Plasma metabolomics in the perioperative period of defect repair in patients with pulmonary arterial hypertension associated with congenital heart disease

Abstract

The quality of life and survival rates of patients with pulmonary arterial hypertension associated with congenital heart disease (CHD-PAH) have been greatly improved by defect-repair surgery and personalized treatments. However, those who survive surgery may remain at risk of persistent PAH, the prognosis may be considerably worse than those unoperated. Dynamic monitoring of clinical measures during the perioperative period of shunt correction is therefore indispensable and of great value. In this study, we explored the plasma-metabolite profiling in 13 patients with CHD-PAH during the perioperative period of defect repair. Plasma was harvested at four time points: prior to cardiopulmonary bypass (CPB) after anesthesia (Pre), immediately after CPB (T0), 24 h (T24), and 48 h (T48) after defect repair. Untargeted metabolomics strategy based on UPLC Q-TOF MS was used to detect the metabolites. A total of 193 distinguishing metabolites were determined at different time points, enriched in pathways such as oxidation of branched-chain fatty acids. We found that 17 metabolite alterations were significantly correlated with the reduction in mean pulmonary arterial pressure (MPAP) at T48 versus Pre. Gradients in diastolic pulmonary arterial pressure (DPAP), bicarbonate in radial artery (aHCO₃), bicarbonate in superior vena cava (svcHCO₃), and the partial pressure of dissolved CO₂ gas in radial artery (aPCO₂) were positively correlated with MPAP gradient. Notably, these clinical-measure gradients were correlated with alterations in shunt-correction-associated metabolites. In total, 12 out of 17 identified metabolites in response to defect repair were increased at both T24 and T48 (all P < 0.05, except propionylcarnitine with P < 0.05 at T24). In contrast, galactinol dihydrate, guanosine monophosphate, and hydroxyphenylacetylglycine tended to decline at T24 and T48 (only galactinol dihydrate with P < 0.05 at T48). In conclusion, 17 metabolites that respond to shunt correction could be used as suitable noninvasive markers, and clinical measures, including DPAP, aHCO₃, svcHCO₃, and aPCO₂, would be of great value in disease monitoring and evaluating future therapeutic interventions.

Keywords: congenital heart disease; defect-repair surgery; metabolites; metabolomics; perioperative period; pulmonary arterial hypertension; pulmonary circulation.

Fig. 1. Overall comparative analysis of the metabolic profiling in 13 CHD–PAH patients during the perioperative stage.

a Schematic of the study design. Blood was drawn at four different time points (before cardiopulmonary bypass (CPB) after anesthesia (Pre), immediately after cardiopulmonary bypass (T0), 24 h (T24), and 48 h (T48) after surgery). Plasma was collected for metabolomics. b Score plots of PLS-DA analysis on the plasma metabolic profiles at four timepoints (green for Pre, blue for T0, red for T24, and yellow for T48). c Heatmap showing the distinguished metabolites (VIP > 1) with fold change of both T48 vs. Pre and T24 vs. Pre >2 or <0.5 in plasma from 13 CHD–PAH patients at the indicated time points. CPB cardiopulmonary bypass.

Fig. 2. Pathway enrichment of distinguished metabolites.

a Venn diagram depicts overlap in upregulated metabolites at T24 versus Pre and upregulated metabolites at T48 versus Pre. Fold change >1.5. b Venn diagram depicts overlap in downregulated metabolites at T24 versus Pre and downregulated metabolites at T48 versus Pre. Fold change <0.67. c The enrichment of top 20 metabolic pathways according to the identified distinguished metabolites at four time points by MetaboAnalyst (v5.0).

Fig. 3. Correlation of mean pulmonary arterial pressure gradient with the content alteration in metabolites.

a–h The content alteration of (a) propionylcarnitine; (b) butenylcarnitine; (c) isobutyryl-L-carnitine; (d) hexanoylcarnitine; (e) PC(16:0/22:4(7Z,10Z,13Z,16Z)); (f) 7-methylguanine; (g) bilirubin and (h) PC(14:0/22:5(4Z,7Z,10Z,13Z,16Z)) at T48 (48 h post CPB) versus Pre (before CPB) correlated with gradient in mean pulmonary arterial pressure (MPAP) during the same time period. i The enrichment of metabolic pathways according to the 17 identified MPAP-gradient-correlated metabolites by MetaboAnalyst (v5.0). Δ denotes the alteration of the indicated clinical parameter or metabolite. Δt48MPAP represent the gradients of mean pulmonary arterial pressure (MPAP) between T48 and Pre. ΔHMDB0000824, ΔHMDB0013126, ΔHMDB0000736, ΔHMDB0000705, ΔHMDB0007988, ΔHMDB0000897, ΔHMDB0000054 and ΔHMDB0007980 represents the change of metabolite propionylcarnitine, butenylcarnitine, isobytyryl-L-carnitine, hexanoylcarnitine, PC(16:0/22:4(7Z,10Z,13Z,16Z)), 7-methylguanine, bilirubin and PC(14:0/22:5(4Z,7Z,10Z,13Z,16Z)) at T48 versus that at Pre, respectively.

Fig. 4. Correlation of gradients in clinical characteristics.

a Correlation of gradients in clinical characteristics, including hemodynamics and clinical parameters of blood-gas analysis between time point of T48 and Pre; SBP: systolic blood pressure, DBP: diastolic blood pressure, MBP: mean blood pressure, SPAP: systolic pulmonary arterial pressure, DPAP: diastolic pulmonary arterial pressure, MPAP: mean pulmonary arterial pressure, svcHCO₃: bicarbonate in superior vena cava, aHCO₃: bicarbonate in radial artery, aPCO₂: the partial pressure of dissolved CO₂ gas in radial artery. All the significant correlations between clinical characteristics were shown as a circle in the square and nonsignificant correlations were shown as blank square, The higher the positive correlation is, the darker the color of blue will be, and the higher the negative correlation is, the darker the color of red will be; clinical characteristics significantly correlated with MPAP gradient were denoted with an asterisk (P < 0.05). b–f Scatter plot depicts the correlation of gradient in (b) DPAP; (c) SPAP; (d) aHCO₃; (e) svcHCO₃, and (f) aPCO₂ with MPAP gradient between the time point of T48 and Pre, respectively. Δ denotes the alteration of the indicated clinical parameters. Δt48MPAP, Δt48DPAP, and Δt48SPAP represents the gradient of mean pulmonary arterial pressure, diastolic pulmonary arterial pressure, and systolic pulmonary arterial pressure between T48 and Pre, respectively. Δt48aHCO₃, Δt48svcHCO₃, and Δt48aPCO₂ represents the change of bicarbonate in radial artery, bicarbonate in the superior vena cava, and partial pressure of dissolved CO₂ gas in radial artery between T48 and Pre, respectively.

Fig. 5. Correlation of metabolite alteration with clinical-parameter gradients in arterial blood-gas analysis.

a–c Correlation of propionylcarnitine variation with gradients in (a) bicarbonate in radial artery (aHCO₃), (b) bicarbonate in the superior vena cava (SvcHCO₃), (c) partial pressure of dissolved CO₂ gas in radial artery (aPCO₂) between the time point of T48 and Pre. d–f Correlation of butenylcarnitine variation with gradients in (d) aHCO₃, (e) SvcHCO₃, and (f) aPCO₂ between the time point of T48 and Pre. g–i Correlation of isobutyryl-L-carnitine variation with gradients in (g) aHCO₃, (h) SvcHCO₃, (i) aPCO₂ between the time point of T48 and Pre. j–l Correlation of hexanoylcarnitine variation with gradients in (j) aHCO₃, (k) SvcHCO₃, (l) aPCO₂ between the time point of T48 and Pre. Δ denotes the alteration of the indicated clinical parameter or metabolite. Δt48aHCO₃, Δt48svcHCO₃, and Δt48aPCO₂ represents the gradients of bicarbonate in radial artery, bicarbonate in the superior vena cava, and partial pressure of dissolved CO₂ gas in radial artery between T48 and Pre, respectively. ΔHMDB0000824, ΔHMDB0013126, ΔHMDB0000736, and ΔHMDB0000705 represents the change of metabolites propionylcarnitine, butenylcarnitine, isobytyryl-L-carnitine, and hexanoylcarnitine at T48 versus that at Pre, respectively.

Fig. 6. Plasma-metabolite content of CHD–PAH patients in the perioperative period.

a–h The plasma content of (a) propionylcarnitine; (b) butenylcarnitine; (c) isobutyryl-L-carnitine; (d) hexanoylcarnitine; (e) 7-methylguanine; (f) bilirubin; (g) PC(14:0/22:5(4Z,7Z,10Z,13Z,16Z)); and (h) PC(16:0/22:4(7Z,10Z,13Z,16Z)) of 13 CHD–PAH patients at the indicated time point in the perioperative period. Data are depicted as mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, as analyzed by the one-way analysis of variance (ANOVA) or Friedman test as appropriate.