Anaemia as predictor of gastrointestinal bleeding in atrial fibrillation patients undergoing percutaneous coronary artery stenting

Sir, We read with great interest the excellent review by Foley et al. who evaluate the risks of gastrointestinal bleeding in patients undergoing percutaneous coronary intervention (PCI) in relation to anti-platelet therapy.1 The authors emphasize the importance of several gastrointestinal bleeding major risk factors including use of dual anti-platelet therapy or non-steroidal anti-inflammatory drugs, increasing age, previous ulceration, helicobacter pylori infection, use of non-steroidal anti-inflammatory drugs and co-morbidity including sepsis and mechanical ventilation. However, they neither mention baseline anaemia nor triple antithrombotic drug use (aspirin, clopidogrel and warfarin) as conditions that could increase the rate of bleeding events after PCI-stenting. In a recent analysis of a cohort of patients with atrial fibrillation (AF) undergoing PCI-stenting with additional indication for long-term oral anti-coagulation, we recently reported that triple antithrombotic use was the main predictor of late major bleeding (MB).2 Also, there are now some data that suggest anaemia as a high-risk marker for hemorrhagic complications in PCI-stenting.3 However, the impact of anaemia on the occurrence of MB and their clinical presentation in AF patients with an indication for oral anti-coagulation undergoing PCI-stenting has not been previously studied. We studied 278 consecutive AF patients with indication for oral anti-coagulation who underwent PCI-stenting in two Spanish tertiary centres. Patients were classified as anaemic using the World Health Organization definition (Hb < 12.0 g/dl in women and <13.0 g/dl in men).4 During the follow-up (median 19.2 months [IQR 10.3–38.2]) patients were contacted by telephone to obtain the occurrence of MB.5 Baseline patient characteristics, as well as procedural variables, are shown in Table 1. Anaemia was presented in 114 (41%) patients. A total of 38 patients (14.4%) suffered MB over the study period. Overall, anaemic patients had higher rate of MB than non-anaemic patients (20.0% vs. 10.7%; p = 0.035). Gastrointestinal MB events (GI-MB) were more common amongst patients with baseline anaemia (12.4% vs. 3.1%; P = 0.004), whilst non-GI-MB were similar in both groups (7.7% vs. 7.5%; p = 0.958) (Fig. 1). The majority of MB occurred within first year of follow-up (63% at 6 months and 76% at 12 months of follow-up). Patients who had MB complications were more often taking triple anti-thrombotic therapy (57%). Almost all INR values on the day of MB were within the therapeutic range. MB complications according to clinical presentation and anaemic status. A total of 38 patients (14.4%) suffered MB events over the study period: gastrointestinal 18, intracranial 6, over arterial puncture site 4, genitourinary 3, retroperitoneal 2, decrease in haemoglobin levels (>4 g/dl) 2, pulmonary 1, otorhinolaryngology 1 and gynaecological 1. Baseline characteristics and procedural characteristics PCI, percutaneous coronary intervention; CABG, coronary artery bypass surgery; LVEF, left ventricle ejection fraction BMI; ClCr, creatinine clearance; STEMI, ST segment elevation myocardial infarction; NSTEMI, non-ST elevation myocardial infarction. Baseline characteristics and procedural characteristics PCI, percutaneous coronary intervention; CABG, coronary artery bypass surgery; LVEF, left ventricle ejection fraction BMI; ClCr, creatinine clearance; STEMI, ST segment elevation myocardial infarction; NSTEMI, non-ST elevation myocardial infarction. Kaplan–Meier survival analysis of MB and GI-MB events at 2 years showed that anaemia was associated with lower event-free survival period (77% vs. 92%; log rank test, p = 0.006) and (89% vs. 98%; log rank test, p = 0.001), respectively (Fig. 2a and b). a Kaplan–Meier survival analysis of MB complications in patients with indication for OAC undergoing percutaneous coronary intervention, in relation to anaemia (log rank test p = 0.006). b Kaplan–Meier survival analysis of gastrointestinal MB complications in patients with indication for OAC undergoing percutaneous coronary intervention, in relation to anaemia (log rank test p = 0.001). In Cox multivariate analyses, anaemia (HR 2.15, 95%CI 1.08–4.30; p = 0.030), previous stroke (HR 2.51, 95%CI 1.23–5.11; p = 0.012), triple antithrombotic therapy (HR 2.17, 95%CI 1.11–4.25; p = 0.024) and age (HR 1.05, 95%CI 1.002–1.009; p = 0.039) were independent predictors for MB; whereas anaemia (HR 4.73, 95%CI 1.53–14.61; p = 0.007), hypertension (HR 3.57, 95%CI 1.21–10.52; p = 0.021) and previous CABG (HR 4.81, 95%CI 1.08–21.36; p = 0.039) were independent predictors for GI-MB. Thus, anaemia is a common comorbidity among AF patients with indication for oral anti-coagulation who undergo PCI-stenting, and strongly associated with increased MB complications. These findings emphasize the importance of a thorough search for predisposing bleeding sites and/or haemorrhagic diatheses.