Evidence for Biologic Drug Modifying Anti-Rheumatoid Drugs and Association with Cardiovascular Disease Risk Mitigation in Inflammatory Arthritis

INTRODUCTION

Systemic inflammatory arthritides are associated with increased risk of atherosclerosis and ultimately cardiovascular (CV) risk.^1–8 Key inflammatory cytokines prominent in these conditions also play a major role in atherosclerosis development and endothelial dysfunction. Endothelial dysfunction is an early event in the development of atherosclerosis and can be detected before structural changes in vessels on angiography.⁹ In the heart, endothelial dysfunction can manifest as coronary vascular dysfunction, associated with both systemic inflammation and may precede or coexist with high-risk coronary atherosclerosis. The endothelium, serving as a major regulator of vascular homeostasis, provides a permeability barrier for the vasculature, maintains a nonthrombogenic surface, regulates vascular tone and tissue flow, and inhibits vascular smooth muscle cell growth. In the presence of inflammatory cytokines, endothelial cells are activated which facilitates leukocyte adhesion and migration into the vessel wall, production of prothrombotic substances, and vasoconstriction, thereby creating a proatherogenic and procoagulant environment.^10,11 At the individual patient level, this proatherogenic and procoagulant environment promote the development of acute thrombi that can cause a myocardial infarction (MI). Consequently, inflammation is a key mediator of a constellation of abnormalities that initiate and accelerate the progression of atherosclerosis.^12–14

A complex interplay exists between underlying systemic inflammation and classic cardiovascular risk factors, such as obesity, diabetes mellitus (DM), hypertension (HTN), and dyslipidemia in patients with inflammatory arthritis. The increased CV mortality attributed to inflammation, observed in patients with inflammatory arthritis is independent of traditional CV risk factors.^15–25 Over the last 3 decades, biologic disease-modifying drugs (bDMARDs) and more recently, small molecules that target specific pathways involved in the pathogenesis of these conditions enable tight control of disease activity and in some cases remission. We will focus our review on the inflammatory arthritides, namely rheumatoid arthritis (RA), psoriatic arthritis (PsA), and spondyloarthritis (SpA). Across these conditions, the pathways targeted include, CD20, interleukin (IL)-1, IL-6, IL-17, IL-12/IL-23, TNFα, the T-cell costimulatory pathway through CD80/86, and more recently inhibition through the Janus kinases (JAK).

While nuances remain in the decision of overall therapy for each disease and the individual patient, these therapies are well tolerated and safe with regards to infection with regular monitoring.^26,27 However, the data on CV risk, including both benefit and possible harm, remain an area of active investigation. Rheumatoid arthritis (RA) has the most robust data on the bDMARDs and modulation of CV risk compared to the other inflammatory arthritides, PsA and SpA. For psoriatic conditions, the majority of the data on CV risk mitigation comes from the psoriasis literature among which 25–30% have concurrent joint involvement. The aim of this review is to provide an overview of the data on CV risk modulation within each class of bDMARDs with a focus on studies that address the impact on surrogate markers of CVD and cardiovascular outcomes.

TNFα INHIBITORS

TNFα is one of the key pathologic cytokines implicated in the pathogenesis of increased CV risk within systemic inflammatory arthritides. TNFα and related proinflammatory cytokines, implicated in the initial insult of endothelial dysfunction and injury to the vascular endothelium, is considered one of the earliest events in atherosclerosis. TNFα inhibitors (TNFi) are commonly used throughout the inflammatory arthritides of which 5 options are available in this class, adalimumab, certolizumab, etanercept, golimumab, and infliximab (Table 1). In RA, epidemiological evidence suggests that treatment with TNFi is largely associated with a reduction in the incidence of cardiovascular disease compared to subjects on nonbiologic DMARDs (nbDMARDs) such as methotrexate and sulfasalazine.^28,29 One of the earlier large-scale studies examining the effect of TNFi on cardiovascular mortality in patients with RA found TNFi use led to a reduction of CV mortality by 35% compared to a standard treatment regimen.³⁰ A large British registry of patients with RA treated with TNFi compared to bDMARD-naïve patients on nbDMARDs were examined over time for incident MI. A total of 252 verified first MIs were analyzed and among these, there were 58/3058 patients receiving nbDMARD and 194/11200 patients receiving TNFI (median follow-up per person was 3.5 years and 5.3 years, respectively). This was associated with a reduced risk of MI among subjects on TNFi compared to nonbiologic nbDMARD therapy, adjusted hazard ratio for MI of 0.61 (95% CI 0.41 to 0.89) TNFi vs nbDMARD. There was not a statistically significant difference observed between mortality²⁹ and in the same cohort, assessment of ischemic stroke did not reveal a difference in the occurrence between TNFi-treated patients compared to conventional therapy.³¹ The QUEST-RA cross-sectional study examined over 4000 patients on a range of DMARD therapies and observed that prolonged exposure to TNFi was associated with a lower risk of CV events, with a greater relative risk reduction compared to other nbDMARDs.³²

Table 1.

Evidence of CV risk mitigation across bDMARDs and small molecules for the treatment of inflammatory arthritis

Pathway	Therapies	RA	SpA	PsA
CD20/B-cell depletion	Rituximab	D
CTLA4	Abatacept	C		D
IL-1	Anakinra	C
IL-6	Sarilumab, tocilizumab	C
IL-17A	Ixekizumab, secukinumab		D	C
IL-12/23	Ustekinumab			D
IL23	Guselkumab			D
JAK	Baricitinib, tofacitinib, upadacitinib	B		B (FDA approved for tofacitinib only)
TNF-a	Adalimumab, certolizumab, etanercept, golimumab, infliximab	A	C	B

A, multiple robust large observational studies; B, moderate observational data of varying sizes; C, few small studies; D, limited or awaiting data.

Gray box = treatment not FDA approved for condition.

Data on the impact of bDMARDs on CV risk among SpA are sparse. Patients with ankylosing spondylitis (AS) AS, as part of a cohort including RA and PsA, from 2001 to 2015 in an Australian-based rheumatic disease cohort were included to evaluate the risk of CV events and the association with TNFi.³³ A small percentage of patients were on other non-TNF biologics (among the patients on biologics, 5.2% on non-TNF versus 94.8% on TNFi), and 36.8% were biologic-naïve at enrollment; 55.6% of the participants were on methotrexate at study enrollment and 39% on prednisone. TNFi was associated with a reduction in major CV events, (HR 0.85, 95% CI 0.76–0.95) which was abrogated in those who stopped therapy (HR 0.96, 95% CI 0.83–1.11). No differences in CV event rate were observed between the inflammatory conditions after statistical adjustment.³³

Due to the relatively low prevalence of inflammatory arthritides compared to other conditions such as HTN, obtaining sufficient numbers of hard clinical endpoints is one of the challenges in studying CV risk in rheumatic conditions. Thus, surrogate CV endpoints are typically used for studies in the inflammatory arthritides to understand associations between bDMARDs with CV risk. As described previously, endothelial dysfunction is a key early event in the development of atherosclerosis. In two small short-duration mechanistic studies, infliximab was demonstrated to improve flowmediated vasodilation in patients with RA [n = 11]³⁴ and patients with AS (n = 12)³⁵ after 12 weeks of therapy. Conversely, another study that examined pulse wave velocity (PWV), a surrogate measure of arterial stiffness, was not shown to be significantly improved in patients with AS after 24 weeks³⁶ or after 6 and 12 months³⁷ of TNFi therapy. In PsA, TNFi compared to other nbDMARDs led to a greater reduction in the development of carotid atherosclerotic plaques after 4 years of treatment, 40.4% of those on TNFi vs 15.8% on nbDMARD.³⁸ A single-center observational study examined whether TNFi had any effect of coronary artery plaque formation or progression via the evaluation of coronary atherosclerosis by coronary computed tomographic angiography (CCTA) in patients with RA. TNFi therapy was associated with lower CV risk in patients who had evidence of high-risk plaque features at baseline but not in patients without and was associated with the transition to more stable plaque features (OR 4.0 [95% CI 1.015–15.32]). Furthermore, TNFi therapy was associated with low-attenuation plaque regression/loss. These results suggest that TNFi use may alter coronary plaque features and reduce the risk of new coronary plaque in patients with the evidence of atherosclerosis at baseline.³⁹ Lastly, although some controversy exists, given the association of adverse outcomes in patients with heart failure and TNFi in the general population, TNFi is not currently recommended in patients with inflammatory arthritis with a known history of Class III or Class IV heart failure.⁴⁰

INTERLEUKIN-1 AND INTERLEUKIN-6 INHIBITION

IL-1 and IL-6 are pivotal cytokines in innate immunity and over the past 10 years, increasing evidence have demonstrated an important role for IL-1 and IL-6 in CVD for the general population.^41–44 IL-6 signaling is implicated in plaque initiation and destabilization as well as adverse outcomes in acute coronary syndrome.⁴⁵ Tocilizumab is a humanized monoclonal antibody to the soluble IL-6 receptor and is the most commonly prescribed agent with the most data for CV effects. However, additional options such as sarilumab a fully human monoclonal antibody to the IL-6 receptor are also now approved.⁴⁶ Both agents are approved bDMARDs utilized in RA predominantly after inadequate response to TNFi.⁴⁷ Both IL-1 and IL-6 blockade are not FDA approved agents utilized in the SpA or PsA. Thus, the data on CV risk reduction regarding IL-6 are focused on RA.

Lipid parameters are known to be altered by IL-6 inhibition. The AMBITION study was a randomized clinical trial that examined the efficacy and safety of tocilizumab monotherapy versus methotrexate in patients with active RA. In patients undergoing tocilizumab therapy, total cholesterol, LDL, HDL, and triglycerides increased within the first 24 weeks when compared to methotrexate. However, after week 24 these levels returned to their baseline. Additionally, the ratio of LDL to HDL cholesterol was also unchanged, suggesting that these transient changes in lipid parameters did not result in a more proatherogenic profile.^48,49 Another small study that examined the fractional catabolic rate of LDL in 11 patients with severe RA undergoing tocilizumab therapy found that lower LDL levels were due to the hypercatabolism of LDL particles and IL-6 inhibition reversed this catabolism; these data suggest that hepatic IL-6 signaling plays a role in these lipid alterations.⁵⁰ In fact, some data may suggest that IL-6 inhibition provides greater CV risk reduction than TNFi. A meta-analysis examined the comparative effects of TNFi compared to non-TNFi bDMARDS and nbDMARDs on CV risk in patients with RA. Interestingly, tocilizumab was associated with a decreased risk of major adverse CV events (MACE) (OR 0.59 [95% CI 0.34–1.00]) compared to patients on TNFi, although rates of stroke were similar (OR 0.98 [95% CI 0.59–1.61]).⁵¹

Few studies directly compare bDMARDs and CV risk reduction and existing data on this topic are predominantly in RA. In one systematic review and meta-analysis, a comparison was made between bDMARDs by drug class and overall MACE. In this study, tocilizumab was associated with a reduced risk of MACE compared to TNFi as a group (OR 0.59). In these studies, this protective effect was observed predominantly among patients with pre-existing CVD.⁵¹

Data on IL-1 inhibition in RA are limited although there are data to support its salutary effect on CV risk in the general population.⁴¹ In a cross-over trial of 80 patients with RA randomized to anakinra or placebo, and after 48 hours to the other treatment, investigators found that IL-1 inhibition was associated with improved coronary flow reserve assessed by doppler flow of the left anterior descending artery (LAD). Data from the general population demonstrated inflammation as an independent risk factor for CVD and that reducing inflammation via IL-1 inhibition, reduced CV risk. The Canakinumab Anti-Inflammatory Thrombosis Outcomes Study (CANTOS) trial was the first large-scale cardiovascular study to show that blocking interleukin-1ß prospectively reduced incident CV events without modifying other CV risk markers, e.g. lipid levels.⁴¹ Further, subgroup analyses demonstrated that the magnitude of clinical benefit was directly related to the degree of reduction in IL-6 levels achieved by individual trial participants.⁵² These data suggest that IL-6 may be the primary target and has been further supported by recent data from the RESCUE trial showing that ziltivekimab, a novel IL-6 inhibitor, reduced biomarkers of systemic inflammation and thrombosis central to atherosclerosis in patients with chronic kidney disease (CKD) and elevated hsCRP.⁵³ Based on these results there is now a large double-blinded randomized control trial, Ziltivekimab Cardiovascular Outcomes Study (ZEUS), which will formally test whether reducing circulating IL-6 leads to a reduction in CV event rates with ziltvekiumab compared to placebo in patients with chronic kidney disease and elevated hsCRP.⁵⁴

ABATACEPT

Abatacept is a recombinant fusion CTLA4 protein that inhibits the CD80/CD86 costimulation important in T-cell activation. Similar to tocilizumab and anakinra, abatacept is often used after patients have an inadequate response to TNFi and other nbDMARDS. Limited data exist on the effects of abatacept on CV risk or surrogate markers of CV risk. In the AMPLE study, comparing patients with RA initiating abatacept versus TNFi therapy, HDL function was observed to improve with both treatments.⁵⁵ In one comparative effectiveness study of Medicare participants abatacept treatment initiators served as the reference group to compare the incidence of MI. In this study, TNFi initiators had a similar risk of incident MI (adjusted HR 1.3 95% CI 1.0, 1.6) compared to abatacept, and tocilizumab initiators had a lower risk (adjusted HR 0.64 95% CI 0.41, 0.99).⁵⁶

RITUXIMAB

Rituximab is currently only FDA approved for RA and not for other systemic arthritides. It is a chimeric monoclonal antibody specific for B-cell specific cell-surface marker, CD20, and thus depletes B-cells. Given the mechanism and long half-life, it is typically dosed every 6 to 12 months. Limited data exist on the cardiovascular benefit of B-cell depleting therapy. In a small analysis, 6 patients with RA who were refractory to TNFi therapy, rituximab was able to improve endothelial function detected by flowmediated dilation.⁵⁷ Larger studies on B-cell inhibition are needed; B cells are also known to play atheroprotective roles and it is thus plausible that CV risk modulation could be neutral or increased.⁵⁸

INTERLEUKIN-17 AND IL12/23 INHIBITION

Recent data have emerged on the pathogenic role of the Th17/IL17 axis in psoriatic disease and spondyloarthropathies. Although high levels of IL-17A and Th17 cells have been reported in RA, the response to therapy has been mixed and thus are not approved agents for RA. Several anti-IL-17 inhibitors have been developed and are approved for psoriatic arthritis and spondyloarthropathies. These include antiIL-17A monoclonal antibodies, secukinumab and ixekizumab. IL-12/23 inhibition also blocks the Th17 response and is achieved through targeting p40, a subunit shared by IL-12 and IL-23.⁵⁹ IL-12/23, e.g. ustekinumab, and IL-23 inhibitors, e.g. geselkumab, are only approved for psoriatic disease as secondary endpoints were not met for spondylarthritis. Thus far, the data on IL-17 and IL-23 inhibition and cardiovascular benefit have been reported in psoriatic disease and not for PsA specifically or for SpA.

While psoriasis was not a focus of this review, we will briefly cover data on bDMARD studies in psoriasis, for treatments also approved for PsA, examining cardiac imaging data as a surrogate marker of CV risk. Two randomized double-blind placebocontrolled trials examined changes in aortic inflammation as measured using FDG PET of the aorta before and after treatment with the IL-12/23 blocker, ustekinumab, the Vascular Inflammation in Psoriasis (VIP)-U study and another with IL-17A blocker, secukinumab, VIP-S; all subjects had moderate to severe psoriasis. Treatment with ustekinumab in the VIP-U study resulted in a reduction in aortic inflammation measured by FDG PET at 12 weeks (6.6% versus 12.1% in placebo, p = 0.001) yet differences were not seen at 52 weeks; similarly, no differences were observed with secukinumab therapy at 52 weeks,^60,61 suggesting the effects of these treatments on CV risk may be neutral.

Similar to the theme of CV biomarkers, another prospective study examined the role of ustekinumab (IL-12/23 inhibitor) in comparison with TNFα inhibition or cyclosporine therapy in 150 subjects with moderate psoriasis and measured their effects on left ventricular remodeling, the coronary microcirculation (coronary flow reserve (CFR) by LAD doppler echocardiography), arterial stiffness, and biomarkers of oxidative stress and inflammation. There was successful resolution of psoriasis skin disease among all therapies; interestingly, IL-12/23 inhibition displayed a greater improvement in left ventricular strain, arterial stiffness and Doppler assessed LAD CFR.⁶² In addition to potential benefit, safety concerns have also been raised predominantly in psoriasis but are relevant in this discussion. An IL-12/23 inhibitor, briakinumab (an IL-12/23 inhibitor) which has the same mechanism of action as guselkumab and ustekinumab approved for PsA, was halted due to safety concerns after one of the 4 clinical trials reported an increase in MACE events.^63,64 In phases II and III placebo-controlled studies for ustekinumab, a total of 5 MACE events occurred all in the ustekinumab arm and these events occurred in patients with at least three cardiovascular risk factors. Unfortunately, the lack of a control group will preclude definitive assessment with long-term follow-up.^65,66 Furthermore, in a recent nationwide cohort study of patients with PsA, a total of 9510 new users of bDMARDs were included and the risk of MACE was assessed from 2015 to 2019. The majority of patients initiated a TNFi (7,289) compared to IL12/23 (1,058), IL-17 (1163), and apremilast (1885) users with a total of 51 (0.4%) MACE events captured. After propensity score weighting, the risk of MACE was greater in patients with IL12/23 (HR 2.0, 95%CI 1.3–3.0) and IL17 (HR 1.9, 95%CI 1.2–3.0) inhibitors compared to TNFi, with no significant increased risk with apremilast.⁶⁷ On the other hand, a meta-analysis encompassing these studies has been performed that incorporates nine independent double-blind, randomized, clinical trials to further assess an association of MACE with psoriatic treatment which did not reveal any significantly observed differences.^68,69

It is clear that further data are required to clarify the role of these agents in CV risk mitigation. In atherogenic mouse models, IL-17 has been shown to be proatherogenic and functional blockade of IL-17A reduced plaque vulnerability and inflammatory cellular infiltration and cytokine expression.^70–72 Conversely, IL-17 has also been proposed in other models to have antiatherogenic roles.^73–75 These results highlight the context dependence and further highlight that we should exhibit caution in the extrapolation of data to a specific inflammatory disease.⁷⁶

JANUS KINASE INHIBITORS

Janus kinase (JAK) is a family of nonreceptor tyrosine kinases and includes JAK1, JAK2, JAK3, and TYK2 (tyrosine kinase 2) which transmits signals through the signal transducer and activator of transcription (STAT). There are 3 JAK inhibitors approved for use in autoimmune diseases: tofacitinib, baricitinib, and upadacitinib and all are oral small molecules. The JAK-STAT kinase system is implicated as an important pathway for RA, PsA, and SpA and functions upstream of the cytokine cascade.⁷⁷Tofacitinib is currently approved in RA and PsA, and is a nonspecific JAKi, targeting JAK 1, 2, and 3. Baricitinib targets JAK1 and 2. The selective JAK1 inhibitor, upadacitinib, is approved in RA and is currently being evaluated in PsA.⁷⁸ Currently, data regarding the cardiovascular safety of JAKi in these inflammatory conditions remain inconclusive. On one hand, there have been six phase III studies and two open-label long-term extension studies of tofacitinib in patients with RA which have shown overall low incidence of cardiovascular events.⁷⁹ A meta-analysis of 26 RCTs did not demonstrate significant differences in CV events rate in RA patients treated with JAK inhibitors.⁸² Tofacitinib is associated with increases in the lipid parameters; yet the relevance of this on cardiovascular risk is less clear. A small mechanistic prospective demonstrated that tofacitinib could positively affect atherosclerosis by regression of carotid intima media thickness in patients who had elevated levels of baseline.⁸⁴ However, more recently, the FDA issued a black box warning of all JAK inhibitors which have added ambiguity after the results of the ORAL-Surveillance study was released.⁸⁴ The black box warning is specifically for increased risk of adverse cardiovascular events, malignancy, thrombosis and death in RA. ORAL Surveillance was a prospective, phase 3b/4 randomized, open- label, non-inferiority, study that compared tofacitinib and TNFi, which was mandated by the FDA as part of a post-marketing safety study. This trial enrolled RA patients with inadequate response to MTX, age>50, and at least one cardiovascular risk factor (cigarette smoking, hypertension, hyperlipidemia, DM, family history of premature CAD, prior history of CAD, or extra-articular disease associated with RA). The primary analysis demonstrated increased incidence rates for VTE and MACE when compared with the TNFi group and reached the threshold for a safety signal. This in turn, led to the FDA issuing a black box warning for tofacitinib for the treatment of RA. Increased risk of MACE was higher among patients >65 years, those who had ever smoked or aspirin users. Malignancy rates were also numerically higher compared to the TNF inhibitor group. Whether these results are specific to tofacitinib or specific to RA remains unclear. A meta-analysis that included ORAL-Surveillance analyzed a total of 66 RCTs across inflammatory conditions found that JAK inhibitors had a numerically higher rate of VTE when compared with controls (OR: 1.65; 95% CI: 0.97–2.79) and primarily driven by the studies which had follow-up of greater than 12 months. The increased risk of VTE was observed when compared with active comparators but not with placebo. The results of MACE were similar but did not reach statistical significance.⁸⁵ In a study using real-world data comparing tofacitinib vs TNFi with similar parameters as ORAL-Surveillance, no increased CV risk was observed in the overall population; a non-significant trend toward increased CV risk was observed among RA patients initiating tofacitinib with CV risk factors, similar to ORAL-Surveillance.⁸⁶ Thus, future clinical trials and mechanistic investigations are needed to understand the role of JAK inhibition in promoting or mitigating cardiovascular disease.

SUMMARY

Overall, the current evidence suggests that the use of bDMARDs mitigate CV risk in patients with inflammatory arthritis; however, the benefit may not be true for all bDMARDs and small molecules. The mixed findings are likely the result of the sample study, the difference in surrogate endpoints and MACE definitions, and heterogeneity of the patient population, and the different pathways targeted. Patients with systemic arthritides and established CV risk factors or known CVD are likely a different patient population compared to systemic arthritis patients without established CV risk factor. As more patients are prescribed these newer therapies, more data will allow us to decipher how to select the best therapy for the individual patient. The emergence of noninvasive cardiovascular imaging modalities that are linked to CV outcomes, such as cardiac PET may serve as important tools that can be employed to help address some of these questions. The integration of a care model with both cardiorheumatology, a growing field, embedded within routine rheumatologic care will allow a patient-centered approach with risk counseling and individualization of therapy.^8,80,81,83 The rapid expansion of bDMARDs and small molecules for the treatment of inflammatory arthritis has provided a growing armamentarium for the rheumatologist. Along with new CV imaging tools, the specificity of each inflammatory pathway and the effect on coronary vascular health and cardiovascular risk can now be addressed.

KEY POINTS

• Overall current evidence suggests that bDMARDs mitigate CV risk in inflammatoryarthritis.

• RA is the most well-studied but emerging evidence for SpA and PsA suggest similarfindings.

• The role of JAK inhibition on CV Risk mitigation is inconclusive.

CLINICS CARE POINTS

• Biologic dMARDS are highy efficacious to control disease activity in inflammatory arthrits.

• The translation of bDMARD to migitation of CV risk is overall favorable although currently mixed findings suggest that not all bdMARDs and small molecules benefit CV risk mitigation equally.