Spontaneous Echo Contrast Caused by Platelet and Leukocyte Aggregates?
Background and Purpose—Spontaneous echocardiographic contrast (SEC) is correlated to clinical thromboembolic events. We sought to determine the origin of SEC by utilizing direct analysis of left atrial blood.
Methods—We examined the blood of 13 patients with and 19 without SEC. Blood samples were taken from the femoral vein and artery and from the right and left atria after transseptal puncture. Samples were incubated with fluorescence-labeled antibodies directed against the platelet (CD41a-PE, CD42b-PE, and CD62p-FITC) and leukocyte membrane epitopes (CD45-APC and CD14-FITC). The expressed epitopes were analyzed by dual laser flow cytometry immediately after blood withdrawal.
Results—In the peripheral blood of both groups, more activation and aggregation were found in the venous blood than in the arterial blood (CD41a, P=0.007; CD14neutro, P=0.017; and leukocyte-platelet aggregates [LTAg], P=0.002). In patients without SEC, the degree of activation and aggregation of the cardiac samples closely resembled the results of the peripheral samples. The degree of activation and aggregation was significantly higher in the right atrium than in the left atrium (LTAg, P<0.01; leukocyte activation, P<0.01; CD41a, P<0.01; CD62p, P<0.02). In contrast, in patients with SEC the parameters of platelet and leukocyte activation as well as LTAg was significantly higher in the left atrium than in the right atrium of the same patient (all P<0.01). A correlation between the amount of SEC and platelet-monocyte aggregates could be found (r=0.92, P<0.0001).
Conclusions—The hypothesis that platelet aggregates are involved in the pathogenesis of SEC is supported by the fact that platelets in the left atrium of patients with SEC showed more activation.
Spontaneous echocardiographic contrast (SEC) is a smokelike echo phenomenon with a swirling pattern of blood flow that can be observed in transesophageal or transthoracic echocardiography, most often within the left atrium.
SEC is significantly associated with a history of embolism and/or left atrial thrombi detected by echocardiography, severity of mitral stenosis, reduced cardiac index, advanced age, constant atrial fibrillation, hypertension, heart failure, and large left atrial diameter. SEC is a cardiac factor most strongly associated with left atrial appendage thrombus and embolic events. SEC formation is promoted by reduced blood flow velocity and enlarged left atrium but is diminished by mitral regurgitation.1 2 3 4 5
The mechanisms leading to formation of SEC are still under discussion. Blood cell aggregates are thought to be the cause of SEC, although there is disagreement about whether red cell or platelet aggregates cause this phenomenon.
An in vitro ultrasound study by Mahony et al,6 which sought to investigate the relationship between red cell aggregation and whole-blood echogenicity, demonstrated that red blood cell aggregates are not visible on ultrasound at physiological hematocrits. Although warfarin therapy is associated with resolution of left atrial thrombus and a lower risk of thromboembolic events, neither warfarin nor aspirin is effective in suppressing left atrial SEC in nonrheumatic atrial fibrillation.7
We therefore postulated that platelets and potentially also leukocytes are activated in the left atrium of patients with SEC, which might result in platelet and leukocyte aggregates.
Subjects and Methods
This study was approved by the Ethics Committee at Leipzig University. After informed consent, 32 patients who underwent transseptal cardiac catheterization for clinical reasons were classified into 2 groups according to the presence or absence of SEC. Group I comprised 12 patients with mitral stenosis and 1 with atrial septal defect (n=13); group II, 12 patients with aortic stenosis, 5 with atrial septal defect type II, and 2 with mitral regurgitation (n=19). The presence or absence of SEC was assessed by transesophageal echocardiography. Group II patients without SEC served as controls. Patient characteristics and concomitant medications are shown in Tables 1⇓ and 2⇓. Heparin therapy was discontinued 6 hours before catheterization so that at the time of catheterization no prolongation of the partial thromboplastin time was observed. No patient showed right-heart SEC. Five additional patients with SEC were treated with 50 mg tirofiban intravenously over 5 minutes during transesophageal echocardiography.
Blood Sampling and Measuring
Blood was drawn from the femoral artery, the femoral vein (both through a 5F Cordis sheath), the right atrium immediately before, and the left atrium immediately after transseptal puncture through the same catheter. The first 10 mL of blood from each sample was discarded.
Samples were incubated with fluorescence-stained antibodies against the platelet and leukocyte membrane receptors that are known to be activation markers.8 Thrombin receptor activating peptide-6 (TRAP-6)–stimulated samples served as positive controls; isotypic controls were used for negative control. Blood sample preparation was based on the consensus protocol for flow cytometric analysis of platelet function,9 with slight modifications.10
The following monoclonal antibodies were used in this study: anti-CD41a (GP IIb/IIIa) and anti-CD42b receptor (GP Ib), both conjugated with PE, and anti-CD62p (Beckman Coulter), conjugated with FITC. CD41a or CD42b was used to identify platelets and assess the degree of platelet activation. Platelet activation was further measured by expression of CD62p, which is acknowledged to be a platelet activation marker.8 Monoclonal antibodies CD45 (panleukocyte antigen, Medac), conjugated with APC, and CD14 (LPS-receptor, Becton-Dickinson) conjugated with FITC, were used for identification of leukocytes as well as to determine their activation status. A total of 100 μL of the citrated whole-blood sample was diluted with 900 μL PBS (Sigma Chemical Co); 60 μL of the diluted blood was incubated with 30 μL TRAP-6 for positive control, and 60 μL was left unstimulated (incubation with 30 μL PBS). The samples were incubated for 10 minutes at room temperature before they were divided into 2 aliquots: for determination of the surface density of CD41a (Gp IIb/IIIa) and to measure the surface density of antigens on leukocytes and their platelet aggregates, 15 μL of the diluted blood was incubated for 10 minutes with a cocktail of anti-CD41a-PE, anti-CD14-FITC, and anti-CD45-APC; to determine the surface density of CD42b receptor (vWF receptor) and CD62p (P-selectin), 15 μL of the diluted sample was incubated with anti-CD42b receptor-PE and anti-CD62p-FITC. All antibodies were used in saturating concentrations. They were titrated under a maximal stimulated condition (TRAP-6), except for anti-CD42b receptor-PE, which was titrated with use of unstimulated blood. After incubation with the antibodies, 0.5 mL refrigerated PBS was added to all samples; the samples were placed on ice and analyzed immediately afterward. Fixation of the samples was omitted to avoid fixation artifact, and freshly stained samples were measured within 30 minutes after sampling.
Samples were analyzed on a dual laser flow cytometer (FACScalibur, Becton-Dickinson). Platelets were detected in a dot-plot log forward-angle light scatter (forward-angle light scatter) versus log 90°-angle light scatter (90°-angle light scatter) and forward-angle light scatter versus log CD41a-PE or CD42b receptor-PE fluorescence in the orange channel. Forward-angle light scatter, 90°-angle light scatter, and green (FITC), orange (PE), and red (APC, excited by the second laser) fluorescence were acquired on a logarithmic scale. To ensure day-to-day sample reproducibility, the setup of the flow cytometer was controlled daily with calibration beads for 488-nm and 630-nm excitation (Spherotech). When necessary, the instrument was recalibrated. An appropriate threshold was set in the forward-angle light scatter and the fluorescence channels to exclude debris and electronic noise. To exclude red blood cells, a gate was acquired. Fluorescence compensation was determined by using single-antibody-stained samples. Cells were measured at a count rate of approximately 5000 events at low-sample pressure to analyze platelets and at high-sample pressure for the leukocytes and aggregates (>1000 events per second). The data were analyzed with the Cellquest software package (version 3.1f, Becton-Dickinson). Cell populations were differentiated as described by our laboratory before.10
Echocardiography and Video Intensitometry
Patients were scanned with transesophageal echocardiography with a HP Sonus 5500 echocardiograph. The gain settings of the echocardiograph were identical in all patients and left unchanged throughout the study. For video intensitometric measurement, the position of the echoscope was kept constant for 10 seconds. In a window of measurement (2×2 cm) within the left atrium, the area under the intensity curve was measured video intensitometrically over 10 seconds from the S-VHS videotapes.11 Because the information of the videotape enters the intensitometer directly, no manipulations can be performed.
In an in vitro model, videointensity was measured over a period of 10 seconds with the transducer held on a water tank containing a monovette. This monovette was filled with 7 mL PBS. Three milliliters of blood were added and stimulated with 5 mL TRAP-6 (to induce aggregates), and 2 mL anti-CD62p were added thereafter. Videointensites were measured before and after the addition of anti-CD62p. This experiment was repeated 8 times, and care was taken that the suspension was standardized by using a vortex machine.
For all parameters measured, fluorescence intensity (mean±SD) was tabulated (Table 3⇓). The ratio of left atrial to right atrial intensity was calculated and plotted as percent. Thus, every patient served as his or her own control. After the normal distribution of data had been confirmed with the Kolmogorov Smirnov test, an unpaired t test was applied to calculate differences between groups. For comparison between area under the video intensitometric curve and the amount of platelet-monocyte aggregates, a linear regression analysis was performed (Graph Pad Prism, Graph Pad Software Inc).
The results of this case control study relating to surface antigen expression on platelets and leukocytes as well as to platelet-leukocyte aggregates in the right and left atria are shown in Table 3⇑.
In patients with SEC, the activation status of platelets and leukocytes as well as the degree of aggregates was higher in the left than in the right atrium of the same individual. Enhanced platelet activity was measured by increased surface antigen expression of CD62 (P-selectin) and CD41a (GP IIb/IIIa) and by a decrease of CD42b (vWF). Increased leukocyte activity was detected by a rise in leukocyte antigen expression measured in the left atrium compared with the right atrium. Furthermore, a higher degree of platelet monocyte aggregation and platelet neutrophil aggregation were detected as indicators for an enhanced activation status (Figure 1⇓).
We found a higher degree of platelet and leukocyte activation and platelet-leukocyte aggregation in venous compared with arterial blood. Nearly identical values between femoral and cardiac samples were found in patients without SEC (Table 3⇑).
In patients with SEC, the samples of the left atrium showed a significantly higher degree of platelet-leukocyte activation and aggregation than did those of the arterial femoral blood sample. The right atrial sample and the venous femoral sample in patients with SEC were almost identical (Table 3⇑).
A correlation between the video intensitometric amount of SEC (area under the videointensity curve) and the amount of platelet-monocyte aggregates could be found (Figure 2⇓). In 5 patients treated with tirofiban, only a slight decrease in videointensity (area under the curve for 10 seconds) could be found (−12±7%).
In vitro treatment of blood with TRAP-6 induced aggregates (Figure 3⇓). Addition of anti-CD62p led to a significant reduction of aggregates and video intensitometric indication of SEC.
In this study we found a higher degree of platelet and leukocyte activation as well as platelet-leukocyte aggregation in venous blood. This can be explained by the lower flow rate in the venous system and might be one reason for the higher incidence of thrombosis within the venous system compared with the arterial system. In addition, the lung acting as a microfilter clears potential aggregates and releases potent mediator substances such as NO and PGI2, which might lead to a lower number of aggregates in the arterial system.
Studies that examined the relationship between SEC and hematological parameters found a positive correlation between the incidence of SEC with fibrinogen level and hematocrit but no correlation with platelet counts.6 Therefore, the hypothesis was raised that left atrial SEC is caused by erythrocyte aggregation. An in vitro ultrasound study12 came to the conclusion that echogenic contrast appears to be primarily caused by the interaction of red blood cells and plasma proteins at low flow and low shear rate conditions. In this study, we did not investigate red blood cells or interactions of red blood cells with leukocytes or other blood components.
Another study13 identified β-thromboglobulin and von Willebrand factor as the independent associates of left atrial thrombosis, ahead of the presence of SEC, which indicates the participation of platelets in the occurrence of SEC. Hwang et al14 sought to examine a correlation among echocardiographic variables, hematologic parameters, or platelet aggregability and the occurrence of SEC. Platelet aggregability was evaluated by a turbidometric method, using different concentrations of activating agents (adenosine diphosphate and collagen). No significant difference was found in platelet aggregability between patients with left atrial SEC and patients without left atrial SEC who were not receiving antiplatelet or anticoagulant therapy. These results do not contradict the results of this study, because no direct spontaneous blood activity was measured by using surface receptors on platelets and leukocytes. Furthermore, no direct puncture of the left atrium was performed.
Ileri et al,15 using thrombin–antithrombin III complexes and prothrombin fragments 1+2 as in vivo hemostatic markers, obtained results similar to those in this study. They found that there is a hypercoagulable status in patients with mitral stenosis and sinus rhythm when SEC is present. Pongratz et al16 investigated the activation status of platelets in the peripheral blood of patients with atrial fibrillation. A significantly higher amount of circulating platelets expressing P-selectin and CD63, and more leukocyte-platelet conjugates were found in patients positive for both SEC and left atrial thrombus or embolic events. An increased spontaneous platelet aggregation in the presence of SEC was described by Rohmann et al,17 who measured the half-maximal formation of platelet aggregates on stimulation with ADP in peripheral blood.
In this study we found a profound increase in left atrial platelet and leukocyte activation in patients with SEC, as indicated by increased surface antigen expression on platelets and leukocytes. Furthermore, as a result of an increased activation status, we found a higher degree of platelet leukocyte aggregation in the left atrium of patients with SEC.
These results support the hypothesis that the origin of SEC is platelet aggregates. The formation of platelet-leukocyte aggregates may contribute to further activation of circulating platelets through the release of specific mediator substances (eg, thromboxane A2 or ADP). Activated monocytes produce tissue factor in high quantities, which is the main initiator of the extrinsic coagulation cascade. Therefore, thrombus formation is the likely result that might lead to thromboembolic complications. Anticoagulative therapy with phenprocoumon or warfarin is likely to reduce thrombus formation because of the activated coagulation cascade following tissue factor release from activated monocytes. In addition to the activation of monocytes and the adhesion of activated platelets that we observed, another sequela of platelet activation is the formation of platelet aggregates. These platelet aggregates will not be reduced by oral anticoagulants, which might explain the increased incidence of thromboembolic complications even in the presence of a sufficient anticoagulant therapy.7
It has been demonstrated18 that antiplatelet therapy, but not anticoagulation, is able to reduce platelet-monocyte interaction and consequently reduce the degree of monocyte activation. Therefore, excessive release of tissue factor is prevented, and the initiation of the coagulation cascade via tissue factor is reduced by antiplatelet therapy. Antiplatelet therapy is able to reduce the formation of platelet aggregates, thereby inhibiting the second possible pathway of thrombus formation, as well as the pathway mediated via the coagulation cascade. We conclude that combined antiplatelet therapy, by inhibiting several pathways of platelet activation simultaneously, may aid in the treatment of SEC. Combined antiplatelet therapy inhibits or at least diminishes both pathways of thrombus formation at the same time, whereas oral anticoagulation is able to inhibit only the coagulation cascade leaving activated platelets and the platelet-induced tissue factor unimpaired.
One could argue that the cytometric data of this study do not prove that platelets are a component of SEC and that platelets are a “bystander” of SEC. However, patients with denser SEC did have a greater density of activated platelets. This is a strong argument in favor of the hypothesis of this study.
Due to the small sample size of this case-control study, results need to be confirmed in a larger patient cohort possibly treated in a randomized fashion with anticoagulants, antiplatelet drugs, and (if possibly available) a human anti-CD62p. One could argue that blood withdrawal from the left atrium via a Brockenborough catheter is a platelet-activating procedure. However, the identical procedure was performed for blood withdrawal from the right atrium. Thus, activation via the catheter should have been the same for all central blood samples. The finding that right atrial and right femoral activation markers were nearly identical makes an excessive artificial activation due to the manipulation with the catheter highly unlikely. This study was performed in patients on therapy; thus, influences of this therapy cannot completely be ruled out. Also, the number of patients on therapy at the time of catheterization was relatively small and not biased in the direction of the results obtained. A positive control, in which patients on glycoprotein IIb/IIIa antagonists were investigated, was not performed in this study because this therapeutic modification might cause a substantial bleeding risk in patients undergoing transseptal puncture for diagnostic reasons. Other factors that might possibly play a role in this setting, such as red blood cells, were not analyzed in this study.
This is the first report to analyze left atrial blood in patients with spontaneous echo contrast. It demonstrated a profound increase in the activation status of platelets and leukocytes compared with right atrial blood. The amount of SEC correlated with the number of leukocyte-platelet aggregates. It seems conceivable that instead of conventional anticoagulation, therapy aimed at the prevention of platelet-leukocyte activation might be more successful in reducing the increased incidence of thromboembolic events in this patient cohort.
- Received October 4, 2000.
- Revision received December 15, 2000.
- Accepted January 11, 2001.
- Copyright © 2001 by American Heart Association
Chimowitz MI, DeGeorgia MA, Poole RM, Hepner A, Armstrong WM. Left atrial spontaneous echo contrast is highly associated with previous stroke in patients with atrial fibrillation or mitral stenosis. Stroke.. 1993;24:1015–1019.
Rinder HM, Bonan JL, Rinder CS, Ault KA, Smith BR. Activated and unactivated platelet adhesion to monocytes and neutrophils. Blood.. 1991;78:1760–1769.
Heppell RM, Berkin KE, McLenachan JM, Davies JA. Haemostatic and haemodynamic abnormalities associated with left atrial thrombosis in non-rheumatic atrial fibrillation [see comments]. Heart.. 1997;77:407–411.
May AE, Neumann FJ, Gawaz M, Ott I, Walter H, Schömig A. Reduction of monocyte-platelet interaction and monocyte activation in patients receiving antiplatelet therapy after coronary stent implantation. Eur Heart J.. 1997;18:1913–1820.