- Email: [email protected]
Role of concealed and supernormal conductions during atrial fibrillation in the preexcitation syndrome
During Atrial Fitkillation in the Preexcitation Syndrome Peng-Sheng Chen, MD, and Eric N. Prystowsky, MD
The hypothesis that retrograde concealed and anterograde supernormal conductions over the accessory pathway are determinants of anterograde accessory pathway conduction during atrial fibrillation (AF) was prospectively tested. For 17 patients during AF, 90 f 13 (mean f standard deviation) consecutive beats were analyzed for morphology and VV interval preceding each QRS complex. In 15 patients with both preexcited and normal QRS complexes, after normal complexes, the cycle length preceding the first preexcited QRS complex was 389 f 69 ms, which was longer than the average cycle length of consecutive preexcited complexes (325 f 55 ms; p = 0.001). The mean difference was 61 f 53 ms. After preexcited QRS complexes, the cycle length preceding the first normal QRS complex averaged 423 f 65 ms, which was significantly longer than the average cycle length of 345 f 47 ms between consecutively conducted normal QRS complexes (p
here are >4 factors that may affect accessory pathway conduction during atria1 fibrillation (AF): accessorypathway refractorinessl; concealed conduction2m4;supernormal conductiorG; and irregularity of the atria1 input to the accessorypathway.7 The ability of concealed and supernormal conductions to influence anterograde conduction over the accessorypathway during AF has not been studied in detail. Evidenceto support a role for concealedconduction is the occurrence of consecutive runs of normal QRS complexesinterspersedbetween runs of preexcited QRS complexes.2One could postulate that repetitive retrograde conduction into the normal atrioventricular (AV) conduction systemafter activation of the ventricle over the accessory pathway could yield a series of preexcited QRS complexes,and a run of normal QRS complexescould occur along with retrograde concealment in the accessorypathway after activation of the ventricle over the normal conducting system. If this is true, then the ventricular cycle length at the transition from the first to the secondQRS morphology should be longer than the averageventricular cycle length of the secondQRS morphology to allow time to recover from the effects of concealed conduction. Regarding supernormality, if an accessorypathway conducts only during the supernormal period, then the preceding cycle length closedby the appearanceof this accessorypathway should be confined to a “phase” of supernormality,5>6rather than varying according to the atria1 input. To test these hypotheses,we prospectivelyanalyzed the morphology and preceding cycle lengths of each QRS complex during AF in 17 patients with preexcitation syndrome who had AF induced during electrophysiologic study. We found data to support concealed and supernormal conductions as important factors in accessory pathway conduction during AF.
T
METHODS From the Division of Cardiology, Department of Medicine, Duke University Medical Center, Durham, North Carolina. This study was performed during the tenure of Clinician ScientistAward 88-414from the American Heart Associationto Dr. Chen. Manuscript receivedJune 7, 1991;revisedmanuscript receivedand acceptedJuly 12, 1991. Addressfor reprints: Peng-ShengChen, MD, Division of Cardiology, Department of Medicine, 8411, UCSD Medical Center, 225 DickinsonStreet, San Diego, California 92103-8411,
Patients with preexcitation syndrome undergoing electrophysiologicstudy at our institution between May 1987 and May 1988 were prospectivelyinvestigated. In all, 17 patients fulfilled the following entrance criteria and were included in the study: (1) AF lasted long enough to have 260 QRS complexes available for analysis;and (2) intraoperative epicardial or endocardi-
ACCESSORY PATHWAY CONDUCTION DURING ATRIAL FIBRILLATION
132%)
es, and between consecutivenormally conducted QRS complexes and between consecutive preexcited QRS complexeswere obtained. T tests were usedto compare the differencesbetween the means.g
TABLE I Patient Characteristics No. of Accessory Pathways
No. of pts. Men/women History of syncope Clinical atrial fibrillation Shortest preexcited RR interval (ms) Accessory pathway ERP Total complexes analyzed Narrow QRS Wide QRS Preexcited by dominant AP Preexcited by other AP
1
2
9 613 3 3
8 612 4 3
201 i: 25 (158-234)
208 -+ 24 (167-240)
249 f 28 (215-290)*
250 + 20 (220-280)
821 (average 91 +- 11)
706 (average 88 k 15)
103 (12.5%)
59 (8.4%)
718 (87.5%)
592 (83.8%)
0 (0%)
55 (7.8%)
*Not obtained in 1 patient owing to recurrent atrial fibrillation AP = accessory pathway: ERP = effective refractory period.
during the study.
al mapping studies, or both, to confirm the location of the accessorypathway(s). Electrophysiologic evaluation: Written and informed consent was obtained before each study. The methods of study of patients with preexcitation syndrome at our institution have been previously described in detail.* In brief, patients were studied in the postabsorptive, nonsedatedstate, and all antiarrhythmic medications were discontinued for 25 drug half-lives. Five surface electrocardiographic (I, II, III, Vi and Vg) and 9 intracardiac leads were routinely recorded. Refractory periods and conduction characteristics of the AV node and the accessorypathway(s), as well as atria and ventricle, were obtained, and atria1 mapping to locate the site of the accessorypathway was performed during reentrant tachycardia and right ventricular pacing. AF was induced by rapid atria1 pacing if it did not occur spontaneously during the course of the study. Data were simultaneously recordedon a TEAC XR-5 10 cassette data recorder and written at 100 mm/set using a Gould ES- 1000 16 channel recorder. Data analysis: After the study, the data were examined to select a period during AF that had the most variations of QRS morphologies. Consecutiveintervals were then measuredbetweenthe ventricular activations recorded by the right ventricular catheter. We choseto measure the W rather than RR interval, becausethe beginning of the QRS complex may be indistinct and the peak of the QRS complex may vary during AF, especiallyin patients with multiple accessorypathways. For each patient, the maximums, minimums, means and standard deviations of the W intervals between the transitions from normal to preexcited QRS complexes and from preexcited to normal QRS complex1330
THE AMERICAN JOURNAL OF CARDIOLOGY VOLUME 68
RESULTS There were 17 patients (12 men and 5 women) included in the study. An average of 90 f 13 (mean f standard deviation) QRS complexeswere analyzed for each patient. Among the beats analyzed, 11% were nonpreexcited and 89% were preexcited complexes. Multiple accessorypathways were presentin 8 patients. The other 9 patients had 1 accessorypathway, and none of these 9 had multiple wide QRS complex morphologiesduring AF. Characteristics of the patients are listed in Table I. Analysis of concealed conduction: Data that support the presenceof concealed retrograde conduction came from analysisof the results of the preceding cycle length and its relation to the anterograde QRS morphology in 15 patients with both normal and preexcited QRS complexesduring AF. In patients with >l preexcited QRS morphology, only the dominant morphology was included. Figure 1 demonstrates these measurements during AF in a patient with a posteroseptalaccessorypathway. The wide QRS complexes(initial and last 3 beats) are preexcited. The narrow QRS complexes (middle 8 beats) are not preexcited. The numbers below the QRS complexes represent the intervals between consecutive right ventricular activations. The preceding cycle length of beat no. 2 is 220 ms, that of beat no. 3 is 190 ms, and so forth. Note that the W intervals during the transitions from the preexcited to the normal complex and from the normal to the preexcited complex were longer than other W intervals on the same tracing. Fifteen patients with both normal and preexcited QRS complexespresent during AF were included for analysis (Table II). Two patients were not included because 1 of the pathways conducted only during the supernormal period. A total of 56 episodeswere observed when nonpreexcited preceded preexcited QRS complexes, and 48 were identified when preexcited prececled nonpreexcited complexes.The interval between the transition from normal to preexcited QRS complexes averaged389 f 69 ms, which was significantly longer than the averagecycle length of 325 f 55 ms between consecutivepreexcited QRS complexesin the sameepisodes. Mean difference was 61 f 53 ms. However, these results did not imply that the VV intervals at transition were always longer than the maximal consecutively preexcited VV intervals. On the contrary, in 13 of 15 tracings 11 consecutivelypreexcited VV interval was longer than the minimal VV interval observed at NOVEMBER 15, 1991
the transition from normal to preexcited complex in the TABLE II Preceding Cycle Lengths at Transition Between same tracing. Preexcited and Nonpreexcited Complexes In 3 of 15 tracings, only 1 normally conductedbeat Average Dominant was seen;therefore, measuring the W interval between Preexcited AP Patient Normal to normally conducted beats was not possible. In the reLocation Preexcited Preexcited to Normal Normal No. maining 12 tracings, the cycle lengths at transition 1 525 439 520 447 PS from the preexcited to normal complex averaged423 f 2 332 269 473 325 RF 65 ms, which was significantly longer than the average 3 385 291 -* -* LF 4 305 343 373 345 IS cycle length of 345 f 47 ms between consecutivenor5 348 324 329 301 LF mally conducted complexes (Table 11). Mean differ6 391 241 -* -* LF 7 362 309 362 358 LF encewas 86 f 45 ms. In 8 of 12 tracings, > 1 consecu8 295 261 384 318 PS tive normally conducted W interval was longer than 9 470 359 447 333 RF the minimal VV interval observed at the transition 10 423 380 383 333 PS 11 444 377 545 404 LF from preexcited to normal complex in the sametracing. 12 347 370 442 375 PS Despite the statistically significant differences be13 492 345 -* -* LF tween the preceding W intervals at the transition from 14 392 312 425 337 PS 15 323 260 391 269 LF normal to preexcited complexesand those of preexcited Mean 2 SD 389269 3252557 423-c65 3452471: complexes,exceptionswere found in patients 4 and 12 *As only 1 normally conducted beat was observed, the average of the normal beats is (Table II). In these 2 patients, we found a shorter prenot available for comparison. tp
FIGURE 1. Episode of atrial fibrillation observed in patient with posteroseptal accessory pathway. Surface electrocardiographic leads I, II, Ill, Vl and Vg, and right ventricular (RV), high right atrium &IRA), His bundle (HBE) and proximal conmary sinus (PCS) electrograms were simultaneously recorded. Numbers represent intervals between right ventricular electrograms in ms. See text for details.
ACCESSORY PATHWAY CONDUCTION DURING ATRIAL FIBRILLATION
1331
length similar to or less than the shortest QRS complexes of the dominant morphology, and were associated with narrow rangesof preceding W intervals that varied between 355 and 360 ms in patient 16 (Figure 2), and 181 and 185 ms in patient 17 (Figure 3). This narrow range of W intervals is compatible with supernormal conduction.5$6
ed to normal complexes.These values representin part the contribution of retrograde concealment on subsequent AV conduction through the accessorypathway and AV node, respectively. The large standard deviation can be explained by the fact that the effects of retrograde concealmentare determined not only by the timing of retrograde activation in the accessorypathway and AV node, but also by the timing of the atria1 input in the accessorypathway and AV node. In addiDISCUSSION Major findings of the study: In this study we dem- tion, repetitive anterograde concealment of atria1 input onstrated that during AF, the transition from either the to the accessorypathway and AV node may further normal to preexcited or the preexcited to normal QRS influence the conduction over these pathways, resulting complex was associatedwith longer W intervals than in large variations of the preceding cycle lengths. Obvithe mean VV intervals. Assuming the larger intervals ously, there are multiple mechanisms that determine were due solely to retrograde concealment, which of whether anterograde conduction will occur over the course we can only hypothesize, retrograde conceal- normal systemor accessorypathway, or both, and it is ment prolonged the W interval by 61 f 53 ms during impossible to determine for any given beat the importhe transition from normal to preexcited complexes, tance of retrograde concealedconduction or other facand by 86 f 45 ms during the transition from preexcit- tors that govern which system will be used for ante600
!
,
G P B a L
0 500
-
.
E
.
>
6
P B ti E Q
400
.
l
9
.
-
co
.
..
. ’
l
al, l ‘.” d . . . ..*-% . 0 as*amw#. .. . . A A
300 0
I
I
20
40
6 0 A A
*
PS NORMAL FUSION LL
60
FIGURE 2. Preceding W intervals of consecuttve QRS complexes during ahial fibdiation in patient 16. Note that at transition of QRS complex morphdogi from that associated with posterosepta accessory pathway to normal QRS complex and from normal QRS complex to that associated with posteroseptal accessory pathway, preceding cycle lengths are lamger than average. Shorter than average VV intervals preceding comptexes conducted through iefl lateral accessory pathway are compatible with supemonn al conductiin. LL = teft lateral; PS = posteroseptal.
CONSECUTIVE COMPLEXES
400
I
z.3 I B
0 0
a z E
0
“00 300-
O0 0 OO
> 4
0 200
0
00 OS,O&O&&
$?@O,mOBo 0 o”
o0
00
oa 08 0oe
:oo”,
0,”
0
@O
3
00
d
z E ti I
ce:
100
0 LP 0 LL
I 0
SO
CONSECU’RVE 1332
0 oo,@o
0 O0
100
COMPLEXES
THE AMERICAN JOURNAL OF CARDIOLOGY VOLUME 68
NOVEMBER 15, 1991
FIGURE 3. Preceding W interval of consecutive QRS complexes during at&l fibrillation in patient 17. No normally conducted QRS complexes present. Majority of QRS compiexes are compatibie with conduction over left lateral free wall accessery pathway. Two QRS complexes had morphotogy compatibte with conduction over left posterior free wall accessory pathway. These QRS compiexes were observed only with short preceding cycle lengths, compattbte with supernormal conduction. LL = left lateral; LP = left postenor.
regrade conduction Becausemultiple mechanismsare mality is demonstrablein variousexcitabletissue,14-16 involvedin accessorypathway conduction,the transi- one might expectto observethis phenomenonin accestional VV intervalsmay not alwaysbe longer than the sory pathways. In this study, 2 accessorypathwayswere found to nontransitionalW intervalsin individualtracings. conduct with short precediig W intervals.There apAtrPoventricular GorduGtion during atrial fibriNa* tiGne AF is associated with irregular ventricular pearedto be a phaseof supernormalityof only 4 to 5 responsesin patientswith and without accessorypath- ms duration.Thesefindingsare compatiblewith superways. The mechanismof the irregularity is not com- normal conduction over the accessorypathway durpletely understood,If the conductionis purely deter- ing AF, Other possible causes of multiple morphologies obmined by the refractory periodsof the normal system and accessorypathway, then during AF, the ventricles served during atrial fibrillation: Ventricular premature shouldbe activatedshortly after eachmomentat which complexesand aberrant conductioncan result in wide the respectiveconduction system becomesexcitable, QRS complexbeatsresemblingpreexcitation.Without Becausethe frequency of stimuli applied to the AV direct accessorypathwayrecordings,it is not possibleto node and accessorypathway is high, the cycle length rule out this possibilitycompletely.However,there are variation should be small. However,this clearly is not 22 observationsthat stronglysuggestthat the different what has been observedduring AF. Experimentalevi- wide complexmorphologiesin patients 16 and 17 repdencedemonstratesthat anterogradeconcealedconduc- resentedconductionover 2 accessorypathways. First, tion is also a factor that determinesthe conductionof the nondominantwide QRS complexmorphologywas the AV node,I *J2 In additionto AV noderefractoriness consistentwith a distinctivepreexcitedmorphology.17 and anterogradeconcealedconduction,the directionof Second,at surgery, a secondaccessorypathway was atria1 input to the AV node may also be a factor in identified that was located at the site predicted at electrophysiologicstudy. Further researchof accessodeterminingAV conduction.I3 The aforementionedstudies”-‘3 were performedin ry pathway conductionshould use computerizedmap patients or animals with conductionover the normal ping techniquesto demonstrateventricular preexcitaAV ndeiHis Purkinje system,In patientswith pmexci- tion overthe secondaccessorypathway and to correlate tation syndrome,there is 4 1 accessorypathway pres- conductionwith precedingcyclelengthsand atria1actient. Thus, in addition to the abovefactors, an inter- vation proximal to the accessorypathway. Thesestudaction between the normal and abnormal pathways ies shouldhelp to identify the mechanismsthat govern through retrogradeconcealedconductionmay alsobe a accessorypathway conductionduring AF. Study limitation: One major limitation of this study factor that determinesthe ventricular responseduring AI?. Although there are good data to support the hy- is that thesepatientsmay not be representativeof all pothesisthat retrograde concealedconduction is de- patients with preexcitationsyndrome.These patients monstrableduring regular rhythm,2-4there is little evi- were referredto surgeryfor symptomaticsupraventiicdenceto support the claim that this factor is operative ular arrhythmias and thus representa highly selected during AF. The data presentedin this report are impor- patient cohort.There was a high proportionof multiple tant becausethey lend support to the hypothesisthat pathways. It is thus possiblethat the results of this retrogradeconcealedconductionis an important factor study may be biasedby the selectionprocess. that determines the anterograde accessorypathway Clinical significance: The single most important conduction. prognosticindicator in patientswith preexcitationsynSupemormirlGonduetiom Supernormalconduction drome is the shortestpreexcitedRR interval during over an accessorypathway is a well-recognizedphe- AF.18Becauseof the imperfectcorrelationbetweenthe nomenon5JjIn thesereports,a “phase” of supernormal shortestpreexcitedRR intervaland the anterogradeefconductionwasusuallypresent.This phaseis definedas fective refractory periods of the accessorypathway,18 a rangeof At-A:! intervalsresultingin conductionover factors other than accessorypathway refractoriness the accessorypathway inside the refractory period of probably contribute to conductionover the accessory the latterb6In patientswith manifestventricularpreex- pathway during AF and therefore alter the potential citation, the supernormalphase can be demonstrated risk of these patients.In this report, we summarized by progressivelyshorteningthe AI-A2 intervaland ob- our observationson accessorypathway conductiondurservingthat the QRS complexafter the A2 is preexcit- ing AF in patientswith preexcitationsyndromereferred ed with relativelylong AL-A2 intervalsand nonpreex- for surgicaltreatment.Although there was a significant cited with shorterintervalswhenthe accessorypathway correlation between accessorypathway refractoriness “effectiverefractory period” is reached,and that with and the shortestRR intervalduring AF,r other factors further shorteningof the Al-A2 intervalsthere is recur- such as retrogradeconcealmentand supernormalconrence of ventricular preexcitation.5Becausesupernor- duction are also important in determining accessory ACCESSORY PATHWAY CONDUCTION DURING ATRIAL FIBRILLATION
1333
pathway conduction, The role of supernormality in uncovering an otherwise undetected accessorypathway is important if operative therapy is contemplated. Aclm~wledgment: We wish to thank Richard Page, MD, and Joyce Pressley,MPH, for their assistancein collecting the data, The surgical ablation of the accessory pathways was performed by James Lowe, MD, and his associates.
REFERENCES 1. Wellens HJ, Durret D. Wolff-Parkinson-White syndromeand atrial fibrillation, relation betweenrefractory period of accessorypathway and ventricular rate during atria1 fibrillation. Am J Cardiol 1974;34;777-782. 2. Klein GJ, Yee R, SharmaAD. Concealedconductionin accessoryatrioventricular pathways; an important determinant of the expressionof arrhythmias in patietlts with Wolff-Parkinson-White syndrome.Circulation 1984;70:402-411. 3, Prystowsky EN, Pritchett ELC, Gallagher JJ. Concealedconductionpreventing anterogradepreexcitation in Wolff-Parkinson-White syndrome.Am JCardiol 1984;53:960-961.
4, Svinarich JT, Tai DY, Mickelson J, Keung E, Sung RJ. Electrophysiologic demonstration of concealed conduction in anomalous atrioventricular bypass tract% f Am Co11 Cardiol 1985:5:898~903. 5, Chang M-S, Miles WM, Prystowsky EN. Supernormalconductionin acccssory atrioventricular comiections.Am J Cardial 1987;59:852-856. & Przybylski J, Chiale PA, SanchezRA, Pastori JD, FrancesHG, Elizari MV, RosenbaumME. Supernormalconduction in the accessorypathway of patients
1334
THE AMERICAN JOURNAL OF CARDIOLOGY VOLUME 68
with overt or concealedventricular pre-excitation. J Am Call Cardiol 1987;9: 1269-1278.
7. Kirsh JA, SahakianAV, BaermanJM, Swiryn S. Ventricular responseto atria1 fibrillation: role of atrioventricular conduction pathways. J Am Colt Card!01 1988;12:1265-1272. 8. Prystowsky EN. Diagnosisand managementof the preexcitation syndromes, Curr Probl Cardiol 1988;13:227-310.
9. Wilkinson L. SYSTAT: The Systemfor Statistics. SYSTAT: Evanston,Illinois, 1988673-675. 10. Gallagher JJ, Selle JG, SvensonRH, Fedor JM, Zimmern SH, Scaly WC, Rob&k FR. Surgical treatment of arrhythmias. Am j Cardiol 19&8;611 2lA-44A. 11. SbderstrGmN. What is the reasonfor the ventricular arrhythmia in oasesof auricular fibrillation. Am Hearf J 1950;40:212. 12, Moe GK, Abildskov JA. Observationson the ventricular dysrhythmia associated with atrial fibrillation in the dog heart. Circ Res 1964;14:447-460, 13. JanseMJ. Influence of the direction of the atrial wave front on A-V nodal transmissionin isolated hearts of rabbits, &r Res 1969;25:439-449. 14. Hoff HE, Nahum LH. The supernormalperiod ih the mammalianventricle. Am J Physiol 1938;124:591-595. 15. Spear JF, Moore EN. Supernormalexcitability and conductionia the His= Purkinje systemof the dog. Circ Res 1974;35:782-792. 16. Soloff LA, Fewell WJ. The supernormalphaseof ventricular excitation in man. Its bearing on the genesisof ventricular premature systoles,and a note on atrioventricular conduction.Am &wrt J 1960;59:869-874. 17. Fananapazier L, German LD, Gallagher JJ, Lowe JE, Ptystowaky EN, Importance of preexcited QRS morphologyduring induced atria1 fibrillation to the diagnosis and localization of multiple accessory pathways. Qrrulatlon 1990;81:578-585. 18. Klein GJ, BashoreTM, Sellers TD, Pritchett ELC, Smith WM, Gallagher JJ. Ventricular fibrillation in the Wolff-Parkinson-White syndrome.N Ettgl J ,wed 1979;301:1080-1085.
NOVEMBER 15, 1991
The hypothesis that retrograde concealed and anterograde supernormal conductions over the accessory pathway are determinants of anterograde accessory pathway conduction during atrial fibrillation (AF) was prospectively tested. For 17 patients during AF, 90 f 13 (mean f standard deviation) consecutive beats were analyzed for morphology and VV interval preceding each QRS complex. In 15 patients with both preexcited and normal QRS complexes, after normal complexes, the cycle length preceding the first preexcited QRS complex was 389 f 69 ms, which was longer than the average cycle length of consecutive preexcited complexes (325 f 55 ms; p = 0.001). The mean difference was 61 f 53 ms. After preexcited QRS complexes, the cycle length preceding the first normal QRS complex averaged 423 f 65 ms, which was significantly longer than the average cycle length of 345 f 47 ms between consecutively conducted normal QRS complexes (p
here are >4 factors that may affect accessory pathway conduction during atria1 fibrillation (AF): accessorypathway refractorinessl; concealed conduction2m4;supernormal conductiorG; and irregularity of the atria1 input to the accessorypathway.7 The ability of concealed and supernormal conductions to influence anterograde conduction over the accessorypathway during AF has not been studied in detail. Evidenceto support a role for concealedconduction is the occurrence of consecutive runs of normal QRS complexesinterspersedbetween runs of preexcited QRS complexes.2One could postulate that repetitive retrograde conduction into the normal atrioventricular (AV) conduction systemafter activation of the ventricle over the accessory pathway could yield a series of preexcited QRS complexes,and a run of normal QRS complexescould occur along with retrograde concealment in the accessorypathway after activation of the ventricle over the normal conducting system. If this is true, then the ventricular cycle length at the transition from the first to the secondQRS morphology should be longer than the averageventricular cycle length of the secondQRS morphology to allow time to recover from the effects of concealed conduction. Regarding supernormality, if an accessorypathway conducts only during the supernormal period, then the preceding cycle length closedby the appearanceof this accessorypathway should be confined to a “phase” of supernormality,5>6rather than varying according to the atria1 input. To test these hypotheses,we prospectivelyanalyzed the morphology and preceding cycle lengths of each QRS complex during AF in 17 patients with preexcitation syndrome who had AF induced during electrophysiologic study. We found data to support concealed and supernormal conductions as important factors in accessory pathway conduction during AF.
T
METHODS From the Division of Cardiology, Department of Medicine, Duke University Medical Center, Durham, North Carolina. This study was performed during the tenure of Clinician ScientistAward 88-414from the American Heart Associationto Dr. Chen. Manuscript receivedJune 7, 1991;revisedmanuscript receivedand acceptedJuly 12, 1991. Addressfor reprints: Peng-ShengChen, MD, Division of Cardiology, Department of Medicine, 8411, UCSD Medical Center, 225 DickinsonStreet, San Diego, California 92103-8411,
Patients with preexcitation syndrome undergoing electrophysiologicstudy at our institution between May 1987 and May 1988 were prospectivelyinvestigated. In all, 17 patients fulfilled the following entrance criteria and were included in the study: (1) AF lasted long enough to have 260 QRS complexes available for analysis;and (2) intraoperative epicardial or endocardi-
ACCESSORY PATHWAY CONDUCTION DURING ATRIAL FIBRILLATION
132%)
es, and between consecutivenormally conducted QRS complexes and between consecutive preexcited QRS complexeswere obtained. T tests were usedto compare the differencesbetween the means.g
TABLE I Patient Characteristics No. of Accessory Pathways
No. of pts. Men/women History of syncope Clinical atrial fibrillation Shortest preexcited RR interval (ms) Accessory pathway ERP Total complexes analyzed Narrow QRS Wide QRS Preexcited by dominant AP Preexcited by other AP
1
2
9 613 3 3
8 612 4 3
201 i: 25 (158-234)
208 -+ 24 (167-240)
249 f 28 (215-290)*
250 + 20 (220-280)
821 (average 91 +- 11)
706 (average 88 k 15)
103 (12.5%)
59 (8.4%)
718 (87.5%)
592 (83.8%)
0 (0%)
55 (7.8%)
*Not obtained in 1 patient owing to recurrent atrial fibrillation AP = accessory pathway: ERP = effective refractory period.
during the study.
al mapping studies, or both, to confirm the location of the accessorypathway(s). Electrophysiologic evaluation: Written and informed consent was obtained before each study. The methods of study of patients with preexcitation syndrome at our institution have been previously described in detail.* In brief, patients were studied in the postabsorptive, nonsedatedstate, and all antiarrhythmic medications were discontinued for 25 drug half-lives. Five surface electrocardiographic (I, II, III, Vi and Vg) and 9 intracardiac leads were routinely recorded. Refractory periods and conduction characteristics of the AV node and the accessorypathway(s), as well as atria and ventricle, were obtained, and atria1 mapping to locate the site of the accessorypathway was performed during reentrant tachycardia and right ventricular pacing. AF was induced by rapid atria1 pacing if it did not occur spontaneously during the course of the study. Data were simultaneously recordedon a TEAC XR-5 10 cassette data recorder and written at 100 mm/set using a Gould ES- 1000 16 channel recorder. Data analysis: After the study, the data were examined to select a period during AF that had the most variations of QRS morphologies. Consecutiveintervals were then measuredbetweenthe ventricular activations recorded by the right ventricular catheter. We choseto measure the W rather than RR interval, becausethe beginning of the QRS complex may be indistinct and the peak of the QRS complex may vary during AF, especiallyin patients with multiple accessorypathways. For each patient, the maximums, minimums, means and standard deviations of the W intervals between the transitions from normal to preexcited QRS complexes and from preexcited to normal QRS complex1330
THE AMERICAN JOURNAL OF CARDIOLOGY VOLUME 68
RESULTS There were 17 patients (12 men and 5 women) included in the study. An average of 90 f 13 (mean f standard deviation) QRS complexeswere analyzed for each patient. Among the beats analyzed, 11% were nonpreexcited and 89% were preexcited complexes. Multiple accessorypathways were presentin 8 patients. The other 9 patients had 1 accessorypathway, and none of these 9 had multiple wide QRS complex morphologiesduring AF. Characteristics of the patients are listed in Table I. Analysis of concealed conduction: Data that support the presenceof concealed retrograde conduction came from analysisof the results of the preceding cycle length and its relation to the anterograde QRS morphology in 15 patients with both normal and preexcited QRS complexesduring AF. In patients with >l preexcited QRS morphology, only the dominant morphology was included. Figure 1 demonstrates these measurements during AF in a patient with a posteroseptalaccessorypathway. The wide QRS complexes(initial and last 3 beats) are preexcited. The narrow QRS complexes (middle 8 beats) are not preexcited. The numbers below the QRS complexes represent the intervals between consecutive right ventricular activations. The preceding cycle length of beat no. 2 is 220 ms, that of beat no. 3 is 190 ms, and so forth. Note that the W intervals during the transitions from the preexcited to the normal complex and from the normal to the preexcited complex were longer than other W intervals on the same tracing. Fifteen patients with both normal and preexcited QRS complexespresent during AF were included for analysis (Table II). Two patients were not included because 1 of the pathways conducted only during the supernormal period. A total of 56 episodeswere observed when nonpreexcited preceded preexcited QRS complexes, and 48 were identified when preexcited prececled nonpreexcited complexes.The interval between the transition from normal to preexcited QRS complexes averaged389 f 69 ms, which was significantly longer than the averagecycle length of 325 f 55 ms between consecutivepreexcited QRS complexesin the sameepisodes. Mean difference was 61 f 53 ms. However, these results did not imply that the VV intervals at transition were always longer than the maximal consecutively preexcited VV intervals. On the contrary, in 13 of 15 tracings 11 consecutivelypreexcited VV interval was longer than the minimal VV interval observed at NOVEMBER 15, 1991
the transition from normal to preexcited complex in the TABLE II Preceding Cycle Lengths at Transition Between same tracing. Preexcited and Nonpreexcited Complexes In 3 of 15 tracings, only 1 normally conductedbeat Average Dominant was seen;therefore, measuring the W interval between Preexcited AP Patient Normal to normally conducted beats was not possible. In the reLocation Preexcited Preexcited to Normal Normal No. maining 12 tracings, the cycle lengths at transition 1 525 439 520 447 PS from the preexcited to normal complex averaged423 f 2 332 269 473 325 RF 65 ms, which was significantly longer than the average 3 385 291 -* -* LF 4 305 343 373 345 IS cycle length of 345 f 47 ms between consecutivenor5 348 324 329 301 LF mally conducted complexes (Table 11). Mean differ6 391 241 -* -* LF 7 362 309 362 358 LF encewas 86 f 45 ms. In 8 of 12 tracings, > 1 consecu8 295 261 384 318 PS tive normally conducted W interval was longer than 9 470 359 447 333 RF the minimal VV interval observed at the transition 10 423 380 383 333 PS 11 444 377 545 404 LF from preexcited to normal complex in the sametracing. 12 347 370 442 375 PS Despite the statistically significant differences be13 492 345 -* -* LF tween the preceding W intervals at the transition from 14 392 312 425 337 PS 15 323 260 391 269 LF normal to preexcited complexesand those of preexcited Mean 2 SD 389269 3252557 423-c65 3452471: complexes,exceptionswere found in patients 4 and 12 *As only 1 normally conducted beat was observed, the average of the normal beats is (Table II). In these 2 patients, we found a shorter prenot available for comparison. tp
FIGURE 1. Episode of atrial fibrillation observed in patient with posteroseptal accessory pathway. Surface electrocardiographic leads I, II, Ill, Vl and Vg, and right ventricular (RV), high right atrium &IRA), His bundle (HBE) and proximal conmary sinus (PCS) electrograms were simultaneously recorded. Numbers represent intervals between right ventricular electrograms in ms. See text for details.
ACCESSORY PATHWAY CONDUCTION DURING ATRIAL FIBRILLATION
1331
length similar to or less than the shortest QRS complexes of the dominant morphology, and were associated with narrow rangesof preceding W intervals that varied between 355 and 360 ms in patient 16 (Figure 2), and 181 and 185 ms in patient 17 (Figure 3). This narrow range of W intervals is compatible with supernormal conduction.5$6
ed to normal complexes.These values representin part the contribution of retrograde concealment on subsequent AV conduction through the accessorypathway and AV node, respectively. The large standard deviation can be explained by the fact that the effects of retrograde concealmentare determined not only by the timing of retrograde activation in the accessorypathway and AV node, but also by the timing of the atria1 input in the accessorypathway and AV node. In addiDISCUSSION Major findings of the study: In this study we dem- tion, repetitive anterograde concealment of atria1 input onstrated that during AF, the transition from either the to the accessorypathway and AV node may further normal to preexcited or the preexcited to normal QRS influence the conduction over these pathways, resulting complex was associatedwith longer W intervals than in large variations of the preceding cycle lengths. Obvithe mean VV intervals. Assuming the larger intervals ously, there are multiple mechanisms that determine were due solely to retrograde concealment, which of whether anterograde conduction will occur over the course we can only hypothesize, retrograde conceal- normal systemor accessorypathway, or both, and it is ment prolonged the W interval by 61 f 53 ms during impossible to determine for any given beat the importhe transition from normal to preexcited complexes, tance of retrograde concealedconduction or other facand by 86 f 45 ms during the transition from preexcit- tors that govern which system will be used for ante600
!
,
G P B a L
0 500
-
.
E
.
>
6
P B ti E Q
400
.
l
9
.
-
co
.
..
. ’
l
al, l ‘.” d . . . ..*-% . 0 as*amw#. .. . . A A
300 0
I
I
20
40
6 0 A A
*
PS NORMAL FUSION LL
60
FIGURE 2. Preceding W intervals of consecuttve QRS complexes during ahial fibdiation in patient 16. Note that at transition of QRS complex morphdogi from that associated with posterosepta accessory pathway to normal QRS complex and from normal QRS complex to that associated with posteroseptal accessory pathway, preceding cycle lengths are lamger than average. Shorter than average VV intervals preceding comptexes conducted through iefl lateral accessory pathway are compatible with supemonn al conductiin. LL = teft lateral; PS = posteroseptal.
CONSECUTIVE COMPLEXES
400
I
z.3 I B
0 0
a z E
0
“00 300-
O0 0 OO
> 4
0 200
0
00 OS,O&O&&
$?@O,mOBo 0 o”
o0
00
oa 08 0oe
:oo”,
0,”
0
@O
3
00
d
z E ti I
ce:
100
0 LP 0 LL
I 0
SO
CONSECU’RVE 1332
0 oo,@o
0 O0
100
COMPLEXES
THE AMERICAN JOURNAL OF CARDIOLOGY VOLUME 68
NOVEMBER 15, 1991
FIGURE 3. Preceding W interval of consecutive QRS complexes during at&l fibrillation in patient 17. No normally conducted QRS complexes present. Majority of QRS compiexes are compatibie with conduction over left lateral free wall accessery pathway. Two QRS complexes had morphotogy compatibte with conduction over left posterior free wall accessory pathway. These QRS compiexes were observed only with short preceding cycle lengths, compattbte with supernormal conduction. LL = left lateral; LP = left postenor.
regrade conduction Becausemultiple mechanismsare mality is demonstrablein variousexcitabletissue,14-16 involvedin accessorypathway conduction,the transi- one might expectto observethis phenomenonin accestional VV intervalsmay not alwaysbe longer than the sory pathways. In this study, 2 accessorypathwayswere found to nontransitionalW intervalsin individualtracings. conduct with short precediig W intervals.There apAtrPoventricular GorduGtion during atrial fibriNa* tiGne AF is associated with irregular ventricular pearedto be a phaseof supernormalityof only 4 to 5 responsesin patientswith and without accessorypath- ms duration.Thesefindingsare compatiblewith superways. The mechanismof the irregularity is not com- normal conduction over the accessorypathway durpletely understood,If the conductionis purely deter- ing AF, Other possible causes of multiple morphologies obmined by the refractory periodsof the normal system and accessorypathway, then during AF, the ventricles served during atrial fibrillation: Ventricular premature shouldbe activatedshortly after eachmomentat which complexesand aberrant conductioncan result in wide the respectiveconduction system becomesexcitable, QRS complexbeatsresemblingpreexcitation.Without Becausethe frequency of stimuli applied to the AV direct accessorypathwayrecordings,it is not possibleto node and accessorypathway is high, the cycle length rule out this possibilitycompletely.However,there are variation should be small. However,this clearly is not 22 observationsthat stronglysuggestthat the different what has been observedduring AF. Experimentalevi- wide complexmorphologiesin patients 16 and 17 repdencedemonstratesthat anterogradeconcealedconduc- resentedconductionover 2 accessorypathways. First, tion is also a factor that determinesthe conductionof the nondominantwide QRS complexmorphologywas the AV node,I *J2 In additionto AV noderefractoriness consistentwith a distinctivepreexcitedmorphology.17 and anterogradeconcealedconduction,the directionof Second,at surgery, a secondaccessorypathway was atria1 input to the AV node may also be a factor in identified that was located at the site predicted at electrophysiologicstudy. Further researchof accessodeterminingAV conduction.I3 The aforementionedstudies”-‘3 were performedin ry pathway conductionshould use computerizedmap patients or animals with conductionover the normal ping techniquesto demonstrateventricular preexcitaAV ndeiHis Purkinje system,In patientswith pmexci- tion overthe secondaccessorypathway and to correlate tation syndrome,there is 4 1 accessorypathway pres- conductionwith precedingcyclelengthsand atria1actient. Thus, in addition to the abovefactors, an inter- vation proximal to the accessorypathway. Thesestudaction between the normal and abnormal pathways ies shouldhelp to identify the mechanismsthat govern through retrogradeconcealedconductionmay alsobe a accessorypathway conductionduring AF. Study limitation: One major limitation of this study factor that determinesthe ventricular responseduring AI?. Although there are good data to support the hy- is that thesepatientsmay not be representativeof all pothesisthat retrograde concealedconduction is de- patients with preexcitationsyndrome.These patients monstrableduring regular rhythm,2-4there is little evi- were referredto surgeryfor symptomaticsupraventiicdenceto support the claim that this factor is operative ular arrhythmias and thus representa highly selected during AF. The data presentedin this report are impor- patient cohort.There was a high proportionof multiple tant becausethey lend support to the hypothesisthat pathways. It is thus possiblethat the results of this retrogradeconcealedconductionis an important factor study may be biasedby the selectionprocess. that determines the anterograde accessorypathway Clinical significance: The single most important conduction. prognosticindicator in patientswith preexcitationsynSupemormirlGonduetiom Supernormalconduction drome is the shortestpreexcitedRR interval during over an accessorypathway is a well-recognizedphe- AF.18Becauseof the imperfectcorrelationbetweenthe nomenon5JjIn thesereports,a “phase” of supernormal shortestpreexcitedRR intervaland the anterogradeefconductionwasusuallypresent.This phaseis definedas fective refractory periods of the accessorypathway,18 a rangeof At-A:! intervalsresultingin conductionover factors other than accessorypathway refractoriness the accessorypathway inside the refractory period of probably contribute to conductionover the accessory the latterb6In patientswith manifestventricularpreex- pathway during AF and therefore alter the potential citation, the supernormalphase can be demonstrated risk of these patients.In this report, we summarized by progressivelyshorteningthe AI-A2 intervaland ob- our observationson accessorypathway conductiondurservingthat the QRS complexafter the A2 is preexcit- ing AF in patientswith preexcitationsyndromereferred ed with relativelylong AL-A2 intervalsand nonpreex- for surgicaltreatment.Although there was a significant cited with shorterintervalswhenthe accessorypathway correlation between accessorypathway refractoriness “effectiverefractory period” is reached,and that with and the shortestRR intervalduring AF,r other factors further shorteningof the Al-A2 intervalsthere is recur- such as retrogradeconcealmentand supernormalconrence of ventricular preexcitation.5Becausesupernor- duction are also important in determining accessory ACCESSORY PATHWAY CONDUCTION DURING ATRIAL FIBRILLATION
1333
pathway conduction, The role of supernormality in uncovering an otherwise undetected accessorypathway is important if operative therapy is contemplated. Aclm~wledgment: We wish to thank Richard Page, MD, and Joyce Pressley,MPH, for their assistancein collecting the data, The surgical ablation of the accessory pathways was performed by James Lowe, MD, and his associates.
REFERENCES 1. Wellens HJ, Durret D. Wolff-Parkinson-White syndromeand atrial fibrillation, relation betweenrefractory period of accessorypathway and ventricular rate during atria1 fibrillation. Am J Cardiol 1974;34;777-782. 2. Klein GJ, Yee R, SharmaAD. Concealedconductionin accessoryatrioventricular pathways; an important determinant of the expressionof arrhythmias in patietlts with Wolff-Parkinson-White syndrome.Circulation 1984;70:402-411. 3, Prystowsky EN, Pritchett ELC, Gallagher JJ. Concealedconductionpreventing anterogradepreexcitation in Wolff-Parkinson-White syndrome.Am JCardiol 1984;53:960-961.
4, Svinarich JT, Tai DY, Mickelson J, Keung E, Sung RJ. Electrophysiologic demonstration of concealed conduction in anomalous atrioventricular bypass tract% f Am Co11 Cardiol 1985:5:898~903. 5, Chang M-S, Miles WM, Prystowsky EN. Supernormalconductionin acccssory atrioventricular comiections.Am J Cardial 1987;59:852-856. & Przybylski J, Chiale PA, SanchezRA, Pastori JD, FrancesHG, Elizari MV, RosenbaumME. Supernormalconduction in the accessorypathway of patients
1334
THE AMERICAN JOURNAL OF CARDIOLOGY VOLUME 68
with overt or concealedventricular pre-excitation. J Am Call Cardiol 1987;9: 1269-1278.
7. Kirsh JA, SahakianAV, BaermanJM, Swiryn S. Ventricular responseto atria1 fibrillation: role of atrioventricular conduction pathways. J Am Colt Card!01 1988;12:1265-1272. 8. Prystowsky EN. Diagnosisand managementof the preexcitation syndromes, Curr Probl Cardiol 1988;13:227-310.
9. Wilkinson L. SYSTAT: The Systemfor Statistics. SYSTAT: Evanston,Illinois, 1988673-675. 10. Gallagher JJ, Selle JG, SvensonRH, Fedor JM, Zimmern SH, Scaly WC, Rob&k FR. Surgical treatment of arrhythmias. Am j Cardiol 19&8;611 2lA-44A. 11. SbderstrGmN. What is the reasonfor the ventricular arrhythmia in oasesof auricular fibrillation. Am Hearf J 1950;40:212. 12, Moe GK, Abildskov JA. Observationson the ventricular dysrhythmia associated with atrial fibrillation in the dog heart. Circ Res 1964;14:447-460, 13. JanseMJ. Influence of the direction of the atrial wave front on A-V nodal transmissionin isolated hearts of rabbits, &r Res 1969;25:439-449. 14. Hoff HE, Nahum LH. The supernormalperiod ih the mammalianventricle. Am J Physiol 1938;124:591-595. 15. Spear JF, Moore EN. Supernormalexcitability and conductionia the His= Purkinje systemof the dog. Circ Res 1974;35:782-792. 16. Soloff LA, Fewell WJ. The supernormalphaseof ventricular excitation in man. Its bearing on the genesisof ventricular premature systoles,and a note on atrioventricular conduction.Am &wrt J 1960;59:869-874. 17. Fananapazier L, German LD, Gallagher JJ, Lowe JE, Ptystowaky EN, Importance of preexcited QRS morphologyduring induced atria1 fibrillation to the diagnosis and localization of multiple accessory pathways. Qrrulatlon 1990;81:578-585. 18. Klein GJ, BashoreTM, Sellers TD, Pritchett ELC, Smith WM, Gallagher JJ. Ventricular fibrillation in the Wolff-Parkinson-White syndrome.N Ettgl J ,wed 1979;301:1080-1085.
NOVEMBER 15, 1991