Anti-Cytokine Immune Therapy and Atherothrombotic Cardiovascular Risk

Abstract

Accumulating observations in humans and animals indicate that inflammation plays a key role in atherosclerosis development and subsequent complications. Moreover, the use of loss- or gain-of-function genetically modified, atherosclerosis-prone mice has provided strong experimental evidence for a causal role of innate and adaptive immunity in atherosclerosis, and has revealed the pathogenic activity of pro-inflammatory cytokines, including TNF-α, IL-1β, IL-6, and IL-18, and the athero-protective effect of anti-inflammatory cytokines, including IL-10 and TGF-β. For the last fifteen years, treatments using monoclonal antibodies specifically targeting cytokines, commonly referred as biological therapies, have transformed the treatment of chronic inflammatory diseases such as rheumatoid arthritis or psoriasis, both conditions associated with increased cardiovascular risk. Analyzing the impact of anti-cytokine therapies on the cardiovascular outcomes of patients with chronic inflammatory diseases provides insight into the clinical relevance of experimental data on the role of inflammation in atherothrombotic cardiovascular diseases. CANTOS provided the first evidence that targeting inflammation in humans with atherosclerosis could improve clinical outcomes. Treatment with the anti-IL-1β antibody canakinumab significantly reduced recurrent cardiovascular events in individuals with stable coronary artery disease well-treated with standard-of-care measures. Other clinical studies support the protective effects of treatment with anti-TNF-α and anti-IL-6 receptor monoclonal antibodies on cardiovascular risk. Blockade of the IL-23/IL-17 axis, however, warrants caution as a cardiovascular intervention. Targeting this pathway has improved psoriasis, but may augment cardiovascular risk in certain patients. Thus, careful consideration of the cardiovascular risk profile may influence the choice of the most appropriate treatment for patients suffering from chronic inflammatory diseases.

Introduction

Since the beginning of the 20th century, animal experiments have been instrumental in understanding the pathophysiological mechanisms of atherosclerosis[1] Following initial studies in hypercholesterolemic rabbits, the development of genetically engineered mice (e.g. apolipoprotein E (apoE) or low-density lipoprotein receptor (Ldlr)-deficient mice) permitted identification of the participation of innate and adaptive immune responses in atherosclerosis development and progression. The atherosclerotic process begins with the retention of lipoproteins, including LDL, in the arterial intima, where they can undergo oxidative modifications, providing one stimulus to endothelial activation and recruitment of CD4⁺ T-lymphocytes and monocytes/macrophages that can become foam cells after uptake of cholesterol from lipoproteins. The immune/inflammatory cells recruited into the artery wall communicate amongst themselves and with vascular cells through membrane contacts and secretion of cytokines [2] . Mouse experiments suggest that pro-inflammatory cytokines, including TNF-α, IL-1β, IL-6, IL-12p70, or IL-18 promote atherogenesis, while anti-inflammatory cytokines such as IL-10, TGF-β, and IL-33 protect against atherosclerosis [3] Yet, the clinical relevance of these experimental findings has long remained unclear [4] Among other considerations, the concentrations of cholesterol in typical mouse experiments exceeds by more than 10-fold that currently encountered in human patients. The mice do not receive as background therapy the medications now mandated for humans with established atherosclerosis.

The introduction of monoclonal antibody-based immunotherapy targeting cytokines, and of small molecules that target inflammatory pathways, provides opportunities to address this issue. The CANTOS trial first used this approach to test the effect of canakinumab, a monoclonal antibody that targets IL-1β, in patients with a prior heart attack and above-median hsCRP despite concomitant guideline-directed care [5] Other monoclonal antibody-based immunotherapies have achieved success in the treatment of chronic immune-inflammatory diseases such as psoriasis, rheumatoid arthritis, ankylosing spondylitis, or Crohn’s disease, all conditions characterized by enhanced cardiovascular risk compared to the general population [6 , 7] A large British study using the database of general practitioners reported a significantly increased risk for myocardial infarction in psoriasis patients compared to the general population; the risk related to the severity of the psoriasis. The relative risk, adjusted to traditional cardiovascular risk factors, approached 3 for the age range between 20 - 30 years, compared to the control population [8] Psoriasis patients also have an increased risk of stroke [6] Rheumatoid arthritis patients also have an elevated cardiovascular risk [9, 10] comparable to that of diabetic patients. This risk assumes even greater importance in regards to the extra-articular manifestations of these diseases, affecting the skin, lung, and renal systems [11] In these chronic inflammatory diseases, the enhanced cardiovascular risk persists after adjustment for classic cardiovascular risk factors, suggesting that other contributors, such as chronic systemic inflammation, aggravate cardiovascular disease. This review will discuss the consequences of targeted immune therapies on cardiovascular athero-thrombotic events (Table 1).

Table 1 :

Summary of athero-thombotic cardiovascular effects of anti-cytokine therapy according to the studied population. Athero, atherosclerosis; MACE, major adverse cardiovascular event; MI, myocardial infarction.

	Agent	Animal model or Patients	Pre-clinical data	Pilot clinical data	Outcome trial data	References
TNF-α	Tnf-α-null TnfR1(P55)-null	Mice	Decreased athero. Increased athero.	-	-	Canault, 2004; Ohta, 2005 Schreyer, 1996
	Adalimumab (n=52) Etanercept (n=36) Infliximab (n=32)	Psoriatic arthritis (prospective study)	-	Decreased athero. (carotid artery)	-	Di Minno, 2011
	Etanercept or Infliximab (n=531)	Rheumatoid arthritis (retrospective study)	-	-	Reduced MACE	Jacobsson, 2005
	TNF-α inhibitors (n=959)	Psoriasis Retrospective	-	-	Reduced MACE	Ahlehoff, 2015
	TNF-α inhibitors (n=6864)	Rheumatoid arthritis (retrospective study)	-	-	Reduced MACE in responders to TNF-α inhibitors	Ljung, 2016
	TNF-α inhibitors (n=9148)	Psoriasis (retrospective study)	-	-	Reduced MACE compared to Methotrexate	Wu, 2017
	Infliximab (n=51)	Heart failure (prospective randomized study)	-	-	worsened heart failure	Chung, 2003
IL-17	Il-17a- or Il-17r-null anti-IL-17 or soluble IL-17 receptor A adenovirus-treatment	Mice	Decreased athero.	-	-	Ge, 2013 Van Es, 2009 Erbel, 2009 Smith, 2010
	Il-17a-null Th17 overexpression	Mice	Increased athero. Stabilized plaque	-	-	Danzaki 2012 Taleb, 2009 Gistera, 2013
	Secukinumab (n=587)	Psoriatic arthritis (prospective randomized study)	-	-	Non-significant increased MACE	Mease, 2015
	Secukinumab (n= 587)	Ankylosing spondylitis (prospective randomized study)	-	-	Non-significant increased MACE	Baeten, 2015
IL-12/ IL-23	Il-12-null Il-23-null	Mice	Decreased athero. No effect on athero.	-	-	Davenport, 2002 Engelbertsen, 2018
	Briakinumab (n=2520)	Psoriasis (prospective randomized study)	-	-	Increased MACE	Langley, 2011
	Ustekinumab (n=178)	Psoriasis	-	-	No effect on MACE	Ahlehoff, 2015
	Briakinumab or ustekinumab (n= 3179)	Psoriasis (prospective study)	-	-	Non significant increased MACE	Ryan, 2011
	IL-12/IL23 inhibitors (n=9290)	Psoriasis (retrospective study)	-	-	Increased MACE in patients with high CV risk	Poizeau, 2018
IL-6	Il-6-null IL-6 supplementation	Mice	Increased late athero. Increased athero.			Schieffer, 2004 Huber, 1999
	Tocilizumab	Rheumatoid arthritis (prospective randomized study)		No MACE	No difference in MACE compared to TNF-a inhibitors-treated group BUT increased LDL and HDL	Kim, 2017 Kleveland, 2016
IL-1β	Il-1β null mice Il-1β supplementation	Mice Pigs	Decreased athero. Accelerated athero.			Kirii, 2003 Shimokawa, 1996
	Canakinumab	Post-MI patients + High hsCRP (prospective randomized study)			Decreases MACE	Ridker, 2017

TNF-α blockade

Human atherosclerotic plaques contain TNF-α [12] and TNF-α plasma levels associate with recurrence of major cardiovascular events (MACE) in patients with acute coronary syndromes, even after adjustment to conventional cardiovascular risk factors [13] The deletion of the Tnf-α gene reduced the development of atherosclerosis in apoE^−/− mice fed a high-fat diet [14] as well as in C57B16 mice fed a high-fat diet containing cholate [15] However, study of mice deficient for the p55 type 1 TNF-α receptor (TNF-R) have yielded conflicting results [16] For the last twenty years, anti-TNF therapy has afforded a great advance in the treatment of rheumatic diseases, using anti-TNF-α monoclonal antibodies that bind specifically to human TNF-α with high affinity, and neutralize its biological activity (infliximab, adalilumab, certolizumab pegol, golimumab) or soluble TNF receptor fusion proteins (etanercept). In psoriatic arthritis patients, anti-TNF-α monoclonal antibodies reduce the development of carotid atherosclerotic plaques, measured by ultrasound in non-randomized observations. After more than 4 years of treatment, 15.8% of the patients treated with anti-TNF-α antibodies presented carotid lesions vs. 40.4% of patients receiving DMARDs (Disease-Modifying Antirheumatic Drugs) and non-selective immunomodulators including sulfasalazine, methotrexate, cyclosporine, and leflunomide (P<0.0001) [17] Positive vascular effects of anti-TNF-α antibodies associated with improved clinical outcomes. Residual confounding, however, limits the rigor of such observational studies. Over a 2-year period, Jacobsson et al. compared a cohort of patients with rheumatoid arthritis non-randomly treated by anti-TNF-α (n=983) to a control population. The incidence of the first cardiovascular event fell significantly among patients receiving anti-TNF-α (after adjustment to age and sex, odds ratio 0.46 (95% CI (0.25-0.85), p = 0.013) [18] Anti-TNF-α antibodies also demonstrated a beneficial effect in a Danish cohort of psoriasis patients who displayed a relative adjusted risk of 0.46 (0.22-0.98, P=0.04) compared to the non-randomized control group treated with other interventions (methotrexate, cyclosporine, retinoid, phototherapy) [19] Notably, at various stages of rheumatoid arthritis disease progression, the beneficial effect attributed to anti-TNF-α treatment on the cardiovascular risk associated with improvement in joint response. Patients who responded positively to anti-TNF-α treatment, as assessed by reduced joint symptoms, showed decreased cardiovascular risk that approximated that in the general population. Non-responding patients had high remaining cardiovascular risk [20] Altogether these clinical data from non-randomized observational studies suggest that TNF-α acts as a pro-atherogenic cytokine, and that its pharmacological blockade might reduce the risk of atherothrombotic complications.

Based on experimental and clinical data, European experts recommended the use of either methotrexate or a TNF-α-blocking agent for treatment of patients with severe psoriasis at high cardiovascular risk [21] The putative protective effects of these two treatments on cardiovascular risk likely differ. Binding of adenosine to A2 and A3 receptors may contribute to the anti-inflammatory actions of methotrexate [22] The effects of methotrexate on cardiovascular risk vary from one study to another. Prodanovich et al. followed for 5 years a population of patients presenting with either psoriasis (n=7615) or rheumatoid arthritis (n=6707) and reported a significant reduction of MACE under methotrexate, mostly for low cumulative doses, compared to patients receiving other DMARDs on a non-randomized basis (OR 0.50 (0.31-0.79), P<0.01) after adjustment to the conventional cardiovascular risk factors [23] However, the CORONA registry that followed more than 10,000 patients with rheumatoid arthritis during a 24-month period showed no such protective effect [24] An American registry study including patients presenting with severe psoriasis reported that those treated with anti-TNF-α therapy had fewer MACE than those non-randomly receiving methotrexate [25] More recently, a CIRT-randomized trial reported the lack of a protective effect of low doses of methotrexate (15-20 mg/week) on MACE in patients with previous MI or multivessel coronary artery disease in addition to either type 2 diabetes or metabolic syndrome, or both. All participants took folic acid daily (N=4786). The study stopped prematurely for futility [26] In contrast to CANTOS, the entry criteria for CIRT did not specify selection by hsCRP. CIRT enrollees had normal hsCRP concentrations at entry (median hsCRP, 1.5 mg/L). The low-dose methotrexate treatment had no effect on CRP nor on other inflammatory biomarkers, including plasma IL-1β and IL-6. This disparity in baseline inflammatory burden may account for the ineffectiveness of low-dose methotrexate in forestalling cardiovascular events [27] Indeed, the consistent apparent cardiovascular benefit of low-dose methotrexate in observational studies may reflect an indirect effect of the agent in quelling inflammation due to the primary rheumatologic or cutaneous disease, removing a driver for atherothrombosis. Indeed, systemic or remote extravascular inflammation can promote local inflammation in the “prepared soil” of the atheroma, a phenomenon we have referred to as an “echo effect” [9, 28]

Overall, these data suggest that anti-TNF-α therapy provides better cardioprotection than methotrexate, although the randomized, double-blind design of CIRT and its rigorous adjudication of cardiovascular events provides a much more reliable assessment of cardiovascular outcomes than registry or other observational assessments. Yet, despite considerable preclinical data and pilot human studies, TNF neutralization did not only fail to improve heart failure outcomes in non-arthritic patients, but may have caused harm [29] in properly powered studies. Thus, patients with heart failure should not receive anti-TNF-α therapy. This recommendation emerged from the results of the ATTACH study. This multicenter, randomized controlled study showed that high doses of anti-TNF-α exacerbate the risk of death and/or of re-hospitalization in patients with moderate to severe heart failure (FEVG<35% stage III or IV NYHA) [30] The mechanisms behind these deleterious effects of TNF-α inhibitors remain uncertain. Activation of the two surface receptors for TNF-α, TNFR1 and TNFR2, display opposite effects. In experimental heart failure, signaling through TNFR1 engagement had cardiotoxic activity, while TNFR2 ligation mediated cardioprotective effects by dampening NF-κB activation and inflammation, as well as limiting cardiac apoptosis [31, 32] .

Patients with lupus, obesity, insulin resistance/metabolic syndrome, or type 2 diabetes who have high cardiovascular risk exhibit elevated plasma TNF-α. Yet, the use of anti-TNF agents in these conditions yielded inconsistent results, perhaps depending on patient characteristics. In addition, some cases of lupus-like diseases following anti-TNF-α therapy have been reported [33] Moreover, obesity associates with a 60% higher risk of failing anti-TNF-α therapy in patients with rheumatoid arthritis or psoriasis [34]

IL-17 blockade

Immune cells, including CD4⁺ T lymphocytes, NK cells, NKT cells, as well as vascular cells, produce the cytokine IL-17, whose major isoform is IL-17A. IL-17A participates in host defense against bacterial and fungal infections [35] In humans, mutations that cause defects in IL-17A or dysfunctional IL-17 receptors increase susceptibility to infections [36] IL-17 also contributes to several chronic inflammatory diseases, including psoriasis, rheumatoid arthritis, and Crohn’s disease [37 , 38] .

CD4⁺ T helper lymphocytes that produce IL-17, denoted Th17 cells, express the transcription factor STAT3 and the nuclear receptor RORγt. In vitro, the combination of TGF-β and IL-6 promotes the differentiation of naïve CD4⁺ T lymphocytes into Th17 cells [39] IL-6 can induce the expression of the IL-23 receptor, and IL-23 in turn amplifies the expression of its own receptor [40] and stimulates the expansion of Th17 cells [41] IL-6, IL-21, and IL-23 induce IL-17 production by activation of STAT3 [42] STAT3 directly controls the expression of other transcription factors that participate in Th17 cell differentiation, including RORγt [43] .

The role of IL-17 in atherosclerosis remains controversial. Several studies reported a pathogenic effect of IL-17 in atherosclerotic mice, either deficient for the Il-17a gene [44] or the IL-17 receptor [45] or treated with anti-IL-17 monoclonal antibodies [46] or adenovirus producing a soluble IL-17 receptor A [47] The pro-atherogenic mechanisms of IL-17 may relate to enhanced recruitment and activation of myeloid cells in the intima, associated with increased release of chemokines and pro-inflammatory cytokines [48] .

Other studies, however, have shown protective effects of IL-17A by use of loss- [49] or gain-of-function experiments [50 , 51] attributed to decreased endothelial expression of VCAM-1 and induction of a plaque with characteristics associated with stability in humans. Moreover, IL-17 can stimulate collagen synthesis by vascular smooth muscle cells, and increased levels of IL-17 in human carotid plaques associated with characteristics of plaque stability [50 , 51] . Moreover, in a cohort of more than 1,000 patients with acute coronary syndrome, below the median plasma IL-17 predicted the recurrence of MACE within one year, after adjustment for cardiovascular risk factors, treatments, and CRP [52] Altogether, these data suggest that the role of IL-17 in atherosclerosis plaque formation might depend on context. Patients with high cardiovascular risk treated with inhibitors of the IL-17 pathway may thus merit intensive preventive efforts.

Today, anti-IL-17 immunotherapy is used to treat patients with arthritis or psoriasis. The first study evaluating treatment with an anti-IL-17A monoclonal antibody, secukinumab, or placebo in psoriatic arthritis patients during a 52-week monitoring period reported one cardiovascular event in the placebo group (n=202) vs. one death of cardiovascular origin and 15 arterial atherothrombotic events (8 cardiac disorders and 7 cerebral disorders) in the group treated with anti-IL-17A antibody (n=587) [53] . Secukinumab was also tested in ankylosing spondylitis. No MACE were observed in the placebo group (n=196) whereas patients treated with anti-IL-17A antibody (n=587) sustained 8 events (4 cardiac disorders, 2 cerebral disorders, and 2 gut disorders) [54] Even though the difference in adverse cardiovascular outcomes was not statistically significant in these studies, they sound a note of caution, especially in light of conflicting preclinical data. The ongoing post-marketing registry results should soon shed light on the long-term cardiovascular effects anti-IL-17 immunotherapies, and should provide more definitive information regarding cardiovascular safety than the smaller studies that may be underpowered for this assessment.

IL-23 blockade

The cytokine IL-23, a member of the IL-12 family consisting of the two subunits IL-12p40 and IL-23p19, may participate in several chronic inflammatory diseases [55] IL-23 maintains Th17 functions [56] IL-23 localizes in atherosclerotic plaques of mice and humans [57] Studies in atherosclerotic mice have yielded conflicting findings. IL-23 receptor deficiency in Ldlr^−/− mice had no effect on atherosclerosis [58] However, another mouse study suggested that IL-23 protected against atherosclerosis by controlling gut dysbiosis and pro-atherogenic bacteria expansion, and by maintaining intestinal barrier function [59] IL-23’s protective effects depended in part on the production of IL-22.

Monoclonal antibodies that target IL-12p40/IL-23p19 have undergone evaluation to treat severe forms of psoriasis. The anti-p40 antibodies that target both IL-12 and IL-23 are briakinumab and ustekinumab. Briakinumab, a fully human monoclonal antibody, showed very promising early results but was withdrawn from the market due to frequent MACE during phase I and II (27/2520 patients) [60] In a meta-analysis of 22 randomized controlled trials that compared treatments with anti-TNF-α agents (etanercept, infliximab, adalimumab) or anti p40-antibodies (briakinumab, ustekinumab), only one MACE was reported in the placebo group (n=1812) and one in the group of patients treated with anti-TNF-α (n=1474), while the group treated with anti-IL12/IL-23 antibodies sustained 10 MACE (n=3179) [61] Even though the numerical imbalance in MACE was not statistically significant between the groups receiving the biologic drug and the placebo groups, the trend for increased MACE resembled that which was reported with IL-17 pathway inhibitors. Here again, the statistical power may not have sufficed for safety evaluation. Another meta-analysis using a statistical methodology better adapted to identify rare events (Peto Method) found an increased risk of MACE under anti-IL-23 with an OR at 4.23 (CI 1.07–16.75 ; p=0.04) [62] A recent study presented at the French dermatology meeting on the SNIRAM (Système National d’Information Inter-régimes de l’Assurance Maladie) database confirmed this increase of MACE in individuals receiving anti-IL-23 therapy, mainly in patients with high cardiovascular risk [63] A study evaluating the effect of ustekinumab in patients with moderate-to-severe psoriasis reported no MACE, but the size of study was very small (n=40) [64] A phase III study that compared anti-IL-17 receptor A antibody (brodalumab) with anti-p40 (ustekinumab) immunotherapy in psoriasis [65] found no difference in MACE between groups. Finally, in a pooled safety analysis of 4 trials, Papp et al. reported a comparable overall rate of MACE in ustekinumab-treated patients (0.44 /100 PY) and those receiving anti-TNF agents for moderate-to-severe psoriasis (range 0.36–0.84 /100 PY) [66, 67]

Three monoclonal antibodies that selectively inhibit the IL-23p19 subunit, guselkumab, tildrakizumab, and risankizumab, are in development for the treatment of moderate-to-severe psoriasis. In a phase 2 trial, selective blockade of IL-23 with risankizumab associated with better outcomes than ustekinumab, but the trial was not large enough or of sufficient duration to draw conclusions about the occurrence of MACE [64]

RORγt, a key regulator of Th17 cell differentiation, may be a drug target for IL-23/Th17-related autoimmune diseases. Several small-molecule inhibitors of RORγt, RORγt “inverse agonists,” reported by groups in industry and academia [68] may represent new therapeutics in arthritis and psoriasis.

IL-6 blockade

The central cytokine IL-6 mediates acute and chronic inflammatory responses, and regulates the acute phase response in hepatocytes which includes CRP synthesis [69] Vascular endothelial and smooth muscle cells produce IL-6, as do macrophages and T and B lymphocytes [2 , 70] The vascular effects of IL-6 vary according to experimental conditions. In C57B1/6 and in apoE^−/− mice, injection of recombinant IL-6 at supra-physiological doses worsened atherosclerosis [71] However, at an early stage, the deletion of the Il-6 gene in apoE^−/− mice had no effect on the disease compared to control animals, while it enhanced atherosclerosis at later stages [72] The late protective effects of IL-6 in mouse experiments might result from the synthesis induction of endogenous cytokine antagonists such as IL-1ra and soluble TNF-α receptor [73]

Human atherosclerotic plaques contain IL-6 [74] and high IL-6 plasma levels associate with worse outcomes in people without manifest cardiovascular disease [75] and in patients with acute coronary syndromes [76 , 77] A large-scale collaborative genetic association analysis of 34 studies including 25,458 coronary heart disease cases and 100,740 controls of IL6R gene polymorphisms with cardiac events found that the IL6R rs7529229, which associates with elevated plasma levels of soluble IL-6R and lower CRP levels, also associated with reduced coronary artery disease events [78] These concordant findings from two independent genetic studies establish a causal relationship between IL-6 signaling and atherosclerosis in humans. Thus, IL6R represents an attractive potential therapeutic target for coronary artery disease.

Tocilizumab, an IL-6 receptor-blocking monoclonal antibody, has received approval to treat rheumatoid arthritis. This molecule affects the lipid profile adversely. Tocilizumab increases LDL-cholesterol (+18%), triglycerides, and HDL-cholesterol (+9%) levels [79] Tocilizumab also reduces lipoprotein Lp(a) levels. Using three American databases, a recent study sought to compare the incidence of MACE during an average monitoring of 1 year in patients with rheumatoid arthritis treated either with tocilizumab (n=9218) or an anti-TNF-α antibody (n=18810). When applying a propensity score, no difference in MACE was found between the two groups [80] Similar results were reported in Italy [81] and in Japan [82] Overall, it appears that IL-6 blockade does not increase cardiovascular risk and possibly even protects against MACE occurrence, although it entails a significant increase in LDL-cholesterol and triglyceride concentrations. In a small, randomized, double-blinded, placebo-controlled trial, patients with MI were treated with a single dose of tocilizumab (n=58) or placebo (n=59). No MACE occured 30 days after enrollment [83 , 84] The ongoing phase II ASSAIL-MI trial in patients with acute coronary syndrome will evaluate the ability of a single administration of tocilizumab to reduce myocardial damage.

IL-1β blockade

The inactive pro-form of IL-1β undergoes cleavage by caspase 1 to attain biological activity [85] A number of pro-inflammatory cytokines induce IL-1β production including TNF-α [86] Cholesterol crystals can co-activate the NLRP3 inflammasome that controls caspase 1 [87 ] Indeed, the inflammasome represents an attractive future target for the interruption of the IL-1β-IL-6 pathway upstream. In animal experiments, IL-1β exerts pro-atherogenic effects. Il-1β genetic deficiency in mice reduced atherosclerosis development [88] Conversely, repeated perivascular administration of recombinant IL-1β accelerated vascular disease in pigs [89] by increasing the production of pro-inflammatory cytokines, chemokines including CCL-2, and by stimulating endothelial expression of adhesion molecules such as VCAM-1 [90 , 91]

A human monoclonal antibody that targets interleukin-1β, canakinumab, has approval for the treatment of rare inflammasome gain-of-function diseases (e.g. Muckle-Wells syndrome), and has seen use for rheumatic inflammatory chronic diseases, such as juvenile chronic arthritis or Familial Mediterranean Fever [92 , 93] In a phase IIb randomized trial including diabetic patients at high cardiovascular risk, Ridker et al. showed that treatment with canakinumab did not alter cholesterol plasma levels, but reduced CRP and IL-6 plasma concentrations in a dose-dependent manner [94] More recently, in the international randomized trial CANTOS (N=10,061) with a median follow-up of 3.7 years, the same group showed for the first time that anti-cytokine therapy could reduce major cardiovascular events in patients with stable coronary artery disease and high-sensitivity CRP > 2mg/L despite optimal medical treatment. In an on-treatment analysis, participants who achieved hsCRP concentrations less than 2 mg/L or IL-6 levels below the study median value of 1.65 ng/L after 3 months showed the most benefit from canakinumab [95] Canakinumab treatment also lowered cancer incidence and mortality in exploratory analyses, reduced gout attacks by about half, and improved heart failure outcomes as well [96, 97] These benefits came at the cost of a small but significant increase in infections. CANTOS provided the first proof in humans that targeting inflammation can reduce cardiovascular event rates independently of lipid lowering. The cost-effectiveness of canakinumab used to treat CVD was evaluated in the perspective of the Australian [98] and US public healthcare systems referring to the orphan drug pricing [99] Canakinumab will not be commercialized for cardiovascular indications, rendering such pharmacoeconomic analyses moot.

Targeting the NLRP3 inflammasome could be another option to block the IL-1β, as well as IL-18, pathways. Colchicine, which acts by preventing cytoskeletal microtubule formation and inhibition of the NLRP3 inflammasome, has been tested in the Low-Dose Colchicine (LoDoCo) study for secondary prevention of CV events in 532 patients [100] Results were promising, but there was a high rate of gastrointestinal intolerance to colchicine. Colchicine is currently being tested in much larger studies of patients with stable CAD (Low-Dose Colchicine2 study and Colchicine Cardiovascular Outcomes Trial). Finally, specific small-molecule inhibitors of NLRP3, like MCC950, have also been developed to treat a variety of inflammasome-driven diseases. There are currently plans to assess MCC950 in atherosclerotic patients.

Perspectives

Monitoring of cardiovascular side effects of immunotherapies has assumed increasing importance for the future as novel anti-cytokine monoclonal antibodies in development will undergo testing in patients with chronic inflammatory diseases or cancers. The pipelines of pharmaceutical and biotechnology companies include anti-IL-4, -IL-13, -IL-18, -INF-γ, -IL-21, and -IL-22 antibodies [101] Antibodies that target co-stimulatory molecules such as CD40, CTLA-4, or PD-1/PDL-1 are also under evaluation for treatment of chronic inflammatory diseases, after proving their efficacy in cancer [102] Checkpoint inhibitors, however, can cause often fatal fulminant myocarditis [103] Although beyond the scope of this review, it is noteworthy that cell-directed immunotherapies using depleting monoclonal antibodies against leukocyte-restricted cell surface antigens can treat hematopoietic cancers and auto-immune diseases. The anti-CD20 antibody rituximab that depletes mature B lymphocytes has wide use in rheumatology and haematologic diseases. In mice, B cell depletion can reduce the development and progression of atherosclerosis [104] and limit post-myocardial infarction cardiac remodeling [105] An ongoing phase I/II clinical study is treating patients with acute MI treated with rituximab (ClinicalTrials.gov ID: ). Finally, other immunomodulatory strategies directed at enhancing immune tolerance by low dose IL-2-induced expansion of CD4+ regulatory T cells [106] , or by vaccination against LDL-derived peptides [107] are currently being explored.

This plethora of immunomodulatory therapies provides an exciting vista for the future [27] We have now crossed the threshold of clinical translation. The task remains to balance the risk of impaired host defenses and tumor surveillance against anti-inflammatory efficacy. The information reviewed above indicates that we will have many opportunities to harness immunomodulatory strategies to lower cardiovascular risk. Only large prospective randomized trials will likely move the field forward from concept, laboratory investigation, and small clinical studies to the many patients who remain at cardiovascular risk despite the best of current care.

Highlights.

• In the era of immunotherapy, monitoring of cardiovascular side effects is essential to optimize patient outcomes

• CANTOS trial provided evidence that anti-IL-1β antibody treatment reduced recurrent cardiovascular events in patients with stable coronary artery disease well-treated with standard-of-care measures.

• Several clinical studies support the protective effect of treatment with anti-TNF-α and anti-IL-6 receptor monoclonal antibodies on cardiovascular risk in patients with rheumatoid arthritis or psoriasis.

• Beneficial effects of antibodies targeting the IL-23/IL-17 axis has been reported in patients with psoriasis, but caution regarding their cardiovascular safety is warranted, especially in patients at high cardiovascular risk.

Non standard abbreviations and Acronyms

CRP

C-Reactive Protein

DMARDs

Disease-Modifying Antirheumatic Drugs

IFN

Interferon

IL-

Interleukin-

LDL

Low density lipoprotein

MACE

Major Adverse Cardiovascular Event

TNF

Tumor necrosis factor

TGF

Transforming Growth Factor