 90.5% of patients reported a VAS score ≥ 10 mm, and 87.9% had a PtGA score ≥ 15 mm (Table 4). Statistically significant improvements were sustained in HAQ ≥ 0.5, VAS pain score ≥ 10 mm, and PtGA ≥ 15 mm independent from monotherapy or combination therapy or line of treatment.

Table 4 Change from the baseline in HAQ, VAS pain, and PtGA scores for the overall study population at study timepoints
Full size table

Significant associations were determined between drug persistence and clinical outcome parameters (DAS28 score and EULAR response) both at week 24 and week 52. However, statistically significant associations between drug persistence, HAQ, VAS pain, and PtGA scores were determined only at week 52 (p < 0.001) (Table 5).

Table 5 Associations between drug persistence, and clinical and patient reported outcome measures (PROs) at the study timepoints
Full size table

The between-group comparison and regression analysis of multiple variables did not show an association with TCZ-SC persistence with the type and line of treatment or the participation in a PSP but discovered a significant result regarding the presence of extra-articular manifestations. Glucocorticoids were administered to 60.4% of the baseline population, decreasing to 55.9% at week 52, with 14.9% being able to reduce or discontinue their glucocorticoid dose at week 24.

Safety

Overall, 18.5% of study participants (n = 41) experienced AE(s), 3.2% (n = 7) SAE(s), while 6.3% (n = 14) had ADR(s) (Supplementary Material Table 3). When comparing monotherapy vs. combination-therapy groups, a statistically significant higher incidence of injection site reactions (ISR) and deaths was observed in the monotherapy group compared to the combination therapy group (12.1% vs. 0.7%, respectively, for ISR, p = 0.013; 6.1% vs. 0%, respectively, for deaths, p = 0.033) (Supplementary Material Table 4). Regardless of the incidence of AEs, in the majority of participants, TCZ treatment did not affect total cholesterol or LDL-C levels (Supplementary Material Table 4). In the full population, the most common AEs encountered were dyslipidemia (2.3%), rash (1.8%), and leukopenia (1.4%) (Supplementary Material Table 5). Regarding laboratory evaluations, clinically significant abnormal results were only noted in a few study participants throughout the study. The most common clinically significant laboratory abnormalities at both week 24 and week 52 involved liver enzymes and lipid levels (reported in ≤ 1% of participants (Supplementary Material Table 6).

Per treatment group (monotherapy or combination treatment), adverse events of special interest (AESI) or SAEs were rarely reported (≤ 4.1% of participants per treatment group) (Supplementary Material Table 7 and Table 8). Regarding persistence (persistent vs. non-persistent participants), the incidence of AEs, treatment-related AEs, SAEs, ISRs, and deaths were statistically significantly higher in the non-persistent group of participants compared to the persistent group (Supplementary Material Table 9 and 10).

The annualized rate for AEs, SAEs, and AESI per 100 patient-years was low at 1.51, 1.43, and 2.13, respectively (Supplementary Material Table 11).

AEs were the main reason for TCZ discontinuation (permanent or temporary) in 8.1% of the study participants (Supplementary Material Table 12). The most frequently reported AEs leading to TCZ-SC discontinuation were rash, leukopenia, and diarrhea. Only permanent tocilizumab dose discontinuations were noted in the non-persistent and the monotherapy treatment groups (Supplementary Material Table 4 and 10).

Discussion

The present study contributed insights on the real-life use, safety, and efficacy of TCZ-SC as well as patient-perceived effect of treatment on quality of life from a cohort of patients who were enrolled in Greece and initiated treatment with TCZ-SC within 4 weeks from enrollment. The majority of patients were persistent to TCZ-SC throughout the study duration. The overall persistence to TCZ-SC was 89.6% at week 24 and 85.1% at week 52, with a recurrence-free rate of 88.5%. These rates are comparable to the literature-reported data for persistence to bDMARDs under real-life conditions [19, 23, 32, 33]. The high rates of persistence were sustained throughout the study up to week 52 across treatment groups, namely, for the monotherapy group 78.4%, combination therapy group 88.5%, and in biologic naïve patients 87.2%. Higher persistence rates for a systematic review of 43 observational studies, conducted both in EU and US, that investigated the adherence and persistence to biologics in chronic inflammatory diseases, such as RA, psoriasis, and psoriatic arthritis (PsA), reported that the range of continuation (persistence) to treatment at 12 months ranged from 32.0 to 90.9%, with higher treatment continuation rates when co-administration with methotrexate or other DMARDs was reported [33].

For many chronic immune system diseases, the availability of IV and SC administration modes aims to ensure that patients are provided with options that are acceptable in order to attain long-term persistence to treatment [34]. Patient preference to treatment administration mode is subjective and can be influenced by factors such as age and, among older individuals, ease of administration (dexterity) [35, 36]. No issues with ease of treatment administration were reported. Based on the review of relevant literature references, factors that contribute towards preference to SC administration include the convenience and comfort of at-home treatment, lack of a requirement to prepare the medicine, the reduced chance of medication error, and ready-to-use syringes with the correct dose, whereas IV infusion preference is mainly driven by fear for self-injection, the feeling of safety due to hospital administration, and opportunity for interactions with the medical team [34, 37,38,39,40].

Persistence to treatment is a critical component of effectiveness. The clinical outcome measurements indicated sustained statistically significant improvement of disease severity across a range of disease-related parameters in addition to acceptable tolerability to TCZ-SC in patients with moderate-to-severe RA. These data are in agreement with real-world published evidence [18, 41,42,43,44,45]. In the present study, the majority of patients achieved remission according to the DAS28 score and a good EULAR response. During the study, no unexpected adverse events and no safety concerns were identified relevant to the available safety information on TCZ [18, 41,42,43,44,45]. The incidence of adverse events related to lipid levels, hematology, and infections was low, which is in agreement with published data [43, 46, 47]. The percentage of TCZ-SC discontinuation due to safety reasons was comparable to published evidence from real-life studies [43]. There were significant associations between the persistence and clinical outcome parameters both at weeks 24 and 52, as well as PROs at the end of the study.

This study had some inherent limitations that pertain to its design and concern the lack of control group and formal sample size calculation. The effect of missing data in the analysis of retrospective observational studies cannot be ruled out. Compliance to SC treatment, which could interfere with the study assessments, as opposed to supervised intravenous infusions was not determined. In addition to the aforementioned factors, the conduct of the study in a single country poses limitations on the external validity of its results. Treatment decisions were at the discretion of the treating physicians. However, the findings of this study contribute valuable information regarding real-life patterns of use of TCZ-SC in Greece. The investigators’ decision-making to administer TCZ-SC was based on current medical practice and preceded the consideration of the patient’s eligibility for enrollment into the study.

Conclusion

The present study contributes national data from Greece on the real-life safety and effectiveness of TCZ-SC when used either as monotherapy or combination therapy in patients with moderate to severe RA, as well as on patient persistence to treatment under conditions of routine clinical practice. In agreement with published data, TCZ-SC is an effective, safe, and acceptable treatment option for RA patients. Patient acceptability to treatment is a critical component for long-term disease management in patients with chronic immune conditions. The findings of this study could prompt further investigation to determine the optimal TCZ treatment administration on eligible RA patients in Greece under real-life conditions with the aim of delivering personalized treatment options that can maximize persistence to treatment.