- Email: [email protected]
Electrocardiographic identification of prior myocardial infarction during right ventricular pacing — Effect of septal versus apical pacing
International Journal of Cardiology 177 (2014) 977–981
Contents lists available at ScienceDirect
International Journal of Cardiology journal homepage: www.elsevier.com/locate/ijcard
Electrocardiographic identification of prior myocardial infarction during right ventricular pacing — Effect of septal versus apical pacing Stylianos Tzeis a,⁎,1, George Andrikopoulos a,1, Stefan Asbach b,1, Verena Semmler c,1, Carsten Lennerz c,1, Ulrich Solzbach d,1, Hrvoje Vrazic e,1, Axel Kloppe f,1, Norbert Klein g,1, Sokratis Pastromas a,1, Jürgen Biermann b,1, Christof Kolb c,1, on behalf of the SPICE study investigators a
Henry Dunant Hospital, Department of Cardiology, Athens, Greece Heart Centre Freiburg University, Department of Cardiology and Angiology I, Freiburg, Germany c Deutsches Herzzentrum München, Klinik für Herz- und Kreislauferkrankungen, Abteilung für Elektrophysiologie, Faculty of Medicine, Technische Universität München, Munich, Germany d Ostalbklinikum, Abteilung für Innere Medizin II, Aalen, Germany e Dubrava University Hospital, Department of Internal Medicine, Division of Cardiology, University of Zagreb, Zagreb, Croatia f Klinikum Lüdenscheid, Abteilung für Innere Medizin III, Lüdenscheid, Germany & Berufsgenossenschaftliches Universitätsklinikum Bergmannsheil, Medizinische Klinik II, Ruhr Universität Bochum, Bochum, Germany g Universitätsklinikum Leipzig, Abteilung für Kardiologie und Angiologie, Leipzig, Germany b
a r t i c l e
i n f o
Article history: Received 12 August 2014 Received in revised form 24 September 2014 Accepted 28 September 2014 Available online 5 October 2014 Keywords: Electrocardiography Pacing Myocardial infarction Cabrera sign Septal site
a b s t r a c t Background: Electrocardiographic (ECG) identification of prior myocardial infarction (MI) during right ventricular (RV) pacing is of clinical importance. Proposed ECG criteria have been evaluated only during apical pacing. We evaluated the effect of pacing site on the predictive performance of ECG signs of prior MI. Methods: The present study is a secondary analysis of a prospective, multicenter study which randomized recipients of an implantable cardioverter defibrillator to an apical versus septal RV lead positioning. ECGs of patients with or without prior MI were analyzed for the presence of the following criteria: Cabrera sign, Chapman sign, QR pattern in leads I, aVL, V5 or V6, QR in inferior leads and notching in the descending slope of the QRS complex in inferior leads. Results: The MI group included 89 patients (55.1% apically paced), while 99 patients had no prior MI (50.5% apically paced). In the total population, the Cabrera sign presented the highest specificity (97%) and diagnostic accuracy (62.2%), with a sensitivity of 23.6%. The Cabrera sign was the only significant predictor of a prior MI [OR = 9.9, (95%CI:2.8–34.5), p b 0.001], among all ECG markers. Pacing site did not significantly influence the sensitivity and specificity of the Cabrera sign for detection of prior MI. Conclusions: In our study, the Cabrera sign was the only ECG marker that predicted the presence of prior MI during ventricular paced rhythm. Septal RV lead positioning did not affect the predictive performance of the Cabrera sign. © 2014 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Electrocardiographic (ECG) documentation of prior myocardial infarction (MI) is challenging in the setting of ventricular pacing. The confounding effect of ventricular paced rhythm is multifactorial, including abnormal depolarization and repolarization patterns, fusion beats resulting in a pseudoinfarction pattern, presence of retrograde P waves and distortion of initial ECG forces during unipolar pacing [1]. However, the identification of previously unrecognized MI is important,
⁎ Corresponding author at: Henry Dunant Hospital, 107 Mesogion Av., 115 26, Athens, Greece. Tel.: +30 2106972000; fax: +30 2106972200. E-mail address: [email protected] (S. Tzeis). 1 This author takes responsibility for all aspects of the reliability and freedom from bias of the data presented and their discussed interpretation.
http://dx.doi.org/10.1016/j.ijcard.2014.09.187 0167-5273/© 2014 Elsevier Ireland Ltd. All rights reserved.
especially among asymptomatic patients or diabetics, since silent MIs account for 9–37% of all non-fatal MIs with a considerable impact on mortality [2–4]. In order to address this need, several ECG criteria have been proposed for the documentation of previous MI in patients with right ventricular (RV) paced rhythm. The most widely studied markers are the Cabrera sign, the Chapman sign, qR pattern in leads I, aVL, V5 or V6, qR pattern in inferior leads and notching of the descending limb of the QRS complex in the inferior leads [5–7]. Reported findings on the sensitivity and specificity of ECG manifestations of prior MI during pacing are conflicting and have been acquired only in patients paced at the RV apex[8–11]. Therefore, there is lack of data on the ECG diagnosis of MI for other pacing sites than the RV apex. The need to ascertain the predictive value of the existing ECG signs in the context of non-apical RV pacing stems from the popularity
978
S. Tzeis et al. / International Journal of Cardiology 177 (2014) 977–981
of alternate pacing sites and the accumulating evidence on the detrimental effect of apical pacing [12]. In the present study we sought to evaluate the diagnostic utility of several ECG signs of prior MI during RV pacing in patients randomized to an apical versus septal RV lead location. 2. Methods 2.1. Study design The present study is a secondary analysis of electrocardiographic data collected for the ECG substudy of the SPICE study. The SPICE study was a prospective, randomized, multicenter, controlled study which evaluated the feasibility, safety and efficacy of implantable cardioverter defibrillator (ICD) lead positioning in the mid-septum as compared to the RV apex. The methodology and the main endpoints of the study have been previously published [13,14]. The ECG substudy was specifically designed to investigate the ECG characteristics of septal pacing in the setting of a randomized, prospective trial, where the lead positioning was documented fluoroscopically in both right and left anterior oblique projections and was reviewed by an independent adjudication committee blinded to the randomized site. The study was approved by the review committee of all participating institutions. Informed consent was obtained from each patient. The study protocol conforms to the ethical guidelines of the 1975 Declaration of Helsinki as reflected in a priori approval by the human research committee of all participating institutions. The following surface 12-lead ECGs were recorded: (a) intrinsic ECG before implantation and (b) RV only paced ECG (DDD mode with shortest programmable atrioventricular delay or VVI mode at a rate faster than the spontaneous rate) to ensure recording of fully paced complexes and avoid fusion.
3. Results 3.1. Study population The study population included a total of 188 patients. The MI group included 89 patients (67.2 ± 9.5 years, 86.5% males, 55.1% apical lead positioning), while the non-MI group included 99 patients (62.5 ± 14.1 years, 71.7% males, 50.5% apical lead location). A significantly lower percentage of MI patients as compared to patients without prior MI received a biventricular defibrillator (24.7% vs 42.4%, respectively, p = 0.01) and had left bundle branch block (LBBB) during intrinsic rhythm (19.3% vs 36.5%, respectively, p = 0.01). In the MI group, 44 (49.4%) patients had an anterior/anterolateral MI, 20 (22.5%) patients had an inferior MI, while in 25 (28.1%) patients the ECG localization of MI was not feasible due to coexistent LBBB (17 patients) or inability to satisfy the posed diagnostic criteria of q wave prior MI (8 patients). The proportion of apically paced patients was similar between low EF (≤ 35%) as compared to high EF (N35%) groups (52.5% vs 51.9%, p = 1.0). Furthermore, the proportion of apically paced patients was similar between NYHA I–II as compared to NYHA III–IV categories (52.2% vs 52.6%, p = 1.0). 3.2. Electrocardiographic signs of prior MI in the total patient population
2.2. Electrocardiogram evaluation All paper-recorded ECGs were scanned and subsequently used for on-screen measurements of selected parameters. The evaluation was performed by two reviewers blinded to the location of the RV lead and to the patients' characteristics. In case of disagreement in ECG interpretation the opinion of a third reviewer, unaware of patients' allocation, was taken into consideration. The presence of the following ECG signs, proposed to be associated with the presence of old MI during ventricular paced rhythm, was evaluated in each of the RV paced ECGs (Fig. 1): (a) Cabrera sign: defined as the presence of a prominent, shelf-like or downward notching, of N0.03 second duration, at the ascending limb of the S wave [6]. In our study, we ruled in the presence of a Cabrera sign if two precordial leads displayed a positive sign. (b) Chapman sign: defined as the presence of a prominent notch (N0.03 second duration) in the ascending limb of the R wave in leads I or aVL or V6 [7]. (c) QR pattern (including qR, Qr and QR but not QS) in leads I, aVL, V5 or V6. (d) notching in the descending slope of the QRS complex in the inferior leads. (e) QR pattern (including qR, Qr and QR but not QS) in leads II, III, or aVF.
2.3. Study groups The MI group included ischemic heart disease patients with a positive history of MI and/or presence of Q waves or QS complexes pathognomonic of a prior MI based on the 2012 consensus definition of MI [15]. The non-MI group included ICD recipients without ischemic heart disease, which was ruled out by coronary angiography and/or myocardial perfusion scintigraphy. Classification of MI location was based on the presence of pathognomonic Q waves in two or more contiguous leads (anterior/anterolateral MI pertaining to leads I, aVL, V1–V6; inferior MI pertaining to leads II, III, aVF).
The sensitivity, specificity and total diagnostic accuracy of the evaluated ECG signs in the total patient population are presented in Table 1. A QR pattern in inferior leads was not found in any patient, neither with nor without old MI. In univariate logistic regression analysis, we examined the association of each ECG sign with the existence of an old MI. A positive Cabrera sign was a significant predictor of the presence of prior MI ([OR = 9.9, (95%CI: 2.8–34.5), p b 0.001]). A positive Cabrera sign in only one precordial lead was also shown to significantly predict the presence of an old MI [OR = 3.4, (95%CI: 1.6–7.0), p b 0.001]. No other ECG sign was demonstrated to be a significant predictor of a prior MI. 3.3. Electrocardiographic signs of prior MI in different pacing sites (Figs. 2 and 3) A significantly higher percentage of patients paced from the RV septum as compared to the RV apex presented a QR pattern in lead I (58.4% vs 24.2%, respectively, p b 0.001) and a QR pattern in lead aVL (55.7% vs 4.0%, respectively, p b 0.001). No statistically significant difference was observed between the apical and the septal pacing site in the occurrence rate of the Cabrera sign in two precordial leads (14.1% vs 11.2%, respectively), Cabrera sign in one precordial lead (24.2% vs 21.3%, respectively), Chapman sign (13.1% vs 14.6%, respectively) and notching in the downward slope of the inferior leads (9.1% vs 10.1%, respectively).
2.4. Statistical methods Continuous data are presented as mean ± standard deviation, while categorical data as count and percentages. Pearson's chi-square test for categorical variables and Student's t-test for continuous variables were employed to compare parameters between groups of interest. In order to evaluate the inherent statistical validity of the tested dichotomous ECG criteria we calculated the sensitivity, specificity and overall diagnostic accuracy, using the following formulas: (a) sensitivity = true positive / (true positive and false negative), (b) specificity = true negative / (true negative and false positive), (c) overall diagnostic accuracy = (true positive and true negative) / (true positive and false positive and true negative and false negative). Logistic regression analysis was used to evaluate the association between a continuous or categorical explanatory variable and a single categorical response variable. Adjusted odds ratios (OR) and confidence intervals (CIs) were calculated from the logistic regression variable estimates. The fit of the binary logistic regression model was assessed using the Hosmer and Lemeshow goodness-of-fit test. All tests were considered to be significant at the 0.05 level. Statistical analyses were performed with SPSS statistical software (version 16.0, SPSS, Chicago, IL, U.S.A.).
Table 1 Sensitivity, specificity and total diagnostic accuracy of the evaluated electrocardiographic criteria of prior MI in the total patient population. ECG sign
Sensitivity
Specificity
Total diagnostic accuracy
Cabreraa Single Cabrerab Chapman QR pattern in lead I QR pattern in lead aVL QR pattern in any of I, aVL, V5, V6 Notching in downward slope of QRS in II, III, aVF
23.6% 33.7% 15.7% 43.8% 30.3% 50.6% 12.4%
97.0% 86.9% 87.9% 62.6% 73.5% 55.6% 92.9%
62.2% 61.7% 53.7% 53.7% 52.9% 53.2% 54.8%
a b
Defined as presence of positive Cabrera sign in two precordial leads (V1–V6). Defined as presence of positive Cabrera sign in one precordial lead (V1–V6).
S. Tzeis et al. / International Journal of Cardiology 177 (2014) 977–981
979
lateral leads exhibited a significantly higher sensitivity for detecting the presence of prior MI during ventricular pacing from the RV septum as compared to the RV apex with a significant reduction in the respective specificity (Table 2). There was no significant difference in the sensitivity and specificity of the Cabrera sign, Chapman sign and notching in the downward slope of the QRS in inferior leads in the apical as compared to the septal pacing site. 3.4. Factors associated with positive Cabrera sign among MI patients In the MI group, we assessed whether baseline ejection fraction and/or location of MI (anterior/anterolateral vs inferior) influence the occurrence of a positive Cabrera sign. Baseline ejection fraction, accounted as a continuous variable, was not shown to significantly predict a Cabrera sign among MI patients. However, the location of MI was a significant predictor of a Cabrera sign. Patients with an anterior/anterolateral MI had a significantly higher likelihood of a positive Cabrera sign as compared to those with an inferior MI [OR = 5.1, (95%CI: 1.1–25.1), p b 0.05]. 4. Discussion
Fig. 1. Examples of the studied electrocardiographic criteria in paced patients (paper speed 50 mm/s): (A) positive Cabrera sign in precordial leads: prominent notch (shelflike or downward) notching of N0.03 second duration of the ascending limb of the S wave in more than 1 precordial lead (V3 and V4 in the present case), (B) Chapman sign: prominent notch (N0.03 second duration) in the ascending limb of the R wave in leads I, aVL, V5 or V6 (lead I in the illustrated ECG), (C) QR pattern (including qR, Qr and QR but not QS) in leads I, aVL, V5 or V6 (qR in lead I in the illustrated ECG), (D) notching in the descending slope of the QRS complex in the inferior leads (lead III in the illustrated ECG).
The sensitivity and specificity of the examined electrocardiographic signs of prior MI during ventricular pacing from the right apex and the septum are presented in Table 2. The presence of QR pattern in the
In our study population, the Cabrera sign was the only significant predictor of the presence of a prior MI, among the examined ECG signs, with a high specificity but a low sensitivity. Septal pacing did not significantly influence the predictive performance of the Cabrera sign as compared to apical pacing. The existing evidence on the predictive value of ECG criteria of old MI during ventricular pacing is contradictory and has been derived only from patients paced at the RV apex. Kochiadakis et al. reported that the Chapman and Cabrera signs had the highest overall diagnostic accuracy, with high sensitivity (91.1% and 86.6% respectively), but low specificity (57.6% for both signs) for diagnosing chronic MI in a study of 71 patients subjected to temporary pacing during coronary angiography [9]. However, the external validity of these findings is limited, since several patient categories frequently encountered in clinical practice were excluded from the study population. In addition, accurate identification of Cabrera sign was not ensured based on the ECG criteria
Fig. 2. Twelve lead electrocardiogram of a patient with an old anterior infarction during ventricular pacing from the RV apex (VVI mode — 110 pulses per minute — paper speed: 50 mm/s). The ECG displays the following ECG signs of old MI: (a) Cabrera sign in the precordial leads V2 to V4, (b) Cabrera sign in the inferior lead II, (c) qR pattern in lead I and (d) notching at the downward slope of the QRS complex at the inferior leads II, III and aVF. Black arrows point to the P waves which are dissociated from the paced ventricular QRS complexes. The presence of ventriculoatrial dissociation excludes potential confounding effect of retrograde P waves on the morphology of the paced QRS complex.
980
S. Tzeis et al. / International Journal of Cardiology 177 (2014) 977–981
Fig. 3. Twelve lead electrocardiogram of a patient with an old anterior infarction during RV septal pacing (VVI mode — 110 pulses per minute — paper speed: 50 mm/s). The ECG displays the following signs of old MI: (a) Cabrera sign in the precordial leads V2 to V6, (b) qR pattern in lead I and aVL (c) notching at the downward slope of the QRS complex at lead II.
illustrated in the respective figure [9]. Furthermore, all Q waves N0.03 second in inferior leads were considered as a positive ECG marker, although only Qr or qR patterns but not QS complexes are of value in the diagnosis of previous MI in paced ventricular rhythm. Contrary to those results, Barold et al. have reported that the Cabrera sign has a specificity of almost 100%, if notching is properly defined, with a low sensitivity of 25% to 50% according to the size of the MI [5, 6]. In an older study, Kindwall et al., reported a sensitivity of 45% and a specificity of 92% for the Cabrera sign with an increased sensitivity if left axis deviation was present [11]. More recently, Théraulaz et al. in a retrospective study of unselected patients with permanent pacemakers reported a moderate sensitivity for Cabrera sign (63.6%) and poor sensitivity for other diagnostic criteria (9.1% to 40.9%), with a relatively high specificity for all ECG criteria ranging from 81.6% to 100% [10]. The discrepancies in the results of the previous relevant studies are attributed to several limitations including inconsistencies in the methods used to confirm the presence of chronic MI, ascertainment
Table 2 Sensitivity and specificity of the evaluated electrocardiographic criteria of prior MI in relation to pacing site. ECG sign
Cabreraa Single Cabrerab Chapman QR in lead I QR in lead aVL QR in any of I, aVL, V5, V6 Notching in downward slope of QRS in II, III, aVF a b
Sensitivity
Specificity
Apex
Septum p value Apex
Septum p value
28.6% 38.8% 14.3% 26.5% 4.1% 28.6% 12.2%
17.5% 27.5% 17.5% 65.0% 62.5% 77.5% 12.5%
93.9% 83.7% 87.8% 46.9% 50.0% 32.7% 91.8%
0.22 0.26 0.68 b0.001 b0.001 b0.001 0.97
100% 90.0% 88.0% 78.0% 96.0% 78.0% 94.0%
0.08 0.35 0.97 0.002 b0.001 b0.001 0.68
Defined as presence of positive Cabrera sign in two precordial leads (V1–V6). Defined as presence of positive Cabrera sign in one precordial lead (V1–V6).
bias due to lack of blinding during data analysis [9], selected patient populations and heterogeneous synthesis or absence of control groups. Another major caveat is the presence of inter-observer variability in the identification of the ECG criteria and the inconsistency in their definition, mainly that of the Cabrera sign. Inconsistent definition of the Cabrera sign pertains to the number of leads (one versus two) and the duration of the notching needed to rule in its presence. Furthermore, inaccuracy in the recognition of the Cabrera sign is partly due to subjective differentiation between slight slurring in the upstroke of the ascending limb, occasionally encountered in otherwise healthy individuals, versus the prominent, shelf-like or downward notching in the case of true positive Cabrera sign [5,8]. In the present study we aimed to address several caveats that have been commented previously. The ECG evaluation was blinded and the studied ECG criteria were strictly defined to avoid overdiagnosis. In order to avoid inconsistency in case definition, we included in the MI group patients with history of ischemic heart disease, and either a positive history of MI or pathognomonic ECG changes as recommended by the recently issued universal definition of MI [15]. The control group included patients without ischemic heart disease which was ruled out by coronary angiography and/or myocardial perfusion scintigraphy. The inclusion of patients with impaired systolic function of nonischemic cause in the control group allowed us to evaluate the likelihood of false positive ECG signs potentially associated with the presence of scar or fibrosis among patients with non-ischemic dilated cardiomyopathy. Finally, each RV lead position was assessed blindly by an adjudication committee using as fluoroscopic criteria not only the left but also the right anterior oblique projections, since the latter has been shown to be helpful in ensuring proper mid-septal lead placement [16]. Based on our results, the most useful ECG criterion of old MI during RV pacing is the Cabrera sign, since it was the single predictor of prior MI during ventricular paced rhythm, with a sensitivity of 23.6%, a specificity of 97% and a diagnostic accuracy of 62.2%. Our findings are in
S. Tzeis et al. / International Journal of Cardiology 177 (2014) 977–981
concordance with those of Barold et al. [5,8]. Furthermore, when Théraulaz et al. analyzed their data “with a very strict definition for Cabrera sign” they reported a sensitivity of 34%, a specificity of 97% and an overall accuracy of 63% [10]. These values are almost identical to our findings. Therefore, it could be concluded that a strictly defined Cabrera sign (prominent shelf-like or downward notching of N0.03 second duration, present in two leads) has a low sensitivity with a high specificity. As we also showed, a less strict assessment leads to overdiagnosis with an increased sensitivity but reduced specificity. Furthermore, based on our results, even when the Cabrera sign was considered positive when present in only one precordial lead it still remained a significant predictor of prior MI. We also demonstrated that ejection fraction did not significantly influence the presence of a positive Cabrera sign. Therefore, if we consider ejection fraction as a surrogate endpoint of the size of MI, we could presume that the latter does not influence the likelihood of a Cabrera sign. However, this is a hypothesis-generating finding and future studies with objective quantification of MI size are needed. In our study, the QR pattern in lateral leads, the Chapman sign and the notching in the downward slope of the inferior leads demonstrated a poor sensitivity and total diagnostic accuracy, without significantly predicting the presence of a prior MI. We could not identify the presence of a QR pattern in inferior leads in any of our patients with old MI, which further supports the very low occurrence rate of this ECG manifestation [17]. These results are in accordance with those of Théraulaz et al. [10]. To our knowledge, this is the first study to evaluate ECG signs of MI in patients subjected to non-apical RV pacing. Our data showed that septal pacing does not influence the sensitivity and specificity of the Cabrera sign, Chapman sign and notching at the descending slope of the inferior leads. However, the septal pacing site significantly increased the sensitivity but decreased the specificity of a QR pattern in lateral leads. Our data are in accord with those of Barold et al. who reported that displacement of the catheter away from the RV apex may produce QR ventricular complexes in leads I and aVL in the absence of an MI [8]. The absence of a predictive value of a QR pattern in lateral leads for detection of prior MI in our total patient population, combined with the increased falsepositive rates during septal pacing, limits the clinical value of this marker for identification of prior MI during ventricular paced rhythm. Several limitations of the study should be commented on. Although patients were included in the study prospectively, the overall population is not large. Lead location was verified by a blinded adjudication committee based on fluoroscopic criteria, though without the implementation of any imaging tool, such as echocardiography or computerized tomography scanning. In addition, although the recently issued universal definition of MI was used for identification of MI patients, no additional myocardial imaging was implemented for confirmation of case definition. 5. Conclusions The Cabrera sign exhibits a high specificity, albeit low sensitivity, for identification of prior MI during ventricular pacing, either from the apex or from the RV septum. No other ECG sign was shown to predict the presence of prior MI. Therefore, physicians should be familiar with the Cabrera sign, which although rare to encounter, can be used for the identification of an old MI when assessing a 12-lead ECG during ventricular pacing. The clinical value of this marker is evident in everyday practice, especially when screening patients at increased risk of silent MI, such as diabetics. Financial support — conflicts of interest The SPICE trial was financially supported by St. Jude Medical Germany, Eschborn, Germany. S.T. has received travel support and lecture honoraria from St. Jude Medical and Medtronic, G.A. has received travel support and lecture
981
honoraria from St. Jude Medical and Medtronic. S.A. has received lecture honorary/travel support from Biotronik, Boston Scientific, Medtronic and St. Jude Medical; he performs/has performed clinical studies supported by Biotronik, Medtronic, Sorin and St. Jude Medical. V.S. has received travel support from Sorin and St. Jude Medical. C.L. has received travel support from St. Jude Medical. U.S. none. H.V. has received lecture honorary/travel support from Boston Scientific, Medtronic, Biotronik; he performs/has performed clinical studies supported by Biotronik. He was supported by an educational fellowship of the European Heart Rhythm Association. A.K. has received lecture honorary/travel support from Biotronik, Boston Scientific, Medtronic and St. Jude Medical; he is a consultant to Medtronik and St. Jude Medical and performs/has performed clinical studies supported by Biotronik, Medtronic, Boston Scientific and St. Jude Medical. N.K. has received lecture honorary/travel support from Boston Scientific, Sorin and St. Jude Medical; he is a consultant to Boston Scientific and St. Jude Medical and performs/has performed clinical studies supported by Biotronik, Boston Scientific, Medtronic, Sorin and St. Jude Medical. S.P. none. J.B. has received lecture honorary/travel support from Biotronik, Boston Scientific, Medtronic and St. Jude Medical; he performs/has performed clinical studies supported by Biotronik, Medtronic, Sorin and St. Jude Medical. C.K. has received lecture honorary/travel support from Biotronik, Boston Scientific, Medtronic, St. Jude Medical and Sorin; he is a consultant to Biotronik, Boston Scientific and Sorin and performs/has performed clinical studies supported by Biotronik, Medtronic, Sorin and St. Jude Medical.
References [1] Barold SS, Levine PA, Ovsyshchert IE. The paced 12-lead electrocardiogram should no longer be neglected in pacemaker follow-up. Pacing Clin Electrophysiol 2001; 24:1455–8. [2] Burgess DC, Hunt D, Li L, Zannino D, Williamson E, Davis TM, et al. Incidence and predictors of silent MI in type 2 diabetes and the effect of fenofibrate: an analysis from the Fenofibrate Intervention and Event Lowering in Diabetes (FIELD) study. Eur Heart J 2010;31:92–9. [3] Toma M, Fu Y, Ezekowitz JA, McAlister FA, Westerhout CM, Granger CB, et al. Does silent MI add prognostic value in ST-elevation MI? Insights from the Assessment of Pexelizumab in Acute MI (APEX-AMI) trial. Am Heart J 2010;160:671–7. [4] Davis TM, Coleman RL, Holman RR, UKPDS Group. Prognostic significance of silent MI in newly diagnosed type 2 diabetes mellitus: United Kingdom Prospective Diabetes Study (UKPDS) 79. Circulation 2013;127:980–7. [5] Barold SS, Falkoff MD, Ong LS, Heinle RA. Electrocardiographic diagnosis of MI during ventricular pacing. Cardiol Clin 1987;5:403–17. [6] Cabrera E, Friedland C. Wave of ventricular activation in left branch block with infarct; new electrocardiographic sign. Arch Inst Cardiol Mex 1953;23:441–60. [7] Chapman MG, Pearce ML. Electrocardiographic diagnosis of MI in the presence of left bundle-branch block. Circulation 1957;16:558–71. [8] Barold SS, Herweg B, Curtis AB. Electrocardiographic diagnosis of MI and ischemia during cardiac pacing. Cardiol Clin 2006;24:387–99. [9] Kochiadakis GE, Kaleboubas MD, Igoumenidis NE, Skalidis EI, Simantirakis EN, Chrysostomakis SI, et al. Electrocardiographic appearance of old MI in paced patients. Pacing Clin Electrophysiol 2002;25:1061–5. [10] Theraulaz D, Zimmermann M, Meiltz A, Bloch A. Value of the 12-lead resting electrocardiogram for the diagnosis of previous MI in paced patients. J Electrocardiol 2007; 40:496–503. [11] Kindwall KE, Brown JP, Josephson ME. Predictive accuracy of criteria for chronic MI in pacing-induced left bundle branch block. Am J Cardiol 1986;57:1255–60. [12] Shimony A, Eisenberg MJ, Filion KB, Amit G. Beneficial effects of RV non-apical vs. apical pacing: a systematic review and meta-analysis of randomized-controlled trials. Europace 2012;14:81–91. [13] Kolb C, Tzeis S, Andrikopoulos G, Asbach S, Lemke B, Hansen C, et al. Rationale and design of the SPICE study-septal positioning of ventricular ICD electrodes. J Interv Card Electrophysiol 2011;31:247–54. [14] Kolb C, Solzbach U, Biermann J, Semmler V, Kloppe A, Klein N, et al. Safety of midseptal electrode placement in implantable cardioverter defibrillator recipients — results of the SPICE (septal positioning of ventricular ICD electrodes) study. Int J Cardiol 2014;174(3):713–20. [15] Thygesen K, Alpert JS, Jaffe AS, Simoons ML, Chaitman BR, White HD. Joint ESC/ACCF/ AHA/WHF Task Force for the Universal Definition of Myocardial Infarction. Third universal definition of MI. Eur Heart J 2012;33:2551–67. [16] Osmancik P, Stros P, Herman D, Curila K, Petr R. The insufficiency of left anterior oblique and the usefulness of right anterior oblique projection for correct localization of a computer tomography-verified right ventricular lead into the midseptum. Circ Arrhythm Electrophysiol 2013;6:719–25. [17] Barold SS. Diagnosis of inferior MI based on the paced QRS complex. J Electrocardiol 2007;40:161–3.
Contents lists available at ScienceDirect
International Journal of Cardiology journal homepage: www.elsevier.com/locate/ijcard
Electrocardiographic identification of prior myocardial infarction during right ventricular pacing — Effect of septal versus apical pacing Stylianos Tzeis a,⁎,1, George Andrikopoulos a,1, Stefan Asbach b,1, Verena Semmler c,1, Carsten Lennerz c,1, Ulrich Solzbach d,1, Hrvoje Vrazic e,1, Axel Kloppe f,1, Norbert Klein g,1, Sokratis Pastromas a,1, Jürgen Biermann b,1, Christof Kolb c,1, on behalf of the SPICE study investigators a
Henry Dunant Hospital, Department of Cardiology, Athens, Greece Heart Centre Freiburg University, Department of Cardiology and Angiology I, Freiburg, Germany c Deutsches Herzzentrum München, Klinik für Herz- und Kreislauferkrankungen, Abteilung für Elektrophysiologie, Faculty of Medicine, Technische Universität München, Munich, Germany d Ostalbklinikum, Abteilung für Innere Medizin II, Aalen, Germany e Dubrava University Hospital, Department of Internal Medicine, Division of Cardiology, University of Zagreb, Zagreb, Croatia f Klinikum Lüdenscheid, Abteilung für Innere Medizin III, Lüdenscheid, Germany & Berufsgenossenschaftliches Universitätsklinikum Bergmannsheil, Medizinische Klinik II, Ruhr Universität Bochum, Bochum, Germany g Universitätsklinikum Leipzig, Abteilung für Kardiologie und Angiologie, Leipzig, Germany b
a r t i c l e
i n f o
Article history: Received 12 August 2014 Received in revised form 24 September 2014 Accepted 28 September 2014 Available online 5 October 2014 Keywords: Electrocardiography Pacing Myocardial infarction Cabrera sign Septal site
a b s t r a c t Background: Electrocardiographic (ECG) identification of prior myocardial infarction (MI) during right ventricular (RV) pacing is of clinical importance. Proposed ECG criteria have been evaluated only during apical pacing. We evaluated the effect of pacing site on the predictive performance of ECG signs of prior MI. Methods: The present study is a secondary analysis of a prospective, multicenter study which randomized recipients of an implantable cardioverter defibrillator to an apical versus septal RV lead positioning. ECGs of patients with or without prior MI were analyzed for the presence of the following criteria: Cabrera sign, Chapman sign, QR pattern in leads I, aVL, V5 or V6, QR in inferior leads and notching in the descending slope of the QRS complex in inferior leads. Results: The MI group included 89 patients (55.1% apically paced), while 99 patients had no prior MI (50.5% apically paced). In the total population, the Cabrera sign presented the highest specificity (97%) and diagnostic accuracy (62.2%), with a sensitivity of 23.6%. The Cabrera sign was the only significant predictor of a prior MI [OR = 9.9, (95%CI:2.8–34.5), p b 0.001], among all ECG markers. Pacing site did not significantly influence the sensitivity and specificity of the Cabrera sign for detection of prior MI. Conclusions: In our study, the Cabrera sign was the only ECG marker that predicted the presence of prior MI during ventricular paced rhythm. Septal RV lead positioning did not affect the predictive performance of the Cabrera sign. © 2014 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Electrocardiographic (ECG) documentation of prior myocardial infarction (MI) is challenging in the setting of ventricular pacing. The confounding effect of ventricular paced rhythm is multifactorial, including abnormal depolarization and repolarization patterns, fusion beats resulting in a pseudoinfarction pattern, presence of retrograde P waves and distortion of initial ECG forces during unipolar pacing [1]. However, the identification of previously unrecognized MI is important,
⁎ Corresponding author at: Henry Dunant Hospital, 107 Mesogion Av., 115 26, Athens, Greece. Tel.: +30 2106972000; fax: +30 2106972200. E-mail address: [email protected] (S. Tzeis). 1 This author takes responsibility for all aspects of the reliability and freedom from bias of the data presented and their discussed interpretation.
http://dx.doi.org/10.1016/j.ijcard.2014.09.187 0167-5273/© 2014 Elsevier Ireland Ltd. All rights reserved.
especially among asymptomatic patients or diabetics, since silent MIs account for 9–37% of all non-fatal MIs with a considerable impact on mortality [2–4]. In order to address this need, several ECG criteria have been proposed for the documentation of previous MI in patients with right ventricular (RV) paced rhythm. The most widely studied markers are the Cabrera sign, the Chapman sign, qR pattern in leads I, aVL, V5 or V6, qR pattern in inferior leads and notching of the descending limb of the QRS complex in the inferior leads [5–7]. Reported findings on the sensitivity and specificity of ECG manifestations of prior MI during pacing are conflicting and have been acquired only in patients paced at the RV apex[8–11]. Therefore, there is lack of data on the ECG diagnosis of MI for other pacing sites than the RV apex. The need to ascertain the predictive value of the existing ECG signs in the context of non-apical RV pacing stems from the popularity
978
S. Tzeis et al. / International Journal of Cardiology 177 (2014) 977–981
of alternate pacing sites and the accumulating evidence on the detrimental effect of apical pacing [12]. In the present study we sought to evaluate the diagnostic utility of several ECG signs of prior MI during RV pacing in patients randomized to an apical versus septal RV lead location. 2. Methods 2.1. Study design The present study is a secondary analysis of electrocardiographic data collected for the ECG substudy of the SPICE study. The SPICE study was a prospective, randomized, multicenter, controlled study which evaluated the feasibility, safety and efficacy of implantable cardioverter defibrillator (ICD) lead positioning in the mid-septum as compared to the RV apex. The methodology and the main endpoints of the study have been previously published [13,14]. The ECG substudy was specifically designed to investigate the ECG characteristics of septal pacing in the setting of a randomized, prospective trial, where the lead positioning was documented fluoroscopically in both right and left anterior oblique projections and was reviewed by an independent adjudication committee blinded to the randomized site. The study was approved by the review committee of all participating institutions. Informed consent was obtained from each patient. The study protocol conforms to the ethical guidelines of the 1975 Declaration of Helsinki as reflected in a priori approval by the human research committee of all participating institutions. The following surface 12-lead ECGs were recorded: (a) intrinsic ECG before implantation and (b) RV only paced ECG (DDD mode with shortest programmable atrioventricular delay or VVI mode at a rate faster than the spontaneous rate) to ensure recording of fully paced complexes and avoid fusion.
3. Results 3.1. Study population The study population included a total of 188 patients. The MI group included 89 patients (67.2 ± 9.5 years, 86.5% males, 55.1% apical lead positioning), while the non-MI group included 99 patients (62.5 ± 14.1 years, 71.7% males, 50.5% apical lead location). A significantly lower percentage of MI patients as compared to patients without prior MI received a biventricular defibrillator (24.7% vs 42.4%, respectively, p = 0.01) and had left bundle branch block (LBBB) during intrinsic rhythm (19.3% vs 36.5%, respectively, p = 0.01). In the MI group, 44 (49.4%) patients had an anterior/anterolateral MI, 20 (22.5%) patients had an inferior MI, while in 25 (28.1%) patients the ECG localization of MI was not feasible due to coexistent LBBB (17 patients) or inability to satisfy the posed diagnostic criteria of q wave prior MI (8 patients). The proportion of apically paced patients was similar between low EF (≤ 35%) as compared to high EF (N35%) groups (52.5% vs 51.9%, p = 1.0). Furthermore, the proportion of apically paced patients was similar between NYHA I–II as compared to NYHA III–IV categories (52.2% vs 52.6%, p = 1.0). 3.2. Electrocardiographic signs of prior MI in the total patient population
2.2. Electrocardiogram evaluation All paper-recorded ECGs were scanned and subsequently used for on-screen measurements of selected parameters. The evaluation was performed by two reviewers blinded to the location of the RV lead and to the patients' characteristics. In case of disagreement in ECG interpretation the opinion of a third reviewer, unaware of patients' allocation, was taken into consideration. The presence of the following ECG signs, proposed to be associated with the presence of old MI during ventricular paced rhythm, was evaluated in each of the RV paced ECGs (Fig. 1): (a) Cabrera sign: defined as the presence of a prominent, shelf-like or downward notching, of N0.03 second duration, at the ascending limb of the S wave [6]. In our study, we ruled in the presence of a Cabrera sign if two precordial leads displayed a positive sign. (b) Chapman sign: defined as the presence of a prominent notch (N0.03 second duration) in the ascending limb of the R wave in leads I or aVL or V6 [7]. (c) QR pattern (including qR, Qr and QR but not QS) in leads I, aVL, V5 or V6. (d) notching in the descending slope of the QRS complex in the inferior leads. (e) QR pattern (including qR, Qr and QR but not QS) in leads II, III, or aVF.
2.3. Study groups The MI group included ischemic heart disease patients with a positive history of MI and/or presence of Q waves or QS complexes pathognomonic of a prior MI based on the 2012 consensus definition of MI [15]. The non-MI group included ICD recipients without ischemic heart disease, which was ruled out by coronary angiography and/or myocardial perfusion scintigraphy. Classification of MI location was based on the presence of pathognomonic Q waves in two or more contiguous leads (anterior/anterolateral MI pertaining to leads I, aVL, V1–V6; inferior MI pertaining to leads II, III, aVF).
The sensitivity, specificity and total diagnostic accuracy of the evaluated ECG signs in the total patient population are presented in Table 1. A QR pattern in inferior leads was not found in any patient, neither with nor without old MI. In univariate logistic regression analysis, we examined the association of each ECG sign with the existence of an old MI. A positive Cabrera sign was a significant predictor of the presence of prior MI ([OR = 9.9, (95%CI: 2.8–34.5), p b 0.001]). A positive Cabrera sign in only one precordial lead was also shown to significantly predict the presence of an old MI [OR = 3.4, (95%CI: 1.6–7.0), p b 0.001]. No other ECG sign was demonstrated to be a significant predictor of a prior MI. 3.3. Electrocardiographic signs of prior MI in different pacing sites (Figs. 2 and 3) A significantly higher percentage of patients paced from the RV septum as compared to the RV apex presented a QR pattern in lead I (58.4% vs 24.2%, respectively, p b 0.001) and a QR pattern in lead aVL (55.7% vs 4.0%, respectively, p b 0.001). No statistically significant difference was observed between the apical and the septal pacing site in the occurrence rate of the Cabrera sign in two precordial leads (14.1% vs 11.2%, respectively), Cabrera sign in one precordial lead (24.2% vs 21.3%, respectively), Chapman sign (13.1% vs 14.6%, respectively) and notching in the downward slope of the inferior leads (9.1% vs 10.1%, respectively).
2.4. Statistical methods Continuous data are presented as mean ± standard deviation, while categorical data as count and percentages. Pearson's chi-square test for categorical variables and Student's t-test for continuous variables were employed to compare parameters between groups of interest. In order to evaluate the inherent statistical validity of the tested dichotomous ECG criteria we calculated the sensitivity, specificity and overall diagnostic accuracy, using the following formulas: (a) sensitivity = true positive / (true positive and false negative), (b) specificity = true negative / (true negative and false positive), (c) overall diagnostic accuracy = (true positive and true negative) / (true positive and false positive and true negative and false negative). Logistic regression analysis was used to evaluate the association between a continuous or categorical explanatory variable and a single categorical response variable. Adjusted odds ratios (OR) and confidence intervals (CIs) were calculated from the logistic regression variable estimates. The fit of the binary logistic regression model was assessed using the Hosmer and Lemeshow goodness-of-fit test. All tests were considered to be significant at the 0.05 level. Statistical analyses were performed with SPSS statistical software (version 16.0, SPSS, Chicago, IL, U.S.A.).
Table 1 Sensitivity, specificity and total diagnostic accuracy of the evaluated electrocardiographic criteria of prior MI in the total patient population. ECG sign
Sensitivity
Specificity
Total diagnostic accuracy
Cabreraa Single Cabrerab Chapman QR pattern in lead I QR pattern in lead aVL QR pattern in any of I, aVL, V5, V6 Notching in downward slope of QRS in II, III, aVF
23.6% 33.7% 15.7% 43.8% 30.3% 50.6% 12.4%
97.0% 86.9% 87.9% 62.6% 73.5% 55.6% 92.9%
62.2% 61.7% 53.7% 53.7% 52.9% 53.2% 54.8%
a b
Defined as presence of positive Cabrera sign in two precordial leads (V1–V6). Defined as presence of positive Cabrera sign in one precordial lead (V1–V6).
S. Tzeis et al. / International Journal of Cardiology 177 (2014) 977–981
979
lateral leads exhibited a significantly higher sensitivity for detecting the presence of prior MI during ventricular pacing from the RV septum as compared to the RV apex with a significant reduction in the respective specificity (Table 2). There was no significant difference in the sensitivity and specificity of the Cabrera sign, Chapman sign and notching in the downward slope of the QRS in inferior leads in the apical as compared to the septal pacing site. 3.4. Factors associated with positive Cabrera sign among MI patients In the MI group, we assessed whether baseline ejection fraction and/or location of MI (anterior/anterolateral vs inferior) influence the occurrence of a positive Cabrera sign. Baseline ejection fraction, accounted as a continuous variable, was not shown to significantly predict a Cabrera sign among MI patients. However, the location of MI was a significant predictor of a Cabrera sign. Patients with an anterior/anterolateral MI had a significantly higher likelihood of a positive Cabrera sign as compared to those with an inferior MI [OR = 5.1, (95%CI: 1.1–25.1), p b 0.05]. 4. Discussion
Fig. 1. Examples of the studied electrocardiographic criteria in paced patients (paper speed 50 mm/s): (A) positive Cabrera sign in precordial leads: prominent notch (shelflike or downward) notching of N0.03 second duration of the ascending limb of the S wave in more than 1 precordial lead (V3 and V4 in the present case), (B) Chapman sign: prominent notch (N0.03 second duration) in the ascending limb of the R wave in leads I, aVL, V5 or V6 (lead I in the illustrated ECG), (C) QR pattern (including qR, Qr and QR but not QS) in leads I, aVL, V5 or V6 (qR in lead I in the illustrated ECG), (D) notching in the descending slope of the QRS complex in the inferior leads (lead III in the illustrated ECG).
The sensitivity and specificity of the examined electrocardiographic signs of prior MI during ventricular pacing from the right apex and the septum are presented in Table 2. The presence of QR pattern in the
In our study population, the Cabrera sign was the only significant predictor of the presence of a prior MI, among the examined ECG signs, with a high specificity but a low sensitivity. Septal pacing did not significantly influence the predictive performance of the Cabrera sign as compared to apical pacing. The existing evidence on the predictive value of ECG criteria of old MI during ventricular pacing is contradictory and has been derived only from patients paced at the RV apex. Kochiadakis et al. reported that the Chapman and Cabrera signs had the highest overall diagnostic accuracy, with high sensitivity (91.1% and 86.6% respectively), but low specificity (57.6% for both signs) for diagnosing chronic MI in a study of 71 patients subjected to temporary pacing during coronary angiography [9]. However, the external validity of these findings is limited, since several patient categories frequently encountered in clinical practice were excluded from the study population. In addition, accurate identification of Cabrera sign was not ensured based on the ECG criteria
Fig. 2. Twelve lead electrocardiogram of a patient with an old anterior infarction during ventricular pacing from the RV apex (VVI mode — 110 pulses per minute — paper speed: 50 mm/s). The ECG displays the following ECG signs of old MI: (a) Cabrera sign in the precordial leads V2 to V4, (b) Cabrera sign in the inferior lead II, (c) qR pattern in lead I and (d) notching at the downward slope of the QRS complex at the inferior leads II, III and aVF. Black arrows point to the P waves which are dissociated from the paced ventricular QRS complexes. The presence of ventriculoatrial dissociation excludes potential confounding effect of retrograde P waves on the morphology of the paced QRS complex.
980
S. Tzeis et al. / International Journal of Cardiology 177 (2014) 977–981
Fig. 3. Twelve lead electrocardiogram of a patient with an old anterior infarction during RV septal pacing (VVI mode — 110 pulses per minute — paper speed: 50 mm/s). The ECG displays the following signs of old MI: (a) Cabrera sign in the precordial leads V2 to V6, (b) qR pattern in lead I and aVL (c) notching at the downward slope of the QRS complex at lead II.
illustrated in the respective figure [9]. Furthermore, all Q waves N0.03 second in inferior leads were considered as a positive ECG marker, although only Qr or qR patterns but not QS complexes are of value in the diagnosis of previous MI in paced ventricular rhythm. Contrary to those results, Barold et al. have reported that the Cabrera sign has a specificity of almost 100%, if notching is properly defined, with a low sensitivity of 25% to 50% according to the size of the MI [5, 6]. In an older study, Kindwall et al., reported a sensitivity of 45% and a specificity of 92% for the Cabrera sign with an increased sensitivity if left axis deviation was present [11]. More recently, Théraulaz et al. in a retrospective study of unselected patients with permanent pacemakers reported a moderate sensitivity for Cabrera sign (63.6%) and poor sensitivity for other diagnostic criteria (9.1% to 40.9%), with a relatively high specificity for all ECG criteria ranging from 81.6% to 100% [10]. The discrepancies in the results of the previous relevant studies are attributed to several limitations including inconsistencies in the methods used to confirm the presence of chronic MI, ascertainment
Table 2 Sensitivity and specificity of the evaluated electrocardiographic criteria of prior MI in relation to pacing site. ECG sign
Cabreraa Single Cabrerab Chapman QR in lead I QR in lead aVL QR in any of I, aVL, V5, V6 Notching in downward slope of QRS in II, III, aVF a b
Sensitivity
Specificity
Apex
Septum p value Apex
Septum p value
28.6% 38.8% 14.3% 26.5% 4.1% 28.6% 12.2%
17.5% 27.5% 17.5% 65.0% 62.5% 77.5% 12.5%
93.9% 83.7% 87.8% 46.9% 50.0% 32.7% 91.8%
0.22 0.26 0.68 b0.001 b0.001 b0.001 0.97
100% 90.0% 88.0% 78.0% 96.0% 78.0% 94.0%
0.08 0.35 0.97 0.002 b0.001 b0.001 0.68
Defined as presence of positive Cabrera sign in two precordial leads (V1–V6). Defined as presence of positive Cabrera sign in one precordial lead (V1–V6).
bias due to lack of blinding during data analysis [9], selected patient populations and heterogeneous synthesis or absence of control groups. Another major caveat is the presence of inter-observer variability in the identification of the ECG criteria and the inconsistency in their definition, mainly that of the Cabrera sign. Inconsistent definition of the Cabrera sign pertains to the number of leads (one versus two) and the duration of the notching needed to rule in its presence. Furthermore, inaccuracy in the recognition of the Cabrera sign is partly due to subjective differentiation between slight slurring in the upstroke of the ascending limb, occasionally encountered in otherwise healthy individuals, versus the prominent, shelf-like or downward notching in the case of true positive Cabrera sign [5,8]. In the present study we aimed to address several caveats that have been commented previously. The ECG evaluation was blinded and the studied ECG criteria were strictly defined to avoid overdiagnosis. In order to avoid inconsistency in case definition, we included in the MI group patients with history of ischemic heart disease, and either a positive history of MI or pathognomonic ECG changes as recommended by the recently issued universal definition of MI [15]. The control group included patients without ischemic heart disease which was ruled out by coronary angiography and/or myocardial perfusion scintigraphy. The inclusion of patients with impaired systolic function of nonischemic cause in the control group allowed us to evaluate the likelihood of false positive ECG signs potentially associated with the presence of scar or fibrosis among patients with non-ischemic dilated cardiomyopathy. Finally, each RV lead position was assessed blindly by an adjudication committee using as fluoroscopic criteria not only the left but also the right anterior oblique projections, since the latter has been shown to be helpful in ensuring proper mid-septal lead placement [16]. Based on our results, the most useful ECG criterion of old MI during RV pacing is the Cabrera sign, since it was the single predictor of prior MI during ventricular paced rhythm, with a sensitivity of 23.6%, a specificity of 97% and a diagnostic accuracy of 62.2%. Our findings are in
S. Tzeis et al. / International Journal of Cardiology 177 (2014) 977–981
concordance with those of Barold et al. [5,8]. Furthermore, when Théraulaz et al. analyzed their data “with a very strict definition for Cabrera sign” they reported a sensitivity of 34%, a specificity of 97% and an overall accuracy of 63% [10]. These values are almost identical to our findings. Therefore, it could be concluded that a strictly defined Cabrera sign (prominent shelf-like or downward notching of N0.03 second duration, present in two leads) has a low sensitivity with a high specificity. As we also showed, a less strict assessment leads to overdiagnosis with an increased sensitivity but reduced specificity. Furthermore, based on our results, even when the Cabrera sign was considered positive when present in only one precordial lead it still remained a significant predictor of prior MI. We also demonstrated that ejection fraction did not significantly influence the presence of a positive Cabrera sign. Therefore, if we consider ejection fraction as a surrogate endpoint of the size of MI, we could presume that the latter does not influence the likelihood of a Cabrera sign. However, this is a hypothesis-generating finding and future studies with objective quantification of MI size are needed. In our study, the QR pattern in lateral leads, the Chapman sign and the notching in the downward slope of the inferior leads demonstrated a poor sensitivity and total diagnostic accuracy, without significantly predicting the presence of a prior MI. We could not identify the presence of a QR pattern in inferior leads in any of our patients with old MI, which further supports the very low occurrence rate of this ECG manifestation [17]. These results are in accordance with those of Théraulaz et al. [10]. To our knowledge, this is the first study to evaluate ECG signs of MI in patients subjected to non-apical RV pacing. Our data showed that septal pacing does not influence the sensitivity and specificity of the Cabrera sign, Chapman sign and notching at the descending slope of the inferior leads. However, the septal pacing site significantly increased the sensitivity but decreased the specificity of a QR pattern in lateral leads. Our data are in accord with those of Barold et al. who reported that displacement of the catheter away from the RV apex may produce QR ventricular complexes in leads I and aVL in the absence of an MI [8]. The absence of a predictive value of a QR pattern in lateral leads for detection of prior MI in our total patient population, combined with the increased falsepositive rates during septal pacing, limits the clinical value of this marker for identification of prior MI during ventricular paced rhythm. Several limitations of the study should be commented on. Although patients were included in the study prospectively, the overall population is not large. Lead location was verified by a blinded adjudication committee based on fluoroscopic criteria, though without the implementation of any imaging tool, such as echocardiography or computerized tomography scanning. In addition, although the recently issued universal definition of MI was used for identification of MI patients, no additional myocardial imaging was implemented for confirmation of case definition. 5. Conclusions The Cabrera sign exhibits a high specificity, albeit low sensitivity, for identification of prior MI during ventricular pacing, either from the apex or from the RV septum. No other ECG sign was shown to predict the presence of prior MI. Therefore, physicians should be familiar with the Cabrera sign, which although rare to encounter, can be used for the identification of an old MI when assessing a 12-lead ECG during ventricular pacing. The clinical value of this marker is evident in everyday practice, especially when screening patients at increased risk of silent MI, such as diabetics. Financial support — conflicts of interest The SPICE trial was financially supported by St. Jude Medical Germany, Eschborn, Germany. S.T. has received travel support and lecture honoraria from St. Jude Medical and Medtronic, G.A. has received travel support and lecture
981
honoraria from St. Jude Medical and Medtronic. S.A. has received lecture honorary/travel support from Biotronik, Boston Scientific, Medtronic and St. Jude Medical; he performs/has performed clinical studies supported by Biotronik, Medtronic, Sorin and St. Jude Medical. V.S. has received travel support from Sorin and St. Jude Medical. C.L. has received travel support from St. Jude Medical. U.S. none. H.V. has received lecture honorary/travel support from Boston Scientific, Medtronic, Biotronik; he performs/has performed clinical studies supported by Biotronik. He was supported by an educational fellowship of the European Heart Rhythm Association. A.K. has received lecture honorary/travel support from Biotronik, Boston Scientific, Medtronic and St. Jude Medical; he is a consultant to Medtronik and St. Jude Medical and performs/has performed clinical studies supported by Biotronik, Medtronic, Boston Scientific and St. Jude Medical. N.K. has received lecture honorary/travel support from Boston Scientific, Sorin and St. Jude Medical; he is a consultant to Boston Scientific and St. Jude Medical and performs/has performed clinical studies supported by Biotronik, Boston Scientific, Medtronic, Sorin and St. Jude Medical. S.P. none. J.B. has received lecture honorary/travel support from Biotronik, Boston Scientific, Medtronic and St. Jude Medical; he performs/has performed clinical studies supported by Biotronik, Medtronic, Sorin and St. Jude Medical. C.K. has received lecture honorary/travel support from Biotronik, Boston Scientific, Medtronic, St. Jude Medical and Sorin; he is a consultant to Biotronik, Boston Scientific and Sorin and performs/has performed clinical studies supported by Biotronik, Medtronic, Sorin and St. Jude Medical.
References [1] Barold SS, Levine PA, Ovsyshchert IE. The paced 12-lead electrocardiogram should no longer be neglected in pacemaker follow-up. Pacing Clin Electrophysiol 2001; 24:1455–8. [2] Burgess DC, Hunt D, Li L, Zannino D, Williamson E, Davis TM, et al. Incidence and predictors of silent MI in type 2 diabetes and the effect of fenofibrate: an analysis from the Fenofibrate Intervention and Event Lowering in Diabetes (FIELD) study. Eur Heart J 2010;31:92–9. [3] Toma M, Fu Y, Ezekowitz JA, McAlister FA, Westerhout CM, Granger CB, et al. Does silent MI add prognostic value in ST-elevation MI? Insights from the Assessment of Pexelizumab in Acute MI (APEX-AMI) trial. Am Heart J 2010;160:671–7. [4] Davis TM, Coleman RL, Holman RR, UKPDS Group. Prognostic significance of silent MI in newly diagnosed type 2 diabetes mellitus: United Kingdom Prospective Diabetes Study (UKPDS) 79. Circulation 2013;127:980–7. [5] Barold SS, Falkoff MD, Ong LS, Heinle RA. Electrocardiographic diagnosis of MI during ventricular pacing. Cardiol Clin 1987;5:403–17. [6] Cabrera E, Friedland C. Wave of ventricular activation in left branch block with infarct; new electrocardiographic sign. Arch Inst Cardiol Mex 1953;23:441–60. [7] Chapman MG, Pearce ML. Electrocardiographic diagnosis of MI in the presence of left bundle-branch block. Circulation 1957;16:558–71. [8] Barold SS, Herweg B, Curtis AB. Electrocardiographic diagnosis of MI and ischemia during cardiac pacing. Cardiol Clin 2006;24:387–99. [9] Kochiadakis GE, Kaleboubas MD, Igoumenidis NE, Skalidis EI, Simantirakis EN, Chrysostomakis SI, et al. Electrocardiographic appearance of old MI in paced patients. Pacing Clin Electrophysiol 2002;25:1061–5. [10] Theraulaz D, Zimmermann M, Meiltz A, Bloch A. Value of the 12-lead resting electrocardiogram for the diagnosis of previous MI in paced patients. J Electrocardiol 2007; 40:496–503. [11] Kindwall KE, Brown JP, Josephson ME. Predictive accuracy of criteria for chronic MI in pacing-induced left bundle branch block. Am J Cardiol 1986;57:1255–60. [12] Shimony A, Eisenberg MJ, Filion KB, Amit G. Beneficial effects of RV non-apical vs. apical pacing: a systematic review and meta-analysis of randomized-controlled trials. Europace 2012;14:81–91. [13] Kolb C, Tzeis S, Andrikopoulos G, Asbach S, Lemke B, Hansen C, et al. Rationale and design of the SPICE study-septal positioning of ventricular ICD electrodes. J Interv Card Electrophysiol 2011;31:247–54. [14] Kolb C, Solzbach U, Biermann J, Semmler V, Kloppe A, Klein N, et al. Safety of midseptal electrode placement in implantable cardioverter defibrillator recipients — results of the SPICE (septal positioning of ventricular ICD electrodes) study. Int J Cardiol 2014;174(3):713–20. [15] Thygesen K, Alpert JS, Jaffe AS, Simoons ML, Chaitman BR, White HD. Joint ESC/ACCF/ AHA/WHF Task Force for the Universal Definition of Myocardial Infarction. Third universal definition of MI. Eur Heart J 2012;33:2551–67. [16] Osmancik P, Stros P, Herman D, Curila K, Petr R. The insufficiency of left anterior oblique and the usefulness of right anterior oblique projection for correct localization of a computer tomography-verified right ventricular lead into the midseptum. Circ Arrhythm Electrophysiol 2013;6:719–25. [17] Barold SS. Diagnosis of inferior MI based on the paced QRS complex. J Electrocardiol 2007;40:161–3.