Changes in Phenylacetylglutamine Levels Provide Add-On Value in Risk Stratification of Hypertensive Patients: A Longitudinal Cohort Study

Background: Despite antihypertensive treatment, some high-risk hypertensive patients still experience major adverse cardiovascular events (MACEs). Current risk stratification tools may underestimate the presence of metabolites in hypertension and thereby risk of MACEs.

Objectives: We aimed to explore the potential value of gut microbiota-derived metabolite phenylacetylglutamine (PAGln) in risk stratification of hypertension.

Methods: We measured plasma PAGln levels using liquid chromatography tandem mass spectrometry in 1543 high-risk hypertensive patients, dividing them into a discovery cohort (n = 792) and a validation cohort (n = 751). After follow-up, the Kaplan-Meier curve and the Cox regression model were utilized to determine the correlation between PAGln and MACEs (death, non-fatal ischemic stroke and hemorrhagic stroke, non-fatal acute coronary syndrome and unplanned revascularization). We examined the predictive performance of PAGln in different subgroups and evaluated the incremental predictive value of PAGln as an addition to the ASCVD risk assessment model.

Results: Among all high-risk hypertensive patients, 148 patients experienced MACEs after a mean follow-up of 3.02 years. In both cohorts, after adjusting other confounding risk factors, PAGln remained an independent risk factor the MACEs in hypertensive patients. Patients with plasma PAGln ≥ 1.047 μmol/L have a higher risk of MACEs. PAGln concentration provided incremental predictive value to the ASCVD risk model, with better performance in the discovery cohort. It was most effective in female, patients with a systolic blood pressure (SBP) ≥ 130 mmHg and taking angiotensin-converting enzyme inhibitors (ACEIs).

Conclusions: PAGln was associated with an increased risk of MACEs in hypertension, especially in women or in subgroups with SBP ≥ 130 mmHg and taking ACEIs. PAGln should be considered as an independent predictor in risk stratification to improve prognosis.