A pacing site is considered to be part of the macro–re-entrant circuit if the first post-pacing interval (PPI) after entrainment is ≤30 ms of the tachycardia cycle-length (TCL) (1- 3). It has been noted earlier, however, that the PPI-TCL may be misleadingly long if the pacing cycle-length (PCL) is ≥40 ms below the TCL (2- 4).

Typical atrial flutter (AFL) is due to right atrial macro–re-entry, and the cavotricuspid isthmus (CTI) is a critical component of the circuit and the primary target for ablation (1). We aimed to evaluate the prevalence and mechanism of a misleading long PPI-TCL in patients presenting with typical AFL for CTI ablation.

This study was conducted in patients presenting with typical AFL for ablation therapy. A retrospective analysis addressed the prevalence of a misleading long PPI-TCL (>30 ms) upon entrainment from the CTI. Thereafter, we conducted a prospective study to systematically evaluate the PPI at different PCLs from pre-specified CTI sites.

Institutional ethics committees approved the study protocol, and subjects gave written informed consent. For the retrospective analysis, we reviewed data from 243 consecutive subjects referred for AFL ablation to the Brigham and Women's Hospital. A total of 60 patients met the following pre-defined criteria: 1) sustained AFL during the electrophysiological (EP) procedure; 2) atrial activation from the proximal to the distal coronary sinus; 3) documented entrainment with concealed fusion from the CTI; and 4) AFL termination during CTI ablation.

The prospective study was conducted at the University Medical Center in Göttingen. Patients with persistent AFL were eligible if the 12-lead electrocardiogram was suggestive of typical AFL. Prior atrial ablation or surgery were exclusion criteria.

All catheters were introduced through femoral veins. Bipolar intracardiac electrograms were filtered between 30 and 500 Hz and recorded at a sampling rate of 1,000 Hz (Prucka CardioLab EP system, GE Healthcare, Waukesha, Wisconsin).

Diagnostic catheters were positioned as shown in (Figure 1). An open-irrigated 3.5-mm-tip catheter (Celsius Thermocool, Biosense Webster, Diamond Bar, California) was utilized for pacing and ablation. In the retrospective analysis, an 8-mm-tip catheter was also used.

Entrainment from the CTI was performed with ≥15 bipolar stimuli (pulse strength: 10 mA at 2 ms). In the retrospective analysis, stimuli were delivered with a PCL 10 to 40 (21 ± 9) ms below the TCL. In the prospective study, entrainment was performed from 4 pre-specified CTI locations (Figure 1) with a PCL 10 ms, 20 ms, 30 ms, and 40 ms below the TCL. Reproducibility of repeated PPI measurements was evaluated by determining the standard deviation (SD) of 3 consecutive measurements from 1 pre-specified site at a constant PCL. In 2 patients, we utilized a monophasic action potential (MAP) recording and pacing catheter (EasyMap MAP, MedFact, Lörrach, Germany). Details on MAP catheter design and application have been described elsewhere (5). Conventional CTI ablation was performed as described earlier (6).

Local atrial activation was defined as the first sharp peak in the corresponding bipolar electrogram or as sharpest deflection of the local MAP upstroke. TCL was the mean value of 5 consecutive atrial cycles during AFL immediately before entrainment. TCL variability was quantified by TCL range (difference between maximum and minimum cycle length of 5 consecutive AFL cycles) and coefficient of variation (standard deviation of 5 consecutive AFL cycles divided by mean TCL × 100%).

PPI was defined as the interval from pacing impulse onset of the last captured beat during entrainment to subsequent local atrial activation at the pacing site. PPI was only measured if: 1) local electrograms from the pacing site indicated stable catheter position; 2) ≥5 of the last atrial cycles before cessation of pacing were captured and accelerated to the PCL; 3) surface electrocardiograms and intracardiac electrograms were suggestive of entrainment with concealed fusion (2); and if 4) AFL resumed after pacing without change in TCL or activation sequence.

Segmental conduction time was measured from local atrial activation (or stimulus artifact) at the beginning to local activation at the end of the 5 atrial segments, respectively (Figure 1). Segmental conduction delay was the difference between segmental conduction time during AFL and during entrainment.

Continuous variables were tested for normal distribution, were compared using the independent Student t test, and are expressed as mean ± SD unless stated otherwise. Categorical values are stated as absolute and relative frequencies and were compared using Fisher exact test. The effects of site of entrainment and PCL on PPI-TCL were assessed by 2-way repeated-measures analysis of variance followed by Tukey test, if applicable. Differences in conduction delay between atrial segments were compared using 1-way repeated-measures analysis of variance followed by Tukey test, if applicable. All tests were 2-tailed. Values of p < 0.05 was considered statistically significant.

A long PPI-TCL from the CTI was found in 18% of the 60 patients with typical AFL (Figure 2). (Figure 3) shows a representative example. Patient characteristics were similar between subjects with short or long PPI-TCL, but individual TCL variability was greater in patients with a long PPI-TCL (Table 1).

Twenty patients were eligible for enrollment and gave written informed consent. Four subjects were excluded because AFL was not present in the EP lab. Thus, data from 16 patients (age 69 ± 9 years, 94% male) were analyzed. Two patients were taking amiodarone, 1 dronedarone, and 1 flecainide. Mean TCL of AFL was 240 ± 28 ms. In 3 patients, the pacing protocol was not completed because AFL terminated with entrainment. In all remaining subjects, AFL terminated with CTI ablation.

Entrainment site and stimulation rate both influenced the PPI-TCL (p < 0.001) (Figure 4). Overall, a PPI-TCL >30 ms was found in 20%, 20%, 37%, and 45% of the cases for a TCL-PCL of 10 ms, 20 ms, 30 ms, and 40 ms, respectively. PPI values at constant PCL varied considerably among different CTI sites but were highly reproducible at single sites (3 ± 2 ms).

Subjects of the prospective cohort with at least 1 long PPI-TCL upon entrainment from the CTI (n = 11) had a similar rate of antiarrhythmic drug use (25% vs. 25%, p = 1.0) and comparable TCL (238 ± 29 ms vs. 245 ± 31 ms, p = 0.68) but tended to have greater TCL variability as compared with patients with only short PPI-TCL (individual TCL range: 10 ± 3 ms vs. 7 ± 2 ms, p = 0.064; TCL coefficient of variation: 1.7 ± 0.5 ms vs. 1.2 ± 0.2 ms, p = 0.070).

Segmental atrial conduction times during AFL are summarized in (Figure 1). Segmental conduction delays with entrainment in cases of long PPI-TCL are shown in (Figure 5). With different entrainment sites, the greatest delay was not bound to one specific segment but was always located in the segment activated first by the orthodromic paced wave front. Accordingly, MAP catheter utilization showed local stimulus response latency (Figure 6).

This study provides several important findings: In patients with typical AFL, long PPI-TCL values >30 ms after entrainment from the CTI were observed:

It was noted earlier that entrainment from a site within the re-entrant circuit may be misleading when the PCL is ≥40 ms below the TCL (2- 4). Misleading long PPI values in our study were common even when the PCL was 10 to 20 ms below the TCL. Fatemi et al. (7) found PPI-TCL values >20 ms after entrainment of typical AFL from the CTI in 27% of the cases using a PCL 30 ms below the TCL. Long PPI-TCL values were more common in subjects on amiodarone. In our analysis, a long PPI-TCL was associated with greater TCL variability. Prior investigations have related increased TCL variability during typical AFL with the loss of a fully excitable gap (8).

It has been postulated that misleading long PPI-TCL values may be caused by rate-dependent slowing of conduction or alterations of the activation path (3- 4). It is also known that conduction significantly prolongs in the human atrium with premature stimuli delivered during the relative refractory period (9). This conduction delay is caused by local pacing latency and results from incomplete recovery of excitability of the tissue that is immediately adjacent to the pacing electrodes (5,10).

Pacing latency explains why the greatest conduction delay in our study was not confined to one distinct anatomical site but was always located in the segment immediately downstream from pacing. Other segments were less delayed, probably because additional time was available for recovery of repolarization. MAP recordings provided further support for this hypothesis. We cannot exclude, however, that other potential mechanisms, e.g., rate-dependent slowing of impulse propagation or altered pathways of conduction, may contribute in some patients.

Considerable spatial differences exist in right atrial refractoriness, which may explain why PPI-TCL values from the CTI region showed significant dependence on pacing location in our study: At sites with a long PPI-TCL, the local refractory period may have been relatively longer, which produced more pacing latency. One may speculate that refractory periods are shorter at the mid CTI close to the inferior vena cava (where shorter PPI-TCL values were found in our study).

Antiarrhythmic medication was not discontinued, but use was low and not different between subjects with short or long PPI-TCL values. MAP recordings were acquired in subjects without anti-arrhythmic medication.

We did not map the re-entry pathway, but the inclusion criteria make arrhythmia mechanisms other than typical AFL unlikely. Intra-isthmus re-entry may have similar electrocardiographic appearance as typical AFL (11). However, intra-isthmus re-entry is a rare finding that is mostly linked to a history of prior CTI ablation, which was excluded in our prospective cohort.

Entrainment from the CTI is commonly utilized to rapidly confirm or exclude CTI-dependent AFL. Our results demonstrate that a misleading long PPI-TCL is common, even if the PCL is within 20 ms of the TCL. Entrainment from near the inferior vena cava in the mid CTI may help to minimize the chances of a misleading PPI value. Although we identified limitations of entrainment during typical AFL, the findings may also apply to other macro–re-entrant tachycardias.