ELECTROPHYSIOLOGY
Can transventricular intracardiac impedance measurement discriminate haemodynamically unstable ventricular arrhythmias in human?
1 Department of Cardiology, Castle Hill Hospital, East Yorkshire, UK; 2 Biotronik UK; 3 Biotronik GmbH & Co KG, Erlangen, Germany
Manuscript submitted 8 December 2005. Accepted after revision 4 November 2006.
* Corresponding author: Princess Alexandra Hospital, Ipswich Road, Woolloongabba, Brisbane, QLD 4012, Australia. Tel: +00617 32402537; fax: +00617 32407630. E-mail address: gerald_kaye{at}health.qld.gov.au
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
Aims To measure changes in transventricular impedance during arrhythmias.
Methods and results Patients were studied during electrophysiological studies. A quadrapolar catheter was positioned at the right ventricular apex (RVA) and a decapolar catheter within the coronary sinus (CS). Transventricular impedance was measured by injecting a subthreshold biphasic rectangular pulse of 600 µ A between poles 1 of the CS catheter and pole 1 of the RVA catheter and the voltage measured between CS pole 10 and RVA catheter pole 4. Stroke impedance (SZ), surface ECG, intracardiac electrogram (IEGM), and invasive femoral artery blood pressure (FAP) were recorded. Twenty-eight patients were analysed, 5 with inducible, haemodynamically unstable ventricular tachycardia (VT) (HUSVT), 5 with stable VT (HSVT). During HUSVT, the SZ value reduced to 22% (range 0.15–0.32 P < 0.001) in comparison with sinus rhythm. For HSVT, the SZ value reduced to 58% (range 0.33–0.88) P < 0.01, significantly different from HUSVT (P < 0.01). There was a good correlation between reduction of SZ and arterial pulse pressure (PP) during arrhythmias (r = 0.95).
Conclusion Changes in SZ strongly correlated with PP amplitude. Transventricular impedance fell significantly during unstable arrhythmias and may be useful as a sensor capable of haemodynamic discrimination.
Key Words: Transventricular intracardiac impedance, Ventricular tachycardia, Haemodynamic stability
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
Intracardiac impedance was first investigated as a means of assessing the contractile capacity of the heart more than half a century ago. In 1953, Rushmer et al1
demonstrated that changes in impedance related to various phases of the cardiac cycle. There is a strong relationship between stroke volume and changes in intracardiac impedance.2,3 Currently, implantable cardioverter defibrillators (ICDs) identify arrhythmias on the basis of changes in the heart rate and no assessment of haemodynamic stability is made. Occasionally, inappropriate shock delivery can occur for haemodynamically stable arrhythmias, which can be painful and frightening. A device capable of monitoring the haemodynamic stability of arrhythmias would give additional information and potentially reduce the frequency of inappropriate intracardiac shocks. Also, such a form of haemodynamic monitoring could help to assess the optimal pacing mode of biventricular devices.
Haemodynamic sensors have been used with variable success in the past. The ideal might include a lead-based pressure sensor in the right ventricle and current technology is such that long-term pressure sensing may now be feasible.4 In a previous study by Khoury et al.,5 impedance amplitude was measured using a multipolar catheter within the right ventricle with simultaneous right ventricular pulse pressure (PP) during stable and unstable arrhythmias. Impedance fell significantly during unstable arrhythmias and the study showed 100% specificity at detecting haemodynamic instability. This clearly suggested that intracardiac impedance could be used as a haemodynamic sensor. However, current commercial systems use unipolar impedance, measuring changes in impedance between the tip of the pacing lead and the generator can. This reflects alterations in RV contractility. (Inos, Protos and Cylos, Biotronik GmbH, Berlin, Germany).3,6
Using temporary pacing leads, we have previously investigated the role of unipolar intracardiac impedance to discriminate haemodynamically stable from unstable arrhythmias,7 but have been unable to reproduce the result shown by Khoury using a multipolar catheter. Thus far a reliable change in impedance during arrhythmias has proven elusive.
These investigations have raised the question of what is the optimal impedance configuration capable of acting as a reliable haemodynamic sensor. Unipolar impedance would tend to reflect right ventricular dynamics and may not give an accurate reflection of left ventricular changes. The effect of predominantly left ventricular function might require measuring impedance across both ventricles (transventricular impedance).
The aim of this study was to investigate transventricular impedance during arrhythmias in humans.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
Patients
A total of 37 patients was included in the study. Nine were excluded due to difficulty placing the coronary sinus (CS) electrodes. Twenty-eight patients (25 males) had their data analysed. The protocol was passed by the local Ethics Committee and complied with the Declaration of Helsinki. Written informed consent was obtained in all cases. Procedures were performed in the cardiac electrophysiology laboratory with the patients supine and in the non-sedated post-absorptive state. All patients underwent routine ventricular tachycardia (VT) stimulation studies as part of their risk assessment for sudden death. A 6 French pacing/recording electrode (Bard, Crawley, UK) was positioned at the right ventricular apex (RVA) from the right femoral vein and a 7 French decapolar deflectable catheter (Daig, a division of St Jude Medical, St Paul, MN, USA) was positioned in the CS as far distal as possible with X-ray confirmation (Figure 1).
|
Pacing protocols
During routine clinical VT stimulation up to two extrastimuli were delivered during sinus rhythm (if applicable), and during RVA pacing at 600 and 400 ms. As an additional part of this study, ventricular drive pacing for a period of 1 min at a rate of 400 ms (150 ppm) was also performed.
Transventricular intracardiac impedance
The methodology for impedance measurement has been previously described.6 In brief, impedance was measured using an external Inos2CLS implantable pacemaker directly connected to both catheters. The device was not modified. The impedance data were transmitted via telemetry to a pacemaker programmer (PMS1000 + with customized software), which in turn generated an impedance-proportional analogue output signal.
Impedance was measured by injecting subthreshold biphasic rectangular pulses of 600 µ A current between poles 1 (distal) of the CS catheter and pole 1 (distal) of the RVA catheter and the voltage was measured between CS pole 10 (proximal) and RVA catheter pole 4 (proximal) (Figure 2). The current pulse duration for each polarity was 15 µ s. Pulses were repeated every 8 ms.
|
The absolute impedance varies for individual patients. The measured level depends on the mutual position of the RV and left ventricular (LV) lead, which differs between patients but remains stable, if the lead position is not changed. Initial gain adjustment is necessary for each patient before the measurements. After adjustment the required gain did not change for the complete procedure and no further adjustments were performed.
Stroke impedance (SZ) was derived and defined as the difference between maximum impedance at an end systolic time window and the minimum impedance at an end diastolic time window (Figure 3). Impedance was recorded and continuously displayed in real time together with a surface ECG, intracardiac electrogram (IEGM) and invasive femoral artery blood pressure (FAP). Femoral artery blood pressure was measured using a 4F desi-valve sheath inserted into the femoral artery under local anaesthesia and connected to a fluid-filled transducer (Medex Medica, Rossendale, Lancs, UK, model no. MX 9604).
|
The ECG, IEGM, impedance, and FAP were recorded synchronously during the study at a sampling rate of 1000 Hz per channel. Impedance and blood pressure channels were filtered by a low-pass filter (0–500 Hz). Pacing was performed through electrodes other than those used for impedance sensing (i.e. using the Inos2CLS), in order to prevent recording of artefacts.
Haemodynamic instability was defined arbitrarily as a reduction in systolic pressure to < 90 mmHg, a change in or loss of consciousness or the requirement for emergency resuscitation.
Data analysis
Stroke impedance, SZ was analysed during sinus rhythm, RVA pacing, and during induced ventricular tachycardia. The Student's t-test was used for comparisons between sinus rhythm and each arrhythmia. Among the included arrhythmia groups a one-way analysis of variance was performed (SPSS, version 13.0, SPSS Inc., Chicago, IL, USA). Levene test was used for confirming homogeneity of variances. Post hoc comparisons were made according to Bonferroni. Significance level was 0.05.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
A total of 28 (25 males) patients' data, mean age 61 ± 11 years were analysed. The mean left ventricular ejection fraction was 36%.
Five patients had inducible, haemodynamically unstable VT (HUSVT) and 5 had stable VT (HSVT). Two patients also had an inducible supraventricular tachycardia. Ventricular pacing at 150 ppm was performed in 22 patients, 19 of them were haemodynamically unstable (HUSVpace), three stable (HSVpace).
The results of changes in transventricular stroke impedance (SZ) are summarized in Table 1 and Figure 4. Compared with sinus rhythm there was a significant reduction in average SZ (minimum–maximum) during HUSVT to 22% (0.15–0.32) and HUSV pace to 19% (0.03–0.82) P < 0.001 compared with a reduction for HSVT to 58% (0.33–0.88) P < 0.01. The reductions for both the HUSVT group and the HUSVpace group were significantly larger than for the HSVT group (P < 0.01). No significant difference was found between the HUSVT and HUSV pace groups. The two groups HSVpace [three patients, SZ reduction to 67%, 0.21–1.38, and PP reduction to 41%, 0.4–0.42] and supraventricular tachycardia (SVT) (two patients, SZ reduction to 50%, 0.49–0.51, and PP reduction to 76%, 0.74–0.77) have not been included into the statistics and the further discussion due to low patient numbers. The greatest reduction in SZ occurred during HUSVT and HUSV.
|
|
As illustrated in Figure 4, the changes in PP and SZ are closely related. Figure 5 shows the linear regression of SZ and PP for all patients with stable and unstable VT. The correlation coefficient is r = 0.95 (P < 0.001, n = 10).
|
For calculation of sensitivity and specificity for both VT–groups, we selected the optimal cut-off point of SZ-reduction to 32%. With this threshold we achieve 100% sensitivity and specificity (n = 10, 5 HUSVT, and 5 HSVT). If we also add the HUSV pace results, for the same threshold we get 88% sensitivity, 100% specificity, 100% positive predictive value, and 63% negative predictive value.
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
Unipolar intracardiac impedance measured between the pacing poles of an implanted lead and the pacing generator is currently used in commercial systems as a rate sensor assessing changes in the inotropic state of the myocardium.3
Previous studies have shown a strong correlation between stroke volume and impedance,2 suggesting that impedance could act as a haemodynamic sensor. Most studies have examined impedance changes either at rest or during exercise where changes in heart rate are not extreme. To increase the capabilities of modern implantable defibrillators any impedance sensor would need to be effective across a wide range of heart rate changes. In addition, impedance reflects a complex series of changes with the cardiac cycle and it is not currently known which configuration gives the optimal and most stable impedance signal which accurately reflects changes in intracardiac haemodynamics.
Khoury et al.5 used a multipolar temporary electrode placed in the right ventricle and showed a fall in impedance amplitude during arrhythmias in humans. The greatest fall occurred in patients with haemodynamically unstable arrhythmias suggesting that impedance could be used as a haemodynamic sensor. However, we have performed similar studies measuring unipolar impedance using temporary pacing leads in the RVA. Although we found that the greatest reduction occurred during ventricular fibrillation, an event associated with complete haemodynamic collapse, changes during unstable and stable VT and SVT were less predictable.7 These changes did not allow reliable haemodynamic differentiation between arrhythmias.8
There may be a number of reasons for this lack of reliability. A major issue is related to catheter movement, which can significantly alter impedance measurements. During ventricular fibrillation, where ventricular contraction is effectively zero catheter tip movement may not be an issue and this was reflected in a reproducible reduction in impedance amplitude. However, during ventricular and supraventricular arrhythmias, where there may be strong local contractions in the RV apex, even though overall cardiac output may be reduced, catheter tip movement may be a significant factor producing variations in impedance waveform. This may be particularly important during rapid changes in heart rates. In commercial systems with permanent leads, fibrosis of the tip ensures that the lead movement is directly related to movement of the heart. It is likely therefore that any impedance signal will be more stable under these conditions.
Impedance data from the right ventricle may additionally be affected by the complex right ventricular geometry, producing a false positive indication of sufficient pumping activity despite haemodynamic instability. This can originate from strong local wall motion around some of the impedance-measuring electrodes and can occur even when no blood is effectively pumped by these contractions. Unipolar impedance measures changes within a small distance from the catheter tip and local contractions may therefore be of more influence.
Multipolar impedance with a multi-channel catheter in the right ventricle assesses volume and output changes within the right ventricle.5 Although there is evidence that right ventricular changes during arrhythmias reflect left ventricular dynamics, multipolar impedance can be distorted by atrial activity due to the location of the poles close to the tricuspid valve.3,9 Unipolar impedance works commercially and provides a simpler waveform for analysis. Ideally, as the left ventricle has a more regular geometry than the right and is less deformed during contractions, an impedance sensor that strongly reflects LV changes is desirable. Transventricular impedance, measuring impedance changes across the left and right ventricles, may be less sensitive to local wall motion artefacts and be suitably insensitive to atrial distortion.
The development of biventricular pacemakers allows the future possibility of an impedance sensor using the CS pacing lead to measure transventricular impedance. An experimental dog study has shown a good correlation between transventricular impedance measurement and stroke volume.10 We, therefore, investigated configurations other than in the right ventricle in humans, which may more accurately represent changes within the left ventricle.
In the present study, the configuration used produced satisfactory signals and showed a good correlation between SZ and arterial pulse pressure. The results implied that as the arrhythmia became ‘increasingly unstable’ there was a corresponding reduction in SZ. During unstable VT and right ventricular apical pacing at 150 ppm the fall in SZ was accompanied by a corresponding decrease in pulse pressure. During RVA pacing, there was a large fall in arterial PP comparable with that seen during unstable VT. It is well recognized that pacing the RVA has a deleterious effect on left ventricular output, and this study emphasizes that the fall in output would appear to be similar to that seen during ventricular arrhythmias.
Overall, however, in the current study the diagnostic window between stable and unstable VT was small and may not allow automatic haemodynamic differentiation based on impedance alone. This effect also may have been partly due to small patient numbers in this study.
We conclude that there was good linear correlation between the stroke impedance SZ and arterial PP for stable and unstable VT, and that SZ for HUSVT was significantly different from both sinus rhythm and HSVT. This would support the use of impedance as a sensor for determining changes in the haemodynamic stability of arrhythmias. However, in the present study the changes during arrhythmias were variable and impedance alone may not allow real time discrimination of haemodynamics. The results are encouraging, however, and further studies are required to determine the optimal impedance signal to act as a reliable haemodynamic sensor. Although there is no requirement for a sensor as far as VF is concerned, as all ICDs are fully sensitive in detecting VF, a haemodynamic sensor particularly for biventricular devices could be a useful adjunct in patient management. Further studies are needed to investigate the long-term sensitivity and reliability of impedance measurement before any future clinical applications are implemented.
Limitations of the current study
The major limitation was small patient numbers. During the study period, the results of MADIT II, were published and as a result the number of patients undergoing routine VT stimulation significantly reduced. The positive induction rate for VT stimulation studies is low at best and, therefore, high patient numbers were difficult to recruit.
All patients were studied in the supine position and the effect of changes in posture on impedance signal was not assessed and, therefore, these changes may not reflect ‘real life’.
Temporary pacing electrodes were used. Although active fixation temporary electrodes are now available these were not available at the start of the study. It is possible that changes in the impedance signal may have given a more predictable result in the presence of active fixation electrodes.
|
Acknowledgements |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
We thank Dr Tim Houghton for his efforts in recruiting patients to this study.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
[1] Rushmer RF, Crystal DK, Wagner C, Ellis RM. Intracardiac impedance plethysmography. Am J Physiol 1953; 174: 171–4.
[2] Salo RW, Wallner TG, Pedersen BD. Measurement of ventricular volume by intracardiac impedance—theoretical and empirical approaches. IEEE-Trans Biomed Eng 1986; 33: 189–95.
[3] Osswald S, Cron T, Gradel C, Hilti P, Lippert M, Strobel J, et al. Closed-loop stimulation using intracardiac impedance as a sensor principle: correlation of right ventricular dP/dtmax and intracardiac impedance during dobutamine stress test. Pacing Clin Electrophysiol 2000; 23: 1502–8.[CrossRef][Medline]
[4] Magalski A, Adamson P, Gadler F, Boehm M, Steinhaus D, Reynolds D, et al. Continuous ambulatory right heart pressure measurements with an implantable haemodynamic monitor: a multicenter 12 month follow-up study of patients with chronic heart failure. J Card Fail 2002; 8: 63–70.[CrossRef][Web of Science][Medline]
[5] Khoury D, McAllister H, Wilkoff B, Simmons T, Rudy Y, McCowan R, et al. Continuous right ventricular volume assessment by catheter measurement of impedance for antitachycardia system control. Pacing Clin Electrophysiol 1989; 12: 1918–26.[CrossRef][Medline]
[6] Irnich W. Impedance, conductance, resistance: definitions and measurement principles. In Winter UJ, Klocke RK, Kubicek WG, Niederlag W (Eds.). Thoracic impedance measurements in Clinical Cardiology 1994; Stuttgart George Thieme pp. p4–10.
[7] Arthur W. Can the real time measurement of intracardiac impedance discriminate haemodynamically stable from unstable arrhythmias? 2003; Thesis,Newcastle upon Tyne, UK University of Newcastle.
[8] Kaye G, Arthur W, Edgar D, Lippert M, Czygan G. The use of unipolar intracardiac impedance for discrimination of haemodynamically stable and unstable arrhythmias in man. Europace 2006; 8: 988–93.
[9] Park CH, Nishimura K, Katano M, Matsuda K, Okamoto Y, Ban T. Analysis of right ventricular function during bypass of the left side of the heart by afterload alterations in both normal and failing hearts. J Thorac Cardiovasc Surg 1996; 111: 1092–1102.
[10] Zima E, Lippert M, Czygan G, Merkely B. Determination of left ventricular volume changes by intracardiac conductance using a biventricular electrode configuration. Europace 2006; 8: 537–44.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||