Cystatin C and bleedings during hospitalization of noninvasively treated patients with acute coronary syndromes

Abstract To assess relationship between blood level of a marker of renal function cystatin C (CC) and bleeding events in patients with acute coronary syndromes (ACS). CC level was measured in 160 patients from the ACS registry conducted in a noninvasive hospital in Moscow. Each month we included in this registry 25 consecutively hospitalized patients and recorded deaths and bleeding events (major and moderate by TIMI and/or GUSTO criteria) during hospitalization. CC measurement was made by ELISA in stored blood samples taken within first 24 hours after admission. The cut-off level for CC (> 1.53 mg/l) was established by ROC analysis. There were 10 deaths (6.3%) and 11 bleeding events (6.9%) during hospitalization (median 20 days). Inhospital mortality was 45.5 and 3.4% % among patients with and without bleeding, respectively (=0.0001). Bleeding events occurred significantly more often among patients with CC level >1.53 mg/l compared to those with CC level less or equal 1.53 mg/l (13.8 and 2.1%; respectively, p=0.008). Multifactorial regression analysis identified the following independents predictors of bleeding: thrombolysis at admission (odds ratio [OR] 9.32; 95% confidence interval [95%CI] 1.96-44.29; =0.005), CC level >1.53 mg/l (OR 7.96; 95% CI 1.34-47.26; p=0.023), initial white blood cells count (OR 1.31; 95%CI 1.10-1.57; =0.003). When CC was excluded from regression analysis creatinine clearance entered the list of independent predictors of bleeding (OR 0.95; 95% CI 0.90-0.99; =0.023). In this group of noninvasively treated patients with ACS high CC level (>1.53 mg/l) was independently related to major and moderate bleedings during hospitalization. If CC was excluded from analysis, another marker of renal function (creatinine clearance) became an independent predictor of inhospital bleeding.