Do electrical parameters of the cardiac cycle reflect the corresponding mechanical intervals as the heart rate changes?

Transcription

1 Europace (2010) 12, doi: /europace/euq068 CLINICAL RESEARCH Pacing and CRT Do electrical parameters of the cardiac cycle reflect the corresponding mechanical intervals as the heart rate changes? Eraldo Occhetta 1 *, Giorgio Corbucci 2, Miriam Bortnik 1, Cristina Pedrigi 3, Salah A.M. Said 4, Herman T. Droste 4, Robert Hofmann 5, and Paolo Marino 1 1 SCDU Cardiologia 1, Azienda Ospedaliero-Universitaria Maggiore della Carità, Novara, Italy; 2 Medtronic SQDM, Maastricht, The Netherlands; 3 Divisione Clinica di Cardiologia, IRCS S. Raffaele, Milano, Italy; 4 Department of Cardiology, District Hospital Strekziekenhuis Midden-Twente, Hengelo, The Netherlands; and 5 Cardiovascular Division, City Hospital Linz, Linz, Austria Received 2 November 2009; accepted after revision 11 February 2010; online publish-ahead-of-print 26 March 2010 Aims The evoked QT interval can be detected beat by beat through an implanted pacing system. The correlation between the right ventricular paced QT interval and the left ventricular systolic interval is not known. The aim of our study was to collect data on the correlation between QT and systolic and diastolic indexes at different heart rates in patients with dual-chamber rate-responsive pacemakers.... Method and The study involved 13 patients [ years; ejection fraction (EF) %] with standard indication for results dual-chamber pacing. Patients were evaluated at rest in the supine position. The AV delay was set at 130 ms, and the pacing rate was increased from 90 to 130 bpm (10 bpm steps for 3 min). At the end of each 3 min step, QT intervals were automatically evaluated in real time by means of pacemaker telemetry. We also evaluated heart performance by means of echo-2d (end-diastolic/end-systolic volumes, EF) and echo-doppler measurements [left ventricular ejection time (LVET) and diastolic filling time (LVDFT), aortic velocity time integral, and systolic volume] and systemic arterial pressure. The QT interval progressively decreased from to ms as the pacing rate was increased from 90 to 130 bpm. The correlation between the QT interval and LVET as a function of the pacing rate was R 2 ¼ 0.966, indicating a good and relatively parallel trend in these two parameters. The correlation between RR-QT (reflecting electrical diastole) and LVDFT (reflecting mechanical diastole) was R 2 ¼ The index LVET/QT (ratio between mechanical and electrical systole) was constant in the range bpm, but significantly decreased at 130 bpm: the mechanical LVET shortens more than the electrical QT does at the highest heart rates.... Conclusion In paced patients at rest and during artificially increased heart rates, QT interval dynamics is closely correlated with changes in ejection time, thus constituting an electrical parameter of systolic time. A similar correlation exists between RR-QT, as a diastolic electrical interval, and the DFT Keywords QT interval Ejection time Rate-responsive pacing Cardiac resynchronization therapy Introduction The cardiac ventricular muscle begins to contract a few milliseconds after the onset of the electrical action potential and continues to contract for a few milliseconds after the termination of the electrical action potential. Therefore, the duration of the contraction of the cardiac muscle is mainly a function of the duration of the electrical action potential. 1 The ratio of contraction time (systolic interval, SI) to relaxation time (diastolic interval, DI) increases as the heart rate increases. At a normal heart rate of 75 bpm, the period of contraction constitutes 34% of the cardiac cycle (RR interval). At 200 bpm, * Corresponding author. SSD Elettrofisiologia e Cardiostimolazione Dipartimento Cardiovascolare, Azienda Ospedaliero-Universitaria Maggiore della Carità, Corso Mazzini 18, Novara, Italy. Tel: ; fax: , occhetta@r-j.it; eraldo.occhetta@maggioreosp.novara.it Published on behalf of the European Society of Cardiology. All rights reserved. & The Author For permissions please journals.permissions@oxfordjournals.org.

2 QT and systolic ejection time 831 it increases to 53% of the cardiac cycle. 1 This implies that in some conditions, the heart may not remain relaxed long enough to allow complete filling of the cardiac chambers before the next contraction. Thus, estimation of the SI, if related to the DI and/or to the RR interval, could provide important information on the beat-to-beat performance of the heart. In the past, many authors estimated the SI as the interval between the Q-wave and the second heart sound (Q-S2). 2 4 The QT interval might provide an alternative electrical estimate of the SI, without the need for additional mechanical sensors. 5,6 The present study was aimed at collecting data on the QT interval, SI and other electrical and echo-derived parameters, and indexes of the cardiac cycle at different pacing rates in patients implanted with dual-chamber pacemakers. Methods All patients with standard indication for dual-chamber pacing, advanced AV block, or a PR interval exceeding 150 ms at rest, without other major cardiopathies, and implanted with a DDDR pacemaker Diamond II (Vitatron, The Netherlands) were eligible for enrolment into the study. The Diamond II was a DDDR pacemaker with two rate-responsive sensors: an activity sensor and a physiological electrical sensor (the QT interval). The QT interval is derived from the paced QRS-T complex: the pacemaker measures the interval between the ventricular pulse and the maximum slope of the T-wave sensing. A total of 13 patients were included in the study: age years (range years), 10 men, 3 women. Their basal ejection fraction (EF) was %. After informed consent had been obtained, patients were evaluated at rest in the supine position. The pacemaker was programmed in the DDD pacing mode. The AV delay was set at 130 ms to ensure ventricular stimulation and automatic QT measurement during the tests. The lower pacing rate was increased from 90 to 130 bpm in 10 bpm steps. Each rate was applied for 3 min to attain stable haemodynamic conditions and stable QT intervals. At the end of each 3 min period, the performance of the heart at that specific pacing rate was evaluated by means of echo-doppler measurements. The parameters evaluated were the following: left ventricular ejection fraction (LVEF), LV end-systolic volume (LVESV), and LV end-diastolic volume (LVEDV), evaluated by means of the apical four-chamber view; LV diastolic filling time (LVDFT), evaluated as the time interval corresponding to the trans-mitral flow-velocity profile; LV ejection time (LVET), evaluated as the time interval corresponding to the transaortic flow-velocity profile; aortic velocity time integral (AoVTI), evaluated through the transaortic flow-velocity profile; systolic volume (SV), evaluated through the transaortic flow integral and the aortic cross-sectional area derived from the ring diameter; systolic blood pressure (SBP) and diastolic blood pressure (DBP), measured through a standard arm cuff. The evoked right ventricular QT interval was collected beat by beat in real time by means of pacemaker telemetry, through dedicated software installed in the device for this purpose. The stored value corresponded to the average value of 10 consecutive samples. The resolution of automatic QT measurement was 1.6 ms and that of the programmed cycle (RR) was 6.4 ms. Indexes of cardiac function derived from electrical/mechanical SI and DI were also calculated from the parameters and were evaluated at each pacing rate: correlation between QT interval (reflecting electrical systole) and LVET (reflecting mechanical systole); index LVET/RR (ratio between systolic time and pacing cycle); index LVET/QT (ratio between mechanical and electrical systole); correlation between RR-QT (reflecting electrical diastole) and LVDFT (reflecting mechanical diastole); index LVDFT/(RR-QT) (ratio between mechanical and electrical diastole); index LVET/LVDFT (ratio between mechanical systole and diastole). Statistical analysis Results are expressed as mean + standard deviation. Statistical analysis was performed by using two-tailed paired Student s t-test and the analysis of variance (ANOVA) as applicable. A P-value, 0.05 was considered significant. Results All patients completed the test without suffering any side effects. Table 1 shows the data of the electrical parameters and the related indexes: the analysis of variance showed statistically significant variations (P, ). The QT interval progressively decreased from to ms as the pacing rate was increased from 90 to 130 bpm. At each step, the decrease was statistically significant in comparison with the previous one (Table 1). The interval (RR-QT), which was obtained by subtracting the QT interval from the cardiac cycle, and which was taken to indicate the diastolic part of the cardiac cycle from an electrical point of view, progressively and significantly decreased at each step as the heart rate increased (Table 1). Table 1 Electrical parameters and related indexes Pacing rate (bpm) QT (ms) P-value RR 2 QT (ms) P-value QT/RR (%) P-value QT/(RR 2 QT) (%) P-value , , , , , , , , , , , , , , , ,0.0001

3 832 E. Occhetta et al. Table 2 Haemodynamic parameters Pacing rate (bpm) LVEDV (ml) P-value vs. 90 bpm LVESV (ml) P-value vs. 90 bpm LVEF (%) P-value vs. 90 bpm Pacing rate (bpm) AoVTI (cm) P-value vs. 90 bpm SV (cm 3 ) P-value vs. 90 bpm LVET (ms) P-value vs. 90 bpm , Pacing rate (bpm) LVDFT (ms) P-value vs. 90 bpm SBP (mmhg) P-value vs. 90 bpm DBP (mmhg) P-value vs. 90 bpm , Increasing trends were observed both in the index QT/RR, indicating the ratio between the electrical SI and the cardiac cycle, and in the index QT/(RR-QT), indicating the ratio between the electrical SI and the complementary electrical DI (Table 1). At each step, a statistically significant difference was seen in both indexes in comparison with the previous step. Left ventricular end-diastolic volume and LVESV decreased as the pacing rate increased, though the variations were not statistically significant at each step, but only at the highest pacing rates (Table 2). Left ventricular ejection fraction remained relatively constant in the heart rate range tested, showing non-significant variations (Table 2). Aortic velocity time integral and SV significantly decreased over the last two steps of the test (120 and 130 bpm) compared with the basal condition at 90 bpm (Table 2). Left ventricular ejection time progressively decreased as the pacing rate increased, the variations being statistically significant at each of the last three steps in comparison with the basal one (Table 2). Left ventricular diastolic filling time progressively decreased as the pacing rate increased; a statistically significant variation was observed with the pacing rates in the range bpm compared with the basal 90 bpm (Table 2). Systolic blood pressure and DBP did not show any statistically significant variation during the study (Table 2). The correlation between the QT interval (reflecting electrical systole) and LVET (reflecting mechanical systole) as a function of the pacing rate was R 2 ¼ 0.966, indicating a good and relatively parallel trend in these two parameters (Figure 1). The index LVET/RR (ratio between systolic time and pacing cycle) showed an initial increase up to the pacing rate of 110 bpm, followed by a plateau up to 120 bpm and a decrease at 130 bpm (Figure 2). The ratio QT/RR continuously increased with pacing rate (Figure 2) and was highly correlated with the trend of the LVET/RR (R 2 ¼ 0.81). Figure 1 The trends of QT and ET intervals at rest as a function of pacing rate in the range bpm. Both parameters have a linear trend. As shown by the corresponding trend lines, the slope of ET is bit higher than that of QT ( vs , respectively). ET shortens more than QT as the pacing rate increases. As a consequence, the index LVET/QT (ratio between mechanical and electrical systole) was relatively constant in the range bpm, with a relevant decrease at 130 bpm: the mechanical ejection time shortens more than the electrical QT does (Figure 2). The correlation between RR-QT (reflecting electrical diastole) and LVDFT (reflecting mechanical diastole) was also high (R 2 ¼ 0.975). Figure 3 shows the trends of LVDFT and RR-QT: the slope of RR-QT is higher than that of LVDFT. The diastolic index LVDFT/(RR-QT) (ratio between mechanical and electrical diastole) showed a linear increase in the range bpm (Figure 4). Finally, the index LVET/LVDFT (ratio between mechanical systole and diastole) showed a pretty constant trend in the range bpm (Figure 5, panel A). Panel B shows the ratio of the corresponding electrical parameters QT/(RR-QT). The trends of LVET/LVDFT and QT/(RR-QT) are discordant.

4 QT and systolic ejection time 833 Figure 2 The ratio ET/RR increases as the pacing rate increases up to 110 bpm. It reaches a plateau in the range bpm and tends to decrease at higher pacing rates. To describe this curvilinear behaviour, a parabolic trend line was used for approximation. The ratio QT/RR shows a good correlation (R 2 ¼ 0.81) with ET/RR, but it continues to increase at higher pacing rates in the range bpm. As a consequence, the ratio ET/QT is relatively constant in the range bpm and then it decreases. Figure 5 The ratio ET/DFT, between systolic and diastolic intervals (A), tends to remain constant in the range bpm. The corresponding electrical ratio QT/(RR-QT) (B) increases as pacing rate increases from 90 to 130 bpm. Figure 3 The electrical diastolic time (RR-QT) and the corresponding mechanical diastolic time (LVDFT) show a linear decrease as the pacing rate increases in the range bpm. The slope of diastolic filling time is a bit less steep than that of RR-QT. Diastolic filling time tends to shorten more than RR-QT as the pacing rate increases. Figure 4 The ratio DFT/(RR-QT) progressively increases as the pacing rate increases from 90 to 130 bpm, showing that the electrical diastole shortens more than the corresponding mechanical diastole as the pacing rate increases. Discussion The main finding of our study is the correlation between QT interval changes and the corresponding variations of the mechanical ET. In addition, thanks to the relationship between QT and ET intervals, the mechanical diastolic time (DFT) could also be estimated through RR-QT. This finding can play a role for real-time management of the upper rate limit: each implanted device could stop the rate increase when the estimated DI becomes too short. Proper haemodynamic functioning of the heart depends on the correct temporal relationship between systole and diastole. The duration of the diastole must be long enough to allow adequate ventricular filling; otherwise, the subsequent systole will not have the right ejection efficiency. In the decompensated heart, or in one with left ventricular dysfunction, the systolic shortening that takes place as the heart rate increases is more limited compared with the normal heart. Consequently, the frequency at which haemodynamic efficiency is lost is decidedly lower. This impairment is even more marked in cases of intraventricular electrical conduction disorders or during right ventricular apical stimulation; indeed, both of these conditions lengthen systolic duration further, leading to a significant increase in electromechanical latency and shortening the effective systolic ejection time. 2,3 In rate-responsive cardiac pacing, an upper limit frequency guided by a movement sensor or a biological sensor may not be adequate if the contractile efficiency of the left ventricle is poor. In such cases, the effective duration of the systole is shortened, since the pre-ejection phase is pathologically lengthened. Consequently, the effective duration of the diastole may also be reduced, thereby hindering proper ventricular filling and hence impairing the efficiency of the subsequent ejection.

5 834 E. Occhetta et al. During rate-responsive cardiac pacing, whether it be of the VVIR, DDDR, or biventricular type, it may be important to have reliable markers that estimate the duration of the systole and diastole. Indeed, if these two parameters are known in real time for each cardiac cycle, an automatic algorithm could be programmed in order to limit inappropriate increases in the pacing rate whenever the critical ratio might generate haemodynamic inefficiency. The pacemakers used in our study were designed some years ago and were endowed with an algorithm for the automatic determination of the QT on the paced complex. 5,6 We demonstrated that the electrical parameter QT can be a true expression of the mechanical systole at least in a certain range of heart rates; by subsequently subtracting the QT interval from the RR cycle, the effective duration of the diastole can then be determined, at least in terms of changes as heart rate changes. Measuring these parameters, we were able to observe that there really is a critical point in the pacing rate beyond which haemodynamic performance, as expressed by SV and AoVTI, can be impaired. In pacemakers of the past, the electrical parameter QT served as an indirect sensor of the autonomic state and was used as a rateresponsive biological sensor. However, although its specificity value was very high with regard to the metabolic demands of the organism imposed by either physical effort or emotional stress, its sensitivity and speed of response were poor. 7 It therefore had to be integrated with a movement sensor; through a mechanism of cross-checking and blinding, the double-sensor corrected the inadequacy of the single electrical sensor. Nevertheless, owing to the complexity of the algorithm, and the development of biological sensors that offered simpler determination and greater sensitivity (i.e. minute ventilation), QT was subsequently abandoned. 8 The results of our study, however, suggest that it could be taken up again and used in new systems resynchronization pacing (CRT, cardiac resynchronization therapy) as a haemodynamic sensor that can monitor in real time the changes of systolic and diastolic time, to avoid high pacing rates that could impair the haemodynamic performance of the heart. This approach was never used before, and some previous attempts failed to find a reliable solution to assess automatically the upper rate limit. 9 The change in trends of LVET and QT at the higher heart rates may provide the rationale to explain pacemaker syndrome and tachycardia-induced cardiomyopathy. Study limitations The measurement of the evoked QT interval is not new and its application to rate response has been documented in the past. Nevertheless, our findings may be relevant for heart failure population; however, we could not enrol these patients because there are no CRT devices equipped with the feature for the endocavitary QT measurement. For this reason, we consider this study a starting point to understand the correlation between electrical parameters and the mechanical function of the heart. In addition, the paced QT interval reflects the pacing myocardial coupling, conduction time from right to left ventricle, the ejection time, and part of the relaxation time, so it does not measure only the ET. All the measurements were done at rest. The critical concerns related to this condition are the following: (i) the left ventricular filling volume and resistance to ejection are different during exercise than at rest; (ii) the shortened AV delay used in this study may have relevant effects on trans-mitral flow, thereby affecting the DFT; (iii) the rate increase through electrical pacing may provide information on the effect of heart rate increase per se, but we could not evaluate the effect of heart rate increase due to exercise. The hypothesis that evoked QT interval may be valuable, as a potential surrogate of SI needs further investigation both at rest and during exercise. Whether the evoked endocavitary QT interval can give similar information as the spontaneous endocavitary QT interval could not be investigated because the device was not equipped with a feature to detect the spontaneous QT interval. Conclusions In paced patients at rest and during artificially increased heart rates, QT interval dynamics are closely correlated with changes in ejection time, thus constituting an electrical parameter of systolic time. A similar correlation exists between RR-QT, as a diastolic electrical interval, and the DFT. At relatively high pacing rates (.120 bpm), the mechanical ejection time shortens more than the electrical QT interval. Conflict of interest: G.C. is an employee of Medtronic BRC, Maastricht, The Netherlands. References 1. Guyton AC. Textbook of medical physiology. 4th ed. Philadelphia: W.B. Saunders Company, p Meiler SE, Boudoulas H, Unverferth DV, Leier CV. Diastolic time in congestive heart failure. Am Heart J 1987;114: Weissler AM, Harris WS, Schoenfeld CD. Systolic time intervals in heart failure in man. Circulation 1968;37: Chung CS, Karamanoglu M, Kovàcs SJ. Duration of diastole and its phases as a function of heart rate during supine bicycle exercise. Am J Physiol Heart Circ Physiol 2004;287:H Fananapazir L, Rademaker M, Bennett DH. Reliability of the evoked response in determining the paced ventricular rate and performance of the QT or rate responsive (TX) pacemaker. PACE 1985;8: Milne JR, Ward DE, Spurrell RA, Camm AJ. The ventricular paced QT interval - the effects of rate and exercise. PACE 1982;5: Lau CP, Leung SK, Lee ISF. Delayed exercise rate response kinetics due to sensor cross-checking in a dual sensor rate adaptive pacing system: the importance of individual sensor programming. PACE 1996;19: Cole CR, Jensen DN, Cho Y, Portzline G, Candidas R, Duru F et al. Correlation of impedance minute ventilation with measured minute ventilation in a rate responsive pacemaker. PACE 2001;24: Payne G, Spinelli J, Garrat CJ, Skehan JD. The optimal pacing rate. An impredictable parameter. PACE 1997;20: