There is accumulating evidence that inflammation is a key driver of atherosclerosis development and thrombotic complications. This pathophysiological mechanism explains, at least in part, the increased cardiovascular risk of patients with immune-mediated arthritis. Experimental and clinical studies have shown that tumour necrosis factor (TNF) plays a pathological role in both vascular and joint diseases, suggesting that TNF inhibitors (TNFis) may limit cardiovascular events in patients with rheumatoid arthritis (RA), psoriatic arthritis (PsA) or spondyloarthritis (SpA). This review summarizes studies exploring the effects of TNFis on cardiovascular outcomes in patients with RA, PsA or SpA. Clinical studies suggest that TNFis reduce vascular inflammation and may improve (or prevent worsening of) endothelial dysfunction and arterial stiffness. There is evidence that TNFis reduce the incidence of cardiovascular events in patients with inflammatory arthritis compared with non-biological treatments, particularly in patients with rheumatoid arthritis. Fewer studies have compared the effects of different classes of biological therapy on outcomes, but found no significant difference in the risk of cardiovascular events between patients taking TNFis and other biological therapy. In contrast, patients at high cardiovascular risk may derive greater benefit from a TNFi than from a Janus kinase inhibitor (JAKi). The cardiovascular impact of JAKis is still under debate, with a recent safety warning. Targeted control of inflammation is a key strategy to reduce the risk of major adverse cardiovascular events in patients with inflammatory arthritis. Cardiovascular evaluation and risk stratification, using a multidisciplinary approach involving rheumatology and cardiology teams, are recommended to guide optimal immunomodulatory treatment.
Inflammation, as a key driver of atherosclerosis development and thrombotic complications, might explain increased cardiovascular (CV) risk in patients with immune-mediated arthritis.
Tumour necrosis factor (TNF) plays a pathological role in both vascular and joint diseases, thus TNF inhibitors (TNFis), aimed at suppressing inflammation, may limit CV events in patients with immune-mediated arthritis.
This review summarizes studies exploring the effects of TNFis on CV outcomes in patients with rheumatoid arthritis, psoriatic arthritis or spondyloarthritis, suggesting that TNFis reduce vascular inflammation and may improve endothelial dysfunction and arterial stiffness.
There is evidence that TNFis reduce the incidence of CV events in patients with inflammatory arthritis compared with non-biological treatments, and that patients at high CV risk may derive greater benefit from a TNFi than from a Janus kinase inhibitor, although TNFis should not be used in patients with severe heart failure (left ventricular ejection fraction < 35%, New York Heart Association Class III or IV).
Targeted control of inflammation is a key strategy to reduce the risk of major adverse CV events in patients with inflammatory arthritis; CV evaluation and risk stratification are recommended to guide optimal immunomodulatory treatment.
Introduction
Over the past 25 years, accumulating evidence has demonstrated that immuno-inflammatory responses play a critical role in atherosclerosis pathophysiology, from the earliest development of a fatty streak to plaque growth and eventual thrombotic rupture [1]. In response to oxidised lipoprotein accumulation in the intima, innate immune cells, mainly monocytes/macrophages, are recruited and locally activated [1]. In addition, T and B cells participate in both vascular disease development and complications [2]. Immune cells produce chemokines and cytokines, including tumour necrosis factor (TNF) and interleukins (ILs), which are key drivers of vascular disease progression [1].
Anzeige
The pathophysiological link between inflammation and atherosclerosis may explain, at least in part, the significantly increased risk for cardiovascular (CV) diseases in patients with inflammatory diseases affecting the skin (i.e. psoriasis) [3], the gut (i.e. Crohn’s disease) [4], or the joints (i.e. rheumatoid arthritis [RA], psoriatic arthritis [PsA] and spondyloarthritis [SpA]) [5‐7].
In the context of arthritis, biological disease-modifying antirheumatic drugs (bDMARDs) are recommended in patients with uncontrolled disease despite therapy with conventional synthetic DMARDs (csDMARDs) [8‐11]. TNF inhibitors (TNFis) were the first bDMARDs available and are the most widely prescribed class of bDMARDs for inflammatory arthritis [12‐16].
There is accumulating evidence that TNFis ameliorate CV risk in patients with chronic inflammatory rheumatic disorders, and this has been the topic of several reviews over the years. However, few review articles have attempted to draw conclusions about the clinical implications of these data for practical application in the management of patients.
Large, prospective observational cohort studies recently highlighted that the impact of immunomodulatory treatment on CV outcomes differs between bDMARDs. For example, the ORAL Surveillance study reported an increased CV risk with the Janus kinase inhibitor (JAKi) tofacitinib when compared with TNFis [17]. Therefore, effective management of patients with inflammatory arthritis requires CV risk assessment and multidisciplinary collaboration, particularly between cardiologists and rheumatologists, to propose a more personalised and integrative treatment of these patients [18].
Anzeige
This structured review aims to summarise the effects of TNFis on CV outcomes in patients with RA, PsA or SpA, and to discuss the implications of these findings for clinical practice in light of recent data and the need for multidisciplinary care.
Methods
A literature search was undertaken of the PubMed database for articles published between 1 January 2003 and 7 May 2024 (last 15 years) using the following search terms: (rheumatoid arthritis OR psoriatic arthritis OR spondyloarthritis) AND (infliximab OR etanercept OR adalimumab OR certolizumab pegol OR golimumab) AND (major cardiovascular event OR cardiovascular event OR acute stroke OR myocardial ischemia OR myocardial infarction OR atherosclerosis cardiovascular disease OR stroke OR death).
Results were limited to clinical trials with control groups (active comparators, placebo or usual care) or observational research (cross-sectional, cohort or case–control studies), and excluded correspondence, editorials, comments, protocols, guidelines and reviews. Article titles were reviewed for relevance to the research question, and articles excluded if they did not present clinical data, were focused on treatments other than TNFis, reported results in children, were not published in English and were > 15 years old. In addition, the programmes of the American College of Rheumatology (ACR) and European Alliance of Associations for Rheumatology (EULAR, formerly the European League Against Rheumatism) conferences in 2022 and 2023 were searched for relevant abstracts, as was the ClinicalTrials.gov registry.
Additional ad hoc searches were also undertaken as needed to investigate a particular issue raised in the evidence review, and studies identified from the bibliographies of articles were also included as appropriate. The most relevant of these articles were selected for discussion in this review, based on study population (limited to studies with > 20 patients) and the focus of the study (relationship between TNFi use and CVD).
This article is based on previously conducted studies and does not contain any new studies with human participants or animals performed by any of the authors.
Results
Search Results
The PubMed search identified 192 articles of interest, of which 155 were potentially relevant based on their titles, and the congress search identified 19 potential abstracts of interest (total 174 articles); 30 were randomised controlled trials (RCTs), 41 registry studies, 71 observational studies (58 on atherosclerosis development), 13 pooled analyses and 19 conference abstracts. Literature selection is summarised in Figure S1 in the electronic supplementary material.
Effects of TNFis on Atherosclerosis
Inflammation and Atherosclerosis
The atherosclerotic process begins with the retention of lipoproteins, mostly low-density lipoprotein (LDL), in the arterial intima, where they undergo oxidative modification [1]. This stimulates endothelial activation and recruitment of monocytes/macrophages that can become foam cells after uptake of cholesterol from lipoproteins [19].
Adaptive immunity also participates in atherosclerosis development. CD4+ T cells are capable of differentiating into a range of T helper (Th) cells (which are pro-atherogenic) or regulatory (Treg) cells (which dampen down the inflammatory effect and are protective against atheroma development; Fig. 1) [1]. When progenitor Th cells encounter antigenic epitopes on lipoproteins, they acquire their final effector phenotype. Similarly, different types of B cells show pro- and anti-atherogenic effects. B2 cells aggravate atherosclerosis through Th1 cell polarisation, whereas immunoglobulin M-secreting B1 cells are atheroprotective [2].
Fig. 1
The types of T and B cells involved in atherosclerosis and the cytokines they secrete [1]. Reproduced from Kong et al. Signal Transduct Target Ther 2022;7(1):131 (https://doi.org/10.1038/s41392-022-00955-7) under a CC-BY 4.0 license (creativecommons.org/licenses/by/4.0). Breg regulatory B cells, Ig immunoglobulin, IL interleukin, IFN interferon, TGF transforming growth factor, Th T helper cells, TNF tumour necrosis factor, Treg regulatory T cells
×
The immune/inflammatory cells recruited into the artery wall communicate amongst themselves and with vascular cells through membrane contacts and secretion of cytokines [20]. In vivo study findings suggest that pro-inflammatory cytokines, including TNF, IL-1β and IL-6, promote atherogenesis, while anti-inflammatory cytokines, such as IL-10 and transforming growth factor-β, protect against atherosclerosis [21].
The actions of TNF are mediated by binding to TNF receptor 1 (TNFR1) or 2 (TNFR2) [22]. TNF exists in soluble and transmembrane-bound forms; the transmembrane form primarily interacts with TNFR2, although it may also activate TNFR1, whereas the soluble form only interacts with TNFR1 [22]. TNFR1 is present in most cells, while TNFR2 is present mainly in immune cells and the heart [23].
In animal models, TNF contributes to atherosclerosis by enhancing LDL uptake by endothelial cells, disrupting endothelial barrier functions, inhibiting nitric oxide (NO)-mediated vasodilation, promoting superoxide formation and vascular smooth muscle cell proliferation, promoting leukocyte adhesion to the endothelial surface, interfering with insulin signalling and contributing to endothelial cell apoptosis [23]. In addition, TNF-α promotes metalloproteinase production, which is involved in fibrous cap degradation and plaque vulnerability [20]. The deletion of the TNF gene reduced the development of atherosclerosis in several separate animal models, including Apoe–/– mice fed a high-fat diet [24], and C57BL/6 mice fed a high-fat diet containing cholate [25]. This may help to explain the ‘lipid paradox’ in patients with inflammatory conditions, whereby these patients develop CV disease at lower lipid levels than people without these conditions do [26].
Human atherosclerotic plaques contain TNF [27]. Increased plasma TNF levels are associated with recurrent major adverse CV events (MACE), i.e. myocardial infarction (MI) or CV death, in patients with acute coronary syndromes, even after adjustment for conventional CV risk factors [28]. In a genome-wide study of 30,912 people in Europe, genetically predicted TNF levels were associated with a more than twofold increase in the risk of coronary artery disease (CAD; odds ratio [OR] 2.25, 95% confidence interval [CI] 1.50–3.37) and stroke (OR 2.27, 95% CI 1.50–3.43) [29].
Effects of TNFis on Endothelial Function
Endothelial function is impaired in patients with inflammatory arthritis [30]. In addition, the risk of CV disease is correlated with endothelial dysfunction in patients with RA, but the effect appears to differ between macrovascular and microvascular beds.
Observational studies generally indicate that TNFis are associated with a short-term improvement in macrovascular endothelial function in patients with RA or ankylosing spondylitis (AS; as measured by brachial flow-mediated dilation [FMD]), which may or may not be sustained (see Table S1 in the electronic supplementary material for studies identified in our search) [31‐33]. However, a study in patients with PsA showed no change in FMD at 2 or 5 years after starting treatment with a TNFi [34].
A meta-analysis examining the effect of TNFis on endothelial function in patients with RA found a significant enough effect of these agents to reject the null hypothesis, but the standardised mean difference versus baseline was 0.987 (95% CI 0.64–1.33) and there was significant heterogeneity between studies, including in measurement technique, sample size, TNFis and concomitant treatments, patient characteristics (e.g. disease duration) and duration of follow-up [35].
Data on the effects of TNFis on microvascular endothelial function are mixed (see Table S1 in the electronic supplementary material) [36, 37]. In a study that included patients with RA, PsA or AS, the digital reactive hyperaemia index (RHI) was significantly improved versus baseline in patients receiving methotrexate monotherapy or TNFi ± methotrexate. The effect of treatment on RHI was more marked in patients receiving methotrexate monotherapy than in those receiving a TNFi ± methotrexate [36]. However, in this study, the TNFi was added to methotrexate in patients with a poor response, so the greater effect of methotrexate monotherapy on RHI may have been related to the fact that these patients were methotrexate-naïve, whereas the TNFi group were poor responders [36]. On the other hand, a separate study in patients with AS and high disease activity showed that 3 months of treatment with TNFis significantly improved microvascular flow parameters [38].
All of these studies on microvascular and macrovascular endothelial function were observational and conducted in a relatively small number of patients; therefore, further research in larger studies is required to determine the effect of TNFis on endothelial function in large and small vessels.
Effects of TNFis on Atherosclerosis Progression and Arterial Stiffness
Patients with inflammatory arthritis show accelerated progression of atherosclerosis based on measures of carotid intima-media thickness (IMT) [39]. Moreover, the presence and severity of carotid IMT is significantly correlated with the development of MACE in patients with RA [40].
Studies identified in our search that evaluated the effects of TNFis on carotid IMT reported varied results (see Table S1 in the electronic supplementary material). Three studies (all in patients with RA) showed improvement [41‐43], two studies (one in patients with AS and the other in patients with AS or RA) showed IMT remaining stable [33, 44], and two studies (both in patients with PsA) showed atherosclerosis progression [34, 45]. One of the studies in patients with PsA reported a slowing of atherosclerosis progression measured by total plaque area (rather than carotid IMT) in men (but not women) receiving TNFis compared with non-biological medications, after controlling for other risk factors [46]. A meta-analysis of studies in AS patients suggested that control of inflammation was the key factor in mitigating carotid IMT, with increased IMT in AS patients vs controls overall, but not in populations with a Bath Ankylosing Spondylitis Activity Index (BASDAI) of < 4 or with ≥ 50% of patients receiving TNFis [47].
Previous systematic reviews and meta-analyses report similar heterogeneity in results [48, 49]. One concluded that, on balance, there is more evidence in favour of a beneficial effect of TNFis on carotid IMT than there is against it [49], while a more recent (2023) meta-analysis concluded that carotid IMT remains stable during TNFi treatment, and does not worsen [48], suggesting a slowing of atherosclerosis progression.
There are some issues associated with evaluating carotid IMT, including that it is operator-dependent and techniques may vary between centres. Thus, future studies on the impact of TNFis on carotid IMT must include rigorous protocols and use experienced operators [30].
Generally consistent results have been seen in studies examining the effect of TNFis on measures of arterial stiffness, such as pulse wave velocity (PWV), Young’s elastic modulus (YEM) or augmentation index (AIx) [49]. Consistent with previous reports [49], the observational studies identified in our search found that PWV, YEM and AIx were generally improved or remained unchanged during treatment with TNFis in patients with RA, AS or PsA (see Table S1 in the electronic supplementary material) [33, 41, 44, 50‐52]. In one study, the aortic stiffness index did not change over 12 months in patients with RA receiving TNFis, but worsened in patients taking csDMARDs [53].
Effects of TNFis on Arterial Inflammation
Our search identified three studies which used 18F-fluorodeoxyglucose positron emission tomography/computed tomography (18F-FDG PET/CT) to investigate the effects of TNFis on vascular inflammation in large arteries, one observational study in patients with PsA [46] and one RCT and one observational study in patients with RA [54, 55].
The study in patients with PsA reported a significant reduction in vascular inflammation over 1 year in patients using TNFis (n = 21), but no change in those using non-biological treatments (n = 11), and the improvement was significant after adjustment for risk factors (p = 0.02) [46].
The RCT in patients with RA compared a strategy of adding a TNFi to methotrexate or sulfasalazine + hydroxychloroquine to methotrexate in patients who did not achieve disease control on methotrexate alone [55]. Each strategy resulted in a similar decrease in arterial inflammation after 24 weeks, and no relationship was noted between disease activity and the reduction in inflammation [55]. Consistent results were seen in the observational study in patients with RA [41]. That study also found that vascular inflammation decreased over 6 months of treatment to a similar extent in patients with RA starting csDMARD therapy and in those with established RA receiving adalimumab, suggesting that any anti-inflammatory treatment is likely to affect vascular inflammation [54].
Effects of TNFis on Cardiovascular Outcomes
Randomised Controlled Trials
Of the RCTs investigating CV event rates, 10 were in patients with RA [17, 56‐64], and two in patients with PsA [65, 66].
The 10 studies in patients with RA (Table 1) were comparisons of TNFis with csDMARDs (one study) [63], other TNFis (two studies) [60, 64], an IL-6 inhibitor (IL-6i; tocilizumab) (one study) [59], or JAKis (six studies) [17, 56‐58, 61, 62].
Table 1
Randomised comparative studies reporting cardiovascular endpoints in patients with rheumatoid arthritis receiving tumour necrosis factor inhibitors
Age ≥ 50 years, active RA despite treatment with MTX + ≥ 1 additional CV risk factor
Tofacitinib 5 mg bid (n = 1455) vs 10 mg bid (n = 1456) vs TNFi (adalimumab or etanercept; n = 1451)
5.5
HR for nonfatal MI with tofacitinib (combined doses) vs TNFis was 2.20 (95% CI 1.02–4.75), but HRs for any MI, fatal MI, any stroke, fatal or nonfatal stroke were not significantly different
HR for MACE with tofacitinib (combined doses) vs TNFis was greater in those with vs without a history of ASCVD (1.98, 95% CI 0.95–4.14 and 1.14, 95% CI 0.73–1.78, respectively)
Age ≥ 50 years, active RA despite treatment with MTX + ≥ 1 additional CV risk factor
Tofacitinib 5 mg bid (n = 1455) vs 10 mg bid (n = 1456) vs TNFi (adalimumab or etanercept; n = 1451)
5.5
Risk of MACE was higher in patients with residual disease activity (HDA, MDA and LDA) than those in remission across all groups, and the risk of MACE increased by 6% with each 5 mg/L increment in serum CRP
Age ≥ 50 years, active RA despite treatment with MTX + ≥ 1 additional CV risk factor
Tofacitinib 5 mg bid (n = 1455) vs 10 mg bid (n = 1456) vs TNFi (adalimumab or etanercept; n = 1451)
5.5
Risks of MACE, MI and all-cause mortality were significantly higher in patients receiving tofacitinib vs TNFi in high-risk patients (age ≥ 65 years or ever smokers) but not in low-risk patients (age < 65 years and never smokers)
AEs adverse events, ASCVD atherosclerotic cardiovascular disease, bid twice daily, CI confidence interval, CRP C-reactive protein, csDMARD(s) conventional synthetic disease-modifying antirheumatic drug(s), CV cardiovascular, HDA high disease activity, HHF hospitalisation for heart failure, HR hazard ratio, IL-6i interleukin-6 inhibitor, JAKi Janus kinase inhibitor, LDA low disease activity, MACE major adverse cardiovascular event, MDA moderate disease activity, MI myocardial infarction, MTX methotrexate, OL open-label, PY person-years, RA rheumatoid arthritis, TNFi tumour necrosis factor inhibitor
aAll placebo recipients and patients in the adalimumab group who did not achieve the efficacy endpoint were switched to upadacitinib at week 26, and non-responders in the upadacitinib group were switched to adalimumab
Key findings from these studies (Table 1) are that there was no difference in the rate of mortality or CV death between treatment with infliximab versus csDMARDs [63], and no difference in the rate of serious CV adverse events (AEs) in studies comparing one TNFi with another [60, 64].
Comparisons between TNFis and other targeted therapies are more mixed. There was no difference in the risk of MACE, MI, stroke, CV or all-cause mortality or hospitalisation for heart failure (HF) between a TNFi (etanercept) and an IL-6i (tocilizumab) in the ENTRACTE study [59], and no difference in the risk of MACE between a TNFi (adalimumab) and a JAKi (upadacitinib) in the SELECT-COMPARE study [57, 58]. On the other hand, in the ORAL Surveillance study, the risk of nonfatal MI was increased in patients receiving a JAKi (tofacitinib) compared with a TNFi (adalimumab or etanercept) [17, 56]. The difference in the risk of MACE was more marked in high-risk patients and in those with a history of atherosclerotic CV disease (ASCVD) [56, 62, 67‐69].
Interestingly, the incidence rate of MACE in the TNFi arm of the ENTRACTE study was 1.70 (95% CI 1.35–2.10) per 100 person-years (PY) [59], whereas it was 0.73 (95% CI 0.52–1.01) per 100 PY in the ORAL Surveillance study [17]. The reason for such a difference is unclear since the two studies had similar patient populations, follow-up duration and definitions for MACE (see Table S2 in the electronic supplementary material).
It is notable that, in the ORAL Surveillance study, the presence of residual disease activity was associated with an increased risk of MACE, irrespective of whether patients received a JAKi or a TNFi [61], reinforcing the relationship between inflammatory activity and CV events and highlighting the need to achieve remission in patients with RA to minimise the risk.
Pooled Analyses
A number of pooled analyses have examined the effects of TNFis on CV outcomes of patients with inflammatory arthritis [70‐84].
RA
Preliminary data from a pooled analysis of 19 RCTs reported a slightly and non-significantly higher incidence of MACE in patients receiving tocilizumab or JAKis compared with TNFis, with an OR of 1.077 (95% CI 0.79–1.44; p = 0.673) for tocilizumab vs 1.27 for TNFis (95% CI 0.90–1.80; p = 0.177) [83].
The other pooled analyses examined event rates with individual TNFis: all CV events with etanercept (one study) [71] and MACE with certolizumab pegol (two studies) [73, 78]. The analysis of etanercept pooled data from the open-label extension studies was conducted in North America in patients with early (n = 558) or longstanding (n = 714) RA, following patients for up to 10 years [71]. During etanercept treatment, CV events occurred in 6.8% of patients with early RA and 8.8% of patients with longstanding RA; MI was the most common CV event, occurring in 2.9% of patients with early RA and 3.4% with longstanding RA [71]. In the analysis of certolizumab pegol data, certolizumab pegol was associated with a numerically slightly higher incidence of MACE compared with placebo (0.71 [95% CI 0.55–0.99] vs 0.54 [95% CI 0.07–1.94] MACE per 100 PY) [73]. The mean exposure duration was 152 days in the certolizumab pegol group and 110 days in the placebo group, with a total exposure duration of 1302 PY and 373 PY, respectively [73].
SpA
There are limited observational data on MACE incidence specifically in SpA patients. However, data from the Veterans Affairs database in the US presented as a conference abstract suggested that early TNFi initiation (i.e. within 12 months of diagnosis) was associated with an increased risk of MACE (HR 1.22, 95% CI 1.03–1.45) compared with no TNFi initiation [84]. However, the authors acknowledge that the analysis may be influenced by confounding [84].
PsA
Two pooled analyses examined the effect of TNFis on CV events specifically in patients with PsA or plaque psoriasis [76, 77]. Champs and colleagues examined the effect of biological agents on CV events during the early placebo-controlled phases of the randomised comparative studies, and found that these were low and not significantly different versus placebo (Fig. 2). However, the authors acknowledge that the placebo-controlled phases of these studies were probably too short to detect any major differences in CV event rates [77].
Fig. 2
The difference in the risk of a major adverse cardiovascular event per person-year during treatment with different biologicals versus placebo during randomised controlled trials in patients with psoriatic arthritis or plaque psoriasis (from a meta-analysis of 77 randomised controlled trials) [77]. CI confidence interval, IL-12i interleukin-12 inhibitor, IL-17i interleukin-17 inhibitor, IL-23i interleukin-23 inhibitor, PY person-year, TNFi tumour necrosis factor inhibitor
×
Yang and colleagues analysed CV event rates in five trials comparing TNFis with either topical/phototherapy or methotrexate in patients with psoriasis with or without PsA, and reported a significantly reduced risk with TNFis overall (relative risk [RR] 0.60, 95% CI 0.48–0.74) versus topical/phototherapy (RR 0.58, 95% CI 0.43–0.77) and versus methotrexate (RR 0.67, 95% CI 0.52–0.88) [76]. Similar results were seen with the risk of MI, in which the overall RR was 0.74 (95% CI 0.60–0.84), RR versus topical/phototherapy was 0.73 (95% CI 0.59–0.90), and RR versus methotrexate was 0.65 (95% CI 0.48–0.89) [76].
Mixed Populations
The ATTACH study found an increased risk of HF-related hospitalisation and death among patients with moderate-to-severe HF receiving infliximab [85], so TNFis are contraindicated in patients with New York Heart Association class III or IV HF. In response to these concerns, two pooled analyses examined the rate of congestive HF in global clinical trials with adalimumab [70, 79], with the more recent analysis (in 2020) including data from 77 clinical trials across a range of indications (33 in RA, three in PsA and eight in SpA) [79]. Over 56,916 PY of adalimumab exposure, congestive HF developed at an incidence rate of 0.2 per 100 PY overall (n = 29,967) in patients with RA (n = 15,512), 0.1 per 100 PY in patients with non-radiographic axial SpA (n = 863), < 0.1 per 100 PY in those with AS and 0 per 100 PY in patients with PsA or peripheral SpA [79]. These data indicate that the risk associated with adalimumab is comparable with that reported with other TNFis.
An analysis of event rates among patients with a range of inflammatory arthritis disorders (n = 2303) receiving golimumab reported an incidence that was numerically slightly lower with the TNFi than with placebo. The incidence of MACE (MI, stroke or sudden cardiac death) was 0.25 (95% CI 0.07–0.63) or 0.48 (95% CI 0.23–0.89) per 100 PY with golimumab 50 or 100 mg, respectively, compared with 0.59 (95% CI 0.07–2.13) per 100 PY with placebo [72]. Duration of exposure was not reported [72]. Pooled analyses in patients with a range of indications suggest that the incidence of MACE with certolizumab pegol or golimumab tends to be higher in patients with RA than in those with PsA or SpA [78, 81].
Safety follow-up data from clinical trials in patients with RA, PsA or AS reported that six of the 2226 patients receiving golimumab had died from cerebrovascular/CV causes after 3 years of follow-up [74], and seven of 2228 after 5 years of follow-up [75].
Observational Studies
Multiple observational studies prior to 2003 (our search start date) suggest that treatment with TNFis reduces the risk of CV disease in patients with RA compared with other forms of treatment [86]. Observational and registry data published since 2003 support these findings by showing at least no increase in the risk of CV events or CV death with use of TNFis [87‐90], or a decrease in the risk of these events relative to non-use or csDMARD use in patients with rheumatoid arthritis [91‐96]. The magnitude of the risk reduction increases with duration of TNFi use [94].
However, the risk of acute coronary events remains higher among patients with rheumatoid arthritis receiving TNFis than it is in the general population [97]. In these studies, the crude incidence rate for MACE during use of any TNFi ranged from 5.8 to 24.9 per 1000 PY (Fig. 3A) [98‐102] for acute MI or acute coronary syndrome from 2.9 to 14 per 1000 PY (Fig. 3B) [87, 92, 93, 97, 99, 100, 103, 104], and for stroke from 1.7 to 7.3 per 1000 PY (Fig. 3C) [88, 99, 100]. Crude incidence rates for these endpoints tended to be a little higher with infliximab than with other TNFis (Fig. 3A–C). In registry studies, there was no significant difference in the rate of serious CV events between patients taking TNFis or JAKis [82, 89, 98, 99, 102].
Fig. 3
Crude incidence rate of events per 1000 person-years during treatment with biologicals, Janus kinase inhibitors or conventional disease-modifying antirheumatic drugs in observational studies in patients with rheumatoid arthritis. (A) Major adverse cardiovascular events; (B) myocardial infarction or acute coronary syndrome; and (C) stroke [87, 88, 92, 93, 97‐100, 103]. ACS acute coronary syndrome, ADA adalimumab, csDMARD conventional synthetic disease-modifying antirheumatic drug, CTZ certolizumab pegol, ETN etanercept, GOL golimumab, IFX infliximab, JAKi Janus kinase inhibitor, MACE major adverse cardiovascular events, MI myocardial infarction, PY person-years, TNFi tumour necrosis factor inhibitor
×
There are comparatively fewer observational data on the impact of TNFis on CV outcomes in patients with PsA or SpA. Overall, individuals with PsA or SpA tend to have a lower risk of CV events than patients with RA [105], which makes demonstration of a treatment effect more difficult in these patient groups. The Australian Rheumatology Association Database of patients with RA, PsA or SpA showed a significantly reduced risk of CV events in patients who were taking TNFis compared with biological-naïve patients (hazard ratio [HR] 0.85, 95% CI 0.76–0.95; p = 0.006) after adjustment for other risk factors [91]. However, a case–control study in patients with PsA or AS found no significant association between the type of treatment a patient was receiving (csDMARD or TNFi) and the risk of MI [106]. A separate retrospective cohort study identified a significantly lower risk of MI among psoriasis patients taking TNFis compared with those not taking these agents (HR 0.26, 95% CI 0.12–0.56; p = 0.0005), but not in patients with PsA (HR 0.86, 95% CI 0.28–2.70; p = 0.80) nor in the combined group of PsA + plaque psoriasis patients (HR 0.76, 95% CI 0.47–1.24; p = 0.27) [107]. In a separate report, the same researchers found the risk of MI significantly decreased with etanercept among patients with PsA or plaque psoriasis (HR 0.53, 95% CI 0.31–0.92); the reduction in the risk of MI with other TNFis was similar in magnitude but did not reach statistical significance (HR 0.53, 95% CI 0.27–1.06) [108].
A clearer association between CV event incidence and TNFi therapy is seen in patients with SpA. A cross-sectional study found a lower risk of ischaemic heart disease in AS patients taking TNFis than in those not taking these agents [109]. Similarly, a retrospective cohort study in 450 patients with axial SpA indicated a lower incidence of CVD in those exposed to TNFis compared with those not exposed (2.6% vs 6.5%; p = 0.046), despite a longer follow-up period in the TNFi group [110]. A significantly lower risk of CV disease in the TNFi-exposed group was seen in a univariate analysis (HR 0.35, 95% CI 0.13–0.93; p = 0.034), and in multivariate analysis after adjustment for traditional CV risk factors (HR 0.30, 95% CI 0.10–0.85; p = 0.024) [110]. Similarly, a retrospective analysis of event rates among 2616 patients with SpA reported a significantly lower risk of MACE (HR 0.37, 95% CI 0.17–0.80; p = 0.01) and cerebrovascular events (HR 0.21, 95% CI 0.06–0.78, p = 0.02) in those taking TNFis compared with those not taking TNFis [111]. The largest analysis included 22,929 patients with SpA enrolled in the French National Health Insurance database [112]. Over a mean of 6.4 years of follow-up, TNFi use was associated with a significantly reduced cumulative incidence of MACE (HR 0.59, 95% CI 0.45–0.78, p < 0.001) [112].
Observational data do not necessarily bear out the results of the ORAL SURVEILLANCE study regarding the lower risk of MACE with TNFis versus JAKis. In the German RABBIT register, patients at high CV risk like those in ORAL SURVEILLANCE developed MACE at an incidence rate of 1.03 (95% CI 0.74–1.40) per 100 PY in the group receiving TNFis or 0.92 (95% CI 0.62–1.33) per 100 PY in those receiving JAK inhibitors [89]. The highest MACE incidence rates were seen in patients receiving csDMARDs (1.48 [95% CI 1.05–2.01] per 100 PY) followed by those receiving bDMARDs other than TNFis (1.17 [95% CI 0.80–1.65] per 100 PY) [89]. The RELATION study of the French national healthcare database also reported a similar rate of MACE in patients receiving a JAKi (tofacitinib) versus TNFis (HR 1.06, 95% CI 0.68–1.64; p = 0.8127) [102], as did a case–control study using US insurance claims data (adjusted OR for JAKis vs TNFis of 0.86, 95% CI 0.37–1.99) [113]. On the other hand, a preliminary report from the BIOBADASAR 3.0 registry suggested that patients with RA or PsA who develop MACE during treatment do so earlier if they are taking JAKis compared with TNFis (median MACE-free interval of 0.5 years in the JAKi group versus 1.8 years in the TNFi group) [82].
Implications for Practice
There is no doubt that the presence of inflammatory arthritis increases the risk of CV disease; RA is equivalent to diabetes in terms of its contribution to CV risk [114]. Therefore, it is important for patients with such conditions to be carefully assessed and treated to minimise this risk, according to EULAR recommendations [115].
Control Disease Activity
The magnitude of the CV risk in patients with inflammatory arthritis is related to the extent of systemic inflammation [26, 116, 117]. Therefore, treatment to limit disease activity, as advocated by current guidelines [8‐11, 18, 115, 118], is likely to have CV benefits as well as disease control and symptom relief.
Anzeige
Active management to achieve remission, or at least low disease activity, is best practice and will require treatment with a targeted therapy for many patients. Among the available biological agents, TNFis are associated with a reduction in the risk of CV events compared with non-use of these agents in patients with RA, PsA and AS [91], although there is considerably more evidence of benefit in patients with RA [91‐94, 119].
Data suggest that there is no difference in the risk of CV events between patients taking TNFis and other types of biological therapy, including IL-6is [57‐59, 98‐100, 120, 121]. In contrast, based on data from the ORAL Surveillance study, patients at high risk of CV events may derive greater benefit from a TNFi than a JAKi [62]. Despite some controversies and based on the results of the ATTACH trial described above, TNFis should not be used in patients with severe heart failure (left ventricular ejection fraction < 35%, New York Heart Association Class III or IV) [85].
Risk Assessment
Careful CV risk assessment is important in patients with inflammatory arthritis, with guidelines recommending risk assessment at least once every 5 years, and after major changes in inflammatory therapy [115]. Screening for multimorbidity is feasible in clinical practice [122], and improves patient management [123]. However, CV risk assessment information tends to be inconsistently measured and recorded. In one UK study, only 3% of patients with RA had a CV risk score recorded in their primary care medical record within 12 months of diagnosis [124].
A range of risk assessment tools are available online (Table 2) [125], and can help guide decisions for risk factor management. The Systematic COronary Risk Evaluation (SCORE) tool is widely used [126], and can be adjusted to account for the patient’s underlying inflammatory disease [115]. Alternatively, the Joint British Societies 3 (JBS3) and QRISK2 tools include RA as a risk factor, and do not need adjustment [127, 128].
Table 2
Cardiovascular risk assessment tools available online for healthy individuals
ACC American College of Cardiology, AHA American Heart Association, ASCVD atherosclerotic cardiovascular disease, CVD cardiovascular disease, ESC European Society of Cardiology, JBS3 Joint British Societies 3, RA rheumatoid arthritis, SCORE Systematic COronary Risk Evaluation, SLE systemic lupus erythematosus, UK United Kingdom, USA United States of America
Anzeige
Risk Factor Management
Risk factor management is a key element of reducing the risk of CV events in patients with inflammatory arthritis, as hypertension, hyperlipidaemia, diabetes and overweight/obesity contribute significantly to the risk in these groups [91, 109]. EULAR guidelines recommend approaching CV risk management in patients with inflammatory disease as per recommendations in the general population [115].
Blood Pressure
While some studies suggest that TNFis increase the risk of incident hypertension [3, 74, 75], others do not [129‐131]. Moreover, the risk of hypertension development may be related to disease activity, since high levels of rheumatoid factor are an independent predictor of hypertension development [131].
Non-steroidal anti-inflammatory drugs (NSAIDs) and corticosteroids may raise BP and increase the incidence of hypertension [132, 133], and NSAIDs may attenuate the effect of antihypertensive therapy [134], so the use of NSAIDs and corticosteroids should be avoided or minimised wherever possible.
Lipids
CV disease develops at lower lipid levels in patients with inflammatory arthritis than it does in the general population [26]. In general, control of inflammation has benefits for the lipid profile [115]. TNFis are associated with changes in lipid levels that are largely favourable [135], increasing levels of high density lipoprotein (HDL) cholesterol [135], improving the function of HDL cholesterol [136] and reducing levels of atherogenic very-LDL particles [137]. While levels of LDL and total cholesterol may increase during treatment with TNFis, the proatherogenic ratios of total/HDL cholesterol or LDL/HDL cholesterol decrease overall as a result of the effect on HDL [135]. Treatment with IL-6is or JAKis are also associated with increases in LDL and HDL cholesterol, but without a favourable overall effect on these ratios [138, 139].
When monitoring lipid levels in patients receiving TNFis, physicians need to consider the ratios of LDL/HDL or total/HDL cholesterol, and initiate lipid-lowering therapy in response to these ratios, rather than in response to total or LDL cholesterol levels in isolation.
Diabetes and Blood Glucose
Inflammatory arthritis is associated with insulin resistance [140], but treatment with biological agents reduces insulin resistance [141, 142]. In patients with RA, improvements in insulin resistance during treatment with TNFis parallel decreases in inflammatory markers, such as C-reactive protein or erythrocyte sedimentation rate (ESR) [143].
Data indicate that TNFis are unlikely to negatively impact glycaemic control, and may even improve it, in patients with diabetes or prediabetes [140, 144].
Risk Management After MACE
Not all CV events can be prevented in patients with inflammatory disease, so a key clinical question is whether or not to modify treatments when a patient experiences a MACE, especially since these patients are now at high risk of a subsequent event [145]. An analysis of data in patients with RA, PsA or psoriasis indicated that patients with RA were at the highest risk of a subsequent event [105]. In this analysis, most patients persisted with the same therapy after a CV event, and the rate of drug retention was highest for TNFis (72.8%), intermediate for csDMARDs (69.9%) and lowest for non-TNFi biologicals (56.0%) [105]. The incidence of subsequent events was lowest in the group receiving TNFis (crude incidence rate 75.2 [95% CI 54.4–96.0] per 1000 PY), intermediate in those receiving csDMARDs (83.6 [95% CI 53.3–113.9] per 1000 PY), and highest in those receiving non-TNFi biologicals (122.4 [95% CI 60.6–184.3] per 1000 PY), although there was no significant between-group difference in the risk of a subsequent event after adjustment for other risk factors [105].
Ideally, a patient with inflammatory arthritis who experiences a serious CV event should receive multidisciplinary care, involving the rheumatology, cardiology and primary care teams, with careful consideration of future risk factor management as well as effective control of their inflammatory condition.
Conclusions
The weight of evidence indicates that treatment with TNFis to achieve remission or at least low disease activity is likely to minimise the risk of a serious CV event in patients with RA, PsA or SpA. Suppression of inflammation is the key mechanism for this benefit, mediating an anti-atherogenic effect, so for these patients, it is necessary to achieve maximal control of inflammatory disease. Other cornerstones for minimising CV risk are regular and appropriate CV risk assessment, effective management of all modifiable risk factors, use of treatments with the best CV benefit/risk balance (avoiding NSAIDs, reducing or even stopping corticosteroid therapy and using treatments with proven CV benefits), and effective collaboration between all relevant healthcare providers.
Anzeige
Acknowledgements
Medical Writing, Editorial, and Other Assistance
We would like to thank Catherine Rees who wrote the outline and subsequent drafts of this manuscript on behalf of Springer Healthcare. This medical writing assistance was funded by Celltrion Healthcare.
Declarations
Conflict of interest
Jérôme Avouac has received honoraria from Galapagos, Lilly, Pfizer, Abbvie, Bristol-Myers Squibb, Sanofi, Roche-Chugai, Nordic Pharma, Medac, Novartis, Biogen, Fresenius Kabi, Janssen, Celltrion and MSD; and research grants from Bristol-Myers Squibb, Pfizer (Passerelle), Novartis (Dreamer), Fresenius Kabi, Galapagos and Nordic Pharma. Hafid Ait-Oufella is the co-founder of POLYGON Therapeutics, a biotech company that is developing therapeutic monoclonal antibodies to limit CV disease burden. Caroline Habauzit and Salim Benkhalifa are employees of Celltrion Healthcare. Bernard Combe has received honoraria from Abbvie, Celltrion, Chugaï, Lilly, Nordic and Pfizer.
Ethical Approval
This article is based on previously conducted studies and does not contain any new studies with human participants or animals performed by any of the authors.
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License, which permits any non-commercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc/4.0/.
Bestimmte mutierte Blutzellen erhöhen nicht nur das Risiko für hämatologische Neoplasien, sondern sind offenbar auch bei soliden Tumoren prognostisch ungünstig. Dafür sprechen aktuelle Erkenntnisse, insbesondere beim nichtkleinzelligen Lungenkarzinom.
Ergebnisse eines kürzlich publizierten „natürlichen Experiments“ in Wales legten nahe, dass eine Herpes-Zoster-Impfung das Demenzrisiko senkt. Jetzt hat das Studienteam ähnliche Daten aus Australien publiziert, die in die gleiche Richtung zeigen. Offene Fragen bleiben allerdings so oder so.
Immer mehr Menschen überleben kritische Erkrankungen. Aber Beatmung, Sedierung und die Eindrücke der Intensivstation hinterlassen Spuren. Das Post-Intensive-Care-Syndrom kann die Folge sein. Es ist nicht nur eine Herausforderung für Kliniken, sondern auch Hausarztpraxen. Mit Allgemeinmediziner Prof. Dr. med. Konrad Schmidt sprechen wir in dieser Folge darüber, wie die Überlebenden wieder ins Leben finden können.
In diesem CME-Kurs können Sie Ihr Wissen zur EKG-Befundung anhand von zwölf Video-Tutorials auffrischen und 10 CME-Punkte sammeln. Praxisnah, relevant und mit vielen Tipps & Tricks vom Profi.
