sequential regional phase mapping of radionuclide gated biventriculograms in patients with sustained...

American Heart Journal Founded in 1925

March 1982 Volume 103, Number 3

CLINICAL INVESTIGATIONS

Sequential regional phase mapping of radionuclide gated biventriculograms in patients with sustained ventricular tachycardia: Close correlation with electrophysiologic characteristics

Radionuclide (RNA) gated studies were performed during sinus rhythm and during spontaneous or induced sustained ventricular tachycardia (VT) in six patients with clinical VT. Fourier analysis of time-activity variation was used to calculate a RNA phase value for each pixel in the image. Color coding of each pixel according to its calculated phase resulted in a RNA phase map of the ventricles. The following results were considered to be consistent with the known electrophysiology of VT: (1) the phase map correlated with QRS morphology and axis in most but not all tachycardias; (2) earliest phase usually demonstrated the VT origin to be at the border of the ventricular wall motion abnormality; (3) endocardial mapping (available in one patient) showed close correlation with RNA phase mapping; (4) in three patients with ischemic heart disease, VT with left bundle branch block (LBBB) pattern had earliest LV phase along the septum; and (5) for one patient imaged during two different VT morphologies, the tachycardias had earliest phase at different borders of the same wall motion abnormality with differing progression of phase across the ventricles. RNA phase mapping of VT is feasible and appears to provide data consistent with the electrophysiology of this arrhythmia. (AM HEART J 103:319, 1982.)

Steven Swiryn, M.D., Dan Pavel, M.D., Ernest Byrom, Ph.D., Robert A. Bauernfeind, M.D., Boris Strasberg, M.D., Edwin Palileo, M.D., Wilfred Lam, M.D., Christopher R. C. Wyndham, M.D., and Kenneth M. Rosen, M.D. Chicago, Ill.

Mapping of ventricular tachycardia (VT) has been employed as a guide to surgical therapy for this rhythm disturbance.1-4 More generally, mapping has provided insight into the pathophysiology of VT.5 Currently, mapping requires invasive techniques either in the electrophysiology laboratory or in the

From the Section of Cardiology, Department of Medicine, and the Section of Nuclear Medicine, Department of Radiology, Abraham Lincoln School of Medicine, University of Illinois College of Medicine.

Supported in part by National Heart, Lung and Blood Institute’s Institu- tional Training Grant HL 07387, Research Grants HL 18794 and HL 23566, and by grants from the Eleanor B. Pillsbury Resident Trust Fund and the Banes Estate.

Received for publication Sept. 8, 1981; accepted Oct. 15, 1981.

Reprint requests: Steven Swiryn, M.D., Cardiology Section, University of Illinois Hospital, P.O. Box 6998, Chicago, IL 60680.

0002-8703/82/030319 + 14$01.40/O g 1982 The C. V. Mosby Co.

operating room. In the electrophysiology laboratory, this requires placement and manipulation of electrode catheters in the left ventricle (LV). In the operating room, this extends the required duration of cardiopulmonary bypass and sometimes it is difficult to induce the tachycardia while on bypass.

Our group and others have recently reported that Fourier analysis (and related techniques) of gated radionuclide angiography (RNA) provides noninvasive data reflecting abnormalities of ventricular activation.6-10 In the present study, we report results of Fourier analysis of RNA gated studies acquired during sustained VT in patients with clinical VI?. These data show that such studies are feasible and seem consistent with the known electrophysiology of this arrhythmia.

319

320 Swiryn et al. March. 1982

Amarlcan Haart Journal

METHODS

Patient selection. Criteria for inclusion in the study group were the following: (1) history of ECG-documented paroxysmal sustained VT; (2) reliable induction and termination of VT in the electrophysiology laboratory (five patients) or spontaneous VT (one patient); (3) hemodynamically stable, well tolerated, sustained VT, (4) signing of informed consent; and (5) acquisition of RNA gated study during sinus rhythm and again during VT.

RNA acquisition. ECG-gated equilibrium RNA cardiac studies were performed using an in vivo red blood cell labeling method with Tc-99m.” A portable Anger camera (Technicare Sigma 420) with high sensitivity collimator (five patients) or large field of view Anger camera (Searle LFOV) with low energy all-purpose collimator (one patient) was positioned for left anterior oblique (LAO) view with the patient supine. Each study was acquired over a 6- to lo-minute period and stored in 64 frames of 32 X 32 matrix resolution. Data were recorded and ana- lyzed on an Informatek Simis 4 computer. For five patients imaged in the electrophysiology laboratory, the sinus rhythm RNA acquisition was done first, followed immediately by RNA acquisition during induced VT. The sixth patient was studied in the nuclear medicine laboratory during spontaneous VT and again during sinus rhythm 6 days later. All patients were taking antiarrhythmic drugs during both the sinus rhythm and VT studies, which did not prevent induction of sustained VT but decreased the tachycardia rate and therefore made the arrhythmia hemodynamically suitable for study. There was no difference in drug regimen for any patient between the two RNA acquisitions.

Electrophysiologic study. The five patients imaged in the electrophysiology laboratory were undergoing serial drug testing for determination of prophylactic drug therapy of recurrent VT. 12-16 In each case, imaging was done after completion of that day’s electrophysiologic study (EPS) on an unsuccessful drug regimen. Reinduction and termination of VT for imaging was accomplished using standard catheter stimulatory techniques. In one patient (No. 2) endocardial mapping was done in the electrophysiology laboratory using previously described techniques.5 This map was redrawn by one of us (C.W.) in a LAO view for purposes of comparison to the RNA phase map. The electrical map was drawn without knowledge of the results of RNA phase mapping.

Routine processing of RNA studies. Ejection fraction (EF) was calculated using a program based on three regions of interest: end-diastolic (ED), end-systolic (ES), and perisystolic (PS) background.17 Regions of interest were drawn semiautomatically, with the septum and aortic valve plane completed by hand using the amplitude image as a guide. Regional wall motion (RWM) was

evaluated using the display of superimposed ED and ES contours and three functional images: the stroke volume (SV) image, the paradox image (ES minus ED), and the regional EF image.

RNA phase analysis. The amplitude and phase of the first Fourier harmonic*8 of the time-activity curve in each

picture element (pixel) were computed. From these values, three displays were generated: (1) the amplitude image by color coding of the value of amplitude for each pixel; (2) the phase image by color coding of the value of phase for each pixel; and (3) the phase distribution histogram representing the number of pixels in a region of interest with each value of phase. Further analysis generated displays showing the location of areas of earliest and latest phase within any area of interest. The phase image is masked by new regions of interest corresponding to the left (LV) and right (RV) ventricles. These regions are determined using thresholding by the amplitude image with verification against a gradient image (image of spatial gradients of the ED image). Pixels with values of amplitude less than 6% of the maximum were excluded, unless they were within the periphery of the ventricle as indi- cated by the gradient image. In the phase distribution histogram, the x-axis represents channels with a width of 3 degrees for phase angle. The y-axis represents the number of pixels in each channel. A flexible cyclic color scalelg could be translated along the phase distribution histogram with corresponding changes of colors on the phase image for study of specific features and visualization in a cine- matic format for display of phase. In this color scale, light blue, dark green, light green, yellow, brown, red, purple, and white (in that order) cover 36-degree (10%) segments of the cycle with a ninth color (dark blue) used to cover the remaining 72 degrees between white and light blue.

RNA areas of earliest and latest phase. Areas of earliest and latest phase within each ventricle were determined automatically using the following algorithm. The phase distribution histogram was scanned from the left- ward extreme (earliest phase) and, comparably, from the rightward extreme (latest phase) moving toward the center until a value of phase was reached which, when used as a cutoff value, included a minimum of 20 pixels of earliest or latest phase, respectively. To exclude outlying pixels which might represent noise, a minimum requirement for area was included. Therefore any pixel not spatially contiguous to at least three other earliest (or latest) pixels was discarded. The remaining earliest and latest pixels were defined and displayed as new regions of interest within silhouettes of the ventricular regions of interest as areas of earliest and latest phase, respectively. For purposes of this study, an area of earliest RV or LV phase was

considered “earlier” if the cutoff value used for its definition was lower than the cutoff value used for the opposite ventricle. This relationship (the relative timing of earliest events in the two ventricles) was quantified by defining AEarly as the cutoff value for the LV area of earliest phase minus the cutoff value for the RV area of earliest phase. AEarly was calculated for each patient during sinus rhythm and again during VT.

RESULTS

Patient poputatton. Seven tachycardia morphologies were imaged in six patients. One patient (No. 1) was a 20-year-old woman with idiopathic cardio-

Volume 103

Number 3 Regional phase RNA mapping in VT 321

Fig. 1. Functional images in patient No. 1. Shown on the left are the stroke volume image (above) and paradox image (below) from patient No. 1 during sinus rhythm. Shown on the right are the comparable images during VT. Superimposed upon each image is the outline of the LV end-diastolic region of interest. Note that both functional images were normal during sinus rhythm. During VT, the stroke volume image is markedly abnormal and the paradox image demonstrates an area of inferoapical “dyskinesia.” RNA phase analysis during VT showed this inferoapical area to be an area of early phase near the origin of the VT, not an area of late phase, as would be expected for dyskinesia.

myopathy. Cardiac catheterization showed normal wall motion and coronary anatomy. This patient was studied during spontaneous VT while taking oral aprindine, 100 mg each 12 hours. The sinus rhythm study was acquired 6 days later on the same drug regimen, when sinus rhythm had been reestablished for 2 days. The other five patients ranged in age from 60 to 67 years. Four were men and one was a woman. All had advanced ischemic heart disease (IHD), with critical obstruction of two or three coronary vessels and a LV aneurysm. These five patients were studied in the electrophysiology laboratory, with acquisition of the sinus rhythm study followed by induction of VT and acquisition of the tachycardia study. These five patients were also studied on antiarrhythmic drugs which included oral quinidine sulfate, 1600 mg/day (two patiepts); oral disopyramide phosphate, 1200 mg/day (one patient); intravenous aprindine, 300 mg (one patient); and oral quinidine sulfate, 1600 mg/day

plus intravenous procainamide hydrochloride, 1000 mg (one patient).

Ventricular tachycardia. All patients in this series had paroxysmal sustained VT, and thus RNA studies were acquired during sustained VT. Cycle lengths during the acquisition in sinus rhythm ranged from 1080 to 540 msec (mean f SD, 739 + 188 msec), corresponding to heart rates of 56 to 111 beats/min. Cycle lengths during the acquisition in VT ranged from 450 to 315 msec (389 -t 46 msec), corresponding to heart rates of 133 to 190 beats/min. Three ventricular tachycardias had left bundle branch block (LBBB) morphology while four had right bundle branch block (RBBB) morphology. By study design, all were hemodynamically stable and well tolerated.

Cuff blood pressures were recorded for each patient during sinus rhythm and during VT. In one patient, VT had been sustained for several days before study, Her blood pressure (BP) was 120/70

322 Swiryn et al. March, 1982

American Heart Journal

Fig. 2, A through D. RNA phase images and phase distribution histograms of each patient for sinus rhythm and for VT. In each panel the label “THOO” denotes a color scale in which each of eight colors covers a range of 10% of the cycle, or 36 degrees. Below THOO is a number denoting the position in degrees to which the center of the color scale (between yellow and brown) has been translated. The eight colors, in order of increasing (later) phase are light blue, dark green, light green, yellow, brown, red, purple, and white. A ninth color, dark blue, is used to fill the remaining 72 degrees of the cycle. Below these color scale data on the right of each panel are the phase images of the individual LV (aboue) and RV (below) regions of interest, created by color coding of the value of phase for each picture element (pixel). To the left of each phase image is the corresponding phase distribution histogram, representing the number of pixels in the ventricle with each value of phase. The x-axis represents channels with a width of 3 degrees for phase angle. The y-axis represents the number of pixels in each channel. Color coding is the same as for the phase image. A, Patient No. 1 during sinus rhythm. B, Patient No. 1 during VT. C, Patient No. 2 during sinus rhythm. D, Patient No. 2 during VT. See text for detailed descriptions of each panel.

mm Hg both during VT and later during sinus rhythm. In the other five patients, BP measure- ments were available immediately upon induction of VT. All these patients experienced decline in BP upon induction with prompt return to control values, as has previously been reported in patients with paroxysmal supraventricular tachycardia.17

RNA gated study-routine processing. EF during

sinus rhythm ranged from 21% to 50% (35 + 13%). EF during VT ranged from 17 % to 37% (25 f 7 % ) and had fallen in four of the six patients. This result did not quite reach statistical significance (p = 0.08). RWM studies reflected the location and extent of ventricular aneurysms in the five IHD patients. In general, RWM abnormalities, when evaluated by routine processing, were subjectively

Voium. 103


3 dm Fig. 2, E through I. E, Patient No. 3 during sinus rhythm. F, Patient No. 3 during VT. G, Patient No. 4 during sinus rhythm. 25, Patient No. 4 during VT (morphology 1). I, Patient No. 4 during VT (morphology 2). See legend to Fig. 2, A through D, for explanation.

324 Swiryn et al.

Fig. 2, J through M. J, Patient No. 5 during sinus rhythm. K, Patient No. 5 during VT. L, Patient No. 6 during sinus rhythm. M, Patient No. 6 during VT. See legend to Fig. 2, A through D, for explanation.

more extensive and severe during VT. Routinely processed RWM studies were especially interesting in the patient with cardiomyopathy. During sinus rhythm, wall motion was essentially normal. Howev- er, during VT, routine functional images demonstrated what would usually be considered clear evidence of dyskinetic RWM in the LV inferoapical segment (Fig. 1). Examination of the relative timing of LV events with phase analysis showed that, although this area was “dyskinetic” in the sense that it filled and emptied in opposition to the remainder of the ventricle, it was not an area of late phase usually seen with a dyskinetic ventricular aneu- rysrnm (See, for example, Fig. 2, E) but was instead

an area of early phase presumably near the origin of the tachycardia (Fig. 2, B).

RNA phase analysis d&a. Patient No. 1 was a 20-year-old woman with cardiomyopathy but normal wall motion demonstrated during cardiac catheterization. During sinus rhythm, the phase image was essentially normal, with relatively uniform phase across both ventricles (Fig. 2, A). This resulted in phase distribution histograms with single discrete peaks. Note the normal finding of an area of early phase (in dark green) located in the high anterior LV. VT in this patient was characterized eh&roca&iogr~hi4y by R@B pattern with left axis deviation (LAD) (Fig. 3, top panel). Phase

Volume 103


Patient 4a

Patient 6

Fig. 3. Electrocardiograms during VT. Shown are ECG leads I, II, III, V1, and V, during seven VT morphologies in six patients. Note that two different tachycardia morphologies were imaged in patient No. 4.

analysis during VT showed an area of very early phase in the inferoapical LV (Fig. 2, B). From there, phase became progressively later as shown by progression of light blue to green, yellow, brown, red, purple, and white toward the inferoapical RV, to the remainder of the RV, and toward the remainder of the LV. Note that much of the RV (in purple and white) had later phase than the LV (in red), corresponding to electrocardiographic RBBB pattern. This was confirmed by the histograms, which

showed that the major peak representing the majority of pixels over the RV was later than the major peak for the LV. In addition, the apical areas of the ventricles were of relatively earlier phase than the bases, corresponding to electrocardiographic LAD.

Patient No. 2 was a 67-year-old woman with three-vessel coronary obstruction and a discrete in- feroposterior LV aneurysm. During sinus rhythm, the phase image showed a large area of relatively late phase over the lower portion of both the RV and

326 Swiryn et al. March, 1982


ORS Lead II

-flf

1 -40 0 40 80 120 160

blsec. after ORS onset

* Continuous veflffkuhr

ucfivify

0 Apia1 segment

I‘; Mid segment -

:‘: : : :.;

Eoral scqmcnt

Fig. 4. Endocardial ventricular map during VT in patient No. 2. This endocardial map was redrawn in a left anterior oblique view for purposes of comparison to the RNA phase map. In this view, both ventricles are shown schematically in three segments, the apical segment (continuous lines with large numerals), the midsegment (dashed lines with smaller numerals), and the basal segment (dotted lines with smallest numerals). Measure- ments are in msec relative to the onset of the QRS as shown in the lower left corner. Note the area of continuous electrical activity at the posterobasal septum denoted by an asterisk and the adjacent earliest electro- gram recorded at 55 msec before the onset of the QRS. However, endocardial activation then spreads to the RV, with late activation of the remainder of the LV. This is consistent with the LBBB morphology of this VT.

LV projections, perhaps reflecting contribution from normal anterior and apical areas superimposed upon the abnormal wall motion posteriorly (Fig. 2, C). VT in this patient was characterized electrocardiographically by LBBB pattern with LAD (Fig. 3, second panel). Endocardial mapping during VT (Fig. 4) showed continuous activity in the LV aneurysm (posterior and posteroseptal sites) and a single adjacent site at -55 msec with all other LV sites being relatively late (35 msec after onset of QRS or later). On the other hand, RV posteroseptal endocardial activation occurred at 22 msec before onset of QRS. Thus in this patient the activation front during VT crossed the septum to the posteroseptal RV earlier than it spread to the remainder of the LV, accounting for the LBBB pattern with LAD. Phase analysis during this VT (Fig. 2, D) showed an area of earliest phase in the inferior RV (dark green)

followed by progression of phase through much of the RV and to the septal LV (light green and yellow), to the remainder of the RV (brown and red), and finally to the majority of pixels in the LV (red, purple, and white). Note that this pattern was consistent with the QRS morphology and, with the exception of the lack of any feature corresponding to the LV area of continuous activity, remarkably consistent with the endocardial map (Fig. 4).

Patient No. 3 was a 60-year-old man with three- vessel coronary obstruction and an inferolateral LV aneurysm. During sinus rhythm the phase image showed an area of late phase corresponding to the aneurysm (Fig. 2, E). Otherwise the phase image is normal, including a normally located area or earliest phase in the high anterior LV (dark green). VT in this patient was characterized electrocardiographically by RBBB pattern with a northwest QRS axis (Fig. 3, third panel). Phase analysis during this VT showed an area of earliest phase near the aneurysm (light blue) with progression of phase across the LV (green, yellow, brown, and red), and finally to the RV (brown, red, purple, and white; Fig. 2, F). The phase image during this VT was therefore consistent with QRS morphology and with VT origin in or near the aneurysm.

Patient No. 4 was a 67-year-old man with two- vessel coronary obstruction and an extensive inferoapical aneurysm. During sinus rhythm the phase image showed a large area of late phase in the inferoapical LV corresponding to the aneurysm (Fig. 2, G). This patient was imaged during VT of two distinct QRS morphologies. The first VT was characterized electrocardiographically by LBBB pattern with LAD (Fig. 3, fourth panel). Phase analysis during this VT (Fig. 2, H) showed earliest phase at the anteroseptal border of the aneurysm (light blue) with progression of phase clockwise across the LV (green, yellow, brown, and red) and relatively early progression to the RV (light blue, green, yellow and brown). The second VT morphology in this patient was characterized electrocardiographically by RBBB pattern with northwest QRS axis (Fig. 3, fifth panel). Phase analysis during this VT (Fig. 2, II showed earliest phase at the posterolateral border of the aneurysm (light blue) with progression of phase counterclockwise across the LV (green, yellow, brown, and red). Progression of phase to the RV (brown, red, and purple) is quite late.

Patient No. 5 was a 64-year-old man with two- vessel coronary obstruction and a posterior LV aneurysm plus concomitant anterior and septal aki- nesia. During sinus rhythm the phase image showed slightly late phase (yellow) along the LV septum,

Volume 103


Fig. 5. Areas of earliest and latest phase. Shown are outlines of the LV end-diastolic region of interest in patient No. 3. On the left are outlines during sinus rhythm with automatically determined areas of earliest phase (upper panel) and areas of latest phase (lower panel). On the right are comparable areas of earliest and latest phase during VT. Note that the area of earliest LV phase during sinus rhythm (upper left panel) is in the high anterior ventricle, a normal location. The area of earliest LV phase during VT (upper right panel) is abnormally located and is adjacent to the LV aneurysm, which was identified as an area of latest phase during sinus rhythm (lower left panel).

reflecting the anteroseptal RWM abnormality (Fig. 2, J). VT in this patient was characterized electrocardiographically by RBBB pattern with LAD (Fig. 3, sixth panel). The phase image during VT (Fig. 2, K) showed generally earlier RV and LV inferoapical phase (green) and later phase at the base (brown and red), consistent with LAD. Although there were more green (early) pixels in the LV and more red (late) pixels in the RV, the correlation with RBBB morphology is not striking (in contrast to RNA phase images in Fig. 2, B, 2, F, and 2, I). Areas of earliest phase were located in a thin rim (blue) at the posterolateral LV and the RV inflow area (with the LV only slightly earlier).

Patient No. 6 was a 62-year-old man with two- vessel coronary obstruction and both inferior and anteroseptal RWM abnormalities. During sinus rhythm, the phase image showed areas of late phase in each ventricle corresponding to the RWM abnormalities (Fig. 2, I,). Note that the areas of earliest LV phase (blue and dark green) are normally located in the high anterior and the posterolateral ventricle. VT in this patient was characterized electrocardio-

graphically by LBBB pattern with normal QRS axis (Fig. 3, seventh panel), which, except for QRS duration, closely resembled QRS morphology during sinus rhythm. In addition, the QRS was relatively narrow when compared with the other patients in this study, and the patient had been noted clinically to also have VT with RBBB pattern (this second morphology was not imaged). The phase image during VT was similar to that during sinus rhythm (Fig. 2, M). Notable was the alteration of the LV area of earliest phase to a location along the upper septum and the replacement of the early posterolat- era1 area during sinus rhythm by a late area. The LV area of earliest phase was slightly earlier than that for the RV.

RNA areas of earliest phase. During sinus rhythm, automatically identified areas of earliest LV phase were located normally in all patients (high anterior LV) (Fig. 5, top left). During VT, areas of earliest LV phase were located adjacent to the RWM abnormality in four of six VT studies in the five patients with ventricular aneurysms (Fig. 5, top right). In the remaining two patients (both with posterior aneu-

328 Swiryn et al.

60

- 140 )r T - 0 160

- 180

L -240 Sinus Ventricular Rhythm Tachycordia

Fig. 6. Comparison of calculated AEarly during sinus rhythm and VT. On the y-axis is AEarly in degrees calculated as the cutoff value for the LV area of earliest phase minus the cutoff value for the RV area of earliest phase. Each open circle (on the left) denotes the value of AEarly during sinus rhythm (note that the middle open circle denotes the value of AEarly for patient No. 4 with two tachycardia morphologies, and for patient No. 6). Each closed circle (on the right) denotes the value of AEarly during VT. Tachycardias of RBBB and LBBB morphology are denoted by solid and dashed lines, respectively. A value for AEarly of 0 degrees would imply comparable timing of earliest LV and RV events. A positive value for AEarly implies that RV earliest events precede LV earliest events. A negative value for AEarly implies that LV earliest events precede RV earliest events. Note that AEarly decreases during ail seven tachycardias compared to sinus rhythm. AEarly is negative in six of seven tachycardias including two of three with LBBB morphology, implying a LV VT origin.

rysms) LV areas of earliest phase were not adjacent to the RWM abnormality.

Using the cutoff values for the areas of earliest phase as a measure of earliness, the relative timing of earliest events in each ventricle was compared by calculating AEarly (LV cutoff value minus RV cutoff value). Thus when the RV area of earliest phase had earlier (lower) phase, AEarly was a positive number.

March, 1982


When the LV area of earliest phase had earlier (lower) phase, AEarly was a negative number. Dur- ing sinus rhythm AEarly ranged from -11 to 56 (15 f 26 degrees; Fig. 6). During VT AEarly decreased in all patients (the LV was relatively earlier than during sinus rhythm) to a range of -224 to 30 (-84 +- 97 degrees, p < 0.01). As would be expected, this decrease in AEarly was especially marked in VT characterized by RBBB pattern. However, AEarly also decreased during all three tachycardias with LBBB pattern, implying that these also were of LV origin. Note that during six of seven tachycardias, including two of three with LBBB pattern, AEarly was a negative number, implying that LV earliest events preceded RV earliest events. The seventh VT (patient No. 2) was associated with early RV endocardial activation and relatively late spread of activation to most of the LV (Fig. 4).

DISCUSSION

Clinical utility of cardiac electrical mapping. Electri- cal mapping of the activation of the heart during sinus rhythm with normal conductior+ 22 and with bundle branch block,23-25 fascicular block,26 or preex- citation27 has provided insight into physiology and pathophysiology unavailable by any other means. In addition, such studies have provided validation of hypotheses based upon surface ECG28 and, in some cases, represent the strongest gold standard for ECG diagnoses. Electrical mapping of atrial and ventricular activation in patients with paroxysmal supraventricular tachycardia contributes to the definition of tachycardia mechanism and, in patients with preex- citation, is used as a guide for surgical therapy.2g-32 Recently, mapping studies of ventricular endocardial and epicardial activation during sustained VT have been used as a guide for surgical therapy for this rhythm disturbance.l-J Such studies are accomplished with invasive techniques in the electrophysiology laboratory or the operating room.

Feasability of regional phase RNA mapping In hemodynamically stable VT. Fourier analysis of RNA gated studies has been used by our laboratory and others for identification and quantification of.RWM abnormalities.20*33s34 In addition, we and others have reported that abnormal ventricuhu activation also

alters the phase image in characteristic ways.6-10 In the present study, we have applied this technique to patients with sustained VT. We have demonstrated that, in a group of patients with hemodynamically stable VT, equilibrium RNA imaging during VT is feasible. This is obviously not the case for many patients with hemodynamically unstable VT. How-

VOlUme 103


ever, the addition of antiarrhythmic drugs which slow the VT rate may allow imaging in some patients in whom such analysis is otherwise impossible. It is clear that the procedures in the present study are not completely noninvasive, since laboratory induction of VT was used for RNA imaging.

Effects of VT on global and regional LV function assessed by routine RNA processing. Routine processing of RNA gated studies during VT provided important information. LV function generally wors- ened when compared with sinus rhythm, as evi- denced by a marked decrease in EF in those patients not already having a markedly decreased EF during sinus rhythm and by subjectively more abnormal RWM. Since we imaged only patients with hemodynamically stable VT, these results underestimate the deficit in LV function caused by VT in general.

We have previously reported that LV function (and specifically EF) is preserved during paroxysmal supraventricular tachycardia (PSVT) unless the heart rate is extremely fast or underlying heart disease is present .17 Heart rates during VT in most of the patients in the present study were not rapid enough (to a large extent because of antiarrhythmic drug therapy) to account for the observed worsening of function. The severe organic heart disease in these patients, with perhaps some adverse effect of the antiarrhythmic drugs, must be a major factor. In addition, Badke et a1.35 have recently reported that ventricular pacing greatly changed both regional and global LV function “presumably because volume is sequestered and pressure is dissipated into relatively inactive segments that are out of phase with the bulk of contracting myocardium.“35 The loss of atria1 contribution to ventricular filling cannot be a major factor, since function was preserved in patients with PSVT, which is also associated with mis-timing of atrial contraction.

Especially noteworthy was the routine processing of the VT RNA study in patient No. 1 with cardiomyopathy. Subtraction images demonstrated what would usually be considered clear evidence of dyskinetic RWM. Examination of the RNA phase image showed that this area was early, not late as would be expected for regional dyskinesis.20 Although indis- tinguishable by routine RNA processing from a ventricular aneurysm, the pathophysiology of this RWM abnormality was distinctly different and related to the activation sequence of the VT itself, not to decreased LV function in the usual sense. This has been described by Herman and Gorlin36 as “asynchrony.” That the end result of this “asynchrony” is in fact comparably decreased LV func-

tion is demonstrated by the fall in EF from 50% during sinus rhythm to 31% during VT.

Relations of RNA phase analysis of VT to electrophysiologic characteristics of the arrhythmia. Phase mapping during VT was consistent with the known electrophysiology of this arrhythmia in a number of ways. First, the general pattern of the RNA phase map usually reflected both QRS morphology and axis. When this was not the case, the phase map seemed to be more consistent with the probable electrophysiology of the tachycardia than with QRS morphology. For example, the phase image of the first VT morphology in patient No. 4 (LBBB pattern) showed earliest phase in the LV, the expected origin of VT. in a patient with 1HD.S The pitfalls of VT mapping by QRS morphology have been well described.5 Although a LBBB QRS morphology might imply a RV origin of the VT, this is usually not the case in patients with IHD. RNA phase mapping of the first VT morphology in patient No. 4 was encouraging in this regard. All of the ventricular tachycardias imaged in this study were thought (although not proved) to arise in the LV. Calculated AEarly, a measure of the relative earliness of events in the two ventricles, decreased from the sinus rhythm value during all seven tachycardias. AEarly was negative (implying that LV earliest events preceded those in the RV) during six of seven tachycardias including two of three with LBBB pattern. In contrast, eight patients with LBBB previously reported from our laboratory had a positive AEarly which ranged from 23 to 56 (37 + 11) degrees reflecting late LV activation.”

Second, VT could be demonstrated to arise near the border of the RWM abnormality in the majority of instances. This has been shown to be the case for patients with VT complicating LV aneurysms.5 In two patients (No. 2 and 5) the phase map did not conform exactly to this expectation. These patients each had posterior aneurysms not well demonstrated during sinus rhythm in the LAO view chosen. It is possible that other views might have better demonstrated the VT origin. It must be remembered that the RNA phase image is a two-dimensional projection of three-dimensional events. A local volume of the ventricle having early phase over- lapped in the chosen projection by a local volume of late phase may result in an area of intermediate phase and prevent identification of the early event. An example of this may be seen during sinus rhythm in patient No. 2 (Fig. 2, C) where a discrete posterior aneurysm was seen only as an inferoseptal area of slightly late phase. This area is not of phase comparably late to previous results in our laboratory in

330 Swiryn et al.

patients with ventricular aneurysmszO and may be the result of an overlying volume of ventricle having normal phase added to the effect of the aneurysm.

A third consistency of RNA phase mapping with the known electrophysiology of VT was demonstrated by the close concordance of the endocardial map with the phase map in patient No. 2 (Fig. 4). Although not perfect, the RNA phase map did reproduce most of the major features of the endocardial map in this patient. No obvious feature of the phase map corresponds to the continuous activity in the left posteroseptal region recorded by the catheter technique, thus making this the only patient in whom AEarly was positive during VT (implying a RV tachycardia origin). Confirmation of this phase mapping technique clearly requires more correlation with endocardial mapping studies. Again, it must be pointed out that an endocardial map is three- dimensional, while the RNA phase map is two- dimensional, making exact correlation difficult to evaluate.

A fourth finding, the features of VT arising in or near the septum, is demonstrated by the RNA phase map in patient No. 6. This patient had septal wall motion abnormality and clinically documented ventricular tachycardias of both RBBB and LBBB pattern. Imaging during the VT of LBBB morphology and a relatively narrow QRS (Fig. 3), demonstrated early phase along the septum in both the RV and LV regions with the left septum slightly earlier (Fig. 2, M). Josephson et al5 have shown that a relatively narrow QRS during VT implies a septal origin. In all three LBBB pattern tachycardias imaged in this study (all in patients with IHD), the earliest area of LV phase was located near the septum. This is consistent with the data of Joseph- son et a1.5

A fifth feature of the electrophysiology of VT is demonstrated by patient No. 4 (Figs. 3,2, H, and 2, I). This patient was imaged during tachycardias of two different QRS morphologies. Phase analysis showed that both tachycardias arose at the border of the LV aneurysm, with differing progression of phase to the remainder of the ventricles. That areas of earliest phase were located at different borders of the aneurysm seems to be the radionuclide correlate of varying exits from the same reentrant circuit. Production of differing QRS morphologies by tachycardias of differing exit and/or spread of activation from the same reentrant circuit has been shown electrophysiologically.5* 37, 38

Technical considerations of RNA phase mapping in VT. Phase is cyclic with 360 degrees adjacent to 1 degree. Very late phase is close to very early phase

(just as late events in the cardiac cycle are close in time to early events in the next cycle). However, with the cyclic color scale used to generate the phase images in this study, pixels with a phase of 360 degrees and pixels with a phase of 1 degree may be assigned different colors and will appear at opposite ends of the histogram. This has not been a problem with phase images of normal ventricles or even of the very abnormal ventricles in the present study during sinus rhythm. In some phase images during VT, however, there is wraparound from the 360 degree to the 1 degree end of the scale (Fig. 2, D). There are several possible reasons for this. First, mean phase increases with decreasing cycle length so that delayed areas (i.e., the aneurysm) are moved toward the 360 degree “end” of the histogram and may wrap around toward the 1 degree “end.”

Second, the definition of zero phase depends upon what feature of the ECG is used to trigger the start of a new cycle. During VT (with an abnormal QRS complex), although the trigger point is identical for each acquired cycle, its precise timing is not known at present. Triggering late in the QRS might allow areas of ventricle which were depolarized very early to be measured as late in phase. At short cycle lengths, since a given increment in time corresponds to more phase delay, a slight change in the trigger point might result in significant changes in measured phase.

Third, even if the trigger were perfect (at the onset of the QRS), activation during VT which occurs before the QRS, for example, at -55 msec in patient No. 2 (Fig. 4) might still be measured as late in phase, partly depending upon the time lag between electrical and mechanical events. Of course, one could argue that -55 msec (during a tachycardia of cycle length 315 msec) is not very early, but is in fact very late, occurring 260 msec after the onset of the QRS. In practice the continuity and internal consistency of data make confusion of early and late an uncommon problem during electrical mapping. The same is true for RNA phase mapping. It should be pointed out that none of these potential problems affects the pattern and sequence of the RNA phase image. The wraparound which was noted affects only very slightly if at all the location of the area of earliest phase. The point of triggering has no effect on calculation of the first Fourier harmonic (on the relative phase values at different areas).

The calculation of AEarly would be affected. For example, if the blue pixels at the extreme right of the histogram in patient No. 2 during VT (Fig. 2, D) were an artifact of late triggering and should in fact have appeared at the extreme left of the histogram,

VOlume 103


the calculated AEarly would increase because of the addition of earlier RV pixels. This would have resulted in an increase in AEarly compared to sinus rhythm in this patient. Comparison of AEarly during sinus rhythm and VT for the other patients would not be affected significantly by this problem.

Conclusions. Seven morphologies of VT were imaged in six patients. RNA phase mapping seemed to correlate well with the known electrophysiology of this arrhythmia. Further validation of this technique by comparison with endocardial mapping will be helpful in understanding its advantages and limitations.

We are grateful for the generous contribution of original computer programs by F. Bitter of the University of Ulm and by F. Deconinck of the University of Brussels, and for the invaluable assistance of the technical staffs of the Sections of Cardiology and Nuclear Medicine.

REFERENCES

1.

2.

3.

4.

5.

6.

7.

Gallagher JJ, Oldham HN, Wallace AG, Peter RM, Kasell J: Ventricular aneurysm with ventricular tachycardia. Report of a case with epicardial mapping. Am J Cardiol35:696,- 1975. Snurrell RAJ. Yates A. Thorburn CW. Lawton GE. Deuchar DC: Surgical treatment of ventricular’tachycardia after epicardial mapping studies. Br Heart J 37:115, 1975. Fontaine G, Guiraudon G, Frank R, Coutle R, Dragcdanne C: Epicardial mapping and surgical treatment in six cases of resistant ventricular tachycardia not related to coronary artery disease. In Wellens HJJ, Lie KL, Janse MJ, editors: The conduction system of the heart. Philadelphia, 1976, Lea & Febiger, pp 545-563. Horowitz LN, Josephson ME, Harken AM: Ventricular resec- tion guided by epicardial and endocardial mapping for treatment of recurrent ventricular tachycardia. N Engl J Med 302:589, 1980. Josephson ME, Horowitz LN, Farshidi A, Spear JF, Kastor JA, Moore EN: Recurrent sustained ventricular tachycardia. 2. Endocardial mapping. Circulation 57:440, 1978. Swiryn S, Pave1 D, Byrom E, Witham D, Wyndham C, Bauernfeind R, Rosen KM Sequential regional phase mapping of radionuclide gated biventriculograms in patients with left bundle branch block. AM HEART J 102:1000, 1981. Turner DA, Von Behren PL, Ruggie NT, Hauser RG, Denes P, Messer JV, Fordham EW, Groch MW: Identification of initial site of ventricular contraction by least square phase analysis of radionuclide cineangiograms (abstr). Circulation (suppl III) 62:111-148, 1980. -

8. Botvinick E, Dunn R, O’Connell W, Shosa D, Tucker C, Herfkens R: The phase image-Its relationship to patterns of contraction and conduction (abstr). Circulation (suppl III) 62:111-148, 1980.

9. Podrid PJ, Zielonka J, Lown B, Holman BL, Carr S, Young J, Biflock L: The use of radioventriculoeranhv for localizine the

1 . I

site of origin of ventricular tachycardia (abstr). Circulation (suppl III) 62:111-148, 1980.

10. Links JM, Douglass KH, Wagner HN Jr: Patterns of ventricular emptying by Fourier analysis of gated blood-pool studies. J Nucl Med 21:97'S, 1980.

11. Pave1 DG, Zimmer M, Patterson VN: In vivo labeling of red blood cells with 99m Tc: A new annroach to blood ~001 visualization. J Nucl Med l&305, 1977.

12. Fisher JD, Cohen HL, Mehra R, Altschuler H, Escher DJW, Furman S: Cardiac pacing and pacemakers. II. Serial electro-

13.

14.

15.

physiologic-pharmacologic testing for control of recurrent tachyarrhythmias. AM HEART J 93:658,1977. Hartzler GO, Maloney JD: Programmed ventricular stimula- tion in management of recurrent ventricular tachycardia. Mayo Clin Proc 52:731, 1977. Mason JW, Winkle RA: Electrode-catheter arrhythmia induction in selection and assessment of antiarrhythmic drug therapy for recurrent ventricular tachycardia. Circulation 56:971, 1978. Horowitz LN, Josephson ME, Farshidi A, Spielman SR, Michelson EL, Greenspan AM: Recurrent sustained ventricular tachycardia. 3. Role of the electrophysiologic study in selection of antiarrhythmic regimens. Circulation 56:986, 1978.

16. Denes P, WkD, Wyndham C, Dhingra R, Bauernfeind R, Swiryn S, Rosen KM: Chronic long-term electrophysiologic study of paroxysmal ventricular tachycardia. Chest 77:478, 1980.

17.

18.

19.

20.

21.

22.

23.

24.

25.

Swiryn S, Pave1 D, Byrom E, Wyndham C, Pietras R, Bauernfeind R, Rosen KM: Assessment of left ventricular function by radionuclide angiography during induced supraventricular tachycardia. Am J Cardiol47:555, 1981. Feynman RP, Leighton RB, Sands M: “The Feynman lec- tures on physics.” vol. 1. Lecture 50. Reading, Mass., 1963, Addison-Wesley. Denconinck F, Bossuyt A, Hermanne A: A cyclic color scale as an essential requirement in functional imaging of periodic phenomena. Med Physics 6:331, 1979. Pave1 D, Swiryn S, Lam W, Byrom E, Sheikh A, Rosen K: Ventricular phase analysis of radionuclide gated studies (abstr). Am J Cardiol 45:398, 1980. Durrer D, Van Dam RTh, Freud GE, Janse MJ, Meigler FL, Arzbaecher RC: Total excitation of the isolated human heart. Circulation 41:899, 1970. Wyndham CRC, Meeran MK, Smith T, Saxena A, Engleman RM, Levitsky S, Rosen KM Epicardial activation of the intact human heart without conduction defects. Circulation 59:161, 1979. Van Dam RTh: Ventricular activation in human and canine bundle branch block. In Wellens HJJ, Lie KI, Janse MJ, editors: The conduction system of the heart. Philadelphia, 1976, Lea & Febiger, p. 377. Horowitz LN, Alexander JA, Edmunds LH: Postoperative right bundle branch block: Identification of three levels of block. Circulation 62:319, 1980. Wyndham CRC, Smith T, Meeran MK, Mammama R, Levitsky S, Rosen KM Epicardial activation in patients with

26. left bundle branch block. Circulation 61:696, 1980. Wyndham CR, Meeran MK, Smith T, Engelman RM, Levit- sky S, Rosen KM: Epicardial activation in human left anterior fascicular block. Am J Cardiol 44:638, 1979.

27. Gallagher JJ, Kassel J, Sealy WC, Pritchett ELC, Wallace AG: Epicardial mapping in the Wolff-Parkinson-White syndrome. Circulation 57:854, 1978.

29.

30.

28. Tonkin AM, Wagner GS, Gallagher JJ. Coue GD. Kassel J. Wallace AG: Initial forces of veniricular’depolarization in the Wolff-Parkinson-White syndrome. Analysis based upon localization of the accessory pathway by epicardial mapping. Circulation 52:1030, 1975. Durrer D, Roos JP: Epicardial excitation of the ventricles in a patient with Wolff-Parkinson-White syndrome. Circulation 35:15, 1967. Burchell HB, Frye RL, Anderson MW, McGoon DC: Atriov- entricular and ventriculoatrial excitation in Wolff-Parkin- son-white syndrome (Type B). Temporary ablation at sur- gery. Circulation 36:663, 1967. Cobb FR, Blumenschein SD, Sealy WC, Boneau JP, Wagner GS, Wallace AG: Successful surgical interruption of the bundle of Kent in a patient with Wolff-Parkinson-White syndrome. Circulation 38:1018, 1968. Gallagher JJ, Gilbert M, Svenson RH, Sealy WC, Kassel J, Wallace AG: Wolff-Parkinson-White syndrome. The prob-

31.

32.

Swiryn et al.

lem, evaluation, and surgical correction. Circulation 51:767, 1975.

33. Adam WE, Tarkowska A, Bitter F, Stauch M, Geffers H: Equilibrium (gated) radionuclide ventriculography. Cardio- vast Radio1 2:161, 1979.

34. Verba JW, Bornstein I, Alazraki NP, Taylor A, Bhargava V, Shabetai R, Lewinter M: Onset and progression of mechanical systole derived from gated radionuclide techniques and displayed in tine format (abstr). J Nucl Med 20:625, 1979.

35. Badke FR, Boinay P, Cove11 JW: Effects of ventricular pacing on regional left ventricular performance in the dog. Am J Physiol 238:H858, 1980.

36. Herman MV, Gorlin R: Implications of left ventricular asynergy. Am J Cardiol 23:538, 1969.

37. Horowitz LN, Greenspan AM, Spielman SR, Josephson ME: Torsades de pointes: Electrophysiologic studies in patients without transient pharmacologic or metabolic abnormalities. Circulation 63:1120, 1981.

38. Josephson ME, Horowitz LN, Farshidi A, Spielman SR, Michelson EL, Greenspan AM: Recurrent sustained ventricular tachycardia. 4. Pleomorphism. Circulation 59:459, 1979.

Young adults with nonsurgically induced complete heart block (CHB) do not necesseriry have a benign prognosis and pacemaker (FM) implantation may be nacaasary. .No OFW has reported long-term PM follow-up In young adults with CHB. We stud&d 13 paNeM m 15 to 37 years (mean 24 years) at PM tmpfantetion. There were nine famaJe and four mate pat&&s. All were functional class II or 111 (NYHA) before PM impl8&8tlon. Syncope, dlx2i8e55, tam, shortness of breath, and dyspnea on exertion were the mout common symptoms. Ca.r&c caiheterfzation findings (11 of 13 patients) were normal in five, and addt$lod card@ an& wefe pre8ent in six. Hs bundle studies (9 of 13 patients) showed absent AH Interv8~8 in 8U wts, v&h HV intervals not identified in two, 20 to 30 msec in one, and 30 to 50 m8ec in stx p#lwtt& Halter monitor recordlnge (8 of 13 patients) demonstrated CHB in all eight with lrrtermfttbnt second- to third-degree block In two of these patients. Two pOtrent h8d occ8sJo88l ee ventrkuter contractions. Stress exercise tests (9 of 13 patbnte) d8moWWed Incr888ed ve~trlcufar rate response (although subnormal in some pat&eat@; symptom8 devew In severt. &e p8tknt had ventricular ectopy. All 13 patient8 were contacted 3 months to 7 ye8rs (me8n 4. years) after PM implantation. Two patients had died, but the de8th8 were not ret8t8d to PTi# dysfuqctfon. Al patient5 who are currently alive had marked improvement i41 f&mWonat sy~ptom8Wogy and all are currently functional class 1. CHB is not a ben@n condition in young adtdts and may require

PM implantation, whkh improves symptoms and alkws the p8tient to lead a normal Ilfe. (AM HEART J 103:332, 1982.)

Dennis C. Besley, M.D., Gregory J. McWilliams, D.O., Douglas S. Moodie, M.D., and Lon W. Castle, M.D. Cleueland, Ohio

Despite the failure of sequential atrioventricular (AV) conduction in patients with nonsurgically induced complete heart block, many reports have documented prolonged survival. Campbell and

From the Department of Cardiology, The Cleveland Clinic Foundation. Received for publication Jan. 23, 1981; revision received May 4, 1981; accepted June 2, 1981.

Reprint requests: Douglas S. Moodie, M.D., Dept. of Cardiology, The Cleveland Clinic Foundation, 9500 Euclid Ave., Cleveland, OH 44106.

Emanuel’ followed survivals of six adult patienti for 34 to 48 yeara. Five had uncam&cati .courses, and only om h& Stokes-Adams attacks, these occurred intermittently over a period of 48 years; with as long as 19 years separating individual episodes. Patients surviving into the sixth and seventh .decades have been described.2t3 Normal physical working capacity with strenuous 8aer&8 haic- been m&ted. A paGent cited by Gw Bt .&L5 tc&mtisk.ol deliveries without hemodynamic or electrical deteri-

332 0002-8703/82/030332 + 06$00.60/O (:) 1982 The C. V. Mosby Co.

sequential regional phase mapping of radionuclide gated biventriculograms in patients with sustained...

Documents