Validation of admittance computed left ventricular volumes against real-time three-dimensional echocardiography in the porcine heart
New Findings
- •
What is the central question of this study?
How do left ventricular volumes measured in a large animal using admittance catheterization compare with those measured using three-dimensional echocardiography or traditional conductance?
- •
What is the main finding and its importance?
Admittance computed left ventricular volumes were validated in baseline conditions and during inotropic stimulation with dobutamine. These results serve as a foundation for the use of admittance in large-animal experimental models of cardiovascular disease.
The admittance and Wei's equation is a new technique for ventricular volumetry to determine pressure–volume relations that addresses traditional conductance-related issues of parallel conductance and field correction factor. These issues with conductance have prevented researchers from obtaining real-time absolute ventricular volumes. Moreover, the time-consuming steps involved in processing conductance catheter data warrant the need for a better catheter-based technique for ventricular volumetry. We aimed to compare the accuracy of left ventricular (LV) volumetry between the new admittance catheterization technique and transoesophageal real-time three-dimensional echocardiography (RT3DE) in a large-animal model. Eight anaesthetized pigs were used. A 7 French admittance catheter was positioned in the LV via the right carotid artery. The catheter was connected to an admittance control unit (ADVantage; Transonic Scisense Inc.), and data were recorded on a four-channel acquisition system (FA404; iWorx Systems). Admittance catheterization data and transoesophageal RT3DE (X7-2; Philips) data were simultaneously obtained with the animal ventilated, under neuromuscular blockade and monitored in baseline conditions and during dobutamine infusion. Left ventricular volumes measured from admittance catheterization (Labscribe; iWorx Systems) and RT3DE (Qlab; Philips) were compared. In a subset of four animals, admittance volumes were compared with those obtained from traditional conductance catheterization (MPVS Ultra; Millar Instruments). Of 37 sets of measurements compared, admittance- and RT3DE-derived LV volumes and ejection fractions at baseline and in the presence of dobutamine exhibited general agreement, with mean percentage intermethod differences of 10% for end-diastolic volumes, 14% for end-systolic volumes and 9% for ejection fraction; the respective intermethod differences between admittance and conductance in eight data sets compared were 11, 11 and 12%. Admittance volumes were generally higher than those obtained by RT3DE, especially among the larger ventricles. It is concluded that it is feasible to derive pressure–volume relations using admittance catheterization in large animals. This study demonstrated agreements between admittance and RT3DE to within 10–14% mean intermethod difference in the estimation of LV volumes. Further investigation will be required to examine the accuracy of volumes in largest ventricles, where intermethod divergence is greatest.
Introduction
Analysis of pressure–volume relations is considered the best method available to derive relatively load-independent parameters of cardiac function (Burkhoff et al. 2005). Conductance catheterization (CC) is commonly employed for determination of pressure–volume relations in the preclinical setting (Jegger et al. 2006; Pacher et al. 2008). This is enabled by beat-to-beat measurements of blood conductance using a tetrapolar catheter placed in the ventricle. This measured conductance is then converted to volume using a linear equation that relates conductance to volume. This is a source of inaccuracy, because the conductance–volume relation is, in fact, non-linear, and the degree of non-linearity increases with volume. Alternatively, conductance measurements may be made in blood-filled cylindrical cuvette wells of known volumes to generate a conductance–volume calibration curve for the catheter. This curve is then used to convert the measured ventricular conductance to absolute volume. This incorrectly assumes the shape of the ventricle to be a cylinder and that the catheter is perfectly aligned along the centreline in the ventricle. Regardless of how the measured conductance is converted to volume, this volume is a combination of blood and surrounding myocardial volume. A bolus of hypertonic saline must be injected into the ventricle at the end of the experiment to calculate and remove the parallel myocardial volume effectively (Pacher et al. 2008).
Admittance is a new technique that has been proposed to address some of these limitations of the conductance technique. The admittance technique uses a combination pressure–volume catheter that is placed in the ventricle to measure admittance magnitude and ‘phase angle’ signals from the blood pool and surrounding myocardium. The phase angle or delay in the admittance signal is caused by the capacitive property of the myocardium. Given that blood has no capacitive properties at the measurement frequency of 20 kHz, it can be concluded that any measured phase angle is directly proportional to the proximity of the myocardium to the catheter. This enables the real-time removal of myocardial parallel conductance to determine instantaneous blood conductance (Kottam et al. 2006; Porterfield et al. 2009). The blood conductance is then converted to true volume using a non-linear conductance–volume equation developed by Wei and co-workers (Wei et al. 2004; Kottam et al. 2006). This measurement of absolute blood volumes in real time is the biggest advantage of admittance over traditional conductance. Moreover, admittance catheterization (AC) is technically easier to perform than CC, because the former eliminates volume cuvette calibration and parallel conductance determination using hypertonic saline injection (Kottam et al. 2006; Porterfield et al. 2009).
It has been demonstrated in murine models that AC (dynamic parallel conductance Gp) and Wei's equation (dynamic field correction factor) provided more accurate ventricular volume measurements compared with traditional CC (Porterfield et al. 2009). Moreover, AC was found to be more sensitive to inotropic changes in the murine heart (Porterfield et al. 2009). Admittance catheterization has also been used to study right ventricular pressure–volume relations in mice (Tabima et al. 2010). The purpose of the present study was twofold: (i) to examine the agreement of AC volumes with transoesophageal real-time three-dimensional echocardiography (RT3DE)-derived volumes for the left ventricle (LV) in a porcine model; and (ii) to determine the feasibility of deriving LV pressure–volume relations using AC in this model.
Methods
Ethical approval
The study was approved by the Institutional Animal Care and Use Committee of the University of Nebraska Medical Center and was in compliance with the standards set out in the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH publication no. 85-23, revised 2011).
Animal preparation
Eight pigs were used in the study. Prior to each experiment, the animals were fasted overnight and anaesthesia was induced with an intramuscular mixture of Telazol (4.4 mg kg−1; tiletamine HCl and zolazepam HCl, Fort Dodge Animal Health, Fort Dodge, IA, USA), ketamine (2.2 mg kg−1; Bioniche Teoranta, Inverin, Co. Galway, Ireland), and xylazine (2.2 mg kg−1; Lloyd Laboratories, Shenandoah, IA, USA). Intramuscular atropine (0.05 mg kg−1) was used to dry oral and tracheal secretions and to prevent bradycardia during the study. Following placement of a venous line in an ear vein, the animal was intubated, and isoflurane (induction at 4%, maintenance at 1.0–1.8%) was administered in O2 throughout the procedure. The animal was ventilated at a tidal volume of 10 ml (kg body weight)−1 at a rate of 15 breaths min−1. Introducer sheaths were placed in the carotid artery (via cutdown) and bilateral femoral vein and artery (via percutaneous access) for placement of admittance and conductance catheters, balloon catheters, pressure monitoring and drug administration. Continuous monitoring of heart rate, arterial blood pressure and systemic oxygen saturations by pulse oximetry was performed during the study. After completion of the measurements described in the following subsections, the animals were killed by administration of potassium chloride (120 mequiv, i.v.) while still under general anaesthesia, in accordance with American Veterinary Medical Association guidelines.
Admittance catheterization
Admittance catheterization was performed using a 7 French high-fidelity catheter (Transonic Scisense Inc., London, ON, Canada) positioned in the apex of the LV via the right carotid artery sheath under X-ray fluoroscopy. The data were measured and analysed using an ADVantage pressure–volume system (Transonic Scisense Inc.) connected to an FA404 data acquisition system (iWorx Systems, Dover, NH, USA). The admittance catheters did not require any saline calibrations prior to insertion in the animal, according to instructions provided by Transonic Scisense Inc.
Transoesophageal RT3DE
Transoesophageal RT3DE was performed using a commercially available transoesophageal ultrasound transducer (X7-2 X matrix array with PureWave technology, iE33; Philips, Andover, MA, USA) advanced into the oesophagus of the anaesthetized animal. The X7-2 transducer utilizes a cascaded beamformer approach with a 2500 element microbeamformer arranged in 100 groups that sums signals within each group. The transducer was positioned at the level of the midoesophagus and gently rotated as necessary until a four-chamber view of the heart was obtained for image acquisition. In this mode, four wedge-shaped subvolumes were acquired over four consecutive cardiac cycles with electrocardiogram gating during the same breath-hold. Before each acquisition, images were optimized for endocardial visualization by modifying the gain and compression controls. The full-volume acquisitions captured the entire LV to allow dynamic assessment, with careful attention to fitting the LV walls within the scanning sector of the transducer.
Data acquisition
Pressures were obtained from AC, and simultaneously volumes were obtained from AC and RT3DE. A single-lead electrocardiogram (ML132 Bio Amp; AD Instruments, Colorado Springs, CO, USA) was integrated with the RT3DE and ADVantage systems. Volume measurements were made during two separate conditions, namely at rest, and during intravenous infusion of dobutamine at 3 μg kg−1 min−1. Three separate acquisitions (AC and RT3DE) were performed in each condition. Admittance catheterization-derived pressure–volume relations with load alteration were also obtained with inferior vena caval occlusion using a 36–40 mm sizing balloon (AGA Medical, Minneapolis, MN, USA) inflated with isotonic saline solution. All measurements (at baseline and in the presence of dobutamine) were performed at end expiration with breath-holding for 15–20 s and during muscle relaxation with 0.01 mg kg−1 vecuronium bromide administered intravenously.
In a randomly chosen subset of four animals, traditional CC was performed immediately preceding AC measurements at baseline. Conductance catheterization utilized a 7 French high-fidelity catheter (Millar Instruments, Houston, TX, USA) positioned in the LV apex. Specific conductivity of blood was measured with a 5 ml cuvette and parallel conductance determined by the hypertonic saline injection method. The data were measured and analysed using an MPVS Ultra pressure–volume analysis system (Millar Instruments) connected to a PowerLab data acquisition system (AD Instruments).
Data processing
End-systolic (ESV) and end-diastolic volumes (EDV) were measured at baseline and in the presence of dobutamine from AC data and the simultaneously acquired RT3DE image data. Admittance catheterization data postprocessing was performed with Labscribe software (iWorx Systems, Dover, NH, USA). The full volume RT3DE data sets were exported to dedicated analytical software (3DQ, Q-Lab version 8.1; Philips, Andover, MA, USA) for offline semi-automated volumetric quantification of the LV from multiplanar reconstruction views (Soliman et al. 2007). Any beat with distortion of the ventricular endocardial surface on RT3DE was excluded from volumetric analysis. The software utilizes a three-dimensional guided biplane method, whereby two perpendicular two-dimensional planes are placed in the data set to divide the LV accurately along its long axis at the true apex. Ideal midoesophageal four-chamber and two-chamber views of the LV were simultaneously displayed, and the volumes and ejection fraction (EF) were calculated.
An average of two AC beat measurements and an average of two RT3DE full volume measurements in each condition were tabulated for comparison of volumes. Conductance catheterization data post-processing was performed using Labchart 7 Pro software (AD Instruments).
Statistical analysis
Data are expressed as numbers and percentages. Agreements between AC and RT3DE volumes at baseline and in the presence of dobutamine were analysed using Student's t tests, Pearson correlations and the method of Bland–Altman. Agreements between CC and AC measurements from the subgroup of four animals were also analysed. The mean percentage errors of measurements were calculated as the absolute difference between the two sets of observations, divided by the mean of the observations: [(X1 – X2)/mean(X1, X2)]× 100. A P value <0.05 was considered statistically significant. Statistical analysis was performed using Minitab 16.1 (Minitab Inc., State College, PA, USA).
Results
A total of eight pigs (median weight 34.4 kg; range 31.0–37.2 kg) were studied. Good-quality AC data acquisition was feasible in all animals (Fig. 1). Any RT3DE volume data set with a premature ventricular beat, suboptimal LV coverage or obscured endocardial surface during the acquisition was excluded from postprocessing. Admittance catheterization volume measurement with any artifact (e.g. due to catheter contact with the myocardium) was also excluded from postprocessing.
Three sets of loops that indicate admittance-derived volumes at baseline (red), inferior vena caval occlusion at baseline (blue) and vena caval occlusion during dobutamine infusion (green) from one animal
The end-systolic pressure–volume relations during the occlusions are also indicated. Note the beat-to-beat change in left ventricular pressures and volumes during the occlusion.
The averages of two separate volume measurements by each technique were compared. End-diastolic volumes, end-systolic volumes and ejection fractions were tabulated for the AC and RT3DE data. Of the 37 data sets, 20 were obtained at baseline and 17 in the presence of dobutamine (Table 1). More RT3DE data sets had to be excluded from the acquisitions in the presence of dobutamine owing to premature ventricular contractions or stitch artifacts. Haemodynamic data including the AC-derived load-independent parameters, such as the slope of the end-systolic pressure–volume relation (ESPVR) and preload-recruitable stroke work (PRSW) are summarized in Table 2. These were obtained during preload alteration with inferior vena caval occlusion in all animals at baseline and in the presence of dobutamine.
| Conditions | Pig no. | Admittance catheterization* | RT3D TEE* | ||||
|---|---|---|---|---|---|---|---|
| EDV (ml) | ESV (ml) | EF (%) | EDV (ml) | ESV (ml) | EF (%) | ||
| Baseline | 1 | 79.0 | 46.0 | 41.8 | 76.4 | 41.8 | 45.2 |
| 82.0 | 52.0 | 36.6 | 80.3 | 50.3 | 37.4 | ||
| 2 | 78.0 | 42.0 | 46.2 | 75.5 | 42.5 | 43.7 | |
| 79.0 | 45.0 | 43.0 | 75.1 | 36.1 | 51.9 | ||
| 84.0 | 50.0 | 40.5 | 73.0 | 34.2 | 53.2 | ||
| 3 | 92.0 | 55.0 | 40.2 | 70.3 | 47.4 | 32.6 | |
| 90.0 | 56.0 | 41.7 | 75.4 | 42.9 | 43.1 | ||
| 92.0 | 61.0 | 40.2 | 70.6 | 43.0 | 39.2 | ||
| 4 | 83.0 | 45.0 | 45.8 | 75.4 | 42.0 | 41.9 | |
| 84.0 | 47.0 | 44.0 | 71.3 | 33.4 | 53.1 | ||
| 85.0 | 48.0 | 43.5 | 70.6 | 28.7 | 59.4 | ||
| 5 | 67.0 | 36.0 | 46.3 | 64.4 | 33.6 | 47.9 | |
| 6 | 68.0 | 39.0 | 42.6 | 49.5 | 30.6 | 38.2 | |
| 65.0 | 36.0 | 44.6 | 51.9 | 33.8 | 34.8 | ||
| 7 | 48.0 | 21.0 | 56.3 | 46.0 | 21.2 | 53.8 | |
| 51.0 | 23.0 | 54.9 | 40.5 | 21.7 | 46.5 | ||
| 54.0 | 24.0 | 55.6 | 40.0 | 21.6 | 46.0 | ||
| 8 | 57.0 | 23.0 | 59.6 | 54.9 | 23.5 | 57.1 | |
| 60.0 | 26.0 | 56.7 | 58.0 | 23.6 | 59.2 | ||
| 56.0 | 21.0 | 62.5 | 54.1 | 25.7 | 52.6 | ||
| Dobutamine | 1 | 59.0 | 30.0 | 49.2 | 56.4 | 30.1 | 46.7 |
| 61.0 | 30.0 | 50.8 | 59.0 | 28.8 | 51.1 | ||
| 2 | 53.0 | 21.0 | 60.4 | 45.8 | 19.4 | 57.7 | |
| 57.0 | 28.0 | 50.9 | 61.7 | 26.2 | 57.6 | ||
| 3 | 55.0 | 23.0 | 58.2 | 54.2 | 23.1 | 57.4 | |
| 4 | 50.0 | 20.0 | 60.0 | 49.8 | 20.6 | 58.6 | |
| 56.0 | 25.0 | 55.4 | 54.6 | 25.0 | 54.2 | ||
| 5 | 33.0 | 7.0 | 78.8 | 33.2 | 7.2 | 78.2 | |
| 44.0 | 12.0 | 72.7 | 44.8 | 9.9 | 78.0 | ||
| 6 | 54.0 | 17.0 | 68.5 | 50.8 | 14.6 | 71.4 | |
| 54.0 | 18.0 | 66.7 | 48.4 | 13.2 | 72.7 | ||
| 7 | 42.0 | 11.0 | 73.8 | 38.0 | 11.7 | 69.1 | |
| 42.0 | 12.0 | 71.4 | 42.1 | 11.0 | 73.9 | ||
| 36.0 | 11.0 | 69.4 | 35.2 | 11.2 | 68.2 | ||
| 8 | 47.0 | 14.0 | 70.2 | 43.9 | 13.4 | 69.5 | |
| 41.0 | 14.0 | 65.9 | 41.9 | 12.9 | 69.3 | ||
| 47.0 | 16.0 | 66.0 | 44.7 | 12.6 | 71.8 | ||
- Abbreviations: EDV, end-diastolic volume; EF, ejection fraction; ESV, end-systolic volume. *Each volume and EF value in admittance catheterization; RT3D TEE, Real time three dimensional transoesophageal echocardiography and RT3DE is the average of two measurements made using the respective software for each modality.
| Pig no. | Conditions | Heart rate (beats min−1) | Blood pressure (mmHg) | ESPVR slope | PRSW | |||
|---|---|---|---|---|---|---|---|---|
| Peak-systolic | End-diastolic | Linear | Quadratic | Slope | r value | |||
| 1 | Baseline | 103 | 110 | 5 | 0.77 | 3.23 | 32.17 | 0.90 |
| Dobutamine | 122 | 146 | 14 | 4.31 | 18.90 | 95.60 | 0.99 | |
| 2 | Baseline | 112 | 96 | 5 | 1.05 | 2.15 | 45.53 | 0.95 |
| Dobutamine | 117 | 126 | 8 | 3.65 | 16.20 | 95.05 | 0.99 | |
| 3 | Baseline | 99 | 87 | 9 | 1.62 | 4.21 | 38.93 | 0.99 |
| Dobutamine | 128 | 122 | 13 | 3.69 | 14.69 | 91.10 | 1.00 | |
| 4 | Baseline | 104 | 125 | 11 | 1.57 | 4.41 | 56.00 | 0.98 |
| Dobutamine | 126 | 154 | 17 | 4.55 | 20.41 | 95.03 | 1.00 | |
| 5 | Baseline | 85 | 80 | 5 | 1.55 | 3.85 | 46.90 | 1.00 |
| Dobutamine | 107 | 107 | 8 | 7.28 | 22.70 | 63.30 | 1.00 | |
| 6 | Baseline | 92 | 96 | 13 | 1.22 | 4.76 | 52.30 | 0.96 |
| Dobutamine | 106 | 126 | 16 | 2.75 | 10.71 | 93.77 | 1.00 | |
| 7 | Baseline | 86 | 94 | 6 | 2.90 | 7.70 | 42.20 | 1.00 |
| Dobutamine | 103 | 116 | 8 | 7.25 | 30.24 | 77.44 | 1.00 | |
| 8 | Baseline | 101 | 92 | 9 | 3.67 | 7.37 | 63.27 | 1.00 |
| Dobutamine | 130 | 110 | 13 | 6.99 | 17.65 | 89.10 | 1.00 | |
- Abbreviations: ESPVR, end-systolic pressure–volume relation; PSRW, preload-recruitable stroke work. All haemodynamic variables are the average of three measurements in each of the conditions.
Bland–Altman analysis showed fair to good agreements between AC and RT3DE for end-diastolic and end-systolic volumes and ejection fractions obtained at baseline and in the presence of dobutamine infusion (Fig. 2). The mean percentage intermethod differences were 10% for end-diastolic volume, 14% for end-systolic volume and 9% for ejection fraction. The AC-derived volumes were generally higher than those obtained by RT3DE, especially in ventricles with a larger absolute volume. Eight sets of volumetric data compared between AC and CC (Table 3 and Fig. 3) demonstrated fair to good agreement. The mean percentage difference between AC and CC was 11% for EDV and ESV, and 12% for EF. Comparison of measurements between AC and RT3DE using Student's paired t testing showed significant differences for EDV (P < 0.0001) and ESV (P= 0.0003), while EF did not show a significant difference (P= 0.76). Similar comparisons between AC and CC demonstrated no significant differences for EDV (P= 0.5), ESV (P= 0.87) or EF (P= 0.28). Figure 4 illustrates correlations between AC and RT3DE, and between AC and CC for EDV, ESV and EF.
Bland–Altman plots of agreements between admittance catheterization (AC) and RT3DE for left ventricular volumes measured at baseline and during dobutamine infusion
Abbreviations: EDV, end-diastolic volume; EF, ejection fraction; ESV, end-systolic volume; RT3D TEE, Real time three dimensional transoesophageal echocardiography.
| Pig no. | Conductance catheterization (CC) | Admittance catheterization (AC) | ||||
|---|---|---|---|---|---|---|
| EDV (ml) | ESV (ml) | EF (%) | EDV (ml) | ESV (ml) | EF (%) | |
| 4 | 83 | 45 | 45.8 | 70 | 39 | 41.9 |
| 84 | 47 | 44 | 74 | 42 | 53.1 | |
| 6 | 68 | 39 | 42.6 | 72 | 37 | 38.2 |
| 65 | 36 | 44.6 | 59 | 36 | 34.8 | |
| 7 | 48 | 21 | 56.3 | 57 | 27 | 53.8 |
| 51 | 23 | 54.9 | 57 | 25 | 46.5 | |
| 8 | 57 | 23 | 59.6 | 63 | 26 | 57.1 |
| 60 | 26 | 56.7 | 64 | 30 | 59.2 | |
- Abbreviations: EDV, end-diastolic volume; EF, ejection fraction; ESV, end-systolic volume.
Bland–Altman plots of agreements between admittance catheterization- and conductance catheterization-derived left ventricular volumes obtained at baseline
Abbreviations: EDV, end-diastolic volume; EF, ejection fraction; ESV, end-systolic volume.
Correlations between AC and RT3DE (top panel) and between AC and CC for EDV, ESV and EF
Abbreviations: AC, admittance catheterization; CC, conductance catheterization; EDV, end-diastolic volume; EF, ejection fraction; ESV, end-systolic volume; RT3DE, real-time three-dimensional echocardiography; RT3D TEE, Real time three dimensional transoesophageal echocardiography.
Discussion
The porcine heart is well suited for the study of human cardiovascular disease because of its large size and the anatomical and physiological similarities. Besides application in preclinical pharmacological and genetic studies (Swindle et al. 2012), pigs are used extensively as experimental models for heart failure, cardiomyopathy, myocardial infarction, myocardial remodelling and ischaemia–reperfusion injury (Dixon & Spinale, 2009; Heusch et al. 2011).
The AC method employed in the present study has been used in murine models to derive pressure–volume relations (Kottam et al. 2006; Clark et al. 2009; Porterfield et al. 2009; Tabima et al. 2010; DeMarco et al. 2011). Superior accuracy was shown for AC-derived volumes in comparison to traditional conductance in the murine heart (Porterfield et al. 2009). Here, we report the feasibility of obtaining LV pressure–volume relations using AC in a large-animal model. Our results demonstrate good correlation of AC volumes with RT3DE volumes. Validation of AC in this model was performed in baseline conditions and during inotropic stimulation with dobutamine. These results would serve as a foundation for utilizing AC in large-animal experimental models of cardiovascular disease.
The traditional CC technology has been used in several small- and large-animal models, as well in as human research studies, to derive instantaneous pressure–volume relations from the LV and to generate load-independent parameters of contractility (Burkhoff et al. 2005; Jegger et al. 2006; Schmitt et al. 2009; Read et al. 2011). The original theory of CC was proposed by Baan and co-workers and relates the measured conductance to blood volume as a linear relation (Baan et al. 1984; Pearce et al. 2010). Conductance catheterization, although popular in basic science laboratories and evolving into the clinical setting (Kasner et al. 2007; Penicka et al. 2010; Read et al. 2011), is known to have several limitations owing to the following technical reasons. The electrical field from the conductance catheter extends into the blood pool and surrounding tissue (myocardium). The conductance method fails to correct for this parallel admittance of the myocardium accurately (Baan et al. 1984; Wei et al. 2007; Pearce et al. 2010), causing the measured blood volume to be increased erroneously, because the catheter would ‘see’ beyond the blood pool (Wei et al. 2007). The hypertonic saline technique used in CC to correct for parallel conductance (Gp) assumes a constant value for Gp (Baan et al. 1984). However, it is known that Gp is a measurement that varies during the cardiac cycle (Wei et al. 2007; Constantinides et al. 2011). The volumes of hypertonic saline required for Gp correction in CC may alter blood resistivity and haemodynamics and produce errors in the calculated volume (Kottam et al. 2006), and this is especially true for smaller animal models. There is questionable accuracy of the circuit model in CC for the blood and myocardium (Wei et al. 2004, 2005, 2007), and the relation between blood conductance and volume is non-linear owing to the non-linear shape of the stimulating electric field. Finally, it has been shown that cuvette calibration used in CC for estimation of ventricular volumes in vivo is not a very reliable measurement (Jacoby et al. 2006; Pacher et al. 2008; Krenz, 2009; Porterfield et al. 2009). Comparison of CC-derived LV volumes with magnetic resonance imaging-derived data found very poor correlation between the two (Jacoby et al. 2006). Others have shown significant underestimation of CC volumes in comparison to magnetic resonance imaging in mice (Nielsen et al. 2007). Above all, CC methods are more time consuming and often cumbersome to use in the clinical setting owing to the need for calibration of measured volumes and multiple catheter manipulations that may be required during measurements.
The scientific basis of admittance technology is that at frequency ranges of about 20 kHz, blood is purely resistive and has no measurable capacitance, but muscle has both capacitive and resistive properties (Raghavan et al. 2009). This capacitance of muscle causes a phase delay in the measured admittance signal, enabling the real-time removal of the muscle signal from the combined blood–muscle admittance signal. Thus, this technique separates the admittance of the muscle from the admittance of blood using electric field theory, creating a new circuit that models the blood as conductive, and the myocardium as both conductive and capacitive. In the AC technique, absolute blood volumes are calculated using an improved volume-conversion equation that accounts for the non-linear electric field within the LV. Admittance catheterization eliminates both volume cuvette calibration and parallel conductance determination using hypertonic saline injection to determine Gp. The phase angle output of AC provides real-time feedback of the catheter position within the LV, which is another benefit compared with traditional CC.
Transoesophageal RT3DE allows measurement of multiphase three-dimensional volumes and is suitable for the acquisition of real-time volume data; therefore, RT3DE was chosen for head-to-head comparison with AC in the present study. Quantification of LV volumes using RT3DE has been shown to be accurate and reproducible (Jenkins et al. 2007), and correlates well with magnetic resonance imaging-derived volumes (Kühl et al. 2004; Jacobs et al. 2006; Mor-Avi et al. 2008; Dorosz et al. 2012). Left ventricular volumes derived by RT3DE are accepted as being more accurate than two-dimensional echocardiographic methods because the former avoids geometric assumptions. However, it is known that RT3DE often underestimates true LV volumes (Dorosz et al. 2012), and this is attributed to difficulties with endocardial border delineation resulting from the lower resolution of RT3DE. Further investigations on the applicability of AC in comparison with LV volumes obtained by magnetic resonance imaging are planned in our laboratory.
Our study has limitations. The admittance technique uses the phase angle measurement to correct for parallel conductance of myocardium. This correction is based on specific values for the electrical conductivity and permittivity of the myocardium, which have been measured and quantified in rodents (Raghavan et al. 2009). We used these rodent values for our porcine model based on the assumption that the electrical properties of the myocardium are uniform for all mammalian species. This assumption is further validated based on studies (Gabriel et al. 1996a,b) indicating that the electrical properties of biological tissues are consistent across species and vary only with the frequency of the electrical current used in the measurement. The AC volumes and RT3DE volumes were not simultaneous on a beat-by-beat basis, because each RT3DE acquisition merged four gated subvolumes consecutively obtained from a fixed transducer position. Several RT3DE data sets had to be excluded in some of the animals owing to suboptimal imaging windows, sternal artifacts and inadequate resolution for clear endocardial definition.
Conclusions
We demonstrate the feasibility of deriving LV pressure–volume relations using AC in large animals. Our results suggest that the early experience with AC is encouraging, and that AC is more user friendly compared with traditional CC. Admittance catheterization-derived LV volumes had acceptable correlation with RT3DE, with 10–14% mean intermethod difference. Further investigation is needed to examine the accuracy of volumes in the largest ventricles, where intermethod divergence was greatest.
Appendix
Acknowledgements
We appreciate the expert technical assistance of Elizabeth Stolze, Lucas Drvol and Gretchen Fry of the Joint Cardiovascular Research Laboratory of the University of Nebraska Medical Center. The study was supported by a grant from the Children's Hospital and Medical Center, Foundation, Omaha, NE, USA. S.K. receives support from the American Heart Association and the American College of Cardiology Foundation. Philips Health Care and Transonic Scisense Inc. provided instrumentation and technical support.
