Monophasic action potentials and activation recovery intervals as measures of ventricular action potential duration: Experimental evidence to resolve some controversies

Monophasic action potentials and activation recovery intervals as measures of ventricular action potential duration: Experimental evidence to resolve some controversies Ruben Coronel, MD, PhD,* Jacques M.T. de Bakker, PhD,*† Francien J.G. Wilms-Schopman, BSc,*† Tobias Opthof, PhD,* André C. Linnenbank, PhD,* Charly N. Belterman, BSc,* Michiel J. Janse, MD, PhD* *From the Department of Experimental Cardiology, Academic Medical Center Amsterdam, Amsterdam, and the † Interuniversity Cardiology Institute of The Netherlands, Utrecht, The Netherlands. BACKGROUND Activation recovery intervals (ARIs) and monophasic action potential (MAP) duration are used as measures of action potential duration in beating hearts. However, controversies exist concerning the correct way to record MAPs or calculate ARIs. We have addressed these issues experimentally. OBJECTIVES To experimentally address the controversies concerning the correct way to record MAPs or calculate ARIs. METHODS Left ventricular local electrograms were recorded in isolated pig hearts with an exploring electrode grid, with a KCl reference electrode on the left ventricular myocardium, the aortic root, or the left atrium. Local activation was determined from calculated Laplacian electrograms. RESULTS With the KCl electrode on the aortic root, local electrograms represented local activation. However, with the KCl electrode on the myocardium remote from the exploring electrode, a combined electrogram emerged consisting of local activation recorded from the grid and remote activation recorded from the reference electrode. The

Action potential duration, refractory period, and dispersion thereof play a major role in arrhythmogenesis.1 A short action potential duration promotes the occurrence of reentrant arrhythmias2; a long action potential duration underlies triggered arrhythmias induced by early after depolarizations.3 Closely colocalized areas with different action potential durations (corresponding to long and short refractory periods) may form the substrate for unidirectional block, a prerequisite for reentry.2 Technically, action potentials are difficult to record in the beating heart, while measuring refractory periods is timeconsuming. As a substitute, monophasic action potential (MAP) measurements are frequently employed,4 – 6 as are activation recovery intervals (ARIs).7,8 MAPs are extracellular signals generated by activation fronts arriving at depolarized myocardium. ARIs are derived from a local unipolar electrogram. Despite the fact that these measurements have been used for many years, there is still controversy about these techniques. Address reprint requests and correspondence: R. Coronel M.D., Ph.D., Department of Experimental Cardiology, Experimental and Molecular Cardiology Groups, Academic Medical Center, Rm M 0108, Meibergdreef 9, 1105 AZ Amsterdam, The Netherlands. E-mail address: r.coronel@ amc.uva.nl. (Received March 28, 2006; accepted May 30, 2006.)

remote, inverted monophasic component did not show propagation and did not correlate with the Laplacian complex. When the KCl electrode was placed on the atrium during AV block, remote atrial monophasic components were completely dissociated from local, ventricular deflections. At left ventricular sites with a positive T wave, the Laplacian signal showed that the end of the T wave was caused by remote repolarization. During cooling-induced regional action potential prolongation, the T wave became negative, whereby the positive flank of the T wave remained correlated with repolarization (recorded with a MAP at the same site). CONCLUSIONS MAPs are recorded from the depolarizing electrode. In both negative and positive T waves, the moment of maximum dV/dt corresponds to local repolarization. KEYWORDS Ventricular repolarization; Monophasic action potential; Activation recovery interval; Laplacian; Unipolar electrogram (Heart Rhythm 2006;3:1043–1050) © 2006 Heart Rhythm Society. All rights reserved.

The MAP is an extracellular recording and only records potentials that are generated by current flowing in the extracellular space. The extracellular current is caused by differences in transmembrane potential generated during propagation of the electrical impulse. As in every extracellular recording, the MAP reflects the potential difference between two recording sites, which are recorded by the exploring (or “different”) electrode and a reference (or “indifferent”) electrode. For a conventional extracellular recording, the exploring electrode is usually made of metal and connected to the positive input of a differential amplifier; the reference is made of metal as well and connected to the negative input. For a unipolar recording, the reference electrode is positioned in nonexcitable tissue at a large distance from the different electrode. In contrast, for a bipolar recording, the reference electrode is positioned next to the exploring electrode (millimeter range) in excitable tissue. For a MAP recording, the exploring electrode usually consists of a 3 M KCl electrode or an Ag/AgCl electrode that is pressed against myocardial tissue, the reference electrode being a “remote” Ag/AgCl electrode. Alternatively, the exploring electrode may be incorporated in a suction cup applied to the myocardium.5 Recently, various papers have been published debating the question of which of the two

1547-5271/$ -see front matter © 2006 Heart Rhythm Society. All rights reserved.

doi:10.1016/j.hrthm.2006.05.027

1044 electrode terminals used in the MAP catheter actually records the MAP.9 –11 Kondo et al9 have recorded MAP signals using a silver wire as the exploring electrode and a 3-M KCl electrode positioned on excitable tissue as a reference electrode.9 To arrive at a signal resembling a conventional MAP, the KCl electrode was connected to the positive input of the amplifier. Although this confuses the definition of which electrode is the reference electrode, the authors did argue that the typical monophasic signal was recorded from the exploring silver wire electrode. Based on the traditional concept of the MAP,6 we hypothesize that the signal recorded in this configuration suggested by Kondo et al9 is composed of two components: (1) a deflection with a morphology similar to an action potential, generated at the site of the reference (KCl) electrode; and (2) local deflections caused by propagation of the electrical impulse and picked up by the exploring electrode. If the differential signal is inverted,9 its shape resembles the action potential. The hypothesis implies that activation mapping with multiple exploring electrodes should make it possible to discriminate between the relative contributions of the signal components. To test this hypothesis, extracellular recordings were made with a multiterminal electrode as the exploring electrode. As a reference electrode, either a stainless steel needle or a 3-M KCl electrode was used, which were placed at different positions on the heart. Interpretation of the end of the repolarization phase of the MAP therefore is influenced by the interpretation of repolarization in the local electrogram. The question whether the upstroke or the downstroke of the T wave corresponds to the moment of repolarization is therefore relevant as well. ARIs are traditionally measured as the interval between the maximum negative slope of the QRS complex and the maximum positive slope of the T wave in the unipolar extracellular electrogram,7,8 the other slope of the T wave representing remote repolarization. However, other studies claim that repolarization times in electrograms with positive T waves are better estimated by choosing the minimal dV/dt of the downslope of the T wave.12–14 The implication of the former is that localized prolongation of the action potential at sites with an initially positive T wave leads to inversion of the local T wave, but this is not true in the latter case. The downslope of a positive T wave then represents remote repolarization of tissue unaffected by the intervention. To resolve this controversy, we tested this implication by simultaneously recording unipolar electrograms and MAPs in isolated blood-perfused porcine hearts from the region perfused by the circumflex (CX) artery and regionally changing perfusion temperature. Furthermore, we calculated the Laplacian signal, indicative of local transmembrane current flow, during the T wave. Our results indicate that the MAP is generated at the site of the depolarized myocardium and that repolarization time in the local unipolar electrogram is to be measured at the positive slope of the T wave.

Heart Rhythm, Vol 3, No 9, September 2006

Methods The experimental protocol complied with the Guide for the Care and Use of Laboratory Animals published by the U.S. National Institutes of Health and was approved by the institutional animal experiments committee.

Experimental setup Pigs (20 –25 kg) of either gender were premedicated with ketamine 350 mg, azaperone 80 mg, and atropine 0.5 mg (intramuscularly). After sedation, the animal was anaesthetized with 20 mg/kg pentobarbital intravenously, intubated, and ventilated with room air and isoflurane. While 1 L of Tyrode’s solution (37°C) was infused, 2 L of blood-Tyrode’s mixture was collected from the superior caval vein. Ventricular fibrillation was induced by a DC current, and the heart was excised rapidly and submerged in ice cold Tyrode’s solution. The aorta was cannulated and connected to a Langendorff setup.15 Both ventricles were drained. Flow was measured continuously; pressure was regulated to obtain a flow of about 100 mL/min. Myocardial temperature was 37°C, and the pH of the perfusate was between 7.35 and 7.45. After defibrillation, total AV block was created in some experiments by crushing the AV-nodal area.

Electrophysiological measurements Recording of electrical activity (three hearts) was done with a multiterminal electrode as the exploring electrode. As a reference electrode either a stainless steel needle or a fine tube filled with 3-M KCl was used. The open end of the tube was brought into contact with the tissue. The multielectrode harbored 208 terminals arranged in a 13 ⫻ 16 matrix, with interelectrode distances of 0.5 mm. Electrode terminals were the cut ends of silver wires with a diameter of 100 ␮m. To detect local activation or local repolarization, Laplacian signals were calculated from the potential difference between the site of recording and its four neighbors (north, south, east, west) in a regular grid. The Laplacian electrogram is the second spatial derivative of the unipolar electrogram and reflects the local transmembrane current. All remote components in the electrograms are canceled.16 Because they accurately reflect local activation only, Laplacian electrograms were used to construct activation maps. During repolarization, the amplitude of Laplacian signals is small and subject to artifacts generated by noise. Therefore, signal averaging was performed from up to 100 subsequent complexes for detection of the Laplacian signal of repolarization. The exploring multielectrode was positioned on the left ventricle (LV) next to the left anterior descending artery (LAD) and halfway between the base and apex. Stimulation was done at a basic cycle length of 600 ms with a bipolar hook electrode at twice the diastolic current threshold. The following protocols were used:

Coronel et al

Measures of Ventricular Repolarization

1045

Stimulation close to the exploring electrodes Two positions for the reference electrode were used: 1. The stainless steel needle electrode attached to the aortic root. 2. The KCl electrode positioned at the base of LV.

Stainless steel reference electrode

RS map

LA LA RA -

Stimulation distant from the exploring electrodes Stimulation was performed at LV base. The following reference electrode positions were applied: 1. The stainless steel reference electrode attached to the aortic root. 2. The KCl reference electrode attached to the aortic root. 3. The KCl reference electrode placed near the site of stimulation. 4. The KCl reference electrode placed on the right atrium. The electrode positions (stimulus electrode, exploring, and reference electrode) and electrode types are indicated in diagrams in the corresponding figures. In a second series of three experiments, the CX artery was cannulated and separately perfused. The temperature of the perfusate was varied over a range of 10°C, from about 40°C to about 30°C. Two stainless steel hook electrodes were introduced: one in the area perfused by the LAD, the other in the CX area. From these electrodes, unipolar electrograms were recorded, with the aortic root electrode as reference. MAPs were recorded with the contact electrode method described by Franz et al,6 as close as possible to the extracellular hook electrode. Intramyocardial temperature was measured by two thermistor needles (diameter 0.5 mm) introduced into the LV wall in the regions perfused by the LAD and the CX.

Data acquisition and analysis The multielectrode was connected to a PC-based data-acquisition system that was able to record 256 channels simultaneously (Biosemi, Amsterdam). Signals were filtered (3 dB points: DC and 400 Hz). A selected 2 seconds of data were stored on the hard disk of the computer. The sampling interval was 0.5 ms. Analysis of the signals was done off-line with the use of a custom-made data analysis program based on Matlab (The MathWorks, Natick, MA).17 For signal averaging, 1-minute long recordings were made with dedicated software (Biosemi).

Results MAP recording Site of stimulation close to the exploring electrodes Figure 1 shows the standard configuration of unipolar recordings from the multiterminal grid electrode positioned on the LV epicardium of a pig heart. Recordings were made with a stainless steel reference electrode connected to the aortic root. The electrogram (lower panel) is recorded at one of the exploring electrode terminals (A) of the multielectrode and consists of an RS complex followed by a T wave. A small stimulus artifact (stim) precedes the RS complex. At all available electrode positions the Laplacian electro-

+

LAD

26

RV LV 24 22

A stim

T-wave

A

20 0.6 mm

5 mV

RS 100 msec

Figure 1 Upper left panel: Experimental set-up. The heart was paced at a basic cycle length of 600 ms at a site (marker) close to the multielectrode (rectangle on LV). Lower left panel: One of the electrograms (A) recorded with the multielectrode. Electrograms were recorded with respect to the stainless steel needle attached to the aortic root. Right panel: Activation map derived from the recorded electrograms. Numbers are activation times in milliseconds with respect to the stimulus. Lines are isochrones. LA ⫽ left atrium; LAD ⫽ left anterior descending artery; LV ⫽ left ventricle; RA ⫽ right atrium; RV ⫽ right ventricle; stim ⫽ stimulus artifact.

gram was calculated, which defined local activation time. The right panel in Figure 1 shows the activation pattern during stimulation near the left lower corner of the multielectrode. Activation progresses from the lower left corner of the electrode grid (20 ms after the stimulus artifact) toward the top right corner in about 6 ms. Recordings of the experiment in Figure 2 (same heart as in Figure 1) are made in a similar way, but this time the reference electrode is a KCl electrode attached to the base of the LV, approximately 5 cm away from the recording electrode. The lower tracing shows the electrogram recorded from the same terminal as in Figure 1A. The signal comprises three deflections. The first one (stim) is the stimulus artifact. The second one, marked RS, is the depolarization generated by propagation of the electrical impulse underneath the electrode. This is evidenced by the activation map in the lower left panel, which is virtually identical to the map in Figure 1. The third deflection, marked MAP, is a remote component generated at the reference electrode. The RS complex is detached from and precedes the MAP deflections because stimulation is done close to the recording electrode. Activation arrives later at the site of the reference electrode, which generates the MAP. The remote character of this deflection is evidenced by the Laplacian tracing (Lapl), which only reveals a deflection at the time of the intrinsic deflection of the RS component. The lower right map gives the times of the maximal negative downslope (horizontal arrow in the lower tracing) of all MAP deflections (the upstroke of the inverted signal). These times are the same at all sites supporting the remote character of this deflection. In Figure 3, the electrogram recorded with the reference electrode connected to the aortic root (upper tracing, different heart from Figures 1 and 2) is subtracted from the

1046

Heart Rhythm, Vol 3, No 9, September 2006

Lapl KCl reference electrode

LA RA

eg ref aorta

A

RS

+

LAD

stim

RV

MAP 5 mV 100 msec

LV

MAP

RS

eg ref KCl base 28 26 all 69 24 22

20

A

A

0.6 mm

Figure 2 Upper left panel: Experimental set-up. The heart (the same heart as in Figure 1) was paced at a basic cycle length of 600 ms at a site (marker) close to the multielectrode (rectangle on LV). Electrograms were recorded with respect to the KCl reference electrode placed at the base of LV. Upper right panel: The tracing marked A is an electrogram recorded at one of the electrode terminals of the multielectrode. The tracing marked Lapl is the Laplacing signal at the same site. Lower left panel: Activation map derived from the Laplacian signals to avoid effects of remote activation. Lower right panel: Map of the time of steepest negative dV/dt of the MAP component. Numbers are activation times in milliseconds with respect to the stimulus in both maps. MAP ⫽ monophasic action potential. Abbreviations as in Figure 1. Note that the monophasic component of the signal does not propagate and does not correspond to Laplacian signal, which is indicative of a remote activity.

electrogram recorded at the same site, with the KCl reference electrode at the base of LV (middle tracing). This yields an electrogram (lower tracing) that reflects the signal generated at the KCl-reference electrode referenced to the aortic root potential. Note that the downslope of the local T wave (top panel) coincides with the remote repolarization (bottom panel, line).

eg ref aorta _ eg ref KCl base

5 mV 100 ms

stim

Figure 3 Extracellular electrogram recorded at one electrode terminal of the exploring multielectrode with respect to the stainless steel reference electrode attached to the aortic root (upper tracing) and with the KCl reference electrode placed at the base of the LV (middle tracing). The lower tracing shows the subtraction of the two signals in the upper panel, which represent the electrogram of the KCl electrode referenced to the aortic root potential. Only the MAP remains in the subtracted signal.

ulus artifact is not visible here because the stimulation site is far away). The “initial” deflection now reflects both the inverted MAP signal and the local RS complex. In contrast

Stainless steel reference electrode

KCl reference electrode Lapl

RA

Site of stimulation remote from the exploring electrode The lower panels in Figure 4 show activation maps recorded with the stainless steel reference (left panel) and KCl reference electrode (right panel) attached to the aortic root (same heart as in Figure 3). The activation maps as well as the electrograms are identical, indicating that the type of reference electrode, when attached to nonexcitable tissue, is irrelevant for the local electrograms recorded at the exploring electrode. Figures 1–3 have demonstrated that the RS complex of a local electrogram becomes detached from the remote monophasic signal when activation progresses from the exploring to the reference electrode site. Bringing the KCL reference electrode into contact with the base of the heart near the site of stimulation (Figure 5, left panel, same heart as in Figure 4) changes the configuration of the signals. As in Figure 2, there are two interfering components (the stim-

RV

Lapl

A

LA

A

LA RA

+ LAD

RV

+ LAD

10 mV LV

LV

200 msec

60

60

A

63

63

A

66

66 69

69

0.6 mm

Figure 4 Lower panels show activation maps derived from signals recorded with the reference electrode (stainless steel or KCl) connected to the aortic root (upper panel). Numbers in the maps are activation times with respect to the stimulus. Lines are isochrones. Tracings marked A are unipolar electrograms recorded at the sites marked by the circles in the activation map. Note that the signal morphology is independent of the type of reference electrode when attached to nonexcitable tissue. Tracings marked Lapl are the calculated Laplacian signals at the same sites. Abbreviations as in Figure 1.

Coronel et al

Measures of Ventricular Repolarization

1047

KCl reference electrode at LA

KCl reference electrode at LV Lapl

Lapl RS

RS LA RA RV

RS

RS

LA

A

+ LAD

RA

A

+ RV

MAP

LV

LAD

LV

10 mV 200 msec

69

virtually constant at about 38°C. Lowering the temperature in the CX region changed the T wave in the electrogram from positive to negative, while the MAPs lengthened. In the LAD region, the configuration of both the electrograms and the MAPs remained constant. Local repolarization time in local electrograms with a positive T wave is therefore associated with the positive slope of the T wave.

69

A

A

72

72

75

A.

75

78

39.6 oC

78

32.9 oC

0.6 mm

Figure 5 Lower panels show activation maps derived from the signals recorded with the KCl reference electrode attached to the aortic root (lower left panel) or to the left atrium (lower right panel). Numbers in the maps are activation times with respect to the stimulus. Lines are isochrones. Tracings marked A are unipolar electrograms recorded at the sites marked by circles in the activation maps. Lapl ⫽ calculated Laplacian signals at the same sites. Left panel: Stimulation was done at a site remote from the different electrode. Note that signal A is composed of interfering signals from both electrode sites, resulting in an inverted MAP-like signal. The activation pattern constructed from the Laplacian electrograms remains unchanged compared with Figure 4. Right panel: In the recordings made with the reference KCl electrode placed on the left atrium, AV dissociation was present. Thus, complexes from the reference and exploring electrodes interfere but are not related. Local activations are indicated by RS and correspond with the Lapl signals in the upper tracing. Signals recorded at the reference electrode reflect MAPs of the atrium and are indicated by arrows.

B. CX

LAD

Figure 6A shows unipolar electrograms and MAPs recorded from the CX regions during the period when the temperature in the CX region was varied from 39.6°C to 32.9°C, while the temperature in the LAD perfused area remained

38.5 38.3 38.1 37.9

oC

oC oC oC

oC oC oC

10 mV

MAP-duration (msec) o

ARI measurements

oC

200 ms

C. to Figure 2, the RS complex overlaps with the initially negative deflection of the inverted MAP because this time stimulation is done close to the reference electrode. Since the slope of the initial deflection of the inverted MAP is shallow (compare Figure 3, lower tracing), the steep negative deflection is the reflection of the superimposed local RS complex. The Laplacian signal (Lapl) shows that the local component is at the RS deflection. The activation map derived from the Laplacian signals (lower panel) is similar to that in Figure 4. Finally, the KCl reference electrode was moved to the left atrial epicardial wall (Figure 5, right panels, same heart as in Figures 1– 4) during AV block. The activation map derived from the Laplacian signals remains virtually the same. In the lower tracing marked A, deflections generated by activation underneath one of the exploring electrodes are indicated by RS. Arrows mark the inverted atrial MAPs generated at the site of the reference electrode. The third MAP is clearly visible; the first one partly overlaps with the T wave, the second one with RS. The experiments were repeated in three hearts and essentially demonstrated the same results.

39.6 38.0 35.7 32.9

o

300 40

o

38

o

35

30

280 260 240 220 2

R = 0.7947

200 180 180

200

220

240

260

280 300 ARI (msec)

Figure 6 A: Superimposed MAPs and electrograms from the same site in the CX-perfused area at 32.9°C and 39.6°C. The minimum dV/dt of the RS complexes is aligned with the upstroke of the MAP. The repolarization phase of the MAP corresponds in both cases with the upstroke of the local T wave. In addition, note that the local T wave is positive at the higher but negative at the lower temperature. B: Superimposed electrograms recorded from the CX-perfused electrode site (upper left) with the corresponding MAPs recorded from the same site (upper right) during the transition from hot to cold perfusion. Note the change in the polarity of the T wave, corresponding to the changes in MAP duration. MAP amplitudes were adjusted to allow comparison. Note that the negative component of the local T wave remains at the same position and indicates repolarization of the remote myocardium where temperature did not change. The positive flank of the T wave moves together with the repolarization phase of the MAP and represents local repolarization. Lower panel: The corresponding signals derived from the LAD perfused area, where temperature changed only 0.6°C. C: Correlation between ARI (repolarization time measured at the maximum dV/dt of the T wave) and MAP duration (repolarization time measured at the minimum dV/dt) in three experiments at various myocardial temperatures (approximately indicated by arrows). Dotted line is line of identity.

1048

Heart Rhythm, Vol 3, No 9, September 2006

Discussion MAP recordings Lapl

5 mV

eg

5 mV

100 msec

Figure 7 Local electrogram (eg) and Laplacian signal (Lapl) recorded from the same site. Signal averaging was applied from about 100 subsequent complexes from all signals recorded from a grid electrode during ventricular pacing. The Laplacian was calculated from the central electrode and four (north, east, south, west) neighboring electrodes at distances of five times the grid constant (2 mm). Note that the Laplacian signal occurs only before the top of the T wave (line), indicating that the end of the T wave is caused by repolarization at remote sites.

In Figure 6B, electrograms or MAPs are superimposed at different temperatures. At the site where temperature was altered, a change in the time of the positive slope of the T wave occurred, whereas that of the negative slope remained unaltered. This led to inversion of the T wave in the course of temperature change. In Figure 6C, data from all three experiments are shown as a plot of ARI against MAP duration. MAP duration was measured as the interval between action potential upstroke and the steepest part of phase 3. ARI was measured as the interval between dV/dt minimum of the RS complex and the dV/dt maximum of the T wave (irrespective of the polarity of the T wave). In all instances, repolarization of the MAP coincided with positive slope of the T wave. In the range of temperatures, the data points are close to the line of identity. In all experiments, the T wave morphology changed in the CX-perfused area (which was subject to temperature change) commensurate with the change in MAP duration, but not in the LAD perfused area. The Laplacian signal (Lapl in Figure 7) was calculated from a grid of electrodes (2 mm interelectrode distance) at a site where the unipolar electrogram showed a positive T wave (e.g., in Figure 7). For calculation of the Laplacian, all neighboring local electrograms were signal averaged during about 100 subsequent beats after ventricular pacing and the Laplacian was calculated from the electrograms at the central electrode and four neighboring electrodes at distances of five times the interelectrode distance. The unipolar electrogram in Figure 7 shows several remote components (stimulus artifact, retrograde P wave) that are not represented in the Laplacian signal. During repolarization, the Laplacian signal occurs entirely before the peak of the T wave, indicating that the end of the T wave in the unipolar electrogram also represents remote activity.

A recording of a MAP is a bipolar recording.6,11 Signals are recorded at both the exploring and the reference electrode, and the question of which of the two is the recording electrode has led to controversy.9,10 This is particularly relevant when the two electrodes are placed at a distance from each other, a method that has been proposed as an alternative for generating MAPs.9 A controversy exists about the site of “origin” of the monophasic complex in a MAP recording. Some investigators state that it is generated at the nondepolarized area if a KCl electrode is used as a “reference” electrode connected to the positive input of the amplifier,9 allowing intramural MAP recordings.18 This is in contrast to the ideas of Franz et al6 and Tranquillo et al.19 The first part of our study shows that the MAP recorded in the way proposed by Kondo et al9 is in fact a superimposition of a local electrogram (at the site of the exploring electrode) and a remote component (at the site of the reference electrode). Inversion of the signal would indeed give rise to a MAP-like complex, of which the monophasic component is generated at the KCl electrode. Laplacian electrograms, a reflection of local transmembrane current,16 allowed us to discriminate between local and remote components of the signals. Furthermore, when the heart is stimulated from a site close to the exploring electrode (Figures 2 and 3) the components become separated in time. The MAP component does not show an activation pattern and is remote, whereas the RS component does show an activation pattern and is local. Placing the reference KCl electrode on the atrium of the heart with total AV block uncovers the dissociation between the two components and now shows inverted MAPs of short duration compatible with atrial action potentials. Figure 8 schematically depicts how the MAP is generated in accordance with our experiments. Rectangles represent myocardial cells. The right cell is located in normal myocardium and is connected to cells that are partly depolarized (to ⫺40 mV) because of the local high extracellular potassium concentration. The upper panel shows current flow (arrows) and the extracellular potential (tracing) when normal myocardium is at rest (⫺90 mV). Intercellular current flow is from the partly depolarized cell in the high potassium area toward the normal cell. In the extracellular space, current is flowing in the opposite direction, causing a sink (⫺) at the high K⫹ site. Because of extracellular resistance, the extracellular current flow results in a negative extracellular potential at that site when DC recordings are made. The lower panel shows the currents and potentials when an activation front arrives at the border of the normal and high potassium area. Cells at the normal site are depolarized, which increases their intracellular potential to about ⫹20 mV. Cells at the high K⫹ site can, however, not be activated and the intracellular potential remains at ⫺40 mV.

Coronel et al

Measures of Ventricular Repolarization

diastole

ARI measurements

0 mV extracellular potential +

-

High K+ area

- 40 mV

+ -90 mV

systole

Normal area extracellular potential

0 mV +

+ - 40 mV

-

1049

-

+20 mV

Normal area

High K+ area

0 mV

Figure 8 Schematic drawing of intercellular and extracellular current flow and the generated extracellular potentials at a site of high K⫹. Rectangles represent myocardial cells. See text for discussion.

The potential difference now results in intercellular current flow from the normal toward the high K⫹ cells. In the extracellular space, the current flows in the opposite direction, from the high K⫹ area toward the normal zone, causing a current source (⫹) at the high K⫹ area. This current generates a positive potential at that site (tracings in the middle). Because the intracellular potential of the normal cell follows the action potential, the extracellular current and hence the extracellular potential at the high K⫹ site will manifest a configuration similar to the action potential. Thus, a MAP arises at the high K⫹ area. Because KCl injection is not feasible in studies on humans, other means of depolarization have been used. Pressure applied by the tip of a Franz electrode has the same effect, as does suction applied on tissue surrounding the exploring electrode.5,6 Figure 3 has demonstrated how the local T wave interferes with the end of the monophasic potential. The influence of the local T wave increases when the monophasic potential is of low amplitude. Kondo et al9 injected KCl into the myocardium to generate MAPs, causing localized edema. A concomitant reduction of the extracellular resistance leads to a decrease of the local extracellular potentials.20 Kondo et al do not mention MAP amplitudes, and it cannot be excluded that their recordings are influenced by this factor. MAPs recorded with their technique may thus show small regional repolarization changes that are, however, caused by changes in the T wave of the electrogram. This does not necessarily point to changes in local repolarization time because that time is related to the upstroke of the T wave. MAPs should, therefore, be recorded with the two electrodes positioned as closely as possible, as pointed out by Franz.10

Yue et al14 have suggested that the time of repolarization in a local unipolar electrogram with a positive T wave is at the maximum downslope of the T wave. They based their statement on a somewhat higher correlation between MAP duration and ARI measurements with the alternative method than the standard method. Any point on the T wave correlates with repolarization, and it is not clear whether the two correlation coefficients were statistically different. We have demonstrated that a positive T wave can be converted to a negative T wave recorded at the same site, when action potential duration at that site is locally prolonged by cooling. Furthermore, we show that, after signal averaging of about 100 T waves, a maximum Laplacian signal indicative of local repolarizing current flow can be detected at the upstroke and not at the downstroke of a positive T wave. Finally, we demonstrate a correlation between MAP duration (recorded in the standard manner) and ARI close to the line of identity. The small deviations may be explained by inaccuracies in the temperature readings, which were derived from a needle inserted at some distance from the recording site to avoid lesion potentials. It may be argued that our method of creating regional repolarization differences produces larger differences than occur under clinical conditions. However, even with relatively small differences in repolarization (caused by small temperature changes around 37°C; see Figure 6) the correlation was close to the line of identity. Because the recording electrodes in the myocardium subject to heating or cooling were remote from the border between the two perfusion zones, electrotonic interference between the two zones is not probable. Moreover, ARI measurements correlated with MAP recordings, which are bipolar in nature and devoid of remote influences. The T wave also represents (transmural) heterogeneity of repolarization.21 Because we did not measure transmural heterogeneity we cannot address this issue. Figure 6B suggests that the moment of repolarization of the isothermic region (LAD) corresponds to the downstroke of the T wave in the CX area, again suggesting that the downslope of the T wave represents remote repolarization. Also, the T wave was almost abolished at a temperature between 35.7°C and 38°C. This indicates that in our model the contribution of macroheterogeneity in repolarization is predominant over transmural heterogeneity in the genesis of the T wave of the local electrogram. To address the controversy on the interpretation of a positive T wave, we deliberately studied sites where the T wave was positive under normal conditions. The question of how an initially negative T wave should be interpreted is therefore not answered by these experiments. However, Figures 6B and 6C demonstrate that after regional cooling a negative T wave developed. Here the upstroke of the T wave also corresponded to the repolarization phase of the MAP.

1050 In conclusion, we present experimental evidence showing that the monophasic component of a MAP is recorded at the depolarizing electrode and that the ARI of an electrogram with a positive T wave, as of those with a negative T wave, should be determined at its maximum dV/dt.

References 1. Chauhan VS, Downar E, Nanthakumar K, Parker JD, Ross HJ, Chan W, Picton P. Increased ventricular repolarization heterogeneity in patients with ventricular arrhythmia vulnerability and cardiomyopathy: a human in vivo study. Am J Physiol Heart Circ Physiol 2006;290:H79 –H86. 2. Mines GR. On circulating excitations in heart muscles and their possible relation to tachycardia and fibrillation. Trans Royal Soc Canada 1914;4:43–53. 3. Verkerk AO, Tan HL, Kirkels JH, Ravesloot JH. Role of Ca2⫹-activated Cl⫺ current during proarrhythmic early afterdepolarizations in sheep and human ventricular myocytes. Acta Physiol Scand 2003;179:143–148. 4. Hoffman BF, Cranefield PF, Lepeschkin E, Surawicz B, Herrlich HC. Comparison of cardiac monophasic action potentials recorded by intracellular and suction electrodes. Am J Physiol 1959;196:1297–1301. 5. Olsson BS. Monophasic Action Potentials of Right Heart: Suction Electrode Method in Clinical Investigations. Göteborg: Elanders Bocktryckeri, 1971. 6. Franz MR. Long-term recording of monophasic action potentials from human endocardium. Am J Cardiol 1983;51:1629 –1634. 7. Haws CW, Lux RL. Correlation between in vivo transmembrane action potential duration and activation-recovery intervals from electrograms. Effects of interventions that alter repolarization time. Circulation 1990;81:281–288. 8. Steinhaus BM. Estimating cardiac transmembrane activation and recovery times from unipolar and bipolar extracellular electrograms: a simulation study. Circ Res 1989;64:462. 9. Kondo M, Nesterenko VV, Antzelevitch C. Cellular basis for the monophasic action potential. Which is the recording electrode? Cardiovasc Res 2004;63: 635– 644. 10. Franz MR. What is a monophasic action potential recorded by the Franz contact electrode? Cardiovasc Res 2005;65:940 –941.

Heart Rhythm, Vol 3, No 9, September 2006 11. Vigmond EJ. The electrophysiologic basis of MAP recordings. Cardiovasc Res 2005;68:502–503. 12. Chen P-S, Moser KM, Dembitsky WP, Auger WR, Daily PO, Calisi CM, Jamieson SW, Feld GK. Epicardial activation and repolarization in patients with right ventricular hypertrophy. Circulation 1991;83:104 –118. 13. Gepstein L, Hayam G, Ben Haim SA. Activation-repolarization coupling in the normal swine myocardium. Circulation 1997;96:4036 – 4043. 14. Yue AM, Paisey JR, Robinson S, Betts TR, Roberts PR, Morgan JM. Determination of human ventricular repolarization by noncontact mapping. Validation with monophasic action potential recordings. Circulation 2004;110:1343–1350. 15. Coronel R, Fiolet JWT, Wilms-Schopman FJG, Schaapherder AFM, Johnson TA, Gettes LS, Janse MJ. Distribution of extracellular potassium and its relation to electrophysiologic changes during acute myocardial ischemia in the isolated perfused porcine heart. Circulation 1988;77:1125–1138. 16. Coronel R, Wilms-Schopman FJG, de Groot JR, Janse MJ, Van Capelle FJL, De Bakker JMT. Laplacian electrograms and the interpretation of complex ventricular activation patterns during ventricular fibrillation. J Cardiovasc Electrophysiol 2000;1119 –1128. 17. Potse M, Linnenbank AC, Grimbergen CA. Software design for analysis of multichannel intracardiac and body surface electrocardiograms. Comp Meth Prog Biomed 2002;69:225–236. 18. Weissenburger J, Nesterenko VV, Antzelevitch C. Transmural heterogeneity of ventricular repolarization under baseline and long QT conditions in the canine heart in vivo: torsades de pointes develops with halothane but not pentobarbital anesthesia. J Cardiovasc Electrophysiol 2000;11:290 –304. 19. Tranquillo JV, Franz MR, Knollmann BC, Henriquez AP, Taylor DA, Henriquez CS. Genesis of the monophasic action potential: role of interstitial resistance and boundary gradients. Am J Physiol 2004;286:H1370 –H1381. 20. Fleischhauer J, Lehmann L, Kléber AG. Electrical resistance of interstitial and microvascular space as determinants of the extracellular electrical field and velocity of propagation in ventricular myocardium. Circulation 1995;92:587– 594. 21. Antzelevitch C, Yan GX, Shimizu W. Transmural dispersion of repolarization and arrhythmogenicity: the Brugada syndrome versus the long QT syndrome. J Electrocardiol 1999;32:158 –165.