Trimethylamine N-Oxide and Risk of Heart Failure Progression: Marker or Mediator of Disease

This article refers to 'Association with outcomes and response to treatment of trimethylamine N-oxide in heart failure: results from BIOSTAT-CHF' by T. Suzuki et al., published in this issue on pages 877–886. Despite significant advances in diagnosis and therapy over the past 20 years, heart failure (HF) patients still have a poor prognosis.1 The diversity in aetiologies, precipitants, and co-morbidities suggests that multiple pathophysiologic targets contribute to HF progression, and substantial heterogeneity in the response to therapy might be expected.2 However, there is limited understanding of the pathophysiology of HF, and current therapies focus rather on the deleterious consequences of HF (e.g. congestion and neurohormonal activation) as opposed to targeting specific organ pathways. Several conceptual frameworks have been proposed to explore the underlying pathogenesis of HF, and recently emerging evidence has suggested a potential impact of gut microbiota and its mediators on the pathophysiology of HF.3-9 The gut hypothesis (Figure 1) implies that decreased cardiac output and elevated systemic congestion can lead to intestinal mucosal ischaemia and oedema,8 leading to increased intestinal permeability for pro-inflammatory mediators that can contribute to the underlying inflammation seen in patients with HF.3, 9 Increased intestinal permeability then leads to the translocation of bacteria across the intestinal barrier and the systemic presence of endotoxin or other bacterial wall compounds, which may contribute to the pathophysiologic progression of HF5, 10 (Figure 1). Due to gut luminal hypoxia, hypercapnia, changes in local pH, redox state, and norepinephrine, all known to be potent activators of bacterial virulence in microbiota,3, 9 the composition of intestinal microbiota in HF may shift rapidly during intestinal ischaemia11 or following an increase in right atrial pressure.12 However, the composition of gut microbiota in HF patients can be modified not only by haemodynamic alterations, but also by dietary habits and therapeutic interventions.4 Two studies reported that patients with HF harbour significantly altered gut microbiota as a result of intestinal overgrowth of pathogenic bacteria.12, 13 Recently, trimethylamine N-oxide (TMAO) has emerged as a key mediator that provides mechanistic link between gut microbiota and multiple cardiovascular diseases, including HF. TMAO is the hepatic oxidation product of the microbial metabolite trimethylamine (TMA).7 TMA is a small organic compound that is generated by the gut microbiota and results from microbial metabolism of dietary nutrients that possess a TMA moiety (such as choline, phosphatidylcholine, and L-carnitine), as a result of variety of TMA microbial enzymes (TMA lyases).7 TMA is subsequently transported via portal circulation to the liver, where it is oxidized via hepatic flavin monooxygenases to form TMAO.6, 7 Different gut microbiome compositions have varying abilities to generate TMAO and higher levels of circulating TMAO can be attributable to a TMA-producing microbiome harbouring TMA lyases.6, 7 Bacterial overgrowth or changes in the composition of gut microbiota in patients with HF due to intestinal ischaemia11 may explain elevated TMAO levels (Figure 1). Although the distribution of TMAO levels in the general population is unknown and standard reference values are not currently available,6, 7, 12, 13 two recent studies have demonstrated that patients with HF have significantly higher plasma levels of TMAO than healthy control subjects, and elevated plasma TMAO levels are predictive of higher mortality risk in patients with chronic HF.6, 14 Furthermore, in patients admitted to hospital with acute decompensated HF, elevated plasma TMAO levels correlate with decreased renal function and can predict an increased risk of death or rehospitalization due to HF.15 Recent mechanistic pre-clinical studies have identified TMAO as a promising molecule involved in disease progression and associated with cardiovascular events, and it is possible that the presence of HF might per se affect gut microbiota composition and lead, in turn, to increased production of TMAO.16 Further research is needed to elucidate whether TMAO contributes directly to the progression of HF or only reflects the presence of a deleterious colonic microbial metabolism, and if TMAO levels may be impacted by HF therapies. In this issue of the Journal, Suzuki et al.17 present data from BIOSTAT-CHF, a prospective contemporary European HF cohort study. BIOSTAT-CHF enrolled 2234 patients with new-onset or progressively worsening HF confirmed by either a left ventricular ejection fraction of ≤ 40% or plasma concentrations of B-type natriuretic peptide and/or N-terminal pro-B-type natriuretic peptide > 400 pg/mL or > 2000 pg/mL, respectively.18 The authors aimed to investigate the association of circulating TMAO levels with outcomes, with a particular focus on the association of serial TMAO levels, and whether TMAO levels are responsive to initiation and up-titration of guideline-directed medical therapies (GDMTs), and further reflect changes in outcomes. The investigators found a strong association between baseline TMAO levels and adverse events (mortality and/or rehospitalization) at 1, 2, and 3 years [hazard ratio (HR) 1.37–1.51, P ≤ 0.019]. Furthermore, when considering serial TMAO measurements, an independent association with survival was seen only in patients with sustained increases in TMAO levels at both enrolment and follow-up (HR 2.21, 95% confidence interval 1.43–3.43, P < 0.001). Conversely, an increased association was not seen in patients that exhibited low TMAO levels at either visit. Another finding of interest was that current GDMTs did not affect TMAO levels. In fact, reductions in TMAO levels were not observed in patients with titration of dosages to greater than 50% of the recommended dosage, and a significant increase in TMAO levels was seen in patients achieving less than 50% optimal titration of either angiotensin-converting enzyme inhibitors/angiotensin receptor blockers or beta-blockers. One of the key findings of the present study is the strong prognostic association between baseline TMAO levels and long-term mortality in patients with worsening signs and symptoms of HF, an association incremental to traditional risk factors, such as ischaemic aetiology, renal indexes and natriuretic peptides (NPs). The results of the present analysis are similar to a previous study enrolling patients with acute HF.15 However, these observations may be an epiphenomenon, since causality has not been clearly established, and TMAO is more likely a marker than a mediator in this relationship. Although TMAO has been shown to possess pro-atherogenic properties,6 which may contribute to the adverse prognosis associated with higher TMAO levels, is less likely to be a reason for adverse outcome in this cohort, since in BIOSTAT-CHF extended risk model TMAO remained an independent predictor of outcome for mortality regardless of aetiology (i.e. ischaemic or non-ischaemic).17 This suggests that other potential mechanisms contribute to progression of HF and the increased mortality risk observed in patients with HF with elevated TMAO levels. Although in a recent animal experimental study19 elevated TMAO levels induced by rich choline diet have been associated with severe left ventricular adverse remodelling/dilatation and myocardial fibrosis, conferring TMAO a direct pathogenic effect, this hypothesis should be further confirmed in clinical studies. The finding that elevated TMAO levels were associated with more advanced left ventricular diastolic dysfunction, despite lack of correlation between TMAO and inflammatory and endothelial dysfunction biomarkers, supports the pathogenic role of TMAO in cardiac remodelling.14 However, our current understanding about the interaction between the gut microbiota or its metabolites and myocardial function in patients with HF remains limited, and other unknown mechanisms could be involved in the translation from a disturbed gut microbiota and elevated TMAO levels into adverse long-term prognosis in HF (Figure 1). Furthermore, although the authors tried to offer some plausible mechanisms related to increased TMAO levels, both related to intestinal congestion, it remains to be determined whether elevated TMAO levels are associated with enhanced gut oedema, increased intestinal permeability, and bacterial translocation in HF. At this point, 'the reverse hypothesis', whether decongestive therapies (intravenous diuretics/vasodilators) would decrease plasma TMAO levels, deserves considerable attention. Notably, the mechanisms contributing to the rise in TMAO in the setting of HF are likely multifactorial and include increased TMA production in the bowel, as a result of increased intake of carnitine from dietary sources or increasing enzymatic activity of intestinal bacteria, increased TMA entry from the bowel to blood, increased hepatic flavin monooxygenase activity, or decreased clearance of TMAO from plasma compartment.7 As different gut microbial compositions generate different levels of TMAO,7 higher blood TMAO levels and an increased mortality risk can be attributed to a specific TMA-producing microbiome possessing TMA lyases. These findings support the idea of a 'high-risk TMAO-generating phenotype' and furthermore that prevention of mortality is feasible through gut microbial modulation.4, 6, 7, 11-13 Reporting serial measurements of TMAO levels, and association between serial changes of TMAO and long-term adverse outcome, is one of the most important strengths of the study, and underlines again the hypothesis that persistent high levels of TMAO are a component of a high-risk phenotype. Notably, during follow-up, TMAO and NP serial changes had distinct trajectories, suggesting that the impact of TMAO on prognosis is probably independent of haemodynamic mechanisms involving increased filling pressure and systemic congestion. Another finding of interest was that current guideline-based HF treatment did not affect TMAO levels, while NP levels decreased following intensification of GDMT. In HF patients, NP levels have been shown to decrease in response to GDMT, and the magnitude of decrease relates to prognosis.20 The observation that GDMT did not reduce TMAO values in HF patients regardless of ejection fraction points to other targets of intervention. Modulation of gut microbiota composition and function through diet, prebiotics and probiotics, and targeted non-lethal antimicrobial enzyme inhibitors may facilitate, in the long term, the capacity to alter host metabolic profile in a desired favourable direction.7 Therapeutic management of HF through targeting gut microbiota is under recent investigation in the GutHeart trial,21 which will explore the hypothesis that the non-absorbable antibiotic rifaximin, by manipulating the bacterial composition of the gut content, will reduce the inflammatory activation translating into positive effects on adaptive remodelling and HF progression. However, future prospective studies will also be needed to determine whether the modulation of TMAO levels or its precursors might represent a novel potential therapeutic approach to change HF prognosis. Conflict of interest: none declared.