Prognostic significance of percutaneous coronary intervention associated blood loss in acute coronary syndrome

Abstract Background Antiplatelet and anticoagulants are one of the mainstay treatment of acute coronary syndrome (ACS), however they are associated with a significant increase of bleeding risk. While anaemia is a recognized predictor of adverse outcomes, it is unknown if a variation of haemoglobin (HB) levels, even without associated anaemia, has the same impact. Purpose The aim of this study was to determine the prognostic impact of HB variation after percutaneous coronary intervention (PCI) in ACS patients. Methods Retrospective analysis of 822 consecutive patients admitted due to ACS and treated with PCI, in a single coronary intensive care unit. Delta HB – ΔHB – (HB at admission – HB 24 hours after PCI) was calculated. Clinical variables and therapeutic strategies were examined. The primary endpoint analysed during follow-up was all-cause mortality. Possible predictors for all-cause mortality were assessed by Cox regression models. When statistically significant values were found in univariate analysis, multivariate analysis was used to determine whether ΔHB was independent from other known factors in predicting the outcome. Results In the studied sample, 75.4% were male. Mean age was 66.4±13.1. ST-elevation myocardial infarction (STEMI) occurred in 45.5%, non-ST segment elevation myocardial infarction in 42.6% and unstable angina in 11 9% of the studied population. Moderate to severe systolic dysfunction was present in 23.5% of the cases. Regarding comorbidities and past medical history, 76% had hypertension (HTN), 30.3% diabetes, 16.4% chronic kidney disease (CKD), 62.2% dyslipidaemia and 10.5% heart failure (HF). Mean HB at admission was 13.8±1.8 g/dL, mean HB after PCI was 12.9±1.9 g/dL and mean ΔHB was 0.9±1.1 g/dL. The mean follow-up was 51.6±30.6 months. In univariate analysis, ΔHB was significantly associated with all-cause mortality (HR 1.15 per 1 g/dL loss, 95% CI 1.01–1.30, p=0.04), as was HB at admission, HB after PCI, age, sex, diabetes, HTN, dyslipidaemia, CKD and moderate to severe systolic dysfunction. In multivariate analysis, ΔHB remained significantly associated with the endpoint and gained even more statistical power (HR 1.25, 95% CI 1.10–1.43, p<0.01). HB at admission and after PCI, age, CKD and moderate to severe systolic dysfunction were also independent predictors of this outcome. Conclusions In our study, irrespective of the admission and discharge HB, ΔHB was associated with more adverse outcomes in patients submitted to PCI. Hence, even patients with a normal HB after PCI have a worse long-term prognosis if a negative variation of HB occurs. This highlights the importance of identifying and optimising all the correctable factors that might lead to an increased bleeding risk. Funding Acknowledgement Type of funding sources: None.