Effectiveness and Safety of Certolizumab Pegol in Axial Spondyloarthritis in a Real-World Setting in Greece: A Sub-Analysis of the Prospective Non-Interventional CIMAX Cohort Study

Objectives: We report the effectiveness and safety of certolizumab pegol (CZP) treatment in a real-world Greek axial spondyloarthritis (axSpA) population, including patients with radiographic (r-axSpA) and non-radiographic (nr-axSpA) disease. Methods: We performed a sub-analysis of the Greek cohort from CIMAX (NCT02354105), a multicentre, non-interventional cohort study that prospectively investigated CZP treatment in patients with axSpA. The primary outcome was change from baseline (CfB) in Bath Ankylosing Spondylitis Disease Activity Index (BASDAI) to Week 52. Results: Across 12 sites in Greece, 126 patients (r-axSpA: 91; nr-axSpA: 35) received ≥1 dose of CZP and were included in the Safety Set (SS), with 120 patients (r-axSpA: 86; nr-axSpA: 34) included in the Full Analysis Set (FAS). The mean (standard deviation [SD]) CfB in BASDAI at Week 52 was −3.8 (2.0) in the overall axSpA population, with numerically greater improvements observed for nr-axSpA patients compared with r-axSpA (nr-axSpA: −4.2 [2.1]; r-axSpA: −3.7 [2.0]). Improvements in the axSpA population, including r-axSpA and nr-axSpA subpopulations, were observed in key secondary and additional outcomes at Week 52. Overall, 14.3% (18/126) of patients in the axSpA population experienced ≥1 adverse event (AE). There were no serious AEs or deaths reported during the study. Conclusions: Patients with r-axSpA and nr-axSpA treated with CZP in clinical practice in Greece showed improvements in disease activity and key symptoms. CZP treatment may therefore help address the substantial health burden associated with axSpA in Greece.


INTRODUCTION
Axial spondyloarthritis (axSpA) is a chronic, inflammatory rheumatic disease that represents a major health and financial burden for patients and society. 1,2 The disease belongs to a group of chronic inflammatory diseases termed spondyloarthritides (SpA), which 163 CIMAX GREEK SUB-ANALYSIS are often seen in carriers of the HLA-B27 antigen. [3][4][5] The chronic inflammation associated with axSpA predominately affects the axial skeleton and entheses, with patients experiencing symptoms of chronic and severe back pain, morning stiffness and fatigue. 6 The disease is also associated with a number of peripheral and extramusculoskeletal manifestations (EMM), including peripheral arthritis, enthesitis, acute anterior uveitis, inflammatory bowel disease (IBD) and psoriasis. 7 These symptoms and manifestations can negatively impact patients' quality of life and productivity. [8][9][10] Patients with axSpA may also have X-ray evidence of structural damage to the sacroiliac joints; these patients are classified as having radiographic disease (r-axSpA, also known as ankylosing spondylitis), while patients who do not have definitive signs of sacroiliac joint damage on X-rays are described as having non-radiographic disease (nr-axSpA). 6 Compared with patients with nr-axSpA, patients who have r-axSpA are more likely to be male and to have experienced a longer time since diagnosis. 6,11,12 The prevalence of r-axSpA in the general adult population of Greece is thought to be between 0.24% and 0.29%, 13,14 with more than 80% of these patients likely to be HLA-B27 carriers. 3 The prevalence of nr-axSpA in Greece is unclear. The Assessment of Spondyloarthritis international Society (ASAS) and the European Alliance of Associations for Rheumatology (EULAR) recommend that patients with axSpA receive non-steroidal anti-inflammatory drugs (NSAIDs) as first-line drug treatment, with biologic disease-modifying anti-rheumatic drugs (bDMARDs) offered as second-line options to patients with persistently high disease activity. 15 The bDMARDs currently approved for patients with axSpA, including patients with both radiographic and non-radiographic disease, are tumour necrosis factor inhibitors (TNFi) and interleukin-17A (IL-17A) inhibitors. 6,16 In 2014-2015, 9,279 SpA patients using pharmacy dispensed prescriptions for bDMARDs were identified in a Greek prescription database covering more than 95% of the Greek population (10,223,000 Greek citizens). 17 The efficacy and safety of the TNFi certolizumab pegol (CZP) has been demonstrated in patients with axSpA in clinical trial settings. 12,[18][19][20] In addition, a recent noninterventional European study (CIMAX) was conducted to assess CZP effectiveness and safety across the axSpA disease spectrum, in a real-world setting. Primary results of the CIMAX study have been published previously. 6 Here, we report results from a sub-analysis of the Greek cohort from CIMAX, which was conducted to assess the effectiveness and safety of CZP treatment in a real-world Greek axSpA population, including patients with r-axSpA and nr-axSpA.

Study design and patients
The CIMAX study (NCT02354105) was a multicentre, non-interventional cohort study conducted between January 2015 and March 2018, that prospectively investigated the effectiveness and safety of CZP in patients with axSpA being treated in clinical practice. The full study design has previously been reported. 6 Patients were recruited from six European countries and the study was conducted at 101 study sites. This included 12 sites in Greece located across Athens, Larisa, Maroussi, Patra and Thessaloniki. 21 Eligible patients were required to have a clinical diagnosis of active axSpA (r-axSpA or nr-axSpA) from their treating physician and to have been newly prescribed CZP as part of their routine clinical care. During the study, patients were permitted to receive concomitant medications, including synthetic DMARDs (sDMARDs) and NSAIDs. The decision to prescribe CZP was made by treating physicians based on the patients' needs and in accordance with local regulations or guidelines. The dose and administration schedule for CZP treatment were selected according to the Summary of Product Characteristics (SmPC). 22 The study was conducted in line with local legal requirements and the Declaration of Helsinki; in Greece, the study was reviewed by an Independent Ethics Committee. Prior to data collection, all patients provided written informed consent through a study-specific Patient Data Consent form. Patients were able to withdraw from the study at any time.

Study assessments
Outcomes pertaining to effectiveness and safety were assessed up to Week 52, with clinical examinations and investigations performed by the treating physician at baseline (assessments performed up to 30 days before or 10 days after the first dose of CZP), Week  peripheral manifestations (including peripheral arthritis and enthesitis) and EMMs (including uveitis, IBD and psoriasis) were reported for Week 52. The proportion of patients achieving ASDAS clinically important improvement (ASDAS-CII; reduction of ≥1.1 from baseline), ASDAS inactive disease (ASDAS-ID; ASDAS <1.3) and ASDAS <2.1 (includes ASDAS low disease and ASDAS-ID) were also reported at Week 52, in accordance with the ASAS/EULAR recommendations. 15 Adverse events (AEs) were recorded when reported by a patient to their treating physician, according to the Medical Dictionary for Regulatory Activities (MedDRA ® ), version 20.1.

Statistical analysis
Effectiveness and safety outcomes were analysed for patients in the overall Greek axSpA population, as well as by disease subgroup (Greek patients with r-axSpA or nr-axSpA) and prior TNFi exposure (Greek patients who were TNFi-naïve or TNFi-experienced). Patients who received ≥1 dose CZP were included in the Safety Set (SS); those with baseline and ≥1 post-baseline BASDAI assessment formed the Full Analysis Set (FAS) and were included in the effectiveness analyses. Patients had to have a baseline BASDAI assessment within a predefined window (up to 30 days before or 10 days after the first dose of CZP). Effectiveness outcomes are reported using descriptive statistics. Categorical variables are reported as the percentage of responders (with 95% confidence intervals [CI]), and continuous variables are reported as mean (standard deviation [SD]). Outcomes are reported using either observed case analysis (with no imputation for missing data), or multiple imputation (MI) or non-responder imputation (NRI). For the MI analysis, categorical age (≤/>45 years), subgroup (r-axSpA or nr-axSpA) and prior TNFi exposure (naïve or experienced) were specified as covariates and data were assumed to be missing at random. Imputed data are reported for Week 52 only. Calculation of ASDAS using MI (including ASDAS improvement and disease states), was based either on CfB in CRP or CfB in ESR; if the CRP value was missing, the observed ESR value was used and if neither were available, imputed CRP values were used. ASDAS CfB was only calculated if same-type values were available at both timepoints, eg, if both ASDAS-CRP(pre) and ASDAS-CRP(post) were available. Statistical analyses were performed using SAS ® (SAS Institute, Cary, NC, USA) Version 9.2.

Patient disposition and baseline characteristics
In total, 127 patients enrolled in the study from Greece. Of these, 126 started the study and 105 completed the study. The 126 patients who started the study received ≥1 dose of CZP and were therefore included in the SS. The most common reasons for study discontinuation were lack or loss of efficacy (9/126 [7.1%]), withdrawal of consent (not due to adverse events; 2/126 [1.6%]) and 'other' reasons (7/126 [5.5%]). There were 120 patients with a baseline and ≥1 postbaseline BASDAI assessment (Figure 1) who were included in the effectiveness analyses (FAS). All patients in the SS and FAS had a confirmed diagnosis of r-axSpA or nr-axSpA

Primary outcome
The mean (SD) CfB in BASDAI was -3.8 (2.0) for the 101/120 patients in the axSpA population who completed the Week 52 visit ( Figure 1A). Numerically greater improvements were observed in the nr-axSpA subpopulation compared with the r-axSpA subpopulation, with nr-axSpA patients demonstrating a mean (SD) CfB of -4.2 (2.1) vs -3.7 (2.0) for patients with r-axSpA. The improvements observed when missing values were imputed (MI) were consistent with these findings, with a CfB of -3.7 (2.0) in the axSpA population, and -4.2 (2.1) and -3.5 (2.0) in the nr-axSpA and r-axSpA subpopulations, respectively ( Table 2). The greatest improvement in BASDAI score between two consecutive study visits was observed between baseline and the first study visit at Week 12, with further improvements seen at subsequent study visits. The mean (SD) CfB for Weeks 12 and 24 in the axSpA population were -2.4 (1.7) and -3.1 (2.2), respectively, with similar changes observed in the r-axSpA (-2. 3 (Figure 1A). All six items of the BASDAI improved to a similar degree at Weeks 12, 24 and 52 for the overall axSpA population (Figure 1B).

Secondary and additional variables
In the axSpA population overall, there were improvements in key secondary outcomes at Week 52, including in the percentage of ASAS20/40 responders (Figure  Figure 2B) of patients, respectively. Improvements in BASFI and PtGADA were also observed over 52 weeks in the axSpA population (Figure 3). At Week 52, the mean (SD) CfB in BASFI was -3.2 (2.2) in the overall axSpA population, and -3.1 (2.3) and -3.5 (2.0) in the r-axSpA and nr-axSpA subpopulations, respectively ( Figure 3A). For PtGADA, the mean (SD) CfB at Week 52 was -4.6 (2.6) for the axSpA population, and -4.3 (2.6) and -5.1 (2.6) for the r-axSpA and nr-axSpA subpopulations, respectively ( Figure 3B). Improvements in the axSpA population, including the r-axSpA and nr-axSpA subpopulations, for additional outcomes, such as ASDAS, PhGADA and total back pain were also observed at Week 52 ( Table 2)

DISCUSSION
The objective of this analysis was to assess the effectiveness and safety of CZP treatment in patients with axSpA, including r-axSpA and nr-axSpA, treated during routine clinical practice in Greece.
In the Greek axSpA population, patients treated with CZP showed improvements in effectiveness outcomes over the 52-week study period. Improvements in disease activity, as measured by BASDAI, were observed in both r-axSpA and nr-axSpA patients. Both subpopulations also showed improvements in the signs and symptoms commonly associated with axSpA (pain, morning stiffness, fatigue, inflammation, and physical function), including through patient-reported outcome measures, such as PtGADA and total back pain. In addition, a low incidence of AEs and a low rate of drug withdrawal due to AEs were reported during the study. Compared with the overall CIMAX cohort, 6 23 Furthermore, diet plays a role in axSpA, since the disease is associated with gastrointestinal problems. 24 The Mediterranean diet is considered to have beneficial effects on gut microbiota, which may reduce disease activity. 25 However, as the study was not designed to compare outcomes between 169 TITLE patients from individual countries, limited conclusions can be drawn from these comparisons and further studies are needed to understand the observed differences between the Greek and overall population. Additionally, in this study observed case analysis TNFiexperienced patients had equal or numerically higher responses to CZP than TNFi-naïve ones. This is an unusual response pattern, as previous studies have found lower CZP response rates in TNFi-experienced patients compared to TNFi-naïve patients. 6 This is likely explained by the way missing data were handled and higher dropout rates in the TNFi-experienced patient subgroup than the TNFi-naïve subgroup. However, analysing the data using NRI demonstrates good levels of response across both patient subgroups. The CIMAX study was conducted at 12 clinical sites in Greece, providing the opportunity to evaluate outcomes in the Greek axSpA patient population in a real-world setting. Compared with a clinical trial, this aspect of the study design minimises issues associated with the generalisability of results to clinical practice. Another strength of CIMAX is the inclusion of patients with both r-axSpA and nr-axSpA, as there is currently limited information in the published literature about the Greek r-axSpA and nr-axSpA populations. The main limitations of this study are the lack of a comparator arm and the fact that this post-hoc analysis was not powered to formally evaluate outcomes in patients from Greece.

CONCLUSION
In conclusion, patients with r-axSpA and nr-axSpA that were treated with CZP in clinical practice in Greece showed improvements related to disease activity and key symptoms, as well as peripheral manifestations. This study illustrates the potential of CZP treatment to address the substantial health burden associated with axSpA in Greece and supports the use of this TNFi for the Greek axSpA population in clinical practice.