Human immunology of heart failure: deconstructing inflammatory risk

Graphical Abstract

Deconstructing inflammatory risk for HF: translational strategies from biology to phenotype

Open in new tabDownload slide

This editorial refers to ‘Risk of heart failure in inflammatory bowel disease: a Swedish population-based study’, by J. Sun et al., https://doi.org/10.1093/eurheartj/ehae338.

Inflammation and heart failure: theoretical advances, practical gaps

Inflammation plays a central role in the pathogenesis and progression of heart failure (HF). Over the past three decades, observational and mechanistic studies have implicated inflammation as both a cause and a consequence of HF.^1,2 Yet, until recently, approaches targeting inflammation in HF failed to yield clinical benefit. Recently, direct and indirect approaches to modulating inflammation in HF have been somewhat more promising. In a secondary analysis of the Canakinumab Anti-Inflammatory Thrombosis Outcomes Study (CANTOS), direct inflammation modulation with canakinumab, an interleukin-1β antagonist, reduced hospitalized HF events.³ Meanwhile, a randomized controlled trial of semaglutide, a glucagon-like peptide 1 receptor agonist, improved HF symptoms in patients with obesity and HF with preserved ejection fraction (HFpEF)⁴; body weight reduction correlated with reduction in C-reactive protein (CRP) and symptomatic improvement,⁵ suggesting metabolic inflammation (‘meta-inflammation’) reduction as a potential contributor to the observed benefits.

These findings provide modest initial support for targeting inflammation in HF, but a gap remains between theory and practice: observational and experimental data implicating inflammation in HF vastly outpace clinical outcome data showing the efficacy of targeting inflammation in HF. So why the gap? The inherent clinical heterogeneity of HF and varied underlying mechanisms preclude a singular, all-encompassing model of inflammation and HF. As such, dissecting this heterogeneity into coherent clinical and mechanistic entities is a must. Over the past decade, a combination of supervised and unsupervised approaches has provided new insights into pathophysiological and clinical heterogeneity of various HF manifestations. Such approaches offer the potential to define enriched HF populations most likely to benefit from a particular intervention—an opportunity underscored by the early success of semaglutide in cardiometabolic HFpEF.⁴

While these approaches to dissecting clinical heterogeneity have yielded early promise in difficult-to-treat conditions such as HFpEF, much remains unanswered regarding heterogeneity of inflammatory triggers of HF. At its core, inflammation is an immune response to something.⁶ Certainly, the ‘something’—antigens, neo-antigens, and/or injurious stimuli inducing damage- and/or pathogen-associated molecular patterns—dictates the resulting immune responses to a degree. For instance, the nature, duration, and tissue specificity of immune responses during and after acute myocardial infarction and related downstream impact on the myocardium^7,8 differ from those occurring in cardiometabolic HFpEF.⁹ Yet, individual-level differences in immune/inflammatory response may also drive differing responses to common triggers: not all with obesity and hypertension progress to HFpEF, and not all with hyperlipidaemia experience myocardial infarction and ischaemic cardiomyopathy.

What causes some people to have persistent and problematic inflammation in response to pathological stressors and tissue injury, whereas others promptly resolve this inflammation to limit tissue damage and related clinical deterioration? Answering these questions is critical and likely to be complex, requiring a combination of mechanistic and clinical approaches.

Clinical phenotypes: chronic inflammation and heart failure

In this context, the study presented by Sun et al.¹⁰ in this issue of the European Heart Journal focuses on HF risk in a human ‘model’ condition marked by persistent inflammation. Given some inconsistencies in prior studies on inflammatory bowel disease (IBD) and HF risk, the investigators utilized a large Swedish nationwide cohort to comprehensively investigate this question. Specifically, they compared incident HF in patients with biopsy-confirmed IBD who were identified between 1969 and 2017 (n = 81 749) vs. several thoughtfully selected control populations. Control populations included up to five matched general population reference controls (n = 382 190) and IBD-free sibling controls (n = 95 239). Following up these populations through to the end of 2019, the investigators found that IBD patients had a higher risk for HF compared with general population controls, with an adjusted hazard ratio of 1.19 (95% confidence interval 1.15–1.23). This elevated risk was observed across IBD subtypes including Crohn’s disease, ulcerative colitis, and unclassified IBD. Sibling-controlled analyses also revealed an elevated HF risk for IBD patients, although the effect size was less, with an adjusted hazard ratio of 1.10 (95% confidence interval 1.03–1.19).

While cohort-related limitations—such as the absence of left ventricular ejection fraction or other HF phenotype-relevant data—precluded some analyses, these limitations are largely inevitable in a nationwide cohort with this size and duration of follow up. The authors did not investigate associations of specific IBD medications or classes of medications with incident HF. While this is also a limitation, it is an understandable one given confounding by unmeasured indication and other issues that can arise when analysing medication use as an exposure in observational cohort studies, rather than in randomized clinical trials. Overall, these limitations are outweighed by the strengths of utilizing such a large cohort with complete long-term follow-up data. This scope also enabled the authors to investigate whether the higher HF incidence among IBD patients changed over time (for instance, from before vs. after the introduction of modern IBD therapy); they observed relatively consistent IBD-associated HF risk across these time periods.

The primary takeaway from the study by Sun et al.¹⁰ is that people with IBD have modestly elevated HF risk, which has remained largely consistent across IBD treatment eras. How does this modest, but probably real, IBD-attributable increase in relative risk for HF impact practice? Certainly, knowing that IBD patients have elevated risk for HF can inform diagnosis and prevention: for IBD patients, providers should maintain a higher index of clinical suspicion for HF in the presence of new symptoms and be proactive in addressing risk factors for HF such as hypertension, obesity, and diabetes. Beyond these general points, one might also expect heterogeneity among IBD patients in IBD-attributable HF risk, whereby those with higher disease severity and chronic inflammatory burden have particularly elevated HF risk—as observed in other chronic inflammatory conditions¹¹—whereas those with mild disease may have blunted IBD-related increases in HF risk. These questions cannot be answered by the present study given the absence of disease severity-related measurements, but warrant future investigation.

Deconstructing inflammatory risk for heart failure: target practice, from target to practice

How do the findings by Sun et al. fit into the broader goal of more precisely defining inflammatory contributors to human HF? Such phenotype-focused approaches contribute one piece to a large and evolving translational puzzle. Inflammation is not monolithic, nor is HF, and there is no single best approach to deconstructing the complex relationship between inflammation and HF. Approaches ranging from fundamental, mechanistic investigation in experimental models to clinical observational studies can provide valuable and synergistic insights into new targets in the inflammation–HF paradigm. This reflects the diversity of potential mediators of inflammation and HF (Graphical Abstract): putative targets exist at multiple levels and warrant evaluation. Future discoveries may be at a genomic level¹² or involve alterations at gene-regulatory, protein translation (or post-translational modification), and/or metabolic levels. They may be immune cell specific or not. They may incorporate several of the above—for instance, metabolic–epigenetic reprogramming of innate immune cells to induce hyperinflammatory responsiveness.¹³ They are also highly likely to differ by specific HF phenotype—HFpEF vs. HFrEF, ischaemic vs. non-ischaemic, and various sub-phenotypes therein.

Given the heterogeneity of these inflammatory triggers and of HF phenotypes,¹⁴ when considering both clinical/translational and experimental approaches, the more precisely the sampling and/or experimental models reflect the HF subtype, the better. For instance, if one is interested in defining inflammatory perturbations underlying cardiometabolic HFpEF, investigations that interrogate clinical samples from those with a combination of metabolic syndrome, hypertension, and HFpEF are more likely to yield coherent insights than those merging several pathophysiologically diverse phenotypes of HF (valvular, ischaemic, etc.) into a single group. Furthermore, pairing carefully selected, phenotype-specific data with experimental models aimed to reflect these specific HF manifestations offers the potential for viable mechanistic insights. This has become increasingly possible with development and validation of a variety of HF experimental models, ranging from hypertensive–metabolic HFpEF to genetic dilated cardiomyopathies to immune checkpoint inhibitor-induced myocarditis. The more closely these model systems parallel the human condition, the more clinical and experimental data can work in synergy to yield human-relevant insights into targets modulating immune/inflammatory axes in HF.

What is the goal of these investigations? Ideally, the discovery and validation of viable targets in the inflammation–HF axis—be they genetic, gene-regulatory, or otherwise—will ultimately translate to practice. As discussed above, aligning clinical and experimental model systems in a coherent, phenotype-specific way is an important step. Just as the concept of inflammatory risk for atherosclerotic cardiovascular disease (ASCVD)¹⁵ has helped open the door for effective inflammation-targeted therapies, elucidating inflammatory contributors to HF may inform future inflammation-targeted approaches. Given the diversity of clinical and mechanistic contributors to similarly diverse phenotypes of HF, this will require precise, phenotypically targeted approaches. Ultimately, if these efforts yield precise, valid targets, they can inform a future of targeting immune dysregulation and inflammation in HF.

Declarations

Disclosure of Interest

M.J.F serves as a Consultant for Abbott Laboratories.

Funding

National Institutes of Health (grant number R01HL156792) to M.J.F. and American Heart Association (strategically focused research network grant numbers 24SFRNPCN1291224 and 24SFRNCCN1276086) to M.J.F

References

Author notes