Author + information
- Received October 23, 2006
- Revision received January 2, 2007
- Accepted January 9, 2007
- Published online May 1, 2007.
- Ward Y. Vanagt, MD⁎,†,⁎ (, )
- Richard N. Cornelussen, PhD⁎,
- Tamara C. Baynham, PhD‡,
- Arne Van Hunnik⁎,
- Quincy P. Poulina, MSc⁎,
- Fawzi Babiker, PhD⁎,
- Julio Spinelli, PhD‡,
- Tammo Delhaas, MD, PhD⁎,† and
- Frits W. Prinzen, PhD⁎
- ↵⁎Reprint requests and correspondence:
Dr. Ward Y. Vanagt, Department of Physiology, Maastricht University, P.O. Box 616, 6200 MD Maastricht, the Netherlands.
Objectives Considering the recent discovery of postconditioning, we investigated whether intermittent dyssynchrony immediately upon reperfusion induces cardioprotection as well.
Background Intermittent dyssynchrony, induced by ventricular pacing, preconditions myocardium.
Methods Isolated ejecting rabbit hearts were subjected to 30-min coronary occlusion and 2-h reperfusion. Control, left ventricular (LV) pacing preconditioning (LVPpreC) (3 × 5-min LV pacing), and LV pacing postconditioning (LVPpostC) (10 × 30-s LV pacing during early reperfusion) groups were studied. Mechanical effects of LV pacing were determined using local pressure-length loops (sonomicrometry), whereas effects on myocardial lactate release and coronary flow were assessed from coronary effluent and fluorescent microspheres, respectively. Anesthetized pigs underwent 60-min coronary occlusion and 3-h reperfusion in control and right ventricular (RV) pacing postconditioning groups (RVPpostC) (10 × 30-s RV pacing during early reperfusion). In all hearts, area at risk and infarct size were determined with blue dye and triphenyltetrazolium chloride staining, respectively.
Results Infarct size, normalized to area at risk, was 47.0 ± 12.3% in control rabbit hearts, but significantly smaller in LVPpreC (17.8 ± 6.4%) and LVPpostC hearts (17.9 ± 4.4%). Left ventricular pacing significantly altered regional mechanical work, but did not affect coronary flow or lactate release. In pigs, infarct size was significantly smaller in RVPpostC (9.8 ± 3.0%) than in control (20.6 ± 2.2%) animals.
Conclusions Intermittent dyssynchrony during early reperfusion reduces infarct size in 2 different animal models. Dyssynchrony-induced postconditioning cannot be attributed to graded reperfusion but may be induced by modulation of local myocardial workload. Dyssynchrony-induced postconditioning opens new possibilities for cardioprotection in the clinical setting.
We recently demonstrated in globally ischemic rabbit hearts that intermittent mechanical dyssynchrony, induced by ventricular pacing at physiological heart rate, limits infarct size as much as ischemic preconditioning and causes up-regulation of pro-brain natriuretic peptide mRNA (1). This protection was most likely due to the dyssynchrony-induced abnormal myocardial stretch and contraction patterns (1), because alterations in myocardial stretch (2,3) and afterload (4) are known preconditioning triggers.
Clinical application of any kind of preconditioning has been limited by the unexpectedness of most ischemic events. The recent discovery that pre-conditioning stimuli can also be cardioprotective, when applied during early reperfusion (postconditioning [5,6]), has started a new era of cardioprotection. The demonstration of ischemic postconditioning in human hearts during percutaneous coronary intervention (PCI) procedures (7) clearly shows the great clinical potential of postconditioning. Also, various pharmacologic agents have been shown to be cardioprotective in animals when applied in the reperfusion period (6,8,9).
The aim of the present study was to investigate whether dyssynchrony can induce postconditioning and, if so, whether this protective effect is equal to that of dyssynchrony-induced preconditioning. Experiments were performed in isolated ejecting rabbit hearts and in anesthetized pigs. The former model allowed some insight into the mechanisms of dyssynchrony-induced postconditioning; the latter was used to evaluate dyssynchrony-induced postconditioning in a clinically relevant larger animal model.
Materials and Methods
Isolated Rabbit Hearts
Animals and instrumentation
Animal handling and treatment were performed according to the Dutch law on animal experimentation. The protocol was approved by the Maastricht University Animal Ethical Committee. Hearts were isolated from female White New Zealand rabbits after sedation (intramuscular ketamine [50 mg/kg]/xylazine [5 mg/kg]) and cervical dislocation by a blow in the neck, and subsequently attached to the perfusion apparatus as previously described in detail (1). The aorta and a pulmonary vein were cannulated to allow antegrade perfusion (ejecting mode) of the isolated heart.
The heart was instrumented with epicardial pacing electrodes on the right atrial (RA) appendage and posterior basal left ventricular (LV) wall, a catheter to measure LV pressure, and 4 sonomicrometer crystals (Sonometrics Corp., London, Ontario, Canada) to measure regional segment length in the anterior and posterior LV wall (1). Epicardial electrographic electrodes were attached to the RA and LV anterior and posterior basal wall for determination of electrical dyssynchrony. The left descending coronary artery was encircled by a snare, and a deflated standard percutaneous transluminal angioplasty (PTCA) balloon catheter was positioned through the snare. Throughout the experiment, preload was kept constant at 6 mm Hg, afterload at 80 mm Hg, and heart rate at 240 beats/min (physiological heart rate for rabbits ).
Study protocol and groups
Rabbits were divided into 3 study groups: control (n = 12), LV pacing preconditioning (LVPpreC) (n = 8), and LV pacing postconditioning (LVPpostC) (n = 12) groups (Fig. 1).
All hearts were subjected to 30-min coronary artery occlusion (induced by inflating the balloon catheter), followed by reperfusion. In the LVPpreC group, ventricular pacing was performed during three 5-min periods, interspersed with 5 min of RA pacing. After the last ventricular pacing period, 10 min of RA pacing preceded the index ischemia. In the LVPpostC group, postconditioning was performed during the first 10 min of reperfusion using 10 cycles of 30-s LV pacing and 30-s RA pacing at 240 beats/min. Left ventricular pacing was performed using simultaneous atrioventricular pacing (atrioventricular interval = 0 ms) to ensure complete ventricular activation initiating from the pacing electrode.
Coronary artery occlusion was followed by 2 h of reperfusion (n = 8 per group) for determination of postischemic cardiac function, global coronary flow, myocardial lactate release, and quantification of infarct size. Additional experiments (control and LVPpostC groups, n = 4 per group) were performed to measure regional myocardial perfusion during the postconditioning period using fluorescent microspheres.
Data, effluent, and tissue collection and processing
Hemodynamic, mechanical, and electrophysiological data were recorded at a sampling rate of 500 Hz using a Sonometrics system (Sonometrics Corp.). Signals were analyzed off-line using Sonometrics, Matlab (The Mathworks Inc., Natick, Massachusetts) and Hemo (homemade) software. Hemodynamic data consisted of aortic and LV pressure and aortic flow. Cardiac output was calculated by adding coronary to aortic flow (ml/min).
Coronary effluent was collected for determination of lactate release as a marker of anaerobic metabolism. All samples were immediately frozen in liquid nitrogen and subsequently stored at −80°C. Determination of lactate was performed by spectrophotometry using a Cobas Bio autoanalyzer (Roche Diagnostics, Almere, the Netherlands) according to Apstein et al. (11) and expressed in U/5 min/g of tissue at risk.
Left ventricular intraventricular electrical dyssynchrony was assessed from the epicardial electrographic electrodes as the posterior-anterior activation delay.
Regional myocardial mechanical work (stroke work) was calculated as the area of the regional LV pressure-segment length loop, derived from the LV pressure signal and sonomicrometry data (1). This measure is sensitive to dyssynchrony of contraction (1,12). Pressure-length loop areas were calculated in the posterior and anterior LV wall to compare regional stroke work in early- and late-activated myocardium during LV posterior wall pacing, respectively.
At the end of the 2-h reperfusion protocol, the coronary artery was reoccluded, and the heart was perfused with 1% methylene blue dye to delineate the area at risk. The heart was stored overnight at −20°C and cut into transverse 2-mm slices, which were stained with 1% triphenyltetrazolium chloride (TTC) at 37°C for 15 min and fixed in 4% formaldehyde solution for 24 h. Finally, in all slices the area of the LV, the region at risk and the infarct region were determined by digital planimetry using Leica Qwin software (Leica, Rijswijk, the Netherlands).
In the additional 4 control and 4 LVPpostC experiments, fluorescent microspheres (15-μm spheres, 50,000 spheres; Dye-Trak, Triton Technology, San Diego, California) were injected through the LV cannula twice at 30-s intervals during RA (2 injections in control group of which the average is presented, 1 injection in LVPpostC group) and LV pacing (1 injection in LVPpostC group) at 10-min reperfusion. These experiments were stopped thereafter.
Tissue from reperfused (LV apex) and nonischemic (outside region at risk during coronary artery occlusion) myocardium were digested for isolation of the microspheres, and fluorescence was determined as described previously (13,14). Absolute regional flow (ml/min/g) was calculated by calibration of total fluorescence to the coronary effluent flow and dividing tissue flow by the weight of the tissue sample.
Data are presented as mean ± SD. Statistical analysis was performed on absolute values. One-way analysis of variance (ANOVA) was performed to compare weight and planimetry data. Two-way ANOVA for repeated measures within each group and between the groups was performed for hemodynamic, sonomicrometry, coronary flow, and lactate release data. If analysis of variance showed a significant difference, post-hoc analysis with the Tukey test was used for further comparison. For fluorescent microsphere data, comparison within and between groups was performed using paired and nonpaired ttests, respectively. A value of p < 0.05 was considered statistically significant.
In Vivo Pig Hearts
Eight female swine used in this study were treated in accordance with National Institutes of Health Guidelines for the Care and Use of Laboratory Animals. The protocol was reviewed and approved by the Guidant Institutional Animal Care and Usage Committee (Guidant Corp., St. Paul, Minnesota).
During thiopental sodium anesthesia, pacing leads (Guidant Corp.) were positioned at the endocardium of the right ventricular (RV) apex and RA. After thoracotomy, myocardial ischemia was created by ligating the left anterior descending coronary artery (LAD). Index ischemia consisted of 60 min of LAD occlusion, followed by 3 h of reperfusion.
In the control group (n = 4), hearts were allowed to beat in sinus rhythm (SR) throughout the experiment (Fig. 2).In the RV pacing postconditioning (RVPpostC) (n = 4) group, postconditioning was performed during the first 10 min of reperfusion using 10 cycles of 30-s RV pacing and 30-s SR. Right ventricular pacing was performed in the VDD-mode (atrial sense-ventricular pace) using an atrioventricular interval of 50% of intrinsic PQ time to obtain complete electrical activation of the ventricles by the pacemaker.
Infarct size determination
After clamping the aorta and reocclusion of the LAD, area at risk was determined by injection of Evan’s blue dye (1 ml/kg body weight). The ventricles were serially sectioned from apex to base in ∼1-cm slices. The heart slices were incubated for 15 min in 1% TTC solution at 37°C. Area of infarct, area at risk, and total LV area were determined using an image processing program (ImageJ, National Institutes of Health, Bethesda, Maryland).
All data are presented as mean ± SD. Groups were compared using a 2-sided ttest. A value of p < 0.05 was considered statistically significant.
Isolated Rabbit Hearts
Body weights (3.1 ± 0.4 kg) and heart weights (8.5 ± 1.0 g) were not significantly different between the 3 study groups. The area at risk of the LV was similar in the 3 study groups (control group 42.1 ± 9.4%, LVPpreC 39.3 ± 12.9%, LVPpostC 41.8 ± 12.5%). Infarct size, normalized to area at risk (Fig. 3),was 47.0 ± 12.3% in the control group, but significantly smaller in the LVPpreC (17.8 ± 6.4%, p < 0.05) and LVPpostC groups (17.9 ± 4.4%, p < 0.05, p = NS between LVPpreC and LVPpostC groups). Infarcted myocardium was uniformly distributed within the LV area at risk in all groups (Fig. 4).
Cardiac dyssynchrony and function
During RA pacing at reperfusion, PQ time was 83 ± 8 ms, and QRS duration was 25.6 ± 4.3 ms. Electrical LV posterior-anterior delay (∼electrical dyssynchrony) was 1 ± 6 ms during RA pacing, and increased significantly during LV pacing (LVPpreC 29 ± 5 ms, LVPpostC 34 ± 6 ms, p = NS between LVPpreC and LVPpostC groups). Because the AV delay was set at 0 ms during LV pacing, the entire ventricular mass was depolarized by the electrical activation wave started at the LV pacing electrode.
Average regional stroke work was 25.9 ± 4.1 mm Hg·mm in posterior and 26.7 ± 11.7 mm Hg·mm in anterior myocardium at the start of the experiment during RA pacing. Left ventricular posterior wall pacing significantly decreased regional stroke work to negative values in the LV posterior wall (early activated during LV pacing) and increased regional stroke work in the late-activated anterior LV wall (Fig. 5).These changes occurred both when LV pacing was performed before ischemia (LVPpreC group) and upon reperfusion (LVPpostC group), albeit that the amplitude of changes was smaller during reperfusion (Fig. 5).
Cardiac output and maximal rate of rise of the LV pressure curve (LVdP/dtmax) were similar in all groups at baseline during RA pacing, and were not significantly different between the groups at any corresponding time point during the protocol (Table 1).Cardiac output and LVdP/dtmaxwere significantly depressed in all groups during and after ischemia as compared with baseline. During early reperfusion in the LVPpostC group, LV pacing did neither significantly change cardiac output nor LVdPdtmaxas compared with RA pacing.
Ventricular fibrillation occurred in 1 rabbit in the control group (at 2 min of reperfusion) and in 1 rabbit in the LVPpostC group (during LV pacing at 1 min of reperfusion, p = NS between groups).
Coronary flow, myocardial perfusion, and lactate release
Coronary effluent flow was similar between the groups at corresponding time points (at baseline RA pacing: control group 71 ± 9 ml/min, LVPpreC 81 ± 10 ml/min, LVPpostC 80 ± 12 ml/min). Left ventricular pacing did not significantly change coronary effluent flow in the LVPpreC group (82 ± 18 ml/min) as compared with baseline RA pacing. During early reperfusion in the LVPpostC group, coronary flow was similar during LV (70 ± 10 ml/min) and RA (69 ± 15 ml/min) pacing.
Regional myocardial perfusion, measured using fluorescent microspheres (Fig. 6),was similar in the control and LVPpostC groups in both reperfused and nonischemic myocardium. In the LVPpostC group, myocardial perfusion was similar during RA and LV pacing.
Lactate release (Fig. 7)was not significantly different between the 3 groups at any corresponding time point and increased significantly during early reperfusion as compared with baseline within each group. Compared with RA pacing, LV pacing did not change lactate release, neither when performed before ischemia (LVPpreC group) nor during the early reperfusion phase (LVPpostC group).
In Vivo Pig Hearts
Heart weight and area at risk were similar in the control (208.8 ± 30.9 g and 55.8 ± 7.9%, respectively) and RVPpostC (218.8 ± 12.5 g and 49.6 ± 4.9%, respectively) groups. Infarct size was significantly smaller in the RVPpostC (9.8 ± 3.0%) than in the control group (20.6 ± 2.2%, p < 0.05) (Fig. 8).
During SR at reperfusion, PQ time and QRS duration were similar in the control (106 ± 4 ms and 44 ± 3 ms, respectively) and RVPpostC groups (110 ± 12 ms and 43 ± 2 ms, respectively). During RV pacing, QRS duration increased to 63 ± 6 ms. Considering an atrioventricular delay of ∼55 ms, it can be envisioned that almost all ventricular tissue was activated from the pacing site, rather than through intrinsic conduction, arriving ∼110 ms after atrial activation. Ventricular fibrillation did not occur in any of the pigs during the experiment.
The current animal studies demonstrate for the first time that myocardial infarct size after regional ischemia is reduced by intermittent dyssynchrony during early reperfusion, as induced by ventricular pacing. Dyssynchrony-induced postconditioning equals the protective potency of dyssynchrony-induced preconditioning, used as a positive control in this study. This complies with the data of the original demonstration of ischemic postconditioning by Zhao et al. (5), and later by others (15), where ischemic postconditioning was as effective as ischemic preconditioning.
Data from the current study illustrate that dyssynchrony-induced postconditioning is not limited to a single species and can be evoked in isolated, buffer-perfused hearts as well as in blood-perfused hearts in vivo.
Possible mechanism of dyssynchrony-induced postconditioning
The finding that switching from RA to LV pacing in the early reperfusion phase does not change coronary flow, lactate release, and myocardial perfusion practically rules out the possibility that dyssynchrony-induced postconditioning is caused by “graded reperfusion,” which is known to reduce infarct size considerably (16,17). Ischemic postconditioning has therefore been suggested to be a “new bottle” for the “old wine” of graded reperfusion (18), and the possibility that ischemic postconditioning is a modified form of graded reperfusion has not been completely ruled out (18,19). Accordingly, low-pressure reperfusion was equally protective as ischemic postconditioning in rat hearts (20). Data from the present study, however, provide evidence that infarct size can be reduced by interventions in the early reperfusion period even if coronary perfusion is restored abruptly and completely.
In the isolated rabbit heart preparation preload, afterload and heart rate are well controlled. In the absence of these possible triggers of postconditioning, intermittent changes in local stretch and loading appear a likely trigger of dyssynchrony-induced postconditioning. The present study and a previous study in dogs (21) show that ventricular pacing induces regional alterations in myocardial loading and stretch during reperfusion, qualitatively similar to those in nonischemic hearts (1). During LV posterior wall pacing, early-activated myocytes contract during early systole while LV pressure is low, resulting in decreased work load in the posterior LV wall. Late-activated myocytes, however, are stretched during early systole and subsequently contract while LV pressure is high, resulting in increased work load in the anterior LV wall. Furthermore, the strong contraction of late-activated myocytes results in systolic stretching of the early-activated myocardium (eccentric contraction ). Previous studies have shown that increased myocardial stretch (2,3) and workload (4) are cardioprotective.
In the present study, the ischemic region comprises myocardium, which is activated early as well as late during LV pacing in rabbits (Fig. 4) and early- to intermediate-activated in pigs (anterior ischemia with RV apex pacing).
The occurrence of dyssynchrony-induced postconditioning and the uniform distribution of infarcted myocardium within the risk area in both animal models suggest that the exact site of pacing is not critical for achieving this effect. This is in agreement with a previous study in globally ischemic hearts where VPpreC decreased infarct size, both in proximity of as well as remote from the ventricular pacing site (1). One possible explanation is that changes in stretch and workload during ventricular pacing are not limited to the late-activated myocardium alone (see the preceding text). Alternatively, remote cardioprotection within the heart, as was previously described for preconditioning (22,23), may play a role in protecting myocardium remote from the triggered zone.
Comments on the experimental set-up
The 2 species were chosen because data on preconditioning and postconditioning from rabbits and pigs translate well to humans (7,23). The isolated rabbit heart setup was used because we previously discovered dyssynchrony-induced preconditioning in this preparation (1). Moreover, the isolated heart preparation has the advantage that the controlled setup rules out several potential triggers of protection (see the preceding text). The experiments in pigs serve as extrapolation of the findings from an isolated to an in vivo model, which is more comparable to the clinical situation.
Several postconditioning studies in isolated hearts (20,24) used ventricular pacing in order to control heart rate throughout the experiment in all experimental groups. Ventricular pacing was not intermittent in these studies, and other studies were performed during atrial pacing (25) or SR (26). It is unclear whether continuous ventricular pacing allowed some degree of cardioprotection, because no data is available comparing SR, atrial and ventricular pacing for rate control during postconditioning studies. However, data from our study indicate that using ventricular pacing for rate control in experimental studies on pre- and postconditioning may introduce a confounding factor.
Regional myocardial perfusion in blood-perfused in vivo rabbit hearts is generally ∼3 ml/min/g, whereas higher values are described in the present study (∼10 ml/min/g). This difference can most likely be ascribed to the lower oxygen content and the lower viscosity of the crystalloid perfusate as compared with blood. This idea is supported by coronary flow values of ∼6.6 ml/min/g in isolated buffer-perfused hearts with a lower hemodynamic load than in the present study (27). The high absolute myocardial perfusion in the present study does not invalidate the observation that cardioprotection by ventricular pacing is achieved in the absence of reduction in myocardial perfusion.
Possible clinical applications
Findings of the present study open new possibilities for myocardial postconditioning in the clinical setting. Recently, ischemic postconditioning has been demonstrated using repetitive re-inflation of the PTCA balloon (7). Ischemic postconditioning reduced infarct size by ∼30% in acute myocardial infarct patients during PCI procedures. This important finding strongly indicates that reperfusion damage occurs in humans (6), and salvage of human myocardium at risk is possible in this clinical setting (7). Intermittent ventricular pacing is easy, controllable, and does not carry the risk of re-inflation induced vascular damage (19). Postconditioning effects have also been shown for various pharmacologic agents (6,8,9), but these drugs may have undesirable side effects.
The present, small study shows no proarrhythmogenic or adverse hemodynamic effects of intermittent ventricular pacing during the early reperfusion phase. Obviously, application of dyssynchrony-induced postconditioning in patients requires careful evaluation in clinical trials.
The study was financially supported by Guidant Corp. (St. Paul, Minnesota). Dr. Prinzen is an advisor to Guidant Corp. and Medtronic Inc. (Minneapolis, Minnesota). Drs. Baynham and Spinelli are employees of Guidant Corp.
- Abbreviations and Acronyms
- left anterior descending coronary artery
- left ventricular
- percutaneous coronary intervention
- percutaneous transluminal coronary angioplasty
- right atrial
- right ventricular
- triphenyltetrazolium chloride
- ventricular pacing postconditioning
- ventricular pacing preconditioning
- Received October 23, 2006.
- Revision received January 2, 2007.
- Accepted January 9, 2007.
- American College of Cardiology Foundation
- Ovize M.,
- Kloner R.A.,
- Przyklenk K.
- Gysembergh A.,
- Margonari H.,
- Loufoua J.,
- et al.
- Zhao Z.Q.,
- Corvera J.S.,
- Halkos M.E.,
- et al.
- Vinten-Johansen J.,
- Yellon D.M.,
- Opie L.H.
- Staat P.,
- Rioufol G.,
- Piot C.,
- et al.
- Chiari P.C.,
- Bienengraeber M.W.,
- Pagel P.S.,
- Krolikowski J.G.,
- Kersten J.R.,
- Warltier D.C.
- Weihrauch D.,
- Krolikowski J.G.,
- Bienengraeber M.,
- Kersten J.R.,
- Warltier D.C.,
- Pagel P.S.
- Prinzen F.W.,
- Hunter W.C.,
- Wyman B.T.,
- McVeigh E.R.
- van Oosterhout M.F.,
- Arts T.,
- Bassingthwaighte J.B.,
- Reneman R.S.,
- Prinzen F.W.
- Vernooy K.,
- Verbeek X.A.,
- Peschar M.,
- et al.
- Argaud L.,
- Gateau-Roesch O.,
- Raisky O.,
- Loufouat J.,
- Robert D.,
- Ovize M.
- Heusch G.
- Tsang A.,
- Hausenloy D.J.,
- Yellon D.M.
- Bopassa J.C.,
- Ferrera R.,
- Gateau-Roesch O.,
- Couture-Lepetit E.,
- Ovize M.
- Yellon D.M.,
- Downey J.M.
- Darling C.E.,
- Jiang R.,
- Maynard M.,
- Whittaker P.,
- Vinten-Johansen J.,
- Przyklenk K.
- Serviddio G.,
- Di Venosa N.,
- Federici A.,
- et al.
- Tsang A.,
- Hausenloy D.J.,
- Mocanu M.M.,
- Yellon D.M.
- Schwanke U.,
- Deussen A.,
- Heusch G.,
- Schipke J.D.