Role of Right Atrial Endocardial Structures as Barriers to Conduction During Human Type I Atrial Flutter
Activation and Entrainment Mapping Guided by Intracardiac Echocardiography
Background The importance of barriers in atrial flutter has been demonstrated in animals. We used activation and entrainment mapping, guided by intracardiac echocardiography (ICE), to determine whether the crista terminalis (CT) and eustachian ridge (ER) are barriers to conduction during typical atrial flutter in humans.
Methods and Results In eight patients, ICE was used to guide the placement of 20-pole and octapolar catheters along the CT and interatrial septum and a roving catheter to nine sites: just posterior (1) and anterior (2) to the CT along the lateral right atrium, at the fossa ovalis (3), and just posterior and anterior to the ER at the low posterolateral (4 and 5), low posterior (6 and 7), and low posteromedial (8 and 9) right atrium. Entrainment was performed, and each site was considered within the flutter circuit if the postpacing interval–flutter cycle length (PPI−FCL) and the stimulus time–activation time (stim time−act time) were <10 msec. Split potentials were recorded along the CT with components activated in a low-to-high pattern and a high-to-low pattern. Conduction times, as percentage of FCL, were significantly different at sites on either side of the CT and ER: site 1 (33±13%) and site 2 (43±12%) (P=.02), site 4 (48±24%) and site 5 (75±8.9%) (P=.02), and site 6 (22±10%) and site 7 (82±5.3%) (P=.0009). During entrainment, no surface fusion was observed at sites 5, 7, or 9. The PPI−FCL and stim time−act time were not significantly different than 0 at sites 2, 7, 5, or 9, indicating that they were within the flutter circuit, whereas sites 1, 3, 4, and 6 were not.
Conclusions ICE enabled the correlation of functional electrophysiological properties with specific anatomic landmarks, identifying the CT and ER as barriers to conduction during human atrial flutter.
As early as 1947, when Rosenblueth developed a canine atrial flutter model by crushing endocardium between the venae cavae, it was realized that atrial flutter likely involved the propagation of a wave front around anatomically defined obstacles.1 More recently, canine models of atrial flutter have demonstrated the importance of anatomic obstacles in the reentry circuit.2 3 4 5 It is now well established that human type I atrial flutter is a reentrant rhythm localized to the right atrium.6 7 8 A protected, narrow isthmus of conduction has been identified in the low right atrium between the inferior vena cava and the tricuspid annulus.6 9 10 11 More recent evidence suggests that this area extends across the floor of the right atrium to the os of the CS.12 13 14 15 16 A line of block has also been observed in the lateral right atrium, extending superior from the orifice of the inferior vena cava.6 17 However, the extent to which these areas of conduction block are related to anatomic architecture has not been established.
Transient entrainment of atrial flutter was first described in 1977.18 Since then, activation and entrainment (or pace resetting) mapping have been used to localize regions of myocardium critical to the maintenance of reentrant circuits.9 19 20 21 22 However, because current imaging modalities, such as fluoroscopy, that are used to guide catheter placement do not allow direct visualization of anatomic structures, such mapping techniques are unable to directly relate functional properties of the circuit to specific anatomic features in the intact human heart. Intracardiac echocardiography is a technique that can be used to visualize various intra-atrial structures that are not visualized on fluoroscopy (eg, the fossa ovalis, CT, eustachian ridge, CS, venae cavae) and allow precise localization of intracardiac catheters relative to these anatomic structures.23 24 25 26
The present study was performed to define the role of right atrial endocardial structures that may serve as barriers to conduction during type I atrial flutter. Specifically, we used activation and entrainment mapping with anatomically determined catheter positioning guided by intracardiac echocardiography to test the hypothesis that the CT and eustachian ridge are anatomic barriers to conduction during typical flutter.
The study included 8 patients (2 women and 6 men), age 52 to 72 years (mean, 66 years), with type I atrial flutter who were referred to the University of California, San Francisco, between August 1994 and December 1994 for curative atrial flutter ablation. All patients had characteristic flutter waves with a superior axis (negative complexes in leads II, III, and aVF). The range of atrial rates was 210 to 270 bpm (cycle length, 220 to 290 msec) with a mean of 234±24 bpm (cycle length, 258±26 msec). All antiarrhythmic drugs, with the exception of digoxin, were stopped at least 24 hours before the study. Six of 8 patients were in typical flutter at the onset of the study. The remaining 2 patients had type I atrial flutter documented clinically before the study and had type I atrial flutter reproducibly induced with atrial burst pacing and atrial extrastimulation. Informed consent was obtained from all patients before the study according to the protocol approved by the Committee on Human Research.
A 10F, 10-MHz ICE catheter (Cardiovascular Imaging Systems, Inc) was inserted through a femoral vein sheath and advanced into the right atrium. Images were obtained with an Insight 3 (Cardiovascular Imaging Systems, Inc) intravascular ultrasound machine and recorded on videotape. The technique of ICE to guide electrophysiology procedures has been previously described.23 24 25 26 ICE was used to identify the endocardial structures of the right atrium, including the orifices of the venae cavae, fossa ovalis, CT, eustachian ridge, CS, and tricuspid annulus and, along with fluoroscopy, to guide the placement of all catheters as well as to manipulate the roving catheter to specific anatomic positions as described.
A custom 8F, 20-pole steerable catheter (Mansfield) with 2-mm interelectrode distance was placed through an 8F long vascular sheath (SRO, Daig) in the femoral vein and positioned along the CT. ICE was used to ensure continuous catheter position firmly against the CT along its entire extent from the superomedial right atrium (junction of the right atrium and superior vena cava) to the inferior right atrium (junction of the right atrium and inferior vena cava) (Figs 1⇓ and 2A⇓). A 7F, octapolar steerable catheter (EP Technologies) (2.5-mm distance between bipole electrodes and 10-mm distance between bipole pairs) was positioned under ICE guidance along the interatrial septum (Figs 1⇓ and 2⇓). An 8F, open-lumen, decapolar catheter (Elecath) (2-mm interelectrode distance and 5-mm intraelectrode distance) was placed via the internal jugular vein into the CS. Positioning of the proximal electrode pair at the os of the CS was confirmed with contrast injection viewed with fluoroscopy. An 8F, quadipolar catheter (EP Technologies) (4-mm electrode tip; other electrodes, 2 mm with 2-mm interelectrode spacing) was used as a roving mapping, pacing, and ablation catheter.
The roving catheter tip was guided precisely to the following nine sites under ICE guidance based on the characteristic fan-shaped artifact cast by the large distal electrode (Figs 1⇑ and 2⇑): (1) approximately 0.5 cm posterior to the CT, on the smooth right atrium; (2) approximately 0.5 cm anterior to the CT, on the trabeculated right atrium; (3) at the fossa ovalis; (4) posterior to the eustachian ridge, far lateral from the CS os, where the CT joins the eustachian ridge; (5) anterior to the eustachian ridge (opposite site 4), between it and the tricuspid annulus, far lateral from the CS os, where the CT joins the eustachian ridge; (6) posterior to the eustachian ridge, midway between the CS os and site 4; (7) anterior to the eustachian ridge (opposite site 6), between it and the tricuspid annulus, midway between the CS os and site 5; (8) posterior to the eustachian ridge at the CS os (posterior to the CS os); and (9) anterior to site 8, between the CS os (anterior to the CS os) and the tricuspid annulus.
Bipolar intracardiac electrograms filtered between 30 and 500 Hz were recorded and stored digitally on a Cardiolab system (Prucka Engineering) simultaneously with 12-lead surface ECGs. All measurements were performed with the Cardiolab system at screen speeds of 400 to 1600 mm/s using on-screen digital calipers.
During atrial flutter, unipolar pacing was performed at each site from the distal electrode of the roving catheter. A large patch skin electrode was used as the indifferent electrode. A custom-made Y-splitter that attached to the lead from the distal electrode allowed simultaneous recording of the distal bipole pair during unipolar pacing. Pacing was performed at 10, 20, and 40 msec less than the spontaneous FCL for at least 10 beats at the lowest capturing current with a 2-msec pulse width. In those patients who were in sinus rhythm at the onset of the study, pacing at the previously documented FCL was performed during sinus rhythm.
“Split potentials” were defined as distinct early and late potentials separated by an isoelectric phase.10 When split potentials were present, separate measurements were made for each component. Fragmented potentials were not considered split, and measurements were made at the onset of the fragmented potentials.
During entrainment, “fusion” was defined as being present, and thus manifest, when a different morphology of the surface flutter wave and/or change in activation sequence of the intracardiac recordings occurred during pacing compared with spontaneous flutter. Entrainment with “concealed fusion” was defined as entrainment with identical flutter wave morphology as during spontaneous flutter. The interval between the last paced beat entraining the tachycardia and the first beat after pacing was carefully measured. This “PPI” was defined as the time necessary to conduct from the pacing site to the reentrant circuit, through the circuit and back to the pacing site (Fig 3⇓). It was measured from the last pacing artifact to the onset of the first atrial signal on the distal electrode pair of the roving catheter (pacing catheter) (Fig 4⇓). The “FCL” was determined from the average of three beats after the first postpacing beat.
Because the onset of the flutter wave is often difficult to discern on the surface ECG—either because of rapid atrioventricular conduction or lack of a true isoelectric point on the surface flutter wave—the electrogram from the coronary sinus os was used as a surrogate marker for the onset of the flutter wave, as previously described.9 10 12 21 27 “Stim Time” was defined as the orthodromic time of conduction from the pacing site to the exit site from the protected isthmus (CS os) during entrainment (Fig 3⇑). The last entrained beat is, by definition, not fused and therefore activated orthodromically.18 Because it is the orthodromic Stim Time that is relevant, the interval from the last pacing artifact to the onset of the last CS os electrogram that is entrained at the pacing rate was used for this measurement (Fig 5⇓).19 21 “Act Time” was defined as the time between the onset of the local electrogram of the roving catheter and activation of the CS os during spontaneous flutter (Fig 3⇑). To distinguish it from Act Time, “conduction time” was defined as the time between the onset of the local electrogram from the electrode pair positioned at the CS os and activation of electrode pairs at each recording site during spontaneous flutter (Fig 3⇑). Conduction times were expressed as the percentage of the FCL to compare values from patients with differing flutter rates.
Sites were considered to be within the reentrant circuit if both the PPI was within 10 msec of the FCL (PPI−FCL≤10 msec) and the Stim Time was within 10 msec of the Act Time (Stim Time−Act Time≤10 msec) (Fig 3⇑). Ten milliseconds was chosen because it is the estimated precision of measurement using the digital calipers (determined by 10 repeated measures of 20 different intervals). Sites were considered to be in a relatively protected isthmus if concealed fusion resulted during entrainment.
All intervals during entrainment were measured three times at each site at the three different pacing rates for each patient. PPI−FCL and Stim Time−Act Time were calculated at each site for each patient as an average of values for each pacing rate at each site.
Ablation was performed between the distal electrode of the roving catheter and a large surface area skin electrode with 550 kHz unmodulated radiofrequency current from a generator with temperature monitoring (EP Technologies). During ablation, radiofrequency power was adjusted to obtain a catheter tip temperature of between 65° and 90°C. Sites for ablation were selected if entrainment demonstrated concealed fusion, a PPI within 10 msec of the FCL, and a Stim Time within 10 msec of the Act Time, indicating the site is within the circuit and in a protected isthmus. If tachycardia terminated with ablation, reinduction of flutter was attempted using atrial burst pacing down to a cycle length at which 2:1 capture occurred and with triple atrial extrastimuli at pacing cycle lengths of 400 and 300 msec. Successful ablation was defined as termination of flutter with the inability to reinduce atrial flutter.
Values are expressed as mean±SD. Statistical comparisons were performed using the Student’s t test, paired and unpaired when appropriate. A value of P<.02 was considered statistically significant.
Four of the eight patients had no structural heart disease with normal left ventricular function and atrial size. Three patients had coronary artery disease and previous myocardial infarctions. One patient had a prior aortic valve replacement for aortic insufficiency. The mean left ventricular ejection fraction, determined by transthoracic echocardiography, was 53±9%, with two patients having ejection fractions of less than 45%. Left atrial enlargement (>4 cm) was present in only four patients, and right atrial dimensions were normal (<3 cm) in all patients, as determined by surface echocardiography. All patients failed at least three antiarrhythmic drugs, and at least one had DC cardioversion attempted before the procedure.
The eustachian ridge, the entire extent of the CT, the fossa ovalis, and the tricuspid annulus were identified by ICE in all patients. The catheters placed along the CT and the interatrial septum were identified along their entire intracardiac length by ICE in all patients. Successful localization of the tip of the roving mapping catheter in relation to each of the endocardial structures at sites shown in Fig 1⇑ was achieved in all patients.
Conduction times from the CS os to each recording site are shown in Tables 1 through 3⇓⇓⇓. Split potentials—composed of distinct early and late potentials separated by an isoelectric phase—were recorded along the length of the CT in seven of eight patients (Table 1⇓). Fig 6A⇓ shows typical split potentials recorded from the CT catheter in one patient. The conduction time from the CS os, expressed as a percentage of the FCL, was significantly longer for the late deflections compared with the early deflections for each of the bipoles from which they were recorded (Table 1⇓). The early component had a low-to-high activation sequence, whereas the late component had a high-to-low activation sequence. The septal catheter was activated in a low-to-high sequence in every patient (Table 2⇓). Fig 6B⇓ shows a typical activation sequence in the septal catheter from one patient. Split or fragmented potentials were not recorded from the septal catheter in any patient.
Conduction times from the CS os to each of the nine ICE-determined, anatomically defined mapping sites are shown in Table 3⇑. Conduction times were significantly longer at site 2, just anterior to the CT on the trabeculated right atrium, compared with site 1, just posterior to the CT on the smooth right atrium. Site 3, at the fossa ovalis, was activated on time with the septal catheter poles and sites 1, 4, 6, and 8. Conduction times from the CS os to sites 4, 6, and 8, confirmed by ICE to be just posterior to the eustachian ridge, were not statistically different from each other or from the septal catheter poles. Sites 5, 7, and 9, confirmed by ICE to be just anterior to the eustachian ridge, were activated in sequence significantly later than any other site (P<.005). In addition, split potentials were recorded in three patients at site 7 (Fig 7⇓).
Activation of the trabeculated right atrium—separated from the smooth right atrium by the CT—occurred significantly later than the smooth right atrium (site 2 versus sites 1 and 3). Activation of sites just posterior to the eustachian ridge occurred significantly earlier than sites directly opposite (sites 4 versus 5 and 6 versus 7) and just anterior to the eustachian ridge.
Criteria for entrainment, as defined in previous reports, were established for each patient from one or more of the nine mapping sites.28 29 Results of entrainment mapping are shown in Table 4⇓ and Fig 8⇓. At site 1, manifest fusion was observed, the PPI was significantly longer than the FCL and the Stim Time was significantly longer than the Act Time in all patients. In no patient were these values less than 15 msec at this site. When entrainment from site 2 was performed, fusion was observed in all patients, but the PPI was not significantly different from the FCL and the Stim Time was not different from the Act Time. In no patient were these values >3 msec at this site. Thus, although site 1 (confirmed by ICE to be just posterior to the CT) was not a part of the flutter circuit, site 2 (confirmed by ICE to be just anterior to the CT) was. Manifest fusion was observed during entrainment at site 3 (at the fossa ovalis) in all patients. In addition, the PPI was significantly longer than the FCL and the Stim Time was significantly greater than the Act Time, indicating that the fossa ovalis was not a part of the flutter circuit.
At sites 4 and 6, confirmed by ICE to be just posterior to the eustachian ridge, fusion was observed, the PPI was significantly longer than the FCL, and the Stim Time was significantly longer than the Act Time in all patients. In no patient were these values <15 msec. Conversely, at sites 5, 7, and 9, confirmed by ICE to be just anterior to the eustachian ridge, no fusion was seen in any patient, the PPI was not different from the FCL, and the Stim Time was not different from the Act Time. In no patient were these values >7 msec at these sites. Thus, sites 4 and 6 were not within the flutter circuit, whereas sites 5, 7, and 9 were within the circuit and in a protected isthmus since no fusion occurred.
At site 8, fusion was not observed in two patients during entrainment. In these two patients, the PPI was not different from the FCL (PPI−FCL=0±0), and the Stim Time was not different from the Act Time (Stim Time−Act Time=5±2 msec). In the other patients, fusion was observed during entrainment, the PPI trended toward being longer than the FCL (PPI−FCL=41±23 msec), and the Stim Time trended toward being longer than the Act Time (Stim Time−Act Time=116±45 msec). Thus, in two patients site 8 was in the flutter circuit, whereas in the others it was not.
All patients underwent successful ablation of their atrial flutter. Seven patients had successful ablation at sites far lateral to the CS os, between the tricuspid annulus and the orifice of the inferior vena cava (site 5 or 7). Only one patient had a successful ablation near the CS os (site 9), despite attempts at this site in five patients. As confirmed by ICE, all patients required a pullback of the ablation catheter tip to the eustachian ridge at the anterior rim of the inferior vena cava during radiofrequency application—in effect, ablating in a line from the tricuspid annulus to the orifice of the inferior vena cava.
Atrial overdrive pacing with and without extrastimuli induced no arrhythmia in five patients after initial termination of atrial flutter by ablation. In three patients, atypical flutter (one nonsustained and two sustained) was induced. These arrhythmias had the opposite activation pattern of typical flutter, with a low-to-high activation pattern in the CT and a high-to-low activation pattern in the septum. In addition, the cycle lengths of these atypical flutters were from 50 to 100 msec longer than each patient’s typical flutter. A single repeat radiofrequency application at the previously successful site rendered these arrhythmias no longer inducible.
In the present study, activation and entrainment mapping guided by ICE were used to identify the CT and eustachian ridge as lines of block during type I (typical) atrial flutter in humans. Three sets of results are provided that demonstrate that the CT and eustachian ridge are barriers to conduction during flutter. First, split potentials were recorded along the length of the CT—with opposite activation sequences of each component—and across the eustachian ridge. Second, activation mapping demonstrated that sites determined by ICE to be just anterior to the CT (site 2) and the eustachian ridge (site 5, 7, and 9) were activated significantly later than sites determined by ICE to be just posterior to these structures (sites 1 and 3 and sites 4, 6, and 8, respectively). Third, entrainment mapping identified sites determined by ICE to be just anterior to the CT and the eustachian ridge as within the reentrant circuit, whereas sites determined to be just posterior to these structures were identified as not being within the circuit. These data indicate that the flutter wave front is directed inferiorly along the trabeculated right atrium anterior to the CT into a narrow isthmus between the eustachian ridge and the tricuspid annulus and then medially between these two structures to exit near the CS os (Fig 9⇓).
ICE reliably identified key right atrial endocardial structures not visible with fluoroscopy, including the CT, eustachian ridge, tricuspid annulus, interatrial septum, fossa ovalis, and CS. Because the tip of the roving mapping catheter casts a characteristic fan-shaped artifact on ICE images, the catheter tip could be precisely guided to these anatomic positions for mapping and entrainment.23 24 25 26 Fluoroscopy provides only a rough estimate of anatomic location and is inadequate for mapping in relation to specific endocardial structures. With ICE, in this study we were able to correlate the findings of activation mapping and entrainment with specific endocardial anatomic structures by confirming the spatial relationship of the catheters to these structures.
Entrainment with concealed fusion has been previously demonstrated in protected isthmuses of conduction for ventricular tachycardia and atrial flutter.9 10 19 27 During entrainment (either manifest or concealed), local fusion occurs at some point. Fusion will be observed on the surface ECG if and only if the area of myocardium activated antidromically is large enough to affect the surface ECG. Concealed fusion occurs when pacing is delivered to areas where functional or fixed barriers prevent capture of a significant area of the chamber (ie, enough to affect the surface ECG) in any direction other than from the normal exit of a protected isthmus.
In the present study, we used ICE to demonstrate that this zone (containing sites 5, 7, and 9), from which concealed fusion is demonstrated during entrainment, is constrained anatomically by the eustachian valve and the tricuspid annulus. The pacing impulse propagates only orthodromically, and the flutter wave is not inscribed until the wave exits this isthmus. Antidromic fusion occurs within the protected isthmus in an area too small to affect the surface electrogram. Pacing anterior to the CT (site 2), however, produces manifest entrainment, even though it is within the flutter circuit. This occurs because it is not in a protected isthmus, allowing activation of a large enough area of atrium to inscribe a flutter/P wave immediately after the pacing impulse is delivered, and manifest fusion occurs.
In addition to evaluating whether entrainment produces fusion, the Act Times from the pacing stimulus (Stim Time) during entrainment and the PPI for the first beat after cessation of entrainment have been previously demonstrated to be useful in evaluating a reentrant circuit (Fig 3⇑).10 19 30 Inoue et al30 observed that the PPI varied depending on where the pacing and recording sites were in relation to the reentrant circuit. Stevenson et al19 demonstrated in both a computer model and in human ventricular tachycardia that the PPI at the pacing site is a useful indication of whether that site is inside or outside of the reentrant circuit. In our study, entrainment from sites just anterior to the CT (site 2) and just anterior to the eustachian ridge (sites 5, 7, and 9) produced PPIs equal to the FCL and were thus determined to be within the circuit (Fig 8A⇑). Entrainment from sites just posterior to the CT (sites 1 and 3) and to the eustachian ridge (sites 4, 6, and 8) produced PPIs longer than the FCL and thus were not within the circuit (Fig 8A⇑).
The Stim Time (Act Time from the stimulus) can also be used to determine sites that are within the reentrant circuit during entrainment.19 31 If ventricular tachycardia is entrained without fusion, the Act Time (local electrogram to QRS) and the Stim Time (stimulus to QRS) both reflect the time necessary for conduction from the pacing site to the QRS-onset site—which is defined as the exit site from a critical isthmus (Fig 3⇑).19 31 However, if local electrograms are recorded from a known exit site from a protected isthmus (or any site known to be in the reentrant circuit), then neither concealed entrainment nor surface electrograms are necessary for analysis of these times of conduction. Several reports have demonstrated that the exit of the narrow isthmus of conduction in type I flutter is the CS os, which correlates with the onset of the flutter wave.9 10 12 21 27 In our study, the first nonpaced cycle length in the CS os electrogram after entrainment from any site in all patients was always that of the FCL, confirming that the CS os is within the reentrant circuit. Therefore, the CS os is valid as the exit site from the protected isthmus and a surrogate for the flutter wave–onset site.30
The use of the stimulus to CS os electrogram (Stim Time) and the activation time from the pacing site to the CS os electrogram (Act Time) as determinates of whether a site is in the reentrant circuit assumes that local conduction velocities are not significantly affected by pacing rates slightly faster than tachycardia rates.19 31 In our study, there were no significant changes in the FCL after pacing, and there were no significant differences in Stim Times at different pacing rates, suggesting that conduction velocities were indeed not affected. Sites anterior to the CT (site 2) and eustachian ridge (sites 5, 7, and 9) had Stim Times equal to Act Times and thus were a part of the circuit, whereas sites posterior to the CT (sites 1 and 3) and eustachian ridge (sites 4 and 6) had Stim Times greater than Act Times and thus were not a part of the circuit (Fig 8B⇑). These findings are consistent with the results from the analysis of the PPIs.
The CT has been previously implicated as a barrier in animal models of atrial flutter.32 33 34 Observations in humans have demonstrated an area of block in the lateral right atrium, but to date the exact anatomic correlate of this electrophysiological event had been unresolved.6 35 36 Cosio et al6 20 35 37 have demonstrated that activation of the septum is in a low-to-high (caudocranial) pattern, whereas activation of the anterior wall is in a high-to-low (craniocaudal) pattern. He and others6 20 35 38 have also reported double potentials recorded along the lateral right atrium similar to those observed in our study. However, the present study demonstrates conclusively, based on ICE, that the CT is the anatomic structure that acts as the barrier to conduction.
Our study also demonstrates that the eustachian ridge forms a barrier to conduction during atrial flutter in the low right atrium; it, along with the tricuspid annulus, forms the borders of a narrow isthmus of conduction. The area of the low right atrium between the tricuspid annulus and the inferior vena cava has been previously demonstrated to be a critical area in atrial flutter.6 9 10 13 16 27 35 With entrainment techniques, Nakagawa et al12 recently reported that a line of block existed between the inferior vena cava and the CS os in two patients. This is consistent with the findings reported in the present study. However, until the present study, the anatomic features that delineate this region, and in particular the role of the eustachian ridge, were not completely known. ICE was used in the present study to identify the eustachian ridge as the anatomic correlate of this line of block extending from the inferior vena cava to the CS os.
The exit of this narrow isthmus between the eustachian ridge and tricuspid annulus is believed to be near the CS os, since electrograms recorded from this site correspond to the onset of the flutter wave.9 10 12 21 27 However, it was not known whether the wave propagates anterior to the CS os (between it and the tricuspid annulus), posterior to the CS os, or both (Fig 9⇑). In the present study, six of eight patients were demonstrated to have an exit site only on the anterior aspect of the CS os (site 9), whereas in two patients the exit was on both sides of the os (sites 8 and 9) (Fig 9⇑). This may have some implications for ablation, since it may be difficult to determine which patients have these “dual” exit sites. Therefore, ablation at the medial aspect of this isthmus, near the CS os, may not be successful in some patients. In fact, in these eight patients, seven were ablated successfully from the lateral approach (including the two with “dual” exits), whereas only one had successful ablation medially.
Although this study has demonstrated that the CT and eustachian ridge are anatomic barriers to conduction during atrial flutter, they may be either fixed or functionally determined. Our study demonstrates that such lines of block, whether fixed or functional, occur along anatomically defined regions. Observations in a few of these patients allow some inference as to whether the block is fixed or functional. During sinus rhythm, in two patients, the CT catheter was activated in a high-to-low sequence, and the split potentials were no longer present. This is not unexpected because the sinus impulse originates superiorly in the CT itself and activates both the anterior wall and septum in a high-to-low pattern. However, pacing from the low right atrium demonstrated late activation of the CT electrodes, after the septal electrodes, and with the reappearance of split potentials along the CT. In addition, pacing from site 4 demonstrated a relatively short time from stimulus to the CS os, whereas pacing from site 5 (just on the anterior side of the eustachian ridge to site 4) demonstrated a much longer Stim Time to the CS os, similar to that during atrial flutter. Although not conclusive, this suggests that the lines of block established by the CT and the eustachian ridge are both fixed, at least in some patients. This is consistent with the findings of Saffitz et al,39 who demonstrated that longitudinal conduction in the CT is 10 times greater than in the transverse direction. This is likely related to the type and distribution of gap junctions in crista myocytes. It is also intriguing to speculate that structural abnormalities of the CT and eustachian ridge, on a microscopic level, are the primary abnormalities in patients with atrial flutter and may explain the occurrence of atrial flutter even in patients with grossly normal atria. This may be analogous to the dog described by Boineau et al33 with congenital thinning in the floor of the right atrium and spontaneous atrial flutter.
From the activation maps obtained in the present study, typical atrial flutter occurs in a counterclockwise rotation (Fig 9⇑). From the CS os, the flutter wave passively activates the interatrial septum and left atrium. After this, the trabeculated right atrium anterior to the CT is activated, directing the impulse into the narrow isthmus of slow conduction between the eustachian ridge and tricuspid annulus. Arribas et al40 have shown that the superior aspect of the circuit is activated along the anterior right atrium, and not along the smooth right atrium below the superior vena cava orifice. With data from their study and data from the present study, the flutter circuit would appear as in Fig 9⇑. How the impulse travels from the exit of the protected isthmus in the low right atrium to the anterior, trabeculated right atrium is not known, but because the septum is not within the reentrant circuit, the impulse must continue around the tricuspid valve. Double potentials have been reported in the area of the triangle of Koch.6 38 In addition, there are reports of an association between atrioventricular nodal reentry tachycardia and inducible atrial flutter, suggesting the possibility of a shared pathway around the tricuspid annulus to the anterior right atrium.15 41 The tendon of Todaro may act as a continued line of block, or the impulse may be more rapidly conducted toward the apex of the triangle of Koch over specialized fibers. Further studies are needed to elucidate this last link in the flutter circuit.
Our data suggest that for successful ablation, a lesion must be applied between the tricuspid annulus and the eustachian ridge. Perhaps the shortest such lesion is a perpendicular one between the tricuspid annulus and the eustachian ridge. However, such a lesion may be difficult to create with current technology because the eustachian ridge is not seen with fluoroscopy. Alternatively, as was done in all patients in this study, a slanted linear lesion can be applied between the tricuspid annulus and the orifice of the inferior vena cava, in a similar anatomic location to that reported by Cosio et al.35 The distance between the tricuspid annulus and the eustachian ridge is variable and can be quite large in some patients. Long pullbacks with a 4-mm-tipped catheter are often required for successful ablation. Catheters with longer tips may be able to produce the necessary lesion with a single radiofrequency application.
Although this study conclusively demonstrated that the CT and eustachian ridge are barriers to conduction, the exact mechanism of block was not identified. Pacing was not performed from all sites in all patients during sinus rhythm. Moreover, pacing was performed only at the FCL during sinus rhythm and not at rates slightly faster than sinus rate. Therefore, the conduction block may be due to disparate refractory periods or use-dependent conduction failure along these anatomic structures rather than being fixed lines of block.
Although this study suggests that ablation directed laterally in the protected isthmus in the low right atrium should be more successful than that directed medially near the CS os, this has not been conclusively shown. The patients were not randomized, and only a small number of patients were included. Moreover, there is no long-term follow-up available on this cohort of patients.
ICE was useful in identifying anatomic features of the right atrial endocardium. With ICE, functional electrophysiological properties were correlated with specific anatomic landmarks. With activation mapping and entrainment guided by ICE, we identified the CT and eustachian ridge as anatomic barriers to conduction during type I atrial flutter. Further study is needed to determine the role of other septal structures, such as the tendon of Todaro and the entire tricuspid annulus in typical atrial flutter and to determine whether the block along the CT and eustachian ridge is fixed or functional.
Selected Abbreviations and Acronyms
|Act Time||=||activation time|
|bpm||=||beats per minute|
|FCL||=||flutter cycle length|
|Stim Time||=||stimulus time|
Dr Olgin is supported by the North American Society of Pacing and Electrophysiology and is the recipient of the Albert S. Hyman NASPE Postdoctoral Fellowship. Dr Kalman is supported by the Ralph Reader Overseas Research Fellowship of the National Heart Foundation of Australia and the Telectronics Traveling Fellowship of the Royal Australian College of Physicians. Dr Lesh is supported in part by National Institutes of Health grant HL-45664.
- Received January 31, 1995.
- Revision received April 26, 1995.
- Accepted May 3, 1995.
- Copyright © 1995 by American Heart Association
Frame LH, Page RL, Boyden PA, Fenoglio JJ, Hoffman B. Circus movement in the canine atrium around the tricuspid ring during experimental atrial flutter and during reentry in vitro. Circulation. 1987;76:1155-1175.
Boyden PA, Frame LH, Hoffman BF. Activation mapping of reentry around an anatomic barrier in the canine atrium: observations during entrainment and termination. Circulation. 1989;79:406-416.
Waldo AL. Atrial flutter: new directions in management and mechanism. Circulation. 1990;81:1142-1143.
Feld GK, Fleck RP, Chen PS, Boyce K, Bahnson TD, Stein JB, Calisi CM, Ibarra M. Radiofrequency catheter ablation for the treatment of human type 1 atrial flutter: identification of a critical zone in the reentrant circuit by endocardial mapping techniques. Circulation. 1992;86:1233-1240.
Touboul P, Saoudi N, Atallah G, Kirkorian G. Electrophysiologic basis of catheter ablation in atrial flutter. Am J Cardiol. 1989;64:795-825.
Nakagawa H, Wang X, McClelland J, Beckman K, Lazzara R, Hazlitt A, Arruda M, Santoro I, Abdalla I, Singh A, Sweidan R, Gossinger H, Hirao K. Line of conduction block extends from inferior vena cava to coronary sinus ostium in common atrial flutter. PACE Pacing Clin Electrophysiol. 1993;16:881. Abstract.
Saoudi N, Atallah G, Kirkorian G, Touboul P. Catheter ablation of the atrial myocardium in human type I atrial flutter. Circulation. 1990;81:762-771.
Lesh MD, Van Hare GF, Fitzpatrick AP, Griffin JC, Chu E. Curing reentrant atrial arrhythmias: targeting protected zones of slow conduction by catheter ablation. J Electrocardiol. 1993;26:194-203.
Waldo AL, McLean WAH, Karp RB, Kouchoukos NT, James TN. Entrainment and interruption of atrial flutter with atrial pacing: studies in man following open heart surgery. Circulation. 1977;56:737-745.
Stevenson WG, Khan H, Sager P, Saxon LA, Middlekauff HR, Natterson PD, Wiener I. Identification of reentry circuit sites during catheter mapping and radiofrequency ablation of ventricular tachycardia late after myocardial infarction. Circulation. 1993;88:1647-1670.
Cosio FG, Lopez GM, Arribas F, Palacios J, Goicolea A, Nunez A. Mechanisms of entrainment of human common flutter studied with multiple endocardial recordings. Circulation. 1994;89:2117-2125.
Chu E, Fitzpatrick AP, Chin MC, Sudhir K, Yock PG, Lesh MD. Radiofrequency catheter ablation guided by intracardiac echocardiography. Circulation. 1994;89:1301-1305.
Kalman JM, Lee RJ, Fisher WG, Chin MC, Lesh MD, Scheinman MM. Radiofrequency catheter modification of sinus node function guided by intracardiac echocardiography. Circulation. 1994;90(suppl I):I-41. Abstract.
Kalman JM, Fitzpatrick AP, Chin MC, Lee RJ, Scheinman MM, Lesh MD. Efficiency of heating with radiofrequency energy is related to stability of tissue contact: evaluation by intracardiac echocardiography. Circulation. 1994;90(suppl I):I-270. Abstract.
Waldo A, Carlson M, Biblo L, Henthorn R. In: Waldo, ed. Atrial Arrhythmias. 1993;210-228.
Okumura K, Henthorn RW, Epstein AE, Plumb VJ, Waldo AL. Further observations on transient entrainment: importance of pacing site and properties of the components of the reentry circuit. Circulation. 1985;72:1293-1307.
Pastelin G, Mendez R, Moe GK. Participation of atrial specialized conduction pathways in atrial flutter. Circ Res. 1978;42:386-393.
Boineau JP, Schuessler RB, Mooney CR, Miller CB, Wylds AC, Hudson RD, Borremans JM, Brockus CW. Natural and evoked atrial flutter due to circus movement in dogs: role of abnormal atrial pathways, slow conduction, nonuniform refractory period distribution and premature beats. Am J Cardiol. 1980;45:1167-1181.
Ferguson TJ, Schuessler RB, Hand DE, Boineau JP, Cox JL. Lessons learned from computerized mapping of the atrium: surgery for atrial fibrillation and atrial flutter. J Electrocardiol. 1993;26:210-219.
Saffitz JE, Kanter HL, Green KG, Tolley TK, Beyer EC. Tissue-specific determinants of anisotropic conduction velocity in canine atrial and ventricular myocardium. Circ Res. 1994;74:1065-1070.
Arribas F, Lopex-Gil M, Nunez A, Puigdueta B, Goicolea A, Cosio F. The upper link of the common atrial flutter circuit. Circulation. 1994;90(suppl I):I-376. Abstract.