Objectives Evaluation of endothelial dysfunction, lipid metabolism, prevalence and development of cardiovascular diseases in patients with giant cell arteritis (GCA).
Methods 138 GCA patients and 100 controls were evaluated for prevalent cardiovascular diseases in 2012. Cholesterol, lipoproteins and triglycerides, intima–media thickness, arterial stiffness, asymmetric and symmetric dimethylarginine were also measured in 2012. Cardiovascular events, mortality and relapse were retrieved by chart review in 2020.
Results Prevalent carotid and vertebral artery disease was higher in GCA patients than in controls (p<0.001). GCA patients had higher levels of total cholesterol, low-density lipoprotein (LDL), intermediate-density lipoprotein, high-density lipoprotein, apolipoprotein A1 and B, and augmentation index (all with p<0.05). Target LDL levels were less frequently achieved at study inclusion by GCA patients (p=0.001), who developed more frequently new cardiovascular events, also with a higher amount, during follow-up (all with p<0.001). Statin treatment in GCA patients was associated with lower levels of asymmetric dimethylarginine, monocytes and C reactive protein (all with p<0.05). Relapse was independently associated with higher risk of future cardiovascular events (OR 5.01 (95% CI 1.55 to 16.22), p=0.007).
Conclusions GCA patients are at a high risk of developing cardiovascular diseases. Of relevance, there was underuse of statins and a large proportion of these patients showed LDL cholesterol concentrations above the treatment targets for high-risk patients. These data underscore the need for improvement of preventive strategies to reduce cardiovascular risk in GCA patients.
- Giant Cell Arteritis
- Cardiovascular Diseases
Data availability statement
All data relevant to the study are included in the article or uploaded as online supplemental information.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
WHAT IS ALREADY KNOWN ON THIS TOPIC
Patients with giant cell arteritis are at an increased risk for cardiovascular diseases while potential risk factors for cardiovascular diseases are scarce.
WHAT THIS STUDY ADDS
Prevalent cardiovascular diseases seem to be high in patients with giant cell arteritis who need an improved preventive strategy to reduce cardiovascular risk.
Cardiovascular diseases in patients with giant cell arteritis may be rather driven by changes of lipid metabolism than of endothelial dysfunction.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
This study indicates an underuse of statins and failure of target levels of low-density lipoprotein suggesting more focus on cardiovascular protection in patients with giant cell arteritis.
Giant cell arteritis (GCA) is a systemic vasculitis primarily of large-sized and medium-sized vessels affecting people aged 50 years or older. Both subtypes, cranial and extracranial GCA, may affect the aorta and its branches leading to typical ischaemic symptoms, such as visual loss or jaw claudication, while atypical ischaemic symptoms, such as intermittent claudication or stroke, may be also possible making prompt diagnosis and treatment initiation essential.1 2 Prevalence of large-vessel involvement, including large-artery stenosis and aortic aneurysm or dissection, seems to be high in GCA, estimated at about 30 events per 1000 person-years and a reported cumulative incidence of nearly 25% in 10 years.2 3 Additionally, large-vessel disease is also associated with higher mortality in GCA. Furthermore, GCA patients are at an increased risk of dying from large-vessel disease and cardiovascular disease (CVD).3–5 Moreover, patients with GCA are at higher risk of having CVD, including myocardial infarction (MCI), stroke and peripheral artery disease, compared with non-vasculitic controls.6 However, biomarkers and predictors of large-vessel disease in GCA are yet largely unknown. Especially, data on lipid parameters and endothelial dysfunction in GCA are very limited although both are important risk factors for atherosclerotic CVD.7–10
Although new diagnostic and therapeutic guidelines of GCA have been published by European Alliance of Associations for Rheumatology (EULAR) and American College of Rheumatology (ACR), their recommendations on specific evaluation and medical treatment of CVD in GCA are limited.11–13 Main factors for this limitation are that long-term studies evaluating cardiovascular outcome and predictors in GCA are scarce and partially divergent, for example, with regard to the use of statins or antiplatelet therapy.14–17 Therefore, the aim of this study was to investigate prevalence of CVD and to explore cardiovascular risk factors in GCA patients.
Materials and methods
Study design and patient cohorts
Patients with a diagnosed GCA between 1993 and 2010 were identified by electronic search and invited to participate in that study in 2012. At study inclusion between January and December 2012, blood sampling for parameters of lipid metabolism, endothelial dysfunction and inflammation, ultrasound and pulse-wave analysis measuring intima–media thickness (IMT) and arterial stiffness, and computed tomography (CT) evaluating aortic damage were investigated. After study inclusion, patients were followed up by clinical routine. Charts review was performed in 2020 retrieving patients’ clinical, radiologic and laboratory parameters prior to study inclusion and recording potential newly developed cardiovascular events, death and relapse of GCA after study inclusion.
Patients with GCA were diagnosed clinically by the treating angiological or rheumatological physician based on clinical parameters, laboratory data, imaging and/or biopsy. The modified ACR criteria proposed by Dejaco et al1 were fulfilled retrospectively in all GCA subjects, while only patients with a diagnosis of GCA of at least 2 years were included assuming that development of CVD, especially aortic damage, is a chronic process. Additionally, all measurements were performed in a phase of inactive GCA and no subject had a disease relapse within a period of at least 6 months prior to study visit. Exclusion criteria for GCA patients were active cancer, infections or other types of vasculitis. Control patients were identified via electronic search who underwent thoracic and abdominal CT scans for staging of inactive cancer or suspected pulmonary embolism in 2012. All controls were age matched to GCA patients with a range of ±3 years, while sex matching was not possible because the pool of controls was insufficient to match for two variables. Healthy individuals were not recruited as controls because we considered it as unethical to expose healthy individuals to the radiation of a CT scan in order to screen for aortic damage, given its low prevalence in the healthy population.18 Active cancer, infections and any autoimmune disease were exclusion criteria for the control group.
Fasting blood samples for evaluation of standard lipid parameters, including total cholesterol (TC), low-density lipoprotein (LDL), high-density lipoprotein (HDL), triglycerides, lipoprotein (a), apolipoprotein A1 and B, as well as of inflammatory parameters, including white cell count, C reactive protein (CRP), erythrocyte sedimentation rate (ESR) and fibrinogen, were obtained from each patient at study inclusion. Additionally, one tube of whole blood was collected at study inclusion and subsequently centrifuged at 4000 g for 10 min at 15°C temperature within 1 hour after blood sampling obtainment for the measurement of additional lipid parameters, asymmetric dimethylarginine (ADMA) and symmetric dimethylarginine (SDMA). The supernatant was collected and divided into aliquots of 1 mL, which were stored at −80°C until final analysis. Concentration of very-LDL (VLDL) and intermediate-density lipoprotein (IDL), size and particles of VLDL, LDL and HDL were measured by nuclear magnetic resonance analysis at Numares AG (Am Biopark, Regensburg, Germany) using the AXINON lipoFIT method with an Avance III HD nuclear magnetic resonance spectrometer (Bruker Optics, Manning Rd, Billerica, Massachusetts, USA), an Ascend 600 MHz magnet (Bruker) and TopSpin 3.2 (Bruker) as well as Axinon Suite V.188.8.131.52 (Numares) software.19 20 Lipid ratios, including monocytes/HDL, lymphocytes/HDL, neutrophils/HDL, triglycerides/HDL, TC/HDL, LDL/HDL and apolipoprotein B/A1 ratio as well as atherogenic coefficient and atherogenic index, were calculated post hoc since these ratios were associated with CVD.21 22 ADMA and SDMA were measured by high-performance liquid chromatography as described by Meinitzer et al.23
IMT and arterial stiffness
Details about measurements of IMT and arterial stiffness have been described previously.24 In brief, IMT of both common carotid, subclavian and common femoral arteries was measured by ultrasound using a linear transducer with 8–13 MHz (Siemens ACUSON S2000, Siemens Healthcare, Henkelstr., Erlangen, Germany) manually on magnified frozen longitudinal images. The presence of carotid IMT of ≥0.9 mm in any common carotid artery was defined as abnormal.10 25 After the measurement of IMT, carotid-femoral pulse-wave velocity (PWV) and augmentation index (Aix) were measured and calculated by automated analysis via photo-plethysmographic device Vascular Explorer (Enverdis, Fürstenwall, Düsseldorf, Germany) using software V.1.0. PWV>10 m/s was defined as pathologic.10
Details about aortic CT for the measurement of aortic diameters and investigation of aortic dissection have also been described previously.24 Briefly, thoracic and abdominal contrast-enhanced multidetector CT was performed and all measurements were traced manually from outside wall to outside wall of the aorta in anteroposterior planes. Diameters above the 90th percentile of the respective aortic region according to reported reference values with respect to age, gender and body surface area were defined as aortic dilatation/aneurysm.26
Charts review of CVD
Charts review from all participating subjects was performed between July and December 2020 via a fully electronic patient information system, called Medical Documentation and Communication network of Styria (MEDOCS), which is installed in the province of Styria, Austria, to provide electronic health data from all public Styrian hospitals and hospital alliances.27 Demographics, laboratory parameters and CVD prior to study inclusion as well as newly developed cardiovascular events, death and relapse of GCA during follow-up were recorded. Additionally, all patients’ medical reports and performed imaging methods prior and after study inclusion (ultrasound, CT, MRI, digital subtraction angiography), which were stored in MEDOCS, were screened for potentially described CVD. CVD and events were divided into coronary artery disease (CAD), carotid and vertebral artery disease (CVAD), upper extremity arterial disease (UEAD), renal artery disease (RAD), mesenteric artery disease (MAD) and lower extremity artery disease (LEAD) according to the European Society of Cardiology (ESC) guidelines 2017.28 Since those guidelines included only extracranial CVAD, we defined in our study CVAD as extracranial and intracranial CVAD, because GCA may also affect intracranial arteries, although to a lesser extent.29 Reported stenosis <50% of a vascular region was also defined as respective CVD. CAD was subdivided into asymptomatic CAD and MCI, CVAD was subdivided into asymptomatic CVAD and stroke or transient ischaemic attack, and RAD as well as MAD were subdivided into asymptomatic and symptomatic cases. UEAD and LEAD were subdivided into cases without symptoms, with intermittent claudication, with chronic limb-threatening ischemia, corresponding to Fontaine stages III and IV, and acute upper or lower limb ischaemia. Additionally, development of any Fontaine stage progression of previously known UEAD or LEAD, development of aortic rupture or dissection and requirement of aortic repair during follow-up were recorded as cardiovascular events. Total number of newly developed cardiovascular events per patient, occurrence of death, including cardiovascular death and occurrence of relapse in GCA patients during follow-up were recorded as further outcome parameters. Relapse was defined as major or minor relapse according to the EULAR recommendations for the management of large-vessel vasculitis.12 Finally, ESC SCORE2/SCORE2-OP was calculated post hoc based on clinical and laboratory data at study inclusion and the achievement of target LDL level according to recent guidelines was recorded.30–32 The end of the follow-up period was the patient’s last documented medical report in MEDOCS.
Categorical parameters were represented by frequency and percentages. Normally distributed parameters were expressed as means±SD and non-normally distributed parameters as median with IQR. Normality of distribution was examined by the Kolmogorov-Smirnov test and visual inspection. A χ2 test was used for categorical variables, two-sided t-test for normally distributed data and Mann-Whitney U test was used for non-normally distributed data. Pearson’s and Spearman‘s correlation coefficients were used for normally and for non-normally distributed variables, respectively. Simple and multiple logistic regression analyses were used to assess the association between lipid metabolism and parameters of endothelial dysfunction with cardiovascular events, death and relapse. Multiple regression analysis was adjusted for important confounding cardiovascular variables, including age, sex, active smoking, arterial hypertension and diabetes mellitus. Given an exploratory study character no adjustment for multiple testing was applied. Statistical significance was assumed for p values <0.05. Statistical analyses were executed via SPSS V.27.0.
Overall, 144 GCA patients and 115 controls were enrolled in this study, but measurement of lipid parameters, ADMA and SDMA was impossible in six patients with GCA and in 15 controls due to an insufficient collected blood volume. Therefore, 138 patients with GCA (106 female, 76.8%) with a mean age (±SD) of 74.5±7.7 years and 100 controls (64 female, 64.0%) with a mean age (±SD) of 73.9±7.8 years were included in this study. Number of statin intake did not differ between GCA patients and controls (p=0.778). Patients with GCA had higher prevalence of hypercholesterolaemia, but not of hypertriglyceridaemia (p<0.001; p=0.234, respectively). Prevalence of overall and asymptomatic CVAD was higher in GCA patients, but not of symptomatic CVAD (p<0.001; p<0.001; p=1.000, respectively). Higher prevalence of overall and asymptomatic CAD and RAD was observed in controls (p=0.002; p=0.037; p=0.016; p=0.044, respectively). There were no significant differences between GCA patients and controls regarding prevalent UEAD, LEAD or MAD. Significantly higher prevalence of dilation/aneurysm of the ascending and thoracic descending aorta were found in GCA patients at study inclusion (p=0.045; p=0.005, respectively), but not of the infrarenal abdominal aorta (p=0.076). Only the diameter of the thoracic descending aorta was significantly larger in GCA patients (p=0.027). Baseline characteristics are shown in table 1.
Cardiovascular events, mortality and relapse after follow-up
Mean follow-up (±SD) duration GCA patients was 87.1±21.7 months and 48.7±35.2 months in controls (p<0.001). Frequency of any newly developed cardiovascular event was significantly higher in GCA patients (p<0.001) who more likely developed any new cardiovascular event (OR 2.65 (95% CI 1.53 to 4.59), p<0.001). Thereof, higher frequencies of newly developed overall and asymptomatic CVAD (p=0.008; p<0.001, respectively), overall and asymptomatic UEAD (p=0.048; p=0.048, respectively), overall LEAD (p=0.035) and overall as well as asymptomatic RAD (p=0.016; p=0.026, respectively) were observed in GCA patients. Aortic repair was performed in one GCA patient (0.7%) during the follow-up without recorded aortic rupture or dissection in both cohorts. A higher number of cardiovascular events was observed during the follow-up in the GCA cohort (p<0.001) (table 2).
Fifteen patients with GCA (10.9%) and 51 controls (51.0%) died during the follow-up (p<0.001). Thereof, three patients with GCA (2.2%) and nine controls (9.0%) suffered a cardiovascular death (p=0.031). Any reported relapse of GCA occurred in 22 patients (15.9%) during the follow-up period. Five GCA patients suffered from major relapse (3.6%), 17 GCA patients suffered from minor relapse (12.3%) (table 2).
Lipids, endothelial dysfunction and inflammation
Patients with GCA had higher levels of TC, LDL, HDL, apolipoprotein A1 and B (all with p<0.001) and of IDL (p=0.025) than controls. Regarding lipid ratios, lower levels of monocytes/HDL, lymphocytes/HDL, neutrophils/HDL, triglycerides/HDL, TC/HDL, LDL/HDL, apolipoprotein B/A1 ratio and atherogenic index (all with p<0.05) with an elevated atherogenic coefficient (p<0.001) were found in GCA patients. Neither the size of VLDL, LDL or HDL nor levels of VLDL and of LDL particles differed between both cohorts. Total and small HDL particles were higher in GCA patients compared with controls (p<0.001; p=0.006, respectively). Risk categories of ESC SCORE2/SCORE2-OP did not differ between GCA patients and controls while the number of high-risk patients according ESC SCORE2/SCORE2-OP was high in both cohorts (75.0% vs 71.0%, p=0.552). Target LDL cholesterol attainment was poor in GCA patients (p=0.001) (table 3).
There was no significant difference of any measured IMT or PWV between GCA patients and controls. The frequencies of carotid IMT≥0.9 mm and of PWV>10 m/s did also not differ between both cohorts. Aix was higher while ADMA and SDMA levels were lower in GCA patients (all with p<0.001) (table 3).
Higher levels of lymphocytes were found in GCA patients (p=0.003), while higher levels of CRP, ESR and fibrinogen were observed in controls (p<0.001; p<0.001; p=0.008, respectively) (table 3).
Associations between lipid metabolism, endothelial dysfunction and outcome within GCA cohort
GCA patients with previously known CVAD significantly more often received statins than GCA patients without CVAD (OR 2.58 (95% CI 1.19 to 5.60), p=0.017). There were no significant differences for statin intake in GCA patients with or without CAD, UEAD, LEAD, RAD or MAD (p=0.133; p=0.734; p=0.690; p=0.605; p=0.999, respectively). GCA patients, who were classified as high-risk patients, have more commonly received statin treatment at study inclusion (OR 3.44 (95% CI 1.23 to 9.65), p=0.019). GCA patients receiving statins had significantly lower values of ADMA, monocytes and CRP than those without statins (p=0.034; p=0.030; p<0.001, respectively). Other parameters of endothelial dysfunction and inflammation did not differ between GCA patients with and without statin treatment (table 4).
Positive correlations between Aix, HDL and apolipoprotein A1 (p=0.014; p=0.005, respectively) and negative correlations between Aix, lymphocytes/HDL, neutrophils/HDL, triglycerides/HDL ratio and atherogenic index were observed (p=0.018; p=0.009; p=0.024; p=0.016, respectively). Femoral IMT correlated positively with apolipoprotein B/A1 ratio (p=0.007) and negatively with apolipoprotein A1 (p<0.05). PWV also correlated negatively with neutrophil/HDL ratio (p=0.02) (online supplemental figure 1).
In simple logistic regression analysis, associations were found between IDL and any newly developed cardiovascular event (p=0.013), small HDL particles and any newly developed relapse (p=0.019), and between SDMA and death (p=0.019). No associations were found between the development of any cardiovascular event, mortality or relapse and the achievement of target LDL level, LDL, HDL, HDL particles, apolipoprotein A1 and B, statin intake, classification as high-risk patient, IMT≥0.9 mm or PWV>10 m/s, Aix and ADMA (table 5). The development of any relapse during follow-up was significantly associated with the development of any cardiovascular event (p=0.005), but not with mortality (p=0.347). In multiple logistic regression analysis, IDL and development of any relapse were independent predictors for the development of any cardiovascular event (OR 0.28 (95% CI 0.09 to 0.85), p=0.025; OR 5.01 (95% CI 1.55 to 16.22), p=0.007, respectively). Small HDL particles and SDMA remained as independent predictors for any newly developed relapse and for mortality (OR 72.98 (95% CI 2.25 to 2372.11), p=0.016; OR 16.24 (95% CI 1.35 to 195.34), p=0.028, respectively).
GCA represents the most common vasculitis of the elderly with high incidences developing CVD.3–5 However, mediators of CVD, especially of atherosclerotic-driven cardiovascular changes, are yet scarcely investigated in GCA. This study observed that patients with GCA have higher prevalence rates of CVAD and aortic dilatations, and more frequently developed CVAD, UEAD, LEAD, RAD and total cardiovascular events than controls during a long-term follow-up period.
Several studies reported earlier that GCA may be associated with an increased risk for CVD including aortic dilatations and aneurysms.6 24 33–35 In the recent study, relative frequencies of each previously known and newly developed cardiovascular subdisease were even higher compared with prior data.6 34 On the other hand, some meta-analyses did not report any association between GCA and atherosclerotic CVD.36 37 One reason for this finding may be different methods for recording CVD. Although the cohorts for the recent study have been derived from a prior study by Jud et al,24 which had investigated the prevalence of aortic dilatation and occurrence of aortic rupture/dissection in GCA but no further CVDs and events, they revealed differences in patients’ characteristics. Due to a more detailed chart review of cardiovascular parameters in the recent study, including screening of all patients’ medical reports, laboratory data and performed imaging methods prior to study inclusion, significant higher rates for some parameters were observed compared with the former study, including the rate of hyperlipidaemia or CVAD. Furthermore, definitions of CVD and diagnostic methods for CVD may influence results. In this study, CVDs were defined according to the ESC guidelines while diagnostic methods vary, and subsequently also their diagnostic sensitivity and specificity. Moreover, the screening frequency may influence the detection rate of CVD further and may be also one reason for the high rate of CVAD in this study. While an ultrasonography of the carotid arteries was performed repeatedly in GCA patients during follow-up, controls received the same diagnostic less frequently. On the other hand, higher prevalence rates for CAD and RAD in controls may be explained by repeated whole-body staging CT for inactive cancer. Nevertheless, the use of patients, predominantly with inactive cancer, as controls may be reasonable since those patients exhibit also an increased risk for CVD.38 This underlines the significance of this study compared with other data and the finding that GCA was a predictor for cardiovascular events.
Hyperlipidaemia was a common finding in GCA patients, while the rate of statin intake did not differ between both cohorts so that bias by statins can be excluded. Several lipoproteins were elevated in GCA patients, who also exhibit a high cardiovascular risk profile according to the ESC-SCORE2/SCORE2-OP. Furthermore, target LDL was achieved in only 13.2% of GCA patients indicating that those patients are insufficiently treated and may benefit from more preventive action. On the other hand, GCA patients revealed also higher levels of HDL and HDL particles, which exhibit cardioprotective effects.39 40 Additionally, only sporadic and contrary associations of lipid parameters were found in regression analysis, as IDL seems to be protective for cardiovascular events and small HDL particles may be a predictor for relapse. Moreover, 95% CIs were very wide. These findings may be explained by a small absolute event number and a relatively small size of the cohort, and seems to be rather coincidental. Therefore, larger prospective studies and randomised-controlled trials are needed investigating lipid metabolism and the impact of lipid-lowering therapy in disease prevention in GCA. Nevertheless, our findings also indicate that the increased prevalence for CVD in GCA may be potentially driven by changes of lipid metabolism, while the impact of endothelial dysfunction may be less distinct. Only Aix, ADMA and SDMA revealed significant differences between both cohorts, while ADMA and SDMA were higher in controls. No other parameter of endothelial dysfunction revealed significance in regression analysis, except for SDMA, which, however, need to be considered cautiously as an independent mortality predictor in GCA, due to the same issue as described with IDL and HDL particles. Potential correlations between lipids and endothelial dysfunction in GCA were only revealed for Aix and some selected lipid parameters, but again with divergent results. Comparably, most inflammatory parameters and many inflammatory lipid ratios were higher in controls, which may be explained by paraneoplastic alterations of inflammatory and lipid parameters in patients with cancer.41 42 For this purpose, the use of inactive cancer patients as controls seems to be inadequate and healthy controls would be better suited. The fact that relapse was still an independent predictor for any cardiovascular event, suggests that inflammatory processes may play a relevant role for CVD in GCA. Furthermore, statin intake was associated with lower levels of ADMA, monocytes and CRP assuming a potential linkage between lipid metabolism, endothelial dysfunction and inflammation also in GCA.
Limitations of this study are the retrospective data collection of CVD, a relatively small study population compared with other observational studies investigating CVDs, an exploratory study character, and the missing systematic screening for all respective CVD at study inclusion and during follow-up. Unknown preexisting cardiovascular changes may be unintentionally attributed as newly developed events and newly developed events, which have been diagnosed at a resident doctor, were not documented in MEDOCS. Additionally, the use of controls, which revealed a significantly lower follow-up duration and consisted mainly of patients with inactive cancer instead of healthy controls was not ideal and may bias the results, especially for the evaluation of biomarkers. Furthermore, there were some significant differences between GCA patients and controls including sex, medications or previously known CVD, which may also bias the results, especially the development of cardiovascular events.
In conclusion, GCA patients have a high-risk cardiovascular profile and are at an increased risk for CVD, potentially due to insufficient LDL lowering treatment. Selected lipid parameters may contribute to CVD in GCA, while larger and prospective studies are needed to elucidate association between lipid metabolism and CVD.
Data availability statement
All data relevant to the study are included in the article or uploaded as online supplemental information.
Patient consent for publication
This study involves human participants and was approved by local ethics committee of the medical university of Graz (EK Nr. 32-469 ex 19/20). Participants gave informed consent to participate in the study before taking part.
We thank Lutz Niggl, Markus Fuhrmann, Katja Barthelmes, Claus Botzler and Julia Pommerenke for their assistance in the measurements of lipid parameters.
Contributors PJ contributed to data collection, analysis and interpretation of results, and draft manuscript preparation. FH contributed to study conception and design, and supervision. AM contributed to analysis and interpretation of results. MB contributed to supervision. CD contributed to analysis and interpretation of results. GS contributed to analysis and interpretation of results, and draft manuscript preparation. PJ acts as guarantor. All authors reviewed the results and approved the final version of the manuscript.
Funding This research was funded by Amgen, Vienna, Austria.
Competing interests None declared.
Patient and public involvement statement Patients were not involved in the research question, study design or outcome measurements.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.