882 JACC Vol. 13, No. 4 March 15, 1989:882-91

The Scintigraphic Characteristics of Ventricular Pre-excitation Through Mahaim Fibers With the Use of Phase Analysis

NORBERTO SCHECHTMANN, MD, FACC, ELIAS H. BOTVINICK, MD, FACC, MICHAEL DAE, MD, MELVIN M. SCHEINMAN, MD, FACC, J. WILLIAM O’CONNELL, BS, JESSE DAVIS, MD, FACC, STUART WINSTON, OD, ALLAN SCHWARTZ, MD, FACC, JOSEPH ABBOTT, MD, FACC San Francisco and Martinez, California

The phase image pattern of blood pool scintigrams was blindly assessed in 11 patients exhibiting conduction through Mahaim pathways, including 6 nodoventricular and 5 fasciculoventricular. These patterns were compared with the phase image findings in normal subjects, patients with left and right bundle branch block in the absence of pre-excitation and patients with pre-excitation through atrioventricular (AV) connections.

In all patients with a Mahaim pathway, the site of earliest phase angle was septal or paraseptal. Phase pro- gression was asymmetric and the pre-excited ventricle demonstrated the earliest mean ventricular phase angle in 10 of 11 patients. This pattern, and the associated ventric- ular phase difference, appeared to vary from that in normal subjects and in those with a septal AV connection, in whom phase progression is generally symmetric.

Scintigraphic phase analysis provided localizing infor- mation and presented patterns consistent with Mahaim pathways. Although not able to differentiate among Ma- haim pathway subtypes, these phase patterns differed from those in normal subjects, those with right and left lateral free wall pathways and most patients with a septal AV pathway. However, the phase pattern of patients with a Mahaim pathway may not differ from that of patients with a septal AV connection displaying an asymmetric pattern of phase progression, or those with left and right bundle branch block in the absence of pre-excitation. Objective, yet imperfect phase measurements supported these differ- ences. Such image findings may complement the often complex electrophysiologic evaluation of patients present- ing with pre-excitation.

(J Am Coil Cardiol1989;13:882-91)

Refined anatomic studies of the cardiac conduction system have demonstrated the presence of Mahaim fibers, acces- sory pathways connecting the distal atrioventricular (AV) node, His bundle and the upper bundle branches to the

From the Departments of Medicine, Cardiovascular Division, Radiology, Section of Nuclear Medicine and the Cardiovascular Research Institute at the University of California, San Francisco and the Cardiovascular Division, Veterans Administration Hospital, Martinez, California. This work was supported in part by a grant from the Fannie Rippel Foundation, Madison, New Jersey. Dr. Schechtmann, Attending Cardiologist at Ramos Mejia Hospital, Buenos Aires, Argentina, performed this work while he was a postdoctoral fellow at the University of California. San Francisco. Dr. Botvinick participated in this study while an Established Investigator of the American Heart Association with funds contributed by local heart associa- tions. He is supported in part by a grant from the George D. Smith Fund, San Francisco. Dr. Winston performed this work while he was a postdoctoral fellow at the University of California, San Francisco.

Manuscript received May 25, 1988: revised manuscript received August 24, 1988, accepted September 6, 1988.

Address wren Elias H. Botvinick. MD, Box 0214, M-1186, Uni- versity of Californiarat San Francisco, San Francisco, California 94143.

01989 by the American College of Cardiology

ventricular myocardium (l-3). Although the anatomic exist- ence of these fibers has been known for some time (Cll), their functional significance, has been more recently docu- mented (12-15). It is now apparent that some Mahaim fibers play an integral part in the facilitation of arrhythmias in a significant number of patients with pre-excitation syndrome (16-24). Further, the association of Mahaim fibers with additional intranodal pathways or AV connections has also been documented (25-30).

Recent studies (31-34) suggest two main anatomic Ma- haim subtypes. These are believed to insert into ventricular tissue after their origin from either the AV node, (nodoven- tricular connections), usually right-sided, or from the bundle of His or bundle branches (fasciculoventricular connec- tions), which may be right- or left-sided. Their presence can be suspected electrocardiographically, but they can only be identified and differentiated from other accessory pathways by their electrophysiologic profile. Study to establish the latter is invasive, complex, time-consuming, often arduous

073%1097189/$3.50

JACC Vol. 13. No. 4 SCHECHTMANN ET AL. 883 March IS, 19x9:882-91 SCINTIGRAPHIC CHARACTERISTICS OF VENTRICULAR PRE-EXCITATION

Table 1. Phase Values in 51 Patients With an Accessory Atrioventricular Connection or Mahaim Pathway

LV$ RVJ

AV Connections (n z 401

“Right n = 8 IO

“Left n = 16 (9)

*Septal n = 16 (10)

- 10.4 +- 7.4”

(46.7 2 8.2”)

-14.1 + 5.2”

(60.7 k 23.2”)

-1.3 ? 10.3b

(63.8 ? 32.0”)

Mahaim pathways (n ; I1 J Right n = 5 (2~

Left n = 6 12)

14.5 2 23.4”

4.5 + 20.2”

-24.1 i 6.7’

(kl.8 ? 7.0”)

-9.7 = 5.2”

(76.7 2 25.6”) ( 1.0 ir 10.9”

166.4 + 30 I”)

-13.7 i 14.3”

13.5 2 21.6”

a$ (LV-RV) ‘6, (LV-RV)

14.4 t 4.7 28.9 + 6.4”

(48.0 ? 12.0”) 167.6 + 10.8”)

-3.8 _t 3.5” -3.6 z 3.9

-14.3 2 6.5”) (-28.2 I 7.8”)

2.1 ? 4.7” 5.5 I 3.4

(3.1 i- 3.5”) (9.3 t 4.8”)

31.3 + 14.2” 36.6 2 15.1’

-8.7 f 1.8” ~9.2 t 9.9

“Data include thore of pre-excitation patients reported in references 48 and 53. Values for Mahaim pathways represent those acquired in association with the

greatest degree of pre-excitation for each patient and include paced values, in parentheses, when available. Owing to the small numbers in each subgroup, paced

values have been integrated into the calculation of mean values in the patients with a Mahaim pathway. LV$ = mean left ventricular phase angle; RV$ = mean

right ventricular phase angle: ~~ (LV-RV) = difference in mean ventricular phase angles: I\& (LV-RV) = ventricular difference in earliest ventricular phase angle

(phase onset).

and, at times, ambiguous. Additionally, controversy exists as to whether nodoventricular Mahaim tracts actually origi- nate more proximally and pass through the right ventricular free wall, inserting more distally into or near the right bundle (atriofascicular tracts) causing earliest ventricular activation in the distal septum or near the right ventricular apex. Scintigraphic analysis may permit differentiation of these possibilities, where proximal activation may be distin- guished from distal activation by the earliest site and pro- gression of phase angle.

The scintigraphic phase image is derived from the first harmonic Fourier transform of the regional equilibrium blood pool time versus radioactivity curve. The regional phase angle can be interpreted to estimate the relative onset of contraction in that location of the cardiac image (35-38). The phase image has been shown to relate the sequential pattern of ventricular contraction and, indirectly, of ventric- ular conduction (35,36,39-54). Owing to the potential clinical importance and frequent electrophysiologic complexity of Mahaim pathways, we sought to characterize the phase patterns and determine their specificity compared with that of phase images in other conduction subgroups.

Methods Patient selection. Among 51 blood pool scintigrams pro-

spectively acquired and blindly analyzed in patients studied during pre-excitation through an accessory pathway, we identified 11 studies that were acquired in patients exhibiting pre-excitation through a Mahaim pathway. Initially, scinti- graphic studies were acquired for evaluation of left ventric- ular function. Subsequently, as the phase method was estab- lished, patients were studied to confirm bypass localization. Each patient was imaged within 48 h of electrophysiologic study. In each case, scintigraphic findings were assessed with regard to their consistency with the documented elec- trophysiologic pathways.

Study protocol: general relations. Electrophysiologic study was performed in all patients assessed for pre- excitation and was evaluated by observers different from those interpreting phase image data. Each observer was unaware of the findings of the other. A 12 lead electrocar- diogram (ECG) was obtained at the time of imaging for correlative purposes and to confirm identity of conduction patterns present at the time of imaging with those assessed electrophysiologically. The six standard leads and a modified V, lead, monitored during imaging, documented the stability of rhythm and conduction during the acquisition period.

The detailed protocol of the electrophysiologic study has been previously presented (27,48,53,55). The presence of an extranodal bypass connection was defined by the criteria suggested by Wellens (56). Patients exhibiting conduction through a Mahaim pathway were classified into those with a nodoventricular or fasciculoventricular connection accord- ing to the criteria suggested by Gallagher et al. (57). The criteria of Benditt et al. (58) were applied to diagnose enhanced AV node conduction, whereas dual AV conduc- tion was diagnosed according to the criteria of Denes et al. (59).

Patient characteristics. Each of the 51 patients had a documented or suggestive history of Wolff-Parkinson-White syndrome and most had rest ECG evidence of ventricular pre-excitation (38,48). On electrophysiologic study, 11 pa- tients (10 men and 1 woman with a mean age of 36 years) had conduction through a Mahaim pathway; 40 patients (26 men and 14 women with a mean age of 37 years) had conduction through an AV connection.

Among those studied during conduction through a Ma- haim pathway, six had a nodoventricular tract that in five inserted into the right ventricle, and five had a fasciculoven- tricular tract that in four inserted into the left ventricle. Four patients were shown to have concomitant fast AV node conduction and in four patients the study was performed

884 SCHECHTMANN ET AL. SCINTIGRAPHIC CHARACTERISTICS OF VENTRICULAR PRE-EXCITATION

MAHAIM PATHWAYS

RIGHT

\ -60

,n=5, n=6, 1 LVil; RV% LV$ RV8

Figure 1. Mean ventricular phase angle. Values for mean left and right ventricular phase angle in patients with a left- or right-sided Mahaim pathway are compared with the same value in normal patients (N) and patients with right (RBBB) and left (LBBB) bundle branch block. Values used here and in other figures relate to those acquired during maximal pre-excitation. NLV$ = mean left ventric- ular phase angle in normal subjects; NRV$ = mean right ventricular phase angle in normal subjects. See Table I for further abbrevia- tions.

after resection or ablation of an associated AV connection. Two of the latter patients underwent imaging before resec- tion as well. Among those with conduction through AV connections, there were 24 with a right or left lateral pathway, including 2 with a co-existent Mahaim pathway, and 16 with a septal AV pathway, including 2 with a co-existent Mahaim pathway. To maximize the degree of pre-excitation, 28 patients, 4 with conduction through Ma- haim and 24 with conduction through AV connections, including 10 with septal pathways, were imaged as well during rapid atrial pacing to the greatest conducted ventric- ular response consistent with maximal pre-excitation and patient comfort.

Scintigraphy. Equilibrium multiple gated blood pool scin- tigrams were acquired and analyzed in multiple projections as previously reported (43,48,60,62). Phase image analysis was performed with use of the fundamental Fourier har- monic applied to the first 25 frames of the blood pool study according to our standard method (43,48,53). A course overview‘ of the pattern of distribution of sequential phase angles was assessed in each projection by a 16 frame movie version of the gray scale coded image. The static phase image coded phase angle in refined gray shades, with in-

80 -

70 -

60 -

50 -

40 -

30 -

20 -

lo-

O-

-10 -

-20 -

JACC Vol. 13, No. 4 March 15. 1989:882-91

MAHAIM PATHWAYS

LEFT RIGHT

??

-N

- RBBB

n=5 n=6

A@(LV-RV)

Figure 2. Difference in mean ventricular phase angle. The difference in mean ventricular phase angle in patients with a left- or right-sided Mahaim pathway is compared with the same value in normal subjects and patients with bundle branch block. The pre-excited ventricle had the smaller phase angle in only one patient. Note the similar values for patients with a left Mahaim pathway and those with right bundle branch block as well as those with right Mahaim pathway and left bundle branch block. In patients with a right-sided Mahaim pathway, the difference in mean ventricular phase angles was significantly different from similar values in normal subjects. See Figure I and Table 1 for abbreviations.

creasingly lighter shades assigned to pixels with increasingly delayed phase angle. The gray scale was adjusted to fit the range of ventricular phase angles. Phase histograms relating phase angle on the abscissa to the number of pixels at each phase angle on the ordinate were then constructed in the “best septal” projection. Movable cursors applied to the histogram highlighted pixels of any selected phase angle interval as small as 2.8”. This permitted quantitation and localization of the site of earliest phase angle with delinea- tion of phase angle progression in each ventricle and calcu- lation of mean left and right ventricular phase angles. Back- ground noise was suppressed by eliminating all pixels occurring at <5% of the peak histogram frequency (43,48,53).

In other projections, when full anatomic isolation of the ventricles was not always possible, regions of interest were approximated and assessed for the pattern of phase angle

JACC Vol. 13. No. 4 March 15. 1989:8X?-Sl

60 t

70 -

60 -

50-

40-

30-

20

t 10

0 1

M&!!vl PATHWAYS ____~.

LEFT RIGHT

.

.

. N

. .

9? REBB .

-20 t .

SCHECHTMANN ET AL. 885 SCINTIGRAPHIC CHARACTERISTICS OF VENTRICULAR PRE-EXCITATION

n=5 n=6

A@o,,,,WW

Figure 3. Interventricular difference in earliest phase angle. The interventricular difference in earliest phase angle in patients with a left- or right-sided Mahaim pathway is compared with the same value in normal subjects and patients with bundle branch block. Again, the earliest phase angle was localized to the pre-excited ventricle in all but one case. whereas the mean values for left- and right-sided Mahaim pathways were not significantly different from those in patients with right and left bundle branch block, respec- tively. See Figure 1 and Table I for abbreviations.

progression (48). Phase images and values best reflecting the location of the bypass pathway were those acquired during maximal pre-excitation and. when available. studies ac- quired during atria1 pacing were assessed (53).

Statistical methods. Because the distribution of phase angles in ventricular regions of interest is nonparametric (43,481, phase angle data were analyzed nonparametrically. Intrapatient comparisons of ventricular phase angles were performed with use of the Wilcoxon paired sample test. Numeric values are presented, plus or minus standard devi- ations. Intragroup comparisons were made with use of the unpaired t test.

Results Wall motion and right and left ventricular ejection frac-

tion were normal in all patients with pre-excitation syn- drome. Rest heart rate at the time of study ranged from 42 to 122 beatsimin. During atrial pacing, the ventricular response was always constant and I:1 with rates varying from 100 to 170 beatsimin. In each case, there was complete observer agreement on the site of earliest phase angle.

Patients with pre-excitation through an extranodal AV connection. Eight among these 40 patients had a right and 16 had a left pathway. In all but 2 cases at rest and in all I4 cases with atrial pacing. aspects of the ipsilateral ventricle were darker than those of the contralateral ventricle. Phase progression radiated homogeneously from the site of earliest phase angle. which always involved a region of the ipsilateral ventricle, sometimes only slightly but always earlier than the septum. Pacing increased all phase angle values as well as the interventricular difference of mean phase angles, while accentuating the visibility of the site of earliest phase angle and increasing the asymmetry of phase progression (53) (Table Il. As a group, the mean ventricular phase angle, the difference between mean left and right ventricular phase angles and the difference between the earliest right and left ventricular phase angles were significantly different from values previously reported in normal subjects (43) (Fig. I to 3).

Among the I6 patients with electrophysiologic evidence of a septal AV connection, IO studied both at rest and with atrial pacing, 14 demonstrated symmetric, homogeneous left and right ventricular phase patterns with earliest phase angle at a site in or near the septum. In two studies, one acquired during atrial pacing, phase progression spread initially through the right ventricle. Delayed phase angle in the region of the right ventricular outflow tract was not unusual. Although pacing increased the earliest phase angle as well as the mean ventricular phase angle. except for the one case noted earlier. neither the site of earliest phase angle, the symmetric pattern of phase progression nor the interven- tricular difference of mean phase angle changed significantly with pacing (Table I). Although the septal site of earliest phase angle corresponded to the electrophysiologic site, these patients could not be scintigraphically differentiated from previously reported normal subjects (43). In three patients phase progression revealed. as well. earliest sites at the lateral aspects of both ventricles (541. Three paced studies in patients with AV pathways were unavoidably gated off‘ the atrial pacing spike. producing an artifactual increase in ventricular phase angles.

Patients with Mahaim fibers. Among these I I patients, there w’ere six nodoventricular pathways, five of which inserted into the right ventricle, and five fasciculoventricular pathways. four of which inserted into the left ventricle (Fig. I to 7). In these I I patients, the sight of earliest phase angle was septal or paraseptal in all cases and in 10 the pre-excited ventricle demonstrated the earliest mean ventricular phase angle (Table I and Fig. 3 to 7). The phase angle progressed asymmetrically with a significant delay in the earliest and mean phase angle of the contralateral ventricle and with a significant difference in ventricular phase onset and mean ventricular phase angle compared with that in previously reported normal subjects (43). Pacing in four patients accen- tuated the phase differences and brought an exaggerated

886 SCHECHTMANN ET AL. SClNTlGRAPHICCHARACTERlSTlCSOFVENTRICULAR PRE-EXCITATION

JACC Vol. 13, No. 4 March 15, 1989:882-91

mssc.

‘I I I

mSec gji ~"vl~g AH 80

B AV115 AVl& HV 25 AV 105

increase in phase angles, measured in one study gated off the atria1 pacing spike. Localization appeared more distal in the septum or apical, in the “best septal” projection in three patients with right nodoventricular pathways. Figures 5 to 7 illustrate serial phase images and associated histograms in patients with normal conduction, right nodoventricular, left fasciculoventricular, right free wall and left lateral AV connections.

Discussion Scintigraphic phase analysis appears to accurately reflect

the sequence of ventricular activation in patients with right and left bundle branch block (42,43) and left anterior hemi- block (44). The site of earliest phase angle has been success- fully correlated with the location of the initial ventricular focus in patients with an artificial cardiac pacemaker (39,45), sustained ventricular tachycardia (46,47) and an accessory AV connection (48-53). Most recently, the phase image pattern has been characterized in patients with a posterior septal bypass pathway (54).

Conduction phase patterns in pre-excitation through Ma- haim fibers. This study sought to identify a phase pattern characteristic of conduction through Mahaim pathways. Other patient groups noted here were in part reported previously (43,48,53) and were included here to compare with the patients with a Mahaim pathway. The results

Figure 4. Electrophysiologic evaluation. A, Multiple surface and endocardial electrograms in a patient with a right nodoventricular pathway. Note the increasing AH interval and decreasing HV with increasing AV interval and pre-excitation at increasing atrial pacing rates. The surface electrocardiogram indicated a left-sided conduc- tion delay. B, Surface and endocardial electrograms in a patient with a left fasciculoventricular pathway. The constant short HV interval at progressively increased pacing rates, lengthening AH and AV intervals and evidence of increased pre-excitation with a right ven- tricular delay in the surface leads gave evidence of the electrophysiologic abnormality. AP = atrial pacing; CS = coronary sinus electrogram; HBE = His bundle electrogram; NSR = normal sinus rhythm; RA = right atrial electrogram: S = stimulus.

include the first description of scintigraphic studies acquired during atrial pacing in patients with an accessory septal pathway. Phase analysis localized the pre-excitation focus to the region of the septum in all 11 patients with a Mahaim pathway and demonstrated correct lateralization in IO of them. The scintigraphic method seeks evaluation of the conduction sequence by extrapolation from the contraction sequence assessed from analysis of changes in the blood pool. Differences related to this comparison or projectional considerations may account for the apparent localization of two Mahaim pathways to a paraseptal location. Addition- ally, the distal site of earliest phase angle in three patients with a right nodoventricular Mahaim pathway suggests its direct insertion into the right bundle (Fig. 5,6), which is not likely related to projection alone. This finding may illustrate the potential value of the method to resolve questions of conduction not otherwise easily approached.

Two patients with both a left AV connection and a Mahaim pathway studied scintigraphically both before and after resection of the AV connection further document the ability of the method to assess the conduction pattern. Thus, although demonstrating earlier mean left ventricular phase angle, both before and after operation, the site of earliest phase angle within the left ventricle clearly shifts from the lateral wall to the septum, with an associated shift in the pattern of phase progression as conduction shifts from a left

JACC Vol. 13, No. 4 SCHECHTMANN ET AL. 887 March 15, 1989:882-91 SCINTIGRAPHIC CHARACTERISTICS OF VENTRICULAR PRE-EXCITATION

Figure 5. Phase analysis in patients with Mahaim path- ways. In the left panel of each row are the phase histograms for the left (white) and right (black) ventric- ular regions of interest in the related phase image shown at right. These were derived from blood pool studies acquired in the “best septal” left anterior oblique pro- jection in a patient with normal conduction (top), a patient with conduction through a documented right nodoventricular accessory connection (middle), and a patient with conduction through a right free wall acces- sory atrioventricular (AV) connection (bottom). In col- umn 1 pixels are enhanced in white if their phase angle lies within the span of the gray bars superimposed on each histogram. Sequential phase angle progression, corresponding to the span of adjacent histogram inter- vals, is shown in subsequent panels. The earliest en- hanced pixels, the site of earliest phase angle, can be differentiated from light regions of late phase angle by comparison of the original, unenhanced phase image (not shown here) with the initial enhanced image. In the patient with the nodoventricular pathway, the site of earliest phase angle is in the distal septum (open arrows), progressing superiorly in the right ventricle and toward the AV groove (black arrowheads), with late left ven- tricular phase angle. This pattern of intraventricular phase progression is quite different from that in the patient below, in whom the site of earliest phase angle is localized to the lateral aspect of the right AV groove (black arrowheads), progressing toward the septum with similar left-sided phase delay. Both of these patterns differ from that in the normal example, in which multiple sites, including the septum, demonstrate similarly early phase angle, with relatively symmetric phase progres- sion to both ventricles.

lateral AV connection to a left fasciculoventricular pathway (Fig. 7).

In patients with bundle brunch block, similarly analyzed and previously reported (43), ventricular phase images are asymmetric with the septum again the site of earliest phase angle. The mean phase angle of the contralateral ventricle is smaller and its earliest phase angle is earlier than that of the ipsilateral ventricle and larger than that of the same ventricle in normal subjects (Table 2). The pattern is similar to that seen with a Mahaim pathway.

Limitations and advantages of the method. Phase analysis is a mathematic approximation of the time versus radioac- tivity curve and is based entirely on sequential volume changes during the cardiac cycle. Other investigators (37,63) have pointed to the potential errors in the application of this method to the assessment of contraction and indirectly to the conduction sequence and to the possible advantages of more complex methods. Yet, although not a perfect method, it is simple and has proved reliable for applications such as those performed here.

Phase angle is not to be equated to the time of onset of contraction; rather. it is influenced by all aspects of the time versus radioactivity curve and particularly its symmetry (64). For these reasons. the results are not surprisingly

influenced by heart rate. Yet, the method has accurately localized the initial focus as well as the pattern of ventricular conduction both in animal studies (65) and in patients with sinus rhythm, paced rhythms and ventricular tachycardia (39,46,47). Although it would be ideal to limit the patients studied here to those within a narrow range of heart rate, this is not practical and probably unnecessary. Pre-excitation is occasionally only manifest and is generally amplified, scin- tigraphically as well as electrophysiologically, with rate acceleration by atria1 pacing (53). Most prominently, the factor of rate appears to affect the absolute phase angle and the standard deviation of mean ventricular phase angle. However, rate does not seem to influence dramatically the interventricular phase difference or the sequence of phase angle progression, and the localization of the site of earliest phase angle parallels that of the pre-excitation pathway even at high heart rates. This result should not be surprising because, although heart rate acceleration alters curve sym- metry and absolute phase angle, these changes are effected globally. Thus, any relative change in phase angle or its sequence should be reflective of the actual timing of the contraction pattern and therefore indicative of true conduc- tion (contraction) abnormalities. Further, our analytic method used in the patients with a Mahaim pathway mini-

888 SCHECHTMANN ET AL. JACC Vol. 13. No. 4 SCINTIGRAPHIC CHARACTERISTICS OF VENTRICULAR PRE-EXCITATION March 15, 1989:882-91

Figure 6. Right nodoventricular accessory connection (same for- mat). Phase images and ventricular histograms in a patient with conduction through a right nodoventricular accessory pathway. The horizontal line in the histogram represents the 5% frequency level. The full pattern of serial phase progression is illustrated (panels I to 6) as windows of phase angle are sampled by the gray histogram sampling bar. The initial phase image (I) shows a dark gray right ventricle with the right (black) phase histogram preceding that of the left. The site of earliest phase angle is again seen in the more distal or apicoseptal region (arrow point) with progression proximally in the right ventricle and toward the AV groove with delayed phase angle of the left ventricle.

mizes errors introduced by curve asymmetry and irregular- ities of heart rate (66).

The results in patients with pre-excitation through CI septal pathway indicate that atria1 pacing does influence the phase angle, increasing that value as the relative time spent in diastole is reduced and the fitted time versus radioactivity curve approaches symmetry. However, in this group, unlike patients with a right or left pathway, the site of earliest phase angle generally remains septal, mean phase angle remains similar in the two ventricles and the pattern of phase progression continues to reflect septal activation, similar to that in normal subjects (Tables 1 and 2). Although this makes it impossible to differentiate septal pre-excitation from normal conduction in most cases, it does permit identification of a septal pathway in the presence of known pre-excitation. Further, a recent study (54) suggests the presence of a characteristic phase

pattern in patients with a posteroseptal pathway. However, no study has yet assessed the ability of phase imaging to differentiate septal from nearby paraseptal bypass pathways. The latter likely present asymmetric phase patterns similar to Mahaim pathways and the minority of septal pathways. Our ability to differentiate among these patterns will cer- tainly influence the specificity of the phase pattern presented by Mahaim pathways.

Additionul errors in the absolute value of the phase angle during atria1 pacing may occur when gating is triggered by the atria1 pacing spike, seen in some patients here. This occurrence shifts the absolute ventricular phase angle to artifactually higher values, although not affecting the se- quence of phase progression or the interventricular phase difference (53).

The phase angle is potentially injluenced as well by severe contraction abnormalities (41,47,67,68). This possi- bility was not a factor here where the phase method was applied to the localization of the initial focus of electrical activation in the presence of normal contraction.

An alternative analytic method may be based on analysis of all 28 frames. Without applying some corrective maneu- ver, this is impractical, owing to the degenerative effect of variable heart rate on terminal curve data even in patients in sinus rhythm (37,38,64,68). As demonstrated by Wendt et al. (63), higher Fourier harmonics provide a better curve fit, permit more accurate assessment of the time of contraction onset and a variety of other functional variables and help to overcome curve-fitting artifacts related to structural overlap and the moving edge. However, such analysis is time and computer intensive and provides relatively subtle image differences from the analysis applied. Although comparative studies have been made in model systems, the methods have never been practically compared and the optimal method has not yet been established.

Oaring to these technical considerations and the variation of henrt rtrtes in this study. great significance cannot be placed on the absolute values or standard deviations of phase angle presented. However, the site of earliest phase angle, the sequence of phase progression and ventricular phase differences have demonstrated their ability to reflect a variety of contraction and related conduction patterns, par- allel the known pattern of conduction in the patients studied here, and appear to lateralize and characterize Mahaim pathways while differentiating them from a variety of other conditions.

Because phase image pathbiqs present relutive regional d#erences. the phase pattern gives no absolute measure of pre-excitation. Similarly. such analysis could not differentiate among Mahaim subtypes or differentiate among patients with Mahaim conduction, those with right or left bundle branch block in the absence of pre-excita- tion or those with probable paraseptal AV connections. Yet, in the presence of known pre-excitation. phase analysis

JACC Vol. 13. No. 4 March IS. 1989:88?-91

SCHECHTMANN ET .AL. 889 SCINTIGRAPHIC CHARACTERISTICS OF VENTRICULAR PRE-EXCITATION

Figure 7. Left atrioventricular (AV) accessory connection and left fasciculoventricular connec- tion. A, Shown according to the same format is the phase pattern in a patient with conduction through a left lateral AV connection. The site of earliest phase angle, in this preoperative study, is in the lateral left ventricular wall, progressing to the septum and the right ventricle. This patient also had a left fasciculoventricular pathway. B, Phase image in this same patient, acquired after surgical resection of the left Kent fiber, now with conduction through the fasciculoventricular con- nection. The gray scale is reversed owing to gating from an atrial pacing spike. The left ven- tricular histogram again precedes that of the right. However, the pattern of phase progression within the left ventricle is altered dramatically. consistent with the origin of the fasciculoventri- cular pathway. The dark region in the right ven- tricular region of the phase image relates to an artifact of thresholding and represents an image area eliminated from analysis. Another patient with similar pathways and clinical management was also studied both before and after resection of the AV connection, again with appropriate image findings.

could differentiate those patients with conduction through Mahaim pathways from many patients with conduction through a septal pathway, and most of those with con- duction through right or left lateral AV connections, and it may complement often complex electrophysiologic evalu- ation.

Clinical applications. A noninvasive method to charac- terize electrophysiologic abnormalities and specifically to localize pre-excitation pathways would be of significant clinical utility. The potential value of such methods relates to their ability to simply, quickly and accurately identify the global ventricular activation sequence, thereby helping to I) clarify ambiguities often present after electrophysio- logic studies; 2) focus, and therefore shorten and simplify the often complex, lengthy and arduous electrophysiologic

study; 3) help resolve continued clinical problems after treatment. occasionally related to the presence of dual pathways: 4) aid identification of accessory pathways con- ducting only in an anterograde fashion: and 5) confirm accessory pathway location when localization and not merely diagnosis is required, as when ablation or surgery is considered. Although the method is being applied in a few centers around the world, the extent of its application remains limited. This limited use likely relates to the lack of general availability of programs to perform refined phase analysis. to the specialized nature of the study and to the too infrequent coincident location of both an advanced cardiac electrophysiology referral center and an interested, involved, informed and equipped cardiovascular imaging service.

890 SCHECHTMANN ET AL. JACC Vol. 13, No. 4 SCINTIGRAPHIC CHARACTERISTICS OF VENTRICULAR PRE-EXCITATION March 15, 1989:882-91

References 25.

1. Lev M, Lemer R. The theory of Kent: histologic study of the normal AV communications of the human heart. Circulation 1975;12: 176-97.

2. Davies MJ. Pathology of Conducting Tissue of the Heart. London: Butterworths, 1971:171.

26.

3. Gallagher JJ. Variants of preexcitation: update 1984. In: Zipes DP, Jahfe J, eds. Cardiac Electrophysiology and Arrythmias. Grune & Stratton, 1984:419-33.

27.

4. Mahaim 1, Benatt A. Nouvelles recherches sur les connexions supe- rieures de la branch gauche du faisceau de His-Tawara avec cloison interventriculaire. Cardiologia 1938:1:61-5.

5. Mahaim I. Kent’s fibers and the A-V paraspecific conduction through the upper connections of the bundle of His-Tawara. Am Heart J 1947;33: 651-5.

28.

29.

6. Mahaim 1, Winston MR. Recherches d’anatomie compare’e et de pathol- ogie experimentale sur les connexions hautes du faisceau de His-Tawara. Cardiologia 1941;5:189-%.

30.

7. Truex RC, Bishof JK, Hoffman EL. Accessory atrioventricular muscle bundles of the developing human heart. Anat Res 1958;131:45-56.

8. Matsuguchi H, Takeshita A, Makino N, et al. Mahaim conduction producing left axis deviation and normal QRS. Br Heart J 1978;40902-8.

9. Ward DE, Camm AJ, Spurrell RAJ. Ventricular preexcitation due to anomalous nodoventricular pathways: report of 3 patients. Eur J Cardiol 1979;9: I I 1-5.

31.

32.

33. 10. Anderson RH, Becker AE, Brechenmacher C, et al. Ventricular pre-

excitation-a proposed nomenclature for its substrates. Eur J Cardiol 1975;3:27-36. 34.

It. Becker AE, Anderson RH, Durrer D, et al. The anatomic substrates of Wolff-Parkinson-White syndrome. Circulation 3978;57:870-81.

12. Lev M, Fox SM, Bharati S, et al. Mahaim and James fibers as a basis for a unique variety of ventricular preexcitation. Am J Cardiol1975;36:880-9.

13. Gallagher JJ, Smith WM, Kasell JH, et al. Role of Mahaim fibers in cardiac arrhythmias in man. Circulation 1981;176-89.

14. Lev M, Leffler WB, Langendorf R, et al. Anatomic findings in a case of ventricular preexcitation (WPW) terminating in complete atrioventricular block. Circulation 1966;34:718-33.

35.

36.

37.

38. IS. Neuss H, Thormann 3. Functional properties of Mahaim fibers. Khn

Wochenschr 1977;55:1031-8. 39. 16. Prystowsky EN, Miles WM. Heger JJ, et al. Preexcitation syndromes:

mechanism and management. Med Clin North Am 1984;68:831-93.

17. Wellens HJJ. Electrical Stimulation of the Heart in the Study and Treatment of Tachycardias. Baltimore: University Park Press, 1971:97- 109.

40.

18. Probst P, Pachinger U, Steinbach R, et al. Preexcitation of the ventricle associated with total intra-His bundle branch block. Am Heart J 1977;94: 96-104.

41.

19. Bardy GH, German LD, Packer DL, et al. Mechanism of a tachycardia using a nodoventricular Mahaim fiber. Am J Cardiol 1984;54: 1140-9.

20. Bhandari A, Morady F, Shen EN, et al. Catheter-induced His bundle ablation in a patient with reentrant tachycardia associated with a nodo- ventricular tract. J Am Co8 Cardiol 1984;4:61 l-9.

21. Okumura M, Koike T, Hattori M, et al. A case of preexcitation syndrome due to possible Mahaim fiber conduction. Respir Circ 1976;24:8916.

22. Ikram H. Direct demonstration of pure infranodal preexcitation (Mahaim conduction) by A-V nodal and His bundle electrocardiography. Angiology 1977;28:376+34.

42.

23. Toboul P, Vexler RM, Chatelain MT. Reentry via Mahaim fibres as a possible basis for tachycardia. Br Heart J 1978;40:806-I I.

24. Morady F, Scheinman MM, Gonzalez R, et al. His-ventricular dissocia- tion in a patient with reciprocating tachycardia and a nodoventricular bypass tract. Circulation 1981;64:83m.

43.

44.

45.

46.

Gomes JA, Haft JI. His bundle electrocardiography in the Wolff- Parkinson-White syndrome: evidence for combination of James and Mahaim conduction. Cardiology 1977;62:355-62.

Motte G, Brechenmacher C, Davy JM, et al. Association de fibers nodoventriculaires et atrioventriculaires il l’origin de tachycardies recip roques. Arch Mal Coeur 1980;73:737-42.

Abbott JA, Scheinman MM, Morady F, et al. Coexistent Mahaim and Kent connections: diagnostic and therapeutic implications. J Am Coil Cardiol 1987;10:364-72.

Tonkin AM, Dugan FA, Svenson RH, et al. Coexistence of functional Kent and Mahaim-type tracts in the pre-excitation syndrome: demonstra- tion by catheter techniques and epicardial mapping. Circulation 1975;52: 193-200.

Sung RI, Styperek JL. Electrophysiologic identification of dual atrioven- tricular nodal pathway conduction in patients with reciprocating tachy- cardia using anomalous bypass tracts. Circulation 1979;60:1464-76.

Gallagher JJ, Scaly WC, Kasell J, et al. Multiple accessory pathways in patients with the preexcitation syndrome. Circulation 1976;54:571-7.

Gomick CC, Benson DW Jr. Electrocardiographic aspects of the preex- citation syndromes. In: Benditt DG, Benson DW, eds. Cardiac Preexci- tation Syndromes: Origins, Evaluation and Treatment. Boston: Martinus Nijhoff, 1988:43-73.

Bardy G, Fedor JM, German LD, et al. Surface electrocardiographic clues suggesting presence of a nodofascicular Mahaim fiber. J Am Co8 Cardiol 1984;3:1161-8.

Ellenbogen KA, Ramirez NM, Packer DL, et al. Accessory nodoventric- ular (Mahaim) fibers: a clinical review. PACE 1986;9:868-79.

Reiter MJ, Smith WM, Gallagher JJ, et al. Clinical spectrum of ventricular tachycardia with left bundle branch block morphology. Am J Cardiol. 1983;3:1161-7.

Verba JW, Bomstein I, Alazraki NP, et al. Onset and progression of mechanical systole derived from gated radionuclide techniques and dis- played in a tine format (abstr). J Nucl Med 1979;20:625.

Links JM, Douglass KH, Wagner HN. Patterns of ventricular emptying iby Fourier analysis of gated blood pool studies. J Nucl Med 1980;21:978- 88. Goris ML. Functional or parametric images. J Nucl Med 1982;23:36&3.

Pave1 DC, Briandet PA. Quo vadis phase analysis. Clin Nucl Med 1983;8: 564-6. Botvinick EH, Frais MA, Shosa DW, et al. An accurate means of detecting and characterizing abnormal patterns of ventricular activation by phase image analysis. Am J Cardiol 1982;50:289-98.

Links JM, Raichlin JS, Wagner HN Jr, et al. Assessment of the sight of ventricular activation by Fourier analysis of gated blood pool studies. J Nucl Med 1985;26:27-33. Botvinick E, Dunn R, O’Connell W, et al. The phase image: its relation- ship to patterns of contraction and conduction. Circulation 1982;65:551- 60. Swiryn S, Pave1 D, Byrom E, et al. Sequential regional phase mapping of radionuclide gated biventriculograms in patients with left bundle branch block. Am Heart J 1981;102:1000-10. Frais MA, Botvinick EH, Shosa DW, et al. Phase image characterization of ventricular contraction in left and right bundle branch block. Am J Cardiol 1982;50:9>105. Dae MW, Botvinick EH, O’Connell W, et al. Assessment of left anterior fascicular block by scintigraphic phase analysis (abstr). Clin Res 1984;32: 4A. Bashore TM, Steme RA, Schalfer PB, et al. The noninvasive localization of ventricular pacing sights by radionuclide phase imaging. Circulation 1984;70:681-8. Swiryn S, Pave1 D, Byrom E, et al. Sequential regional phase mapping of radionuclide gated biventriculograms in patients with sustained ventricu- lar tachycardia: close correlation with electrophysiologic characteristics. Am Heart J 1982;103:319-32.

JACC Vol. 13, No. 4 SCHECHTMANN ET AL. 891 March 15. 1989:882-91 SCINTIGRAPHIC CHARACTERISTICS OF VENTRICULAR PRE-EXCITATION

41.

48.

49

50

51

52

53.

54.

55.

56.

57.

Botvinick E, Scheinman MM, Morady F. et al. The scintigraphic assess- ment of sequential changes in ventricular tachycardia (abstr). J Am Coil Cardiol 1983;1:711.

Botvinick EH, Frais MA, O’Connell W, et al. Phase image evaluation of patients with ventricular pre-excitation syndromes. J Am Coil Cardiol 3: 1984;799814.

Chan WC, Calff V, MacDonald D II, et al. Topography of pre-emptying ventricular segments in patients with Wolff-Parkinson-White syndrome using scintigraphic phase mapping and esophageal pacing. Circulation 1983;67:1139-47.

Nishimura Y. Yamamuro M, Komishi T, et al. Visualization of Kent bundle position in WPW syndrome using phase analysis (abstr). J Nucl Med 1985;26:58.

Nakajima K, Bunko H, Tada A, et al. Nuclear tomographic phase analysis: localization of accessory conduction pathways in patients with Wolff-Parkinson-White syndrome. Am Heart J 1985;109:809-19.

Nakajima K, Bunko H, Tada A, et al. Phase analysis in patients with Wolff-Parkinson-White syndrome with surgically proven accessory con- duction pathways. J Nucl Med 1984;25:7-26.

Botvinick E, Dae M, Scheinman M, et al. Augmented pre-excitation assessed by scintigraphic phase analysis during atrial pacing. Am Heart J 1987;114:738-45.

Oeff M, Abbott J, Griffin J, Herre J, Scheinman M, Botvinick E. Image triangulation of accessory pathways in patients undergoing catheter abloation of posteroseptal pathways (abstr). PACE 1988:11:525.

Morady F, Scheinman MM. Coexistent posteroseptal and right-sided atrioventricular bypass tracts. J Am Coil Cardiol 1985;5:64C-6.

Wellens HJJ. Electrical Stimulation of the Heart in the Study and treatment of Tachycardias. Baltimore: University Park Press, 1987:70.

Gallagher JJ, Pritchett ELC. Sealy WC, et al. The pre-excitation syn- dromes. Prog Cardiovasc Dis 1978:20:285-327,

58.

59.

60.

61.

62.

63.

64.

65.

66.

67.

68.

Gallagher JJ, Pritchett ELC, Smith WM. et al. Characteristics of atrio- ventricular conduction and the spectrum of arrhythmias in Lown-Ganong- Levine syndrome. Circulation 1978;57:454-65.

Denes P. Wu P, Dhinga RC. et al. Demonstration of dual A-V nodal pathways in patients with paroxysmal supraventricular tachycardia. Cir- culation 1973;48:549-55.

Rahimtoola S, Botvinick E, Perez-Gonzalez J, et al. Biplane geometric ventricular volumes from combined first transit equilibrium radionuclide ventriculography (abstr). Circulation 1979;59,6o(suppl II):H-268.

Botvinick EH, Glazer HB, Shosa DW. What is the reliability and the utility of scintigraphic methods for the assessment of ventricular func- tion? In: Rahimtoola SH, ed. Controversies in Coronary Artery Disease, Cardiovascular Clinics. Philadelphia: F.A. Davis, 1983;13:65-90.

Greenberg BH, Drew D, Botvinick EH, et al. Evaluation of left ventric- ular performance by gated radionuclide angiography. Clin Nucl Med 1980;5:245-52.

Wendt RE III. Murphy PH. Clark JW Jr, et al. Interpretation of multigated Fourier functional images. J Nucl Med 1982;27:715-24.

Bacharach S, Green M, Bonow R, et al. A method for objective evaluation of functional images. J Nucl Med 1982;23:285-93.

Dae M. Davis J, Botvinick EH, et al. Comparison of scintigraphic phase maps to epicardial activation maps in dogs (abstr). J Nucl Med 1985;26: P57.

O’Connell W, Shosa D, Frais M, et al. Symmetry violation errors in first Fourier harmonic phase analysis (abstr). Am J Cardiol 1982;49:1045.

Frais M, Botvinick E, Shosa D, et al. Phase image characterization of localized and generalized contraction abnormalities. J Am Coll Cardiol 1984;4:987-96.

Mancini GBJ, Peck WW, Slutsky RA, et al. Analysis of phase angle histograms from equilibrium radionuclide studies: correlation with semi- quantitative grading of wall motion. Am J Cardiol 1985;55:535-43.