j dynamics ventricular repolarisation in congenitallong ...2 m 15 mnth jln syncope > 300 + 3 m...

Dynamics of ventricular repolarisation in thecongenital long QT syndromes

Nicholas J Linker, A John Camm, David E Ward

Br Heart J 1991;66:230-7

AbstractPatients with congenital QT intervalprolongation are at risk of ventriculararrhythmias and sudden death. It hasbeen suggested that the susceptibility toarrhythmias in these syndromes may berelated to the abnormal dynamics ofventricular repolarisation. The dyn-amics of ventricular repolarisation,including assessment of the effect ofchanging heart rate on the QT intervaland the duration of the right ventricularmonophasic action potential, werestudied in eight patients with congenitallong QT syndromes. The effects ofalteredsympathetic tone on these dynamicswere investigated with isoprenaline,propranolol, and left stellate ganglionblock. The rate adaptation of the QTinterval was abnormal in only a fewpatients and in some patients thisfeature may be related to the severity ofthe condition. These abnormalities maybe exaggerated by isoprenaline and les-sened by propranolol and left stellateganglion block. Monophasic actionpotential dynamics were normal in allpatients.The hypothesis that impaired QT rate

adaptation may play a role in the genesisof ventricular arrhythmias in these syn-dromes is not, in general, supported bythe present data. However, in patientswith impaired adaptation the normalisa-tion of QT dynamics after fi blockadeand left stellate ganglion block was cons-istent with the efficacy of these forms oftreatment.

The QT interval of the electrocardiogramreflects the total duration of electrical activityin the ventricles. Two syndromes areassociated with a primary hereditary prolon-gation of the QT interval. In both the Jervelland Lange-Nielsen and the Romano Wardsyndromes there is usually a family history of

Table 1 Patient characteristics

DocumentedPatient Sex Age Diagnosis Symptoms torsade de pointes1 F 17yr RW Syncope x 3 +2 M 15 mnth JLN Syncope > 300 +3 M 3yr JLN Syncope x 15 +4 F 7yr JLN Syncope x 4 +5 M 18mnth JLN None6 F 19yr RW Syncope x I +7 F 20yr RW Syncope x 3 +8 F 43yr RW Syncope > 20 +

RW, Romano Ward syndrome; JLN, Jervell Lange-Nielsen syndrome.

syncopal episodes caused by ventriculartachycardia, the most common of which istorsade de pointes, and a family history ofsudden cardiac death.'`3 It has been suggestedthat the susceptibility to arrhythmias in thesesyndromes may be related to an absence oralteration of the normal QT shortening inresponse to an increase in heart rate.4 Westudied the dynamics of ventricular repolar-isation, as measured by the QT interval andventricular monophasic action potential dura-tion, in patients with congenital QT intervalprolongation and assessed the effects on thesedynamics of alterations in sympathetic tone.

Patients and methodsWe studied eight patients (three male, fivefemale), mean age 14 years (range 15 monthsto 43 years) (table 1). Four patients (numbers2 to 5) had sensorineural deafness consistentwith the Jervell and Lange-Nielsen syndrome;patients 2 and 3 came from the same family asdid patients 4 and 5. Patient 1 had two sisterswho died suddenly aged 16 and 20 years, bothwith a history of syncope. Patients 6, 7, and 8had no family history of syncope or suddendeath. Patient 5 (brother of patient 4) wasfound to have a long QT interval on routineelectrocardiographic screening. Two patients(6 and 8) had previously been diagnosed ashaving either hysterical or epileptiformcollapses on several occasions. Seven patientshad documented torsade de pointes at thetime of syncope.No patient was taking any antiarrhythmic

medication (including ,B blocking agents) atthe time of the study. Table 2 gives QRSdurations, QT intervals, QTc values, andcycle lengths in sinus rhythm. Four elec-trocardiographic leads (I, aVF, Vi, and V6)were recorded at paper speeds of 50 or100 mm/s. This configuration was preset bythe recorder and remained constant through-out the study. The QT interval was measuredfrom lead VI, because this provides a closeapproximation to the maximum QT intervalin any lead.5 During sinus rhythm and atrialpacing the QT interval was measured fromthe start of the QRS complex to the end of theT wave. During ventricular pacing the QTinterval was measured from the ventricularpacing artefact or the beginning of the pacedQRS complex (if there was stimulus latency)to the end of the T wave. The end of the Twave was defined as the point of return of theT wave to baseline.6 There is no accepteddefinition of U waves in the long QT syn-dromes. In the presence of bizarre and often

230

Department ofCardiologicalSciences, St George'sHospital MedicalSchool, LondonN J LinkerA J CammD E WardCorrespondence toDr Nicholas J Linker,Department of CardiologicalSciences, St George'sHospital Medical School,Cranmer Terrace, LondonSW17 ORE.Accepted for publication26 March 1991

on April 20, 2021 by guest. P

rotected by copyright.http://heart.bm

j.com/

Br H

eart J: first published as 10.1136/hrt.66.3.230 on 1 Septem

ber 1991. Dow

nloaded from

http://heart.bmj.com/

Dynamics of ventricular repolarisation in the congenital long QT syndromes

Table 2 QRS duration, QT interval, and QTc values in all patients before and after propranolol

Baseline (sinus rhythm) Propranolol (sinus rhythm)QRS duration(sinus rhythm) QT Cycle length QTc QT Cycle length QTc

Patient (ms) (ms) (ms) (s') (ms) (ms) (s")

1 90 500 930 0-520 480 980 0 4852 80 645 600 0-835 380 600 0-4903 90 440 680 0-535 435 680 0-5304 80 580 680 0-705 520 810 0-5805 85 360 560 0-480 360 560 0-4806 90 500 950 0-515 520 950 0-5357 90 360 480 0-520 390 650 0-4858 110 470 720 0-555 505 810 0-560

changing T wave configuration it is not possi-ble clearly to distinguish between the T and Uwaves. Thus all our measurements were madeto the end of total repolarisation as definedabove. During ventricular pacing the T wavewas always inscribed as a simple curve with noU waves or multiphasic components.The QTc values were calculated from the

formula adapted by Taran and Szilagyi' froma formula devised by Bazett.8 Dimensionalanalysis of the Taran and Szilagyi formulashows the units of QTc to be s'2.9The procedure was performed after the

patient or a parent had given informed con-sent to an electrophysiology study. In allpatients a 6 French gauge bipolar pacing elec-trode was inserted via the right femoral veinand positioned at the right ventricular apex topermit ventricular pacing. In five patients (1and 4 to 7), a 7 French gauge silver/silverchloride monophasic action potential contactelectrode was introduced via the right femoralvein and positioned at the right ventricularapex. In one patient (case 1) a monophasicaction potential electrode was also positionedat the left ventricular apex via the rightfemoral artery. In one patient (case 6) a mono-phasic action potential electrode and a pacingelectrode were positioned in the right atriumto allow atrial measurements to be made. Thesignals from the monophasic action potentialelectrodes were processed by a custom builtAC amplifier with a band width of 0-025 to500 Hz (-3 dB) and recorded on a Siemens-Elema Mingograf 82 chart recorder at a speedof 50 or 100 mm/s together with four surfaceleads (leads I, aVF, VI, and V6). Pacing wasperformed with a Digitimer 4279 programma-ble stimulator. All measurements were madeby two observers. The duration of the mono-phasic action potential was measured at 90%repolarisation (MAP90) because it is difficult todetermine the exact point of full repolarisa-tion.The QT interval and MAP90 duration were

determined during sinus rhythm in allpatients. In all patients ventricular pacing wasperformed for 10 seconds from the right ven-tricular apex at a cycle length that was 50 msshorter than the intrinsic cycle length of thepatient. The QT interval and MAP90 durationof the last paced beat were measured (fig 1).After a 30 second pause, the pacing train wasrepeated with the cycle length 20 ms shorter;the pacing trains were repeated at shortercycle lengths until ventricular exit blockoccurred. All cycles with extrasystoles were

discarded and repeated. Three patients (4, 6,and 8) received an infusion of isoprenaline at4 ig/kg/min and in two of these (6 and 8) theprotocol was repeated. The protocol wasrepeated 10 minutes after the injection of0-2 mg/kg of propranolol in six patients and inone patient (case 8 who had asthma), after5 mg of intravenous atenolol. In two patients(3 and 4) the protocol was repeated after leftstellate ganglionectomy (including intraven-ous propranolol in patient 3) and in onepatient (case 2) the protocol was repeated aftera marcaine induced left stellate ganglionblock.

In patient 1 the recordings were made fromboth right and left ventricles. In patient 6 thebaseline protocol was performed with atrialpacing and recording atrial monophasic actionpotentials in addition to ventricular studies.In patients 6 and 8 the dispersion of rightventricular MAP90 durations was determinedby recording monophasic action potentialsfrom four and six right ventricular sites res-pectively.

It was not possible to perform everyprotocol in all patients because this wouldhave needed prolonged studies in very youngchildren. In one patient who received isopren-aline (case 4) spontaneous torsade de pointesoccurred repeatedly and the protocol was notcompleted. Only two patients underwent leftstellate ganglionectomy and one left stellateganglion block. The dispersion of monophasicaction potentials was not routinely measuredbecause this has been published previously.'0The measurement of atrial monophasic actionpotentials was introduced late in the study aspart of another continuing protocol.Where appropriate all results were com-

pared by a Student's paired t test.

ResultsQT INTERVAL IN SINUS RHYTHMAll patients had a normal baseline QRS dura-tion (mean (SD) 89 (9) ms) with a mean QTinterval of482 (99) ms at a mean cycle length of700 (166) ms (90 (20) beats/minute) and anincreased QTc (0-583 (0-012) s"2) (table 2).After ,B blockade with propranolol or atenololthere was no significant change in the mean QTinterval (449 (66) ms), QTc value (0-518(0 038) s~'% or cycle length (755 (157) ms, 90(20) beats/minute). Three patients were givenisoprenaline (cases 4, 6, and 8), after whichthere was a significant shortening of the QT

231



j.com/

Br H


ber 1991. Dow

nloaded from


Linker, Camm, Ward

'N~~~~~~~~~~~~~~~~~~~~~~TV

''S4.. .. ...t s. '° 6 ° A 4'''-~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~....

'' VF A. A.>Is A..i.,,

A- t~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~....

interval (330 (56) ms p < 005), though thiswas in the presence of a shorter cycle length(438 (91) ins, p = NS) and there was no sig-nificant change in the QTc value (0-498(0 032) sl'). In the two patients (cases 3 and 4)in whom left stellate ganglionectomy was per-

formed there was no difference between valuesbefore and after ganglionectomy in the QTinterval (440 and 580 ms v 460 and 580 ms),cycle length (680 and 560 ms v 560 and780 ms), or QTc value (0-535 and 0 705 s'" v

0-615 and.0665 s").

MONOPHASIC ACTION POTENTIALS IN SINUSRHYTHMIn the patients in whom monophasic actionpotentials were recorded (cases 1 and 4 to 8) thebaseline MAP90 duration was 363 (79) ms.Abnormalities in the configuration of mono-

phasic action potentials, in particular "humps"on phase 3 of the monophasic action potential,were seen in three patients (cases 4,7, and 8). Inone patient (case 4) a spontaneous episode oftorsades de pointe was recorded during theisoprenaline infusion, which showed that thefirst beat of the arrhythmia was not related tothe "hump" (fig 2). After propranolol there wasno significant change in MAP90 duration (353(52) ms). In the three patients who receivedisoprenaline (cases 4, 6, and 8), there was a

significant shortening in MAP90 duration (252

Figure 2 Spontaneoustorsade de pointes inpatient 4 during anisoprenaline infusion. Theinitiating beat of thearrhythmia occurs beforethe right ventricularmonophasic actionpotential, suggesting thatthe abnormalities in theconfiguration of themonophasic actionpotential are not relevantto the initiation of thetachycardia. For furtherdiscussion see text. MAP,right ventricularmonophasic actionpotential recording. Paperspeed 50 mm/s.

(51) ms, p < 0 05), though this was accompan-ied by a decrease in the sinus cycle length.

QT RATE ADAPTATIONThe response of the patients' QT intervals tothe ventricular pacing protocol was as follows.As the pacing rate was increased, there was a

tendency for the QT interval to shorten (fig 3),though there was considerable variation bet-ween patients. In particular, patients 1 and 2showed no adaptation of the QT interval to thefaster pacing rates. In patient 1 (fig 3) the QTinterval at a cycle length of 680 ms (540 ms)was shorter than at a pacing rate of 400 ims(580 ms). In patient 2 there appeared to be norate adaptation, with the QT interval behavingin a haphazard fashion. Although the otherpatients tended to shorten their QT intervalwith decreasing pacing rate, the amount ofdecrease in QT interval was variable. Forexample, patient 8 showed very little change(100 ms decrease in QT interval over a pacingrange of 600 to 260 ms) whereas patient 4showed a 190 ms decrease in QT interval overthe same pacing range.

In the two patients who received isoprena-line and had the protocol repeated, one (patient8) showed no change in QT interval duration ordynamics. The other (patient 6) had normalbaseline QT dynamics and a total loss of QTadaptation during the isoprenaline infusion

----------------..--- ..--------- ------------------ ----.---------------'

A7O

Figure I Recording ofmonophasic actionpotentials and QT intervalduring ventricular pacingin patient 4. Forexplanation ofmeasurement see text.RA1, RA2, right atrialelectrograms; R V, rightventricular electrogram;MAP, right ventricularmonophasic actionpotential recording. Paperspeed 100 mm/s.

1.

232



j.com/

Br H


ber 1991. Dow

nloaded from



Figure 3 Rateadaptation of the QTinterval in all eightpatients with congenitallong QT syndrome. Forexplanation see text.

650, Patient'

-a 600-E

I-0 550-

a 500-

10

0 0

00

.. 0

450;JI I I I I

48' ? 4~ 4? f

650, Patient 2

600-

E 550

0 Baseline ! 500-0 * Propranolol CD

a 400

350

1- 300-Co0 f~~~~~~~~~~~~~~~~~~~~~~~~~

0

00%000

a co0.0

0 0e ga

550-i Patient 3

0

0 00 0

a0 0coo0O2*.0

0 Baseline* Propranolol0 Post-ganglionectomy

650-

600-

E 550-

7 500-

c 450-C

a 400-

350

300

Patient 4 0

00 0

0 0 0.O *

0 00.0 0

o 0o0oodP'59

0 Baseline* Propranolol0 Post-ganglionectomy

I?p 4)R

400i Patient 5 00 0 0

L-350-cI-

9 50

a

2flfl

0000 00

a00

co 0

(V

550-1 Patient 6

-,; 500-E

0 Baseline 9 450

* Propranolol c

a 400-

350-

0 00

0 00 0

00 000 0

00(<Wom o

4501- Patient 7 550- Patient 8

000

000

00 00 0

0000

0

0 00

0

-A 500-E

O Baseline g 450-CDaPropranolol c._

a 400-

00

800* °80 000

*~000 0

0oaQ°

0 Baseline0 Isoprenaline* Atenolol

350-J

0I I

Pacing cycle length (ims)

with no change in QT interval over the 130 ms(55 beats/minute) change in pacing cyclelength. After the isoprenaline infusion thispatient was given an injection of propranololand her QT dynamics reverted to their normalbaseline state.

After ,B blockade there was no significantchange in the paced QT intervals overall,although five patients (cases 1 to 3 and 5 and 6)showed a general shortening of their QTintervals, while patient 7 showed a lengtheningofQT interval and patients 4 and 8 showed no

change (fig 3). There was the same trend for theQT interval to shorten with increasing pacingrate and in particular patients 1 and 2, who

Pacing cycle length (ms)

showed no rate adaptation in the baselinestudy, displayed appropriate QT intervalshortening.The two patients who underwent left stellate

ganglionectomy (cases 3 and 4) had normal QTdynamics during their baseline studies andthere was no change in either QT interval or

QT dynamics after the operation. The one

patient (case 2) who underwent left stellateblock after the propranolol injection showed noadditional change in QT dynamics or QTinterval. In all three patients torsade de pointesdegenerating to ventricular fibrillation was

induced when the left stellate ganglion was

manipulated surgically or with a needle.

O

0 Baseline* Propranolol0 Stellate block

500-

E450-

.c 400-

a350-

0

0 Baseline0 Isoprenaline0 Propranolol

E 400-

.

350

233



j.com/

Br H


ber 1991. Dow

nloaded from


Linker, Camm, Ward

Figure 4 MAPn rateadaptation in patients Iand 4 to 8. Forexplanation see text.

4001 Patient 1

Ec 350-.°

04)

0 300-0

0

0 00 0

000

0 RV MAP baseline0 RV MAP propranolol

450-

400-

350-

300-

250-

200-

Patient 4 0

0

00

0 00

0 esoof0

I I I -° I'

e 0 RV MAP baseline0 RV MAP propranolol

I II I I I

300,1 Patient 5

0 0

00

so"*00 0O@

.

0 RV MAP baseline* RV MAP propranolol

350-

300-

250-

200-

400i Patient 6

000 0

00 0

0 000000

0 00o Xa 000

oq:p

0 RV MAP baseline0 RV MAP isoprenaline0 RV MAP propranolol

250- Patient 70

000

00000o0 0

00a0

350-1 Patient 8

300-

0 RV MAP baseline0 RV MAP propranolol

250

200I I I Il


0

0

000000 00 0

.00

000

00

0 RV MAP baseline0 RV MAP isoprenaline


MONOPHASIC ACTION POTENTIAL RATEADAPTATIONVentricular monophasic action potentials were

recorded during the ventricular pacingprotocol in six patients (cases 1 and 4 to 8). Inall patients, including those with altered QTdynamics, there was a decrease in MAP9Oduration as the pacing rate was increased (fig 4).The range of decrease in MAP,J duration was

variable-for example 50 ms in patient 1 (whohad altered QT dynamics) and 210 ms inpatient 4 (who showed normal QT dynamics).In the five patients in whom the study was

repeated after propranolol (cases 1 and 4 to 7),there was no overall difference in MAP90 dura-tion: three patients (cases 4, 5, and 6) showedshortening of the MAP9O duration, one patient(case 1) showed a slight increase in MAP90duration, and one patient (case 7) showed no

change. After propranolol in all patients, themonophasic action potential dynamicsremained similar to the baseline study.Two patients had the protocol repeated

during an isoprenaline infusion. One (patient6) showed a shortening of the MAP90 durationwhich was associated with 30 ms adaptationwith increasing pacing rate (fig 4). This patienthad no QT adaptation after isoprenaline. Theother (patient 8) showed no change in MAP9o

duration or dynamics after isoprenaline (fig 4).One patient (case 1) had simultaneous rightapical and left apical ventricular monophasicaction potentials recorded (fig 5). Whereas theleft ventricular MAP90 duration was less thanthe right ventricular MAP90 duration, bothshowed similar rate adaptation before and afterpropranolol. One patient (case 6) had rightatrial monophasic action potentials recordedduring atrial pacing (fig 6). Normal adaptationof the atrial MAP90 duration to increasing atrialpacing rate'' was seen in this patient.

DISPERSION OF MONOPHASIC ACTION POTENTIALS

Dispersion of MAP90 duration was assessed inpatients 6 and 8 by measurement of the rightventricular MAP90 duration at four and six sitesrespectively during atrial pacing at a cyclelength of 500 ms. In patient 6 the amount ofdispersion was 50 ms (370 to 420 ms) and inpatient 8 it was 65 ms (315 to 380 ms). In bothpatients the longest MAP90 durations were

recorded from the right ventricular mid sep-tum; however, the shortest MAP90 durationswere recorded from the right ventricular inflowtract in patient 6 and from the free wall of theright ventricle in patient 8. These values fordispersion are excessive compared with normalvalues.'2 '3

Ec 250-0

° 200-

160-

Ec0

, 200-0

1501

234



j.com/

Br H


ber 1991. Dow

nloaded from



Figure 5 Simultaneousright and left ventricularmonophasic actionpotential recordings inpatient 1 before and afterpropranolol.

400-0

0 RV MAP baseline* LV MAP baseline

300-

00

0

00

*00

0 RV MAP propranolol* LV MAP propranolol


DiscussionThe dynamics of the QT intervaevaluated in normal subjects,patients with congenital long QTStudies in patients or subjects u1prolonged QT syndromes showedinterval is dependent on heart ratechanges in the catecholamine sensiheart.'415 Other studies of changeinterval with exercise showed anrelation between the QT interval axin normal subjects.6 17 Studies todynamics of the paced QT intervaland ventricular pacing8 19 showedmal subjects or in patients with corblock the QT interval shortened ving pacing rate. The relation betwefunctions was curvilinear: with a lixbetween heart rates of approxima120 beats/minute.202'

Six of our eight patients with co:prolongation showed an expccurvilinear pattern ofQT rate adal(patient 2) of the two patients whcrate adaptation had a particularly sfestation of the syndrome with ovcopal episodes. His resting QT iconsiderably prolonged (tableshowed no rate adaptation (fig 3).suggested that poor rate adaptatioiinterval in patients with congenitsyndromes may increase the suscventricular arrhythmias.4 The fairate adaptation was considered toreduction or lack of rate-related sithe ventricular action potential. Tan increased likelihood that aextrasystole will occur in this vulneand produce a ventricular arrhyth

350-

0

300-

0

0 0

Pacing cycle length (ms

I In cy I

Pacing cycle length (ins)

is, however, no direct evidence for this hypo-l have been thesis. Only two of the seven patients withbut not in documented torsade de pointes did not havesyndromes. evident rate adaptation. However, an increas-naffected by ing ventricular rate may be associated with anthat the QT increase in U wave amplitude, a feature thatand also on often precedes the development of torsade de

itivity of the pointes.22 Changes in-the U wave amplitude, Ts in the QT wave alternans, and the occurrence of torsadeexponential de pointes often appear shortly after an abruptnd heart rate increase in heart rate,23 as in the present study,evaluate the or can be elicited by other manoeuvres thatwith atrial'8 increase sympathetic tone such as isoprenaline,that in nor- as seen in patient 4 (fig 2), exercise," or themplete heart Valsalva manoeuvre.25vith increas- The recording of monophasic action poten-en these two tials has provided useful insight into thenear relation mechanisms of congenital QT prolongation.itely 70 and Abnormalities of the monophasic action poten-

tial were described by Bonatti and colleages.'0ngenital QT These workers described "humps" arising)nential or from phase 3 of the action potential. Theseptation. One humps were regarded by others as representingshowed no early afterdepolarisations,22 a phenomenon

ievere mani- produced in vivo by administering caesiumrer 300 syn- chloride to canine hearts.' In these studiesinterval was ventricular arrhythmias with similar electro-2) and he cardiographic characteristics to torsade deIt has been pointes were described which were initiatedn of the QT by early afterdepolarisations. Jackman andtal long QT colleagues reported torsades de pointe arising-eptibility to in humans coincidentally with a hump on theilure of QT endocardial monophasic action potential; and itbe due to a has been suggested that early afterdepolarisa-hortening of tions are responsible for the initiation of tor-'hus there is sades de pointe in humans. In their illustrationventricular (fig 50,22), the initiating ventricular extrasystolerable period has a configuration and axis suggesting an originmia.4 There remote from the site of the monophasic action

potential recording.22 This was confirmed bythe fact that the onset of the ventricular

L L premature contraction preceded that of theright ventricular monophasic action potential.We too have noted this phenomenon (fig 2).Thus changes in local right ventricular mono-phasic action potential contour may beirrelevant to the onset of the prematureextrasystole. Though we saw "humps" on themonophasic action potentials of patient 4, it isclear that spontaneous torsade de pointes werenot consequent upon the "hump" on the rightventricular monophasic action potential (fig 2).

R ~',\t N The available data do not seem to support theearly afterdepolarisations hypothesis, but they

400-

E

0L-

, 350-00)

3001J

00

@00

0

Figure 6 Right atrialmonophasic actionpotential recordings duringatrial pacing in patient 6.

235

2s0-Jr



j.com/

Br H


ber 1991. Dow

nloaded from


Linker, Camm, Ward

are consistent with the theory, advanced byHan and colleagues, Kuo and colleagues, andSurawicz that dispersion of repolarisation isresponsible for torsade de pointes in thesepatients.27' When dispersion of repolarisationwas assessed in patients with normal hearts thedegree of monophasic action potential disper-sion was up to 40 ms'2 3 and in patients withcongenital QT prolongation it ranged from 100to 270 ms.10 None the less, existing animalmodels of arrhythmias dependent on increaseddispersion of repolarisation have beenproduced by large temperature gradients,29 orinjections of large doses of contrast into con-ronary arteries,3' conditions seldom encoun-tered in clinical practice. Nor does the produc-tion of early afterdepolarisations by caesiumchloride occur in practice. Clinical observa-tions of patients with torsade de pointes do notnecessarily provide the solution to the questionof mechanism, because most observations canbe explained by both mechanisms. At thepresent time, neither of the two proposedsubstrates has been conclusively shown to bethe cause of torsade de pointes in humans.In normal subjects, the monphasic action

potential duration decreased linearly withdecreasing pacing cycle lengths over a range of800 to 350 ms.32 This has also been shown inguinea pig33 and canine34 ventricles in vitro. It isof note that the decrease in monophasic actionpotential duration in our patients was moreconsistent than changes in QT adaptation.Because the T wave is considered to be thesummation of phase 3 of the action potential ofall individual myocardial fibres, a parallel can beexpected between the duration of the ven-tricular monophasic action potential and that ofthe QT interval.35 None the less, the precisionofthis relation is affected by individual variationin monphasic action potential duration as wellas the precision of measurement of both func-tions, so it is difficult to interpret the dis-crepancy between these functions that was seenin two patients in our study. Furthermore, themonophasic action potential looks only at onesite within the ventricular mass, whereas theQT interval is the summation of an infinitenumber and it is possible that if recordingswere made from several sites, variability inmonophasic action potential adaptation wouldbe found, consistent with the abnormal QTchanges. Unfortunately, the patient who hadthe least QT rate adaptation (case 2) did nothave monophasic action potentials recorded.# Adrenergic blocking agents have long been

used to treat arrhythmias associated withprolongation of the QT interval and wereshown to reduce mortality over a ten yearperiod from 78% in untreated patients to 6% intreated patients.6 Studies in which patientswere given intravenous f blockade have notshown any significant change inQT interval,3738although a tendency for the QTc value toshorten as the heart rate fell was noted. Whileour results did not show any overall change inQT interval during sinus rhythm after ,B block-ade, there was considerable interpatientvariability, with some patients showing a shor-tening ofQT interval and some an increase. It

is of interest that propranolol normalised theQT rate adaptation in the two patients withabnormal baseline adaptation (patients 1 and2). The mechanism of this normalisation isunclear but may be due to a reduction in thedispersion of repolarisation or a suppressanteffect on afterdepolarisations, which has beenshown in vitro.22

It is not surprising, in view of the smallchange seen in the QT interval, that no changein MAP9, duration was noted in our patientsafter, blockade. This accords with the work ofothers who also found no change in MAP90duration after intravenous metoprolol.37Others, however, found that long term treat-ment with atenolol or metoprolol increased theMAP90 duration by 12% in patients withcoronary artery disease.3739There are some limitations in the present

study with the evaluation of QT dynamics.Pacing for 10 seconds does not take intoaccount hysteresis of the QT interval. We havepreviously shown that the decrease in QTinterval with increasing pacing rate occurs in abi-exponential form and takes approximatelythree minutes to complete 90% of its adapta-tion.' If steady state QT intervals wereobtained, however, the study would either beexcessively long or only a few heart rates couldbe assessed. The pacing protocol was the samein all patients and it is surprising that noabnormalities were seen in the other sixpatients. It is unlikely that the observed abnor-malities in rate adaptation were artefactual ordue to inaccuracies in the measurement of theQT interval. All measurements were made bytwo observers, although in patient 1 the altera-tion of a few measurements would result inreasonably linear rate adaptation. For clinicalreasons not all patients completed the fullprotocol.

In summary, in this study of ventricularrepolarisation (monophasic action potentialduration, monophasic action potential con-figuration ("humps"), monophasic actionpotential dispersion, and monophasic actionpotential/QT rate adaptation) in the long QTsyndrome no single abnormality could be iden-tified as being associated with the tendency todevelop torsade de pointes. The meaning of"humps" or early afterdepolarisation seen invivo remains unclear; when such features wereseen in our patients they seemed to beirrelevant to the onset of the arrhythmia.Abnormal rate adaptation was found in only afew patients and is therefore unlikely to bedirectly related to the genesis of arrhythmias.Perhaps this response is found in only the mostseverely affected cases. Dispersion of repolar-isation has long been suggested as a potentialmechanism for ventricular arrhythmias in thissyndrome.4' Our data, albeit limited, are notinconsistent with this suggestion but the merepresence of inhomogeneity of repolarisationduring sinus rhythm cannot be adduced asevidence for the hypothesis. Our study has notclarified the role of differential sympatheticneural activity in this syndrome.36 Otherevidence indicates that it is likely to be ofsecondary importance to a primary cellular

236



j.com/

Br H


ber 1991. Dow

nloaded from



abnormality that continues to exist when sym-pathetic influences are removed.

We thank Dr E A Shinebourne for permission to includepatients 2-5 in this study. NJL is the holder of a SquibbCardiovascular Research Fellowship.

1 Jervell A, Lange-Nielsen F. Congenital deaf mutism,prologned QT interval, and sudden death. Am Heart J1957;54:59-68.

2 Romano C, Gemme G, Pongiglione R. Aritmie cardiacherare dell'eta periatrica: II. Accessi sincopali perfibrillazione ventricolare parossistica. Clin Pediatr(Bologna) 1963;45:656-61.

3 Ward OC. New familial cardiac syndrome in children. J IrMed Assoc 1964;54:103-6.

4 Attwell D, Lee JA. A cellular basis for the primary long Q-Tsyndromes. Lancet 1988;i: 1136-9.

5 Cowan JC, Yusoff K, Moore M, et al. Importance of leadselection in QT interval measurement. Am J Cardiol1988;61 :83-7.

6 Day CP, McComb JM, Campbell RWF. QT dispersion: anindication of arrhythmia risk in patients with long QTintervals. Br Heart J 1990;63:342-4.

7 Taran LM, Szilagyi N. The duration of the electrical systole(QT) in acute rheumatic carditis in children. Am Heart J1947;33: 14-26.

8 Bazett HC. An analysis of the time relationships of the heart.Heart 1920;7:353-70.

9 Ward DE. Prolongation of the QT interval as an indicator ofrisk of a cardiac event. Eur Heart J 1988;9(suppl G):139-44.

10 Bonatti V, Rolli A, Botti G. Recording of monophasic actionpotentials of the right ventricle in long QT syndromescomplicated by severe ventricular arrhythmias. Eur HeartJ 1983;4:168-79.

11 Brorson L, Olsson SB. Right atrial monophasic actionpotential in healthy males. Acta Med Scand 1976;199:433-46.

12 Hoffman BF, Cranefield PF, Lepeschkin E, Surawicz B,Herrlich HC. Comparison of cardiac monophasic actionpotentials recorded by intracellular and suction elec-trodes. Am J Physiol 1959;196:1297-301.

13 Franz MR, Bargheer K, Rafflenbeul W, HaverichA, LichtlenP. Monophasic action potential mapping in humansubjects with normal electrocardiograms: direct evidenceof the genesis of the T wave. Circulation 1987;75:379-86.

14 Rickards AF, Norman J. Relation between QT interval andheart rate: new design of physiologically adaptive cardiacpacemaker. Br Heart J 1981;45:56-61.

15 Horstmann E, Koenn B. Temporal relationship betweenexercise and QT shortening in patients with QT pace-makers. PACE 1989;12:1080-4.

16 Lecocq B, Lecocq V, Jaillon P. Physiologic relation betweencardiac cycle length and QT duration in healthy volun-teers. Am J Cardiol 1989;63:481-6.

17 Romano M, Clarizia M, Onofrio E, et al. Heart rate, PR,and QT intervals in normal children: a 24-hour Holtermonitoring study. Clin Cardiol 1988;11:839-42.

18 Milne JR, Camm AJ, Ward DE, Spurrell RAJ. Effect ofintravenous propranolol on QT interval. A new method ofassessment. Br Heart J 1980;43:1-6.

19 Milne JR, Ward DE, Spurrell RAJ, Camm AJ. Theventricular paced QT interval-the effects of rate andexercise. PACE 1982;5:352-8.

20 Susmano A, Graettinger JS, Carleton RA. The relationshipbetween Q-T interval and heart rate. J Electrocardiol1969;2:269-74.

21 Akhras F, Richards AF. The relationship between QTinterval and heart rate during physiological exercise andpacing. Jpn Heart J 1981;22:345-51.

22 Jackman WM, Friday KJ, Anderson JL, Aliot EM, Clark M,Lazarra R. The long QT syndromes: a critical review,new clinical observations and a unifying hypothesis.Prog Cardiovasc Dis 1988;31 :115-72.

23 Wellens HJJ, Vermuelen A, Durrer D. Ventricular fibrilla-tion occurring on arousal from sleep by auditory stimuli.Circulation 1972;46:661-5.

24 Phillips J, Ichinose H. Clinical and pathologic studies in thehereditary syndrome of a long QT interval, syncopalspells, and sudden death. Chest 1970;588:236-43.

25 Mitsutake A, Takashita A, Kuroiwa A, Nakamura M.Usefulness of the Valsalva maneuver in management ofthe long QT syndrome. Circulation 1981;63:1029-35.

26 Levine JH, Spear JF, Guarnieri T, et al. Cesium chloride-induced long QT syndrome: demonstration of after-depolarizations and triggered activity in vivo. Circulation1985;72: 1092-103.

27 Han J, Goel BG. Electrophysiologic precursors of ven-tricular tachyarrhythmias. Arch Intern Med 1972;192:749-55.

28 Han J, Moe GK. Nonuniform recovery of excitability inventricular muscle. Circ Res 1964;14:44-60.

29 Kuo CS, Munakata K, Reddy CP, Surawicz B. Characteris-tics and possible mechanism of ventricular arrhythmiadependent on the dispersion of action potential durations.Circulation 1983;67:1356-67.

30 Surawicz B. Electrophysiologic substrate of torsade depointes: dispersion of repolarization of early after-depolarizations? JAm Coll Cardiol 1989;14:172-84.

31 Murdock DK, Euler DE, Becker DM, Murdock JD, Scan-lon PJ, Gunnar RM. Ventricular fibrillation during coron-ary angioplasty: an analysis of mechanisms. Am Heart J1985;109:265-73.

32 Franz MR, Swerdlow CD, Liem B, Schaefer J. Cycle lengthdependence of human action potential duration in vivo.Effect of single extrastimuli, sudden sustained rateacceleration and deceleration, and different steady-statefrequencies. J Clin Invest 1988;82:972-9.

33 Savigny LV, Hohnloser S, Antoni H. Effects of changes infrequency on guinea pig ventricular action potentialduration and on QT interval under different experimentalconditions. Basic Res Cardiol 1981;76:276-88.

34 Moore EN, Preston JB, Moe GK. Durations of transmem-brane action potentials and functional refractory periodsof canine false tendon and ventricular myocardium. Com-parisons in single fibres. Circ Res 1965;27:259-73.

35 Olsson SB, Brorson L, Edvardsson N, Varnauskas E.Estimation of ventricular repolarization in man by mono-phasic action potential recording technique. Eur Heart J1985;6(suppl D):71-9.

36 Schwartz PJ, Locati E. The idiopathic long QT syndrome:pathogenetic mechanisms and therapy. Eur Heart J1985;6(suppl D):103-14.

37 Edvardsson N, Olsson SB. Effects of acute and chronic beta-receptor blockade on ventricular repolarisation in man. BrHeart J 1981;45:628-36.

38 Creamer JE, Nathan AW, Shennan A, Camm AJ. Acute andchronic effects of sotalol and propranolol on ventricularrepolarisation using constant-rate pacing. Am J Cardiol1986;57: 1092-6.

39 Way BPJ, Forfar JC, Cobbe SM. Comparison ofthe effects ofchronic oral therapy with atenolol and sotalol on ven-tricular monophasic action potential duration and effectiverefractory period. Am Heart J 1988;1 16:740-6.

40 Lau CP, Freedman AR, Fleming S, Malik M, Camm AJ,Ward DE. Hysteresis of the ventricular paced QT intervalin response to abrupt changes in pacing rate. CardiovascRes 1988;22:67-72.

41 Surawicz B, Knoebel SB. Long QT: good bad or indifferent.J Am Coll Cardiol 1984;4:398-413.

237



j.com/

Br H


ber 1991. Dow

nloaded from


j dynamics ventricular repolarisation in congenitallong ...2 m 15 mnth jln syncope > 300 + 3 m...

Documents