Promises and challenges of targeting inflammation to treat cardiovascular disease: the post-CANTOS era

This editorial refers to ‘Residual inflammatory risk associated with interleukin-18 and interleukin-6 after successful interleukin-1β inhibition with canakinumab: further rationale for the development of targeted anti-cytokine therapies for the treatment of atherothrombosis’, by P.M. Ridker et al.†, on page 2153. Since its publication, the Canakinumab Anti-Inflammatory Thrombosis Outcomes Study (CANTOS)1 was perceived as the final proof of concept of more than two decades of research, pointing to atherosclerosis as a complex inflammatory disorder involving both adaptive and innate immunity.2,3 In addition to this, CANTOS represented the first step in the endorsement of anti-inflammatory therapies on top of intense lipid lowering in the prevention and treatment of cardiovascular disease (CVD). The aggressive control of risk factors has improved the prognosis of CVD. However, many patients still remain at high risk of acute events, clearly defining an area of substantial unmet medical need.4 For these patients, the CANTOS trial offers clinical evidence that targeting specific cytokine pathways, such as the interleukin (IL)-1β pathway, lowers the risk of acute events.1 Thus, inflammation should be incontrovertibly considered as a potential target for intervention. Despite the success of CANTOS, the clinical application of canakinumab faces numerous challenges. The risk reduction can be seen as modest and burdened by an increase in infections. This is associated with the high cost of canakinumab. A series of secondary analyses of CANTOS data identified subgroups of patients who have greater than average reductions in clinical events, revealing both the promises and the challenges of targeting inflammation to treat CVD.5 This highlighted the importance of understanding the mechanistic diversity of inflammatory pathways contributing to atherosclerosis. In this issue of the European Heart Journal, Ridker et al. pointed out the ‘residual risk’ remaining after canakinumab treatment that could be partially explained by the persistence of high levels of IL-6 and IL-18.6 The same authors recently showed that the magnitude of inflammation inhibition achieved by individual participants in the CANTOS trial was a major determinant of the clinical efficacy of canakinumab. In particular, individuals with IL-6 levels equal to or above the study median levels had no significant benefit for any of the endpoints.7 In the study of Ridker et al., among 4848 stable post-myocardial infarction patients enrolled in the CANTOS trial, who underwent IL-18 and IL-6 measurements before and after initiation of canakinumab, both IL-18 and IL-6 levels remained significant predictors of recurrent events. Compared with placebo, canakinumab significantly reduced IL-6 levels in a dose-dependent manner. As expected, no dose of canakinumab significantly altered IL-18 levels. The effects of IL-18 and IL-6 on the risk for future cardiovascular events were fully additive.6 Based on this analysis, the authors recommend further pharmacological development of potential anti-cytokine therapies for atherothrombosis that simultaneously inhibit IL-1β and IL-18, such as upstream Nod-like receptor (NLR)P3 inhibitors, as well as agents that directly target downstream IL-6 signalling. As highlighted by the authors, the NLRP3-inflammasome activates both IL-1β and IL-18. Thus, targeting IL-1β might not affect IL-18, while directly targeting the upstream inflammasome should achieve a greater anti-atherosclerotic effect than the use of IL-1β inhibition alone.8 Indeed, recent studies show that IL-1-induced left ventricular dysfunction is mediated by IL-18 and not by IL-6. Thus, IL-18 rather than IL-6 might be an additional therapeutic target in this setting.9 Moreover, the residual inflammatory risk before and after treatment with canakinumab is larger for IL-6 than for IL-18. This might reflect issues of biological amplification, but also supports the hypothesis that targeted IL-6 inhibition would exceed the beneficial effects observed in CANTOS. IL-6 is highly up-regulated at the site of coronary occlusion in patients with acute coronary syndrome (ACS), and higher circulating levels are associated with a worse outcome.10 These data provide elegant support for the notion that IL-6 is on the causal pathway leading to atherothrombosis. Atherosclerosis is a chronic inflammatory disease on its own, but a flare up of the inflammatory process within the atherosclerotic plaque may lead to plaque rupture and thrombus formation, resulting in myocardial ischaemia and necrosis. The latter is an additional source for a local and systemic inflammatory response. Patients with ACS frequently undergo percutaneous coronary intervention, which may cause iatrogenic myocardial injury, a further source of inflammation: inflammation begets inflammation.11 Lifestyle and diet have a major impact on the immune system and metabolism. Recent evidence indicates that these effects are mostly mediated by the resident microbial communities in the intestinal tract. The gut microbiota acts as an endocrine organ with metabolic capacity to produce multiple compounds that reach the circulation and act on different target organs. Trimethylamine N-oxide (TMAO), derived from gut microbial catabolism of dietary lipids and l-carnitine, represents an independent risk factor for CVD, and its plasma levels also predicted risk of cardiovascular events in patients with previous ACS.12,13 In addition, gut microbes can also modulate the host innate immunity through metabolism-independent pathways, and might directly influence the immune unbalance leading to ACS, as recently demonstrated in the epicardial adipose tissue surrounding coronaries arteries.8 Although the study of Ridker et al.6 is interesting and very carefully presented, a better knowledge of the causes of coronary instability is mandatory for the identification of new therapeutic targets aimed at the preservation of plaque stability. Postmortem and in vivo studies using intravascular imaging point to four pathological pathways to ACS, i.e. plaque rupture with systemic inflammation, plaque rupture without systemic inflammation, plaque erosion, and plaque without thrombus, although these mechanisms may overlap and co-exist in some patients.11 The first phenotype (plaque rupture with systemic inflammation) is the most prevalent.3,8 On one hand, a deeper T-cell perturbation characterizes ACS patients with ruptured culprit plaques, relying on a significant expansion of aggressive T lymphocytes in the absence of an adequate counter-regulatory response that inversely correlates with cap thickness, identifying a more event-prone plaque phenotype.14 These patients have a worse outcome, mainly due to recurrent analogue episodes of coronary instability, and might need a more specific target of adaptive immunity.15 Although IL-1β activity is an integral component of adaptive immune responses, T-cell receptor activation signalling offers several targets for a focused immunomodulatory therapy that prevents the excess activation of adaptive immunity without affecting its protective effects: among them, CD31, an immunomodulatory molecule whose functional domain is transiently shed in ACS, leading to uncontrolled lymphocyte activation; the protein tyrosine phosphatase PTPN22 that modulates early T-cell receptor activation; and regulatory T-cell expansion.3 On the other hand, plaque erosion is the cause of ACS in at least one-third of patients, and its prevalence is probably increasing. In a mechanistic study, the gene/protein expression of hyaluronidase 2 (HYAL2), the enzyme degrading hyaluronan to its pro-inflammatory 20 kDa isoform, and of the hyaluronan receptor CD44 were higher in patients with plaque erosion when compared with those with plaque rupture. Thus, plaque erosion might be characterized by a profound alteration of hyaluronan metabolism and, after further validation, HYAL2 might represent a potentially useful biomarker for the non-invasive identification of this mechanism of coronary instability.16 This paves the way toward new immunomodulatory targets in ACS caused by plaque erosion. Considering the success of anti-IL-1β therapy and the challenges that its clinical application faces, it is logical to look for alternative targets in this pathway that might exhibit advantages over canakinumab. One step upstream of IL-1β, the NLRP3-inflammasome deserves special attention as a potential target in CVD and other inflammatory disorders. Inhibiting IL-1β and IL-18 secretion through the highly selective NLRP3 inhibitor MCC95079 attenuated atherosclerosis development in hypercholesterolaemic mice. MCC95079 also reduced infarct size and cardiac dysfunction after ischaemia/reperfusion in swine models of acute myocardial infarction, and it improved cardiac remodelling and function in mouse models of heart failure.9 NLRP3 antagonists reduce infection-related adverse effects compared with direct IL-1β inhibitors, as they may preferentially block dyslipidaemia-driven IL-1β secretion, while preserving the activity of other pathogen-recognizing inflammasomes.8 The other target of potential value is the downstream cytokine IL-6. IL-1β strongly induces IL-6 expression, and the reduction in IL-6 levels is an important mechanism of atheroprotection in canakinumab-treated patients. However, IL-6 inhibitors face unique challenges in the CVD setting because of the complex effects of this cytokine on metabolism. IL-6 exerts conflicting actions on glucose homeostasis and insulin resistance at multiple levels. Furthermore, IL-6 inhibition frequently leads to an increase in LDL-cholesterol. Nevertheless, efforts are ongoing to develop IL-6-targeted approaches for CVD. The success of the monoclonal antibody tocilizumab, which targets the receptor for IL-6, in the treatment of rheumatoid arthritis has fuelled interest in it as a potential therapeutic strategy in other diseases characterized by increased IL-6 activity. In a small innovative phase II clinical trial, the authors prove that IL-6 blockade by tocilizumab quenches the acute inflammatory response in non-ST-elevation myocardial infarction patients undergoing percutaneous coronary intervention.17 Patients randomized to tocilizumab had lower levels of high sensitivity troponin T (hs-TnT) as compared with those randomized to placebo, suggesting smaller infarct size.18 Once more, we ask which patients might benefit from an anti-inflammatory treatment and what are the best options we can aspire to? IL-1β inhibition with canakinumab has been the first anti-inflammatory approach to successfully complete the transition from pre-clinical to clinical studies in CVD. By doing so, it has validated the inflammatory hypothesis of atherothrombosis and opened the door to the development of novel anti-inflammatory approaches, in both the inflammasome/IL-1β signalling pathway and beyond. While unlikely to become broadly used for cardiovascular prevention, IL-1β-targeted therapies may become a powerful tool in precision medicine strategies. It is tempting to envisage a future when DNA sequencing data, blood biomarkers, and imaging data are used to identify those individuals who remain at cardiovascular risk even with optimal management of traditional risk factors and who may derive the greatest benefit from anti-inflammatory drugs (see Figure 1).11 Green panel: the residual risk after canakinumab treatment might be explained by the persistence of high levels of IL-18 and IL-6. The use of inflammasome inhibitors or anti-IL-18 antibodies might mitigate the detrimental role of IL-18 on top of standard therapies plus canakinumab. Despite the fact that a significant portion of IL-6 is directly related to IL-1/IL-18 secretion, the need for the use of a specific anti-IL-6 antibody relies on the existence of different sources of this cytokine, deriving from NF-κB downstream pathways or from an altered gut microbiota. Pink panel: in the era of precision medicine, several different pathogenetic mechanisms might lead to ACS. Many patients presenting with ACS do not manifest any sign of systemic inflammation, and mechanisms underlying plaque instability might be multiple and scarcely dependent on inflammatory bursts. The figure shows the optical coherence tomography characterization of different culprit plaques, for which a specific anti-inflammatory treatment might not be useful and may also be somewhat detrimental due to the increased risk of infections. DAMPs, damage-associated molecular patterns; IL-1, interleukin-1; IL-6, interleukin-6; IL-18, interleukin-18; NF-κB, nuclear factor-κB; NLRP3, Nod-like receptor P3; PAMPs, pathogen-associated molecular patterns; SMCs, smooth muscle cells; TLR, Toll-like receptor. Green panel: the residual risk after canakinumab treatment might be explained by the persistence of high levels of IL-18 and IL-6. The use of inflammasome inhibitors or anti-IL-18 antibodies might mitigate the detrimental role of IL-18 on top of standard therapies plus canakinumab. Despite the fact that a significant portion of IL-6 is directly related to IL-1/IL-18 secretion, the need for the use of a specific anti-IL-6 antibody relies on the existence of different sources of this cytokine, deriving from NF-κB downstream pathways or from an altered gut microbiota. Pink panel: in the era of precision medicine, several different pathogenetic mechanisms might lead to ACS. Many patients presenting with ACS do not manifest any sign of systemic inflammation, and mechanisms underlying plaque instability might be multiple and scarcely dependent on inflammatory bursts. The figure shows the optical coherence tomography characterization of different culprit plaques, for which a specific anti-inflammatory treatment might not be useful and may also be somewhat detrimental due to the increased risk of infections. DAMPs, damage-associated molecular patterns; IL-1, interleukin-1; IL-6, interleukin-6; IL-18, interleukin-18; NF-κB, nuclear factor-κB; NLRP3, Nod-like receptor P3; PAMPs, pathogen-associated molecular patterns; SMCs, smooth muscle cells; TLR, Toll-like receptor. Conflict of interest: none declared. The opinions expressed in this article are not necessarily those of the Editors of the European Heart Journal or of the European Society of Cardiology.