Sequence of epicardial repolarisation and

Transcription

1 Br Heart J 1988;6: Sequene of epiardial repolarisation and onfiguration of the T wave J CAMPBELL COWAN, COLIN J HILTON, CLIVE J GRIFFITHS, SUDA TANSUPHASWADIKUL, JOHN P BOURKE, ALAN MURRAY, RONALD W F CAMPBELL From the Departments of Cardiology, Cardiothorai Surgery, and Medial Physis, Freeman Hospital, Newastle upon Tyne SUMMARY Epiardial ativation and repolarisation sequenes were investigated in patients with upright or inverted T waves in left ventriular leads of the surfae eletroardiogram. Fifteen patients were studied: 1 were undergoing oronary artery bypass grafting (upright T waves) and five aorti valve replaement (four patients with T inversion). Monophasi ation potentials were reorded intraoperatively from eight to 1 left ventriular sites in eah patient. In patients with upright T waves there was an inverse relation between the duration of the monophasi ation potential and the ativation time (mean slope -1-44). As a onsequene, ativation and repolarisation proeeded in opposite diretions. Dispersion of repolarisation time (14 ms) was less than dispersion of ativation time (23 ms). In patients with T wave inversion aused by aorti stenosis there was no relation between the duration of ation potential and ativation time; the repolarisation sequene resembled the ativation sequene, and the dispersion of repolarisation time was greater than the dispersion of ativation time (31 and 26 ms respetively). These results show that there are epiardial repolarisation gradients in man and that these are related to the onfiguration of the T wave. In patients with upright T waves an inverse relation between the duration of the ation potential and the ativation time redues the dispersion of the repolarisation time. When the T wave was inverted this relation was no longer found and the dispersion of repolarisation inreased. The human ventriular ativation sequene has been extensively studied. Knowledge ofthe normal ativation sequene"2 has ontributed to our understanding of the role of abnormalities of ativation in arrhythmogenesis.' In ontrast, little is known ofthe repolarisation sequene in health or disease. As a onsequene the role of repolarisation abnormalities in arrhythmogenesis remains unertain. Animal studies have shown that inreased dispersion of repolarisation is arrhythmogeni,' but it has been diffiult to assess dispersion of repolarisation in man. The ontribution of regional differenes in repolarisation to the genesis of the T wave of the surfae eletroardiogram is similarly unlear.7 Requests for reprints to Dr J Campbell Cowan, Departnent of Cardiovasular Studies, Medial Shool, The University of Leeds, Leeds LS2 9JT. Aepted for publiation 13 June 1988 Repolarisation times an be aurately determined from monophasi ation potentials.89 The reent development of intraoperative tehniques for reording monophasi potentials from the epiardium142 offers the opportunity to undertake detailed repolarisation mapping in man. This study reports the sequene of epiardial repolarisation over the left ventrile of patients with upright T waves and patients with inverted T waves in the left ventriular leads of the surfae eletroardiogram. Patients and methods PATIENTS We studied 15 patients: 1 undergoing routine oronary artery bypass grafting and five undergoing aorti valve replaement. Patients undergoing bypass grafting had upright T waves in left ventriular leads of their surfae eletroardiogram (leads I, avl, and V3-V6) before operation. Patients undergoing 424 Br Heart J: first published as /hrt on 1 November Downloaded from on 25 November 218 by guest. Proteted by

2 Epiardial repolarisanon sequene replaement of the aorti valve had dominant aorti stenosis (atheter or Doppler gradients ranging from 6 to 11 mm Hg) and evidene of left ventriular hypertrophy but no or minimal aorti regurgitation. There was T wave inversion in the left ventriular leads ofthe preoperative eletroardiogram in four of the five patients with aorti stenosis. In all four patients, leads I, avl, V5, and V6 showed an inverted T wave. T wave onfiguration in leads V3 and V4 was either biphasi or inverted (V3 three patients biphasi, one inverted; V4 two patients biphasi, two inverted). In the fifth patient T waves were upright in the left ventriular leads. Four of the five patients with aorti stenosis underwent oronary angiography and none had linially signifiant oronary disease. All patients were in sinus rhythm and none had bundle branh blok. No patient had a history of arrhythmias or myoardial infartion and none was being treated with digoxin. Eight of the 1 patients undergoing oronary artery bypass grafting were taking p blokers. All patients gave informed onsent for mapping. The study was approved by the ethis ommittee of the hospital. EPICARDIAL MONOPHASIC ACTION POTENTIAL RECORDING We reorded monophasi ation potentials intraoperatively from the epiardium using a hand held probe.'2 The probe was an eletrode assembly mounted on a flexible spatula. The eletrode assembly was made from two silver-silver hloride eletrodes-a entral blunted spike eletrode, whih protruded by 1 mm, surrounded by a flush mounted ring eletrode (diameter 7 mm). Potentials were reorded through a purpose built amplifier (bandpass 7 to 1 Hz). The signals were displayed on a monitor, whih was visible to the surgeon, and reorded on paper (Mingograf) and on magneti tape (Raal Store 7). For ethial reasons, reordings were onfined to the epiardium and the transmural signals were not investigated. ELECTROCARDIOGRAPHIC MONITORING Lead I of the surfae eletroardiogram was monitored ontinuously. In three patients the amplitude ofthe T wave dereased after thoraotomy, but the T wave polarity remained onstant. Beause mapping required ardia manipulation and hanging the orientation of the heart, the surfae eletroardiogram varied during mapping. At intervals during mapping and on its ompletion the surfae eletroardiogram was reorded with the heart in its normal position: T wave onfiguration was unhanged. 425 EPICARDIAL MAPPING The heart rate was maintained onstant by right atrial paing at a fixed rate 1-2 beats/min above the spontaneous rate. Mapping was undertaken on ardiopulmonary bypass to failitate epiardial aess and to minimise the haemodynami onsequenes of ardia manipulation. The patients' temperature was maintained within the range 36-37C, as indiated by an oesophageal thermistor probe. Mapping was started after six minutes of bypass to avoid the transient hanges in the duration of the ation potential that our in the first few minutes of bypass.'2 Potentials were reorded from 1O sites over the left ventrile (fig 1). These sites overed most of the epiardium of the left ventrile. We did not attempt to map the most apial portion of the ventrile, beause mapping this area did not give satisfatory results.'2 We reorded from the 1 sites on four oasions, starting at a different site on eah oasion to ensure that any time dependent hanges would influene eah site to the same extent. Mapping took approximately eight minutes. DEFINITIONS AND MEASUREMENTS Monophasi ation potentials were aeptable if they showed an amplitude at least 4 mv, a smooth repolarisation phase, freedom from baseline artefat, and onstany of onfiguration over at least two onseutive yles. Potentials were timed with respet to a fixed left ventriular referene eletrode (bandpass 3 to 3 Hz) (fig 2). AS Fig 1 Epiardial mapping sites. The left ventrik was onsidered as a grid. The ten sites indiated were mapped. AS, anterior septum; PS, posterior septum. Br Heart J: first published as /hrt on 1 November Downloaded from on 25 November 218 by guest. Proteted by

3 426 Monophasi ation potential Referene Measurements Duration of ation potential Repotarisation time Fig 2 Monophasi ation potentials were timed in relation to afixed ventriular referene eletrode. Repolarisation time at any site is determined by the sum of ativation time and ation potential duration. The ativation time was defined as the interval between referene and the visually determined time of maximum veloity of the upstroke of the monophasi ation potential. The ation potential duration was defined as the interval between the time of maximum upstroke veloity and the time of 9% repolarisation from maximum plateau amplitude to baseline. The repolarisation time was defined as the interval between referene and the time of 9% repolarisation-that is, the sum of the ativation time and the ation potential duration. For eah epiardial site we measured the ativation time, ation potential duration, and repolarisation time as the mean of the four individual probe appliations at that site. The dispersion of eah variable was defined as the differene between the maximum and minimum values of the site means within eah patient. MEASUREMENT OF EPICARDIAL TEMPERATURE We studied the variation in epiardial temperature in four of the patients undergoing bypass grafting by applying a small thermoouple probe diretly to the epiardium. Temperature was reorded at eah ofthe 1 sites used for mapping. The effets of ooling on Cowan, Hilton, Griffiths, Tansuphaswadikul, Bourke, Murray, Campbell the duration of the ation potential were studied in three ofthe four patients. Monophasi potentials and temperature were reorded simultaneously during ooling and before ardioplegia. The probe was applied repeatedly at the same site and the heart rate was kept onstant by atrial paing. STATISTICAL ANALYSIS The duration of the ation potential and the ativation time at different epiardial sites were plotted and the regression line was alulated for eah patient. Ninety five per ent onfidene limits and the statistial signifiane of the slope of the regression lines were alulated. Before omparison, values in different patients were normalised to eliminate differenes aused by referene plaement, the duration ofation potential, and heart rate. For eah patient, the results of all left ventriular sites were pooled to provide a mean ativation time, ation potential duration, and repolarisation time. The deviation of eah site from the mean was then determined, thereby providing a normalised estimate of the relative timings of different sites. The normalised values for sites were pooled to determine a mean ativation and repolarisation sequene for the study group. Results ACTIVATION AND REPOLARISATION SEQUENCES FOR UPRIGHT T WAVES The ativation time, ation potential duration, and repolarisation time were studied in 1 patients with upright T waves. An inverse relation was seen between the duration of the ation potential and the ativation time (fig 3). The slope of the regression line for the duration ofthe ation potential against the ativation time was negative in all 1 patients (table 1). The 95% onfidene limits for the slope varied in different individuals, ahieving statistial signifiane in seven patients and borderline signifiane in two others. Beause the repolarisation time is determined by the sum ofthe ativation time and the ation potential duration, this inverse relation resulted in a relative uniformity ofrepolarisation times-in eight ofthe 1 patients dispersion of repolarisation time was less than dispersion of ation potential duration (table 2). As a further onsequene, the repolarisation time in individual patients was in general found to be independent of the ativation time (fig 4a). The exeption, patient 1, showed a signifiant inverse relation between repolarisation time and ativation time (slope- 19, p < 25), refleting a partiularly steep inverse relation between the ativation time and the ation potential duration (slope - 2 9). Br Heart J: first published as /hrt on 1 November Downloaded from on 25 November 218 by guest. Proteted by

4 - Epiardial repolarisation sequene i ED E. 25- a- L to O 23-._v- 4 Ua - o C I I U 21- I I * T Ativation time (ms) Ativation time (ms) Fig 3 Relation between the duration of the ation potential and the ativation time in a patient with upright T waves. The relation between the duration of the monophasi ation potential and the ativation timefor left ventriular sites in one patient (patient 4, table 1) is given on the left. The linefor the regression of the duration of the ation potential on the ativation time is shown (slope ). The resulting distribution of repolarisation times and ativation times is shown on the right. In view of the variation in the slopes of the regression lines in individual patients, a mean ativation time, duration of ation potential, and repolarisation time were derived for eah site by pooling the standardised data from all 1 patients. Figure 5 shows the relation between the ativation time and the duration of the ation potential for these standardised site means. There was a lose inverse relation between the ativation time and the duration of the ation potential (slope ,95% onfidene limits of slope to , p < 1). In ontrast with the data for individual patients, a statistially signifiant inverse relation between repolarisation Table 1 Regression of ation potential duration on the ativation time in patients with upright T waves Case Correlation 95% onfide No oeffiient Slope imits of slope Probablity to p < to -2-5 p < -1,NS to p < -OS to p < lto-1-21 p>-1,ns to-1-86 p < -1,NS to p < to-1-67 p < to p < I35 to p < -1 time and ativation time was evident in the pooled data, indiating earlier repolarisation at sites of later ativation (slope -44, 95% onfidene limits of slope --15 to --73, p < -1). The fat that a statistially signifiant inverse relation was demonstrated in only one of the 1 patients, when onsidered individually, reflets the satter of results Table 2 Dispersion of ativation time, ation potential duration, and repolarisation time in patients with upright T waves Dispersion (ms) No of Ation Case sites potential No measured Ativation duration Repolarisation I The number of sites giving aeptable potentials varied between eight and 1 in different patients. The dispersion of eah variable is the differene between the maximum and minimum values of that variable. Br Heart J: first published as /hrt on 1 November Downloaded from on 25 November 218 by guest. Proteted by

5 428 iomvj 25ms Fig 4 (a) Monophasi potentials at two epiardial sites in a patient with upright T waves. Potentials were aligned using the referene signals (bipolar reordings, lower traes). Despite differing ativation times, repolarisation time at the two sites was almost synhronous, refleting a shorter ation potential duration at the site with later ativation. (b) Monophasi potentials at two epiardial sites in a patient with inverted T waves aused by aorti stenosis. The duration of the ation potential was independent of the ativation time and hene the site whih ativated later also repolarised later. For larity ofpresentation, the upstrokes ofpotentials have been retouhed. - 1._2 ED.- o- 4 CL 2 ( L Cowan, Hilton, Griffiths, Tansuphaswadikul, Bourke, Murray, Campbll within individuals and the relatively flat slope of the regression line relating repolarisation time and ativation time. The ativation and repolarisation sequenes derived from these individual site means are plotted shematially in fig 6. Ativation ourred earliest in the anterior and posterior septum. It then spread in two general diretions, from eah septal border towards the free wall ofthe ventrile and from apex to base. Dispersion of the mean ativation times over the left ventriular sites was 23 ms. The repolarisation sequene was very different from the ativation sequene. Repolarisation ourred earliest in the posterobasal region-the areas of latest ativation. The areas over the anterior and posterior septum were the last to repolarise; these areas orrespond approximately to the earliest areas of ativation. There was therefore an inverse relation between the sequenes of ativation and repolarisation. The dispersion of the mean repolarisation times over the left ventrile was 14 ms. AORTIC STENOSIS WITH T WAVE INVERSION We studied ativation time, ation potential duration, and repolarisation time in four patients with T wave inversion aused by aorti stenosis. In ontrast with patients with upright T waves, the duration of the ation potential was independent ofthe ativation time (table 3). As a onsequene, the repolarisation time was determined by the ativation time (fig 4b.) 4,1 VI - C Br Heart J: first published as /hrt on 1 November Downloaded from u I- "- -1- o -2-3 a VA oc 1. a 'a o- E: L I~ Ativation time (ms) Ativation time (ms) Fig 5 Mean relation between (a) the duration of the ation potential and the ativation timefor all IO patients with upright T waves. Normalised datafrom individuals were pooled to provide a mean ativation time and the mean duration of the ation potentialfor eah site. (b) The relation between repolarisation time and ativation time. Both figures show the SEMs and the regression lines. on 25 November 218 by guest. Proteted by

6 Epiardial repolarisation sequene Ativation Repolarisation a 1- I a.oo to - C - a 5 Frt r 1 i ~~~~~~~~~~~~~~~~~~~~-i-a 41- ~i ~~~~~~~1 2 25ms Fig 6 Representation of left ventriular ativation and repolarisation sequenes for patients with upright Twaves. Diagrams are based on the means of the normalised datafor all 1 patients with upright T waves. The 1 mapped sites within the grid are indiated by dots. Estimates of ativation time and repolarisation time at unmapped sites were derived as the mean of measurements of all adjaent sites. The shaded areas indiate the extent of depolarised myoardium in suessive 5 ms timeframes, timedfrom the earliest ativation or repolarisation. The anterior septun is on the left and the posterior on the right (as infig 1). The dispersion of ativation was 23 ms and the dispersion of repolarisation was 14 ms. Table 3 Regression of ation potential duration on In three of the four patients dispersion of repolarisation time was greater than dispersion of ation ativation time in patients with aorti stenosis potential duration (table 4). Case Correlation 95% onfiden The mean ativation time, ation potential duration, and repolarisation time for eah site were No oeffiient Slope limits of slope Probability Inverted T waves: derived by pooling the standardised data for all four to --84 p > -1, NS patients (fig 7). There was no signifiant relation to-1-16 p < O1,NS to p > O1, NS between the ation potential duration and ativation *84to --72 p > 1,NS time. As a result, the repolarisation time was signifiantly related to the ativation time (slope -86, Upright T waves: to p < -1 95% onfidene limits of slope -26 to 1 46, p <.1). r I.I I Ativation time (ms) 2 E 1 E -O.. o O.a a -2- i r r Ativation time (ms) Fig 7 Mean relation between (a) the duration of the ation potential and the ativation time for allfour patients with T inversion aused by aorti stenosis. (b) The resultant relation between repolarisation time and ativation time. Both figures show the SEMs and the regression lines. 25ms Br Heart J: first published as /hrt on 1 November Downloaded from on 25 November 218 by guest. Proteted by

7 43 Ativation x--t--..i t-...,. iit 14* - _ Repolarisation Cowan, Hilton, Griffiths, Tansuphaswadikul, Bourke, Murray, Campbell ms Fig 8 Representation of left ventriular ativation and repolarisation sequenes for patients with T inversion aused by aorti stenosis. Diagrams are based on the means of the normalised data of the four patients. The shaded areas indiate the extent of depolarised myoardiun in suessive 5 ms timeframes. The dispersion of ativation was 26 ms and the dispersion of repolarisation was 31 ms. Table 4 Dispersion of ativation time, ation potential duration, and repolarisation time in patients with aorti stenosis Dispersion (ms) No of Ation Case sites potential No measured Ativation duration Re,polarisation Inverted T waves: Upright T waves: Figure 8 shows the mean ativation and repolarisation sequenes. The ativation sequene was similar to the sequene in patients with upright T waves. Ativation ourred earliest on either side of the septum and proeeded from the septum to the free I 35 apatient 1 4 o 'S I ạ 2ft 5. 3 JI1 Jz a I 23- r--r- 112 li 12 E S- o Patient 2 i I ms I ~ ~ I s.4 wall and from the apex to the base. Dispersion of the mean ativation times was 26 ms. The repolarisation sequene differed from that in patients with upright T waves; it resembled the ativation sequene. Dispersion of the mean repolarisation times (31 ms) was slightly greater than dispersion of ativation times. AORTIC STENOSIS WITH UPRIGHT T WAVES We studied one patient with aorti stenosis and upright T waves. This patient resembled the other patients with upright T waves in showing a statistially signifiant inverse relation between ation potential duration and ativation time (table 3). VARIATION IN THE EPICARDIAL TEMPERATURE In the four patients we studied a two way analysis of variane showed no statistially signifiant differene between epiardial sites. The left ventrile was onsidered as a series of four oaxial rings from base * O I Temperature ( C) ho E 32 -.o _ 3 u.2 M 28 C:l Patient * I I I I Fig 9 Relation between the duration of the monophasi ation potential and temperature. The effets of ooling on ation potential duration in three separate patients are shown.. Br Heart J: first published as /hrt on 1 November Downloaded from on 25 November 218 by guest. Proteted by

8 Epiardial repolartsation sequene to apex. The mean temperature ofthe four rings from base to apex was 35-1, 35-1, 35(, and 3417 C (differenes non-signifiant). In three of the four patients the relative magnitude of temperature effets on ation potential duration was determined by reording monophasi potentials during ooling. The ation potential duration was prolonged (fig 9). The slopes of the regression lines relating ation potential duration and temperature varied between 4 and 12 ms prolongation per C of ooling. Disussion This study presents the first maps of the sequene of human epiardial repolarisation. We showed that epiardial repolarisation gradients do exist and that they are related to the onfiguration of the T wave on the surfae eletroardiogram. The relation between ativation and repolarisation sequenes was different in patients with upright T waves and in those with inverted T waves. STUDY LIMITATIONS It is not feasible to reord intraellular potentials from the beating whole human heart. The use of monophasi ation potentials as a substitute depends on ertain assumptions. Monophasi potentials are of muh lower amplitude than their intraellular ounterparts. Though their overall duration orresponds losely with the intraellular potentials,89 the point of9% repolarisation is measured in relation to the maximum plateau amplitude and hene may be amplitude dependent. The 9% repolarisation point is only independent of amplitude if the monophasi potential is proportionately redued from the intraellular potential throughout the whole of repolarisation. Previous studies in isolated rabbit hearts have suggested that this is indeed the ase.9 There are additional problems in studying repolarisation in vivo. Repolarisation, unlike ativation, is a dynami proess, influened by many fators and apable of rapid hange.7 Consequently, the "normal" or baseline repolarisation state is diffiult to establish and to maintain. During open hest surgery differential epiardial ooling or ardiopulmonary bypass may ause hanges in repolarisation. We saw modest evidene of differential apial ooling. Based on the lengthening of the duration of the ation potential during ooling, the observed 4UC temperature differene between base and apex would be expeted to lengthen the ation potential duration at the apex by 2-5 ms relative to the base. This effet is small ompared with the observed differenes in the duration of the ation potential. Cardiopulmonary bypass auses a transient 431 prolongation of the duration ofthe ation potential in the first minutes of bypass.'2 In the present study we started mapping after six minutes of bypass to avoid these transient effets. We adopted a strategy of repeated point mapping to ensure that omparisons between epiardial sites would be independent ofany ontinuing time-dependent influenes. We therefore believe that intraoperative influenes on epiardial repolarisation gradients were relatively small. DISPERSION OF REPOLARISATION In patients with upright T waves we found an inverse relation between the duration of the ation potential and the ativation time. The slope of this relation was signifiantly more negative than -1, refleting inverse ativation and repolarisation sequenes. A similar inverse relation between the duration of the ation potential and the ativation time was reported by Franz et al, who reorded endoardial monophasi ation potentials in seven patients and epiardial potentials in three patients.' By ontrast, in patients with T wave inversion aused by aorti stenosis the duration of the ation potential was independent of the ativation time (table 3). As a result the repolarisation sequene was largely determined by the ativation sequene. Differing relations between the ativation and repolarisation sequenes ofpatients with upright and inverted T waves have been reported before." An important onsequene ofthe observed inverse relation between the duration of the ation potential and the ativation time is a redution in the dispersion of repolarisation. In patients with upright T waves, the dispersion of the repolarisation time was less than the dispersion of duration of the ation potential. By ontrast, in patients with T inversion aused by aorti stenosis, in whom the duration ofthe ation potential was independent of the ativation time, dispersion of repolarisation time was greater than that of the duration of the ation potential. Dispersion of repolarisation has been shown to favour the development of arrhythmias,'s so minimising dispersion may be a protetive mehanism. The observed inverse relation between the ativation time and the duration of the ation potential may therefore represent a fundamental intrinsi antiarrhythmi property of the myoardium. DIFFERENCES IN EPICARDIAL REPOLARISATION AS DETERMINANTS OF T WAVE CONFIGURATION The onordane ofqrs and T waves in the surfae eletroardiogram led to the lassi hypothesis that ativation and repolarisation must proeed in opposite diretions.'4 Suh an inverse relation might Br Heart J: first published as /hrt on 1 November Downloaded from on 25 November 218 by guest. Proteted by

9 432 Cowan, Hilton, Griffiths, Tansuphaswadikul, Bourke, Murray, Campbell result from a transmural endoardial-epiardial THE LINK BETWEEN THE DURATION OF THE gradient, from gradients aross the epiardial surfae, or from both. The simplest explanation for the inverse relation ACTION POTENTIAL AND THE ACTIVATION TIME There is a onsensus, from animal studies, that between the duration of the potential and the ativation time is an indiret assoiation through a third transmural ativation and repolarisation gradients are opposite.'51' Whereas ativation proeeds from fator, suh as site. A posterobasal loation, for endoardium to epiardium, repolarisation proeeds example, might predetermine both a short ation from epiardium to endoardium. Franz et al have potential duration and a late ativation time. The provided evidene ofa similar transmural gradient in question would then be restated as to how ativation man, based on the finding that endoardial repolarisation ours later, in relation to the QT interval of jointly predetermined on a regional basis. time and the duration of the ation potential might be the surfae eletroardiogram, than epiardial Alternatively, there may be a diret ausal relation, repolarisation.' suh that the duration of the ation potential adapts For gradients aross the epiardial surfae the diretly to the ativation sequene. Although the situation is less lear. There is some measure of basis for suh adaptation is unlear, there is indeed agreement that refratory periods'7 and ation potentials'8 are longer at the apex than at the base. alter in response to hanges in the ativation evidene that the duration of the ation potential an Abildskov found that on the epiardium reovery sequene. Alteration in the ventriular ativation from refratoriness resembled the ativation sequene, whether aused by ventriular paing,"2' sequene, whereas there was an inverse relation on intermittent left bundle branh blok, 27 or intermittent pre-exitation' an produe T wave the endoardium." Toyoshima et al found differing hanges relations between ativation and repolarisation on the anterior and posterior epiardial surfaes.' Spah and Barr found that the apex repolarised later than the base,'5 hene suggesting an inverse relation between ativation and repolarisation. In the present study, we have found that the repolarisation wavefront in individual patients is omplex, refleting differenes in repolarisation times between losely adjaent areas of epiardium. A similar omplexity of the repolarisation wavefront has been reported in animals.'82' Pooling data between patients showed that ativation and repolarisation sequenes were opposite--ativation proeeded from septum to free wall and from apex to base, whereas repolarisation proeeded from base to apex and from free wall to septum. By ontrast, in patients with T inversion, the repolarisation sequene broadly resembled the ativation sequene. These findings are onsistent with lassi theories of the genesis of T waves.they suggest that the normal onordane of QRS and T waves is the result of the observed inverse relation between the duration ofthe ation potential and the ativation time. For ethial reasons, our study was onfined to the epiardium and simultaneous endoardial reordings were not undertaken. The relative ontribution of epiardial and transmural gradients to the T wave remains unertain. Earlier estimates of dispersion of repolarisation in man, whih were based on the variation in QT interval over the body surfae, range from 6 to 8 ms in normal subjets. '23 These values are onsiderably greater than the epiardial dispersion we found. This suggests that transmural gradients have a large effet in the overall dispersion of repolarisation. that persist after the restoration ofnormal ativation. By inferene, the duration of the ation potential alters during the period of abnormal ativation, resulting in an abnormal repolarisation sequene when normal ativation is restored. It an be postulated that an intrinsi property of the myoardium enables the duration of the ation potential to adapt to its ativation sequene. This would result in a diret link between the duration of the ation potential and the ativation time and this would provide a possible mehanism for the inverse relation that we found. This work was supported by a Squibb Cardiovasular Fellowship Award (JCC). The Department of Cardiology is supported by the British Heart Foundation. Referenes 1 Durrer D, Van Dam RTh, Freud GE, Janse MJ, Meijler FL, Arzbaeher RC. Total exitation of the isolated human heart. Cirulation 197;41: Cassidy DM, Vassallo JA, Marhlinsli FE, Buxton AE, Untereker WJ, Josephson ME. Endoardial mapping in humans in sinus rhythm with normal ventriles: ativation patterns and harateristis of eletrograms. Cirulation 1984;7: Gallagher JJ, Kasell JH, Cox JL, Smith WM, Ideker RE, Smith WM. Tehniques of intraoperative eletrophysiologi mapping. Am J Cardiol 1982; 49: Han J, Garia de Jalon P, Moe GK. Adrenergi effets on ventriular vulnerability. Cir Res 1964;14: Kuo C-S, Munakata K, Reddy CP, Surawiz B. Charateristis and possible mehanism of ven- Br Heart J: first published as /hrt on 1 November Downloaded from on 25 November 218 by guest. Proteted by

10 Epiardial repolarisation sequene 433 triular arrhythmia dependent on the dispersion of repolarisation proess of epiardial and endoardial ation potential durations. Cirulation 1983;67: ventriular surfaes. Am Heart J 1984;18: Burgess MJ, Green LS, Millar K, Wyatt R, Abildskov 6 Kuo C-S, Atarashi H, Reddy CP, Surawiz B. Dispersion of ventriular repolarisation and arrhythmia: Heart J 1972;84:66-9. JA. The sequene ofnormal ventriular reovery. Am study of two onseutive ventriular premature omplexes. Cirulation 1985;72:37-6. repolarisation on the ventriular surfae in the dog. 18 Autenrieth G, Surawiz B, Kuo C-S. Sequene of 7 Burgess MJ. Relation of ventriular repolarisation to Am Heart J 1975;89: eletroardiographi T wave-form and arrhythmia 19 Abildskov JA. The sequene of normal reovery of vulnerability. Am J Physiol 1979;236: exitability in the dog heart. Cirulation 1975;52: 8 Hoffman BF, Cranefield PF, Lepeshkin P, Surawiz B, Herrlih HC. Comparison of ardia monophasi 2 Toyoshima H, Lux RL, Wyatt RF, Burgess MJ, ation potentials reorded by intraellular and sution Abildskov JA. Sequenes of early and late phases of eletrodes. Am J Physiol 1959;196: repolarisation on dog ventriular epiardium. J Eletroardiol 1981;14: Franz MR, Burkhoff D, Spurgeon H, Weisfeldt ML, Lakatta EG. In vitro validation of a new ardia 21 Shaefer H, Haas HG. Eletroardiography. In: atheter tehnique for reording monophasi ation Hamilton WF, Dow P, eds. Handbook of physiology, potentials. Eur HeartJ 1986;7: Setion 2: Cirulation, volune 1. Washington DC: 1 Franz MR, Bargheer K, Raifienbeul W, Haverih A, Amerian Physiologial Soiety, 1962:364. Lihtlen PR. Monophasi ation potential mapping 22 Sylven JC, Horaek BM, Spener CA, Klassen GA, in human subjets with normal eletroardiograms: Montague TJ. QT interval variability on the body diret evidene for the genesis of the T wave. surfae. J Eletroardiol 1984;17: Cirulation 1987;75: Mirvis DM. Spatial variation of QT intervals in normal 11 Taggart P, Runnalls M, Donaldson R, et al. Use of persons and patients with aute myoardial infartion. J Am Coll Cardiol 1985;5: monophasi ation potential reordings during routine oronary artery bypass surgery as an index 24 Chatterjee K, Harris A, Davies G, Leatham A. of loalised myoardial ishaemia. Lanet 1986;i: Eletroardiographi hanges subsequent to artifiial ventriular depolarisation. BrHeart J 1969;31: Cowan JC, Griffiths CJ, Hilton CJ, et al. Epiardial 25 Della Bella P, Grazi S, Cipolla CM, et al. Inreased repolarisation mapping in man. Eur Heart J 1987;8: ardia eletrial instability onomitant with paing indued repolarisation abnormalities. Br Heart J 13 Franz MR, Bargheer K, Rafflenbeul W, Frimpong- 1987;57: Botaeng K, Lihtlen PR. Diret evidene for genesis 26 Rosenbaum MB, Blano HH, Elizari JO, Davidenko oft-wave alteration in human left ventriular hypertrophy [Abstrat]. Cirulation 1985;72(suppl III):161. ardia memory. Am J Cardiol 1982;5O: JM. Eletrotoni modulation of the T wave and 14 Wilson FN, Maleod AG, Barker PS. The T defletion 27 Denes P, Pik A, Miller RH, Pietras RJ, Rosen KM. A of the eletroardiogram. Trans Asso Am Physiians harateristi preordial repolarisation abnormality 1931;46: with intermittent left bundle branh blok. Ann 15 Spah M, Barr RC. Ventriular intramural and Intern Med 1978;89:55-7. epiardial potential distributions during ventriular 28 Niolai P, Medvedowsky JL, Delage M, Barnay C, ativation and repolarisation in the intat heart. Blahe E, Pisapia A. Wolff-Parkinson-White syndrome: T wave abnormalities during normal path- Cir Res 1975;37: Higuhi T, Nakaya Y. T wave polarity related to the ways ondution. J Eletroardiol 1981;14: Br Heart J: first published as /hrt on 1 November Downloaded from on 25 November 218 by guest. Proteted by