Author + information
- Enrique Calvo, PhD,
- Ana García-Álvarez, MD, PhD and
- Jesús Vázquez, PhD∗ ()
- ↵∗Reprint requests and correspondence:
Dr. Jesús Vázquez, Department of Vascular Pathophysiology, Centro Nacional de Investigaciones Cardiovasculares Carlos III, C/Melchor Fernández Almagro 3, 28029 Madrid, Spain.
Pulmonary hypertension (PH) and subsequent right ventricular (RV) failure are increasingly recognized as global health problems, affecting patients with highly prevalent diseases such as chronic heart failure, chronic obstructive pulmonary disease, collagen vascular disease, or chronic pulmonary embolism (1). Regardless of the etiology, PH has a poor prognosis, closely related to the degree of RV functional impairment. There is therefore a need to identify new tools allowing timely diagnosis of PH and early detection of RV damage.
Recent evidence from in vitro and animal models has revealed that the pathogenesis of PH may involve metabolic reprogramming, including impaired glucose homeostasis and altered mitochondrial metabolism. In addition, clinical studies have revealed alterations to the arginine-NO pathway in lung tissue from a small cohort of patients with advanced PH (2) and in plasma from patients with systolic heart failure and increased pulmonary artery pressure (PAP) (3). These findings suggest that these metabolic alterations may track disease development, highlighting the potential of circulating metabolites as PH biomarkers. However, to date, there have been no published metabolomic studies of plasma from patients with PH.
In this issue of the Journal, Lewis et al. (4) reported on the search for plasma metabolite biomarkers of PH and associated RV–pulmonary vasculature (PV) dysfunction. Rather than using a discovery cohort of individuals with or without known PH, the investigators initially worked with 71 individuals with dyspnea. Study participants were subjected to simultaneous invasive hemodynamic measurements, blood sampling, and radionuclide ventriculography at rest and during cardiopulmonary exercise testing. Five hemodynamic parameters were correlated with the concentration of 105 pre-selected plasma metabolites, analyzed on a targeted profiling platform. Regression analysis of the resulting 5 × 105 matrix identified 21 metabolites significantly associated with 2 or more RV-PV dysfunction parameters. A large fraction of these metabolites belong to categories linked to PH in earlier studies, including purine degradation products and arginine and related metabolites. The increase in the concentration of uric acid and other purine metabolites is not surprising because uric acid is associated with several cardiovascular and related disorders, including hypertension, hypertrophy, metabolic syndrome, and diabetes mellitus. Moreover, as the investigators indicated, purine metabolites are linked to oxidative stress conditions. Regarding arginine metabolites, the researchers found a particularly strong inverse correlation between overall arginine bioavailability (the ratio of arginine to ornithine + citrulline) and the change in mean PAP relative to the change in cardiac output, apparently reflecting a deficiency in NO-mediated vasodilation. Diminished arginine bioavailability is observed in many other end-organ dysfunctions linked to vascular impairment, including renal failure and diabetes mellitus. Therefore, this finding likely reflects the generic importance of the production and balance of NO for the maintenance of RV-PV function.
The study revealed a direct correlation between RV-PV hemodynamic parameters and tricarboxylic acid (TCA) cycle intermediates. Although this finding was not explored further, the researchers discussed the possibility that these intermediates may play a role in modulating vascular tone in humans. Although this intriguing hypothesis is supported by previous evidence that succinate and other TCA intermediates can act as signaling molecules through binding to specific receptors, further studies are required to clarify whether TCA intermediates elicit a compensatory mechanism to maintain blood pressure or are byproducts produced by alterations to mitochondrial metabolism (5).
The most remarkable finding of the present analysis is that RV-PV hemodynamic parameters showed a strong direct correlation with a subset of tryptophan metabolites. This subset did not include metabolites of the tryptophan hydroxylase–mediated pathway (leading to serotonin production). Instead, a striking correlation was found with the levels of the 4 tryptophan metabolites of the indoleamine 2,3-dioxygenase (IDO) pathway: kynurenine, kynurenate, anthranilate, and quinolinate. This finding is particularly interesting because although the enzyme IDO is known to be modulated by proinflammatory cytokines and a growing body of evidence has indicated that its byproduct kynurenine plays a role as vasodilator (6), the tryptophan/IDO pathway had never been associated with PH in humans. The investigators explored the specificity of the association of the IDO pathway with pulmonary dysfunction in 2 elegant experiments. The first analyzed the relative concentrations of the 4 IDO metabolites in radial and proximal pulmonary arteries. This experiment revealed significantly higher levels of the metabolites in the radial artery of individuals with elevated pulmonary vascular resistance (PVR), providing evidence that these metabolites are produced in the lung. In the second experiment, a mouse model of hypoxia-induced PH showed upregulated mRNA levels of IDO and elevated kynurenine/tryptophan ratios in lung, as well as increased levels of IDO metabolites in plasma. These results fit nicely with an earlier report showing that endogenous IDO expression by pulmonary endothelium in a mouse model protected against PH development (7). Based on these findings, Lewis et al. (4) proposed that IDO metabolites are released as vasodilators in response to PH, acting as a compensatory pathway in states of NO deficiency. This intriguing hypothesis raises the question of whether this mechanism is specific to the pulmonary arteries because IDO is also expressed in other tissues.
The biomarker potential of the 4 IDO metabolites was tested in a validation cohort composed of 71 individuals, studied similarly to the discovery cohort. This analysis confirmed a significant association of the IDO pathway with RV-PV dysfunction and showed that IDO metabolites can discriminate between the absence or presence of abnormal pulmonary vascular function (C-statistic >0.66). It should be noted that the threshold used to define the presence of abnormal function (PVR >2.0 Wood units or mean PAP >21 mm Hg) corresponds to “borderline PH”; the clinical relevance of this threshold is still undetermined, and it lies below the currently recommended threshold (PVR >3.0 Wood units and mean PAP >25 mm Hg) (1).
The investigators also analyzed a second validation cohort composed of 19 controls and 11 patients with known World Health Organization group 1 pulmonary artery hypertension. Three of the IDO metabolites showed a strong ability to discriminate between the 2 groups (C-statistic >0.79). These are impressive results, but the small population size means that these promising data are still preliminary. Confirmation of the clinical utility of IDO metabolites as PH biomarkers will require analysis in larger cohorts. Similarly, whether the IDO pathway could constitute a therapeutic target in PH is unknown.
Another limitation of this study is the use of a targeted configuration, in which the set of metabolites analyzed was pre-defined. Compared with the nontargeted, or shotgun, method, the targeted approach is more straightforward, sensitive, and systematic. However, it comes at the cost of producing potentially hypothesis-driven results because the small number of target metabolites covers only a tiny proportion of the estimated human endogenous metabolome (more than 3,000 compounds), and selection is inevitably influenced by prior knowledge of the disease. Despite these limitations, the impressive data gathered by Lewis et al. (4) show the remarkable ability of metabolomics to identify intermediates with important roles in PH. We foresee that continuing improvements in mass spectrometry and the expansion of compound libraries will allow the development of novel approaches that combine the wide coverage of shotgun analysis with the advantages of targeted approaches, as is already happening in the proteomics field. These novel strategies would move the field toward a true metabolome-wide analysis, allowing full integration of phenotype and omics data in the study of disease.
↵∗ Editorials published in the Journal of the American College of Cardiology reflect the views of the authors and do not necessarily represent the views of JACC or the American College of Cardiology.
The authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- 2016 American College of Cardiology Foundation
- Galiè N.,
- Humbert M.,
- Vachiery J.L.,
- et al.
- Shao Z.,
- Wang Z.,
- Shrestha K.,
- et al.
- Lewis G.D.,
- Ngo D.,
- Hemnes A.R.,
- et al.