The potential of cystatin C as a predictive biomarker in pulmonary hypertension

Abstract

Background: Cystatin C is a novel biomarker to identify renal dysfunction and cardiovascular risk.

Objective: The aim of this study was to investigate the role of cystatin C in non-invasive risk prediction in a large cohort of patients with pre-capillary pulmonary hypertension (PH).

Method: We retrospectively analyzed pre-capillary PH patients with available cystatin C and hemodynamic data derived from right heart catheterization.

Results: A total of 398 consecutive patients with confirmed pre-capillary PH were recruited from Fuwai Hospital between November 2020 and November 2021. Over a median duration of 282 days, 72 (18.1%) of these patients experienced clinical worsening. Cystatin C levels significantly correlated with cardiac index (r = -0.286, P < 0.001), mixed venous oxygen saturation (r = -0.216, P < 0.001), and tricuspid annular plane systolic excursion (r = -0.236, P < 0.001), and high cystatin C levels independently predicted a poor prognosis after adjusting potential confounders in different models (all P < 0.05). A three-group non-invasive risk model was constructed based on the combined assessment of the cystatin C and WHO-FC using dichotomous cut-off value. Those patients with higher cystatin C (≥ 1.0 mg/L) and a worse WHO-FC experienced the highest risk of endpoint occurrence. The predictive capacity of this model was comparable to that of an existing invasive risk stratification model (area under curve: 0.657 vs 0.643, P = 0.619).

Conclusions: Cystatin C levels were associated with disease severity and prognosis in patients with pre-capillary PH. A combination of high cystatin C and advanced WHO-FC identifies patients at particularly high risk of clinical deterioration.

Keywords: Biomarker; Cystatin C; Pulmonary hypertension; Renal function; Risk prediction.

Fig. 1

Receiver operator curves for parameters and eGFR derived from different equations. CKD-EPI, Chronic Kidney Disease-Epidemiology Collaboration; Cr, creatinine; Cys, cystatin C; eGFR, estimated glomerular filtration rate measured in ml/min per 1.73 m²; MDRD, Modification of Diet in Renal Disease. Depicts the receiver operator curves illustrating the performance of cystatin C, creatinine and estimated glomerular filtration rate derived from different equations

Fig. 2

Relationship between cystatin C and disease severity. A Correlation (Spearman) with cardiac index, B Correlation (Spearman) with TAPSE, C Correlation (Spearman) with SvO₂, D Cystatin C in different WHO-FC (***P < 0.001), E NT-proBNP (stratified by tertile) and cystatin C (***P < 0.001), F Swedish/COMPERA risk stratification and cystatin C. CI, cardiac index; COMPERA, Comparative, Prospective Registry of Newly Initiated Therapies for Pulmonary Hypertension; NT-proBNP, N-terminal pro-brain natriuretic peptide; SvO₂, mixed venous oxygen saturation; TAPSE, tricuspid annular plane systolic excursion; WHO-FC, World Health Organization functional class. Presents the relationship between cystatin C levels and disease severity in the context of hemodynamic status, functional class, NT-proBNP and risk stratification

Fig. 3

Kaplan–Meier analysis for the probability of endpoint events. A Patients stratified by the cystatin C cut-off; B Patients stratified by the Swedish/COMPERA risk model; C Patients stratified by the combination of cystatin C and World Health Organization functional class. Presents the Kaplan–Meier analysis, which illustrates the probability of endpoint events stratified by cystatin C cut-off (1.0 mg/L), Swedish/COMPERA risk model and the combination of cystatin C and WHO-FC

Fig. 4

Area under the curve for clinical worsening using the Swedish/COMPERA and the non-invasive risk stratification strategy. COMPERA, Comparative, Prospective Registry of Newly Initiated Therapies for Pulmonary Hypertension. DeLong test pairwise comparison: Non-invasive model vs. Swedish/COMPERA: Area under the curve 0.657 vs 0.643, P = 0.619. Showcases the area under the curve analysis for clinical worsening, comparing the predictive capacity of Swedish/COMPERA risk stratification strategy with a non-invasive risk stratification model