This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Geiser, T. Articles by Beer, J. H. Search for Related Content PubMed PubMed Citation Articles by Geiser, T. Articles by Beer, J. H. Pubmed/NCBI databases Compound via MeSH Substance via MeSH

(Stroke. 1998;29:1770-1777.)
© 1998 American Heart Association, Inc.

Original Contributions

Mechanisms of Cerebrovascular Events as Assessed by Procoagulant Activity, Cerebral Microemboli, and Platelet Microparticles in Patients With Prosthetic Heart Valves

Thomas Geiser, MD; Matthias Sturzenegger, MD; Urs Genewein, MD; André Haeberli, MD; Jürg H. Beer, MD

From the Departments of Internal Medicine (T.G., U.G.) and Neurology (M.S.) and the Laboratory of Thrombosis Research (T.G., A.H., J.H.B.), University Hospital, Bern, Switzerland.

Correspondence to Matthias Sturzenegger, MD, Department of Neurology, University Hospital, CH-3010 Bern, Switzerland. E-mail matthias.sturzenegger{at}insel.ch

	Abstract

Top Abstract Introduction Subjects and Methods Results Discussion References

 
Background and Purpose—Cerebrovascular events (CVE) in patients with prosthetic heart valves (PHV) have remained a severe and frequent complication despite oral anticoagulation with or without aspirin. We studied the possible pathophysiological involvement of platelet-derived microparticles (PMP) as a contributing factor for the increased incidence of CVE in patients with PHV.

Methods—We compared in a retrospective, case-control study the clinical outcome after the implantation of the PHV with several different independent morphological and functional methods, including simultaneous transcranial Doppler monitoring of both middle cerebral arteries, PMP detection by flow cytometry with use of platelet-specific antibodies, coagulation markers, and determination of the procoagulant activity by Russell's viper venom time, a phospholipid-dependent coagulation assay.

Results—Eight of 26 patients with PHV had 9 CVE during 136 person-years of observation. Transcranial Doppler monitoring revealed an increased frequency of microembolic signals recorded over a 30-minute period in patients with CVE (75±25; median, 55; range, 27 to 248) compared with those without CVE (23±12; median, 7; range, 0 to 153; P<0.05) or with control subjects (0; P<0.001). Flow cytometry analysis showed increased levels of PMP in patients with compared to those without CVE (4.1±0.6% versus 2.4±0.4% of all fluorescence-positive events gated; P<0.05). Increased procoagulant activity was documented by the shortened Russell's viper venom time expressed as an increased level of platelet equivalents per microliter of plasma in patients compared with control subjects (+24.7±14.9%; P<0.01). Subgroup analysis revealed that patients with CVE had a higher excess of platelet equivalents per microliter of plasma than patients without CVE in relation to the controls (+68.7±36.7%; P<0.05). Mildly elevated thrombin–antithrombin III complexes (2.9±0.7; median, 2.3; normal, <2.0 µg/L) suggested incompletely suppressed thrombin formation, and fibrin generation (fibrinopeptide A) was in the upper normal range (2.1±0.2; median, 1.8; normal, <2.0 ng/mL), despite adequate anticoagulation (INR=3.6±0.1).

Conclusions—Our data show increased microembolic signals, platelet microparticles, and procoagulant activity in symptomatic patients with PHV and provide a potential pathophysiological explanation of CVE.

Key Words: cerebral embolism • coagulation • heart valve prosthesis • platelets • procoagulant

	Introduction

Top Abstract Introduction Subjects and Methods Results Discussion References

 
The incidence of thromboembolic events in patients with prosthetic heart valves (PHV) is in a range of 1% to 5% per year despite oral anticoagulation, which is thought to be optimal with an INR of 3.0 to 4.5. In most cases (approximately 85%) the cerebral circulation is involved.^{1 2 3 4} Approximately 75 000 mechanical PHV annually are implanted worldwide.⁵ As the risk is continuous and cumulative, several thousand strokes occur every year in this patient population, and approximately 5% to 10% of them are fatal.^{1 6}

Several studies have tried to define the optimal range of anticoagulant treatment, optimizing between thromboembolic and bleeding complications. The results demonstrate a U-shaped curve with an optimal therapeutic level of INR 3.5.¹ In addition, combinations with antiplatelet agents at different dosages have been studied, but in the most pertinent studies the annual incidence of thromboembolic events has remained at 2% to 3%.⁴

Further studies should therefore focus on the analysis of the pathophysiological mechanisms leading to cerebrovascular events (CVE) in patients with PHV. Transcranial Doppler ultrasonography is able to detect signals representing microemboli in the basal cerebral arteries.^{7 8 9 10 11} Recent studies^{12 13 14} have shown conspicuously high frequencies of such high-intensity transient signals (HITS) in patients with PHV. The debate over whether these HITS are formed elements (eg, platelet aggregates) or gaseous bubbles resulting from cavitation is still not settled,^{12 13} and their clinical significance is disputed.

Increased shear rates at the PHV can activate platelets¹⁵ and generate platelet-derived microparticles (PMP). Such generated platelet fragments provide negatively charged surfaces and have a potent procoagulant activity.¹⁶ The phospholipids facilitate the assembly of the prothrombinase complex and may accelerate the coagulation cascade and thrombin formation several thousandfold,¹⁷ while multiple receptors (platelet glycoprotein [GP] IIb/IIIa and GP Ib) on PMP function as cross-links for mixed thrombi.¹⁸

We therefore hypothesized that the CVE might result from an increased procoagulant activity mediated by PMP generated at the PHV and analyzed the clinical course after implantation of the PHV in relation to several independent morphological and functional parameters. Microembolic signals in both middle cerebral arteries (MCAs) were measured by transcranial Doppler ultrasound, and PMP were detected and quantified by flow cytometry using 2 well-characterized platelet-specific monoclonal antibodies directed against 2 different platelet glycoproteins. Procoagulant activity was determined by the Russell's viper venom (RVV) time (RVVT), and thrombin and fibrin generation were assessed by the currently available methods.

	Subjects and Methods

Top Abstract Introduction Subjects and Methods Results Discussion References

 
Patients and Controls
Twenty-six patients (aged 61.5±1.8 years; 14 women and 12 men) with PHV gave their informed consent to participate in this retrospective case-control study. The study was approved by the Ethics Committee of the Medical Faculty of the University of Bern (Switzerland).

The 26 patients examined were referred to us at our request from cardiologists in the catchment area of our hospital. (None of our group ordinarily take care of patients with PHV.) There were no special selection criteria other than the following: patients having a PHV with an interval between valve implantation and study enrollment of at least 3 months and patients not having a major cerebrovascular event before PHV implantation. Patients with major CVE prior to the PHV implantation were excluded.

Patients were classified into the following 3 groups: A, those with proven cerebrovascular events; B, those with nonspecific transient neurological symptoms, such as vertigo, blurred vision, or recurrent amnesias; and C, those without cerebrovascular events. Classification was based on the patient's history and the clinical and neuropsychological examination by the same neurologist (M.S.) and performed before further examination.

Cardiovascular risk factors (hypertension, smoking, diabetes, hyperlipidemia, and atrial fibrillation) were distributed similarly in the patients with and without CVE. Patient characteristics and data concerning the PHV implantation are summarized in Table 1. Patients who suffered from cerebral infarction or transient ischemic attacks (TIAs) were extensively evaluated clinically, radiologically, and with blood analysis, including cerebrovascular ultrasound and transesophageal echocardiography studies, to exclude causes of platelet activation other than the PHV.

View this table: [in this window] [in a new window]   Table 1. Characteristics of Patients With PHV

Five patients were analyzed in part in a previous preliminary study, in which markers of plasmatic coagulation and Doppler microemboli counts were measured.¹⁹ The patients agreed to be reexamined according to this more elaborate protocol.

All patients were on a regimen of oral anticoagulant therapy with either acenocoumarol (n=4) or phenprocoumon (n=22), with a target INR of between 3.0 and 4.0.¹ Actual mean INR was 3.6±0.1 throughout the observation period. Further medication consisted of ACE inhibitors (11 patients), diuretics (10 patients), digoxin (7 patients), calcium antagonists (5 patients), ß-blockers (7 patients), nitrates (3 patients), amiodarone (2 patients), antibiotics (2 patients), oral antidiabetics (1 patient) or insulin (1 patient), low-dose estrogen replacement therapy (2 patients), benzodiazepines (2 patients), and thyroxine substitution (2 patients). The distribution of the medication was similar in the patients with and without CVE except for aspirin, which was given to 4 of 12 patients with CVE or nonspecific neurological symptoms and 1 of 14 patients with no CVE. One patient with CVE was on additional dipyridamole.

Control subjects were 26 healthy, age- and sex-matched volunteers who were examined according to the protocol on the same day as the patient. Except for 1 subject with well-controlled, mild hypertension on an ACE inhibitor, none of the control subjects took any medication.

Transcranial Doppler Ultrasonography
Studies were performed according to a protocol described elsewhere.¹⁹ Briefly, patients were examined under standard conditions in a supine position. The recording, performed with a 2-channel transcranial Doppler system (Multi-Dop X/TCD 7, Firma DWL, Elektronische Systeme GmbH) equipped with a specially developed software for automated emboli detection, was started after the subjects had rested for at least 20 minutes in a quiet room. The Doppler frequency spectra of both MCAs were recorded simultaneously and continuously during 30 minutes. HITS were identified by their short (<0.1 second), transient, unidirectional, high-amplitude signal with a narrow spectrum⁸ and counted by 3 different methods: (1) visually (on the monitor displaying the fast Fourier transform of the color-coded Doppler spectra), (2) acoustically (by continous on-line observation by an examiner using headphones), and (3) computer-assisted (with a special software). The detected HITS counts in the left and right MCAs were added, thus reflecting the total over the registered 30 minutes.

Laboratory Studies
Blood was drawn from patients and controls by a specially trained technician through clean venipuncture from an antecubital vein under controlled venous stasis of 60 mm Hg for maximally 45 seconds. Previous studies have shown normal values for fibrinopeptide A and thrombin–antithrombin III complexes for this technique.²⁰ The Monovette system (Sarstedt) was used. A total of 45 mL of blood was drawn. The first 5 mL was discarded, then 5 parts blood was added to 1 part acid citrate dextrose (citric acid 38 mmol/L, sodium citrate 74.8 mmol/L, dextrose 124 mmol/L; pH 5.0) and prepared for flow cytometry and RVVT (see below); 9 mL blood was added to 1 mL citric acid (110 mmol/L) with the inhibitors theophylline (15 mmol/L), adenosine (3.7 mmol/L), dipyridamole (0.198 mmol/L), and phenyl-prolyl-arginine-chloromethylketone (0.62 mmol/L), pH 5.0 (CalBiochem), for the measurement of fibrinopeptide A (RIA, reagents by Imco), prothrombin fragment 1+2 (ELISA, Enzygnost F1+2, Behring), thrombin–antithrombin III complex (ELISA, Enzygnost TAT micro, Behring), D-dimer (Enzygnost D-dimer micro, Behring) and ß-thromboglobulin (RIA, supplied by Amersham).

Fibrinopeptide A was determined by polyclonal antibodies. Cross-reacting fibrinogen was eliminated by bentonite absorption, resulting in a specifity of 100%. Free antigen was separated from bound antigen by the use of a secondary donkey anti-rabbit antibody (IDS Ltd).

In addition, 4.5 mL blood was drawn into 0.5 mL of 106 mmol/L trisodium citrate for the assessment of the prothrombin time, and another 4.0 mL was drawn into 6.4 mg dry EDTA for the determination of leukocyte and platelet count and hemoglobin concentration. The tubes were put on melting crushed ice immediately after blood sampling for no more than 15 minutes, then centrifuged at 4°C for 30 minutes at 2000g. Multiple aliquots were snap-frozen in liquid nitrogen and stored at -70°C until analysis.

For quality control reasons, the coagulation assays were repeated in 4 patients on a separate occasion and gave essentially the same results. In these cases the mean of the 2 examinations was used. Two patients were excluded because of grossly elevated fibrinopeptide A levels (and normal thrombin–antithrombin III complex), suggesting fibrin formation in vitro. Therefore, the above constellation makes an improper blood sampling and an in vitro artifact very likely.

Flow Cytometry
Platelet-rich plasma (PRP) was prepared immediately after venipuncture by centrifugation with 800g for 2 minutes at room temperature. Platelets were counted (Coulter Counter, Coulter Electronics Ltd, or Sysmex K-1000, Toa Medical Electronics) and adjusted to 300 000/µL with sterile-filtered (0.2-µm pore size, Schleicher and Schuell) and degazed PBS. Platelets were fixed within 10 minutes after venipuncture by adding paraformaldehyde (ratio 1:1) in PBS, pH 7.4, to achieve a final concentration of 0.5%. The fixation step was early because several monoclonal antibodies have been reported to induce vesiculation.²¹ For analysis by flow cytometry, the following monoclonal antibodies (Mab) were used: Mab 6D1, directed against GP Ib, and Mab 7H2, directed against GP IIIa (both kindly provided by Dr B.S. Coller, Mount Sinai Hospital, New York, NY) and Mab against P-selectin respectively CD62P (Serotec) in order to include a marker of platelet activation. We also used monoclonal antibodies directed against a red cell antigen (Glycophorin-A, Serotec) and a leukocyte antigen (CD44, Dako) to obtain data on microparticles potentially derived from the other cell lines. Nonspecific mouse IgG1 (Dako) and PBS instead of the primary antibody were used as negative controls. Incubations with the first, GP-specific antibody (Mab 6D1, 7H2), or aCD62, aGlycophorin, and aCD44, were performed for 45 minutes at a final concentration of 10 µg/mL, which is definitely above the saturation level. Fluorescein isothiocyanate (FITC)-labeled goat anti-mouse antibody [F(ab') fragment of affinity-isolated goat anti-mouse immunoglobulins, final concentration 1.5 µg/mL, from Dako)] was then added without washing (so as not to lose any microparticles) and incubated for another 45 minutes. All steps were performed at room temperature. After a 20-fold dilution with PBS (typically, 100 mL sample was diluted with 1.9 mL buffer), the samples were analyzed on a Becton-Dickinson FACScan equipped with 15-mW air-cooled argon laser (Becton-Dickinson), and 10 000 gated events per sample were analyzed with Cell Quest software (Becton-Dickinson). Standard beads of different sizes (2, 0.5, 0.1, and 0.01 µm) were used to calibrate the system (Polysciences Inc). PMP were defined by their size and their specific positive fluorescence for platelet glycoprotein and quantified as described by Warkentin et al.²² Briefly, the number of GP Ib– or GP IIIa–positive particles of <87 relative fluorescence units was expressed as percentage of the total number of GP Ib– or GP IIIa–positive particles gated. This cutoff value was derived from the analysis of PRP of 25 normal donors, in accordance with the findings of Warkentin et al.²² This threshold (87 fluorescence units) separates optimally the well-defined platelet population from microparticles, whereas nonspecific binding of the control mouse Mab sets a much lower fluorescence margin.

Russell's Viper Venom Time
Procoagulant activity was determined by the dilute RVVT, a sensitive, phospholipid-dependent coagulation assay.²² To 100 µL of pooled, ultracentrifuged (2000g for 20 minutes, followed by 15 000g for 20 minutes) and filtered (0.2-µm pore size) plasma from 12 healthy donors, we added 100 µL RVV solution (1 µg/mL) (Pentapharm AG, with high purity grade) and 20 µL of a PMP-containing solution; the assay was then started by recalcification (100 mL of 20 mmol/L CaCl₂). Coagulation times were recorded with use of an ST4 instrument (Diagnostica Stago). To quantify procoagulant activity, a standard curve was obtained with frozen-thawed platelets as a source for PMP, as proposed by Warkentin et al.²² This standard was prepared by freezing platelets (300x10⁹ /L) twice in liquid nitrogen for 20 minutes, followed by thawing (37°C). Final concentrations ranged from 50 000 down to 1 single platelet equivalent of frozen-thawed platelets per microliter of plasma. The RVVTs were then expressed as platelet equivalents per microliter of plasma.

Different concentrations of platelets were measured in addition to platelet-poor plasma (PPP) and high-speed PPP (HS-PPP). PPP was prepared by centrifugation of the whole blood with 2 000 for 20 minutes and HS-PPP by an additional centrifugation with 15 000g for 20 minutes, followed by filtration with a 0.2-µm filter.

Statistical Analysis
Values were analyzed by the Student t test and Wilcoxon-Mann test for paired data, the Mann-Whitney test for unpaired data, and ANOVA, where appropriate. For multiple testing, the Bonferroni-Holm test was used. Differences were considered significant at P<0.05. Values are given as mean±SEM; in some instances the median values were analyzed as well, as indicated.

	Results

Top Abstract Introduction Subjects and Methods Results Discussion References

 
Clinical Course
The total accumulated observation time of the 26 patients evaluated was 136 patient-years. Time span between PHV insertion and enrollment in the study was 5.2±4.6 years (range, 8 months to 20 years). In patient group A, the following 9 CVE had occurred in 8 patients: 3 completed strokes (cerebrovascular infarctions [CVIs]), 1 prolonged reversible ischemic neurological deficit (PRIND), 1 single TIA, and 4 recurrent TIAs (2 or more events), corresponding to an incidence of 6.6 per 100 patient-years for all CVE and of 2.9 per 100 patient-years for definite stroke (CVI and PRIND). In group B, 4 patients complained of nonspecific transient neurological symptoms such as recurrent attacks of vertigo (3 occurrences); recurrent amnesias (1 occurrence); and visual disturbances, such as blurred or flickering vision (2 occurrences). All symptoms occurred exclusively after the valve replacements and at least 3 months before laboratory workup and transcranial Doppler monitoring, despite a level of oral anticoagulation considered to be therapeutically optimal for these patients (INR, 3.0 to 4.5). Extensive clinical, laboratory, and ultrasonographic evaluation in the patients with CVE did not disclose any other sources of thromboemboli or other acute illnesses that may induce platelet activation or PMP generation. Fourteen patients had neither CVE nor nonspecific transient neurological symptoms during the observation time (group C).

Doppler Emboli Monitoring
The majority of patients with PHV (81%) showed a strikingly high frequency of HITS, with a wide range (48±12 emboli per 30 minutes; median, 24; range, 0 to 248). Control subjects had no HITS (P<0.001).

The subgroup analysis of the patients in relation to their clinical events revealed 75±25 HITS (median, 55; range, 27 to 248) in the group of patients with CVE (group A, n=8); 83±41 HITS (median, 63; range, 14 to 191) in the group of patients with nonspecific transient neurological symptoms (group B, n=4); and in sharp contrast, 23±12 HITS (median, 7; range, 0 to 153; P<0.05) in patients without CVE (group C), which is a 3- to 4-fold lower emboli count (Figure 1).

View larger version (18K): [in this window] [in a new window]   Figure 1. Increased emboli counts in patients with CVE (group A), nonspecific transient neurological symptoms (vertigo, amnesia, and visual disturbances) (group B), and without CVE (group C). All patients and controls were monitored for HITS with transcranial Doppler ultrasound during a 30-minute period simultaneously in both MCAs. Data are mean±SEM of the total number of HITS in both MCAs. Statistical differences are indicated (*P<0.05, **P<0.001).

Flow Cytometry
The number of PMP, as quantified from freshly drawn, carefully standardized and fixed PRP, was dramatically increased by 70.1% in patients with documented CVE (group A, n=8) compared to those without CVE (group C, n=14): 4.1±0.6% with CVE and 2.4±0.4% without CVE, using Mab 7H2; P<0.05. The latter group of patients did not differ from the healthy control subjects (2.2±0.2%). In addition, the patients with nonspecific transient neurological symptoms (group B, n=4) did not show significant differences in the number of PMP compared with control subjects or patients without CVE.

Considering all patients together, the number of PMP was slightly higher compared with the controls, but it did not reach significance (2.9±0.3% compared with 2.2±0.2% using anti-GP IIIa and 2.2±0.2% compared with 2.0±0.2% using anti-GP Ib [both NS]; Figure 2). The Mab directed against GP IIIa appeared more sensitive for the analysis of PMP, therefore distinguishing better between the 3 patient groups and the control group.

View larger version (17K): [in this window] [in a new window]   Figure 2. Increased PMP in patients with CVE (group A), nonspecific transient neurological symptoms (vertigo, amnesia, and visual disturbances) (group B), and without CVE (group C). PMP were quantified with flow cytometry using 2 different platelet-specific antibodies (GP IIIa or GP Ib) and are shown as percentage of the total fluorescence-positive events gated. Data are presented as mean±SEM; statistical differences are indicated (*P<0.05).

The binding to P-selectin or CD62P on PMP surfaces was low and similar in patients and controls (0.3% of all events gated). Microparticles derived from red cells and leukocytes were present at very low concentrations as well (0.4±0.08 in patients compared with 0.2±0.02% in controls for glycophorin-A [NS] and 0.1% in both groups for CD44).

Coagulation Markers
To detect increased thrombin or fibrin generation or fibrinolysis, prothrombin fragment 1+2, thrombin–antithrombin III complex, fibrinopeptide A, D-dimer, and ß-thromboglobulin were measured. Prothrombin fragment 1+2, D-dimer, and ß-thromboglobulin were within normal ranges (see Table 2), whereas the thrombin–antithrombin III complex levels (2.9±0.7 µg/L; median, 2.3; normal, <2.0) and fibrinopeptide A levels (2.1±0.2 ng/mL; median, 1.8; normal, <2.0) were slightly elevated despite oral anticoagulation. One patient, however, showed extremely high levels of prothrombin fragment 1+2 (2.42 nmol/L), thrombin–antithrombin III complex (17.86 µg/l), and fibrinopeptide A (3.62 ng/mL); tragically, he developed a CVI 1 day after testing.

View this table: [in this window] [in a new window]   Table 2. Coagulation Markers

There were no significant differences when the results of the different clinical subgroups (groups A, B, and C) were analyzed (Table 2).

Procoagulant Activity (RVVT)
Patients had a significantly higher amount of (frozen-thawed) platelet equivalents per microliter of plasma compared with control subjects (P<0.01; Figure 3), which is also expressed in a significant shortening of the RVVT. These differences appeared most prominent in diluted samples with low platelet concentrations. When the platelets, but not the PMP, were removed by differential centrifugation (PPP), the effect could still be observed, whereas it disappeared after the elimination of PMP by high speed centrifugation and filtration of the plasma (HS-PPP) (Figure 3). The excess of platelet equivalents (comparison of platelet equivalents between patients and control subjects) was most prominent in the group of patients with CVE (group A) (+68.7±36.7%; P<0.05) compared to the patients with nonspecific transient neurological symptoms (group B) or those without CVE (group C), where no significant differences compared to the controls could be found (Figure 3, insert).

View larger version (29K): [in this window] [in a new window]   Figure 3. Increased procoagulant activity in patients with PHV. The RVVT was determined with use of PRP at different platelet concentrations in addition to PPP and HS-PPP, ie, decreasing concentrations of platelet membranes and membrane fragments, respectively, as a source of phospholipids. Filled bars represent patients and hatched bars controls. Values are the mean±SEM of frozen-thawed platelet equivalents per microliter of plasma. Insert, The excess of platelet equivalents (platelet equivalents of patients compared with controls) is shown in the different clinical subgroups (group A, patients with CVE; group B, patients with nonspecific transient neurological symptoms; and group C, patients without CVE) (*P<0.05).

	Discussion

Top Abstract Introduction Subjects and Methods Results Discussion References

 
Ischemic CVE in patients with PHV may have 2 causes: (1) the level of anticoagulation is unintentionally too low or (2) thrombogenic mechanisms are not sufficiently suppressed by the reduction of coagulation factors during adequate coumadin prophylaxis. A number of well-designed and well-conducted studies have shown that it is possible to reduce (but not completely suppress) the incidence of CVE by actually achieving the targeted INR in a high percentage of the prothrombin times measured.^{1 2} However, the precise pathophysiology of thromboembolism in patients with PHV is still not established. A thrombogenic role has been attributed to activation of platelets or plasma coagulation on foreign surfaces, incomplete "wash out" of blood from valves, and increased shear rates with secondary platelet activation.

Therefore, we have taken a stepwise approach to analyze the thrombogenic mechanisms in 26 patients. To potentially identify subclinical events and therefore possibly a subgroup of patients particularly at risk for strokes, we have evaluated all patients and a control group for cerebrovascular microembolic signals (HITS) with transcranial Doppler monitoring. We have tried to analyze the procoagulant activity by simultaneously quantifying the circulating platelet membrane microparticles as well as their functional procoagulant impact on the RVVT, a clotting test particularly sensitive to the presence of phospholipids.

Our clinically observed CVE rate of 6.6 per 100 patient-years and the rate of definitive stroke of 2.9 per 100 patient-years is in the range found in the literature.⁴ If other transient and less specific symptoms are included as well, the rate is in the upper range (11.0 per 100 patient-years). Possible reasons include a relatively high number of patients with mitral valve replacements (n=4) and double valve replacements (aortic and mitral, n=8) known to be at higher risk for thromboembolic complications and a scrutinized neurological evaluation (MS), which may have detected discrete but relevant symptoms other studies might have missed.

The patients with CVE had significantly higher HITS counts during Doppler monitoring compared with the asymptomatic patients. It appears possible that these patients constantly embolize into the cerebral circulation platelet aggregates generated at the site of the PHV by the previously mentioned mechanisms. Other compositions of the embolic particles appear less likely, considering the results of the coagulation studies, which showed no thrombin or fibrin generation. Mixed thrombotic materials at this high embolization rate would be expected to result in positive tests for fibrin formation and/or fibrinolysis (fibrinopeptide A, D-dimer). Recent studies suggest that at least part of HITS represent air bubbles generated by the cavitation effect just behind the valve leaflets.²³ However, these air bubbles are of a size unable to cause strokes. Because air-fluid interfaces activate platelets, this might be another mechanism generating platelet aggregates.^{24 25 26 27 28} Microbubbles may also become stabilized, probably by being coated with membrane fragments (from platelets and other cells) generated at the PHV. Otherwise, it is difficult to understand how bubbles caused by the cavitation affect of PHV should persist until their arrival in the MCA.

Almost all of these HITS, despite occurring at high frequency, are asymptomatic. Thus, postulated aggregates likely disaggregate in the cerebral microcirculation,²⁹ embolize into clinically silent areas,³⁰ or are too small to generate ischemic injury. The stabilized bubbles may also disaggregate in the periphery, as may be the case with HITS composed of particulate materials. All control subjects had HITS counts of 0, whereas even the asymptomatic PHV patients had a mean count of approximately 20 HITS. It remains to be determined whether their increased rate identifies these patients to be at risk for future strokes. Additional important determinants may include size of the emboli (which cannot be estimated by Doppler analysis), their detailed composition, and most importantly, the ability to dissolve the emboli rapidly.

PMP counts were significantly elevated in the subgroup of patients with CVE but not in asymptomatic patients. In sharp contrast to the HITS quantified by Doppler ultrasound, the PMP can be observed in the healthy control population as well; namely, at a level of approximately 2%, which is similar to that in the asymptomatic patients. The low PMP level in controls may represent background noise or an insignificant systemic level, or it may result from sample collection. This observation may reflect a threshold level for thromboembolic complications; however, the almost invariably observed shortened RVVT suggests an increased prothrombinase activity in the plasma of all (symptomatic and asymptomatic) PHV patients. Therefore, very small membrane fragments may have been missed by the higher threshold of flow cytometry. One might hypothesize that the larger microparticles are more thrombogenic. In addition, other qualitative differences of membrane microparticles may also contribute to their procoagulant role in these patients.^{31 32 33} Preliminary results of our group suggest that the orientation of the platelet receptor, and perhaps of the phospholipid bilayer of microparticles as well, is different depending on the mechanism of their formation.³⁴ Therefore, "inside-out" vesicles may provide negatively charged phospholipids to a much greater extent than those that retain the nonactivated orientation. Moreover, in contrast to the platelets, the loss of the aminophospholipid-translocase activity in the PMP may lead to permanent loss of the lipid asymmetry.³⁵ The fact that we found no significant levels of P-selectin- (or CD62P-) positive PMP or platelets might suggest a predominantly mechanical generation rather than the activation-induced vesiculation. Increased levels of PMP have also been reported in patients with a vascular pathology,^{36 37 38} favoring turbulent flow and thus high shear rates as the mechanism for their generation. The levels of ß-thromboglobulin were not increased in our patients and do not support significant platelet release in vivo. On the other hand, reduced levels of PMP (ie, the inability to vesiculate) are associated with a bleeding disorder, Scott syndrome.^{33 39} The latter biological model provides the evidence that a physiological level of PMP appears to be required for normal hemostasis. This interpretation fits well with our findings that the control subjects had no HITS at all but had a constant level of approximately 2% of PMP in their circulation.

It is well established that negatively charged phopholipids serve as a template for the tenase/prothrombinase complex of plasmatic coagulation.^{16 40 41} This can be demonstrated most sensitively by the RVVT, which is shortened by the presence of phospholipids.²² In our hands, as little as the equivalent of 3 frozen and thawed platelets per microliter was sufficient to accelerate the RVV clotting time detectably. RVVT were substantially shortened in patients compared with control subjects, even though the activity of prothrombin and Factors VII, IX, and X was reduced to less than one fifth the normal values because of coumadin therapy.

The fact that the relative shortening of the RVVT in patients compared with controls increases continuously when the platelet count is lowered from 100x10⁹/L to 5x10⁹/L indicates that it is indeed the PMP rather than activated platelets that are responsible for the effect. The argument is further strengthened by the observation that the difference disappears after high-speed centrifugation and ultrafiltration (HS-PPP) to eliminate PMP. The latter method interestingly leads to an even slightly longer RVVT in patients, which might be explained with their lower content of vitamin K– dependent coagulation factors (Figure 3).

Interestingly, we found no evidence that the CVE itself induced a persistent procoagulant state: there was no inverse correlation between the parameters measured and the time elapsed between the CVE and the individual analysis.

Taken together, our data provide a potential pathophysiological model for the thromboembolic events that occur in patients with PHV and may help to explain the incomplete efficiency of therapeutic oral anticoagulation: Elevated concentrations of PMP may be a risk factor and may decrease the threshold for thromboembolic complications; under careful oral anticoagulation only minimal, if any, ß-thromboglobulin release, thrombin and fibrin generation, or fibrinolysis can be observed. Symptomatic patients, however, may have particularly elevated procoagulant activity provided by their PMP or may have additional congenital or acquired risk factors, which then lead to macroscopic emboli formation and cerebrovascular events. A prospective study with a larger and well-defined sample of patients is therefore needed to further elucidate this hypothesis.

It will be interesting to further evaluate platelet inhibitors as adjunct to anticoagulant therapy, particularly those that inhibit activation-induced vesiculation.⁴² Even if they will not inhibit mechanically generated PMP, they may inhibit the shear-induced activation⁴³ and influence the further buildup of PMP or enhance their dissolution in the cerebral circulation, thus reducing their deleterious effects.

	Acknowledgments

 
This study was supported by the Swiss National Science Foundation (grant 31–40 822.94 to Dr Beer) and the Swiss Heart Foundation (grant to Dr Beer). We wish to thank Christa Beer, Daniela Spina, Anita Vogt, Karin Woodtli, and Cornelia Schmid for expert technical assistance; Steven Merlin for his support in the flow-cytometry analysis; and Dr Christoph Minder for his help with statistical analysis. We also thank the patients, their general practitioners, and their cardiologists who agreed to participate in this study.

Received January 22, 1998; revision received April 27, 1998; accepted April 27, 1998.

	References

Top Abstract Introduction Subjects and Methods Results Discussion References

 
1. Cannegieter SC, Rosendaal FR, Wintzen AR, van der Meer FJM, Vandenbroucke JP, Briët E. Optimal oral anticoagulant therapy in patients with mechanical heart valves. N Engl J Med. 1995;333:11–17.[Abstract/Free Full Text]

2. Cortelazzo S, Finazzi G, Viero P, Galli M, Remuzzi A, Parenzan L, Barbui T. Thrombotic and hemorrhagic complications in patients with mechanical heart valve prosthesis attending an anticoagulation clinic. Thromb Haemost. 1993;69:316–320.[Medline] [Order article via Infotrieve]

3. Criscitiello MG, Levine HJ. Thromboembolism and prosthetic heart valves. Hosp Pract. 1992;15:69–96.

4. Turpie AGG, Gent M, Laupacis A, Latour Y, Gunstensen J, Basile F, Klimek M, Hirsh J. A comparison of aspirin with placebo in patients treated with warfarin after heart-valve replacement. N Engl J Med. 1993;329:524–529.[Abstract/Free Full Text]

5. Hanle DD, Harrison EC, Yonagathan AP, Allen TD, Corcoran WH. In vitro flow dynamics of four prosthetic aortic valves: a comparative analysis. J Biomech. 1989;22:597–607.[Medline] [Order article via Infotrieve]

6. Stein B, Fuster V, Halperin JL, Chesebro JH. Antithrombotic therapy in cardiac disease: an emerging approach based on pathogenesis and risk. Circulation. 1989;80:1501–1513.[Free Full Text]

7. Markus HS, Harrison MJ. Microembolic signal detection using ultrasound. Stroke. 1995;26:1517–1519.[Free Full Text]

8. Grosset DG, Georgiadis D, Kelman AW, Lees KR. Quantification of ultrasound embolic signals in patients with cardiac and carotid disease. Stroke. 1993;24:1922–1924.[Abstract/Free Full Text]

9. Dagirmanjian A, Davis DA, Rothfus WE, Deeb ZL, Goldberg AL. Silent cerebral microemboli occurring during carotid angiography: frequency as determined with Doppler sonography. AJR Am J Roentgenol. 1993;161:1037–1040.[Abstract/Free Full Text]

10. Spencer MP, Thomas GI, Nicholls SC, Sauvage LR. Detection of middle cerebral artery emboli during carotid endarterectomy using transcranial Doppler ultrasonography. Stroke. 1990;21:415–423.[Abstract/Free Full Text]

11. Russell D, Madden KP, Clark WM, Sandset PM, Zivin JA. Detection of arterial emboli using Doppler ultrasound in rabbits. Stroke. 1991;22:253–258.[Abstract/Free Full Text]

12. Georgiadis D, Grosset DG, Kelman A, Faichney A, Lees KR. Prevalence and characteristics of intracranial microemboli signals in patients with different types of prosthetic heart valves. Stroke. 1994;25:587–592.[Abstract]

13. Georgiadis D, Kaps M, Siebler M, Hill M, Konig M, Berg J, Kahl M, Zunker P, Diehl B, Ringelstein EB. Variablitiy of Doppler microemboli signal counts in patients with prosthetic cardiac valves. Stroke. 1995;26:439–443.[Abstract/Free Full Text]

14. Grosset DG, Cowburn P, Georgiadis D, Dargie HJ, Faichney A, Lees KR. Ultrasound detection of cerebral emboli in patients with prosthetic heart valves. J Heart Valve Dis. 1994;3:128–132.[Medline] [Order article via Infotrieve]

15. Chow TW, Hellums JD, Moake JL, Kroll MH. Shear stress-induced von Willebrand factor binding to platelet glycoprotein Ib initiates calcium influx associated with aggregation. Blood. 1992;80:113–120.[Abstract/Free Full Text]

16. McGill M, Fugman DA, Vittorio N, Darrow C. Platelet membrane vesicles reduced micro-vascular bleeding times in thrombocytopenic rabbits. J Lab Clin Med. 1987;109:127–133.[Medline] [Order article via Infotrieve]

17. Mann KG. Membrane-bound enzyme complexes in blood coagulation. In: Progress of Hemostasis and Thrombosis. New York, NY: Grune & Stratton; 1984:1–23.

18. Gawaz M, Ott J, Reininger AJ, Heinzmann U, Neumann FJ. Agglutination of isolated platelet membranes. Arterioscler Thromb Vasc Biol. 1996;16:621–627.[Abstract/Free Full Text]

19. Sturzenegger M, Beer JH, Rihs, F. Monitoring combined antithrombotic treatments in patients with prosthetic heart valves using transcranial Doppler and coagulation markers. Stroke. 1995;26:63–69.[Abstract/Free Full Text]

20. Genewein U, Haeberli A, Straub PW, Beer JH. Rebound after cessation of oral anticoagulant therapy: the biochemical evidence. Br J Haematol. 1996;92:479–485.[Medline] [Order article via Infotrieve]

21. Nomura S, Nagata H, Suzuki M, Kondo K, Ohga S, Kawakatsu T, Kido H, Fukuori T, Yamaguchi K, Iwata K. Microparticle generation during in vitro platelet activation by anti-CD9 murine monoclonal antibodies. Thromb Res. 1991;62:429–439.[Medline] [Order article via Infotrieve]

22. Warkentin TE, Hayward CPM, Boshkov LK, Santos AV, Sheppard JA, Bode AP, Kelton JG. Sera from patients with heparin-induced thrombocytopenia generate platelet-derived microparticles with procoagulant activity: an explanation for the thrombotic complications of heparin-induced thrombocytopenia. Blood. 1994;84:3691–3699.[Abstract/Free Full Text]

23. Kaps M, Hansen J, Weiher M, Tiffert K, Kayser J, Droste DW. Clinically silent microemboli in patients with artificial prosthetic heart valves are predominantly gaseous and not solid. Stroke. 1997;28:322–325.[Abstract/Free Full Text]

24. Malmgren R, Thorsen T, Nordvik A, Holmsen H. Microbubble-induced phospholipase C activation does not correlate with platelet aggregation. Thromb Haemost. 1993;69:394–396.[Medline] [Order article via Infotrieve]

25. Softeland E, Framstad T, Nordvik A, Strand I, Thorsen T, Holmsen H. Nitrogen microbubbles induce a disappearance of single platelets (aggregation) with porcine platelets: a comparative study of the effects of anticoagulants and blood collection methods. Thromb Res. 1994;76:61–70.[Medline] [Order article via Infotrieve]

26. Thorsen T, Brubakk A, Ovstedal T, Farstad M, Holmsen H. A method for production of N₂-microbubbles in platelet rich plasma in an aggregometer-like apparatus, and effect on the platelet density in vitro. Undersea Biomed Res. 1986;13:271–288.[Medline] [Order article via Infotrieve]

27. Thorsen T, Dalen H, Bjerkvig R, Holmsen H. Transmission and scanning electron microscopy of N₂-microbubble-activated human platelets in vitro. Undersea Biomed Res. 1987;14:45–59.[Medline] [Order article via Infotrieve]

28. Malmgren R, Thorsen T, Lie RT, Holmsen H. Microbubble-induced serotonin secretion in human platelets. Thromb Haemost. 1991;65:399–402.[Medline] [Order article via Infotrieve]

29. Hennerici MG. High intensity transcranial signals (HITS): a questionable "jackpot" for the prediction of stroke risk. J Heart Valve Dis. 1994;3:124–125. Editorial.[Medline] [Order article via Infotrieve]

30. Brown MM, Markus HS. Transcranial Doppler detection of asymptomatic cerebral microemboli. J Heart Valve Dis. 1994;3:126–127.[Medline] [Order article via Infotrieve]

31. Lee LH, Baglin T. Altered platelet phospholipid-dependent thrombin generation in thrombocytopenia and thrombocytosis. Br J Haematol. 1995;89:131–136.[Medline] [Order article via Infotrieve]

32. Tans G, Rosing J, Thomassen MC, Heeb MJ, Zwaal RFA, Griffin JH. Comparison of anticoagulant and procoagulant activities of stimulated platelets and platelet-derived microparticles. Blood. 1991;77:2641–2648.[Abstract/Free Full Text]

33. Sims PJ, Wiedmer T, Esmon CT, Weiss HJ, Shattil SJ. Assembly of the platelet prothrombinase complex is linked to vesiculation of the platelet plasma membrane: studies in Scott syndrome: an isolated defect in platelet procoagulant activity. J Biol Chem. 1989;264:17049–17057.[Abstract/Free Full Text]

34. Beer JH, Coller BS, Siegenthaler C. The membrane sidedness of platelet microparticles, as judged by an antibody to the C-terminus of GPIIIa, depends on the mechanism of their production. Thromb Haemost. 1991;65:1068a.

35. Comfurius P, Senden JMG, Tilly RHJ, Schroit AJ, Bevers EM, Zwaal RFA. Loss of membrane phospholipid asymmetry in platelets and red cells may be associated with calcium-induced shedding of plasma membrane and inhibition of aminophospholipid translocase. Biochim Biophys Acta. 1990;1026:153–160.[Medline] [Order article via Infotrieve]

36. Gawaz M, Neumann FJ, Ott I, Schiessler A, Schömig A. Platelet function in acute myocardial infarction treated with direct angioplasty. Circulation. 1996;93:229–237.[Abstract/Free Full Text]

37. Lee YJ, Jy W, Horstman LL, Janania J, Reyes Y, Kelley RE, Ahn YS. Elevated platelet microparticles in transient ischemic attacks, lacunar infarcts, and multiinfarct dementias. Thromb Res. 1993;72:295–304.[Medline] [Order article via Infotrieve]

38. Kelton JG, Warkentin TE, Hayward PM, Murphy WG, Moore JC. Calpain activity with thrombotic thrombocytopenia purpura is associated with platelet microparticles. Blood. 1992;80:2246–2251.[Abstract/Free Full Text]

39. Kojima H, Newton-Nash D, Weiss HJ, Zhao J, Sims PJ, Wiedmer T. Production and characterization of transformed B-lymphocytes expressing the membrane defect of Scott syndrome. J Clin Invest. 1994;94:2237–2244.

40. Hemker H, Beguin S. How to measure the relevant effect of anticoagulant treatment on the coagulability of plasma. In: Thrombosis and Its Management. Edinburgh, UK: Churchill Livingstone; 1993:3–14.

41. Bevers E, Comfurius P, Zwaal R. Platelet procoagulant activity: physiological significance and mechanisms of exposure. Blood Rev. 1991;5:146–154.[Medline] [Order article via Infotrieve]

42. Gemmell CH, Sefton MV, Yeo EL. Platelet-derived microparticle formation involves glycoprotein IIb-IIIa. J Biol Chem. 1993;268:14586–14589.[Abstract/Free Full Text]

43. Konstantopoulos K, Kamat SG, Schafer AI, Banez EL, Jordan R, Kleiman NS, Hellums JD. Shear-induced platelet aggregation is inhibited by in vivo infusion of an anti-glycoprotein IIb/IIIa antibody fragment, c7E3 Fab, in patients undergoing coronary angioplasty. Circulation. 1995;91:1427–1431.[Abstract/Free Full Text]

This article has been cited by other articles:

				C. Adiguzel, O. Iqbal, D. Hoppensteadt, W. Jeske, J. Cunanan, E. Litinas, He Zhu, J. M. Walenga, and J. Fareed Comparative Anticoagulant and Platelet Modulatory Effects of Enoxaparin and Sulodexide Clinical and Applied Thrombosis/Hemostasis, October 1, 2009; 15(5): 501 - 511. [Abstract] [PDF]

				R. A. Rodriguez, H. J. Nathan, M. Ruel, F. Rubens, D. Dafoe, and T. Mesana A method to distinguish between gaseous and solid cerebral emboli in patients with prosthetic heart valves Eur. J. Cardiothorac. Surg., January 1, 2009; 35(1): 89 - 95. [Abstract] [Full Text] [PDF]

				L. Hilker, M. Wodny, M. Ginesta, H.-G. Wollert, and L. Eckel Differences in the recovery of platelet counts after biological aortic valve replacement Interactive CardioVascular and Thoracic Surgery, January 1, 2009; 8(1): 70 - 73. [Abstract] [Full Text] [PDF]

				M. Skjelland, A. Michelsen, F. Brosstad, J. L. Svennevig, R. Brucher, and D. Russell Solid Cerebral Microemboli and Cerebrovascular Symptoms in Patients With Prosthetic Heart Valves Stroke, April 1, 2008; 39(4): 1159 - 1164. [Abstract] [Full Text] [PDF]

				L. Guerrieri Wolf, B. P. Choudhary, Y. Abu-Omar, and D. P. Taggart Solid and gaseous cerebral microembolization after biologic and mechanical aortic valve replacement: Investigation with multirange and multifrequency transcranial Doppler ultrasound J. Thorac. Cardiovasc. Surg., March 1, 2008; 135(3): 512 - 520. [Abstract] [Full Text] [PDF]

				R. Dittrich and E. B. Ringelstein Occurrence and Clinical Impact of Microembolic Signals During or After Cardiosurgical Procedures Stroke, February 1, 2008; 39(2): 503 - 511. [Abstract] [Full Text] [PDF]

				F. di Marco, S. Grendene, G. Feltrin, D. Meneghetti, and G. Gerosa Antiplatelet therapy in patients receiving aortic bioprostheses: A report of clinical and instrumental safety J. Thorac. Cardiovasc. Surg., June 1, 2007; 133(6): 1597 - 1603. [Abstract] [Full Text] [PDF]

				F. di Marco, M. Giordan, and G. Gerosa Early antithrombotic therapy after aortic valve replacement with tissue valves: When the practice diverges from the guidelines J. Thorac. Cardiovasc. Surg., June 1, 2006; 131(6): 1223 - 1223. [Full Text] [PDF]

				A. LeGuyader, R. Watanabe, J. Berbe, A. Boumediene, M. Cogne, and M. Laskar Platelet activation after aortic prosthetic valve surgery Interactive CardioVascular and Thoracic Surgery, February 1, 2006; 5(1): 60 - 64. [Abstract] [Full Text] [PDF]

				S. Milo, E. Rambod, C. Gutfinger, and M. Gharib Mitral mechanical heart valves: in vitro studies of their closure, vortex and microbubble formation with possible medical implications Eur. J. Cardiothorac. Surg., September 1, 2003; 24(3): 364 - 370. [Abstract] [Full Text] [PDF]

				R. A. Preston, W. Jy, J. J. Jimenez, L. M. Mauro, L. L. Horstman, M. Valle, G. Aime, and Y. S. Ahn Effects of Severe Hypertension on Endothelial and Platelet Microparticles Hypertension, February 1, 2003; 41(2): 211 - 217. [Abstract] [Full Text] [PDF]

				J. Laas, S. Kseibi, M. Perthel, A. Klingbeil, L'E. El-Ayoubi, and A. Alken Impact of high intensity transient signals on the choice of mechanical aortic valve substitutes Eur. J. Cardiothorac. Surg., January 1, 2003; 23(1): 93 - 96. [Abstract] [Full Text] [PDF]

				D. Russell and R. Brucher Online Automatic Discrimination Between Solid and Gaseous Cerebral Microemboli With the First Multifrequency Transcranial Doppler Stroke, August 1, 2002; 33(8): 1975 - 1980. [Abstract] [Full Text] [PDF]

				K. Moshfegh, M. Redondo, F. Julmy, W. A. Wuillemin, M. U. Gebauer, A. Haeberli, and B. J. Meyer Antiplatelet effects of clopidogrel compared with aspirin after myocardial infarction: enhanced inhibitory effects of combination therapy J. Am. Coll. Cardiol., September 1, 2000; 36(3): 699 - 705. [Abstract] [Full Text] [PDF]

				L. Alberio, O. Safa, K. J. Clemetson, C. T. Esmon, and G. L. Dale Surface expression and functional characterization of alpha -granule factor V in human platelets: effects of ionophore A23187, thrombin, collagen, and convulxin Blood, March 1, 2000; 95(5): 1694 - 1702. [Abstract] [Full Text] [PDF]

				Z. G. Nadareishvili, Z. Choudary, C. Joyner, D. Brodie, and J. W. Norris Cerebral Microembolism in Acute Myocardial Infarction Stroke, December 1, 1999; 30(12): 2679 - 2682. [Abstract] [Full Text] [PDF]

				R. Lagos, F. M. Cabezas, and G. Deklunder Microemboli in Cerebral Circulation and Alteration of Cognitive Abilities in Patients With Mechanical Prosthetic Heart Valves • Response Stroke, May 1, 1999; 30(5): 1150 - 1150. [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Geiser, T. Articles by Beer, J. H. Search for Related Content PubMed PubMed Citation Articles by Geiser, T. Articles by Beer, J. H. Pubmed/NCBI databases Compound via MeSH Substance via MeSH