A national, multicenter, retrospective study evaluating retention rate and efficacy of tocilizumab treatment in patients with active rheumatoid arthritis who had an inadequate response to csDMARDs and/or TNF inhibitors

Background/aim To describe the disease activity and retention rate in rheumatoid arthritis (RA) patients with inadequate response (IR) to conventional synthetic disease-modifying antirheumatic drugs (csDMARDs) and/or tumor necrosis factor inhibitors (TNFis) who were prescribed tocilizumab (TCZ) as first-line or second-line biologic treatment in real-world setting. Materials and methods Data gathered from patients’ files was used in a multicenter and retrospective context. Retention rates and the Disease Activity Score in 28 joints with CRP (DAS28-CRP) were evaluated at time points. The relationship of drug efficacy with factors such as smoking, obesity, and previous use of TNFis was also examined. Results One hundred and twenty-four patients with a median (IQR) RA duration of 3.7 (7.4) years were included. Mean (SD) age was 52.9 (12.9) and 75% of the patients were female. TCZ retention rates in the 6th and 12th months were 94.1% and 86.6%, respectively. In all patients, DAS28-CRP level decreased significantly from baseline to Months 3 and 6. There was an increase in patients with remission and/or low disease activity and a decrease in patients with high disease activity at Month 3 and Month 6 (p < 0.001 for both). Disease activity was similar between subgroups based on body mass index, smoking status, and previous use of TNFis at any time point. Regression analysis showed that absence of concomitant corticosteroid treatment independently was associated with remission/LDA achievement at Month 6 [OR = 0.31, 95% CI (0.14–0.72), p = 0.006], and Month 12 [OR = 0.35, 95% CI (0.13–0.94), p = 0.037]. Overall, 25 mild adverse events were reported. Conclusion TCZ was found to be effective and safe in RA patients with IR to csDMARDs and/or TNFis. The drug retention rate was considered satisfactory with more than half of the patients continuing TCZ treatment at Month 12.


Introduction
Rheumatoid arthritis (RA) is a chronic inflammatory disease characterized by pain and swelling of peripheral joints resulting in a progressive functional disability and joint damage, as well as an increased risk of osteoporosis and cardiovascular morbidity. The goals of therapy for RA are to decrease joint inflammation and pain, preserve the ability of patients to function in activities of daily living and work, and prevent joint deformity and joint destruction.
Today clinical remission or low disease activity (LDA) is recommended to be aimed by current European Alliance of Associations for Rheumatology (EULAR) and American College of Rheumatology (ACR) guidelines when treating patients with RA [1,2]. Initial treatment with conventional synthetic disease-modifying antirheumatic drugs (csDMARDs), such as methotrexate, has become the standard of care. In the last decades, RA management has dramatically changed with the introduction of biologic DMARD (bDMARD) therapies; tumor necrosis factor inhibitors (TNFis) were the first to be developed [3].
However, approximately 30%-40% of patients develop an inadequate response to csDMARDs and TNFis [4,5]. The increasing knowledge about the pathogenesis of RA and potential targets for treatment resulted in the introduction of new bDMARD options (non-TNF biologics) with different mechanisms of action and targeted synthetic DMARDs (tsDMARDs). Current guidelines for RA, ACR and the EULAR recommend adding a bDMARD or tsDMARD if the treatment target is not achieved with the first csDMARD strategy. Additionally, if a bDMARD or tsDMARD has failed, treatment with another bDMARD or tsDMARD should be considered. Patients may be given an agent with another mechanism of action or a second TNFi, if there is a failure of TNFi treatment [1,2].
Tocilizumab (TCZ) is a humanized monoclonal antibody against IL-6 receptor which inhibits IL-6 mediated signaling and reduces RA disease symptoms, prevents joint damage, and improves other symptoms such as fatigue, anemia, bone loss, and depression. Besides, it may prevent the development of type 2 diabetes mellitus and increased cardiovascular risk, which are also found to be related to RA [6]. Its efficacy and safety have been demonstrated in many clinical studies, whether in combination with methotrexate [7][8][9], or in monotherapy [10], and further confirmed with the real-life studies [11][12][13].
In this retrospective study, we aimed to evaluate the retention rate, efficacy, and safety of TCZ in routine clinical practice in RA patients with inadequate response (IR) to csDMARDs and/or TNFis. The association of drug efficacy in the treatment of TCZ with factors such as smoking, obesity, and previous use of TNFis was also investigated in this study.

Materials and methods
This study was designed as a national, multicenter, retrospective, and noninterventional study in which data was retrospectively collected between January 2020 and September 2020 with contributions of 13 different rheumatology centers in Turkey. Adult patients with a diagnosis of RA according to the ACR 2010 classification [14] who initiated TCZ between November 30, 2015, and April 02, 2020, and treated with TCZ at least three months after csDMARDs or a TNFi as first or second-line biologic therapy were included in the study. Secondary data were obtained from the patient files and uploaded into an electronic data collection system using electronic data collection forms.
The primary objective was to evaluate the treatment retention rate of TCZ for from 3rd month. Secondary objectives were to analyze the efficacy of TCZ treatment on the 3 rd , 6 th , 12 th , 18 th , and 24 th (±1 month was considered as the time frame for the 3 rd month and ±2 months for other time points) months by change in Disease Activity Score in 28 joints with CRP (DAS28-CRP) levels from baseline, to evaluate the efficacy of TCZ treatment in patient subgroups planned according to body mass index (BMI; with a cutoff 30 kg/m 2 ), previous use of TNFis, and smoking status. The safety profile of TCZ based on the changes in the lipid profile, liver enzymes, hepatitis B virus (HBV) serology, and latent tuberculosis infection (LTBI) reactivation risk (evaluated with purified protein derivative (PPD) test and/ or QuantiFERON-TB Gold Plus test), as well as adverse events over time were also recorded.
Data were obtained for demographics, smoking status (current or nonsmoker), concomitant diseases, disease duration of RA, previous and concomitant RA treatments, laboratory parameters, C-reactive protein (CRP) levels, DAS28-CRP levels, date of TCZ treatment initiation and discontinuation, route of administration, and adverse events. The number of days for TCZ treatment duration was calculated by subtracting the date of starting TCZ treatment from the date of the last visit entered for each patient, and the number of months was calculated by dividing the total number of days by 30. Analysis of total TCZ usage time and retention rate has been made using the last-observation-carried-forward (LOCF) method and as-observed data.
This was a secondary data use study, and therefore the treatment retention rate of patients with TCZ treatment for one year was hypothesized as 60% based on literature [16]. Precision was estimated as ±12%, and with a 95% confidence interval, an approximate minimum sample size of 100 patients was required. All data gathered during the course of the study were summarized using descriptive statistics; numeric variables were expressed as mean (standard deviation) or median (minimum and maximum or quartiles); categorical variables were expressed as numbers and percentages. Normality was assessed with visual (histogram and probability graphs) and analytical methods (Kolmogorov-Smirnov/Shapiro-Wilk tests). The relationship between categorical variables was examined with McNemar test. Kaplan-Meier estimates were calculated for the retention rate of TCZ. Logistic regression analysis was carried out to identify the predictors of remission or LDA achievement at months 6 and 12. In this analysis, the explanatory variables used were age, gender, duration of RA, concomitant diseases, the previous number of csDMARDs, previous use of TNFis, concomitant use of csDMARDs, concomitant use of corticosteroids, and being TCZ monotherapy. For statistical analysis, PASW 18.0 for Windows was used. A p-value of <0.05 was considered significant.
The efficacy of TCZ treatment was evaluated by changes in DAS28-CRP levels from baseline at related time points. In all patients, DAS28-CRP level decreased significantly from baseline to months 3 and 6 and there was a significant increase in patients with remission and/ or LDA and a decrease in patients with HDA at month 3 and month 6 (p < 0.001 for both, Figure 2a). DAS28-CRP levels were also evaluated at subgroups based on smoking status (smoker and nonsmoker), previous use of TNFis, and BMI (obese: BMI ≥ 30 kg/m 2 and nonobese: BMI < 30 kg/m 2 ). Among the patients, 19.4% were active smokers and 30.6% were obese having a BMI ≥30 kg/m 2 . At baseline, 54.2% of smokers had HDA and 41.6% had MDA, while all of the obese patients had MDA or HDA (36.8% MDA, 63.2% HDA). In both subgroups based on smoking status, patients had significant improvement in disease activity at month 3 (p = 0.013 for smokers and p < 0.001 for nonsmokers) and month 6 (p = 0.041 for smokers and p = 0.005 for nonsmokers) compared to baseline (Figure 2b and 2c). Similar significance was also observed in BMI subgroups: improvement of disease activity at month 3 (p < 0.001 for nonobese patients and p = 0.014 for obese patients) and month 6 (p = 0.001 for nonobese patients) compared to baseline (Figures 2d and  2e). However, disease activity levels did not differ between subgroups based on smoking status and BMI at any time point.
At baseline, 31 patients (25%) had previous use of TNFis. Of them, 61.3% had HDA and 29.0% had MDA, which was slightly different than TNFi-naïve subgroup (60.2% HDA and 38.7% MDA, p = 0.051). Although the percentage of patients with remission and/or LDA increased in time compared to baseline, it did not reach a significance level in TNFi-experienced subgroup (p > 0.05 for all time points). However, regarding TNFi-naïve subgroup, there was a significant increase in patients with remission and/or LDA and a decrease in patients with HDA at month 3 (p < 0.001) and month 6 (p = 0.001) compared to baseline (Figures 3a and 3b). In addition, disease activity levels did not differ between subgroups based on TNFi use at related time points (p > 0.05 for all time points).
Lipid profile and liver enzymes were also analyzed during the study. Although the percentage of patients with abnormal total cholesterol and LDL levels increased slightly at month 3, it did not reveal a significant difference (Table 3, p > 0.05 for both). Furthermore, these increments returned to baseline ratios at month 6. The percentages of patients with normal levels of triglycerides and HDL were also similar at related time points compared to baseline (p > 0.05 for all). Similar trends of increment of transaminase levels at month 3 and decrement at month 6 were observed as well (Table 4). Grade ≥ 1 alanine aminotransferase (ALT) was observed in 5.7% and 17.2% of patients at baseline and month 3, respectively (p = 0.004). All other comparisons regarding Grade ≥ 1 ALT and AST (aspartate aminotransferase) were similar at related time points compared to baseline (p > 0.05 for all).
In total, 25 adverse events were reported during the study (Table 5). TCZ treatment was temporarily discon- tinued in three patients (one for skin rash during TCZ infusion which resolved quickly and two for planned surgeries). There was one death due to the prior malignancy, which was unrelated to the study drug. All other adverse events were mild and resolved completely. The Hepatitis B profile and LTBI reactivation were also evaluated for safety profile. Overall, no hepatitis B infection and LTBI reactivation were recorded during the follow-up.

Discussion
In this real-world data study, TCZ was found to be effective and safe in RA patients with inadequate response to csDMARDs and/or TNFi treatments. The drug retention rate for TCZ was also considered satisfactory with more than half of the patients continuing TCZ treatment at Month 12. In our study, approximately nine out of ten patients were still on TCZ treatment at the 6 th month, maintaining a rate of 86% at 12 months. Flipo et al. found the median retention rate of TCZ treatment until one year was 69%, without any difference between TCZ monotherapy or combination with csDMARDs [12]. Haraoui et al. recently published a real-world study including 1912 patients from 16 countries, in which 78.7% of the patients were found to continue to receive TCZ treatment [17]. In another study, the drug continuation rate of TCZ was found 89.4% at one year, which was concordant with our study [18].
There are several studies demonstrating the efficacy and safety of TCZ. These studies showed the superiority of TCZ compared to placebo [9][10][11][12][13]19], and to TNFis [20] by disease activity measurements. Similar results were observed in a German cohort including patients refractory to csDMARDs or TNFis, and TCZ in combination with csDMARDs or as monotherapy resulted in significantly more patients achieving remission compared with TNFis [21]. In a recent meta-analysis evaluating the efficacy of targeted immune modulators (Janus kinase inhibitors, TNFis, and other non-TNFi therapies) in patients with IR to csDMARDs, it was shown that TCZ had the highest effect to achieve DAS28 remission among all targeted immune modulators compared with a csDMARD alone [22]. Bykerk et al. found a 56.8% achievement of DAS28 remission with TCZ treatment at 24 weeks in a patient cohort with IR to DMARDs and/or TNFis [23]. In a recent study from the Italian biologics' register GISEA, remission was achieved in 51% and 52.3% of the patients using TCZ at 6 and 24 months, respectively [24]. Similarly, in our study, approximately 60% of the patients achieved clinical remission or LDA at month 3, and this finding was sustained for two years.
Obesity and smoking are among the important risk factors for RA [25]. Although poor responses were associated with smoking and increased BMI [26,27], recent  studies showed that TCZ response was not affected by smoking status or BMI of the patient [28][29][30]. In our study, improvements in disease activity were similarly observed in patient subgroups based on these factors. Furthermore, the efficacy of the TCZ treatment in TNFi-experienced subgroup was comparable to that in the TNFi-naïve subgroup. Bykerk et al. showed improvements in DAS28 scores with TCZ treatment in the treated population while better outcomes were observed in TNFi-naive patients than TNFi-exposed patients [23]. As a result, both TNFinaïve and TNFi-experienced patients may benefit from TCZ treatment. Overall, our efficacy results are in line with observations from other studies. Both changes in lipid profile and liver enzymes were acceptable. The treatment with TCZ was well tolerated and overall, the adverse events were as expected from the published records.
Our study has a few limitations. The number of patients refractory to TNFis and at the subgroups were limited. Also, the number of patients on follow-up after 12 months was considerably low as a result of follow-up challenges in routine clinical practice. Additionally, the lack of a comparison group was considered as a limitation of this study. Since this was a retrospective study, data from all patients were not available for several variables.
In conclusion, TCZ drug retention rates of 94.1% and 86.6% at months 6 and 12 respectively were concordant with previously conducted TCZ clinical studies and TCZ seemed to be effectively showing favorable DAS28-CRP response rates in routine clinical practice in patients with moderate to severe RA responding inadequately to csD-MARDs and/or TNFis.

Acknowledgment/disclaimers/conflict of interest
The study was funded by Roche Mustahzarlari San AS., Turkey in accordance with Good Publication Practice (GPP3) guidelines. All authors contributed to the conduction and writing of this study and have approved the final version.
The authors thank the patients and all the investigators who participated in the study, the clinical operations team (Ömer Fatih Karakaş, MD; Gözde Yelkenci; Türkan Duman, IQVIA CRAs), and Ayşe Esra Aydın, MD from Monitor CRO for providing medical writing support, which was funded by Roche Mustahzarlari San AS., Turkey in accordance with Good Publication Practice (GPP3) guidelines.
MET received payment for speakers' bureaus from Ab-bVie, Pfizer, UCB, MSD, Roche, Boehringer Ingelheim, and Pharmactive; and received payment as an instructor for Pharmactive. SY received payment for speakers' bureaus from AbbVie, Abdi İbrahim İlaç, Pfizer, UCB, MSD, Roche, Farmanova. ZY received payment for speakers' bureaus and consulting fees from AbbVie, Pfizer, UCB, MSD, Roche, Amgen, and Lilly. YK received payment for speakers' bureaus and consulting fees from AbbVie and Roche. MÖ received payment for speakers' bureaus and