|
| HOME | SUBSCRIPTIONS | CURRENT ISSUE | PAST ISSUES | CARDIOSOURCE | SEARCH | HELP | FEEDBACK |
|
|
|
|||||||||
|
|||||||||||
|
J Am Coll Cardiol, 2002; 39:896-906 © 2002 by the American College of Cardiology Foundation |
a Department of Cardiology, The Royal Melbourne Hospital and University of Melbourne, Melbourne, Australia
Manuscript received September 11, 2001; revised manuscript received December 3, 2001, accepted December 14, 2001.
* Reprint requests and correspondence: Prof. Jonathan M. Kalman, Department of Cardiology, The Royal Melbourne Hospital, Grattan Street, Parkville 3050, Melbourne, Australia
jon.kalman{at}mh.org.au
 |
Abstract |
|---|
 Top Abstract MethodsResultsDiscussionReferences |
|---|
This study was designed to determine the sensitivity and specificity of concealed entrainment (CE) for the identification of a critical isthmus in the atrium.
BACKGROUND
Isthmus identification during entrainment mapping of macro-reentrant atrial tachycardia (MRAT) relies on the demonstration of CE.
METHODS
Using the model of typical atrial flutter, entrainment was performed in 10 patients at four rates (flutter cycle length [FCL] minus 10/20/30/40 ms) from seven sites: isthmus entrance/exit, low lateral/high lateral/high septal right atrium and proximal/distal coronary sinus. Surface 12-lead electrocardiogram fusion was evaluated by three observers blind to patient status. The extent of antidromic penetration (AP) was measured off the pacing catheter positioned around the tricuspid annulus.
RESULTS
The sensitivity for CE identifying any isthmus site was greatest at FCL–10 (100%), but the specificity was poor (54%). Conversely, specificity was greatest at FCL–40 (98%), but the sensitivity was poor (65%), with manifest entrainment (ME) observed from the isthmus entrance in 70% of episodes. At FCL–30, sensitivity (85%) and specificity (90%) were "balanced," but CE still resulted during entrainment from a non-isthmus site in five of 10 patients. Antidromic penetration increased with pacing CL shortening (p < 0.001) and correlated with the development of ME (p < 0.001). Antidromic penetration was significantly blunted from the isthmus exit compared to all other sites (p = 0.003).
CONCLUSIONS
The sensitivity and specificity of CE for identifying an isthmus in the atrium are critically dependent on the pacing rate and the precise anatomic pacing site within the isthmus. These findings may have implications for the use of entrainment in the mapping of unknown MRAT circuits.
| ||||||||||||||||||||||||||||||||||||||||
Although entrainment mapping has proved to be an important adjunctive tool during ablation of ventricular arrhythmias, the predictive value of CE in isolation for identifying a successful ablation site has been relatively low (6,13). In the landmark study by Stevenson et al. (6), the positive predictive value of the observation of CE associated with a post-pacing interval (PPI)-tachycardia cycle length (TCL) <30 ms for successful ablation of ventricular tachycardia in the presence of scar was only 25%. During atrial macro-reentry there is a relative paucity of data describing the utility of CE for identification of a narrow isthmus.
The aim of the present study was to prospectively characterize the specificity and sensitivity of CE for identifying an isthmus in the atrium. We also explored the effect of entrainment rate, anatomic relationship of the pacing site to the isthmus and extent of antidromic penetration (AP) of the circuit on the demonstration of fusion on the surface electrocardiogram (ECG). We used typical counter-clockwise atrial flutter (AFL) as the model for this study because the critical components of this circuit have been extensively described (10,14–17).
|
Methods |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
Definitions.
Manifest entrainment (ME)
The acceleration of AFL to the particular pacing rate (capture of the reentrant circuit) with evidence of constant fusion in the F-wave morphology as assessed by a standard 12-lead ECG. Upon termination of pacing there is spontaneous resumption of AFL (i.e., the tachycardia has not been terminated), and the last paced beat is always entrained but not fused. The findings of constant fusion and progressive fusion (different degrees of constant fusion at different pacing rates) confirm the presence of a reentrant circuit with an excitable gap. These findings represent fulfillment of the first and second criteria for establishing the presence of transient entrainment (18).
Concealed entrainment
Concealed entrainment is the acceleration of AFL to the particular pacing rate (capture of the reentrant circuit) without evidence of fusion in the F-wave morphology as assessed by a standard 12-lead ECG. The diagnosis of CE can only be made after ME has been demonstrated at another site to prove the existence of a reentrant circuit with an excitable gap.
Isthmus entrance
The isthmus entrance is a site within the cavo-tricuspid isthmus in its lateral region at approximately 6:30 when viewed in the left anterior oblique (LAO) projection.
Isthmus exit
The isthmus exit is a site in the cavo-tricuspid isthmus in its medial region at approximately 5:00 when viewed in the LAO projection. The coronary sinus (CS) ostium was designated 4:00 in this projection.
Electrophysiologic study.
Catheter positioning
Venous access was obtained via the right femoral vein and right internal jugular vein (for the CS catheter). A 20-pole tricuspid annulus (TA) catheter was positioned in the right atrium (RA) parallel to the TA such that the distal pole was located in the medial region of the cavo-tricuspid isthmus (inter-electrode spacing 2 mm and inter-bipole spacing 5 mm: "2–5–2"). A decapolar catheter was positioned in the CS (electrode spacing 2–5–2 mm) with the proximal bipole positioned 1 cm distal to the CS ostium as determined in the LAO projection. A quadripolar catheter was positioned to record a His bundle electrogram. Intracardiac signals and a standard 12-lead ECG were displayed on a computerized mapping system and recorded to optical disk.
Entrainment protocol
The study protocol was performed before RFA. Pacing was performed from seven anatomic sites (two within the isthmus and five outside the isthmus) defined by fluoroscopic visualization of the RA using standard left and right anterior oblique projections.
The cycle length (CL) of the AFL was measured (flutter cycle length, FCL) and entrainment performed at 10, 20, 30, and 40 ms below the FCL (FCL–10, FCL–20, FCL–30 and FCL–40). For each site and CL, pacing was performed three times. Pacing was performed at twice the local capture threshold determined during pacing at a rate faster than the FCL. If the local capture threshold was >2.5 mA, the position of the TA catheter was moved to ensure better atrial wall contact. Each episode of pacing was performed until steady-state entrainment was achieved for 5–8 s as determined by the presence of a consistent relationship between the pacing stimulus artifact and the atrial electrograms and a constant unchanging paced F-wave morphology.
For each episode the post-pacing interval (PPI) was measured as the interval from stimulus artifact to onset of the initial sharp deflection of the first post-pacing electrogram. An offline analysis of the PPI minus the FCL was made after each episode of entrainment. If the F-waves during pacing were obscured by the QRS complex or T-wave, an AV blocking agent (digoxin 250 to 500 µg or atenolol 2.0 to 5.0 mg) was administered intravenously.
At completion of the research protocol all patients underwent conventional isthmus ablation with confirmation of bidirectional isthmus block.
Determination of concealed entrainment
For each episode of entrainment a high quality copy of the ECG was printed showing the F-wave morphology both during and after pacing. The ECG was displayed at 50 mm/s speed with the signals gained to a level of 25 mm/mV to allow a clear appreciation of the F-wave morphology. Electrocardiogram traces were then reviewed by three electrophysiologists blind to the site and rate of pacing. The paced F-waves in each of the 12 surface ECG leads were evaluated for amplitude, duration and morphology and compared to the non-paced F-wave. The episode of entrainment was defined as concealed when there was an exact match in all 12 surface ECG leads. For episodes where there was a disagreement among the three reviewers, the majority opinion was deemed correct. The inter-observer variability for the description of CE was calculated.
Antidromic penetration
The number of bipoles antidromically penetrated by the paced activation wave front was measured from the 20-pole TA catheter for each of the sites excluding the proximal and DCS. A bipolar recording site was considered to have been penetrated by the antidromic wave front when the following criteria were constantly present during entrainment: 1) electrogram advancement by 
10 ms compared with timing during AFL; 2) change in activation sequence (reversal of the activation sequence from the entrained site); and 3) change in electrogram morphology (4).
The distance of antidromic capture (mm) was measured from the center of the stimulating bipole pair to the center of the furthermost antidromically penetrated bipole. All measurements of electrogram timing were made offline using on-screen calipers at 400 mm/s sweep speed. The onset of the bipolar electrogram was taken as the onset of the initial sharp deflection.
Statistics
Continuous data are expressed as mean ± 1 SD. Proportions are expressed as percentages with 95% confidence intervals (CI) (19). To determine the sensitivity and specificity a conventional 2 x 2 matrix (A/B/C/D) was constructed for each rate of entrainment where A = true positive result (CE when pacing an isthmus site); B = false positive result (CE when pacing a non-isthmus site); C = false negative result (ME when pacing an isthmus site); and D = true negative result (ME when pacing a non-isthmus site). Hence sensitivity = A/(A + C) and specificity = D/(B + D). The effect of entrainment CL on sensitivity and specificity was determined using logistic regression analysis for ME versus CE predicted by patient, site and rate. The effect of pacing site and rate on the degree of AP was determined by performing a two-factor (rate and site) analysis of variance. The relationship between the extent of AP and surface ECG fusion (expressed at two levels: CE or ME) was determined by extending the analysis of variance to include fusion as a third factor. Multiple comparisons were performed using Tukey pairwise comparisons. Inter-observer variability for the description of CE versus ME was determined by calculating 
(the kappa coefficient) (20). A p value of <0.05 was considered significant. Statistical analysis was performed using commercially available software: Minitab (Minitab Inc., State College, Pennsylvania) and LogXact (Cytel Software Corp., Cambridge, Massachusetts).
|
Results |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
PPI-FCL.
During entrainment from sites within the flutter circuit, the mean PPI-FCL was 
30 ms for all four pacing rates. The PCS pacing site was within the AFL circuit in five of 10 patients and outside the AFL circuit in the remainder (mean PPI-TCL 58 ms). The DCS pacing site was outside the AFL circuit in eight of eight patients (mean PPI-TCL 97 ms). There was no evidence of significant lengthening of the PPI during entrainment at faster (FCL–40) as compared with slower (FCL–10) rates when pacing from the HSRA (FCL–10 vs. FCL–40: 22 ± 13 vs. 26 ± 20 ms, p = ns), HLRA (25 ± 11 vs. 27 ± 18 ms, p = ns), LLRA (25 ± 18 vs. 29 ± 17 ms), isthmus entrance (21 ± 11 vs. 25 ± 19 ms, p = ns) or in those patients where the PCS was in the circuit (23 ± 16 vs. 26 ± 18 ms, p = ns). However, during entrainment from the isthmus exit a small but statistically significant lengthening in the mean PPI was observed (FCL–10 vs. FCL–40: 14 ± 14 vs. 21 ± 19 ms, p = 0.02). Despite this lengthening, the site was still within the circuit based on a definition of PPI-TCL <30 ms. In addition, lengthening of the mean PPI was observed during entrainment from the PCS in those five patients where this site was outside the flutter circuit (59 ± 13 vs. 73 ± 13 ms, p = 0.04).
Concealed entrainment. The standard 12-lead surface ECG recordings of 272 episodes of entrainment were classified as concealed (C) (n = 107) or manifest (M) (n = 165). In two patients entrainment was not performed from the DCS. The percentage of entrainment episodes classified as CE is presented in Figure 1 according to the site and rate of entrainment pacing.
|
|
|
|
|
|
There was a highly statistically significant relationship between the degree of AP and the documentation of entrainment as either concealed or manifest (p < 0.001). Adjusting for site and rate, episodes resulting in ME were associated with a greater degree of AP (95% CI: 2–6 mm) than those episodes resulting in CE.
Inter-observer error.
There was a high degree of correlation among each of the three observers for the description of the ECG during entrainment as either ME or CE: = 0.8 for observer 1 and 2, = 0.9 for observer 1 and 3 and = 0.9 for observer 2 and 3.
|
Discussion |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
A highly statistically significant relationship was shown to exist between the degree of AP and each of the following factors: pacing rate, pacing site and presence or absence of CE on the surface ECG.
Utility of entrainment mapping: comparison with other studies. Many clinical studies have determined the sensitivity of CE for identifying an isthmus and RFA target site during entrainment mapping of macro-reentrant ventricular tachycardias (6,12,13). The predictive value of CE in isolation for identifying an isthmus has generally been low and has required the addition of other parameters to increase it.
Fewer studies have evaluated CE in the atrium (7,21). Bogun et al. (21) performed entrainment mapping and RFA of atypical AFL in a population without prior cardiac surgery. When pacing was at 10 to 30 ms below the FCL, the predictive value of CE for identifying a successful RFA site was only 45%. This increased to 75% in the presence of matching electrogram-F and stimulus-F intervals or if flutter terminated during entrainment pacing, and to 88% in the presence of split atrial electrograms or diastolic potentials. A limitation of using an unknown circuit such as an atypical right AFL to evaluate the utility of CE to identify an isthmus is that an RF application may be ineffective not only because it is incorrectly located but for other reasons including inadequate lesion size (either width or depth) or inadequate tissue heating. In addition, because RF applications will not be applied at sites demonstrating ME, the true specificity and negative predictive value cannot be calculated.
More recently a number of studies have compared the presence of CE with a circuit defined by an electroanatomic map in patients with "scar-mediated" atrial macro-reentry. Nakagawa et al. (22) performed entrainment mapping in six of 15 patients undergoing electroanatomic mapping of scar-mediated tachycardia. In these patients many sites within the circuit, but outside the isolated channel ("outer loop" site), exhibited entrainment with concealed fusion (same P-wave and activation sequence) and a PPI equal to the TCL. In 15 patients with repaired congenital heart disease and intra-atrial reentrant tachycardia undergoing electroanatomic mapping, Triedman et al. (23) found that the fraction of atrial endocardial points meeting entrainment criteria used in this study (PPI-TCL 20 ms at an entrainment rate of 5 to 25 ms below the tachycardia CL) was generally large, averaging 20% to 30%. Thus, the authors suggested that targeting sites for ablation solely by entrainment may have a low predictive value for success. In neither of these studies did the authors evaluate the effect of entrainment rate or systematically detail the sensitivity and specificity of CE for identification of a channel or isthmus.
By using a macro-reentrant tachycardia with a known circuit in the present study, a more accurate picture of the limitations of CE for the identification of an isthmus site could be obtained.
Is there an "ideal" entrainment rate?. The present study demonstrates that whereas there may be no ideal entraining cycle length, there is instead an optimal CL that produces the best compromise. Furthermore, it may be best to utilize several entraining CL at a designated site in an attempt to improve diagnostic accuracy. Thus, pacing at a long CL (either FCL–10 or FCL–20) was associated with a 23% to 46% false-positive rate attributable to CE from non-isthmus circuit sites. The highest specificity (98%) was associated with pacing at shorter CLs (FCL = 40 ms), but at this rate isthmus entrance sites usually (70%) demonstrated ME. As such, at this CL certain isthmus sites may be inappropriately rejected as potential target sites for RFA. At an intermediate pacing rate of FCL–30, the sensitivity (85%) and specificity (90%) were more "balanced" than at other CLs, but concealed fusion still resulted during entrainment from one of five non-isthmus sites in 50% of patients.
It has been demonstrated during entrainment of ventricular arrhythmias that entraining at faster rates may increase the PPI by decremental conduction within the circuit (6). In the current study, although there was a slight increase in the PPI when entraining at TCL-40 compared with TCL-10, this was not of sufficient magnitude to result in sites within the circuit falsely appearing "out."
Extent of antidromic penetration. The degree of AP significantly increased at all sites during pacing at progressively shorter CL and was significantly associated with the finding of fusion (ME) on the surface ECG (p < 0.001). The demonstration of increasing AP with increasing pacing rate represents fulfillment of the originally described fourth criteria for establishing the presence of transient entrainment, the electrogram equivalent of progressive fusion (4). The current study has quantified this variable (AP) in a known macro-reentrant circuit and systematically analyzed the relationship among AP, entrainment site, rate and presence/absence of surface ECG fusion.
Cosio et al. (17) have previously shown during entrainment of typical AFL from the high anterior RA that surface fusion may occur in the absence of significant AP if there is overlap of orthodromic septal activation by the n–1 wave front with anterior wall activation of the following cycle. This site-dependent effect would also be expected to be more prominent at shorter pacing CL (Fig. 2C).
In the present study, pacing from the isthmus exit produced almost 50% less AP than that measured from the isthmus entrance and other non-isthmus sites at the same rate. In typical AFL the isthmus, and in particular the medial isthmus, has been shown to be an area of slowed conduction, which would be expected to influence the extent of AP (10,15,17,24,25). In contrast, the absolute degree of AP from the lateral isthmus was not different at FCL–40 ms from the non-isthmus sites. Extrapolating these data to unknown atrial circuits implies that during entrainment at shorter CL from isthmus entrance sites, and possibly from some mid-isthmus sites, ME may result due to significant AP.
Limitations. In this study, the extent of AP was measured from a single multipolar catheter (TA catheter). If the AP occurred in a direction at an angle to this catheter, or if there was anisotropy of the extent of AP, this technique would provide an inaccurate assessment.
Conclusions. On the basis of this model, the diagnostic utility of CE for identification of an isthmus in the atrium is highly dependent on the pacing CL and the anatomic location of the site relative to the isthmus exit. Concealed entrainment is frequently observed at long CL from non-isthmus sites within the circuit, but may be absent from entrance isthmus sites at shorter CL. These findings may have implications for mapping of other macro-reentrant tachycardias where the location of the critical isthmus is unknown.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  P. Maury, A. Duparc, A. Hebrard, M. El Bayomy, and M. Delay Prevalence of typical atrial flutter with reentry circuit posterior to the superior vena cava: Use of entrainment at the atrial roof Europace, February 1, 2008; 10(2): 190 - 196. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | SUBSCRIPTIONS | CURRENT ISSUE | PAST ISSUES | CARDIOSOURCE | SEARCH | HELP | FEEDBACK |
