Assessment of QT Dispersion in Symptomatic Patients with Congenital Long QT Syndromes Nicholas J. Linker, MB, ChB, Paolo Colonna, MD, Christopher A. Kekwick, MB, BS, Janice Till, MB, BS, A. John Camm, MD, and David E. Ward, MD
It has been suggested that QT dispersion recorded on the surface electrocardiogram may be a predictor of arrhythmic events in patients with congenital QT prolongation. To evaluate this, 9 patients (6 female, mean age 17.6 years) with congenital long QT syndromes, all of whom had syncope and documented torsades de pointes, were studied. Patients were studied off treatment and during therapy with @blocking agents. Three patients were also studied after left stellate ganglionectomy. An age-matched control group was also studied. Good quality 12-lead electrocardiograms were recorded from all patients. For each lead, QT and RR intervals were measured, and QTc value was calculated. QT and QTc dispersions were calculated for each patient. Patients had a significantly longer mean QT interval compared with that of the control group (450 f 100 vs 359 f 63 ms; p = 0.015) at similar mean RR intervals (736 f 231 vs 783 f 289 ms), with a longer mean QTc value (0.53 f 0.08 vs 0.41 f 0.02 slm; p = 0.004). Patients also had longer QT and QTc dispersions compared with those of the control group (110 f 45 vs 43 f 12 ms[p=O.O04],and O.lOSi0.03~~ 0.05 f 0.02 SIB [p = O&02], respectively). QT and QTc dispersions on and off p-blocking agents were not significantly different. Comparing patients with frequent and those with infrequent symptoms, there was no difference in QT or QTc dispersion either off treatment or during therapy with &blocking agents. There was an increase in mean QT interval and QTc value in patients with frequent symptoms on &blocking agents. On @-blocking agents, patients with frequent symptoms had a significantly longer mean QTc value compared with that of those with infrequent symptoms. There was no change in QT or QTc dispersion, mean QT or RR interval, or mean QTc value afler left stellate ganglionectomy. QT-interval dispersion is observed in patients with congenital long QT syndromes; however, the degree of dispersion is not related to the severity of symptoms, nor influenced by treatment with @ blocking agents. (Am J Cardiol 1992;69:634-638) From St. George’s Hospital and Medical School, Royal Devon and Exeter Hospital, and the Royal Brompton National Heart and Lung Hospitals, London, England. Manuscript received August 27, 1991; revised manuscript received and accepted October 28, 1991. Address for reprints: Nicholas J. Linker, MB, ChB, Department of Cardiological Sciences, St. George’s Hospital and Medical School, Cranmer Terrace, London SW 17 ORE, England.
634
THE AMERICAN JOURNAL OF CARDIOLOGY VOLUME 69
P
atients with congenital QT-interval prolongation have a high incidence of ventricular arrhythmias and suddendeath. Prediction of those patients at risk of these arrhythmias, and the evaluation of asymptomatic family memberswith QT-interval prolongation is difficult. One explanation for the developmentof ventricular arrhythmias in such patients is basedon dispersion of repolarization between different regions of the myocardium. This feature has been observedin patients with congenital QT-interval prolongation by measuring endocardial monophasicaction potentials from multiple siteswithin the ventricles and demonstrating an increase in the dispersion of the duration of the monophasic action potentials.’ Thus, the hypothesishas been advanced that the degree of dispersion between the QT intervals from a standard 1Zlead electrocardiogram may be an indicator of the risk of ventricular arrhythmias in patients with prolonged QT intervals.2 The present study was designedto investigate this hypothesisand to ascertain whether treatment of such patients affects the degree of QT dispersionin patients whosearrhythmias are suppressedand in those in whom they are not. METHODS
Nine patients (6 female, age range 1 month to 43 years), 6 with Roman0 Ward syndrome,and 3 with Jervell Lange-Neilsen syndrome, were studied (Table I). All had QT-interval prolongation on standard 1Zlead electrocardiogram and had documented torsades de pointes. The frequency and severity of attacks varied among patients, with 4 (nos. 1, 2, 3 and 7) having frequent syncopal episodes (> 1/month) off treatment. Two patients were membersof the samefamily (nos. 8 and 4 were grandmother and grandson). In addition, 9 age-matchedcontrol subjectswith no evidenceof heart diseasewere studied. All patients had standard 1Zlead electrocardiograms recorded at 25 rnms-’ off antiarrhythmic agents.All electrocardiogramswere examined retrospectively. No patient had bundle branch block or evidence of any other intraventricular conduction abnormality on the electrocardiogram. There was no significant difference in QRS duration in any patient throughout the study. Electrocardiogramswere enlarged by a factor of 4. QT intervals were measured from the onset of the QRS complex to the end of the T wave (defined as the point of return of the T wave to the isoelectric baseline).QT and RR intervals, and QTc value (using the formula developedby Taran and Szilagyi3 from Bazett’s4 original work, and the preceding RR interval) were calculated for each complex. As many intervals as possiblewere measuredin each individual lead, and mean QT and RR intervals, and QTc value were calculated for each lead. If the end of the T MARCH 1. 1992
wave could not be reliably determined, the lead was not included in the analysis. Mean maximum QT interval and QTc value from any individual lead were also calculated, becausethis is the standard assessmentof the QT interval and QTc value. QT-interval and QTc dispersions were defined as the difference between the minimum and maximum QT intervals and QTc values in any of the 12 leads. All patients were treated with /3-blocking agents (atenolol or propranolol) and had repeat 1Zlead electrocardiograms analyzed. In addition, the 3 patients with Jervell Lange-Neilsen syndrome had electrocardiograms recorded after left stellate ganglionectomy. The patients have been followed up for a mean of 6.5 years (range 8 months to 15 years), during which time 6 patients have remained free of further symptoms. The 3 patients with Jervell Lange-Neilsen syndrome continued to have syncopal arrhythmias, despite left stellate ganglionectomy and treatment with /3 blockers. Two of these patients subsequentlydied suddenly, and the third continues to have symptoms. The Mann-Whitney U and Wilcoxon signed rank tests were used for statistical analysis. RESULTS Comparison between patient group and control subjectw In the patient group, all 12 leads of the elec-
trocardiogram were included in 4 patients, whereas leads Vs and V5 were excluded in patient 1, lead III in patient 3, leads Vs and V=,in patient 4, lead Vi in patient 5, and leads III and aVL in patient 7. In the control group, all 12 leads were included in 6 subjects, whereas lead aVL was excluded in 2, and lead Vs in 1. Study patients had a significantly longer mean QT interval (averaged over all 12 leads) than that of the control group (450 f 100 vs 359 f 63 ms; p = 0.015), and a longer mean QTc value (0.53 f 0.08 vs 0.41 f 0.02 s1i2; p = 0.004), maximum QT interval (504 f 114 vs 376 f 63 ms; p = 0.012) and QTc value (0.59 f 0.08 vs 0.43 f 0.02 s112;p = 0.002), but a similar mean RR interval (736 f 231 vs 783 f 289 ms) (Table II). The degreeof QT dispersionin the patient group was significantly longer than that in the control subjects (109 f 47 vs 43 f 12 ms; p = 0.004), as was QTc dispersion (0.12 f 0.04 vs 0.05 f 0.02 s1i2;p = 0.002) (Figure 1). Comparison
before and during P-blocking
therapy:
On treatment with ,B-blockingagents, all 12 leads were analyzed in 8 patients, with lead III being excluded in patient 3. Mean QT interval was 498 f 98 ms, mean TABLE II QT and RR Intervals
(ms), QTc Values
(s1j2),
TABLE I Patient Patient
1 2 3 4 5 6 7 8
Characteristics
Sex
Age
F
1 month 1 month 3 years
M M
M F F F F F
9
Diagnosis JLN JLN JLN RW RW RW RW RW RW
9 years 15 years 19 years 32 years 36 years 43 years
Symptoms Syncope Syncope Syncope Syncope Syncope Syncope Syncope Syncope Syncope
x4 x4 x 15 x 1 x 1 x2 > 100 x2 > 20
JLN = Jervell Lange-Neilsensyndrome; RW = Romano Ward syndrome.
-
RR interval was 896 f 233 ms, and mean QTc value was 0.53 f 0.09 s1i2.Mean maximum QT interval was 557 f 112 ms, mean maximum RR interval was 941 f 247 ms, and mean maximum QTc value was 0.59 f 0.09 s1i2.There was no significant difference in mean QT and RR intervals, QTc value, maximum QT and RR intervals, and maximum QTc value before and during treatment with /3-blocking agents.There was no significant difference in the degreeof QT dispersion (133 f 71 ms) compared with that at baseline, although the tendency was for an increasein dispersion(Figure 2). In addition, there was no difference in QTc dispersion (0.13 f 0.07 s1j2) compared with that at baseline, although again there was a tendency for an increase. Comparison between symptom groups: Comparing the 4 patients who presented with frequent symptoms (nos. 1, 2, 3 and 7) with the 5 without frequent symptoms, there was no difference between the groups in terms of mean QT interval, QTc value, maximum QT interval, maximum QTc value and RR interval (Table III). In addition, there was no difference in QT or QTc dispersionoff treatment between patients with frequent symptoms (94 f 24 ms and 0.1 f 0.02 s~/~,respectively) and those with infrequent symptoms (121 f 59 ms and 0.13 f 0.05 s~/~,respectively). On treatment with ,&blocking agents, 6 patients had their symptoms controlled and did not experience any further syncope. These patients all had Roman0 Ward syndrome. Comparing these patients with the 3 with Jervell LangeNeilsen syndrome, the mean QTc value of the patients with no symptoms was shorter than that of those who continued to experiencesymptoms (0.48 f 0.06 vs 0.63 f 0.06 ~‘1~;p = 0.039). There was no significant differencein mean QT, RR and maximum QT intervals, and maximum QTc value. There was no difference in QT
and QT and QTc Dispersions
(ms and s.112, respectively)
for All Patients
I
Study Group
Mean QT interval Mean maximum QT interval Mean RR interval Mean maximum RR interval Mean QTc value Mean maximum QTc value QT-interval dispersion QTc-value dispersion
Control Group
No Treatment
B-Blocked
359?63(256-413) 376%63(273-436) 783 ?289(391-1,152) 806&299(405-1,227) 0.41 f 0.02 (0.38-0.45) 0.43 f 0.02 (0.4-0.47) 43 k 12 (27-63) 0.05 k 0.02 (0.03-0.07)
45Ok lOO(314-626) 504 k 114(345-679) 736 k 231 (1,053) 783 f 244 (560-1,114) 0.53 f 0.08 (0.38-0.62) 0.59 z 0.08 (0.43-0.67) 109 r47(35-182) 0.12 k 0.04 (0.05-0.19)
498 e 98(357-677) 557 &112(390-740) 893 +233(593-1,188) 9412 247(615-1,230) 0.53 2 0.09 (0.42-0.7) 0.59 2 0.09 (0.46-0.74) 133 + 71(50-265) 0.13 f 0.07 (0.48-0.3)
QT DISPERSION
IN LONG QT SYNDROMES
-
635
TABLE III Mean QT and RR Intervals (ms), QTc Values W2), and QT and QTc Dispersions (ms and s112, respectively) for Patients with Frequent and Infrequent Symptoms at Baseline, During Treatment with Beta-Blocking Agents and After Left Stellate Ganglionectomv Infrequent Symptoms
Frequent Symptoms
Baseline Mean QT interval Mean RR interval Mean QTc value QT dispersion QTc dispersion
488 -c 114 (332-626) 838 f 269 (535-1053) 0.54 r 0.07 (0.45-0.623) 121 f 59 (35-182) 0.13 f 0.05 (0.05-0.19)
p-blocked Mean QT interval Mean RR interval Mean QTc value QT dispersion QTc dispersion
(n = 6)
(n = 3)
477 2 90 (357-590) 977 + 223 (688-l 188) 0.48 zi 0.06 (0.42-0.6) 153 2 78 (50-265) 0.16 f 0.08 (0.05-0.29)
540 f 119 (463-677) 734 f 184 (593-943) 0.63 f 0.06 (0.5947) 92 +: 32 (55-115) 0.09 f 0.03 (0.59-0.11)
(n = 5)
(n = 4)
402 ir 609 f 0.52 f 94 -t 0.1 f
Left stellate ganglionectomy Mean QT interval Mean RR interval Mean QTc value QT dispersion QTc dispersion
QT -
200
DISPERSION
62 (314-461) 81 (534-680) 0.09 (0.38-0.57) 24 (60-l 15) 0.02 (0.08-0.13)
(n = 3)
536 k 870 f 0.61 f 72 f 0.06 it
(ms)
QTc
DISPERSION
120 (468-698) 171(685-1023) 0.1 (0.49-0.69) 21 (50-91) 0.01 (0.05-0.07)
(s’) - 0.2
X X X X # #
loo-
$2
- 0.1
n :: X i
dii-
X
$ X
z
=I-
0
0
QTc CONTROL GROUP
OT
QT
QT -
300
FIGURE 1. QT-interval and QTc-value pershsin9patientswithcongenitallollg QT~(x)comparedwiththoseof control group (+A
DISPERSION
OTC STUDY GROUP
OTC
(ms)
DISPERSION
(s”) - 0.3
+ +
200
+
-
+ $
+
+
loo-
-37
+ +
-jt Av
- 0.2 +
-ii+
T X
# X
+
+
zp
T
- 0.1
X x
x X
x
QT
OTC
0
0 QT
OTC TH~&PY
636
-
QT QTc BETA-BLOCKADE
LEFT STELLATE GANGLIONECTOMY
THE AMERICAN JOURNAL OF CARDIOLOGY VOLUME 69
MARCH 1, 1992
FIGURE 2. QT-intewal and QTc-value bspershsin9patientswiBtcongenRalloq QTsyndromesandfreqwnt~x)orlnfreqwnt (+) symptoms before and during hatmentwRh@bbckingagents,andin3 after iefl stenate gangBonectc#ny.
dispersion betweenthose with frequent symptoms(92 f 32 ms) and those without symptoms (153 f 78 ms), nor in QTc dispersion (0.09 f 0.03 vs 0.16 f 0.08 s1i2). In patients with continuing symptoms, there was a significant increase in mean QT interval on ~-blocking agents as compared with that at baseline (540 f 119 vs 402 f 62 ms; p = 0.034), and in mean QTc value (0.63 f 0.06 vs 0.52 f 0.09 s1i2; p = 0.034), but no difference in mean maximum QT interval and QTc value, and mean RR interval. Three of the 4 patients with severesymptoms had QTc dispersion >O.lOO s1i2,compared with 3 of the 5 with mild symptoms, a value consideredto be of use in separating those at risk of ventricular arrhythmias.2 Comparison
after
left
stellate
ganglionectomy:
Three patients with Jervell Lange-Neilsen syndrome also had repeat electrocardiograms after left stellate ganglionectomy. Furthermore, all 3 patients were continuing to receive P-blocking agents. In 2 of the 3 patients, all 12 leads were used in the analysis, and in patient 1, leads Vr and V5 were excluded. Comparing their electrocardiogramswith those obtained before left stellate ganglionectomy, there was no significant difference in mean and mean maximum QT intervals and QTc values, and RR interval (Table III). QT and QTc dispersions (Figure 2) were also unchanged (72 f 21 and 0.06 f 0.01 s1i2,respectively). DISCUSSION
The mechanism of QT-interval prolongation, and the development of torsadesde pointes remain matters of some debate. Currently, there are 2 theories for the explanation of QT prolongation. The first theory is that there is dispersion of repolarization within the ventricular myocardium.5 In patients with congenital QT-interval prolongation there is an increasein the dispersionof endocardial monophasic action potential durations from the normal 20 to 406,7to >2501 ms. Experimentally produced dispersion of repolarization has been shown to give rise to ventricular arrhythmias morphologically similar to torsadesde pointes8 The secondtheory is that the development of the arrhythmias is due to early afterdepolarizations. These phenomena, which can be produced in vitro, can also generateventricular arrhythmias morphologically similar to torsades de pointes.‘O They may also be produced in circumstances that reflect the occurrence of the arrhythmia in vivo (i.e., in situations of increased sympathetic nervous system activity).” In the clinical setting, evidence has been presented for both theories. Whichever is correct, there is no doubt that there are regional differencesin the duration of myocardial repolarization as measuredby endocardial monophasic action potentials. It can potentially be extrapolated from this to suggestthat theseregional differences in repolarization may be reflected in regional differences in the surface QT interval. In normal subjects there is somedispersion of QT intervals (
It has been suggested that QT dispersion recorded on the surface electrocardiogram may be a predictor of arrhythmic events in patients with congenital QT prolongation. To evaluate this, 9 patients (6 female, mean age 17.6 years) with congenital long QT syndromes, all of whom had syncope and documented torsades de pointes, were studied. Patients were studied off treatment and during therapy with @blocking agents. Three patients were also studied after left stellate ganglionectomy. An age-matched control group was also studied. Good quality 12-lead electrocardiograms were recorded from all patients. For each lead, QT and RR intervals were measured, and QTc value was calculated. QT and QTc dispersions were calculated for each patient. Patients had a significantly longer mean QT interval compared with that of the control group (450 f 100 vs 359 f 63 ms; p = 0.015) at similar mean RR intervals (736 f 231 vs 783 f 289 ms), with a longer mean QTc value (0.53 f 0.08 vs 0.41 f 0.02 slm; p = 0.004). Patients also had longer QT and QTc dispersions compared with those of the control group (110 f 45 vs 43 f 12 ms[p=O.O04],and O.lOSi0.03~~ 0.05 f 0.02 SIB [p = O&02], respectively). QT and QTc dispersions on and off p-blocking agents were not significantly different. Comparing patients with frequent and those with infrequent symptoms, there was no difference in QT or QTc dispersion either off treatment or during therapy with &blocking agents. There was an increase in mean QT interval and QTc value in patients with frequent symptoms on &blocking agents. On @-blocking agents, patients with frequent symptoms had a significantly longer mean QTc value compared with that of those with infrequent symptoms. There was no change in QT or QTc dispersion, mean QT or RR interval, or mean QTc value afler left stellate ganglionectomy. QT-interval dispersion is observed in patients with congenital long QT syndromes; however, the degree of dispersion is not related to the severity of symptoms, nor influenced by treatment with @ blocking agents. (Am J Cardiol 1992;69:634-638) From St. George’s Hospital and Medical School, Royal Devon and Exeter Hospital, and the Royal Brompton National Heart and Lung Hospitals, London, England. Manuscript received August 27, 1991; revised manuscript received and accepted October 28, 1991. Address for reprints: Nicholas J. Linker, MB, ChB, Department of Cardiological Sciences, St. George’s Hospital and Medical School, Cranmer Terrace, London SW 17 ORE, England.
634
THE AMERICAN JOURNAL OF CARDIOLOGY VOLUME 69
P
atients with congenital QT-interval prolongation have a high incidence of ventricular arrhythmias and suddendeath. Prediction of those patients at risk of these arrhythmias, and the evaluation of asymptomatic family memberswith QT-interval prolongation is difficult. One explanation for the developmentof ventricular arrhythmias in such patients is basedon dispersion of repolarization between different regions of the myocardium. This feature has been observedin patients with congenital QT-interval prolongation by measuring endocardial monophasicaction potentials from multiple siteswithin the ventricles and demonstrating an increase in the dispersion of the duration of the monophasic action potentials.’ Thus, the hypothesishas been advanced that the degree of dispersion between the QT intervals from a standard 1Zlead electrocardiogram may be an indicator of the risk of ventricular arrhythmias in patients with prolonged QT intervals.2 The present study was designedto investigate this hypothesisand to ascertain whether treatment of such patients affects the degree of QT dispersionin patients whosearrhythmias are suppressedand in those in whom they are not. METHODS
Nine patients (6 female, age range 1 month to 43 years), 6 with Roman0 Ward syndrome,and 3 with Jervell Lange-Neilsen syndrome, were studied (Table I). All had QT-interval prolongation on standard 1Zlead electrocardiogram and had documented torsades de pointes. The frequency and severity of attacks varied among patients, with 4 (nos. 1, 2, 3 and 7) having frequent syncopal episodes (> 1/month) off treatment. Two patients were membersof the samefamily (nos. 8 and 4 were grandmother and grandson). In addition, 9 age-matchedcontrol subjectswith no evidenceof heart diseasewere studied. All patients had standard 1Zlead electrocardiograms recorded at 25 rnms-’ off antiarrhythmic agents.All electrocardiogramswere examined retrospectively. No patient had bundle branch block or evidence of any other intraventricular conduction abnormality on the electrocardiogram. There was no significant difference in QRS duration in any patient throughout the study. Electrocardiogramswere enlarged by a factor of 4. QT intervals were measured from the onset of the QRS complex to the end of the T wave (defined as the point of return of the T wave to the isoelectric baseline).QT and RR intervals, and QTc value (using the formula developedby Taran and Szilagyi3 from Bazett’s4 original work, and the preceding RR interval) were calculated for each complex. As many intervals as possiblewere measuredin each individual lead, and mean QT and RR intervals, and QTc value were calculated for each lead. If the end of the T MARCH 1. 1992
wave could not be reliably determined, the lead was not included in the analysis. Mean maximum QT interval and QTc value from any individual lead were also calculated, becausethis is the standard assessmentof the QT interval and QTc value. QT-interval and QTc dispersions were defined as the difference between the minimum and maximum QT intervals and QTc values in any of the 12 leads. All patients were treated with /3-blocking agents (atenolol or propranolol) and had repeat 1Zlead electrocardiograms analyzed. In addition, the 3 patients with Jervell Lange-Neilsen syndrome had electrocardiograms recorded after left stellate ganglionectomy. The patients have been followed up for a mean of 6.5 years (range 8 months to 15 years), during which time 6 patients have remained free of further symptoms. The 3 patients with Jervell Lange-Neilsen syndrome continued to have syncopal arrhythmias, despite left stellate ganglionectomy and treatment with /3 blockers. Two of these patients subsequentlydied suddenly, and the third continues to have symptoms. The Mann-Whitney U and Wilcoxon signed rank tests were used for statistical analysis. RESULTS Comparison between patient group and control subjectw In the patient group, all 12 leads of the elec-
trocardiogram were included in 4 patients, whereas leads Vs and V5 were excluded in patient 1, lead III in patient 3, leads Vs and V=,in patient 4, lead Vi in patient 5, and leads III and aVL in patient 7. In the control group, all 12 leads were included in 6 subjects, whereas lead aVL was excluded in 2, and lead Vs in 1. Study patients had a significantly longer mean QT interval (averaged over all 12 leads) than that of the control group (450 f 100 vs 359 f 63 ms; p = 0.015), and a longer mean QTc value (0.53 f 0.08 vs 0.41 f 0.02 s1i2; p = 0.004), maximum QT interval (504 f 114 vs 376 f 63 ms; p = 0.012) and QTc value (0.59 f 0.08 vs 0.43 f 0.02 s112;p = 0.002), but a similar mean RR interval (736 f 231 vs 783 f 289 ms) (Table II). The degreeof QT dispersionin the patient group was significantly longer than that in the control subjects (109 f 47 vs 43 f 12 ms; p = 0.004), as was QTc dispersion (0.12 f 0.04 vs 0.05 f 0.02 s1i2;p = 0.002) (Figure 1). Comparison
before and during P-blocking
therapy:
On treatment with ,B-blockingagents, all 12 leads were analyzed in 8 patients, with lead III being excluded in patient 3. Mean QT interval was 498 f 98 ms, mean TABLE II QT and RR Intervals
(ms), QTc Values
(s1j2),
TABLE I Patient Patient
1 2 3 4 5 6 7 8
Characteristics
Sex
Age
F
1 month 1 month 3 years
M M
M F F F F F
9
Diagnosis JLN JLN JLN RW RW RW RW RW RW
9 years 15 years 19 years 32 years 36 years 43 years
Symptoms Syncope Syncope Syncope Syncope Syncope Syncope Syncope Syncope Syncope
x4 x4 x 15 x 1 x 1 x2 > 100 x2 > 20
JLN = Jervell Lange-Neilsensyndrome; RW = Romano Ward syndrome.
-
RR interval was 896 f 233 ms, and mean QTc value was 0.53 f 0.09 s1i2.Mean maximum QT interval was 557 f 112 ms, mean maximum RR interval was 941 f 247 ms, and mean maximum QTc value was 0.59 f 0.09 s1i2.There was no significant difference in mean QT and RR intervals, QTc value, maximum QT and RR intervals, and maximum QTc value before and during treatment with /3-blocking agents.There was no significant difference in the degreeof QT dispersion (133 f 71 ms) compared with that at baseline, although the tendency was for an increasein dispersion(Figure 2). In addition, there was no difference in QTc dispersion (0.13 f 0.07 s1j2) compared with that at baseline, although again there was a tendency for an increase. Comparison between symptom groups: Comparing the 4 patients who presented with frequent symptoms (nos. 1, 2, 3 and 7) with the 5 without frequent symptoms, there was no difference between the groups in terms of mean QT interval, QTc value, maximum QT interval, maximum QTc value and RR interval (Table III). In addition, there was no difference in QT or QTc dispersionoff treatment between patients with frequent symptoms (94 f 24 ms and 0.1 f 0.02 s~/~,respectively) and those with infrequent symptoms (121 f 59 ms and 0.13 f 0.05 s~/~,respectively). On treatment with ,&blocking agents, 6 patients had their symptoms controlled and did not experience any further syncope. These patients all had Roman0 Ward syndrome. Comparing these patients with the 3 with Jervell LangeNeilsen syndrome, the mean QTc value of the patients with no symptoms was shorter than that of those who continued to experiencesymptoms (0.48 f 0.06 vs 0.63 f 0.06 ~‘1~;p = 0.039). There was no significant differencein mean QT, RR and maximum QT intervals, and maximum QTc value. There was no difference in QT
and QT and QTc Dispersions
(ms and s.112, respectively)
for All Patients
I
Study Group
Mean QT interval Mean maximum QT interval Mean RR interval Mean maximum RR interval Mean QTc value Mean maximum QTc value QT-interval dispersion QTc-value dispersion
Control Group
No Treatment
B-Blocked
359?63(256-413) 376%63(273-436) 783 ?289(391-1,152) 806&299(405-1,227) 0.41 f 0.02 (0.38-0.45) 0.43 f 0.02 (0.4-0.47) 43 k 12 (27-63) 0.05 k 0.02 (0.03-0.07)
45Ok lOO(314-626) 504 k 114(345-679) 736 k 231 (1,053) 783 f 244 (560-1,114) 0.53 f 0.08 (0.38-0.62) 0.59 z 0.08 (0.43-0.67) 109 r47(35-182) 0.12 k 0.04 (0.05-0.19)
498 e 98(357-677) 557 &112(390-740) 893 +233(593-1,188) 9412 247(615-1,230) 0.53 2 0.09 (0.42-0.7) 0.59 2 0.09 (0.46-0.74) 133 + 71(50-265) 0.13 f 0.07 (0.48-0.3)
QT DISPERSION
IN LONG QT SYNDROMES
-
635
TABLE III Mean QT and RR Intervals (ms), QTc Values W2), and QT and QTc Dispersions (ms and s112, respectively) for Patients with Frequent and Infrequent Symptoms at Baseline, During Treatment with Beta-Blocking Agents and After Left Stellate Ganglionectomv Infrequent Symptoms
Frequent Symptoms
Baseline Mean QT interval Mean RR interval Mean QTc value QT dispersion QTc dispersion
488 -c 114 (332-626) 838 f 269 (535-1053) 0.54 r 0.07 (0.45-0.623) 121 f 59 (35-182) 0.13 f 0.05 (0.05-0.19)
p-blocked Mean QT interval Mean RR interval Mean QTc value QT dispersion QTc dispersion
(n = 6)
(n = 3)
477 2 90 (357-590) 977 + 223 (688-l 188) 0.48 zi 0.06 (0.42-0.6) 153 2 78 (50-265) 0.16 f 0.08 (0.05-0.29)
540 f 119 (463-677) 734 f 184 (593-943) 0.63 f 0.06 (0.5947) 92 +: 32 (55-115) 0.09 f 0.03 (0.59-0.11)
(n = 5)
(n = 4)
402 ir 609 f 0.52 f 94 -t 0.1 f
Left stellate ganglionectomy Mean QT interval Mean RR interval Mean QTc value QT dispersion QTc dispersion
QT -
200
DISPERSION
62 (314-461) 81 (534-680) 0.09 (0.38-0.57) 24 (60-l 15) 0.02 (0.08-0.13)
(n = 3)
536 k 870 f 0.61 f 72 f 0.06 it
(ms)
QTc
DISPERSION
120 (468-698) 171(685-1023) 0.1 (0.49-0.69) 21 (50-91) 0.01 (0.05-0.07)
(s’) - 0.2
X X X X # #
loo-
$2
- 0.1
n :: X i
dii-
X
$ X
z
=I-
0
0
QTc CONTROL GROUP
OT
QT
QT -
300
FIGURE 1. QT-interval and QTc-value pershsin9patientswithcongenitallollg QT~(x)comparedwiththoseof control group (+A
DISPERSION
OTC STUDY GROUP
OTC
(ms)
DISPERSION
(s”) - 0.3
+ +
200
+
-
+ $
+
+
loo-
-37
+ +
-jt Av
- 0.2 +
-ii+
T X
# X
+
+
zp
T
- 0.1
X x
x X
x
QT
OTC
0
0 QT
OTC TH~&PY
636
-
QT QTc BETA-BLOCKADE
LEFT STELLATE GANGLIONECTOMY
THE AMERICAN JOURNAL OF CARDIOLOGY VOLUME 69
MARCH 1, 1992
FIGURE 2. QT-intewal and QTc-value bspershsin9patientswiBtcongenRalloq QTsyndromesandfreqwnt~x)orlnfreqwnt (+) symptoms before and during hatmentwRh@bbckingagents,andin3 after iefl stenate gangBonectc#ny.
dispersion betweenthose with frequent symptoms(92 f 32 ms) and those without symptoms (153 f 78 ms), nor in QTc dispersion (0.09 f 0.03 vs 0.16 f 0.08 s1i2). In patients with continuing symptoms, there was a significant increase in mean QT interval on ~-blocking agents as compared with that at baseline (540 f 119 vs 402 f 62 ms; p = 0.034), and in mean QTc value (0.63 f 0.06 vs 0.52 f 0.09 s1i2; p = 0.034), but no difference in mean maximum QT interval and QTc value, and mean RR interval. Three of the 4 patients with severesymptoms had QTc dispersion >O.lOO s1i2,compared with 3 of the 5 with mild symptoms, a value consideredto be of use in separating those at risk of ventricular arrhythmias.2 Comparison
after
left
stellate
ganglionectomy:
Three patients with Jervell Lange-Neilsen syndrome also had repeat electrocardiograms after left stellate ganglionectomy. Furthermore, all 3 patients were continuing to receive P-blocking agents. In 2 of the 3 patients, all 12 leads were used in the analysis, and in patient 1, leads Vr and V5 were excluded. Comparing their electrocardiogramswith those obtained before left stellate ganglionectomy, there was no significant difference in mean and mean maximum QT intervals and QTc values, and RR interval (Table III). QT and QTc dispersions (Figure 2) were also unchanged (72 f 21 and 0.06 f 0.01 s1i2,respectively). DISCUSSION
The mechanism of QT-interval prolongation, and the development of torsadesde pointes remain matters of some debate. Currently, there are 2 theories for the explanation of QT prolongation. The first theory is that there is dispersion of repolarization within the ventricular myocardium.5 In patients with congenital QT-interval prolongation there is an increasein the dispersionof endocardial monophasic action potential durations from the normal 20 to 406,7to >2501 ms. Experimentally produced dispersion of repolarization has been shown to give rise to ventricular arrhythmias morphologically similar to torsadesde pointes8 The secondtheory is that the development of the arrhythmias is due to early afterdepolarizations. These phenomena, which can be produced in vitro, can also generateventricular arrhythmias morphologically similar to torsades de pointes.‘O They may also be produced in circumstances that reflect the occurrence of the arrhythmia in vivo (i.e., in situations of increased sympathetic nervous system activity).” In the clinical setting, evidence has been presented for both theories. Whichever is correct, there is no doubt that there are regional differencesin the duration of myocardial repolarization as measuredby endocardial monophasic action potentials. It can potentially be extrapolated from this to suggestthat theseregional differences in repolarization may be reflected in regional differences in the surface QT interval. In normal subjects there is somedispersion of QT intervals (
