Shared inflammatory pathways of rheumatoid arthritis and atherosclerotic cardiovascular disease

Abstract

The association between chronic inflammation and increased risk of cardiovascular disease in rheumatoid arthritis (RA) is well established. In the general population, inflammation is an established independent risk factor for cardiovascular disease, and much interest is placed on controlling inflammation to reduce cardiovascular events. As inflammation encompasses numerous pathways, the development of targeted therapies in RA provides an opportunity to understand the downstream effect of inhibiting specific pathways on cardiovascular risk. Data from these studies can inform cardiovascular risk management in patients with RA, and in the general population. This Review focuses on pro-inflammatory pathways targeted by existing therapies in RA and with mechanistic data from the general population on cardiovascular risk. Specifically, the discussions include the IL-1, IL-6 and TNF pathways, as well as the Janus kinase (JAK)–signal transducer and activator of transcription (STAT) signalling pathway, and the role of these pathways in RA pathogenesis in the joint alongside the development of atherosclerotic cardiovascular disease. Overall, some robust data support inhibition of IL-1 and IL-6 in decreasing the risk of cardiovascular disease, with growing data supporting IL-6 inhibition in both patients with RA and the general population to reduce the risk of cardiovascular disease.

Introduction

Targeting inflammation is an established and effective strategy for the treatment of rheumatoid arthritis (RA) that has been in place since the development of the first biologic therapy in the late 1990s^1-3. Historically, the treatment of atherosclerotic cardiovascular disease (ASCVD) has focused on the management of traditional risk factors such as arterial hypertension and hypercholesterolaemia^4,5. However, large-scale randomized controlled trials in the past few years have now established inflammation as a modifiable risk factor to be considered in the prevention of the development and progression of atherosclerosis in the general population^4,6,7. With the increased adoption of targeted biologic therapies to reduce the risk of ASCVD, summarizing and examining these pathways and their roles in the pathogenesis of RA and ASCVD have become increasingly important for the management of both RA and ASCVD.

RA can be considered a human model of inflammation and can inform the impact of chronically increased levels of inflammation in the general population. Indeed, studies on RA have linked inflammation to an 1.5× increased risk of cardiovascular morbidity and mortality^8-11. Evidence is also available showing direct effects of inflammation on the vascular endothelium as well as interactions between systemic inflammation with classic cardiovascular risk factors, further contributing to excess cardiovascular risk^12,13. The inflammatory pathophysiology of RA shares important links with that of atherogenesis and atherothrombosis, with inflammation promoting the progression of atherosclerosis^14-16, leading to a range of cardiovascular outcomes such as venous thromboembolism, heart failure and ASCVD; the latter, ASCVD, is the focus of this Review. From studies of targeted therapies in RA, evidence suggests that targeting certain pathways might have a more beneficial effect on cardiovascular risk than others^17-19. Thus, recent studies in RA in the past few years have informed the cardiovascular literature. In turn, large randomized controlled trials (RCTs) in the general population can help to inform the management of ASCVD in patients with RA, which is otherwise a major challenge in RA owing to the difficulty in achieving the sample size needed for a sufficient number of hard cardiovascular endpoints in RA-only RCTs.

As the convergence of drug targets for RA and cardiovascular disease is a relatively new development, limited reviews are available that summarize the shared pathways and downstream effects in one resource. Thus, in this Review, we seek to address this gap and will focus on shared pathways for existing targeted therapies in RA where mechanistic data are available from both the RA and ASCVD literature, specifically, the IL-1, IL-6, and TNF and JAK–STAT pathways, where emerging data are available²⁰. A more detailed understanding of shared pathways between RA and cardiovascular disease can better inform the management of RA for individuals with or at higher risk of ASCVD, as well as the management of inflammation for individuals in the general population with an increased cardiovascular risk owing to inflammation. In this Review, we provide a comparative overview of the pathways shared by RA and ASCVD and the known mechanistic pathways for the pathogenesis of both conditions using data from selected mechanistic, clinical trials and observational cohort studies. Last, we highlight areas for future research that might inform treatment strategies that can control disease activity in RA and mitigate the risk of ASCVD.

The IL-1 pathway

IL-1 was first characterized in the 1970s as an endogenous pyrogen that induced fevers and a mononuclear cell factor involved in tissue destruction in RA²¹. Eventually, examinations of synovial explants from patients with RA led to the characterization of IL-1 (refs. ^21,22). Although the IL-1 family consists of 11 members, IL-1 typically refers to IL-1α and IL-1β (which are both pro-inflammatory cytokines involved in inflammasome signalling²³). The IL-1 receptor antagonist (IL-1Rα) is an endogenous inhibitor that blocks the actions of IL-1α and IL-1β.

IL-1 in RA

Mice deficient in IL-1Rα develop spontaneous erosive arthritis²⁴. Anakinra, an analogue of IL-1Rα, is approved for use in RA; however, this drug is less frequently used than other biologic therapies owing to its limited efficacy^25,26. The limited efficacy has been attributed in part to a bypass of IL-1-dependent signalling through the adaptor protein MyD88, providing a mechanism for ongoing synovial Toll-like receptor (TLR) signalling²⁷. IL-1 can stimulate receptor activator of nuclear factor kappa-B ligand (RANKL), linked to osteoclast activation, and certain matrix metalloproteinases (MMPs, including MMP1, MMP3 and MMP13), which are both involved in bone and cartilage destruction^28-31 (Fig. 1a). In patients with RA, elevated plasma concentrations of IL-1 correlate with higher disease activity^32,33.

Fig. 1∣. Shared inflammatory pathways of rheumatoid arthritis and atherosclerosis.

a, Various inflammatory pathways, including the IL-1, IL-6 and TNF pathways and the downstream Janus kinase (JAK)- signal transducer and activator of transcription (STAT) signalling pathway, have a role in both rheumatoid arthritis (RA) and atherosclerotic cardiovascular disease (ASCVD). The figure highlights these shared inflammatory pathways and the downstream effects of these pathways on joint damage and coronary artery atherogenesis. b, Various treatments are available that target these shared pathways and are used in the treatment of RA (shown on the left) or have either been shown in clinical trials to reduce the risk of ASCVD or are under evaluation for reducing ASCVD risk in the general population (shown on the right).

IL-1 in ASCVD

The inflammasome, and specifically the IL-1β pathway, has multiple effects in many stages of atherosclerosis and ASCVD³⁴. IL-1β induces the expression of adhesion factors, such as intercellular adhesion molecule (ICAM-1) and vascular cell adhesion molecule (VCAM-1), and chemokines, including monocyte chemoattractant protein (MCP1), on endothelial cells and promotes inflammatory cell recruitment into the vasculature via MCP1 (refs. ^35,36). Enhanced expression of IL-1 itself can form a positive feedback loop and generate the production of IL-6 and MMPs, upon which IL-6 activates the acute phase response³⁷. Certain matrix metalloproteins (MMPs, also known as collagenases), such as MMP1, MMP8 and MMP13, can disrupt the atherosclerotic fibrous cap, increasing susceptibility to rupture and thrombus formation^38,39 (Fig. 1a). In an established atheroma, neutralization of IL-1β promotes monocytes to switch to an anti-inflammatory state and secrete IL-10 (a cytokine associated with beneficial plaque remodelling⁴⁰).

IL-1 has an important role in ischaemia–reperfusion injury and, in mouse models, overexpression of IL-1Rα by gene transfection protects the myocardium by reducing levels of apoptosis and attenuating inflammation⁴¹. These findings and others provided the rationale for further studies evaluating the effect of IL-1 inhibition after acute myocardial infarction in humans^42-47 (Table 1). Promising results have now led to a larger ongoing randomized clinical trial that will evaluate IL-1 blockade versus placebo in the treatment of patients with acute ST-segment elevation myocardial infarction (NCT04463251)^48,49.

Table 1 ∣.

Selected randomized controlled clinical trials of targeted therapies for ASCVD in the general population

Target (drug)	Trial name	Dosing	Follow-up	Population	Primary end point	Outcome
IL-1 (anakinra)	MRC-ILA Heart44	100 mg daily	7 days	NSTE-ACS	Change in hsCRP	Decrease in hsCRP
	VCU-ART45 and VCU-ART 246	100 mg daily	14 days	STEMI	Change in LVESVi	No change in LVESVi; decrease in hsCRP
	VCU-ART347	100 mg daily or 100 mg twice daily	14 days	STEMI	Change in hsCRP	Decrease in hsCRP
IL-1β (canakinumab)	CANTOS6	50 mg, 150 mg, 300 mg every 3 months	3 years	Prior myocardial infarction and hsCRP levels over 2 mg/l	MACE	150-mg dose: decrease in recurrent cardiovascular events; no change in all-cause mortality
IL-6R (tocilizumab)	STAT-MI184	Single dose of 162 mg	30 days	Acute coronary syndrome	MACE	No changes; trial stopped because of lack of effect
	NCT01491074 78	Single dose of 280 mg	1–3 days	NSTEMI	AUC for hsCRP and hsTnT	Reduction in median AUC for hsCRP and hsTNT compared with placebo
	NCT01491074 185	Single dose of 280 mg	1–3 days, 3 and 6 months	NSTEMI	Inflammatory markers (27 cytokines)	Increase in serum levels of MIP1B and IP-10
	ASSAIL-MI79	Single dose of 280 mg	3–7 days	Acute coronary syndrome	Myocardial salvage index from cardiac MRI (3–7 days post-myocardial infarction)	Increase in myocardial salvage; no change in final infarct size
	NCT05350592 186	Single dose of 280 mg	48 h	Acute myocardial infarction and cardiogenic shock	NT-proBNP	Trial underway
IL-6 (ziltivekimab)	ZEUS92	15 mg every month	4 years	ASCVD and chronic kidney disease	MACE	Trial underway
TNF (etanercept)	RENEWAL128	25 mg once, twice or three times a week	24 weeks	NYHA III–IV heart failure	Composite death or heart failure hospitalization	No effect
TNF (infliximab)	ATTACH127	5 mg/kg or 10 mg/kg at 0, 2 and 6 weeks	14 weeks	NYHA III–IV heart failure	Clinical composite score	No effect on primary end point; high dose associated with worse outcome at 28 weeks

Trials for heart failure included only for TNF inhibitors. aPWV, aortic pulse wave velocity; ASCVD, atherosclerotic cardiovascular disease; AUC, area under the curve; FMDBA, brachial artery flow-mediated dilation; hsCRP, high-sensitivity C-reactive protein; hsTNT, high-sensitivity cardiac troponin T; IP-10, interferon gamma-inducible protein; LVESVi, left ventricle end systolic volume index; MACE, major adverse cardiovascular events; MIP1b, macrophage inflammatory protein-1b; MMP8, matrix metalloproteinase 8; NSTE-ACS, non-ST-segment elevation acute coronary syndrome; NSTEMI, non-ST-elevation myocardial infarction; NT-proBNP, N-terminal pro B-type natriuretic peptide; STEMI, ST-elevation myocardial infarction.

Several IL-1-targeting biologic agents are available that have shown beneficial effects in cardiovascular conditions: these include anakinra, a recombinant IL-1 receptor antagonist that blocks IL-1α and IL-1β; rilonacept, a fusion protein that functions as a trap receptor for IL-1α, IL-1β and IL-1Rα; and canakinumab, an IgG human monoclonal that targets IL-1β (Fig. 1b, Table 1). However, only anakinra is approved for the treatment of RA⁵⁰. In early-phase trials, rilonacept had some limited efficacy in RA⁵¹, and canakinumab has some beneficial effects in reducing RA disease activity⁵², but neither was ultimately pursued as a treatment for RA. Canakinumab was the focus of the Canakinumab Anti-inflammatory Thrombosis Outcome Study (CANTOS), the first large-scale proof-of-principle trial to validate the inflammation hypothesis in human atherosclerosis⁶ (Table 2). The CANTOS trial enrolled over 10,000 individuals with a history of myocardial infarction and a high sensitivity C-reactive protein (hsCRP) test result >2 mg/1; the hsCRP test cut-off was chosen as the median cut-off for the general population, pre-selecting for individuals who had a residual inflammatory burden despite guidelines-directed medical therapy. Individuals were randomly assigned to treatment with either canakinumab or a placebo. All participants received post-myocardial infarction standard-of-care treatment, including statin therapy. The results demonstrated an overall 15% reduction in major cardiovascular events, defined as myocardial infarction, stroke, or cardiovascular death, with canakinumab therapy. Additionally, a post hoc analysis of the CANTOS trial found that the magnitude of cytokine reduction was associated with extent of cardiovascular risk reduction; individuals with a greater than average reduction in levels of IL-6 or hsCRP (known as ‘cytokine responders’) had fewer cardiovascular events than the other study participants, with a 30% reduction in cardiovascular and all-cause mortality in this subset^53,54.

Table 2 ∣.

Targeted therapies and effect on ASCVD risk in patients with RA or the general population

Pathway	Target	Drug name	ASCVD in RA	ASCVD general population	Comments
IL-1	IL-1	Anakinra	–	Decreased	Surrogate endpoints only
	IL-1β	Canakinumab	–	Decreased	CANTOS trial6
IL-6	IL-6R	Tocilizumab and sarilumab	Decreased	Decreased	Surrogate endpoints only
	IL-6	Ziltivekimab	–	Unclear (studies ongoing)	ZEUS trial underway92
TNF	TNF	Adalimumab, certolizumab, etanercept, golimumab and infliximab	Decreased	Uncleara	Reduced risk of ASCVD risk in patients with RA (majority observational data only); TARGET trial186
JAK–STAT	All JAK members	Tofacitinib	Increased (safety signal requires further investigation)	–	ORAL Surveillance20
	JAK1 and JAK2	Baricitinib	–	–	–
	JAK1 only	Upadacitinib	–	–	–
		Filgotinib	–	–	Approved in the EU and Japan for the treatment of RA, but not approved in the USA

ASCVD, atherosclerotic cardiovascular disease; RA, rheumatoid arthritis; –, insufficient data to comment.

^aMost studies were in patients with heart failure and not ASCVD.

The IL-6 pathway

IL-6 is a pleiotropic cytokine with broad effects in immunity, inflammation and haematopoiesis. Multiple names were given for IL-6 until sequencing of cDNA clones revealed that these molecules were the same protein^55-57. Interest in targeting IL-6 for the treatment of RA grew with the discovery that IL-6 knockout mice were resistant to antigen-induced experimental arthritis^58,59. In vascular biology, IL-6 is secreted by multiple cell-types implicated in atherothrombosis, including macrophages, monocytes, fibroblasts and endothelial cells³⁷. IL-6 is typically secreted by cells following IL-1 stimulation, and has three predominant signalling mechanisms: ‘classical’ IL-6 signalling, known as cis-signalling, where secreted IL-6 binds to the membrane bound IL-6 receptor (IL-6R), which subsequently binds to a second membrane-bound ubiquitous protein known as gp130 to activate intracellular signalling; trans-signalling occurs when IL-6 binds to soluble IL-6R and is associated with a signal-transducing glycoprotein 130 (gp130) to activate signalling; and trans-presentation is the last mechanism of IL-6 signalling that requires interaction between a dendritic cell and a T cell^60-62.

IL-6 in RA

In the RA synovium, IL-6 is one of the most abundantly expressed cytokines. Levels of both IL-6 and soluble IL-6R correlate with histological evidence of synovial inflammation in patients with RA^63,64. IL-6, through induction of vascular endothelial growth factor (VEGF), can further promote joint inflammation and damage by increasing vascular permeability and the migration of endothelial cells³⁷. IL-6 also induces osteoclast recruitment, contributing to the destruction of bone and cartilage⁶⁵ (Fig. 1a). IL-6 and sIL-6R trans-signalling can influence osteoclastogenesis through different pathways, inducing RANKL and osteoprotegerin (OPG)⁶⁶. Human synovial fluid from patients with RA containing high levels of IL-6 and sIL-6R can stimulate the formation of osteoclast-like cells from murine osteoblastic cells, and this process is inhibited by adding an anti-IL-6R antibody⁶⁷. Two approved therapies for RA, tocilizumab and sarilumab, are monoclonal antibodies that inhibit both the membrane and the soluble form of IL-6R^68,69.

IL-6 might be associated with a higher risk of cardiovascular disease, independent of traditional cardiovascular risk factors, in patients with RA^70,71. Despite observed improvements in articular and systemic inflammation with IL-6 inhibition, treatment was also associated with lipid profile changes across various clinical trials, including increases in LDL cholesterol (LDL-C) of 12–20%, as well as increases in total cholesterol and triglycerides (all typically associated with a pro-atherogenic profile), but also increases in HDL cholesterol (HDL-C; associated with an anti-atherogenic profile)⁷². In the MEASURE trial, the effects of tocilizumab on biomarkers of cardiovascular disease in patients with RA were evaluated⁷³. Although the levels of LDL-C increased with tocilizumab therapy, the levels of more atherogenic lipoproteins, such as oxidized LDL, did not increase and pro-atherogenic proteins such as serum amyloid A, secretory phospholipase A2 and lipoprotein(a), decreased⁷³. Smaller observational studies suggest that IL-6R blockade might also promote the levels and function of anti-atherogenic HDL particles^74,75. The ENTRACTE trial was a non-inferiority phase IV randomized clinical trial that compared the risk of major cardiovascular events in patents with RA receiving tocilizumab treatment versus those patients receiving treatment with the TNF inhibitor etanercept⁷⁶. A total of 3,080 participants with inadequate response to conventional DMARDs and at least one cardiovascular risk factor were enrolled and randomized in a 1:1 fashion. Although circulating concentrations of total cholesterol, LDL-C and triglycerides did increase in the tocilizumab-treatment group compared with the etanercept-treatment group, no concomitant increase in cardiovascular events was observed over the 5-year study period⁷⁶. Together, these data suggest that tocilizumab treatment might have an overall neutral effect on ASCVD risk when compared with etanercept treatment, as indicated by the overall changes in both atherogenic and atheroprotective lipids along with a reduction in systemic inflammation.

IL-6 in ASCVD

A role for IL-6 reduction in decreasing the risk of cardiovascular disease was observed in 2017 in the aforementioned CANTOS trial, which targeted inflammation through IL-1α inhibition⁶. Further analysis of the CANTOS study data revealed that the magnitude of IL-6 reduction correlated with the overall reduction in major cardiovascular events⁷⁷. In fact, in those participants who did not experience reductions in IL-6 levels with canakinumab treatment, the number of cardiovascular events did not decrease and remained similar to the levels observed in the placebo group. IL-6 is also upregulated during myocardial infarction and can affect myocardial remodelling and plaque destabilization⁷⁷ (Fig. 1a). In two separate studies, a single dose of tocilizumab resulted in higher levels of CRP and cardiac troponin T in patients experiencing a non-ST-segment elevation myocardial infarction and in patients following an ST-segment elevation myocardial infarction^78,79. Furthermore, in one of these studies (the ASSAIL-MI trial), a single dose of tocilizumab administered prior to a percutaneous coronary intervention showed the potential to reduce myocardial damage from ischaemia, as indicated by an increase in the myocardial salvage index (the difference between the actual and potential infarct size defined by the initial area at risk by coronary occlusion, as measured by cardiac MRI) and a reduction in circulating CRP concentrations^78,79 (Table 1).

Genomic studies also support a key role for IL-6 in ASCVD. A known variant in IL6R (that leads to a aspartic acid to alanine substitution at position 358 (Asp358Ala) of the encoded protein) hinders the cell surface expression of membrane-bound IL6R, thus reducing IL-6 signalling in individuals with this variant⁸⁰. Mendelian randomization studies reported an association of this variant with reduced risk of ASCVD^81,82, an association that was later confirmed in follow-up genetic association studies that considered a variety of cardiovascular phenotypes, including myocardial infarction and peripheral arterial disease^83-85.

Although mounting data are available suggesting that direct inhibition of IL-6 itself could maximize the anti-inflammatory benefit of inhibiting the IL-6 pathway in atherosclerosis, not all IL-6 antagonists likely share the same profile. Sirukumab, an IL-6 inhibitor that targets the cytokine rather than the cytokine receptor, did not receive US Food and Drug Administration (FDA) approval for the treatment of RA owing to concerns regarding the mortality attributed to cardiovascular events and malignancy in the treatment group of an ongoing clinical trial⁸⁶; the mechanism for this safety signal is not known. By contrast, ziltivekimab, a humanized monoclonal antibody that blocks IL-6, had promising effects on reducing the cardiovascular risk of individuals in a 2021 randomized double-blind placebo-controlled phase II trial, known as RESCUE^87-90 (Fig. 1b, Table 1). RESCUE evaluated the effects of ziltivekimab on biomarkers of inflammation and thrombosis in patients at a high cardiovascular risk, enrolling 264 patients with chronic kidney disease and increased levels of hsCRP (>2 mg/dl)⁹¹. Over a period of 24 weeks, ziltivekimab treatment resulted in a reduction in the levels of various biomarkers of inflammation and thrombosis, including hsCRP, lipoprotein(a), fibrinogen and soluble phospholipase A2 (sPLA2). Moreover, no notable effect was observed for ziltivekimab on the total cholesterol to high-density lipoprotein cholesterol ratio, unlike previous reports for other IL-6 antagonists⁸⁷. Notably, however, prior studies on IL-6 blockade were performed largely in patients with RA, where factors such as higher starting levels of inflammation might influence changes in lipid levels. The ziltivekimab phase II findings laid the foundation for the current phase III study, the ZEUS study, which will compare ziltivekimab with placebo in approximately 6,000 patients with chronic kidney disease and elevated hsCRP levels to determine whether IL-6 inhibition reduces the rate of cardiovascular events⁹² (Table 2).

The TNF pathway

TNF is a pleiotropic cytokine that signals through two different receptors: TNF receptor 1 (TNFR1) and TNF receptor 2 (TNFR2). TNFR1 is ubiquitously expressed and present on most cells, whereas TNFR2 has more limited expression and is mainly expressed on certain immune cells (such as some T cell and B cell subsets) as well as fibroblasts, endothelial cells and neuronal cell subsets⁹³. In general, signalling through soluble TNFR1 triggers pro-inflammatory pathways whereas TNFR2 signalling has an important role in immune modulation⁹⁴.

TNF in RA

TNF is the most well-known and targeted cytokine in RA^25,95. Overexpression of a TNF transgene in mouse models can induce spontaneous erosive inflammatory arthritis and inhibition of TNF via various methods, and can suppress arthritis in this model^96,97. In fact, inhibition of TNF can ameliorate disease in several animal models of arthritis, including antigen-induced arthritis, adjuvant arthritis and the aforementioned human TNF transgene-induced arthritis⁹⁸. In the human synovium, TNF can induce the proliferation of fibroblast-like synoviocytes and trigger cartilage destruction by inducing the production of collagenase by synovial cells, osteoclastogenesis and bone resorption^99-101 (Fig. 1a). TNF inhibition can rapidly decrease levels of IL-6 and acute-phase reactants, suppress leukocyte migration and endothelial cell activation and improve regulatory T cell function^102,103.

Five TNF inhibitors are available for the treatment of RA. In order of development, infliximab is a chimeric anti-TNF antibody comprising both a mouse Fc fragment and a human IgG1 antibody^104,105. Etanercept is a recombinant IgG1 Fc fragment fused to two p75 TNF receptors^106,107. Adalimumab is a fully humanized recombinant IgG1 antibody against TNF¹⁰⁸. Golimumab and certolizumab are the two most recently approved TNF inhibitors and are both humanized anti-TNF monoclonal antibodies^109-111. Multiple observational studies have demonstrated an association between TNF inhibitor use and reduction in cardiovascular events in patients with RA^112-115 and most recently one trial (the TARGET trial) demonstrated a reduction in vascular inflammation with TNF inhibitor therapy¹¹⁶ (Table 2).

TNF in ASCVD and other cardiac conditions

In the general population, TNF inhibition has a more complicated history. Whether TNF inhibition can reduce the risk of cardiovascular disease in the general population has not been studied and limited studies are available regarding the relationship between the TNF pathway and ASCVD risk in the general population^117-119. TNF has multiple effects on the endothelium, including promoting the production of reactive oxygen species (ROS), reducing the bioavailability of nitric oxide and increasing endothelial permeability^117,120,121 (Fig. 1a); furthermore, TNF is present in human atherosclerotic plaques¹²². In some mouse models of atherosclerosis, deletion of Tnf can attenuate the development of disease^123,124. Yet, other mouse models have yielded conflicting results^125,126. In large-scale observational studies, higher plasma concentrations of TNF were associated with the recurrence of cardiovascular events even after adjusting for classic cardiovascular risk factors¹¹⁸. The most-studied effect of TNF inhibition in cardiovascular disease relates to the effects on heart failure^127,128.

In experimental settings, TNF can induce cardiomyopathy and left ventricular dysfunction^119,129. However, in trials of TNF inhibition (with etanercept and infliximab) in patients with established systolic heart failure, treatment had no beneficial effect on recurrent hospitalization or death (Table 1). Further, a higher hospitalization rate was observed in a subgroup of patients that received infliximab 10 mg/kg compared with placebo^127,128. Thus, the data suggested that global TNF inhibition could cause dysregulation in both adverse and protective signalling in a failing heart and this treatment is generally avoided in patients with systolic heart failure¹³⁰. However, it is important to note that these studies were performed in the general population and not in RA specifically, where TNF has a strong pathogenic role.

The JAK–STAT pathway

The JAK–STAT pathway is activated downstream of multiple pro-inflammatory cytokines and is one of the newest targets for RA. Studies are ongoing to address concerns regarding the association between JAK inhibition and risk of cardiovascular events²⁰. The JAK-STAT pathway consists of four JAK proteins (JAK1, JAK2, JAK3 and TYK2) and seven STAT proteins. Activation of JAK–STAT requires a cytokine that binds to its receptor, which results in JAK phosphorylation, STAT binding and STAT activation. The activated STAT protein translocates to the nucleus and binds to enhancers to induce gene transcription¹³¹. Downstream effects of the JAK–STAT pathway are broad, ranging from immune cell development to the induction or prevention of apoptosis^132,133.

JAK inhibition in RA

In RA, the JAK–STAT pathway is constitutively elevated and various proteins that function to inhibit the JAK–STAT pathway are thought to be deficient¹³⁴. Blockade of the JAK–STAT pathway can improve clinical symptoms and slow disease progression in patients with RA¹³⁵. To date, three oral JAK inhibitors are approved for the treatment of RA in the USA: tofacitinib, baricitinib and upadacitinib¹³⁵. Filgotinib is also approved for the treatment of RA in Europe and Japan, but approval was rejected by the FDA in 2020 (ref. ¹³⁶). Each agent differs in terms of its JAK receptor selectivity, although studies in RA have suggested that JAK1 inhibition is largely responsible for the efficacy of these drugs in RA¹³⁷. Tofacitinib is a non-specific JAK inhibitor that has greater selectivity for JAK1 and JAK3 than for JAK2 and TYK2 (refs. ^138-140); baricitinib targets JAK1 and JAK2, with some activity against TYK2 but minimal activity against JAK3; whereas upadacitinib and filgotinib are both JAK1 selective agents^137,141,142. Inhibition of the JAK pathway can affect the signalling of multiple cytokine pathways, including IL-6 and TNF, which depends on the JAK selectivity of the agent¹³⁵ (Fig. 1b).

In 2022, the cardiovascular safety of JAK inhibitors in RA was called into question with results from a post-marketing safety surveillance study of tofacitinib²⁰. The ORAL Surveillance study was a prospective, phase IIIb/IV randomized, open-label, non-inferiority trial of major adverse cardiovascular events and malignancy in patients with RA receiving tofacitinib versus TNF inhibitors (etanercept or adalimumab). The trial, mandated by the FDA, enrolled patients with RA with an inadequate response to methotrexate, who were over 50 years old and had at least one cardiovascular risk factor (such as cigarette smoking, hypertension, hyperlipidaemia, diabetes mellitus, family history of premature coronary artery disease, prior history of coronary artery disease or extra-articular disease associated with RA). Patients were randomly assigned to receive tofacitinib (5 mg or 10 mg twice daily) or a TNF inhibitor. During the interim data analysis, tofacitinib 10 mg twice daily was associated with a higher risk of venous thromboembolism than treatment with a TNF inhibitor. As a result, the tofacitinib dose for all study participants was reduced to 5 mg twice daily. The results of ORAL Surveillance indicated a safety signal for increased major adverse cardiovascular events in tofacitinib versus TNF inhibition, on the basis of a pre-specified non-inferiority criterion (Table 2). Malignancy rates were also higher in the tofacitinib group for both doses than in the TNF inhibitor group. The signal for major adverse cardiovascular events was higher among patients who were older than 65 years, patients who had ever smoked and aspirin users¹⁴³. Whether these results are specific to tofacitinib and not more selective JAK inhibitors such as upadacitinib, or whether the effect is specific to RA and not other indications for JAK inhibitors, remains to be seen. Furthermore, although tofacitinib treatment is associated with an increase in lipid levels^138,140,144, the relevance of this effect on cardiovascular risk is less clear; one small mechanistic prospective study demonstrated that tofacitinib might have beneficial effects on atherosclerosis as measured by regression of carotid intima media thickness in patients who had elevated levels of carotid intima media thickness at baseline¹⁴⁵. Another small study demonstrated that treatment with tofacitinib for 6 weeks was associated with an increase in HDL-C and LDL-C concentrations. Further functional studies of HDL suggested that the treatment led to a more anti-atherogenic profile of HDL¹⁴⁶.

Notably, prior to the ORAL Surveillance study, in six phase III studies and two open-label long-term extension studies of tofacitinib in patients with RA, the overall incidence of cardiovascular events was low¹⁴⁷. In a post hoc analysis that combined the phase III and long-term extension studies, higher levels of HDL-C were associated with lower rates of subsequent cardiovascular events. A meta-analysis of 26 RCTs found no difference in the rate of cardiovascular events among patients with RA being treated with a JAK inhibitor compared with placebo¹⁴⁸. The STAR-RA population-based study evaluated patients initiating treatment with tofacitinib or a TNF inhibitor using real-world observational data from commercial and Medicare data¹⁴⁹. The analysis found no difference in rates of myocardial infarction or stroke incidence in the overall population. However, when the analysis was restricted to patients with cardiovascular risk factors using the same criteria as the ORAL Surveillance study, a similar numerical increase in cardiovascular events was observed^149,150. Thus far, JAK inhibition has been consistently associated with a signal for an elevated risk of cardiovascular events among older individuals with cardiovascular risk factors when compared with TNF inhibition. Further studies are needed to understand the absolute risk for a given patient and the underlying mechanisms¹⁵¹.

JAK inhibition in ASCVD

The JAK–STAT pathway has an important role in atherosclerosis, as is evident from the cytokines impacted by JAK inhibition involved in this disease (such as IL-6 and TNF, as discussed in the previous sections). For instance, IL-6 activates the JAK–STAT pathway after binding to its receptor¹⁵². IFNγ can also augment atherosclerosis and signals through both JAK1 and JAK2 (ref. ¹⁵³). However, not every JAK–STAT process is pro-inflammatory. For instance, although STAT3 can promote the expression of angiotensinogen II (a known mediator of cardiac hypertrophy) by binding to its gene promoter in myocytes^154,155, STAT3 can also mediate cardioprotective signalling in myocytes when activated by IL-10 (ref. ¹⁵⁶). Unlike IL-1, IL-6 and TNF, no large-scale trials have yet assessed the effects of JAK inhibition in the prevention of cardiovascular events in the general population, although a multi-STAT inhibitory strategy has been proposed as a potential therapeutic avenue¹⁵⁷. Future mechanistic investigations to understand the role of JAK inhibition in promoting or mitigating cardiovascular disease will be crucial to understanding how this class of drugs could be used in the future.

Broad inhibition of multiple cytokines

The majority of this Review focuses on the key cytokine pathways involved in both RA and ASCVD that are the subject of approved targeted therapies, whereas in this section we discuss methotrexate, a nonspecific drug that has been studied in both patients with RA and the general population for reducing the risk of ASCVD and that is believed to target the pathways downstream of IL-6, TNF and IL-1 (Fig. 1a). First developed as an anticancer therapy for childhood leukaemia, this drug was also found to inhibit the proliferation of connective tissue, leading to the first study of this drug in RA in 1951 (ref. ¹⁵⁸). Methotrexate is a dihydrofolate reductase inhibitor that, when given at high doses, blocks the synthesis of DNA nucleotides¹⁵⁹. For example, when used as part of a chemotherapy regimen for lymphoma, an individual could receive 5,600 mg of intravenous methotrexate in one dose¹⁶⁰. In RA, this drug is used at much lower doses, for example, 25 mg once a week, together with folic acid supplementation, to reduce the toxicity profile of this drug without reducing its anti-inflammatory efficacy¹⁶¹. Various mechanisms have been proposed for the anti-inflammatory effects of methotrexate; in addition to the anti-metabolite effect of this drug, methotrexate can lower the production of key pathogenic cytokines including IL-1, IL-6 and TNF; other proposed mechanisms include inducing apoptosis of inflammatory cells and increasing adenosine signalling¹⁶².

Methotrexate in RA

Methotrexate is currently the first-line therapy for RA²⁵. Although open-label studies of methotrexate date back to 1968 (ref. ¹⁶³), it was not until 1983 that the first RCT of methotrexate was performed in patients with RA¹⁶⁴. The drug showed promising efficacy in these patients and similar results were obtained in a second larger placebo-controlled trial the following year¹⁶⁵, leading the FDA to approve methotrexate as a therapy for RA in 1988 (ref. ¹⁶⁶).

Studies in RA have shown an association between methotrexate use and reduction in cardiovascular risk by surrogate endpoints. In the TEAR trial, which compared methotrexate monotherapy, methotrexate therapy in combination with etanercept and methotrexate therapy in combination with two other DMARDs (triple therapy) in patients with RA, all three approaches led to an improvement in measures related to the anti-atherogenic functions of HDL (as measured by paraoxonase 1 activity and apolipoprotein AI levels)¹⁶⁷. In other studies, methotrexate therapy has led to improvements in endothelial function biomarkers in patients with RA, including carotid intima-media thickness¹⁶⁸. In a systematic review and meta-analysis that included 10 observational studies of patients with immune-mediated inflammatory diseases taking methotrexate, methotrexate use was associated with a 21% overall lower risk of cardiovascular disease and a 18% lower risk of myocardial infarction¹⁶⁹.

Methotrexate in ASCVD

In the general population, the CIRT (Cardiovascular Inflammation Reduction Trial) was the second large-scale trial to test the inflammation hypothesis of atherosclerosis following the CANTOS trial¹⁷⁰. The rationale was to not only choose an agent that has anti-inflammatory properties without effects on lipids (such as statins) but also a drug that is well-tolerated with a reasonable safety profile. Given the widespread use of methotrexate and long-term data on the efficacy and toxicity of this drug in RA, methotrexate was an ideal agent. The primary end point of CIRT was reduction in the rate of a composite primary end point of non-fatal myocardial infarction, non-fatal stroke or cardiovascular death among patients with either a prior myocardial infarction or multivessel coronary artery disease who additionally had either type II diabetes or metabolic syndrome. Although aimed at enrolling up to 7,000 patients, the trial was stopped early owing to a lack of benefit of methotrexate and included a total of 4,786 patients. No difference was observed for the primary end point when comparing low-dose methotrexate (15–20 mg weekly) with placebo. Furthermore, no differences were observed for CRP, IL-1α or IL-6 levels in the two treatment arms. However, unlike the CANTOS trial, the entry criteria did not include a minimum CRP requirement and the average baseline levels for the population were relatively low at 1.6 mg/l (ref. ¹⁷¹), introducing a potential basement effect. At this level of CRP, inflammation could only be further reduced by a small or insignificant amount, which in turn would preclude the ability to detect associations with reductions in cardiovascular risk.

Other relevant inflammatory pathways

In addition to the IL-6, TNF and IL-1 pathways, other pro-inflammatory pathways might also have a role in in both RA and ASCVD. In this section we discuss pathways for which some data are available on the effects of targeting these pathways on cardiovascular risk in RA or ASCVD but not for both. In patients with RA, targeting the CD80–CD86 costimulatory pathway with abatacept might result in a lower cardiovascular risk compared with targeting the TNF pathway, particularly for patients with diabetes mellitus^172,173. However, limited data are available so far for this drug.

Likewise, the inflammasome, an important component of innate immunity, is implicated in the pathogenesis of ASCVD but is not known to have a major role in RA pathogenesis. The inflammasome is an important component of innate immunity and is implicated in the pathogenesis of atherosclerosis^174,175. Of the five pattern recognition receptors known to function as inflammasome receptors, NLRP3 is the major receptor involved in atherogenesis^174,176. Colchicine is an NLRP3 inflammasome inhibitor and inhibits tubulin polymerization and microtubule formation¹⁷⁷. Colchicine is not used in the treatment of RA but is a mainstay of treatment for gout and familial Mediterranean fever¹⁷⁷. Studies in the general population have demonstrated that colchicine has some efficacy in reducing the risk of ASCVD^7,178,179. In the COLCOT (Colchicine Cardiovascular Outcomes Trial) study, which tested low-dose colchicine as secondary prevention and reduce cardiovascular disease in patients with a recent acute myocardial infarction, colchicine lowered the risk and incidence of urgent hospitalization for angina leading to coronary revascularization. Evidence also exists for the use of colchicine in the treatment of stable coronary artery disease; in the LoDoCo (Low-Dose Colchicine) pilot trial, colchicine (0.5 mg once daily) reduced the rate of cardiovascular events in patients with chronic coronary artery disease¹⁷⁸. As noted in the section on IL-1, therapies that block the IL-1 pathway also target the consequences of inflammasome activation, and some evidence suggest that this approach provides some benefit in the prevention of ASCVD; however, these drugs either have a low efficacy (such as anakinra) or are not currently used in the treatment of RA (such as canakinumab)¹⁸⁰.

Synthesizing the data

IL-1, IL-6 and TNF have important roles in the pathogenesis of RA and ASCVD, and emerging data support the involvement of the JAK–STAT pathway (which functions downstream of these cytokines). Focusing on the pathways of targeted therapies enables us to more precisely understand the downstream effect of blocking these pathways. Although deciphering the exact hierarchy of IL-1, IL-6, TNF and the JAK–STAT pathway within the pathogenesis of RA and atherosclerosis is challenging, most compelling data point to the involvement of IL-6 in both RA and ASCVD (Table 2). Indeed, data from both mechanistic investigations and genomic studies support the important role of IL-6 as a central cardiometabolic signalling cytokine¹⁸¹.

Despite the importance of TNF in RA pathogenesis, and evidence from observational studies of TNF inhibitors reducing the cardiovascular risk of patients with RA, the complex role of TNF in vascular biology and prior clinical trials of TNF inhibitors in patients with heart failure have made this therapy a less attractive option for reducing the risk of cardiovascular disease in the general population. Data from studies of IL-1 inhibitors, as well as of inflammasome inhibition, suggest that the IL-1 pathway probably has a more general upstream role in the inflammatory cascade. This more general role might explain the lower efficacy of IL-1 inhibitors in treating articular signs and symptoms of patients with RA compared with inhibitors of other pathways (such as IL-6 and TNF inhibitors), which have more direct effects on controlling RA inflammation. In terms of cardiovascular effects, in a study comparing the effects of anakinra and tocilizumab in patients with RA, anakinra had a more salutary effect on cardiac function, whereas tocilizumab mainly had beneficial vascular effects¹⁸². Furthermore, specifically targeting IL-1α was associated with a reduction in ASCVD risk in the general population. Based on these data, in the future, IL-1 might play a larger role in reducing cardiovascular risk than in controlling RA disease activity alone¹⁷⁰.

In terms of drugs that more broadly target these cytokine pathways, observational data suggest that methotrexate can reduce the cardiovascular risk of patients with RA, a population exposed to chronic moderate-to-high levels of inflammation. However, methotrexate did not reduce the rate of recurrent cardiovascular events among high-risk individuals in the general population who had a low inflammatory burden in the CIRT study¹⁷¹. Although it is unlikely that another trial will assess the effects of methotrexate on cardiovascular disease in the general population, the lack of benefit observed in the CIRT population (that is, in individuals with a low baseline hsCRP level) in a sense supported the notion of inflammation as an important risk factor for cardiovascular disease. Inhibition of the JAK–STAT pathway impacts both the IL-6, TNF and other pro-inflammatory pathways. However, this broader effect might also lead to the mixed signals regarding its effect on ASCVD risk.

The complex signalling of these pathways creates a challenge for both rheumatologists and cardiologists in that these results cannot be viewed in a vacuum. Each cytokine and pathway has pleotropic functions and can further induce the expression of chemokines, which directly modulate the migration of cells and regulate context-specific aspects of the acute and chronic inflammatory response in the synovium and in the vasculature¹⁸³. This complexity has created a double-edged sword in that blockade of proximal pro-inflammatory mediators might yield differing off-target effects, as was observed in the ORAL Surveillance study²⁰. Furthermore, the effect of targeting these pathways to reduce cardiovascular risk might differ in patients with RA compared with the general population.

Conclusion

The convergence of pathways implicated in RA and ASCVD have led to a paradigm shift with regard to how treatments are considered in both rheumatology and cardiology. With multiple targeted therapies available for the treatment of RA, and limited data regarding which treatment will best control a particular patient’s disease, treatment decisions are currently guided by a patient’s comorbidities and the safety profile of the drugs. Incorporating ASCVD risk in therapy selection for each individual patient could help to inform the optimal therapy beyond simply the control of joint inflammation as we refine our knowledge of the mechanistic interplay between synovial inflammation and atherosclerotic development. An improved understanding of RA and cardiovascular risk will allow consideration of mitigating not just RA disease activity but also cardiovascular risk. IL-6 targeting currently stands as the most promising candidate; for the majority of remaining targets, data regarding the use of these targeted therapies and overall cardiovascular events in the general population remain limited. Returning to the notion of RA as a human model of inflammation, more studies on the pathobiology of atherosclerosis are needed in RA to better understand the downstream cardiovascular effects of targeted therapies and the mechanism behind these effects. Although this Review provides a simplified version of the complex inflammatory pathways shared between the two conditions, it provides a starting point for summarizing to date the inflammatory pathways that converge in RA and ASCVD, informing future direction for studies to improve the management of ASCVD risk. Moreover, this Review highlights a movement towards a growing collaboration between rheumatology and cardiology in both research and clinical care in the years to come.

Key points.

Rheumatoid arthritis (RA) and atherosclerotic cardiovascular disease share pro-inflammatory pathways that promote the pathogenesis of both conditions, including the IL-1 and IL-6 pathways.

Inflammation encompasses numerous pathways, and data from studies of targeted therapies developed for RA can provide insight into the potential effects of these therapies on cardiovascular risk.

Targeting IL-6 is considered more effective in controlling RA disease activity than targeting IL-1 and robust data also suggest that this therapy can reduce atherosclerotic cardiovascular disease risk in the general population.

Targeting some pro-inflammatory pathway(s), such as the TNF pathway, might be more beneficial than targeting others, such as JAK–STAT signalling, in reducing cardiovascular risk.