Author + information
- Received October 11, 2004
- Revision received January 6, 2005
- Accepted January 25, 2005
- Published online August 16, 2005.
- Frank Bogun, MD⁎,⁎ (, )
- Subramaniam Krishnan, MD†,
- Mukarram Siddiqui, MD†,
- Eric Good, DO⁎,
- Joseph E. Marine, MD†,
- Claudio Schuger, MD†,
- Hakan Oral, MD⁎,
- Aman Chugh, MD⁎,
- Frank Pelosi, MD⁎ and
- Fred Morady, MD⁎
- ↵⁎Reprint requests and correspondence:
Dr. Frank Bogun, Division of Cardiology, University of Michigan Health System, 3119 TC, 1500 E. Medical Center Dr., Ann Arbor, Michigan 48109-0366
Objectives The purpose of this study was to correlate infarct age with characteristics of the endocardial electrograms (EGM) obtained in patients undergoing mapping procedures for postinfarction ventricular tachycardia (VT).
Background Experimental studies have demonstrated that infarct age influences EGM duration in the subepicardial left ventricle (LV). The relationship between infarct age and endocardial EGM characteristics has not been investigated in patients with postinfarction VT.
Methods In a consecutive series of 23 patients with a history of remote infarction (range 1 to 31 years) and VT, endocardial LV mapping was performed with an electroanatomical mapping system (CARTO, Biosense Webster Inc., Diamond Bar, California) during sinus rhythm. Electrogram morphology and width were analyzed and correlated with infarct age. Isthmus sites of the VT re-entry circuits were identified by entrainment mapping and related to the results of substrate mapping.
Results There was a significant correlation between infarct age and width of the bipolar endocardial EGM during baseline rhythm in the peri-infarct zone (r = 0.84; p < 0.0001). Increasing infarct age was associated with progressive activation delays in the scar and with isolated potentials separated by an isoelectric interval, the duration of which also correlated with infarct age (r = 0.77; p < 0.001). Among all endocardial sites, the VT isthmus sites displayed the most delay and broadest EGMs during sinus rhythm.
Conclusions The presence of broad, fractionated EGMs and isolated potentials indicates a healed myocardial infarction; the older the infarction, the broader the EGM width. Remodeling over time alters the electrophysiologic properties of the peri-infarct tissue.
Myocardial infarction (MI) results in deposition of fibrous tissue and ventricular remodeling (1). Fibrous tissue invades the surviving myocardium and separates muscle bundles, resulting in slowing of impulse propagation (2). The electrophysiologic correlate of this process is fractionation and broadening of the local bipolar electrogram (EGM) and the formation of isolated potentials separated from the main ventricular EGM by an isoelectric segment (3). In a canine infarct model, there was progressive widening of EGMs recorded in the infarct zone with increasing age of the infarct (3). The effects of infarct age and remodeling on EGM width and on the arrhythmogenic substrate have not been described in humans.
To elucidate potential arrhythmogenic effects of post-infarction remodeling, this study analyzed the relationship of infarct age to EGM characteristics and to the mapping data obtained during ablation of ventricular tachycardia (VT).
Characteristics of subjects
Twenty-three consecutive patients (mean age 68 ± 7 years, mean ejection fraction 0.21 ± 0.11) referred for VT ablation underwent left ventricular (LV) endocardial mapping during baseline rhythm. All patients had a remote history of MI, with the first infarction being anterior in nine patients and inferior in nine patients. In five patients, the date of the infarction could not be determined by history. Ten patients subsequently developed infarctions in other vessel territories (six with prior anterior infarction then had an inferior MI, and four with prior inferior infarction then had an anterior infarction), and nine had recurrent infarctions in the same vessel territory. The infarct age was defined as the number of years from the time of the first infarction to the time of the electrophysiology procedure and ranged from 1 to 31 years, with a mean of 16.2 ± 9.0 years. Sixteen patients received reperfusion and/or revascularization therapy consisting of thrombolysis in three patients, primary angioplasty in four patients, and coronary artery bypass graft surgery in nine patients. The remaining seven patients did not receive revascularization therapy. Coronary angiography revealed single vessel coronary disease in 4 patients, two-vessel disease in 8 patients, and three-vessel disease in 11 patients. In 5 of 23 patients, the infarct age could not be determined on the basis of available clinical information. These patients were not included in the statistical correlation of EGM characteristics and infarct age. Data from these patients were included in the assessment of the duration index at isthmus sites and the analysis of morphologic EGM characteristics. Symptomatic VT was first documented 12 ± 9.5 years after the first MI (range 6 months to 29 years).
Thirteen patients were being treated with amiodarone at the time of the electrophysiology procedure, five patients were being treated with a combination of amiodarone and procainamide (one patient) or mexiletine (four patients), and one patient was being treated with sotalol. All 23 patients were treated with an angiotensin-converting-enzyme inhibitor or angiotensin receptor blocker. Eight patients were being treated with spironolactone, and 21 patients were being treated with a beta-blocker at the time of the electrophysiology study.
Sequence of data acquisition
Programmed right ventricular (RV) stimulation was performed to assess the inducibility of the clinical VT and other tachycardias. An electro-anatomical map of the LV then was constructed, and areas with broad EGMs were marked. With the ablation catheter positioned at the site with the broadest EGM during sinus rhythm, VT was induced by programmed stimulation, and radiofrequency energy was delivered if the site was within the re-entry circuit. The catheter was then moved to sites with narrower EGMs.
The electrophysiology procedures were performed in the fasting state after informed consent was obtained. A quadripolar catheter with 5-mm inter-electrode spacing was advanced to the RV apex or RV outflow tract and programmed stimulation was performed with up to four extrastimuli at two basic drive cycle lengths. In 22 of 23 patients, LV access was obtained with a retrograde aortic approach. A transseptal approach was used in one patient because of bilateral iliac artery occlusions. Once the catheter was positioned in the arterial system, a bolus of 5,000 U of heparin was administered followed by an infusion of 1,000 U/h to maintain an activated clotting time of approximately 250 s.
Mapping and ablation
A three-dimensional mapping system (CARTO, Biosense Webster Inc., Diamond Bar, California) was used to create a voltage map during the baseline rhythm (sinus rhythm in 22 patients and RV pacing in 1 patient) with a Navistar catheter (Biosense Webster Inc.). Bipolar EGMs were recorded between the 4-mm tip electrode and a 2-mm ring electrode separated by 1 mm. The bipolar EGMs were filtered at 10 to 30 to 400 to 500 Hz.
The endocardial sites with the broadest ventricular EGMs and the EGMs with the latest activation during sinus rhythm were sought. Electrogram width was measured with electronic calipers from the onset to the offset of the ventricular EGM (Fig. 1A)or, if an isolated potential was present, to the end of the isolated potential (Fig. 1B). The delay of the EGM was measured from the beginning of the QRS complex to the end of the ventricular EGM or, if an isolated potential was present, to the end of the isolated potential. In the presence of an isolated potential, the duration of the isoelectric segment separating the ventricular EGM from the isolated potential was measured from the end of the ventricular EGM to the beginning of the isolated potential.
The local bipolar EGMs recorded by the distal electrode pair of the mapping catheter were categorized according to previously reported criteria (4):
1. Normal EGMs: sharp biphasic or triphasic spikes with amplitudes ≥3 mV and duration <70 ms, and/or amplitude/duration ratio >0.046 (Fig. 1A).
2. Fractionated EGMs: amplitude ≤0.5 mV, duration ≥133 ms, and/or amplitude/duration ratio <0.005 (Fig. 1C).
3. Isolated potential: a potential separated from the ventricular EGM by an isoelectric segment (Fig. 1B), and/or a segment with low amplitude noise (<0.05 mV) of >20 ms duration at a gain of 40 to 80 mm/mV.
4. The remaining EGMs were defined as abnormal (Fig. 1D).
In a porcine chronic infarction model, Callans et al. (5) found that the area of contiguous bipolar EGMs of ≤1 mV correlated well with the infarct size determined by pathological analysis. Therefore, we defined low-voltage segments as areas with an amplitude <1 mV.
With a user-defined map, each endocardial site was color-coded with respect to the EGM width (Figs. 2A and 3A).The areas covered by EGM durations <133 ms, 133 to 160 ms, 160 to 200 ms, and >200 ms were measured (with rectangular, triangular, and trapezoid shapes covering the area of interest) and compared with the low-voltage area (Figs. 2B, 2C, 3B, and 3C). These EGM duration categories were chosen arbitrarily.
Target sites for radiofrequency ablation were identified with standard mapping criteria described in prior studies (6,7). In addition, to test the hypothesis that the broadest EGMs obtained during sinus rhythm play a critical role in the substrate of inducible VT, the ablation catheter was positioned at sites displaying the widest EGMs during sinus rhythm, and VT was induced by RV programmed stimulation. Radiofrequency energy was delivered if concealed entrainment was demonstrated at that site.
Isthmus sites were defined as sites where there was concealed entrainment and where radiofrequency energy delivery resulted in termination of VT and the inability to re-induce the VT. The EGM width at isthmus sites was measured during sinus rhythm and compared with the maximal EGM width obtained during endocardial LV mapping. The EGM duration index was defined as the ratio of EGM width at a particular site to the maximal EGM duration in that patient during sinus rhythm. The EGM duration index was calculated at isthmus sites in order to assess the relationship of isthmus sites to areas with the maximally delayed activation during sinus rhythm.
Continuous variables were indicated with the mean ± standard deviation. Continuous variables were compared with Student ttest, and categorical variables were compared with chi-square analysis or the Fisher exact test. Pearson correlation coefficient was calculated and indicated as r and p values. A p value of <0.05 was considered significant.
Electrophysiology study and mapping data
A total of 115 distinct VTs (mean cycle length 362 ± 92 ms) were inducible by RV programmed stimulation (5 ± 3 VTs per patient) (Table 1,Fig. 4). Forty-three VTs (mean cycle length 428 ± 85 ms; range: 280 to 580 ms) were reproducibly inducible. In 32 of 43 VTs (74%), an isthmus site could be identified and the targeted VT was successfully ablated. A mean of 164 ± 56 endocardial points (range 62 to 305 points) were recorded to reconstruct the LV geometry with the electro-anatomical mapping system.
The low-voltage area (<1.0 mV) covered 63 ± 35 cm2of the LV endocardium. Contiguous areas with an EGM duration >133 ms covered 19 ± 18 cm2, comprising 37 ± 31% of the low-voltage area. An EGM duration >160 ms covered an area of 6.7 ± 8.0 cm2, comprising 12 ± 11% of the low-voltage area. Areas at which the EGM duration was >200 ms covered 2.7 ± 1.7 cm2, comprising 5 ± 4% of the low-voltage area.
In 1 of 23 patients, the maximal EGM width was <160 ms. In 3 of 23 patients the maximal EGM duration was <200 ms. Of the EGMs that had a duration >133 ms, 95% had a low-voltage amplitude (598 of 630 EGMs; mean amplitude: 0.34 ± 0.33 mV); 96% of EGMs >160 ms had a low amplitude (342 of 358 EGMs; mean amplitude: 0.3 ± 0.28 mV); and 97% of EGMs with a width >200 ms had a low-voltage amplitude (178 of 183 EGMs; mean amplitude: 0.28 ± 0.26 mV).
Infarct age and mapping data
There was a significant positive correlation between the infarct age and the duration of the broadest endocardial EGM detected during sinus rhythm mapping (r = 0.84; p < 0.0001) (Table 1, Fig. 5). There also was a strong correlation between infarct age and maximal delay of local endocardial activation after the QRS complex during sinus rhythm (mean 318 ± 91 ms; range 143 to 494 ms) in the peri-infarct zone (r = 0.80; p < 0.001). The isoelectric segment separating the ventricular EGM from isolated potentials was longer (mean 170 ± 93 ms; range 29 to 344 ms) the older the infarction (r = 0.77; p < 0.001). The duration index at an isthmus site obtained during sinus rhythm was 0.9 ± 0.12.
The maximal EGM duration per patient did not correlate with the ejection fraction (r = 0.13; p = 0.7) or with baseline QRS duration of the patients (r = 0.07, p = 0.8). The maximal EGM duration was similar in the 16 patients who did undergo revascularization after infarction as compared with those that did not (315 ± 92 ms vs. 367 ± 113 ms; p = 0.3). There was no correlation between EGM duration at an isthmus site and VT cycle length for the VTs in which an isthmus could be identified (r = 0.1, p = 0.7). The number of MIs did not correlate with the maximal EGM duration (r = 0.17; p = 0.5). Furthermore, the maximal EGM duration was similar, whether or not patients were on amiodarone at the time of the mapping procedure (349 ± 97 ms vs. 361 ± 105 ms; p = 0.8). The areas of EGM duration >133 ms, >160 ms, and >200 ms did not correlate with infarct age (r = 0.1, = 0.03, = 0.1; p = 0.7, = 0.9, = 0.9, respectively). Infarct location did not correlate with infarct age or maximal EGM duration. In patients with prior anterior infarctions as the first infarction, the maximal EGM width was 316 ± 21 ms vs. 313 ± 114 ms in patients with prior inferior wall infarction (p = 0.9). Infarct age was 18 ± 5 years in patients with prior anterior infarction at the time of the mapping procedure and 13 ± 11 years in patients with prior inferior infarction (p = 0.18).
Contiguous EGMs with a duration >200 ms contained isolated potentials in 86% of cases (Tables 2 and 3,⇓⇓Fig. 4). Sites with isolated potentials were significantly broader and had significantly smaller amplitude compared with sites with normal, abnormal, or fractionated EGMs. The vast majority of sites with an isolated potential were within the low-voltage area. All but one of the isthmus site EGMs fell into the area containing contiguous EGMs with a duration of >200 ms. Among the 23 patients, 39 contiguous areas bounded by EGMs with a duration of >200 ms could be identified (2.3 ± 1.5 per patient).
In patients with post-infarction VT, there was a significant correlation between infarct age and the maximal duration of endocardial EGMs in the peri-infarct zone during sinus rhythm. Critical isthmus sites in the VT re-entry circuit were likely to be located at sites with the latest activation during sinus rhythm. Most critical isthmus sites displayed EGMs that had a duration >200 ms and isolated potentials. These findings are consistent with post-infarction remodeling that results in progressive electrophysiological effects in the anatomical substrate for VT. Furthermore, isolated potentials may identify surviving muscle bundles within the infarct scar that can be localized and targeted for ablation during sinus rhythm.
Postinfarct remodeling and EGM characteristics
Experiments with superfused subacute, healing, and healed canine infarct hearts have demonstrated a relationship between infarct age and EGM characteristics (3). This study demonstrates that the same type of relationship occurs in postinfarction patients who develop VT.
The anatomical basis of fractionated EGMs has been described in human postinfarct papillary muscles that have been harvested after cardiac transplantation. The higher the fractionation of recorded EGMs, the more that surviving myocardial muscle tracts separated by collagenous tissue were identified in the recording area (2). In the canine postinfarction model of Gardner et al. (3), isolated potentials were related to two separated myocardial strands that each gave rise to a single biphasic waveform, the muscle strands being separated by fibrous tissue. Furthermore, there was, over time, a gradual increase of nonmuscular tissue mainly composed of fibrous tissue when controls were compared with subacute, healing, and healed infarcts (3). This was most pronounced in areas where fractionated EGMs were recorded. In the present clinical study, all of the patients had a remote MI. The relationship between infarct age and EGM width in the peri-infarct zone suggests that there may be an ongoing process of collagenous fiber deposition for many years after infarction, resulting in progressive separation of individual muscle bundles and a gradual increase in maximal EGM width. The areas bounded by broad EGMs, however, did not correlate with infarct age and probably depend on variables other than only time.
Postinfarct remodeling and the substrate of VT
The infarct scar is a dynamic tissue with continued collagen turnover for many years after the initial infarction (8). Continued collagen deposition is considered to be one of the major consequences of ventricular remodeling in patients with ischemic cardiomyopathy (9). There is evidence that prevention and attenuation of LV remodeling decreases the occurrence of ventricular arrhythmias and/or sudden death (10,11).
Although our data do not directly prove the relationship between remodeling and ventricular arrhythmias, they support a possible role of the remodeling process in creating the substrate for re-entrant VT. The disruption of side-to-side connections between surviving myocardial strands secondary to fibrosis results in slowing of conduction due to a zig-zag course of activation (2); a milieu that promotes re-entry is thereby created. Consistent with this process, this study showed that critical isthmus sites of VT re-entry circuits were identified at areas showing a high EGM duration index.
A relationship between areas with isolated potentials and ventricular arrhythmias has been proposed in earlier studies (12). This relationship was confirmed in this study. All of the critical isthmus sites that were identified displayed an isolated potential during sinus rhythm and relatively broad EGMs. Histologic analysis of sites with isolated potentials has demonstrated the presence of surviving myocardial bundles surrounded by collagenous tissue (3), making isolated potentials an EGM correlate of the remodeling process.
Wilber et al. (13) reported that in the Multicenter Automatic Defibrillator Implantation Trial (MADIT II), the mortality risk increased as a function of time after MI. Our study provides a possible explanation for an increase in potentially lethal arrhythmias in a time-dependent manner after an MI. Postinfarction remodeling and collagen fiber deposition is a time-dependent process that eventually leads to the extent of complex anisotropy necessary for post-infarction VT. We cannot exclude a time-dependent effect of other factors on the appearance of postinfarction VT, including remodeling with compensatory hypertrophy (14,15), mechanical stretch (11,16), and alterations in the autonomous nervous system (17,18), for example. The latter factors most likely could initiate VT secondary to abnormal automaticity or triggered activity, but would be less likely to set the stage for re-entry as the only mechanism.
This study has several limitations. First, the study was a cross-sectional analysis. A longitudinal analysis with serial mapping procedures over several years would be the preferred method for analyzing the relationship between infarct age and EGM characteristics. Second, although there was no correlation between EGM characteristics and the number of recurrent MIs, the additional infarctions also might have affected the remodeling process.
A third limitation is that the width of the EGM at sites with isolated potentials included both the isolated potential and the ventricular EGMs. It is possible that these EGMs, at times, included both far-field and near-field components. In analyzing broad, complex EGMs that are often composed of multiple high-frequency deflections with isolated potentials separated by long isoelectric intervals, we chose to define them as single EGMs. Because of the potential limitations of this definition of maximal EGM width, however, additional measurements (such as the length of the isoelectric interval between the ventricular EGM and isolated potentials, and the activation delay between the onset of the QRS complex and isolated potentials) were performed. These other measures of EGM width also correlated with infarct age, thereby validating the relationship between maximal EGM width and infarct age.
This study did not include a control group of post-infarction patients without inducible monomorphic VT. Therefore, whether the EGM characteristics found in this study are specific to the substrate of re-entrant VT is unknown.
The LV reconstructions were only partial, because not all possible endocardial sites were sampled. Furthermore, although the data point density was highest in areas of scar, it is possible that points with even later EGMs might have been missed.
Lastly, it is possible that changes in the treatment of acute MI, over time, might have had an influence on the remodeling process and served as a confounding variable.
In this study of patients with post-infarction VT, there was a significant correlation between infarct age and EGM width. Postinfarction ventricular remodeling with continued collagenous fiber deposition most likely accounted for these time-dependent changes in peri-infarct endocardial EGM characteristics. Critical isthmus sites for post-infarction VT are composed of the broadest EGMs mainly displaying isolated potentials with an EGM width of >200 ms. Isolated potentials identify surviving myocardial bundles within the infarct scar and may be useful in identifying the critical substrate of post-infarction VT during sinus rhythm.
Drs. Morady and Oral are consultants for Biosense Webster Inc. (Diamond Bar, California).
- Abbreviations and Acronyms
- left ventricle/ventricular
- myocardial infarction
- right ventricular
- ventricular tachycardia
- Received October 11, 2004.
- Revision received January 6, 2005.
- Accepted January 25, 2005.
- American College of Cardiology Foundation
- Jugdutt B.I.,
- Amy R.W.
- de Bakker J.M.,
- van Capelle F.J.,
- Janse M.J.,
- et al.
- Gardner P.I.,
- Ursell P.C.,
- Fenoglio J.J. Jr..,
- Wit A.L.
- Josephson M.
- Callans D.J.,
- Ren J.F.,
- Michele J.,
- Marchlinski F.E.,
- Dillon S.M.
- Bogun F.,
- Bahu M.,
- Knight B.,
- et al.
- Bogun F.,
- Bahu M.,
- Knight B.,
- et al.
- Beltrami C.A.,
- Finato N.,
- Rocco M.,
- et al.
- Sogaard P.,
- Gotzsche C.O.,
- Ravkilde J.,
- Norgaard A.,
- Thygesen K.
- Klein H.,
- Karp R.B.,
- Kouchoukos N.T.,
- Zorn G.L. Jr..,
- James T.N.,
- Waldo A.L.
- Wilber D.J.,
- Zareba W.,
- Hall W.J.,
- et al.
- Chevalier B.,
- Heudes D.,
- Heymes C.,
- et al.
- White H.D.,
- Norris R.M.,
- Brown M.A.,
- Brandt P.W.,
- Whitlock R.M.,
- Wild C.J.
- Sabbah H.N.,
- Goldberg A.D.,
- Schoels W.,
- et al.