OiWW

F

B(dV

Osoh

M3wolV

Ripf

IOimtmonpcaecica

6Vsm5

1

ptical measurements of intramural action potentials insolated porcine hearts using optrodesei Kong, PhD, Nadia Fakhari, MS, Oleg F. Sharifov, MD, PhD, Raymond E. Ideker, MD, PhD,illiam M. Smith, PhD, Vladimir G. Fast, PhD

rom the Department of Biomedical Engineering, University of Alabama at Birmingham, Birmingham, Alabama.

cmNlAVbo

Citrse

KV

(

ACKGROUND Measurements of intramural membrane potentialVm) would greatly increase knowledge of cardiac arrhythmias andefibrillation. Optrodes offer the possibility for three-dimensionalm mapping, but their signal quality has been inadequate.

BJECTIVE The purpose of this work was to improve optrodeignal quality and use optrodes to measure intramural distributionf action potentials and shock-induced Vm changes in porcineearts.

ETHODS Optrodes were made from seven optical fibers 225 or25 �m in diameter. Fiber ends were polished at a 45° angle,hich improved light collection and allowed their insertion with-ut a needle. Fluorescent measurements were performed in iso-ated porcine hearts perfused with Tyrode’s solution or blood usingm-sensitive dye RH-237 and a 200-W Hg/Xe lamp.

ESULTS The signal-to-noise ratio for 325-�m fibers was 44 � 23n blood-perfused hearts (n � 5) and 106 � 45 in Tyrode’s-erfused hearts (n � 3), which represents an approximately

tlsctdsbnliltmagmc

wto, 2007; accepted July 1, 2007.)

547-5271/$ -see front matter © 2007 Heart Rhythm Society. All rights reserved

lose correspondence between optical and electrical measure-ents of activation times and action potential duration (APD).o significant intramural APD gradients were observed at cycle

engths up to 4 s and in the presence of dofetilide or d-sotalol.pplication of shocks (5–50 V/cm) produced large intramuralm changes (up to �200% action potential amplitude), possi-ly reflecting a combined effect of tissue discontinuities andptrode geometry.

ONCLUSIONS A substantial improvement of optrode signal qual-ty was achieved. Optical measurements of APD and activationimes matched electrical measurements. Optrode measurementsevealed no significant intramural APD gradients. Application ofhocks caused large intramural Vm changes that could be influ-nced by the optrode geometry.

EYWORDS Cardiac excitation; Action potentials; Defibrillation;irtual electrodes; Optical mapping; Fiber optics

Heart Rhythm 2007;4:1430–1436) © 2007 Heart Rhythm Society.

our-fold improvement over previously reported data. There was All rights reserved.

ntroductionptical mapping of electrical activation has become an

mportant research tool in cardiac electrophysiology.1 Thisethod employs fluorescent dyes and arrays of photodetec-

ors to measure changes in membrane potential (Vm) atultiple locations on the heart surface. Optical mapping

ffers several advantages over electrical mapping tech-iques, including the ability to measure the repolarizationhase of the action potential (AP) as well as the Vm changesaused by electrical shocks. In addition, optical mappingllows for simultaneous measurements of Vm and otherlectrophysiological parameters such as intracellular cal-ium concentration.2–5 Optical mapping has been employedn various experimental preparations ranging from singleells to whole hearts to study mechanisms of impulse prop-gation,6,7 cardiac arrhythmias,8–14 and defibrillation.15,16

This work was supported by National Institutes of Health grant nos. HL7748 and HL28429. Address reprint requests and correspondence:ladimir G. Fast, Ph.D., Department of Biomedical Engineering, Univer-

ity of Alabama at Birmingham, 1670 University Blvd., VH B126, Bir-ingham, AL, 35294. E-mail address: fast@crml.uab.edu. (Received May

Although application of optical mapping in cardiac elec-rophysiology has been very successful, it has importantimitations. Unlike electrical mapping, which allows mea-urements from both the heart surface and intramural mus-le layers,17–19 optical mapping was until recently restrictedo the heart surface. However, information on the three-imensional Vm dynamics is critically important for under-tanding the mechanisms of cardiac arrhythmias and defi-rillation. Thus, electrical mapping studies using plungeeedle electrodes demonstrated that the source of ventricu-ar fibrillation may be intramural.20 Microelectrode record-ngs and optical mapping in wedge preparations revealedarge transmural heterogeneities of action potential dura-ion (APD) that may play an important role in arrhyth-ogenesis.21,22 Intramural shock-induced Vm changes

re believed to be critical for defibrillation.16,23 Investi-ation of these phenomena requires three-dimensionalapping of Vm in intact hearts, but no such methods are

urrently available.To circumvent this limitation, a new optical technique

as introduced recently,24,25 which employs fiber-opticechnology. In this method, bundles of optical fibers (called

ptrodes) are used to deliver excitation light to intramural

. doi:10.1016/j.hrthm.2007.07.002

tism(oolnomwhhogrcas

MOTmdiecdoIl

avifBaT1adCaIt

OOaTeessaidwtmwellvaap1sta

msfiowjrd

maaa

FePtirfio

1431Wei Kong et al Measuring APs with Optrodes in Pigs

issue layers and to collect the emitted fluorescence, whichs measured with arrays of photodiodes. The initial optrodetudies demonstrated the feasibility of this approach for Vm

easurements at multiple sites across the left ventricularLV) wall. These studies also exposed several limitations ofptrode recordings that hinder the extension of this technol-gy to fully three-dimensional mapping. One importantimitation of current optrodes is a relatively low signal-to-oise ratio, which is significantly smaller than in surfaceptical mapping. Such low signal quality precludes intra-ural optical measurements in hearts perfused by blood,hich is preferable in long-lasting experiments on isolatedearts.9 Therefore, the main goal of this work was to en-ance the quality of optrode recordings by improving theptrode design and optimizing the optical system. The otheroals were to use the new optrodes to characterize intramu-al distribution of APs in the LV of porcine hearts, toompare optical recordings with electrical measurements,nd to measure intramural Vm changes caused by electricalhocks.

ethodsptical setuphe optical setup for intramural optrode recordings, sche-atically shown in Figure 1A, was similar in its overall

esign to the one described elsewhere,25 with several mod-fications. A stronger 200-W Hg/Xe lamp was used as anxcitation light source, which resulted in increased fluores-ence. The excitation light was filtered at 530–585 nm,eflected by a dichroic mirror (�600 nm), and focused onptrode entrance by an objective lens (5� Fluar, Carl Zeissnc., Thornwood, NY). The emitted fluorescence was col-

igure 1 A: Schematic diagram of the optical system. HF: heat filter; S:lectromechanical shutter; ExF: excitation filter; DM: dichroic mirror;DA: 16 � 16 PDA. B: View of the cross section of an optrode and its

issue end. Fibers are arranged hexagonally with ends spaced at 2 or 3 mmntervals. C: View of the optrode detector end. Blank circles in the middleepresent seven measurement fibers; the two shaded circles depict thebers used for alignment. This linear fiber array is projected onto a columnf nine photodiodes in the PDA.

ected by the same lens, passed through the dichroic mirror, m

nd filtered at �650 nm. The fluorescent light was con-erted to voltages using a 16 � 16 photodiode array (PDA)ntegrated with current-to-voltage converters with 100-M�eedback resistors (model C4675-102, Hamamatsu Corp,ridgewater, NJ). Electrical signals were conditioned using256-channel set of second-stage amplifiers (Innovative

echnologies Inc., Germantown, MD) and digitized at a2-bit resolution and a sampling interval of 0.512 ms usingdata acquisition system that consisted of four 64-channel

ata acquisition cards (PCI-6071E, National Instrumentsorp., Austin, TX) installed in a personal computer. Andditional eight-channel digital card (PCI-6602, Nationalnstruments) was used to control stimulation, light excita-ion, shock application, and change of amplifier parameters.

ptrodesptrodes were made from silicon fibers with an outer di-

meter of 225 or 325 �m (FT-200-URT and FT-300-UMT,horlabs Inc., Newtown, NJ). These fibers have core diam-ters of 200 and 300 �m, respectively, and numerical ap-rtures of 0.50 and 0.39, respectively. Each optrode con-isted of nine fibers. Seven fibers were used to record opticalignals, and two other fibers were used for optrode-PDAlignment. At one end (tissue end), seven fibers were pol-shed at approximately 45° to the fiber axis. This wasifferent from previous work,25 in which flat-cleaved fibersere used. Polishing fibers resulted in a closer contact with

he tissue and better light collection. In addition, optrodesade from polished fibers could be introduced into the heartithout the aid of an insertion needle, which reduced the

xtent of tissue damage. This made possible the use of thearger 325-�m fibers, which resulted in much higher signalevels with less tissue damage in comparison with the pre-ious study.25 Fibers were arranged in a hexagonal patterns shown in Figure 1B with fiber ends spaced 2 or 3 mmpart. At the other end (detector end), fibers were polishederpendicular to the fiber axis and arranged in a line (FigureC). Two additional alignment fibers were prepared in theame fashion but with both ends polished perpendicular tohe fiber axis and placed on both sides of the linear fiberrray.

The optrode detector end and the PDA were mounted onicropositioners. Their mutual alignment was performed by

hining light from a light-emitting diode at the alignmentbers and measuring optical signals at corresponding diodesf the PDA. The optrode-lens and the lens-PDA distances asell as lateral positions of all optical elements were ad-

usted until the maximal cross talk between the diodeseceiving light from the alignment fibers and any otheriode in their immediate vicinity became less than 3%.

The quality of fiber preparation was examined by im-ersing an optrode into a 6-�mol/L solution of rhodamine

nd measuring its fluorescence. The fibers were consideredcceptable if the differences in signal magnitudes betweenll seven fibers were less than 30% of the maximum signal

agnitude.

PTUI1aawwmNtathcaftamwwL

HmRtvptst

o1maTdf2l

OApdotdum(At

ag(dk

enmcsmtwc

ROF

FmTetht

1432 Heart Rhythm, Vol 4, No 11, November 2007

orcine heart preparationhe investigation conformed to the Guide for the Care andse of Laboratory Animals published by the U.S. National

nstitutes of Health (NIH Publication No. 85-23, revised996). Pigs with a weight of �25 kg of either sex werenesthetized with telazol (4.4 mg/kg), xylazine (4.4 mg/kg),nd antropine (0.04 mg/kg). Anesthesia was maintainedith inhalation of isoflurane (1.3–2.5%) in oxygen. If bloodas used for heart perfusion, a warm Tyrode’s solution (inmol/L, NaCl 128.5, glucose 20, KCl 4.7, MgCl2 0.7,aH2PO4 0.5, CaCl2 1.5, and NaHCO3 28) was infused in

he carotid vein, and the blood was drained from the carotidrtery. Heart contraction was stopped by KCl injection or byhe application of a 9-V battery to induce fibrillation. Theeart was quickly removed from the chest and placed in aold physiological solution. After weighing the heart, theorta was cannulated, and the heart was retrogradely per-used with either Tyrode’s solution or blood. The perfusateemperature was maintained at 37°C and was bubbled with

mixture of 95% O2 and 5% CO2. The flow rate wasaintained at approximately 1 mL/min per 1 g of hearteight. At the end of the experiments, the heart waseighted again to estimate the degree of tissue edema. TheV wall was dissected to determine the wall thickness.

To enable external pacing, the sinus node was removed.earts were paced using a bipolar electrode and a custom-ade stimulus generator at a cycle length (CL) of 0.5 s.ectangular shocks with duration of 10 ms were applied via

wo large mesh electrodes sutured to the LV and rightentricular (RV) epicardium using a custom-made currentulse generator. An optrode was inserted through a whole athe center of the mesh LV electrode. Intramural shocktrength was measured using a bipolar plunge needle elec-rode inserted approximately 5 mm away from the optrode.

For optical measurements, hearts were stained with flu-rescent Vm-sensitive dye RH-237. The dye was diluted in0 mL of Tyrode’s solution at a concentration of 0.25mol/L. This solution was injected into a perfusion line atrate of 3.3 mL/min for 3 minutes using a syringe pump.he estimated dye concentration in the perfusion solutionuring injection was 5.5 �mol/L. To prevent motion arti-acts in optical recordings, excitation-contraction uncoupler,3-butanedione monoxime was added to the perfusion so-ution at a concentration of 15–25 mmol/L.

ptrode recordingsn optrode was inserted into the free anterior LV wallerpendicular to the epicardium. Recording sites were at aepth of 2–14 mm (2-mm optrode) or 3–21 mm (3-mmptrode). If necessary, optical signals were low-pass filteredo reduce noise. AP amplitude (APA) was determined as aifference in optical levels measured before and after anpstroke. Activation times (ATo) were measured at a mo-ent of 50% of AP upstroke. The signal-to-noise ratio

SNrms) was calculated from unfiltered signals as the ratio ofPA and root-mean-square noise measured over a 20-ms

ime interval before the AP upstroke. o

The APD was measured as a time interval between the ATnd 80% level of repolarization (APD80). Transmural APDradients were measured in control conditions, at various CLs0.5–4 s) and in the presence of potassium channel blockers-sotalol (125 �mol/L) or dofetilide (2 �mol/L) that arenown to augment intramural APD heterogeneity.26,27

Optical recordings were compared with extracellularlectrical recordings from a plunge needle electrode insertedear an optrode. Electrical activation times (ATe) wereeasured at the moment of minimum derivative of the QRS

omplex. Activation-recovery intervals (ARIs) were mea-ured as the time difference between ATe and the time ofaximum derivative of the T wave.28 The correlations be-

ween ATo and ATe as well as between APD80 and ARIere calculated using linear regression. Shock-induced Vm

hanges were measured as previously described.16

esultsptrode signal qualityigure 2 shows intramural APs recorded using 325-�m fiber

igure 2 Optrode recordings of intramural APs. A: Intramural APseasured with a 325-�m fiber optrode in a porcine heart perfused withyrode solution. Numbers 1–6 designate recordings obtained from sub-picardium through subendocardium. One fiber was outside the muscle inhe cavity (not shown). B: Intramural APs measured in a blood-perfusedeart. C: Reduction of the average signal-to-noise ratio (SNrms) during theime of experiment in one heart.

ptrodes in porcine hearts perfused either by Tyrode’s so-

lrts1htsrwf(

wSs1S

eswitihe11

CEabAdon

eowf1poo

IIscThvot

mdaf

saiTspcaaCr

4(iri

SFcdstbew

ca

Fn(fieb

1433Wei Kong et al Measuring APs with Optrodes in Pigs

ution (panel A) or blood (panel B). Measurements in Ty-ode’s-perfused hearts had substantially higher SNrms thanhe ones obtained in blood-perfused hearts. Recordingshown in panels A and B had SNrms of 154 � 86 and 58 �6, respectively. Average SNrms values obtained from allearts were 106 � 46 for Tyrode perfusion (n � 6 sites inhree hearts) versus 44 � 23 for blood perfusion (n � 12ites in five hearts; P �.001). The highest SNrms in Ty-ode’s-perfused hearts could reach �300. When optrodesith narrower 225-�m fibers were used in Tyrode’s-per-

used hearts (recordings not shown), SNrms was 40 � 8n � 7 sites in earth hearts).

Signal quality decreased over time, likely because of dyeashout and photobleaching. Panel C illustrates changes inNrms measured in a Tyrode’s-perfused heart. It shows thatignal quality decreased by approximately 40% during a-hour time period. In four hearts, the average decrease ofNrms over 1 hour was 33% � 7%.

APs sometimes exhibited a triangular shape, which isxemplified by intramural AP at site 5 in panel A. Such APhape indicates local tissue ischemia.29 Such AP shapesere observed more often in Tyrode’s-perfused hearts than

n blood-perfused hearts. This could be partially attributedo tissue edema that developed during the course of exper-ment, which was more pronounced in Tyrode’s-perfusedearts than in blood-perfused hearts. Over the duration of anxperiment (�4 hours), the heart weight increased by18% � 34% in Tyrode’s-perfused hearts (n � 3) and by 35 �5% in blood-perfused hearts (n � 3).

omparison of optical and electrical recordingsxtracellular electrical recordings of ARIs have been useds an estimate of intramural APD measurements,28 which isased on the assumption that electrical ATs and ARIs matchTs and APDs obtained from Vm measurements. To vali-ate this assumption, we compared optrode measurementsf intramural Vm with electrical measurements from aearby location.

Figure 3A shows examples of simultaneous optical andlectrical recordings at different pacing CLs. ATs measuredptically and electrically were very similar when the CLas changed between 200 and 1000 ms. The average dif-

erence between ATo and ATe measured at all CLs was.3 � 1.4 ms. Similarly, repolarization times changed inarallel, which was reflected in a high correlation betweenptical APD80 and electrical ARI (R2 � 0.98) upon changesf CL (panel B).

ntramural APD heterogeneityntramural heterogeneity of APD, previously reported intudies on single cells and isolated wedge preparations, isonsidered to be an important factor in arrhythmogenesis.o examine the presence of APD gradients in porcineearts, we measured intramural APD in the LV wall atarious CLs and in the presence of drugs. The APD valuesbtained from different tissue depths were combined in

hree groups: subepicardial group (Sub-Epi), with APD p

easurements obtained at a depth of 2–4 mm; midmyocar-ial group (Mid), with data obtained at a depth of 6–10 mm,nd subendocardial group (Sub-Endo), with measurementsrom a depth of 12–14 mm.

In control conditions and a pacing CL of 500 ms, noignificant APD gradients were detected (Figure 4A). Theverage APD80 was 200 � 17, 199 � 18, and 205 � 28 msn the Sub-Epi, Mid, and Sub-Endo regions, respectively.he differences among all regions were not statisticallyignificant. Increasing the CL up to 4000 ms resulted in arolongation of the average APD80 to 229 � 29 (P �.05,omparison with APD80 at 500-ms CL), 225 � 25 (P �.05),nd 230 � 36 ms (P �.05) in corresponding regions withoutn increase in regional APD80 differences. Such a degree ofL-dependent prolongation of APD is similar to previously

eported data obtained from porcine heart preparations.30

Application of d-sotalol (Figure 4B) or dofetilide (FigureC) caused a statistically significant increase of APD80

P �.05 for both drugs) at all intramural regions but did notncrease intramural APD80 heterogeneities even at slow pacingates. The increase of APD under the influence of these drugss consistent with previous publications.26,27

hock-induced �Vmigure 5A shows examples of intramural Vm changesaused by 10-ms shocks with a strength of 8 V/cm applieduring the AP plateau phase in one heart. Shocks producedignificant polarizations at the majority of intramural loca-ions. The �Vm magnitude was highly variable, rangingetween approximately 5% APA and 100% APA at differ-nt locations. In most cases, the sign of �Vm was reversedith changing the shock polarity.Qualitatively similar findings were obtained in other

ases. Panel B depicts maximal and minimal �Vm measuredt different shock strengths in five preparations. Similar to

igure 3 Comparison of optical and electrical recordings. A: Simulta-eous recordings of optical AP (Vm) and extracellular electrical potentialVe) at different CLs. The optical recordings were made using a 225-�mber; electrical signals were made using a closely adjacent plunge needlelectrode. B: Correlation between APD80 and ARI with variation of CLetween 200 and 1000 ms. The straight line depicts linear regression.

revious �Vm measurements in wedge preparations,16

stfba

DFmfirsu

bac

OOiwSowt2fitotifibpp

emPcceep

3dwSodisr

CEpAsmrmisdii

Fcmdo�

1434 Heart Rhythm, Vol 4, No 11, November 2007

hock-induced polarizations were asymmetric, with nega-ive �Vm exceeding positive �Vm by approximately two-old. Increasing shock strength resulted in an increase ofoth maximal and minimal �Vm that reached magnitudes ofpproximately 100% APA and 200% APA, respectively.

iscussioniber-optic technology allows for multisite intramural Vm

easurements in the heart. Extension of this technology toull three-dimensional mapping depends on improvementsn existing optrode techniques with regard to signal quality,eliability, and ease of use. The present work achieved aignificant improvement in the quality of optrode recordings

igure 4 Intramural APD distribution. A: APD80 measured in subepi-ardial (Sub-Epi), midmyocardial (Mid), and subendocardial (Sub-Endo)uscle layers at pacing CLs of 0.5, 1, and 4 s in control conditions. The

ata were obtained from five hearts perfused with either Tyrode’s solutionr blood. B and C: APD80 in different tissue layers in the presence of 125M of d-sotalol (B) and 2 �M of dofetilide (C).

sing a standard optical design. These optrodes were tested r

y measurements of intramural APD, comparison of opticalnd electrical recordings, and shock-induced �Vm in por-ine hearts.

ptrode signal qualityne of the most important parameters of optical recordings

s the SNrms. In comparison with our previous study,25

hich employed the same overall optical system design, theNrms value was approximately doubled when usingptrodes made from 225-�m fibers and increased four-foldhen using the 325-�m fibers. The gain in SNrms was due

o several factors. One of them was the use of a brighter00-W Hg/Xe lamp instead of a 100-W Hg lamp. Anotheractor was the use of fibers polished at a 45° angle. Thismproved the contact of fiber ends with adjacent tissue and,herefore, light collection. Another advantage of suchptrodes is that they can be introduced into tissue withouthe aid of an insertion needle used previously. This resultedn a lower degree of tissue damage even when 325-�mbers were used. The diameter of such an optrode variedetween 325 and 975 �m along its length, which is com-arable to the diameter of the plunge needle electrodes usedreviously.31

Gains in signal quality allowed optical measurementsven in blood-perfused hearts, in which optical signals areuch weaker due to strong light absorption by hemoglobin.erfusion with Tyrode’s solution is known to cause a slowhange of electrophysiological tissue properties during theourse of experiment,32 which is partially caused by tissuedema. Perfusion with blood significantly reduced tissuedema and thus provided for better preservation of tissueroperties.

SNrms decreased over time at the rate of approximately3% per hour. This is consistent with the rate of signalecay measured using surface optical mapping3 and reflectsashout and photobleaching of the voltage-sensitive dye.ignificantly faster signal deterioration was reported previ-usly in optrode measurements.24 A higher rate of signalecay in that study was probably caused by a much strongerntensity of excitation light and longer light exposures as-ociated with averaging of multiple consecutive beats thatesulted in faster dye photobleaching.

omparison of optical and electrical recordingslectrical mapping with plunge needle electrodes was usedreviously to map intramural activation, repolarization, andRI as a substitute for APD measurements.33,34 This as-

umes a close correspondence between electrical and opticaleasurements, but direct experimental verification of this

elationship is lacking. Furthermore, previous measure-ents using optrodes and plunge electrodes reported signif-

cant differences between electrically and optically mea-ured ATs as well as between APDs and ARIs.35 Suchifferences were explained by a larger interrogation volumen electrical recordings in comparison with optical record-ngs. An alternative explanation could be the spatial sepa-

ation between optical and electrical recording sites. Results

otaft

ISsttscciultoidhcvAwoa(s

OMuiupetmit

papdccafwmdwclsp

FRs .

1435Wei Kong et al Measuring APs with Optrodes in Pigs

btained here are more consistent with the second explana-ion. Both ATs and APD measurements were very similarnd highly correlated with changes of CL when measuredrom closely adjacent sites. Such close correspondence jus-ifies the use of ARI measurements as a substitute for APD.

ntramural APD heterogeneitypatial heterogeneities of repolarization and APD are con-idered important factors in arrhythmogenesis.21,22 One ofhe major reasons for electrophysiological heterogeneity ishe existence of “M cells” in the midmyocardium that re-pond to slowing in pacing rate and inhibition of potassiumurrents by prolonging APD to a much greater extent thanells in epicardial and endocardial regions.36 Until recently,ntramural APD heterogeneities were most often measuredsing microelectrodes and surface optical mapping in iso-ated cells and in wedge preparations.21,22,37 Optrodes offerhe possibility of characterizing electrophysiological heter-geneities in the intact hearts. The optrode measurements ofntramural APD performed here did not reveal significantifferences in APD values across the LV wall in porcineearts. Slowing of pacing rate and blocking of potassiumurrents with d-sotalol or dofetilide, which were used pre-iously to expose M cells,26,27 did not increase intramuralPD differences either. The lack of M cells is consistentith previous measurements in isolated tissue preparationsf porcine LV myocardium.30 As argued previously,38 theirbsence may be related to the young age of the animals�8–10 weeks) that were used in that and in the present

igure 5 Optical measurements of shock-induced �Vm. A: Intramural Vecording 1 was obtained from subepicardium, and recording 6 from sub

hock strength. Solid lines depict second-order polynomial fits of the data

tudies. t

ptrode measurements of shock-induced �Vmeasurements of intramural Vm changes are critical for

nderstanding the mechanism of defibrillation. Previously,ntramural �Vm were measured only in wedge preparationssing surface optical mapping.16 Although these studiesrovided valuable information about shock effects, theirxtrapolation to whole hearts is limited by the differences inissue properties and boundary conditions at the cut trans-ural surface of wedge preparations versus the ones in the

ntact LV, which necessitates mapping of intramural �Vm inhe intact hearts.

The optrode measurements of �Vm in porcine heartserformed in the present study revealed similarities as wells important differences from �Vm measurements in wedgereparations. Similar to wedge preparations, shocks pro-uced intramural �Vm. In contrast, the magnitude of thesehanges was significantly larger in the optrode recordings inomparison with wedge mapping. This can be attributed tohigher spatial resolution of optrode recordings and, there-

ore, to a smaller degree of spatial averaging in comparisonith surface optical mapping. In wedge mapping experi-ents, strong shocks produced globally negative �Vm that

id not change sign upon changing the shock polarity,hich was attributed to “virtual electrodes” formed at mi-

roscopic tissue discontinuities.16 In contrast, positive po-arizations were detected by optrodes even at the strongesthocks, and the �Vm sign reversed with change in the shockolarity. Such differences in �Vm behavior may be due to

dings in control and during application of 8-V/cm shocks of two polarities.rdium. B: Dependences of minimal and maximal �Vm magnitude on the

m recorendoca

he differences between wedge preparations and whole

haotetTi

CTpdfirct

AWsa

R

1

1

1

1

1

1

1

1

1

1

2

2

2

2

2

2

2

2

2

2

3

3

3

3

3

3

3

3

3

1436 Heart Rhythm, Vol 4, No 11, November 2007

earts. Alternatively, the discrepancies between the optrodend wedge mapping data could be due to the effect ofptrodes on �Vm. It was shown in computer simulationshat, with appropriate orientation of the shock field, plungelectrodes provide a substrate for shock-induced polariza-ions.39 The optrodes may possibly cause the same effect.he role of optrodes in shock-induced �Vm requires further

nvestigation.

onclusionshe quality of optrode recordings was substantially im-roved, which should facilitate the development of three-imensional mapping. The optrodes can be used for high-delity intramural measurements of activation andepolarization times. Measurements of shock-induced Vm

hanges, however, may be affected by the optrode itself;his question requires further investigation.

cknowledgmentse would like to thank Dr. Alexey Zaitsev for his help in

etting up a system for blood perfusion and Reuben Collinsnd Frank Vance for their technical support.

eferences1. Rosenbaum DS, Jalife J, eds. Optical mapping of cardiac excitation and arrhyth-

mias. Armonk, NY: Futura, 2001.2. Choi BR, Salama G. Simultaneous maps of optical action potentials and calcium

transients in guinea-pig hearts: mechanisms underlying concordant alternans.J Physiol 2000;529:171–188.

3. Fast VG, Ideker RE. Simultaneous optical mapping of transmembrane potentialand intracellular calcium in myocyte cultures. J Cardiovasc Electrophysiol2000;11:547–556.

4. Laurita KR, Singal A. Mapping action potentials and calcium transients simul-taneously from the intact heart. Am J Physiol 2001;280:H2053–H2060.

5. Kong W, Walcott GP, Smith WM, Johnson PL, Knisley SB. Emission ratiom-etry for simultaneous calcium and action potential measurements with coloadeddyes in rabbit hearts: reduction of motion and drift. J Cardiovasc Electrophysiol2003;14:76–82.

6. Fast VG, Kléber AG. Microscopic conduction in cultured strands of neonatal ratheart cells measured with voltage-sensitive dyes. Circ Res 1993;73:914–925.

7. Rohr S, Kucera JP, Fast VG, Kleber AG. Paradoxical improvement of impulseconduction in cardiac tissue by partial cellular uncoupling. Science 1997;275:841–844.

8. Choi BR, Liu T, Salama G. The distribution of refractory periods influences thedynamics of ventricular fibrillation. Circ Res 2001;88:E49–E58.

9. Zaitsev AV, Guha PK, Sarmast F, Kolli A, Berenfeld O, Pertsov AM, de GrootJR, Coronel R, Jalife J. Wavebreak formation during ventricular fibrillation inthe isolated, regionally ischemic pig heart. Circ Res 2003;92:546–553.

0. Akar FG, Rosenbaum DS. Transmural electrophysiological heterogeneities un-derlying arrhythmogenesis in heart failure. Circ Res 2003;93:638–645.

1. Omichi C, Lamp ST, Lin SF, Yang J, Baher A, Zhou S, Attin M, Lee MH,Karagueuzian HS, Kogan B, Qu Z, Garfinkel A, Chen PS, Weiss JN. Intracel-lular Ca dynamics in ventricular fibrillation. Am J Physiol 2004;286:H1836–H1844.

2. Katra RP, Laurita KR. Cellular mechanism of calcium-mediated triggered ac-tivity in the heart. Circ Res 2005;96:535–542.

3. Weiss JN, Qu Z, Chen PS, Lin SF, Karagueuzian HS, Hayashi H, Garfinkel A,Karma A. The dynamics of cardiac fibrillation. Circulation 2005;112:1232–1240.

4. Lakkireddy V, Bub G, Baweja P, Syed A, Boutjdir M, El-Sherif N. The kineticsof spontaneous calcium oscillations and arrhythmogenesis in the in vivo heart

during ischemia/reperfusion. Heart Rhythm 2006;3:58–66.

5. Efimov IR, Cheng Y, Van Wagoner DR, Mazgalev T, Tchou PJ. Virtual3

electrode-induced phase singularity. A basic mechanism of defibrillation failure.Circ Res 1998;82:918–925.

6. Fast VG, Sharifov OF, Cheek ER, Newton J, Ideker RE. Intramural virtualelectrodes during defibrillation shocks in left ventricular wall assessed by opticalmapping of membrane potential. Circulation 2002;106:1007–1014.

7. Frazier DW, Krassowska W, Chen P-S, Wolf PD, Daniely ND, Smith WM,Ideker RE. Transmural activations and stimulus potentials in three-dimensionalanisotropic canine myocardium. Circ Res 1988;63:135–146.

8. El-Sherif N, Chinushi M, Caref EB, Restivo M. Electrophysiological mechanismof the characteristic electrocardiographic morphology of torsade de pointestachyarrhythmias in the long-QT syndrome: detailed analysis of ventriculartridimensional activation patterns. Circulation 1997;96:4392–4399.

9. Chattipakorn N, Fotuhi PC, Chattipakorn SC, Ideker RE. Three-dimensionalmapping of earliest activation after near-threshold ventricular defibrillationshocks. J Cardiovasc Electrophysiol 2003;14:65–69.

0. Pogwizd SM. Nonreentrant mechanisms underlying spontaneous ventriculararrhythmias in a model of nonischemic heart failure in rabbits. Circulation1995;92:1034–1048.

1. Yan GX, Shimizu W, Antzelevitch C. Characteristics and distribution of M cellsin arterially perfused canine left ventricular wedge preparations. Circulation1998;98:1921–1927.

2. Akar FG, Yan GX, Antzelevitch C, Rosenbaum DS. Unique topographicaldistribution of M cells underlies reentrant mechanism of torsade de pointes in thelong-QT syndrome. Circulation 2002;105:1247–1253.

3. Trayanova N. Concepts of ventricular defibrillation. Phil Trans R Soc Lond A2001;359:1327–1337.

4. Hooks DA, LeGrice IJ, Harvey JD, Smaill BH. Intramural multisite recording oftransmembrane potential in the heart. Biophysical J 2001;81:2671–2680.

5. Byars JL, Smith WM, Ideker RE, Fast VG. Development of optrode for trans-mural multisite optical recordings of Vm in the heart. J Cardiovasc Electro-physiol 2003;14:1196–1202.

6. Sicouri S, Moro S, Elizari MV. d-Sotalol induces marked action potentialprolongation and early afterdepolarizations in M but not empirical or endocar-dial cells of the canine ventricle. J Cardiovasc Pharmacol Ther 1997;2:27–38.

7. Obreztchikova MN, Sosunov EA, Plotnikov A, Anyukhovsky EP, Gainullin RZ,Danilo P, Yeom ZH, Robinson RB, Rosen MR. Developmental changes in IKrand IKs contribute to age-related expression of dofetilide effects on repolariza-tion and proarrhythmia. Cardiovasc Res 2003;59:339–350.

8. Haws CW, Lux RL. Correlation between in vivo transmembrane action potentialdurations and activation-recovery intervals from electrograms. Effects of inter-ventions that alter repolarization time. Circulation 1990;81:281–288.

9. Holley LK, Knisley SB. Transmembrane potentials during high voltage shocksin ischemic cardiac tissue. PACE 1997;20:146–152.

0. Rodriguez-Sinovas A, Cinca J, Tapias A, Armadans L, Tresanchez M, Soler-Soler J. Lack of evidence of M-cells in porcine left ventricular myocardium.Cardiovasc Res 1997;33:307–313.

1. Kovoor P, Campbell C, Wallace E, Byth K, Dewsnap B, Eipper V, Uther J, RossD. Effects of simultaneous insertion of 66 plunge needle electrodes on myocar-dial activation, function, and structure. PACE 2003;26:1979–1985.

2. Arbel ER, Prabhu R, Ramesh V, Pick R, Glick G. A perfused canine right bundlebranch-septal model for electrophysiological studies. Am J Physiol 1979;236:H379–H384.

3. Newton JC, Smith WM, Ideker RE. Estimated global transmural distribution ofactivation rate and conduction block during porcine and canine ventricularfibrillation. Circ Res 2004;94:836–842.

4. Zhang YZ, He B, Wang LX. Effect of an increase in coronary perfusion ontransmural ventricular repolarization. Physiol Res 2007;56:285–290.

5. Caldwell BJ, Legrice IJ, Hooks DA, Tai DC, Pullan AJ, Smaill BH. Intramuralmeasurement of transmembrane potential in the isolated pig heart: validation ofa novel technique. J Cardiovasc Electrophysiol 2005;16:1001–1010.

6. Antzelevitch C, Shimizu W, Yan GX, Sicouri S, Weissenburger J, NesterenkoVV, Burashnikov A, Di Diego J, Saffitz J, Thomas GP. The M cell: itscontribution to the ECG and to normal and abnormal electrical function of theheart. J Cardiovasc Electrophysiol 1999;10:1124–1152.

7. Sicouri S, Fish J, Antzelevitch C. Distribution of M cells in the canine ventricle.J Cardiovasc Electrophysiol 1994;5:824–837.

8. Antzelevitch C. Are M cells present in the ventricular myocardium of the pig?A question of maturity. Cardiovasc Res 1997;36:127–128.

9. Langrill DM, Roth BJ. The effect of plunge electrodes during electrical stimu-lation of cardiac tissue. IEEE Trans Biomed Eng 2001;48:1207–1211.