Author + information
- Received May 25, 2016
- Revision received September 9, 2016
- Accepted September 12, 2016
- Published online December 21, 2016.
- S2405500X16303383-1b8d1c1816ab1a361822801bd3e46a97Christopher F. Liu, MD∗ (, )
- S2405500X16303383-f8ac21e5196b04ae4006fce3efef6373James E. Ip, MD,
- S2405500X16303383-cf7e5aecfbdbe3631c701787f9188db5Jim W. Cheung, MD,
- S2405500X16303383-d4df08ec044eeb4d29207340b64bd639George Thomas, MD,
- S2405500X16303383-e5c507ecc37e2985d3787b02e0335246Steven M. Markowitz, MD and
- S2405500X16303383-42788dcc14f53b095e52aafabc7d11f6Bruce B. Lerman, MD
- ↵∗Reprint requests and correspondence:
Dr. Christopher F. Liu, Division of Cardiology, Weill Cornell Medical College, 520 East 70th Street, Starr-4, New York, New York 10021.
Objectives This study sought to evaluate the utility of ventriculoatrial (VA) conduction patterns in response to adenosine in predicting inducibility of orthodromic reciprocating tachycardia (ORT).
Background Adenosine is known to consistently block atrioventricular (AV) nodal conduction. We hypothesized that persistent VA conduction despite administration of adenosine would have a high predictive value for identifying the presence of a retrograde accessory pathway (AP) and associated ORT.
Methods A total of 168 patients undergoing electrophysiological study for supraventricular tachycardia (SVT) had assessment of VA conduction during ventricular pacing and adenosine administration. Standard pacing maneuvers were then used for induction and diagnosis of the SVT mechanism.
Results Absence of VA block to adenosine (doses up to 24 mg) had 88% sensitivity and 91% specificity for identifying ORT (positive predictive value 76%, negative predictive value 96%). Four patients with adenosine-induced VA block and inducible ORT had decremental APs. Adenosine caused VA block in 6 patients with eccentric VA activation due to atypical AV nodal conduction, and concentric VA conduction persisted in all 12 patients with a septal AP. Adenosine unmasked free-wall APs in 10 patients by blocking AV nodal conduction, shifting VA activation from concentric to eccentric.
Conclusions The response of VA conduction to adenosine is a highly sensitive and specific method for detecting retrograde AP conduction and inducible ORT. Adenosine-induced VA block rules out inducible ORT due to a nondecremental AP. In cases of VA fusion, adenosine-induced block of AV nodal conduction can delineate the location of the AP atrial insertion site.
During sustained narrow complex tachycardia, the participation of a retrograde accessory pathway (AP) can be demonstrated using well-described pacing maneuvers that differentiate orthodromic reciprocating tachycardia (ORT) from atrioventricular nodal re-entrant tachycardia (AVNRT) and atrial tachycardia (AT) (1–3). Demonstrating the presence or absence of a retrograde-conducting AP may help to narrow the differential diagnosis before induction of tachycardia, and pacing schemes have been described for this purpose (4,5). However, these maneuvers are primarily accurate with respect to septal APs that are associated with more rapid conduction than the atrioventricular (AV) node and have limitations when AP conduction is masked by competing AV nodal conduction (ventriculoatrial [VA] fusion), or when assessing for the presence of free-wall APs. Furthermore, the predictive value of these pacing maneuvers for inducible ORT has not been evaluated.
The multiple effects of adenosine on supraventricular tachycardia (SVT) include a negative dromotropic effect on the AV node (with a lack of effect on typical APs), transient negative chronotropic effect on automatic rhythms, and termination of AT due to cyclic adenosine monophosphate–mediated triggered activity (6). Adenosine is commonly used for rapid termination of AV nodal-dependent SVTs (7), and adenosine-induced PR jump during sinus rhythm suggests dual AV nodal physiology (8). Based on the differential effects of adenosine on the AV node and typical AV APs, we hypothesized that the response to adenosine during ventricular pacing would be instrumental in revealing the presence of a retrograde-conducting AP, particularly in cases with competing AV nodal conduction. In the present study, we sought to systematically assess this utility of adenosine and to determine its predictive value in the diagnosis of ORT.
We prospectively examined 168 consecutive patients who underwent invasive electrophysiological (EP) evaluation for SVT. The presence of structural heart disease was based on diagnosis of cardiomyopathy or significant valvular disease. Coronary artery disease was considered to be clinically significant if there was >60% stenosis or previous myocardial infarction. All procedures were performed following institutional guidelines of the Weill Cornell Medical Center/New York-Presbyterian Hospital, and all patients gave written informed consent. The study was approved by the Institutional Review Board of Weill Cornell Medical College.
All patients underwent EP study in a post-absorptive state. Initial catheters included quadripolar 6-F catheters positioned at the His-bundle position and right ventricular (RV) apex. Right atrial electrogram recordings were obtained with either a quadripolar catheter positioned in the high right atrium or a 7-F duodecapolar catheter positioned along the tricuspid annulus. A 6-F decapolar catheter was positioned in the coronary sinus (CS) to record left atrial and left ventricular activity. Surface and intracardiac electrocardiograms were recorded (Prucka CardioLab EP System, GE Healthcare, Waukesha, Wisconsin).
Step 1: Adenosine assessment of VA conduction
During the baseline state, RV pacing was performed to evaluate for the presence of concentric or eccentric VA activation. In order to isolate the effects of adenosine on retrograde AV nodal conduction and AP conduction, pacing was performed at a cycle length of 600 ms. If VA conduction was not present at baseline, isoproterenol was infused to achieve 1:1 VA conduction. Concentric VA conduction was present when earliest atrial activation occurred along the septum (at the His or CS ostium), whereas VA conduction was considered eccentric when right or left atrial activation preceded septal activation. During RV pacing, adenosine was administered as a rapid bolus through a femoral venous sheath, followed by a 20-ml normal saline flush. The initial dose of adenosine was 12 mg and then 24 mg if VA block was not achieved. Therefore, the response of VA conduction to adenosine was binary: either transient VA block or persistent VA conduction. In the latter case, it was noted whether there was a change in VA interval or a change in VA activation sequence, for example, from concentric to eccentric.
Step 2: Programmed stimulation and pacing maneuvers
After baseline adenosine testing, a standard EP study was performed, which included burst pacing and extrastimulus testing to assess myocardial refractory and conduction properties and to induce sustained SVT. If sustained SVT was not inducible in the baseline state, isoproterenol was infused at 2 μg/min and titrated per operator discretion. Programmed stimulation was performed until sustained SVT was induced. During sustained SVT, standard pacing maneuvers (including ventricular overdrive pacing and ventricular extrastimuli) were performed to establish 1 of 3 SVT mechanisms: AVNRT, ORT, or AT (1–3).
Continuous variables are expressed as mean ± SD or median (interquartile range), depending on normality of distribution. For 2-group comparisons of continuous variables, the 2-tailed Student t test was used for normal data, and the 2-tailed Mann-Whitney U test was used for non-normal data. Comparisons of more than 2 groups were made by analysis of variance (with Scheffé post hoc analysis for pairwise comparisons), or by the Kruskal-Wallis test for non-parametric data. Comparisons of categorical variables were made using the Fisher exact test for 2 groups and the chi-square test for more than 2 groups. Statistical calculations were performed using Medcalc 13.2 (Medcalc Software, Belgium). A value of p < 0.05 was considered statistically significant.
Overall, 85 patients (51%) were male; mean age was 55 ± 17 years. Thirty-eight (23%) had structural heart disease and 28 (17%) had coronary artery disease. Median left ventricular ejection fraction was 60% (interquartile range: 58% to 65%). A total of 175 SVTs were induced: 87 (50%) AVNRT, 32 (18%) ORT, and 56 (32%) AT. Eleven patients (6.5%) had >1 inducible SVT (5 had AVNRT and AT, 6 had multiple ATs). Twenty-four patients (14%) had manifest pre-excitation. The clinical characteristics of the 168 patients are summarized in Table 1. Thirteen patients (7.7%) had a concealed AP. Compared with patients having AVNRT or AT, those with ORT were younger (p < 0.001) and were more often male (p < 0.001).
VA Conduction Response to Adenosine
A total of 145 patients had evaluation of VA conduction (patients who came to the laboratory in SVT were excluded from this phase of the protocol). Isoproterenol was required to facilitate consistent VA conduction in 9 patients, 3 of whom ultimately had ORT using a decremental AP. Excluding the 5 patients without VA conduction despite isoproterenol and the 1 patient with severe asthma, adenosine was given to 139 patients during RV pacing. The responses were interpretable in 134 (2 were excluded because of frequent ectopy or echo beats, and 3 were excluded because of initiation of atrial fibrillation, precluding definitive assessment of VA response). Adenosine 12 mg caused VA block in 85 (63%) patients. The remaining 49 (37%) patients were given adenosine 24 mg. Taken together, adenosine (up to 24 mg) caused VA block in 97 (72%) patients (Figure 1), 96% (93 of 97) of whom did not have inducible ORT (77 had AVNRT, and 16 had AT only). Persistent VA conduction was seen in the remaining 37 (28%) patients, 76% (28 of 37) of whom were ultimately confirmed to have inducible ORT. Of the 9 patients with persistent VA conduction after adenosine but no inducible ORT, all had inducible AVNRT. Therefore, with an adenosine dose up to 24 mg, the persistence of VA conduction was 88% sensitive and 91% specific for inducible ORT, with a positive predictive value of 76% and negative predictive value of 96%.
Four patients (12.5%) with inducible ORT demonstrated VA block with adenosine (2 with adenosine 12 mg and 2 with adenosine 24 mg). These 4 patients had decremental APs. One was a septal AP that was partially ablated during a prior procedure (previously non-decremental); the remaining 3 cases were left free-wall APs with inconsistent retrograde AP conduction at baseline and required isoproterenol to elicit reliable conduction.
VA Activation Patterns and Adenosine-Induced Changes
Before adenosine administration, VA conduction was concentric in 117 patients (87%) and eccentric in 17 (13%). Among the 108 patients with no inducible ORT, 102 (94%) had concentric VA activation, whereas 6 (6%) patients had eccentric VA activation pattern (earliest activation in the proximal–mid CS). All 6 patients demonstrated VA block with adenosine, and all were demonstrated to have AVNRT with a left-sided retrograde atrionodal extension. Three of these tachycardias had a short RP interval (Figure 2), whereas 3 had a long RP interval (Figure 3).
Among the 32 patients with ORT, AP locations were septal in 12 (38%), left free wall in 16 (50%), and right free wall in 4 (12%). Overall, 62% (20/32) of the patients with ORT had concentric VA activation at baseline, whereas 38% (12/32) had eccentric VA activation. Of the 20 patients with ORT and baseline midline conduction, adenosine induced a change in VA pattern from concentric to eccentric in 50% (10 patients: 7 with left lateral APs, 2 with left posterior APs, and 1 with a right free-wall AP). Depending on the relative conduction velocities of the retrograde AV node and AP, 2 patterns were seen in these patients before adenosine effect: concentric VA pattern (activation via the AV node) (Figure 4A) or fused VA pattern (activation via the AV node and AP) (Figure 4B). Both of these patterns changed to clear eccentric pattern with adenosine effect because of adenosine’s negative dromotropic effects on AV nodal conduction.
All 12 patients with septal APs had persistent concentric VA conduction with adenosine. However, in 2 of these patients, adenosine caused a clear delay in the atrial timing in the His-bundle recording (Figure 5), suggesting block in the retrograde AV nodal fast pathway, and unmasked a relatively slower-conducting septal AP. All 12 patients with septal APs had inducible ORT.
ECG Pre-excitation and Retrograde AP Conduction
Of the 24 patients with baseline manifest pre-excitation, 5 (21%) were found to have VA block with adenosine; all 5 had no inducible ORT and no evident retrograde AP conduction seen with pacing maneuvers. Two of these 5 anterograde-only APs were ablated because of high-risk features (short anterograde effective refractory period [ERP] of the AP and/or rapid AP conduction during atrial fibrillation).
Atrial fibrillation occurred after adenosine administration in 6 patients (4.3%): 2 patients with AVNRT, 3 patients with ORT, and 1 patient with AT. All episodes were hemodynamically stable. Three of these patients (2.2%) required uncomplicated intraprocedural direct current cardioversion. No other adverse reactions requiring treatment occurred.
The principal findings of this study are that persistent VA conduction with up to 24 mg of adenosine has approximately 90% sensitivity and specificity for identifying a retrograde-conducting AP and inducible ORT. The corollary also is true, that is, adenosine-induced VA block has a high negative predictive value for non-inducible ORT. Finally, adenosine-induced retrograde block of AV nodal conduction can unmask and simplify localization of retrograde AP conduction patterns in cases with VA fusion. A proposed algorithm incorporating adenosine testing of VA conduction in the diagnosis of SVT is summarized in Figure 6.
Persistent VA Conduction With Adenosine
A fundamental element of any SVT diagnostic protocol is to determine whether a retrograde-conducting AP is present. This is commonly tested by observing for non-decremental eccentric VA activation (9), as well as extranodal responses to differential (basal vs. apical) RV pacing (4) and/or para-Hisian pacing (5). However, these maneuvers have recognized limitations. The accuracy of basal versus apical pacing and para-Hisian pacing for identifying an extranodal AP depends on proximity of the AP to the basal pacing site as well as the relative conduction velocities of the AP and AV node. Therefore, these maneuvers are less accurate in the setting of rapid retrograde AV nodal conduction or with left free-wall APs (5). In the present study, we showed that persistent VA conduction in response to adenosine is a highly sensitive and specific method for detecting the presence of a retrograde-conducting AP and inducible ORT. These results were applicable to all AP locations and were independent of retrograde VA velocities over the AV node or AP. Furthermore, to our knowledge, this is the first study that evaluated the association of a baseline maneuver suggestive of AP conduction with inducibility of ORT.
Our protocol was designed to simplify the approach to adenosine dosing, with the understanding that variable doses are required to block the AV node among subjects (10). A small subset of patients with AVNRT were found to have persistent VA conduction despite a large dose of adenosine (24 mg). This may be due to relative adenosine resistance of the retrograde fast AV nodal pathway compared with the anterograde fast pathway, which has been observed infrequently in patients with AVNRT (11). Notably, the 3 cases of atypical decremental APs in this study were adenosine sensitive, as was the case of an AP injured by previous ablation. These findings are also consistent with the results from previous studies of these rare APs (12,13).
Eccentric VA Activation Sequence Due to AV Nodal Conduction
The presence of eccentric VA activation has been considered a simple and reliable indicator of an AP (9). However, in atypical forms of AVNRT, retrograde conduction over a slow AV nodal pathway often demonstrates eccentric VA activation (14). Less commonly, retrograde conduction over a left-sided fast AV nodal pathway connection can also be observed (15–17). In the present prospective study, 6% of patients without an AP were found to have an eccentric VA activation sequence; all cases demonstrated VA block with adenosine. Therefore, excluding the rare exception of a decremental AP, adenosine-induced block of eccentric VA activation rules out the presence of an AP. This finding may be particularly helpful in cases in which results from ventricular extrastimuli or ventricular overdrive pacing are difficult to interpret or provide conflicting information (Figure 3).
Adenosine Unmasks Retrograde AP Presence and Location
During ventricular pacing, AP conduction may not be apparent when AV nodal conduction is rapid and the atrium is either entirely or predominantly activated via the AV node. This masking effect of retrograde AV nodal conduction is a major limitation of basal versus apical pacing and para-Hisian pacing (4,5). Ventricular extrastimulus testing may produce progressive decrement in the AV node and allow for demonstration of AP conduction. However, if the retrograde ERP of the AP is longer than the retrograde AV nodal ERP, the AP would block before the AV node, leaving the issue unresolved (18). In the present study, we found 2 patterns of adenosine unmasking of AP conduction. With free-wall APs, adenosine block of retrograde AV nodal conduction caused VA activation to change from concentric or fused to a clearly eccentric pattern, revealing AP conduction and a free-wall location (Figure 4). With septal APs, adenosine block of the AV node caused delay in atrial activity recorded on the His-bundle catheter and clarified the retrograde pattern over the AP (Figure 5). Importantly, all changes in VA activation were observed in association with prolongation of the VA interval in the His-bundle recording, suggesting block in retrograde AV nodal conduction. These findings illustrate the unique value of adenosine to transiently and selectively block the AV node in order to assess for AP conduction and location without additional pacing maneuvers. Furthermore, when baseline VA fusion is demonstrated with adenosine, the optimal approach to ablating the retrograde AP should involve mapping either during ORT or during RV pacing at rates that produce block in the AV node.
Although adenosine-induced AV nodal blockade may demonstrate or unmask a retrograde AP, this finding does not prove AP participation in SVT. Although all cases of retrograde-conducting AP in this study were ultimately found to be implicated in ORT, there may be cases in which an AP does not participate in SVT or the patient may have inducible AVNRT or AT in addition to ORT.
As discussed earlier, decremental APs were found to be adenosine sensitive and therefore are a notable but rare exception to the negative predictive value of VA block (no inducible ORT). Another limitation is that doses of adenosine >24 mg were not administered to the 37 patients with persistent VA conduction (9 without ORT). It is possible that higher doses would have resulted in AV nodal block in some patients whose retrograde nodal conduction was relatively resistant to adenosine, and the positive predictive value of adenosine could have been further enhanced. Other theoretical concerns regarding routine adenosine use include potential bronchospasm and coronary steal (none was seen in this study). Finally, our study did not directly compare the results of adenosine testing with other pacing maneuvers for diagnosis of retrograde AP conduction, with the understanding that there is no universally recognized gold standard maneuver. As is the case with other pacing maneuvers, the response to adenosine testing should be viewed as one finding among others, which may support or disprove the presence of an AP.
COMPETENCY IN MEDICAL KNOWLEDGE: During invasive EP diagnosis of SVT, adenosine-guided assessment of VA conduction predicts ORT inducibility and facilitates AP localization by blocking competing AV nodal conduction.
TRANSLATIONAL OUTLOOK: Future studies should directly compare the accuracy of adenosine testing with other pacing maneuvers using a randomized approach.
The authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- accessory pathway
- atrial tachycardia
- atrioventricular nodal re-entrant tachycardia
- coronary sinus
- effective refractory period
- orthodromic reciprocating tachycardia
- right ventricle
- supraventricular tachycardia
- Received May 25, 2016.
- Revision received September 9, 2016.
- Accepted September 12, 2016.
- American College of Cardiology Foundation
- Knight B.P.,
- Ebinger M.,
- Oral H.,
- et al.
- Gonzalez-Torrecilla E.,
- Arenal A.,
- Atienza F.,
- et al.
- AlMahameed S.T.,
- Buxton A.E.,
- Michaud G.F.
- Martinez-Alday J.D.,
- Almendral J.,
- Arenal A.,
- et al.
- Hirao K.,
- Otomo K.,
- Wang X.,
- et al.
- Lerman B.B.,
- Belardinelli L.
- Glatter K.A.,
- Cheng J.,
- Dorostkar P.,
- et al.
- Toal S.C.,
- Vajifdar B.U.,
- Gupta A.K.,
- Vora A.M.,
- Lokhandwala Y.Y.
- Gallagher J.J.,
- Gilbert M.,
- Svenson R.H.,
- Sealy W.C.,
- Kasell J.,
- Wallace A.G.
- DiMarco J.P.,
- Sellers T.D.,
- Lerman B.B.,
- Greenberg M.L.,
- Berne R.M.,
- Belardinelli L.
- Efimova E.,
- Riahi S.,
- Fiedler L.,
- et al.
- Lerman B.B.,
- Greenberg M.,
- Overholt E.D.,
- et al.
- Chen S.A.,
- Tai C.T.,
- Chiang C.E.,
- et al.
- Hwang C.,
- Martin D.J.,
- Goodman J.S.,
- et al.
- Engelstein E.D.,
- Stein K.M.,
- Markowitz S.M.,
- Lerman B.B.
- Ip J.E.,
- Dobesh D.P.,
- Liu C.F.,
- Lerman B.B.