Effects of Myocardial Contractility on Microemboli Production by Mechanical Heart Valves in a Bovine Model

Abstract

Microemboli caused by mechanical heart valves have the potential to cause cerebrovascular events. We investigated the effects of myocardial contractility and heart rate on microemboli production in association with conventional and experimental mechanical heart valves implanted in the mitral position in a bovine model.

In 10 calves, the mitral valves were replaced with mechanical valves. Doppler recordings were analyzed for high-intensity transient signals, which are ultrasound reflections from circulating microemboli. The animals were studied at rest, during pacing at 160 bpm, after dobutamine infusion, and after esmolol infusion. The incidence of high intensity transient signals was expressed as signal frequency (signals per hour) and as signal rate (signals per 100 heart cycles). With a 68% increase in the heart rate, signal frequency increased by 135%, but signal rate increased by only 41%. With a 144% increase in myocardial contractility, signal rate increased by 264%. With a 31% decrease in contractility, signal rate decreased by 62%.

We conclude that microemboli production by mechanical heart valves varies with myocardial contractility and heart rate. The fact that contractility affects the incidence of high-intensity transient signals suggests that the microemboli are gaseous in nature, that their production is pressure driven, and that cavitation is a possible cause. It is likely that mechanical heart valve design is responsible for the quantity of microemboli production.

Microemboli caused by mechanical heart valves (MHVs) may be a cerebrovascular risk factor. Minimization or elimination of such microemboli seems desirable for the improvement of current heart valve technology. Doppler ultrasonography is widely used to assess blood velocity and can also detect microemboli in the circulation, ¹ especially the cerebral circulation. ² When an embolus passes through the ultrasonic beam, the reflected Doppler signal is of high-intensity and is transient in nature, thus the acronym ‘HITS’. These unique signals (signatures of microemboli) are observed under various conditions. Gaseous emboli can be detected during extracorporeal circulation in multiple vascular structures or during the decompression phase after deep-sea diving (the latter occurrence was indeed the first context in which HITS were described). ^3–5 In either situation, HITS are most likely to be ultrasound reflections from microbubbles formed under nonphysiologic conditions. In addition, HITS may be detected in the circulation when solid emboli are released from carotid plaques or during atrial fibrillation.

The nature of MHV-related microemboli is still debated. ^6–8 According to Levy and associates, ⁹ the process starts when mitral blood flow is reversed during valve closure; this regurgitant blood is ejected into the atrial cavity at high speed, where it forms vortices capable of causing cavitation. The speed of the regurgitant blood appears to depend on the ventricular pressure gradient and, therefore, on the inotropic state of the heart. Whatever their mechanism, MHV-related microemboli may have an important clinical impact. Many reports highlight the role of circulating microemboli as a risk factor for cerebrovascular events, including stroke. ¹⁰ Some authors have reported a permanent decline in cognitive function in patients with mechanical heart valve prostheses. ¹¹ Because of the large number of MHV implants worldwide and the potential deleterious effects of long-term microembolization on brain function, further knowledge is needed about the production of microemboli by mechanical valves. We investigated the effects of myocardial contractility on MHV-related microemboli production in calves.

Materials and Methods

The calf model used for this study has been previously described. ¹² Ten calves, aged 4 to 10 months and weighing 80 to 100 kg, received either a commercially available clinical quality 29-mm bileaflet mitral MHV (n = 3) or an experimental trileaflet 29-mm mitral MHV (n = 7) via a left thoracotomy after initiation of CPB. An 18-mm, 4-MHz custom Doppler perivascular probe was secured around the brachiocephalic artery. Permanent epicardial pacing leads were attached to the right and left atria. A radiofrequency transmitter (model D70; Data Sciences International [DSI], Minneapolis, Minn), designed for measurement of 2 pressures and ECG, was implanted with 1 catheter placed in the internal thoracic artery and the other in the left ventricle (LV) via the LV apex. Vascular access ports were implanted in the neck via the jugular vein for permanent venous access for blood sampling and inotrope delivery. The heart was restarted, the calf weaned from bypass, and the chest closed (described previously in detail). ¹²

The calves were studied 4 to 6, 8 to 10, and 12 to 14 weeks postoperatively. Doppler recordings were made with a computer data-acquisition device (Multi-Dop X4; DWL Electronische Systeme GmbH, Sipplingen, Germany) and with specialized software (TCD8, Version 8.00M; DWL Electronische Systeme GmbH, Sipplingen, Germany). The Doppler spectra were stored and later analyzed offline by 2 experienced, independent observers who were blinded to the experimental conditions. High-intensity transient signals were defined as signals that had an initial unidirectional spectral deviation, had a signal power of ≥9 dB relative to the background power, and lacked a cyclic appearance. The LV dP/dt and the heart rate (HR) were recorded from telemetrically transmitted data from the DSI transmitters. Production of HITS was studied during 3 interventions—during rapid pacing at 160 bpm, after dobutamine infusion, and after esmolol infusion—and was compared to HITS production under resting (baseline) conditions. Rapid pacing was accomplished by an external pacemaker (Stat-Pace; Seecor Inc., Fort Worth, Tex), which was connected to the permanently externalized atrial pacing leads. The dobutamine infusion was titrated (20 to 40 mg/hr) to obtain a twofold increase in the dP/dt relative to the baseline level. The esmolol infusion was titrated (2,000 to 4,000 mg/hr) to lower the dP/dt to half of the baseline level. Under each experimental condition, Doppler recording was performed for 15 minutes. Each animal was tested 3 times at 2-week intervals. The HITS rate was assessed as HITS per hour and as HITS per 100 heart cycles.

Statistical analysis was performed with SPSS, Version 8 for Windows (SPSS Inc., Chicago, Ill). A two-tailed Student's t-test was performed for unpaired samples of equal variance for a rejection region of α < 0.05. Values are expressed as mean ± standard deviation.

Results

In the 10 calves that underwent MHV implantation, 27 successful recordings were performed (Table I). Myocardial contractility (dP/dt) did not significantly change during rapid pacing, but increased by 144% during dobutamine administration (p<0.05), and decreased by 31% during esmolol administration (p<0.05) (Fig. 1).

Table I. HITS Rate, Heart Rate, and dP/dt (Mean ± Standard Deviation) under the 4 Experimental Conditions

Fig. 1 Mean index of left ventricular contractility (dP/dt) under 4 inotropic conditions, expressed in mmHg per second.

Rapid pacing resulted in a 135% increase in the HITS frequency (Fig. 2). When HITS frequency was normalized per 100 cycles, pacing the heart still had a significant but less pronounced effect on the HITS production (41% more than at baseline, p<0.001) (Fig. 3).

Fig. 2 Mean HITS (high-intensity transient signal) count in events per hour under 4 inotropic conditions.

Fig. 3 Mean HITS (high-intensity transient signal) count in events per 100 cardiac cycles under 4 inotropic conditions.

Dobutamine administration yielded a significant (264%) increase in the HITS rate (p<.0005), and esmolol administration resulted in a significant (62%) decrease in HITS rate (p<.0005).

Discussion

The bovine model used in this study has been described previously and has proved suitable for analyzing the properties and functioning of MHVs intended for human use. ¹² Doppler techniques for the detection of circulating microemboli are widely used clinically to detect microemboli in the cerebral circulation, where only a portion of all microemboli are present. In the bovine model, the Doppler probe is placed around the brachiocephalic artery, which enables detection of most microemboli produced by the MHV that enter the cerebral circulation.

In this study, we used standard HITS detection criteria, which are easily applied when the Doppler signal is set at low gain and the ultrasonic beam is correctly aligned. In some situations—especially when the global energy of the reflected signal is high, as during states of hypercontractility (elevated dP/dt)—HITS detection is less accurate, which can lead to discrepancies in observer counts and increased standard deviations.

If, as hypothesized, microemboli are produced by the activity of the MHV, one would expect to find a relationship between the heart rate, which reflects valve activity, and the HITS rate. Our results support this hypothesis, as the HITS rate and heart rate were strongly correlated. This explains why the heart rate accounts for 59% of the variance in the HITS rate. During pacing, the small but significant change in the HITS rate relative to baseline, in the presence of a stable dP/dt under both conditions, shows that contractility is not the only contributing factor. Other characteristics, not accounted for by the change in heart rate during pacing, also affect microemboli production.

Our experiments showed that a change in the contractile state of the heart, as demonstrated by the change in the dP/dt during esmolol and dobutamine infusions, affects the HITS rate and, in this manner, microemboli production.

In vitro studies using high-speed video photography have shown microbubble formation on the atrial side of mitral MHVs during closure. ¹³ This bubble formation depends on the valve structure and the pressure dynamics, which can induce cavitation and rapid growth of the microbubbles in the bloodstream. ^14,15 The microembolic process starts during valve closure, when high-speed regurgitant blood is ejected into the atrial cavity, producing vortices that may cause cavitation. ⁹ The speed of the regurgitant blood is theorized to depend on the ventricular pressure gradient and, therefore, on the heart's inotropic state. Our data support this hypothesis by showing that the HITS rate increases with an elevated dP/dt and decreases with a lowered dP/dt.

These data give rise to a further question: what factors, aside from heart rate and contractility, affect microemboli production? It is possible that the additional variance is due to a measurement error. It is also possible that the reduced stroke volume at an elevated heart rate could cause a perivalvular hemodynamic change that results in additional microemboli production. At the elevated heart rate, the energy of valve closure may be greater because of an increased afterload. This increased energy would cause larger pressure differentials and, therefore, additional microemboli production. Experiments designed to assess the HITS rate over a range of heart rates would provide further insight into this theory.

Cavitation-produced microbubbles are composed of water vapor and blood gases (i.e., nitrogen, oxygen, and carbon dioxide). The vapor is unstable and rapidly returns to liquid as soon as the pressure is again normal. Because the blood gases are less soluble and can survive much longer, bubbles can be found far from their site of origin in the circulation. The results of clinical studies support the hypothesis of gaseous composition for MHV-related microemboli by demonstrating: 1) the lack of effect of the degree of anticoagulant, antithrombotic, or antiplatelet therapy on the HITS rate, ^16,17 2) the changes in the HITS count with modification of the inspired gas concentration in oxygen, ^18,19 and 3) the presence of abnormal, sparkling, intracavitary echoes upstream during MHV closure. ^20,21 These echoes are similar to the bright ones generated in cardiac chambers by the injection of ultrasonographic contrast agents that are, themselves, microbubbles stabilized with a special coating.

In summary, we found that MHV-related microemboli production is influenced by the heart rate and myocardial contractility. The fact that contractility is the main factor to affect the HITS rate strengthens the hypothesis that microemboli are gaseous and are caused by a cavitation effect. Therefore, further investigation of MHV design is warranted, as these gaseous microemboli are unaffected by anticoagulant therapy and may be a likely source of increased risk for an adverse cerebrovascular event.