Is warfarin associated with higher risk of thrombus in left atrial appendage than direct oral anticoagulants?

Funding Acknowledgements

Type of funding sources: None.

Background

Thrombus in left atrial appendage (LAA) is associated with cardiogenic ischemic stroke, and anticoagulation therapy is utilized to prevent thromboembolism.

Purpose

This study is to investigate the incidence of thrombus in left atrial appendage detected by transoesophageal echocardiography (TOE) under anticoagulation therapy with warfarin or direct oral anticoagulants (DOAC).

Methods

Between 2005 and 2016, the patients who underwent TOE under anticoagulation therapy due to atrial fibrillation more than one month were enrolled in this study. The patients were divided into 2 groups according to whether treated with warfarin or DOAC and baseline characteristics and incidence of LAA thrombus were assessed.

Results

Among the 313 study patients, 243 (78%) were treated with warfarin. The patients treated with warfarin were elder (median 73 y vs. 67 y [interquartile range 66 – 78 y vs. 58 – 72 y], P < 0.01), the body mass index was lower (23.2 vs. 24.0 [21.0 – 25.4 vs. 21.3 – 26.9], P = 0.03). The prevalence of male sex (64% vs. 71%, P = 0.26) were similar between the 2 groups. The previous history of hypertension (69% vs. 59%, P = 0.10), diabetes (24% vs. 19%, P = 0.32), vascular disease (30% vs. 26%, P = 0.52), and ischemic stroke were similar between the 2 groups (30% vs. 23%, P = 0.22). The prevalence of CHA2DS2-VASc score > 1 (84% vs. 59%, P < 0.01) and the d-dimer level (0.7 vs. 0.5 mcg/ml [0.5 – 1.8 vs. 0.5 – 0.5 mcg/ml], P < 0.01) were higher in the warfarin groups than those of the DOAC. The velocity of LAA was slower in the warfarin group than those of DOAC (35 vs. 55 cm/s [21 – 54 vs. 38 – 68 cm/s], P < 0.01). The incidence of detection of LAA thrombus was 19% in the warfarin group and 3% in the DOAC group (P < 0.01). In the warfarin group, the PT-INR were lower in the patients with LAA thrombus (1.38 vs. 1.66 [1.11 – 1.92 vs. 1.34 – 2.03], P = 0.03).

Conclusions

The higher risk of ischemic stroke and out of range PT-INR may be the cause of the higher incidence of LAA thrombus in the patients treated with warfarin than those with DOAC.

Abstract Figure.

Prognosis and cause of death in patients with left atrial appendage thrombus treated with or without anticoagulation therapy

Background

Thrombus in left atrial appendage (LAA) is associated with cardiogenic ischemic stroke; however, little is known about prognosis of patients with LAA thrombus detected by transoesophageal echocardiography (TOE).

Purpose

This study is to investigate the prognosis of patients with LAA thrombus and their cause of death.

Methods

Between 2005 and 2016, the patients who were performed TOE in our hospital were enrolled in this retrospective observational study. Five-year stroke free and survival curves were constructed by Kaplan-Meir method and cause of death were assessed.

Results

Among the 1263 study patients, LAA thrombus was detected in 146 (12%) patients. The patients with LAA thrombus were elder (74 y [66–79 y] vs. 70 y [62–76 y], P<0.001), than those without LAA thrombus, respectively. The prevalence of male sex (67% vs. 69%, P=0.63) were similar between the 2 groups. The prevalence of CHA2DS2-VASc score ≥2, d-dimer (1.7 mcg/ml [0.9–3.5 mcg/ml] vs. 0.8 mcg/ml [0.5–2.2 mcg/ml], P<0.001), and plasma brain natriuretic peptide (315 pg/ml [128–515 pg/ml] vs. 126 pg/ml [47–284 pg/ml], P<0.001) were higher in the patients with LAA thrombus than those without (89% vs. 78%, P=0.003). The LAA velocity was slower in the patients with LAA thrombus than those without (23 cm/s [15–34 cm/s] vs. 51 cm/s [35–72 cm/s], P<0.001). The prevalence of receiving anticoagulation therapy before (34% vs. 24%, P=0.01) and after (98% vs. 66%, P<0.001) TOE 1 month were higher in the patients with LAA thrombus than those without. The 5-year stroke free rate was lower in the patients with LAA thrombus than those without (82% vs. 93%, P<0.001); however, the 5-year survival were similar between the 2 groups (84% vs. 84%, P=0.93) (Figure). The cause of death as ischemic stroke was only 7% (1/14) and 3% (3/94), (P=0.43); the cardiac cause (14% vs. 43%, P=0.07) and the malignancy (35% vs. 29%, P=0.75) were the frequent cause of death in the patient with LAA thrombus and those without, respectively.

Conclusions

The patients who were detected thrombus in the LAA had higher incidence of ischemic stroke; however, the 5-y survival were similar. The ischemic stroke was not major cause of death in the patients with and without LAA thrombus. The higher rate of receiving anticoagulation therapy may be one of the causes of the discrepancy.

Figure 1

Funding Acknowledgement

Type of funding source: None

P684 Paradoxical cerebral infarction due to massive pulmonary embolism in extracorporeal cardiopulmonary resuscitation and surgical embolectomy

Introduction

Paradoxical cerebral infarction is a mechanism of acute ischemic stroke; however, definitive images to diagnose paradoxical embolism are not often obtained. We report a case of paradoxical cerebral embolism complicated with cardiac arrest due to massive pulmonary embolism.

Case report

A 40-year-old man presented due to sudden-onset chest pain, and was admitted to our hospital. He was restless and had cold sweat; we could not measure blood pressure. Electrocardiography showed wide QRS complex with right bundle branch block, and T wave inversion in leads V1 and III. Transthoracic echocardiography showed diffuse severe left ventricular hypokinesis, with slightly better inferior wall motion compared to other segments. Few minutes after arriving, he experienced cardiac arrest; chest compression was initiated. He was transported to the catheter laboratory, and veno-arterial extracorporeal membrane oxygenation was initiated subsequently. To diagnose the cause of arrest, we performed coronary angiography, which revealed no occluded coronary artery. Pulmonary angiograms showed bilateral proximal pulmonary artery occlusion with massive thrombi (panel A). Surgical embolectomy was performed after cardiac team discussion. After ICU admission post-surgery, pericardial effusion was increased, and the blood drained continuously from the chest tube; a large amount of blood transfusion was required. Reopen chest haemostasis was utilised. After the second ICU admission, anisocoria was observed; subsequent computed tomography showed low density and midline shift in almost the entire left cerebral hemisphere (Panel B). Carotid duplex ultrasound revealed a large thrombus saddled at the left carotid artery bifurcation (Panel C and D). We rechecked the transthoracic echocardiogram at arrival to reveal the cause of the cerebral infarction, which showed the thrombus to be at the ascending aorta (Panel E). We thought that the thrombi had moved from the lower limb to the right atrium. The massive pulmonary embolism increased the pulmonary artery and right atrial pressure, resulting in the lower pressure of the left atrium compared to that of the right atrium. The thrombi passed through the patent foramen ovale into the left atrium, moved into the left ventricle, and embolised the left internal carotid artery (Panel F). He expired due to severe neurologic injury from brain herniation.

Conclusion

In this case, although the pulmonary embolism was massive and led to cardiac arrest, the deteriorated haemodynamics improved by extracorporeal cardiopulmonary resuscitation and surgical embolectomy. However, we could not rescue the patient because of the severe neurological injury due to paradoxical embolism. Paradoxical cerebral infarction in pulmonary embolism is rare; however, we should pay careful attention to early detection of paradoxical cerebral infarction in pulmonary embolism and treatment for return of the patient to the former lifestyle.

Abstract P684 figure

P765 Does detection of thrombus in left atrial appendage increase risk of ischemic stroke and mortality?

Background

Left atrial appendage (LAA) thrombus is one of causes of cardiogenic stroke and detection of LAA thrombus by transoesophageal echocardiography (TOE) strongly suggest cardiogenic stroke. It was reported that cardiogenic stroke patients had higher in-hospital mortality about 19%; however, little is known about LAA thrombus and mortality after indexed detection of LAA thrombus. We investigated LAA thrombus detection and their prognosis including ischemic stroke and survival.

Methods

The patients who were performed TOE between 2005 and 2016 in our hospital were enrolled in this study. Patients were divided into 2 groups based on thrombus detection in the LAA, and baseline characteristics and outcomes including prevalence of 5-y stroke-free and survival from the indexed TOE were compared.

Results

Among the 1260 study patients, the follow-up duration was median 971 d (interquartile range 345 d – 2017 d), and 67% of the patients were performing TOE for atrial fibrillation (AF), 20% for cerebral infarction, and 14% for valvular heart disease. Non-valvular AF was seen in 64% of the study patients and rheumatic AF was in 2%. The age (74 y [66 y – 79 y] vs 70 y [62 y – 76 y], p < 0.001), the prevalence of male sex (67% vs 69%, p = 0.63), and hemoglobin level (13.9 g/dl [12.5 – 15.1 g/dl] vs 13.8 g/dl [12.4 – 14.9 g/dl], p = 0.49) were similar between the patients with LAA thrombus and those without. The CHA2DS2-VASc score (p = 0.008), the prevalence of receiving anticoagulation before TOA (34% vs 24%, p = 0.01), those of after TOA (98% vs 66%, p < 0.001), serum creatinine (0.92 mg/dl [0.80 – 1.10 mg/dl] vs 0.85 mg/dl [0.71 – 1.00 mg/dl], p < 0.001), d-dimer level (1.7 mcg/ml [0.9 – 3.5 mcg/ml] vs 0.8 mcg/ml [0.5 – 2.2 mcg/ml], p < 0.001), and plasma brain natriuretic peptide (315 pg/ml [128 – 515 pg/ml] vs 126 pg/ml [47 – 284 pg/ml], p < 0.001) were higher in the patients with LAA thrombus than those without. The 5-y ischemic stroke-free rate was lower in the patients with LAA thrombus than those without (p < 0.001) (Figure, Panel A); however, the 5-y survival was similar between the 2 groups (p = 0.93) (Panel B).

Conclusions

The patients who were detected thrombus in the LAA had higher incidence of ischemic stroke, but the survival rate were similar. The higher rate of receiving anticoagulation therapy in the patients with LAA thrombus may be the cause of this discrepancy. Further studies are necessary to clarify this issue.

Abstract P765 Figure

P4548Early drop in systolic blood pressure and worsening renal function in the elderly acute heart failure: how does heart rate interact?

Background

Renal dysfunction is a frequent finding in patients hospitalized for acute heart failure (AHF). Worsening renal function (WRF) during hospitalization was found to be related with a poor outcome independently of baseline renal function. Early drop in systolic blood pressure (SBP) has shown to predict WRF in AHF. However, there have been few studies that reported the impact of on-admission heart rate (HR) on the relationship between early SBP drop and WRF in the elderly AHF.

Purpose

We assessed the hypothesis that early SBP drop predict WRF in the elderly patients with AHF, and investigated that on-admission HR might have an interaction with that relationship.

Methods

SBP and HR were measured on admission and 6 times during 48 hours in the 245 elderly AHF inpatients (82.9±6.0 years old, male 49.4%). WRF was defined as a serum creatinine increase of ≥0.3 mg/dL by Day 5. Early drop in SBP was calculated as the difference between admission and the lowest value measured during the first 48 hour of hospitalization.

Results

Early SBP drop (51.3 vs 32.5mmHg, p<0.01) and on-admission HR (79.3 vs 89.6bpm, p<0.05) were significantly different between the group with WRF (n=36) and the group without WRF (n=209). In the multiple logistic regression analysis adjusted for the confounders including age, gender, hypertension, left ventricular ejection fraction, total cholesterol, BNP, baseline creatinine, beta-blockade use, intravenous loop diuretic, isosorbide dinitrate and carperitide use, early SBP drop (OR: 1.003, 95% CI: 1.003–1.03, p<0.04) and on-admission HR (OR: 0.98, 95% CI: 0.96–0.99, p<0.01) were significantly associated with WRF. The interaction term of early SBP drop by on-admission HR did not have a significant association with WRF (p=0.3).

Conclusions

In the elderly AHF patients, exaggerated early SBP drop and lower on-admission HR were shown as significant independent predictors of WRF. These two factors were additively associated with WRF. Too much reduction in SBP and that in HR might be harmful to renal circulation in AHF.