Prognostic value of gut microbe‐generated metabolite phenylacetylglutamine in patients with heart failure

Abstract

BackgroundPhenylacetylglutamine (PAGln) is a phenylalanine‐derived metabolite produced by gut microbiota with mechanistic links to hear failure (HF)‐relevant phenotypes. We sought to investigate the prognostic value of PAGln in patients with stable HF.Methods and resultsFasting plasma PAGln levels were measured by stable‐isotope‐dilution LC‐MS/MS in patients with stable HF from two large cohorts. All‐cause mortality was assessed at 5‐year follow up in the Cleveland Cohort, and HF, hospitalization, or mortality were assessed at 3‐year follow up in the Berlin Cohort. Fasting PAGln levels were measured by stable‐isotope‐dilution LC‐MS/MS. Within the Cleveland cohort, median PAGln levels were 4.2[IQR 2.4‐6.9] μM. Highest quartile of PAGln was associated with 3.09‐fold increased mortality risk compared to lowest quartile. Following adjustments for traditional risk factors, as well as race, eGFR, NT‐proBNP, hsCRP, LV ejection fraction, ischemic etiology, and heart failure drug treatment, elevated PAGln levels remained predictive of 5‐year mortality in quartile comparisons (adjusted Hazard Ratio [95%CI] for Q4vQ1: 1.64 [1.07‐2.53]). In the Berlin cohort, a similar distribution of PAGln levels was observed (median 3.2 μM [IQR 2.0‐4.8] μM), and PAGln levels were associated with a 1.94‐fold increase in 3‐year HF hospitalization or all‐cause mortality risk (adjusted HR [95%CI] for Q4vQ1: 1.94 [1.14‐3.28]). Prognostic value of PAGln appears to be independent of trimethylamine N‐oxide levels.ConclusionHigh levels of PAGln are associated with adverse outcomes independent of traditional cardiac risk factors and cardio‐renal risk markers.This article is protected by copyright. All rights reserved.

Unravelling the gut hypothesis of heart failure with gut microbiota‐derived metabolites

“…Perhaps the most interesting aspect of the present study by Tang et al 3 is the finding that both gut-derived metabolites PAGln and TMAO are maintained in a model to predict all-cause mortality, in the Cleveland Clinic cohort as well as in the Berlin cohort of (stable) HF patients. It should be noted that model 3 in the present study consists of traditional cardiac risk factors and TMAO levels, and no longer of the other variables of model 2 (estimated glomerular filtration rate and N-terminal pro-B-type natriuretic peptide as most contributing), so we may not be informed on all contributing factors.…”

“…S4, and although it is not clear how, the PAGln effects in patients with normal kidney function seem less severe, with the result that kidney function is attenuating the effects of PAGln in multivariable analysis. 3 In a recent large Chinese study on PAGln and prognosis in HF patients, also referenced by the authors, the interaction between PAGln and creatinine levels on endpoints was more clearly demonstrated. 10 The prognostic value of PAGln in the study by Wei et al 10 was considerably adjusted by kidney function but still retained its predictive value for 2-year all-cause mortality.…”

“…In the present study that includes TMAO, but analyses the additive prognostic effect of PAGln, there is a clear correlation between the two metabolites with an r-coefficient of 0.55, and the distribution of patients around the median PAGln and TMAO levels in Figure 3 is not random, with most patients having either low or high levels of both PAGln and TMAO. 3 These findings alone would reflect the presence of either a normal or increased intestinal permeability and one may think that differences in outcome explained by either one of the two metabolites are only represented in one third of patients in whom the levels of PAGln and TMAO are somewhat discordant. On the other hand, the prediction of outcomes is much worse when both metabolites are present in higher concentration, and this would not reflect a single postulate of increased intestinal permeability as the sole cause of increased mortality.…”

“…1,12 Although the population of HF patients have an increased likelihood of intestinal permeability problems, and are expected to have increased levels of metabolites, outcomes of the present and previous studies may have been influenced by selection of patients from populations undergoing elective coronary artery angiograms. 3,5,12 To circumvent the question whether there are mainly ischaemic events to be expected from the two metabolites, the question can be asked whether the metabolites act on non-ischaemic HF patients. For TMAO, the prediction on mortality was also valid for HF patients without coronary artery disease.…”

“…1,2 In this issue of the Journal, the paper of Tang et al associates two gut-derived metabolites phenylacetylglutamine (PAGln) and trimethylamine-N-oxide (TMAO) with prognosis in stable heart failure (HF) patients. 3 The gut hypothesis of heart failure Heart failure may be the cause of altered intestinal function, followed by increased permeability of both the small and large intestine, accompanied by increased colonization of pathogenic bacterial species (Salmonella, Shigella, Yersinia and Campylobacter) and fungal species (Candida). 4 In the study by Pasini et al 4 , well-fed stable HF patients were divided into two groups with lower and higher New York Heart Association (NYHA) class, with mean left ventricular ejection fractions of 39% and 35%, respectively.…”