Growth differentiation factor-15 and plasminogen activator inhibitor-1 are possible predictors of cardiovascular events in atrial fibrillation patients after elective PCI

Abstract

Abstract Introduction Patients (pts) with AF (atrial fibrillation), undergoing elective PCI (percutaneous coronary intervention), receiving combined antithrombotic therapy (AT) have high risk of stroke/systemic embolism, coronary events and bleedings. Minimizing the duration of combined AT is the main way to improve the safety of treatment. However, due to the high risk of thrombotic events, patient stratification is necessary. Biomarkers of high risk cardiovascular events (CVE) could be the rational reason to prolong combined AT for some pts. The search for them is relevant. There is data on relationship between growth differentiation factor-15 (GDF-15) level, plasminogen activator inhibitor-1 (PAI-1) level and adverse CVE. Purpose To investigate the prognostic value of GDF-15 and PAI-1 levels in AF pts after elective PCI during 1 year follow up. Methods 150 pts (104 males), aged 70.8±8.5 years, with AF after elective PCI were enrolled in the prospective study. All pts received direct oral anticoagulants and double (89.3%) or single (10.7%) antiplatelet therapy. Median duration of follow up period was 11.5 months [IQR 8.0; 12.0]. Efficacy endpoint of the study was the sum of CVE: ACS (acute coronary syndrome), ischemic stroke (IS), VTE (venous thrombotic events), cardiovascular death and need for unplanned PCI. Plasma samples for GDF-15 and PAI-1 were taken before PCI and analyzed using ELISA. Results AF pts receiving combined AT demonstrated high thrombotic risk (CHA2DS2-VASc, Med 5 [IQR 4; 6]), high rate of comorbidity (index Charlson, Med 7 [IQR 5; 9]). Frequency of CVE during follow up period was 16% (ACS 2; fatal IS 2; VTE 2, include 1 fatal pulmonary embolism; cardiovascular death 2; pts needed an unplanned PCI 16). Median GDF-15 level was 1270 pg/ml [IQR 953; 1778]. Median PAI-1 level was 10.15 U/ml [6,3; 17,0]. By linear correlation analysis (r=0.24; p=0.0037) and by regression model (F-ratio 157.4, p<0.0001) GFD-15 and PAI-1 are exhibited a relation. According to ROC analysis GDF-15 level >1191 pg/ml (sensitivity 83.0, specifity 50.0) and PAI-1 level >13.2 U/ml (sensitivity 58.3, specifity 68.6) increases the probability of the development of CVE (AUC=0.647, p=0.0076, CI 0.565–0.723; AUC=0.652, p=0.0135, CI 0.569–0.728 respectively). Kaplan-Meier curves demonstrated significant difference in CVE free survival between the patients with GDF-15 level > and ≤1191 pg/ml (75.9% and 94.0% respectively, p=0.0032 [HR 4.36, CI 1.50–7.48]); and with PAI-1 level > and ≤13.2 U/ml (74% and 89% respectively, p=0.0112 [HR 2.73, CI 1.28–6.98]). The multivariable Cox regression model demonstrated that GDF-15 (p=0.0363) and PAI-1 (p=0.0074), were independently associated with an increased risk of CVE (op=0.0022). Conclusions GDF-15 and PAI-1 are correlated, and they are new markers of CVE in AF pts after elective PCI. Funding Acknowledgement Type of funding sources: None.