CLINICAL STUDY
QRST subtraction combined with a pacemap catalogue for the prediction of ectopy source by surface electrocardiogram in patients with paroxysmal atrial fibrillation
Kee-Joon Choi, MD*,
Dipen C. Shah, MD ,*,
Pierre Jais, MD*,
Meleze Hocini, MD*,
Laurent Macle, MD*,
Christophe Scavee, MD*,
Rukshen Weerasooriya, MD*,
Florence Raybaud, MD*,
Jacques Clementy, MD* and
Michel Haissaguerre, MD*
* Hôpital Cardiologique du Haut-Lévêque, Bordeaux-Pessac, France
Hôpital Cantonal Universitaire de Geneve, Geneva, Switzerland
Manuscript received March 5, 2002;
revised manuscript received July 2, 2002,
accepted August 19, 2002.
* Reprint requests and correspondence: Dr. Dipen C. Shah, Centre de Cardiologie, Hopital Cantonal Universitaire de Geneve, 24, Rue Micheli du Crest, CH 1211 Geneva 14, Switzerland.
Dipen.Shah{at}hcuge.ch
 |
Abstract
|
|---|
OBJECTIVES: This study evaluated the use of ectopic P-wave morphology to localize pulmonary vein (PV) and non-PV sources of atrial ectopics in patients with paroxysmal atrial fibrillation (PAF).
BACKGROUND: The vectorial information embodied in the morphology of ectopic P waves is concealed by overlying T waves.
METHODS: The P-wave morphology of 56 ectopics was prospectively analyzed in 44 patients with PAF (age, 52 ± 12 years; 36 male) by subtracting the adjacent QRST from the QRST-ectopic P-wave complex using custom-designed software. Subtraction fidelity was validated in 15 other patients (55 ± 19 years, 11 male) by comparing drive beats with simulated ectopics (S2 from the same site) unmasked by subtracting overlying QRST. An algorithm combined with PV pacemaps was used to predict PV sources. Subtracted ectopic P-wave morphologies after PV disconnection were compared with PV and non-PV site pacemaps. Localization was confirmed by mapping and successful ablation.
RESULTS: A  10-lead electrocardiogram (ECG) match was observed in 92% of 644 simulated ectopics (coupling intervals: 190 to 520 ms). In PAF patients, 37 spontaneous ectopics originated from the PV, while 19 were noted after PV disconnection. Using the P-wave algorithm alone, correct prediction of PV origin was achieved in 30/37 ectopics (81%). Combination with PV pacemaps allowed correct prediction in 34/37 (92%). After PV disconnection, ECG localization predicted successful ablation sites in 16/19 (84%).
CONCLUSIONS: Comparison of subtracted ectopic P waves with a pacemap catalogue provides a simple and accurate 12-lead ECG-based method for localization, which can facilitate ablation of arrhythmia triggers irrespective of origin from the PV or elsewhere.
|
Abbreviations and Acronyms
| | AF | | atrial fibrillation | | IVC | | inferior vena cava | | PAF | | paroxysmal atrial fibrillation | | PV | | pulmonary vein | | SVC | | superior vena cava |
|
Most episodes of paroxysmal atrial fibrillation (PAF) have been reported to be initiated by supraventricular ectopics, frequently originating from the pulmonary veins (PV) but also from other thoracic veins as well as the atria (1,2). While the morphology of atrial ectopics could allow localization and facilitate intracardiac mapping and ablation of PAF, the discrimination of atrial ectopic morphology is difficult because most ectopics are premature enough to coincide with the preceding T-wave (so-called "P on T" ectopics) and are, therefore, concealed by the higher voltage and long duration of ventricular repolarization. These atrial extrasystoles are coincident with but independent of ventricular repolarization and can be exposed by peeling away ventricular activity, in other words, by subtracting the adjacent QRST from the subsequent QRST-ectopic P-wave complex (3,4). We hypothesized that the ectopic P-wave morphology derived by QRST subtraction and compared with the corresponding pacemap for each vein in each patient is helpful in the localization of PV ectopy for rapid targeting of the arrhythmogenic PV and the localization of non-PV sources of ectopics after PV disconnection.
|
Methods
|
|---|
QRST subtraction.
QRST subtraction was performed by subtracting the adjacent (preferably the immediately preceding) QRST from the subsequent QRST-ectopic P-wave complex using a special software developed for this purpose (Bard Electrophysiology, Boston, Massachusetts) (Fig. 1) . Surface electrocardiogram (ECG) signals filtered through a bandwidth of 0.5 to 200 Hz and digitized at a frequency of 1 kHz were slaved in real time to another Bard Lab system Duo. A template from the onset of the QRS complex to the end of the T-wave (defined as its return to the T-P baseline) was manually selected by calipers. The template QRST complex was automatically aligned with the subsequent QRST-ectopic P-wave complex based on an algorithm that detected the onset of the QRS complex. A correlation coefficient was derived between the template QRS complex and that of the beat to be subtracted. If a coefficient lower than 90% was obtained and a prominent residual QRS complex was observed, an earlier QRS was used as a new template for subtraction. A QRS with a pattern suggestive of aberration of intraventricular conduction was not used as templates, and subtraction was also not performed in the presence of noisy tracings or significant baseline fluctuations. Only surface ECG signals were transferred to this Lab system Duo console, therefore, automatically blinding the operator to intracardiac signals.
View larger version (36K):
[in this window]
[in a new window]
|
Figure 1 Electrocardiogram shows one ectopic (*), which initiates atrial fibrillation (B). The QRST complex preceding ectopy was used as a template (A) for QRST subtraction from the subsequent QRST-ectopic P-wave complex (C). The QRS complexes of the template and that of the beat preceding the ectopy were compared by a correlation coefficient calculated for each lead (shown above each trace) and, in the example shown here, indicate a high degree of matching. The subtracted ectopic P-wave morphology is flat in lead I, broadly positive in V1, and broad, notched, high-amplitude (100 µV) in lead II with the amplitude ratio of lead III/II >0.8, suggesting that the left superior pulmonary vein (PV) is the ectopic source. During intracardiac mapping of this patient, reversal in activation sequence of atrial (arrow, closed head) and PV potentials (arrow, open head) was observed in the circumferential recording of left superior PV during ectopics (D). This finding confirms the ectopic source predicted by postsubtraction P-wave analysis. CSd = distal coronary sinus; map = mapping catheter in right superior PV; PV 1-2 until PV 10-1 = bipoles from the circumferential catheter in left superior PV.
|
|
Validation of QRST subtraction
The fidelity of QRST subtraction was validated in 15 patients (55 ± 19 [mean ± SD] years, male/female 11/4) undergoing catheter ablation. Twelve-lead ECGs were obtained during bipolar pacing from 7 ± 2 (range, 4 to 9) right and left atrial sites. P-on-T ectopics were simulated with an S1S2 protocol decremented from a coupling interval coinciding with the end of the QT interval down to the atrial refractory period. Both morphology and amplitude of the paced S1 beat during the drive cycle were compared at standard amplification in all 12 leads by two observers to the derived S2 beat unmasked by subtracting the adjacent QRST complex.
Individual pacemap catalogue
Pulmonary vein pacemapping was performed to obtain an individual P-wave catalogue for each vein in each patient for comparison with the spontaneously occurring subtracted P waves in order to compensate for interindividual anatomic variations and improve prediction accuracy. Pacing was performed distal to the PV-left atrial junction of each PV at an output equal to two times the threshold. The PV-left atrial junction was defined by the maximum change in diameter on selective PV angiography, and ablation was performed at this level in order to disconnect the vein. A slow enough pacing rate was chosen to enable complete separation of the P-wave from the ventricular repolarization of the preceding beat.
Prediction and confirmation of PV ectopic sources
In 37 paroxysmal atrial fibrillation (AF) patients (51.3 ± 11 years, male/female 31/6) with ectopics during sinus rhythm before PV disconnection, the ectopic P-wave morphology was prospectively analyzed by QRST subtraction. By analyzing the morphology of the postsubtraction ectopic P waves, we predicted the PV source based on an algorithm previously developed by our group (5). The P waves comprising the individual pacemap catalogue were then compared (at standard amplification of about 1 mV/10 mm), with postsubtraction ectopic P waves in 12 leads, by two experienced independent observers, and at least an 11/12-lead match was required to indicate congruence. During intracardiac mapping, the origin of ectopic foci was indicated by either a reversal of the activation sequence of atrial and PV potentials, or distal-to-proximal venous activation during ectopy or earliest activation during spontaneous initiation of AF (Fig. 1D). Elimination of ectopy by disconnection of the suspected culprit PV confirmed its origin.
Localization and confirmation of non-PV foci after PV disconnection
In 18 patients (54.3 ± 11.3 years, male/female 15/3) in whom ectopics or AF initiation was observed after disconnection of all PVs, ectopic P-wave morphologies were analyzed by QRST subtraction in conjunction with pacemapping from the four PVs. The ectopic foci were considered to be ostial (proximal to the ablation line) if the postsubtraction P-wave morphology matched a P wave in the individual PV pacemap catalogue (with 11/12 match). If the postsubtraction P-wave morphology differed from the pacemap catalogue of all PVs, the ectopy was considered to originate from a non-PV focus. Pacing was then performed from several sites of suspected non-PV foci (coronary sinus ostium, superior vena cava [SVC], inferior vena cava [IVC], left atrial appendage, left atrial roof between two superior PVs, and mid or distal coronary sinus) to compare with the subtracted P-wave morphology. Ablation was performed at the site of earliest activation, and ECG-based localization was compared with the successful ablation sites. Correct surface ECG localization was defined as an area within 1 cm of the successful ablation site but which did not transgress anatomical boundaries—such as the interatrial septum.
|
Results
|
|---|
Validation of QRST subtraction.
A total of 644 atrial extrastimuli were delivered at coupling intervals ranging from 190 to 520 ms and, therefore, coinciding with all phases of the QT interval. Comparison of the 12-lead morphology of subtracted S2 with paced S1 beats revealed that 82% of subtracted P waves matched the drive cycle P- wave in all 12 (67%) or 11 (another 15%) ECG leads. A 10/12-lead match was observed in 10%, and a 9/12-lead or lesser match in 6%. Surface ECG P-wave changes suggestive of aberrant atrial conduction were observed in 2% of extrastimuli, mostly at the shortest coupling intervals.
Prediction of PV ectopic sources
There were 41 ectopic P-wave morphologies in 37 patients. Thirty-seven of these ectopics originated from the PVs. Using the P-wave algorithm alone, correct prediction of the origin of PV ectopy was achieved in 30/37 ectopics (81%, 14/14 in right superior PV, 12/18 in left superior PV, 2/2 in left inferior PV, and 2/3 in right inferior PV) (Fig. 2). Ectopics from the left superior PV were erroneously ascribed to the right superior PV (n = 3) because of an unusually high-amplitude positive P-wave in lead I, or to the left inferior PV (n = 3) because of an unusually low-amplitude P-wave in lead II and III. Comparison with PV pacemapping allowed correct prediction in 34/37 (92%) ectopic morphologies with an 11/12 lead match in 24 cases and a 12/12 lead match in 10 cases (Fig. 3).
View larger version (37K):
[in this window]
[in a new window]
|
Figure 3 Ectopic P-wave morphology before (A) and after (B) QRST subtraction. The subtracted ectopic P-wave morphology shows prominent positive P-wave in lead I and high-amplitude (100 µV) P-wave in lead II, with the amplitude ratio of lead III/II <0.8. Though this suggests right superior PV as the source of ectopy, comparison with PV pacemapping from left superior pulmonary vein (PV) (C) and right superior PV (D) indicates a 12/12-lead match with left superior PV pacing. Ectopics were eliminated by the disconnection of left superior PV.
|
|
Localization of non-PV foci after PV disconnection
A total of 19 morphologies of ectopic P waves were observed in 18 patients: either just after PV ablation (n = 12) or during a subsequent ablation session (n = 7). Comparison of subtracted ectopic morphology with PV pacemapping localized four ectopic foci to the ostia of previously ablated PVs, and all were successfully ablated. Comparison with PV pacemapping indicated 15 foci of non-PV origin. Pacemapping from non-PV sites matched 12 of the subtracted ectopic P waves (left atrium in five, right atrial septum in one, SVC in four, IVC in two). Five left atrial sites were situated in the high anterior left atrium close to Bachmann’s bundle, the site between two right PVs, left atrial septum, and two in the low posterior left atrium (Fig. 4). One patient showed two different morphologies of ectopics, which originated from the septal and lateral side of the IVC. For ectopy originating from SVC, the postsubtraction P wave was taller than the P-wave in sinus rhythm in the inferior leads but had a similar morphology in V1. For IVC ectopics, a tall P-wave in lead I and negative P-wave in the inferior leads with deeper negativity in lead III than in lead II was observed (Fig. 5). Both localizations allowed successful ablation.
View larger version (34K):
[in this window]
[in a new window]
|
Figure 4 Ectopic P-wave morphology before (A) and after (B) QRST subtraction in a patient with frequent ectopics, sometimes in a bigeminal pattern, after pulmonary vein (PV) disconnection. Postsubtraction ectopic P-wave morphology shows biphasic pattern (negative-positive) in lead II, III, and avF, and broad positivity in V1, which is clearly dissimilar to pacemapping P-wave morphologies from PVs (D: right superior PV, E: right inferior PV, F: left superior PV, G: left inferior PV), suggesting a non-PV focus. Pacemapping from several left atrial sites demonstrated that ectopic morphology closely matched the pacemap from the low posterior left atrium between two inferior PVs (C). Successful ablation was performed at a site very close (within 5 mm) to this pacemapping site.
|
|
View larger version (45K):
[in this window]
[in a new window]
|
Figure 5 Ectopic P-wave morphology before (A) and after (B) QRST subtraction in cases with ectopics from superior vena cava (SVC) (a) and inferior vena cava (IVC) (b). Subtraction results showed a P-wave taller than sinus rhythm in inferior leads but with similar morphology in V1 for ectopy originating from SVC (a). For IVC ectopics, a tall P-wave in lead I and negative P waves in the inferior leads with deeper negativity in lead III than in lead II was observed (b). a’ and b’ are P-wave morphologies during pacemapping from the medial aspect of the SVC and of the IVC, respectively.
|
|
In three other cases, comparison with pacemapping suggested origin of the ectopic in the posterior left atrial wall in two patients, but the localization could not be confirmed by ablation. In one patient taking amiodarone, the subtracted morphology did not match the pacemapping result from any site including the successful ablation site, possibly because of aberrant intraatrial conduction at short coupling intervals.
In summary, ECG-based localization correlated with the successful ablation site in 16/19 (84%) of ectopics after PV disconnection.
|
Discussion
|
|---|
major findings.
To our knowledge, this is the first report of the use of QRST subtraction for the localization of ectopic sources in paroxysmal AF patients. A simple and quick software-based method of 12-lead QRST subtraction in conjunction with a catalogue of PV pacemapping allows rapid and accurate localization of PV ectopic foci to facilitate rapid targeting of arrhythmogenic PV. Furthermore, after disconnection of all PVs, the same technique is useful to rapidly detect and localize non-PV ectopics.
QRST subtraction and P-wave algorithm derived by PV pacing
QRST subtraction in order to eliminate obscuring ventricular activity from the surface ECG has been reported by several authors (6–8). This technique has been used to detect atrioventricular dissociation during wide QRS tachycardia (6) or to perform spectral analysis of AF cycle length (7,8). Recently, Sippensgroenewegen et al. (4) reported that the discrete isolation of ectopic atrial activity from the overlying T-wave was possible by an automatic 62-lead QRST subtraction method during sinus rhythm and programmed right atrial stimulation, and also in five patients with atrial premature beats or atrial tachycardia. Clinical application of a 12-lead QRST subtraction technique combined with a pacemap catalogue in patients with paroxysmal AF has not previously been evaluated with regard to the ability to predict ectopic sources.
In this study we preferentially used the immediately preceding QRST as a template. This choice of template allowed the best possible matching of beat-to-beat ventricular activation to minimize distortion of the derived atrial extrasystole. Compared with averaging several beats preceding the index beat, changes in the QRST complex by alterations in autonomic tone, respiratory effort, and patient movement could be reduced as well as allowing practically real time subtraction.
We validated this technique of QRST subtraction by confirming that simulated atrial extrasystolic beats concealed within the ST-T segment of preceding ventricular beats (produced by appropriately timed S2 beats) could be accurately unmasked by eliminating the overlying QRST. The subtracted beats were shown to be accurate as assessed by comparing the morphology of S1 beats with S2 (3).
From PV pacemapping results, Yamane et al. (5) proposed an algorithm using three criteria to distinguish right from left PV origin: 1) positive P-wave in lead aVL and the amplitude of P-wave in lead I 50 µV indicated right PV origin; 2) a notched P-wave in lead II as a predictor of left PV origin; and 3) the amplitude ratio of lead III/II and the duration of positivity in lead V1. In addition, superior PV could be distinguished from inferior according to the amplitude in lead II (100 µV).
Prediction of PV ectopic sources
Systematic disconnection of all PVs from the left atrium guided by a circumferential mapping catheter during sinus rhythm is presently being performed to treat AF (9,10). A manifestly arrhythmogenic PV is first targeted because frequent initiation of sustained AF prolongs the ablation procedure by disrupting mapping-guided ablation of left atrium-PV inputs. The prediction of an arrhythmogenic PV on the basis of surface ECG before endocardial mapping can allow rapid preferential targeting of dominantly arrhythmogenic PV, so that the ablation procedure can be continued in stable sinus rhythm.
Moreover, younger patients without organic heart disease have fewer ectopic initiating sources, and the technique described in this study may allow rapid noninvasive recognition of this subset of patients and ablation of the origin of their ectopy, particularly if there are no arrhythmias during intracardiac mapping.
In this study, correct prediction of the origin of PV ectopy was achieved in 81% of PV ectopics based only on the P-wave algorithm. When QRST subtraction was used in conjunction with pacemapping from the PV, 92% of PV ectopics were correctly localized.
Localization of non-PV foci after PV disconnection
Recurrence of paroxysmal AF after PV disconnection is a significant clinical problem. Nonpulmonary vein ectopic foci including ostia of previously ablated PV and atrial foci are frequent sources of recurrence (10). However, localization of non-PV foci is difficult and time consuming, particularly because of the limited catheter coverage of the left atrium and their capricious behavior with frequent sustained AF. In our study, ectopic P-wave morphology derived from QRST-wave subtraction and compared with pacemapping from the PVs and common sites of non-PV foci was helpful for detecting and localizing ectopics after PV disconnection in 84% of patients. This technique would also be relevant in order to distinguish recurrences from atrial as opposed to ostial or intra-PV foci and to recognize multisource ectopy based solely on ECG criteria (11,12).
Study limitations
Though a systematic evaluation of the accuracy of using correlation coefficients for QRS matching was not performed, the validation data confirmed that accurate derived (subtracted) P waves could be obtained using these methods. Individual variations in PV location and, also atrial anatomy limited the predictive accuracy for correct localization for ectopic PV foci based solely on the P-wave algorithm. Ectopics from the left superior PV were erroneously ascribed to the right superior PV or to the left inferior PV in six patients in our study. This limitation could be overcome or reduced by combination with pacemapping results from PVs. Pacemapped P-wave morphology from the site of origin of the ectopic focus may be different from the spontaneous ectopic morphology in some cases due to differences in coupling interval, especially in patients on antiarrhythmic drugs. Subtraction to derive the paced premature beat in a short-coupled S1S2 will be needed to overcome this limitation. Because the exact level of demarcation of PV from non-PV foci is undetermined, the distinction can most practically be made after initial PV isolation, as was done in this study. Furthermore, in this study, subtraction for ostial ectopic P waves confirmed the similarity of these P waves with intra-PV pacing, and though unipolar pacing and/or stimulation with a smaller bipole may improve results, there are likely to be inherent limits to the resolution of surface ECG localization. Anatomically placed and relatively large ablation lesions at sites indicated by surface ECG analysis may, therefore, be the only solution to eliminate capricious triggering arrhythmias.
Conclusions
This technique of comparing an unmasked ectopic P-wave with a catalogue of pacemaps from the commonest sites of atrial ectopic origin provides a simple and accurate 12-lead surface ECG-based method for localization, which can facilitate the curative ablation of these frequently unpredictable arrhythmia triggers irrespective of whether they originate from the PVs or elsewhere.
|
References
|
|---|
1. Haissaguerre M, Jais P, Shah DC, et al. Spontaneous initiation of atrial fibrillation by ectopic beats originating in the pulmonary veins. N Engl J Med. 1998;339:659–666[Abstract/Free Full Text]
2. Chen SA, Hsieh MH, Tai CT, et al. Initiation of atrial fibrillation by ectopic beats originating from the pulmonary veins. Circulation. 1999;100:1879–1886[Abstract/Free Full Text]
3. Shah DC, Haissaguerre M, Yamane T, et al. Unmasking the ECG morphology of short coupled atrial ectopics by adjacent QRST subtraction. (abstr)Pacing Clin Electrophysiol. 2001;24:651
4. Sippensgroenewegen A, Mlynash MD, Roithinger FX, et al. Electrocardiographic analysis of ectopic atrial activity obscured by ventricular repolarization: P-wave isolation using an automatic 62-lead QRST subtraction algorithm. J Cardiovasc Electrophysiol. 2001;12:780–790[Medline]
5. Yamane T, Shah DC, Peng JT, et al. Morphological characteristics of P wave during selective pulmonary vein pacing. J Am Coll Cardiol. 2001;38:1505–1510[Abstract/Free Full Text]
6. Slocum J, Byrom E, McCarthy L, et al. Computer detection of atrioventricular dissociation from surface electrocardiograms during wide QRS complex tachycardias. Circulation. 1985;72:1028–1036[Abstract/Free Full Text]
7. Holm M, Pehrson S, Ingemansson M, et al. Non-invasive assessment of the atrial cycle length during atrial fibrillation in man: introducing, validating and illustrating a new ECG method. Cardiovasc Res. 1998;38:69–81[Abstract/Free Full Text]
8. Bollmann A, Kanuru NK, MaTeague KK, et al. Frequency analysis of human atrial fibrillation using the surface electrogram and its response to ibutilide. Am J Cardiol. 1998;81:1439–1445[CrossRef][Medline]
9. Haissaguerre M, Jais P, Shah DC, et al. Electrophysiological end point for catheter ablation of atrial fibrillation initiated from multiple pulmonary venous foci. Circulation. 2000;101:1409–1417[Abstract/Free Full Text]
10. Haissaguerre M, Shah DC, Jais P, et al. Mapping-guided ablation of pulmonary veins to cure atrial fibrillation. Am J Cardiol. 2000;86(Suppl):9K–19K
11. Tsai CF, Tai CT, Hsieh MH, et al. Initiation of atrial fibrillation by ectopic beats originating from the superior vena cava: electrophysiological characteristics and results of radiofrequency ablation. Circulation. 2000;102:67–74[Abstract/Free Full Text]
12. Tang CW, Scheinman MM, Van Hare GF, et al. Use of P-wave configuration during atrial tachycardia to predict site of origin. J Am Coll Cardiol. 1995;26:1315–1324[Abstract]
|
Top
Abstract
Methods