Monitoring biventricular pacing parameters depending on the left ventricle lead configuration

Abstract

OBJECTIVES:

To evaluate whether pacing or sensing configuration has an effect on pacing parameters or their time progression. Three left ventricular (LV) pacing parameters were monitored – the LV pacing threshold, pacing impedance and intrinsic R-wave amplitude.

METHODS:

Data were collected at three intervals: during implantation; between the second and fifth month after implantation (first follow-up); and between the eighth and 15th month after implantation (second follow-up). Repeated-measures ANOVA was used for the statistical analysis.

RESULTS:

The impedance, but not its time progression, was significantly higher for the LV tip to LV ring configuration than for other configurations. R-wave amplitude and impedance increased significantly (without dependance on configurations) between implantation and first follow-up, as expected. The time progression of any parameter was not dependent on configuration of the LV lead.

CONCLUSIONS:

LV tip to LV ring is the best configuration for maintaining a high impedance level. It is better to maintain an individual approach for pacing threshold and R-wave amplitude, and their settings.

Cardiac resynchronization therapy (CRT) is a standard procedure in the treatment of severe heart failure. Randomized studies (1–4) have documented the positive impact on the quality of life and prognosis of patients with severe heart failure and ventricular dyssynchrony. Currently, the most common form of CRT is biventricular pacing, ie, pacing of both ventricles. Heart failure is, in most cases, associated with intraventricular conduction delay, which manifests as an extension of QRS width on the surface electrocardiogram (ECG). Using biventricular devices, this problem has been solved by creating a second pacing spot in the lateral part of the left ventricle (LV). The LV is then paced from two spots: from the lead placed in the right ventricle and from the lead placed on the lateral wall of the LV. The stimuli are encountered in the LV, which reduces the duration of ventricular activation (QRS duration) (5–7). Devices used for biventricular pacing are either implantable CRT pacemakers (CRT-P) or implantable CRT defibrillators (CRT-D).

Configuration

CRT devices have three leads. The configuration (pacing vector) can be set by choosing the pacing electrodes. Due to the fact that CRT devices serve both purposes, ie, sensing an intracardial electrogram and pacing, the configurations can be further divided into sensing and pacing configurations.

The number of combinations depends on the type of pacing device and lead system. Compared with unipolar lead systems, bipolar systems offer more combinations because either the distal or proximal electrode can be set as the cathode. Current bipolar lead systems offer up to six configurations. Configurations are illustrated and briefly described in Figure 1.

Figure 1).

Configurations for biventricular pacing. Figure shows three leads and the device (can). Arrows illustrate stimulating vectors. 1) Left ventricle (LV) tip to can – from the distal LV electrode to the device. 2) LV ring to can – from the proximal LV electrode to the device. 3) LV tip to LV ring – from the distal LV electrode to the proximal LV electrode. 4) LV ring to LV tip – from the proximal LV electrode to the distal LV electrode. 5) LV tip to right ventricle (RV) ring – from the distal LV electrode to the proximal RV electrode (sometimes called LV tip to RV coil or LV tip to RV). 6) LV ring to RV ring – from the proximal LV electrode to the proximal RV electrode (sometimes referred to as LV ring to RV coil or LV ring to RV) (17)

The recently developed LV lead systems are quadripolar; therefore, they provide 10 LV vectors. In the past few years, the first studies regarding the quadripolar leads were published (8,9). Why is it useful to have more than six combinations and how is the setting of configuration used? All publications that involve the optimization of pacing parameters – eg, pacing threshold (10) or the prevention of phrenic nerve stimulation (11–13) – involved an individual patient approach using the method of changing settings and determining which configuration is optimal. But is it necessary to maintain the individual approach, or is there a general optimal configuration yet to be found? Can we find a statistically optimal configuration for the majority of patients? This configuration may be useful as the first-choice setting during implantation. Currently, there is no evidence that a particular configuration is optimal to be set after implantation.

The aim of the present study was to find a configuration that is generally optimal for maintaining one of the three most important pacing parameters (R-wave amplitude, impedance and pacing threshold).

METHODS

The following three parameters were measured: R-wave amplitude, impedance and pacing threshold. Values were measured in the following intervals: during implantation; between the second and fifth month after implantation (first follow-up); and between the eighth and 15th month after implantation (second follow-up). The design of the present study is an additional data analysis. The data were collected retrospectively from medical documentation at three clinics. The treatment of patients was not influenced and the principles of confidentiality and randomization were respected. All measurements were performed using the implantable device and no external equipment was used. There is a difference in the procedure between measurement of R-wave amplitude, and measurement of pacing threshold and impedance. R-wave amplitude is obtained from the intracardial electrogram record, which is measured using a sensing configuration. In contrast, the two other parameters – pacing threshold and impedance – are measured using ventricular pacing. In this case, the pacing configuration is used and the differences in tissue response to the pulse are measured. In the present study, parameters were evaluated separately; therefore, it was assumed that the different measurement procedures do not influence results of the study.

The statistical population consisted of 33 patients who were indicated for resynchronization therapy in accordance with guidelines (14). Originally, 45 patients were randomly selected from medical documentation, but because some configurations occurred in only a few cases, the study population was reduced to 33 patients – the patients with the infrequent configurations were excluded from the data set for statistical analysis. The LV tip to LV ring and LV tip to right ventricle ring configurations were most frequent ones.

Three different devices were used in the data set: Cognis 100-D (Boston Scientific, USA; 72%), Guidant H230 (Guidant Corporation, USA; 13%) and Lumax 340 HF-T (Biotronik, Germany; 15%). Another repeated-measures ANOVA was used to avoid the effect of the device on measured parameters. Results of this analysis are presented in the data preprocessing section. None of the results were statistically significant; therefore, it can be said that the devices are comparable with each other.

Data preprocessing

The impact of the stimulation impulse on the myocardium depends on its energy. The relationship between pulse amplitude and pulse duration for constant energy is not linear; therefore, it is not easy to compare pacing thresholds with different pulse lengths. For better analysis of pacing thresholds with different pulse duration, the pacing threshold was recalculated to the minimum energy needed for the contraction of the myocardium. This value, which has been labeled ‘minimal stimulation energy’, is based on the following equation, in which E_ms is the minimal stimulation energy (μJ), U_sp is the amplitude of pacing threshold (V), t_sp is the stimulation pulse duration (ms) and R is the impedance measured in the LV (Ω):

$${\text{E}}_{\text{ms}}=({{\text{U}}_{\text{sp}}}^2\times {\text{t}}_{\text{sp}})/\text{R}$$

Nonparametric Lilliefors test was used for normality testing. The null hypothesis was rejected in most cases at a significance level of P=0.05 and, therefore, the data were transformed. The natural logarithmic transformation was used. The equation between values is following: new value = ln (original value). After transformation, some differences occurred, but only in a small number of cells, and they were not extreme (mostly P>0.02). The equality of variances was tested using the Brown-Forsyth test. Results only showed a significant difference in one case, which was not extreme (P>0.01). Therefore, parametric methods were used for the following analysis.

Repeated-measures ANOVA was used for statistical analysis. Application of one statistical test prevents the increased probability of type I error that can manifest when several two-tailed tests are used. ANOVAs were computed using Statistica software (StatSoft Inc, USA), in which the test of sphericity and its correction are a part of the test. Therefore, the sphericity were always tested and, if necessary, corrected during the ANOVA test.

ANOVA for repeated measures yields three sets of results, called effects. The first set of results is the effect of the group, which relates to mean values in groups without regard to time. This means that each patient is represented as the mean value of all three terms. In the graphical representation of means, this effect is shown as a vertical distance between curves. The second set of results is the effect of time, which relates to the time progression of parameters regardless of group association – it brings together all configurations in one and tests whether a parameter changes over time. The third set of results is the effect of the interaction between group and time. This effect relates to time progression differences with regard to group association; therefore, it highlights the groups (configurations) with different time progression. If the interaction does not exist, the curves of the graphical representation of means are parallel.

Another preanalysis was performed to test whether different types of devices influence any of the measured parameters. The same statistical method – repeated-measures ANOVA – was used. The device was set as the grouping factor to determine whether the impedance, R-wave amplitude and pacing threshold or their time progression depend on the device. The results were: impedance P>0.19 (device) and P>0.57 (time progression); minimal stimulation energy P>0.051 (device) and P>0.84 (time progression); and R-wave amplitude P>0.11 (device) and P>0.79 (time progression). The P values from the results of this preanalysis were not lower than the significant value 0.05; therefore, for purposes of the present study, it can be said that different types of CRT devices do not influence parameters measured. However, care must be taken when interpreting results for minimal stimulation energy due to its low P value, which is close to the cutoff for statistical significance.

RESULTS

For impedance, the effect of time and effect of group were statistically significant (P<0.00001 and P<0.0002). This suggests that the impedance has significant progression in time and that at least one group (configuration) has a significantly different mean value than other groups. Fisher’s post-hoc test showed that the impedance had a significantly different mean value for the LV tip to LV ring configuration. This is also shown in Figure 2 as a displacement of the curve for the LV tip to LV ring configuration. Fisher’s test for the effect of time shows that the impedance changes significantly between implantation and first follow-up, and between implantation and second follow-up. On the other hand, the effect of interaction between group and time is statistically nonsignificant (P=0.42 shows that the time progression of impedance does not depend on configuration, ie, time progression is similar for all configurations).

Figure 2).

Time progression of impedance means for each configuration with the 95% CIs. LV Left ventricle; RV Right ventricle

The effect of group and the effect of time were not statistically significant for minimal stimulation energy (group effect P>0.12; time effect, P>0.24). This indicates that minimal stimulation energy (calculated from the pacing threshold) does not have significant progression in time and that no configuration has a significantly different mean value than other groups. The interaction between group and time for minimal stimulation energy is also statistically nonsignificant, with the P value being quite lower, but not <0.09. This indicates that the time progression of minimal stimulation energy (pacing threshold) does not depend on configuration (ie, it is similar for all configurations). Results for minimal stimulation energy (pacing threshold) are presented in graphical form in Figure 3.

Figure 3).

Time progression of minimal (min) stimulation energy (pacing threshold) means for each configuration with the 95% CIs. LV Left ventricle; RV Right ventricle

For the amplitude of the R-wave, the effect of time was statistically significant (P<0.02), ie, the amplitude of the R-wave exhibited significant progression in time. Fisher’s least significant difference test has been used for detailed assessment of the output, and the results show a significant increase between implantation and the first follow-up, and between implantation and the second follow-up. The two other effects (group and interaction) are statistically nonsignificant. Results for R-wave amplitude are shown in graphical form in Figure 4.

Figure 4).

Time progression of R-wave amplitude means for each configuration with 95% CIs. LV Left ventricle; RV Right ventricle

DISCUSSION

Based on the results, the time progression of any parameter (impedance, pacing threshold, R-wave amplitude) is not influenced by the LV lead configuration. The lowest P value was for minimal stimulation energy (pacing threshold), but did not exceed statistical significance.

In contrast, significantly higher impedance (if the time progression is disregarded) appears within the group of patients with LV tip to LV ring configuration. In this configuration, the distance between the electrodes is the shortest of all configurations considered – the distance may have an impact on this higher value. Impedance is one of the basic parameters for setting the pacing devices. This parameter is related to the battery discharging and the battery longevity.

High impedance is associated with lower current drain; therefore, the pulse generator longevity increases with impedance. The stability of high impedance leads has been documented for >10 years (15). Our results show that the configuration may also be an important factor in maintaining higher impedance, which extends the longevity of the power source. The stability of impedance that was discussed (15) is not the objective of the study, but the graphical representation of time progression (Figure 2) does not show any changes in impedance that would be a sign of instability.

For the past few years, the feasibility of optimizing LV lead position, atrioventricular delay and interventricular delay by intracardiac impedance has been investigated (16). Existing research dealt only with devices during implantation and only with one configuration. However, in the future, the optimization of ingrown leads may be performed, and different measurements of impedance (using different configurations) may then be investigated. The result observed in the present study that impedance is higher for configurations with the shortest distance between electrodes can be used in further research such as this.

Other measured quantities (ie, the R-wave amplitude and pacing threshold) do not depend on the configuration.

Results also show the expected output that impedance and R-wave amplitude significantly change over time (without depending on configuration). The changes were expected to be more pronounced due to ingrowth processes during the first months (between implantation and first follow-up), as was observed.

CONCLUSION

The present study shows that the configuration has an impact on the impedance of lead system. The LV tip to LV ring configuration shows a higher impedance of the lead system than the other configurations evaluated. Higher impedance, combined with low energy output programming, can increase the pacemaker longevity. For this reason, it may be used as the first-choice setting during implantation.

However, configuration does not have an impact on pacing threshold and amplitude of R-wave in intracardial ECG record, nor does the configuration influence the time progression of any measured parameter.