antitachycardia pacing for ventricular tachycardia using ...pacericd.com/documents/articles/atp for...

www.pace

ricd.c

om

REVIEW

Antitachycardia Pacing for Ventricular TachycardiaUsing Implantable Cardioverter Defibrillators:Substrates, Methods, and Clinical ExperienceMICHAEL O. SWEENEYFrom CRM Research, Cardiac Arrhythmia Service, Brigham and Women’s Hospital and Harvard Medical School,Boston, Massachusetts

IntroductionAntitachycardia pacing (ATP) refers to the use

of pacing stimulation techniques for termination oftachyarrhythmias. Such techniques can be auto-matically applied using implantable cardioverterdefibrillators (ICDs) and offer the potential forpainless termination of ventricular tachycardia(VT). Reduction in painful shocks may improvepatient quality-of-life (QOL) and extend ICD pulsegenerator longevity. Numerous older studies haveconsistently demonstrated that ATP can reliablyterminate ∼ 85%-90% of slow VT (cycle lengths[CL] < 300–320 ms) with a low risk of acceleration(1%-5%).1 More recently similar high success andlow acceleration rates for fast VT (CL 320–240 ms)have been demonstrated. These observations haverepositioned the ICD as primarily an ATP devicewith defibrillation backup only as needed.

Physiology of Pacing TerminationTachycardias that require reentry to persist,

are susceptible to termination with pacing. Thesine-qua non of a reentrant arrhythmia is theability to reproducibly initiate and terminatethe tachycardia by critically timed extrastimuli.2Therefore, the possibility of successful termina-tion of tachycardias with pacing can be anticipatedon the basis of the mechanism. It is therefore usefulto consider the origin of ventricular tachycardiasthat are commonly encountered in the ICD patientpopulation.

Mechanism of Sustained Monomorphic VT inCoronary Artery Disease

The pathophysiological basis for sustainedmonomorphic VT due to prior myocardial infarc-tion is well understood. The mechanism of ar-rhythmogenesis in this setting is reentry.3−6 Theanatomic substrate for reentry is the interlacing of

Address for reprints: Michael O. Sweeney, M.D., CRMResearch, Cardiac Arrhythmia Service, Brigham and Women’sHospital, 75 Francis Street, Boston, MA 02115. Fax: 617-277-6289; e-mail: [email protected]

Received January 2, 2004; revised April 12, 2004; accepted May5, 2004.

viable myocardium and connective tissue (scar) atsites of prior myocardial infarction.7,8 This spe-cific pathological condition is the basis for low am-plitude, fractionated endocardial electrograms atthe sites of origin of VT.7 Poor cellular coupling atsites where fractionated electrograms are recordedresults in slow propagation of impulses neces-sary for initiation and maintenance of sustainedVT.7,8 Such abnormalities of conduction, alongwith altered refractoriness, enhanced automatic-ity and areas of inexcitablity form the electrophys-iological substrates for reentry caused by priormyocardial infarction.2 The evidence for reentryobtained from electrophysiology studies includesreproducible initiation and termination of tachy-cardia by critically timed extrastimuli, response ofthe tachycardia to stimulation or drugs, and activa-tion mapping demonstrating reentrant excitation.2

Thus, the interlacing of viable and nonviablemyocardium is capable of satisfying the three con-ditions required for initiation and maintenance ofa reentrant rhythm: (1) at least two functionally(or anatomically) distinct potential pathways thatjoin proximally and distally to form a closed cir-cuit of conduction; (2) unidirectional block in oneof these potential pathways; (3) slow conductiondown the unblocked pathway, allowing the previ-ously blocked pathway time to recover excitabil-ity. The reentry wavefront circulates around ar-eas of functional or anatomically fixed conductionblock. Areas of fixed conduction block are mostoften due to inexcitable scar tissue. Occasionally,an anatomic obstacle such as the mitral annulusforms a border zone for reentry circuits in adjacentinfarcted myocardium.9

The electrophysiological substrate for reen-try due to prior myocardial infarction is remark-ably durable. Long-term follow-up of patientswho present with sustained monomorphic VThas demonstrated a 3%-5% per year incidence ofrecurrent VT up to 15 years after presentation.2Sustained monomorphic ventricular tachycardiasinduced early after myocardial infarction amongpatients with spontaneous VT can be reproduciblyinduced up to a year later, regardless of whetherthe induced VT ever occurred spontaneously.10

Sustained monomorphic VT induced among

1292 September 2004 PACE, Vol. 27

www.pace

ricd.c

om

ANTITACHYCARDIA PACING IN ICDS

patients with prior myocardial infarction, reducedleft ventricular ejection fraction, and no his-tory of spontaneous sustained VT is reproducibly(> 90%) inducible up to 6 years later.11 These stud-ies establish that the substrate for reentry after my-ocardial infarction can remain anatomically per-sistent for many years.

Relationship Between Monomorphic VT andVentricular Fibrillation

The underlying mechanism for ventricularfibrillation (VF) not associated with acute my-ocardial infarction is poorly understood. Ambula-tory monitoring has clearly demonstrated that sus-tained monomorphic VT precedes some episodesof VF.12−14 Destabilization of monomorphic VTmay be related to ischemia, left ventricular dys-function, electrolyte imbalances, activation of thesympathetic nervous system, or other poorly un-derstood factors. Analysis of stored electrogramsretrieved from ICDs has yielded further insightsregarding the initiation of VF in the setting ofchronic coronary artery disease. The occurrenceof spontaneous monomorphic VT was observed tobe higher among patients who presented with VT(54%) versus those who presented with VF (18%)in one study.15 Abrupt onset VF (not preceded bymonomorphic VT) was recorded in 11% of the pa-tients who presented with VF. These observationssuggest that in some patients spontaneous VF isa primary event, rather than a destabilization ofmonomorphic VT. An example of VF initiated bysustained VT is shown in Figure 1. The clinicalimportance of this observation is that terminationof VT by ATP may prevent VF in some patients.

Mechanism of Sustained Monomorphic VT inNonischemic Dilated Cardiomyopathy

Sustained monomorphic VT in nonischemicdilated cardiomyopathy (NDCMP) is less com-mon than in ischemic cardiomyopathy. The patho-physiological basis for sustained monomorphicVT associated with NDCMP is poorly understoodcompared to chronic coronary artery disease, andprobably more diverse.

Autopsy series have demonstrated visuallyevident left ventricular scars (replacement fibrosis)in patients with NDCMP.16 Interlacing of replace-ment fibrosis and viable myocardium can producefractionated, broad, low amplitude, endocardialelectrograms compatible with slow conductionzones as seen in chronic myocardial infarction.17

These are capable of sustaining reentry.18 How-ever, most patients with NDCMP have relativelynormal endocardial activation and electrograms,not significantly different than normal individu-als. Only those rare patients with NDCMP and

sustained monomorphic VT have fractionated en-docardial electrograms.2,19,20

The electrophysiological mechanisms of ven-tricular arrhythmia in nonischemic idiopathicdilated cardiomyopathy were studied by intraop-erative mapping just prior to explantation amongpatients undergoing cardiac transplantation byPogwizd et al.21 Zones of functional conductiondelay and block were demonstrated in the epi-cardium and less often, in the midmyocardiumand endocardium. Extensive interstitial fibrosisin continuous linear bundles extending from theendocardium to the midmyocardium was consis-tently found in these locations. However, ventricu-lar premature beats and nonsustained VT inducedby programmed stimulation were found to ariseprimarily in the subendocardium by a focal mech-anism without evidence of macroreentry. Thesesites of initiation were consistently distant fromzones of functional conduction delay and blockthat did not contribute to VT initiation. The inves-tigators hypothesized that focal initiation of VTcould be due to triggered activity (delayed after-depolarizations [DADs], or early afterdepolariza-tions [EADs]) citing the observation that triggeredactivity can be initiated in the myocardium ofNDCMP.22

Implications of VT Substrate for ClinicalApplication of ATP in ICD Patients

Monomorphic VT associated with chronic is-chemic heart disease is most commonly due toclassic reentry and is therefore susceptible to ter-mination by ATP. Monomorphic VT is less com-monly due to reentry and occurs with lower fre-quency in NDCMP. These fundamental differencesin substrate are important for interpretation of clin-ical trials of ATP in ICD patients since nonreen-trant VT would not be expected to respond to ATP.

Termination of Reentrant VT by a PacingStimulus

Theoretically, any reentrant VT can be termi-nated by a critically timed pacing stimulus thatdepolarizes the excitable gap. This principle isshown schematically in Figure 2. In panel A, thehead of the wave of depolarization, denoted by(1), is followed by an area of absolute refractori-ness, denoted by (2), relative refractoriness, de-noted by (3), and an excitable gap, denoted by(4), that is fully excitable. In panel B, the stimulusenters the excitable gap both antegradely and ret-rogradely. Retrogradely it collides with the head,which is extinguished. Antegradely, it “preex-cites” the tail, perpetuating the reentry wavefront.Panel C shows a more premature stimulus. Ter-mination occurs because the impulse collides ret-rogradely with the preceding tachycardia impulse

PACE, Vol. 27 September 2004 1293

www.pace

ricd.c

om

SWEENEY

Figure 1. Tracings are from a 79-year-old male with ischemic cardiomyopathy and history of cardiac arrest. The patienthad a Medtronic 7271 ICD system, and the preonset electrogram storage feature was enabled. The ICD stores up to15 seconds of electrogram prior to onset of the episode. (Panel A) Interval plot associated with episode of ventricularfibrillation. The interval plot shows each VV interval (X-axis) with its corresponding interval value in milliseconds(Y-axis). Time zero is at episode detection. Note that detection is triggered by short VV intervals with wide cycle lengthvariability (150–320 ms) in the VF zone (< 320 ms). The VV intervals proceeding detection are stable at 400–410 msand dissociated from the AA intervals (1,000 ms). This is consistent with stable monomorphic VT. (Panel B) Storedlocal bipolar atrial and far-field ventricular electrograms confirm sustained monomorphic ventricular tachycardiathat degenerates to ventricular fibrillation (arrow). Had preonset electrogram storage not been enabled, the onset ofVF would have been assumed to be abrupt.


www.pace

ricd.c

om


Figure 2. Termination of reentry by a pacing stimulus. See text for details. Adapted from Figure21–1, Prysytowsky EN, Klein GJ.75

and blocks antegradely owing to encroachment onthe refractory period of the preceding wavefront.

Commonly, circulating reentrant VT wave-fronts propagate through regions of the infarct thatcontain surviving myocyte bundles. This is easiestto understand by considering a simple, theoretical,single-loop reentrant circuit as shown in Figure 3.The reentry wavefront circulates around a cen-tral region of inexcitable scar tissue, or functionalblock, which is surrounded by viable myocardium.

However, the true configuration of reentrantVT circuits is often considerably more complex asrepresented in Figure 4. In each panel the grayregions are regions of conduction block, whichcould be due to fibrous scar or collision of ex-citation wavefronts. Black arrows indicate wave-fronts propagating in “bystander” regions of thescar that are not in the reentry circuit. In theleft hand panel, the circulating reentry wavefrontpropagates along the border of the infarct region. Inthe middle panel, most of the reentry circuit is con-tained within the infarct region. In the right handpanel, a portion of the reentry circuit is containedwithin the infarct region, but following exit of theexcitation wavefronts from the infarct regions, twowavefronts travel along the border of the infarct toreach the proximal entrance to the reentry circuitpath through the infarct, forming a double loop

Figure 3. Simple, theoretical, single-loop, two-dimensional reentrant circuit. The reentry wavefrontcirculates around a central region of inexcitable scartissue, or functional block, which is surrounded byviable myocardium. See text for details. Adapted fromStevenson WG.76

(figure-of-8) reentry circuit. The reentry circuit of-ten contains zones of slow conduction.

Factors that Influence the Ability of Pacing toInteract with Tachycardia

In order for pacing stimuli to terminate reen-trant VT they must first interact with the VT cir-cuit. Josephson2 has summarized the major factorsthat influence the ability of pacing stimuli to inter-act with the VT circuit. These are (1) the tachy-cardia CL, (2) the presence and duration of an ex-citable gap in the VT circuit, (3) the conductiontime from the stimulation site to the site of impulseformation, and (4) the refractoriness at the stimula-tion site and site of impulse formation. The tachy-cardia CL is probably most important. In general,as CL decreases the probability of termination bypacing stimuli decreases. This is because shorterCLs have a “protective” effect on the VT circuit.The duration of the excitable gap shortens withCL, therefore the “window of vulnerability” is re-duced. Shorter CLs also reduce the probability thatremotely delivered pacing stimuli can overcomeconduction time from the stimulation site to theVT circuit before local refractoriness is reached atthe stimulation site.

Factors that Influence Termination of VT byPacing

Assuming conditions exist for pacing stimulito interact with the VT circuit, the probability ofsuccessful termination is influenced by several fac-tors 2,23 including (1) timing of the stimulus, (2)coupling interval, rate and number of pulses in thestimulus drive train, (3) proximity of the stimulat-ing site to the circuit (closer is better), (4) barriers(functional or anatomic) to the circuit.

Two-dimensional models of reentrant VT cir-cuits such as shown in Figures 3 and 4 aresufficient for explaining classical resetting, en-trainment, and abrupt termination (Type-1 breaks)by pacing stimuli.24 However, reentrant circuitsare complex three-dimensional structures with


www.pace

ricd.c

om

SWEENEY

Figure 4. Complex, theoretical, multiple pathway, three-dimensional reentrant circuit. See textfor details. Adapted from Stevenson WG.76

multiple fixed barriers and pathways. Such spa-tial heterogeneity complicates the understandingof how pacing stimuli interact with the tachycar-dia circuit. Insights regarding such complexitiescan be gained by analyzing Type-2 breaks in re-sponse to ATP, where VT persists or changes toanother VT for one or more beats before termina-tion. Sharma et al.25 analyzed the incidence andcharacteristics of Type-2 breaks in response to ATPusing stored EGMs from ICDs. Type-2 breaks wereobserved in ∼ 32% of patients accounting for ∼10% of episodes and were associated with a 150%increase in VT CL variability after ATP delivery.Such oscillation in VT CL has been described inspontaneous termination of VT.26 ATP affected ei-ther VT CL or morphology, or both in 80% of Type-2 breaks. It is possible that changes in VT CL and/ormorphology during Type-2 breaks reflect modi-fication of the reentrant path by the ATP wave-front within a complex, three-dimensional sub-strate. Approximately 9% of all Type-2 episodesmay be spontaneously terminating nonsustainedVT since ATP did not affect these episodes in anyway.

Antitachycardia Pacing Nomeclature andModalities

A single critically timed extrastimulus mayterminate reentrant VT but the efficacy is low.23

Multiple stimuli delivered in the form of pacingdrive trains increase the probability of interact-ing with the VT circuit and the likelihood of ter-mination. The building blocks of ATP stimula-tion patterns are burst and ramp pacing (Fig. 5).A burst stimulation pattern consists of a train ofpacing pulses with an equal interstimulus inter-val. A ramp stimulation pattern consists of a trainof pacing pulses with an automatically decrement-ing interstimulus interval. Either stimulation pat-tern may be applied with “rate adaptation” whichmeans that the interval from the last sensed ven-tricular event during VT to the first pacing stimu-lus is a programmable percentage of the detected

VT CL. There are an almost limitless number ofvariations on these themes that probably have littleclinical advantages and have been reviewed else-where.27

Application of ATP in ICDsClinical Considerations

As reviewed earlier, termination of reentranttachycardia requires critical timing of the pacingstimulus. Success or failure of the pacing stimu-lus to interact with and terminate tachycardia, isinfluenced by several factors, some of which arespecific to the pacing scheme itself. These includethe timing of the stimulus, and the coupling in-terval, rate and number of pulses in the stimulustrain. It is for this reason that different schemesmay have different efficacy in terminating specifictachycardias in the individual patient. The flex-ibility to program different pacing schemes en-ables one to overcome, in part, the factors thatconstrain the ability of some pacing stimuli to in-teract with the tachycardia, such as proximity ofthe stimulating site to the circuit and functionalor anatomic barriers to the circuit. Convention-ally, ATP stimuli are delivered from the tip of anendocardial electrode positioned at the apex ofthe right ventricle. The pacing stimuli must cap-ture the local myocardium and produce an exci-tation wavefront that propagates through the my-ocardium to the left ventricle, where, in the caseof coronary disease, most reentrant circuits arelocated.

Some generalizations regarding the clinicalapplication of ATP for terminating VT are possible.Acceleration to a faster monomorphic VT, poly-morphic VT, or VF, is uncommon when ATP is ap-plied to slow VT (CL < 300–320 ms). The lack of aconsistent definition of acceleration renders com-parison between studies difficult. A >10%-25%change in CL, or the transformation of stable VTto polymorphic VT or VF is generally acceptedas acceleration. Most studies have consistently re-ported acceleration rates in the range of < 10% for


www.pace

ricd.c

om


Figure 5. (Top) Simple burst-pacing scheme. (Bottom) Simple ramp-pacing scheme. See text for details.

slow VT regardless of the pacing scheme. From apractical standpoint, acceleration is a variation onATP termination failure that, in addition to invok-ing a painful shock, may result in hemodynamicinstability due to ventricular rate, delay in defini-tive shock therapy, or both, and may have impor-tant clinical consequences such as syncope. Theprobability of ATP termination failure and accel-eration increase as VT CL decreases. Early recog-nition of this relationship resulted in a historicalhesitancy to apply ATP for fast VT (CL < 300–320ms). This consternation has been largely resolvedby recent large clinical trials and similar high suc-cess rates and low acceleration rates can be an-ticipated with properly applied ATP for fast VT.Additionally, ramp pacing is more likely to resultin acceleration for fast VTs than burst pacing; this

does not appear to be generally true for slow VTs.An example of acceleration by ATP is shown inFigure 6.

Comparisons of ATP Modalities for VT

As commonly occurs, the emergence of a newtechnology stimulates clinical research centeredon application of that technology. Accordingly,when ventricular pacing capability in ICDs ap-peared in the late 1980s and early 1990s a seriesof small clinical investigations compared differ-ent ATP modalities for termination of VT. Thesesmall studies are difficult to compare due to enroll-ment bias, small numbers, nonrandomized treat-ment assignments, variable validation of treatedrhythms, and differences between pacing schemes


www.pace

ricd.c

om

SWEENEY

Figure 6. Acceleration of fast VT by ATP.

and substrates for VT. Nonetheless, it is instructiveto selectively review some of this data as it pro-vides a consensus on the success and limitationsof ATP in a diversity of ICD patient populations.

The relative efficacy of the two most com-monly applied ATP schemes, burst and ramp, hasbeen studied (Tables I and II) for induced and spon-taneous VT. In three of four studies, there was nodifference in the ability of either scheme to termi-nate VT, which was successful in about 70%-75%of episodes.28−30 In one study, ramp pacing wassignificantly more effective than burst pacing, butthe overall success of either scheme was signifi-cantly lower at 50%-60% than that observed in theother studies.31 There was no difference in the like-lihood of pacing acceleration of induced VT usingburst or ramp pacing in all four studies; the inci-dence ranged from 3% to 21%. Similarly, Fisheret al.32 demonstrated similar efficacy of burst and

Table I.

Comparisons of Burst vs Ramp-Pacing for Induced VT

Termination AcclerationInduced (%) Burst (%) Burst

Reference VT vs Ramp vs Ramp

(28) All cycle lengths 76 vs 68 3 vs 11VT CL < 300 ms 86 vs 38* 0 vs 12VT CL > 300 ms

(29) All cycle lengths 70 vs 72 21 vs 18VT CL < 300 ms 45 vs 36 55 vs 55VT CL < 300 ms 76 vs 80 13 vs 9

(30) All cycle lengths 65 vs 72 4 vs 3(31) All cycle lengths 49 vs 75 11 vs 12

VT CL < 330 ms 55 vs 28* 14 vs 7*VT CL > 330 ms 57 vs 39* 9 vs 6

*significant difference.

ramp pacing for induced VT, but noted that burstpacing required a significantly fewer number of at-tempts for termination.

The relative efficacy of burst versus ramp pac-ing for terminating induced VT was then comparedon the basis of VT CL. Again, in two of three stud-ies, burst and ramp pacing were equivalently ef-fective in terminating both slower VTs, typicallydefined as having CLs greater than 300–330 ms,and more rapid VTs, defined as having CLs lessthan or equal to 300–330 ms.28,29 In contrast, onestudy showed that ramp pacing was more effec-tive than burst pacing for both slower and fasterVTs.31 Another important observation from thesestudies is that, generally, the incidence of pacingacceleration of VT is higher with shorter VT CLs,regardless of the pacing scheme. In these studies,the incidence of acceleration for slower VTs rangedfrom 4% to 13%, whereas the incidence of accel-eration for faster VTs ranged from 0% to 55%.

As there is no reason, a priori, to believe thatthe response of spontaneously occurring VTs to

Table II.

Comparisons of Burst vs Ramp-Pacing Spontaneous VT

Termination AcclerationSpontaneous (%) Burst (%) Burst

Reference VT vs Ramp vs Ramp

(28) All cycle lengths 96 vs 93 0.02 vs 0.01VT CL < 300 ms 86 vs 38* 0 vs 12VT CL > 300 ms 71 vs 77 4 vs 10

(33) All cycle lengths 85 vs 90 8 vs 3(34) All cycle lengths 88 vs 94* 4 vs 3

VT CL < 300 ms 86 vs 77* 7 vs 18*VT CL > 300 ms 88 vs 96* 3 vs 2

*significant difference


www.pace

ricd.c

om


pacing stimuli should be the same as that observedfor induced VTs, this comparison was the sub-ject of three studies. In two of three studies, therewas no difference in the ability of either burstor ramp pacing to terminate spontaneous VT ofany CL, which was successful in about 85%-95%of episodes.28,33 In one study, ramp pacing wasmore successful but the difference was small.34

As was observed for induced VTs, there was nosignificant difference in the incidence of acceler-ation of spontaneous VT for either burst or ramppacing.

The relative efficacy of burst versus ramp pac-ing for terminating spontaneous VT was then com-pared on the basis of VT CL in two of these stud-ies. Interestingly, for slower VTs with CLs > 300ms, there was no difference in efficacy betweenburst and ramp pacing. However, unlike inducedVTs, for faster VTs with CLs < 300 ms, burst pac-ing was more likely to terminate spontaneous VTsthan ramp pacing.28,34 This raises the intriguingquestion of why the response of induced versusspontaneous VT to ATP should be different. In ei-ther case, the incidence of acceleration was low,but more likely with ramp pacing for faster VTs inone study.34

Peters et al.35 reported a reduced likelihoodof ATP success and increased risk of accelerationwith ramp versus burst-pacing schemes. In con-tradistinction, Nasir et al.35A in 1997 observed nodifference in termination success, failure, or risk ofacceleration between burst or ramp pacing. Theseresults were unchanged when ramp pacing andscanning were used in all possible combinations.Though the pacing schemes were nonrandomized,the consistent success, failure, and accelerationrates across pacing schemes suggested that there isprobably no important clinical difference betweenthe bewildering array of pacing schemes availableto the clinician.

Empiric Versus Untested Programming of ATP

An important practical question is whetherATP schemes should be tested on inducedVTs prior to application for spontaneous VTs.Schaumann et al.36 in 1998 showed that there wasa high level of concordance between the effective-ness of ATP tested against induced VTs in theelectrophysiology laboratory and spontaneouslyoccurring clinical VTs (Table III). Similarly, the in-cidence of acceleration was equivalently low inboth circumstances. A single ATP scheme (threesequences of an 8–10 burst train of autodecremen-tal ramp at 81% of VT CL) was prospectively eval-uated in 200 patients. This scheme was demon-strated to be successful at terminating induced VTin 54 patients, and either unsuccessful or untested(due to noninducibility of VT) in the remain-

Table III.

Comparison of Specific ATP Schemes for Induced VersusSpontaneous VT

Termination (%) Accleration (%)Reference Tested vs Nontested Tested vs Nontested

(77) 79 vs 76 N/A(36) 95 vs 90 2 vs 5

ing 146 patients. During average follow-up of 20months, > 90% of 5,165 spontaneously occurringVT episodes in both groups were successfully ter-minated with ATP. Acceleration rates were simi-larly low (2%-5%).

Clinical Application of ATP for Fast VT

Until recently, ATP was conventionally ap-plied to only slower, presumably hemodynami-cally tolerated VTs, typically with CLs > 300–320 ms. Though several of the small studies citedearlier showed that many fast VTs (CL < 300–300 ms) can reliably be terminated by ATP, con-cerns about loss of consciousness during pacingattempts that delay definitive shock therapy at ex-tremely rapid ventricular rates or acceleration toan even faster VT or VF. Fast VT is therefore typi-cally detected by ICDs as VF and terminated withpainful shocks even though ATP might be success-ful. The PainFREE Rx studies tested the hypothesisthat ATP is safe and effective for FVT and reducesshock burden in ICD patients (Table IV).

The PainFREE Rx studies evaluated ATP atvery rapid rates within the VF treatment zone. FastVT (FVT) detection required 12/16 (PainFREE Rx I)or 18/24 (PainFREE Rx II) intervals with CL < 320+ last 8 consecutive intervals ≥240 ms. If 1 or moreof the last 8 intervals was < 240 ms, the episodewas detected as VF. In PainFREE Rx I, 220 patientswith coronary artery disease and ICDs for stan-dard indications received nonrandomized, empiri-cal, standardized ATP therapy for all FVT episodesprior to shock delivery.37 The pacing scheme con-sisted of two sequences of an 8-pulse burst-pacingtrain at 88% of the FVT CL. ATP terminated 396FVT episodes (89%) with an overall efficacy of77% when adjusted for multiple episodes/patientand a 4% incidence of acceleration.

These observations were validated and ex-tended in PainFREE Rx II, which was the first large-scale, randomized trial to compare use of ATP ver-sus shocks to treat FVT (CL 320–240 ms).38 Sixhundred thirty-seven patients were randomized toreceive shocks or ATP (1 sequence of an 8-pulseburst-pacing train at 88% of the FVT CL). Fifty-six


www.pace

ricd.c

om

SWEENEY

Table IV.

Studies of ATP for Fast VT (CL 320–240 ms)

FVT% Termination (%, Accelerationof all adjusted for multiple FVT episodes

Reference N episodes episodes/patient with ATP

(37) 220 (nonrandomized) 40 77 4(38) 637 (randomized) 34 72 1.2

percent of true ventricular episodes were classifiedas slow VT (CL > 320 ms), 34% as FVT (CL 240–320 ms), and 10% as VF (CL < 240 ms). ATP termi-nated 82% of FVT episodes (72% when adjustedfor patients with multiple episodes). The relativereduction of shocked episodes by programmingATP for FVT was 71%. Acceleration occurred in1.2% of episodes in the ATP group but was neverassociated with syncope. The PainFREE Rx stud-ies conclusively demonstrate that ATP is highlyeffective for FVT and results in a > 70% relativereduction in the proportion of shocked episodeswithout increasing time to termination or risk ofacceleration.

Clinical Factors Affecting SuccessfulTermination of VT with ATP

The basic electrophysiological factors that in-fluence the ability of pacing stimuli to interactwith a reentrant VT circuit and terminate tachy-cardia have been extensively investigated and de-scribed previously. However, relatively little isknown about clinical factors that influence thesuccess of ATP among ambulatory ICD patients.Hamill et al.31 reported that lower ejection frac-tion, longer VT CL, and coronary artery diseasewere multivariate predictors of ATP success forinduced VT. Fries et al.39 reported lowest relativeATP success and highest VT occurrence rates wereobserved in the morning hours (6 AM to noon).There was a trend towards highest risk of accelera-tion during ATP in this time period as well. Peterset al.35 observed that ATP was less successful inwomen, when there was more severe left ventricu-lar dysfunction, and in the presence of antiarrhyth-mic drugs. Kouakam et al.40 observed that fast si-nus rates immediately preceding the onset of VT,and absence of β-blocker therapy were the only in-dependent predictors of ATP failure. A more com-prehensive analysis from PainFree Rx II evaluatedpredictors of ATP success for FVT (CL 320–240ms).41 The strongest predictor of ATP efficacy wasejection fraction (EF < 30% ATP efficacy 52% vsEF > 30%, efficacy = 72%, P = 0.01). History of

sustained VT, inducible VT, New York Heart As-sociation (NHYA) class, age, sex, history of infarc-tion, and infarct location were not predictive ofATP success.

An interesting recent development in the clin-ical application of ATP is stimulation site of origin.It is important to note that the pathophysiologicalmechanism of reentrant VT is not dependent on, orinfluenced by, site of origin of the VT circuit. Froma practical perspective, site of origin might be veryimportant since the majority of VT circuits arisein the left ventricle and pacing stimuli are conven-tionally delivered from the right ventricular apex.Since distance and conduction time between stim-ulation site and site of origin affect the ability ofpacing stimuli to interact with the reentrant cir-cuit, ATP delivered from the left ventricular pac-ing lead, or biventricular pacing leads in cardiacresynchronization therapy/defibrillation systems(CRTD) might improve efficacy compared to rightventricular ATP.

The relative efficacy of right ventricular ver-sus biventricular ATP was evaluated in the InSyncICD OUS (Outside United States) Study.42 ATP ter-mination success was 2.4 times greater with biven-tricular versus right ventricular ATP and appearedto be associated with fewer accelerations for bothslow VT and fast VT. A preliminary report from theVENTAK CHF/CONTAK CD Study showed thatbiventricular ATP was more successful in patientsrandomized to CRT pacing therapy.43 This effectwas influenced by left ventricular pacing lead lo-cation (improving in lateral locations, worseningin anterior locations) and improved over time inthe patients who were receiving CRT.

These data are insufficient to support defini-tive conclusions regarding the role of alternate siteATP for terminating VT. Due to technical limita-tions, the CRTD ICDs in both studies were onlycapable of right ventricular or biventricular stim-ulation, and therefore provide no insights on apossible role for isolated left ventricular stimula-tion. From a theoretical perspective, it is not im-mediately obvious that left ventricular stimulation


www.pace

ricd.c

om


should improve ATP success in coronary arterydisease, since many reentrant VT circuits arise inthe interventricular septum which is closer to aRV stimulation site than a left ventricular free wallstimulation site. Conduction delay out of left ven-tricular stimulation sites due to interposed infarc-tion and fibrosis might modify any advantage re-lated to proximity to site of VT origin, and thiseffect may be different in the right ventricle. Fur-thermore, a recent case report described the re-producible initiation of monomorphic VT by leftventricular pacing but not right ventricular pac-ing that could be reliably terminated by RV ATPbut not by LV ATP.44 This suggests the possibilitythat local tissue anisotropy might affect the abil-ity of site-specific stimulation wavefronts to inter-act with the reentrant VT circuit. Recent studieshave shown that pacing-site dependent changes inventricular activation sequence can alter ventricu-lar repolarization and refractoriness.45 How thesefactors might influence the relative efficacy of leftventricular ATP is unknown.

Effect of ATP on Quality of Life in ICD PatientsThe issue of QOL in the ICD patient pop-

ulation has been extensively evaluated.46−50 Al-though ICD therapy is generally well toler-ated by most patients, approximately 30%-50%experience some degree of psychological distressfollowing implantation.51 One of the principal lim-itations of ICD therapy is the discomfort asso-ciated with high-voltage shocks. Several studieshave noted a direct correlation between poor QOLscores and the experience of ICD shocks.46−49

Intuitively, the absence of discomfort is theprincipal merit of ATP from the patient’s perspec-tive. Early nonrandomized studies showed thatATP reduced shocks52 for VT. However, the as-sumption that every episode of VT successfullytreated with ATP would otherwise have neces-sitated a shock is erroneous and overestimatesthe magnitude of shock reduction. This is be-cause ATP is delivered immediately when detec-tion criteria are satisfied, whereas shocks are de-layed due to capacitor charging. Therefore, ATPmay appear to have successfully treated VT thatwould have terminated spontaneously. This situ-ation was observed in PainFREE Rx II where 33%of fast VT episodes in the shock arm terminatedspontaneously during capacitor charging.38 Ac-cordingly, only randomized comparisons of ATPversus shocks for VT can account for this poten-tial bias and accurately quantify shock reductionand possible effects on QOL.

PainFREE Rx II was the first randomized trialthat tested the hypothesis that reduction in painfulshocks by ATP would improve QOL. Among pa-tients with at least one episode of FVT, those

treated with ATP experienced significantly moreimprovement in physical functioning, role physi-cal, social functioning, and vitality, as well as inoverall mental health compared to patients treatedwith shocks.53

Programming ATP and Patient SelectionPatient and Rhythm Selection for ATP

An important and unresolved issue is opti-mal application of ATP in different ICD patientpopulations. In general, secondary prevention pa-tients have a greater frequency of spontaneous ven-tricular arrhythmia than primary prevention pa-tients. However, differences in the incidence ofspecific ventricular rhythms (VT, fast VT, and VF),response to therapy (ATP or shocks) and suscepti-bility to spurious therapies due to SVT are incom-pletely characterized. Several studies have prelim-inarily addressed these issues.

Wilkoff et al.54 analyzed the frequency andcharacteristics of spontaneous VT and VF betweenpatients with a primary versus secondary preven-tion indication for ICD therapy in the MIRACLEICD study of CRTD. Primary prevention patientshad a lower frequency of appropriate VT and VFepisodes (0.12 vs 0.53 episodes/month) at signif-icantly faster CLs (303 ± 53 ms vs 367 ± 54 ms,P < 0.0001). Primary prevention patients also had asignificantly higher percentage of device-classifiedVF (40% vs 14%, P < 0.0001). The absolute rateof inappropriate detections in the primary pre-vention group was lower but constituted a muchhigher portion of all episodes for that group (32%vs 14% for the secondary prevention group). Mostinappropriate detections in the primary preven-tion group were due to sinus tachycardia and weretreated as VT.

Russo et al.55 examined spontaneous thera-pies in primary prevention patients. Over 21 ±18 months, 23% patients had appropriate thera-pies and 14% had inappropriate therapies for SVT.Clinical VT rates were higher than SVT rates (211± 38 beat/min vs 179 ± 14 beats/min). Only 10%of the patients with appropriate therapies had VTrates < 190 beats/min. The authors concluded thatalthough there was some overlap in VT and SVTrates, VT rates less than 190 beats/min were un-common and avoidance of programming to nom-inal VF detection rates may reduce inappropriateshocks for SVT.

These early observations provoke examina-tion of tachyarrhythmia detection and therapyprogramming based on indication for ICD ther-apy. “Overtreatment” in primary prevention pa-tients is an important concern, potentially at thecost of spurious therapies for inappropriate ven-tricular detections due to SVT. Though similar


www.pace

ricd.c

om

SWEENEY

proportions of primary and secondary preventionpatients have appropriate detections for poten-tially life-threatening VT, fast VT and VF the fre-quency is lower in primary prevention.

Since the relative frequency of specific ven-tricular rhythms is similar between primary andsecondary prevention patients, an equivalentefficacy of ATP could be anticipated assumingsimilar arrhythmia substrate (i.e., reentrant VT).Therefore, it is reasonable to conclude that if anyVT therapy is to be prescribed in either group,it should include ATP with the expectation that70%-90% of episodes will be painlessly termi-nated. The more difficult issue is whether anyslow VT therapy should be prescribed in primaryprevention patients, particularly those in whomprogrammed stimulation has not been performed.Elimination of slow VT detection might reducespurious therapies for some specific SVTs (such assinus tachycardia) but might not be as effective forothers, such as atrial fibrillation with a rapid ven-tricular response. The zeal for reducing the proba-bility of spurious therapies by eliminating a slowVT detection zone must be balanced against therisk of failing to treat unanticipated VT. This issuewas indirectly addressed by a retrospective studyby Bansch et al.56 The risk of VT above the VTdetection interval ranged between 2.7% and 3.5%per year during the first 4 years after ICD implanta-tion. Fifty-four (88.5%) of the VT episodes abovethe VT detection interval were associated with sig-nificant symptoms and 10% of patients had to beresuscitated. Risk factors for VT above the initialVT detection interval were heart failure, lower EF,spontaneous or inducible monomorphic VT, anduse of Class III antiarrhythmic drugs. The risk ofrecurrent VT above the VT detection interval was11.8%, 12.5%, and 26.6% during the first, second,and third year after the first occurrence above theVT detection interval. This suggests that elimina-tion of a slow VT zone in some patients will resultin clinically consequential undertreatment of slowVT.

Future Developments in ATPReducing Risk of Syncope During Attempts atPainless Termination of VT

Syncope prior to termination of VT by ICDtherapies remains a significant problem. The in-cidence of syncope during spontaneous VT in ICDpatients ranges from 2%-4% at 6 months37,57 to10%-15% at 2–3 years.57,59 Syncope, despite suc-cessful termination of potentially lethal VT, mayresult in bodily trauma. Syncope is also sociallydisabling and is the principal concern underlyingdriving restrictions in ICD patients.57,60

The hemodynamic response to sustained VTis heterogeneous. Arterial pressure decreases andcardiac filling pressures increase at the onset ofVT.61 If VT sustains, arterial pressure may par-tially recover. If arterial blood pressure does notrecover sufficiently, reductions in cerebral bloodflow and oxygen saturation result in syncope.62

The extent of this hemodynamic recovery is prob-ably governed by multiple interlinked factors.These include heart rate, loss of atrioventricular(AV) synchrony, left ventricular EF, ventriculardyssynergy, ischemia, and alterations in auto-nomic activity.63−66 Of these, heart rate and auto-nomic response are probably the most importantdeterminants of hemodynamic outcome.

The multiplicity of factors that may determinethe hemodynamic response to VT likely explainsthe heterogeneous response observed among pa-tients in clinical studies.67 Readily available clin-ical variables do not reliably predict likelihoodof syncope during VT. Generally, the risk of syn-cope during VT relates to tachycardia rate,57,64,68

although the correlation is modest. In two studies,the likelihood of syncope during induced VT wascorrelated with heart rate but only when VT rateexceeded 200 beats/min.69,70 Thus, it may be thatat slower VT rates, other factors play a more im-portant role in hemodynamic outcome. AlthoughVT duration prior to termination and severity ofleft ventricular dysfunction are commonly thoughtto relate to risk of syncope, this is not consis-tently supported by clinical investigation 68−70 orexperimental studies.64,66 Demographic, clinicaland variables derived from electrophysiologicalstudy were not predictive of syncope during spon-taneous device therapies.59,71 In contrast, Banschet al.57 found that low baseline left ventricular EFand induction of fast VT (CL < 300 ms) conferredan increased risk of syncope during appropriateICD therapies.57

The clinical management of syncope risk inICD patients is further confounded by several ob-servations. First, many episodes of syncope appearto be due to failed therapies for relatively slowVT. Olatidoye et al. observed that 62% of synco-pal events were due to VT acceleration by ATP orlow energy cardioversion, whereas 23% were dueto VT alone and 15% were due to VF.59 Second,some patients with syncopal VT may have otherVT events not resulting in syncope.37,59,71 Third,some patients with nonsustained VT experiencenear-syncope.37,58

These data suggest that VT rate and durationand other common clinical variables may not besufficiently robust discriminators for the risk ofsyncope and hence, ICD design and programming.One potential approach to this problem is ATP dur-ing capacitor charging. Theoretically, this should


www.pace

ricd.c

om


reduce the delay between ATP termination fail-ure and definitive shock therapy. This might inturn reduce the risk of syncope in some patients.Weber et al.72 evaluated consciousness during in-duced rapid VTs (CL 300–220 ms) in 20 patientsrandomized to either immediate shock or a sin-gle ATP attempt prior to shock. ATP terminated55% of episodes of induced rapid VTs whereas car-dioversion shocks were 100% successful. No pa-tient suffered syncope during capacitor chargingfor a shock or during successful or failed ATP at-tempt. This preliminary data suggests that a singleattempt at ATP during capacitor charging wouldnot result in syncope in the study population.

Selecting Fast Ventricular Rhythms for ATP

Current ICDs use rate-based detection thatmay not reliably discriminate between monomor-phic fast VT (CL 320–240 ms) and VF. FVT isoften detected as VF and treated with shockseven though most episodes of FVT are pace-terminable.37,38 In PainFREE Rx II, FVT detectionrequired 18/24 intervals with CL < 320 + last 8consecutive intervals ≥240 ms. Stored EGM anal-ysis of 564 episodes of device detected VF revealedthat 68/132 (52%) were actually FVT and 64/132(48%) were true VF. Therefore, more than 50% ofpotentially pace-terminable FVTs are misclassifiedas VF using this rate-based detection technique.These observations suggest that alternative detec-tion methods that consider CL variability or mor-phology73 are needed to discriminate FVT from

true VF in order to permit broader application ofATP.

SummaryAntitachycardia pacing reliably terminates ∼

85%-90% of slow VT (cycle lengths [CL] < 300–320 ms) with a low risk of acceleration (1%-5%).Similar high success and low acceleration ratesfor fast VT (CL 320–240 ms) have recently beendemonstrated. These results appear to be consis-tent across different substrates (ischemic vs non-ischemic dilated cardiomyopathy) and probablyrelate to a common mechanism (reentry) ATP-responsive VTs. These observations have reposi-tioned the ICD as primarily an ATP device withdefibrillation backup only as needed. Reductionin painful shocks may improve patient QOL andextend ICD pulse generator longevity.

Some general recommendations on program-ming ATP schemes are possible. For VT CL > 300–330 ms, burst and ramp pacing are equivalentlyeffective for terminating VT and equivalently lowrisk for causing acceleration. For VT CL < 300–330ms, burst pacing is more effective and less likely toresult in acceleration than ramp pacing. In eithercase, the risk of acceleration is inversely related tothe VT CL. “Less aggressive” burst stimulation (forexample, 91% of VT CL vs 81% of VT CL) is moreeffective and causes less acceleration, especiallyfor fast VT (CL < 320 ms).74 “Tailoring” of ATPto specific induced VTs is not necessary in mostsituations.

References1. Fisher JD, Mehra R, Furman S. Termination of ventricular tachy-

cardia with bursts of rapid ventricular pacing. Am J Cardiol 1978;41:94–102.

2. Josephson ME. Clinical cardiac electrophysiology: Techniquesand interpretations. 3rd ed. Philadelphia: Lippincott, Williamsand Wilkins; 2002.

3. Josephson ME, Almendral JM, Buxton AE, et al. Mechanisms ofventricular tachycardia. Circulation 1987; 74:41–47.

4. de Bakker JMT, van Capelle FJL, Janse MJ, et al. Reentry as a causeof ventricular tachycardia among patients with chronic ischemicheart disease: Electrophysiologic and anatomic correlation. Cir-culation 1988; 77:589–606.

5. de Bakker JM, van Capelle FJL, Janse MJ, et al. Macroreentry in theinfarcted human heart: Mechanism of ventricular tachycardiaswith a focal activation pattern. J Am Coll Cardiol 1991; 18:1005–1014.

6. Pogwizd SM, Hoyt RH, Saffitz JE, et al. Reentrant and focal mech-anisms underlying ventricular tachycardia in the human heart.Circulation 1992; 86:1872–1887.

7. Gardner PI, Ursell PC, Fenoglio JJ Jr., et al. Electrophysiologicand anatomic basis for fractionated electrograms recorded fromhealed myocardial infarcts. Circulation 1985; 72:592–611.

8. de Bakker JMT, van Capelle FJL, Janse MJ, et al. Slow conduc-tion in the infarcted human heart: “zigzag” course of activation.Circulation 1993; 88:915–926.

9. Wilber DJ, Kopp DE, Glascock DN, et al. Catheter ablation of themitral isthmus for ventricular tachycardia asssociated with infe-rior infarction. Circulation 1995; 92:3481–3489.

10. Roy D, Marchand E, Theroux P, et al. Long-term reproducibil-ity and significance of provocable ventricular arrhythmias aftermyocardial infarction. J Am Coll Cardiol 1986; 8:32–37.

11. Brembilla-Perrot B, Houriez P, Claudon O, et al. Long-term

reproducibility of ventricular tachycardia induction with electro-physiological testing in patients with coronary heart disease anddepressed left ventricular ejection fraction. PACE 2000; 23:47–53.

12. Lahiri A, Balasubramian V, Rafferty EB. Sudden death duringambulatory monitoring. Br Med J 1977; 1:1676–1678.

13. Gradman AH, Bell PA, DeBusk RF. Sudden death during ambu-latory monitoring. Circulation 1977; 55:210–211.

14. Pratt CM, Francis MJ, Luck JC, et al. Analysis of ambulatory elec-trocardiograms in 15 patients during spontaneous ventricular fib-rillation with special reference to preceding arrhythmic events. JAm Coll Cardiol 1982; 2:789–797.

15. Raitt MH, Dolack GL, Kudenchuk PJ, et al. Ventricular arrhyth-mias detected after transvenous defibrillator implantaiton amongpatients with a clinical history of only ventricular fibrillation: Im-plication for use of implantable defibrillator. Circulation 1995;91:1996–2001.

16. Roberts WC, Siegel RJ, BM M. Idiopathic dilated cardiomyopathy:Analysis of 152 necropsy patients. Am J Cardiol 1987; 60:1304–1315.

17. de Bakker JM, van Capelle FJ, Janse MJ, et al. Fractionated elec-trograms in dilated cardiomyopathy: Origin and relation to ab-normal conduction. J Am Coll Cardiol 1996; 27:1071–1078.

18. Wu TJ, Ong JJC, Hwang C, et al. Characteristics of wavefrontsduring ventricular fibrillation in human hearts with dilated car-diomyopathy: Role of increased fibrosis in the generation of reen-try. J Am Coll Cardiol 1998; 32:187–196.

19. Cassidy DM, Vassallo JA, Miller JM, et al. Endocardial cathetermapping in humans in sinus rhythm: Relationship to underly-ing heart disease and ventricular arrhythmias. Circulation 1986;73:645–652.

20. Delacretaz E, Stevenson WG, Ellison KE, et al. Mapping and ra-diofrequency catheter ablation of the three types of sustained


www.pace

ricd.c

om

SWEENEY

monomorphic ventricular tachycardia in nonischemic heartdisease. J Cardiovasc Electrophysiol 200; 11:11–17.

21. Pogwizd SM, McKenzie JP, Cain ME. Mechanisms underly-ing spontaneous and induced ventricular arrhythmias in pa-tients with idiopathic dilated cardiomyopathy. Circulation 1998;98:2404–2414.

22. Vermeulen JT, Tan HL, Rademaker H, et al. Electrophysiologicand extracellular ionic changes during acute ischemia in the fail-ing and normal rabbit myocardium. J Mol Cell Cardiol 1996;28:123–131.

23. Fisher JD, Kim SG, Matos JA, et al. Comparative effectivenessof pacing techniques for termination of well-tolerated sustainedventricular tachycardia. PACE 1983; 65:915–922.

24. Callans J, Hook BG, Josephson ME. Comparison of resetting andentrainment of uniform sustained ventricular tachycardia. Fur-ther insights into the characteristics of the excitable gap. Circu-lation 1993; 87:1229–1238.

25. Sharma V, DeGroot PJ, Wathen MS. Incidence and characteristicsof Type-2 breaks in response to antitachycardia pacing therapyin implantable cardioverter defibrillator patients. J CardiovascElectrophysiol 2003; 14:1156–1162.

26. Callans DJ, Marchlinksi FE. Characterization of spontaneous ter-mination of sustained ventricular tachycardia associated withcoronary artery disease. Am J Cardiol 1991; 67:50–54.

27. Fisher JD, Johnston DR, Kim SG, et al. Implantable pacers fortachycardia termination: Stimulation techniques and long-termefficacy. PACE 1986; 9:1325–1333.

28. Gillis AM, Leitch J, Sheldon RS, et al. A prospective random-ized comparison of autodecremental pacing to burst pacing indevice therapy for chronic ventricular tachycardia secondary tocoronary artery disease. Am J Cardiol 1993; 72:1146–1151.

29. Calkins H, El-Atassi R, Kalbfleisch S, et al. Comparison of fixedburst versus decremental burst pacing for termination of ventric-ular tachycardia. PACE 1993; 16:26–32.

30. Kantoch MJ, Green MS, Tang AS. Randomized cross-over eval-uation of two adaptive pacing algorithms for the termination ofventricular tachycardia. PACE 1993; 16:1664–1672.

31. Hamill SC, Packer DL, Stanton MS, et al. and the MulticenterPCD Investigator Group. Termination and acceleration of ven-tricular tachycardia with autodecremental pacing, burst pacing,and cardioversion in patients with an implantable cardioverterdefibrillator. PACE 1995; 18:3–10.

32. Fisher JD, Zhang Z, Kim SG, et al. Comparison of burst pacing, au-todecremental (ramp) pacing, and universal pacing for termina-tion of ventricular tachycardia. Archives des Maladies du Coeuret des Vaisseaux 1996; 89:135–139.

33. Newman D, Dorian P, Hardy J. Randomized controlled compari-son of antitachycardia pacing algorithms for termination of ven-tricular tachycardia. J Am Coll Cardiol 1993; 21:1413–1418.

34. Schaumann A, Poppinga A, von zur Muehlen F, et al. Anti-tachycardia pacing for ventricular tachycardias above and below200 beats/min: A prospective study for ramp versus can mode.(abstract) PACE 1997; 20:1108.

35. Peters RW, Zhang X, Gold MR. Clinical predictors and efficacy ofantitachycardia pacing in patients with implantable cardioverterdefibrillators: The importance of the patient’s sex. PACE 2001;24:70–74.

35A. Nasir N, Pacifics A, Doyle TK, et al. Spontaneous ventriculartachycardia treated by antitachycardia pacing. Cadence Investi-gators. Am J Cardiol 1997; 790:820–822.

36. Schaumann A, von zur Muhlen F, Herse B, et al. Empirical versustested antitachycardia pacing in implantable cardioverter defib-rillators: A prospective study including 200 patients. Circulation1998; 97:66–74.

37. Wathen MS, Sweeney MO, DeGroot P, et al. for the PainFREERx Investigators. Shock reduction using antitachycardia pacingfor rapid spontaneous ventricular tachycardia in patients withcoronary artery disease. Circulation 2001; 104:796–801.

38. Wathen MS, Sweeney MO, Khalighi K, et al. Prospective, random-ized comparison of shock versus antitachycardia pacing for fastventricular tachycardia (188–250 bpm) in ICD patients. In: NorthAmerican Society of Pacing and Electrophysiology (NASPE) An-nual Scientific Sessions; 2003; Washington, DC; 2003.

39. Fries R, Heisel A, Nikoloudakis N, et al. Antitachycardia pacingin patients with implantable cardioverter-defibrillators: Inversecircadian variation of therapy success and acceleration. Am J Car-diol 1997; 80:1487–1489.

40. Kouakam C, Lauwerier B, Klug B, et al. Effect of elevated heart ratepreceding the onset of ventricular tachycardia on antitachycardia

pacing effectiveness in patients with implantable cardioverter de-fibrillators. Am J Cardiol 2003; 92:26–32.

41. Volosin KJ, Khalighi K, Machado C, et al. Ejection fraction pre-dicts success of antitachycardia pacing for fast ventricular tachy-cardia. Circulation 2003; 108:IV-544.

42. Krater L, Lamp B, Heintze J, et al. Influence of antitachy pacinglocation on the efficacy of ventricular tachycardia termination.J Am Coll Cardiol 2002; 39:124A.

43. Lozano IF, Higgins S, Hummel J, et al. The efficacy of simultane-ous right and left ventricular antitachycardia pacing (BiV ATP)in heart failure patients with an AICD indication improves withtime. (abstract) PACE 2003; 26:984.

44. Guerra J, Wu J, Miller JM, et al. Increase in ventricular tachycardiafrequency after biventricular implantable cardioverter defibrilla-tor upgrade. J Cardiovasc Electrophysiol 2003; 14:1245–1247.

45. Medina-Ravell VA, Lankipalli RS, Yan GX, et al. Effect of epi-cardial or biventricular pacing to prolong QT interval and in-crease transmural dispersion of repolarization. Circulation 2003;107:740–746.

46. Sears SF, Conti JB. Understanding implantable cardioverter de-fibrillator shocks and storms: Medical and psychosocial consid-erations for research and clinical care. Clin Cardiol 2003; 26:107–111.

47. Schron EB, Exner DV, Yao Q, et al. for the AVID Investigators.Quality of life in the antiarrhythmics versus implantable defibril-lators trial. Impact of therapy and influence of adverse symptomsand defibrillator shocks. Circulation 2002; 105:589–594.

48. Namerow PB, Firth BR, Heywood GM, et al. for the CABG PatchTrial Investigators and Coordinators. Quality of life six monthsafter CABG surgery in patients randomized to ICD versus noICD therapy: Findings from the CABG Patch Trial. PACE 1999;22:1305–1313.

49. Irvine J, Dorian P, Baker B, et al., for the CIDS Investigators. Qual-ity of life in the Canadian Implantable Defibrillator Study (CIDS).Am Heart J 2002; 144:282–289.

50. Exner DV. Quality of life in patients with life threatening arrhyth-mias: Does choice of therapy make a difference? Am Heart J 2002;144:208–211.

51. Sears SF, Todaro JF, Urizar G, et al. Assessing the psychosocialimpact of the ICD: A national survey of implantable cardioverterdefibrillator health care providers. PACE 2000; 23:939–945.

52. Leitch JW, Gillis AM, Wyse DG, et al. Reduction in defibrillatorshocks with an implantable device combining antitachycardiapacing and shock therapy. J Am Coll Cardiol 1991; 18:145–151.

53. Wathen MS, Khaligi K, Voloson KJ, et al., for the PainFREE Rx IIInvestigators. Antitachycardia pacing of fast ventricular tachycar-dia improves quality of life in implantable cardiac defibrillatorpatients. Circulation 2003; 108:IV-615.

54. Wilkoff B, Hess M, Young JD, et al., on behalf of the MIRACLE ICDInvestigators. Differences in ICD utilization by primary and sec-ondary prevention patients with cardiac resynchronization ther-apy indication. (abstract) PACE 2003; 26:972.

55. Russo AM, Hsia H, Callans D, et al. Importance of avoidingnominal programming to prevent inappropriate implantable car-dioverter defibrillator shocks. J Am Coll Cardiol 2002; 39:79A.

56. Bansch D, Castrucci M, Bocker D, et al. Ventricular tachycardiasabove the initially programmed tachycardia detection interval inpatients with implantable cardioverter-defibrillators: Incidence,prediction, and significance. J Am Coll Cardiol 2000; 36:557–565.

57. Bansch D, Brunn J, Castrucci M, et al. Syncope in patients withimplantable cardioverter-defibrillator: incidence, prediction andimplications for driving restrictions. J Am Coll Cardiol 1998;31:608–615.

58. Grimm W, Flores BF, Marchlinski FE. Symptoms and electro-cardiographically documented rhythm preceding spontaneousshocks in patients with implantable cardioverter-defibrillators.Am J Cardiol 1993; 71:1415–1418.

59. Olatidoye AG, Verroneau J, Kluger J. Mechanisms of syncope inimplantable cardioverter-defibrillator recipients who receive de-vice therapies. Am J Cardiol 1998; 82:1372–1376.

60. Curtis AB, Conti JB, Tucker KJ, et al. Motor vehicle accidents inpatients with an implantable cardioverter-defibrillator. J Am CollCardiol 1995; 26:180–184.

61. Crozier IG, Ikram H, Niholls MG. Hemodynamic and hormonechanges during induced ventricular tachycardia secondary tocoronary artery disease. Am J Cardiol 1989; 63:618–621.

62. Singer I, Dawn B, Edmonds H Jr., et al. Syncope is predictedby neuromonitoring in patients with ICDs. PACE 1999; 22:216–222.


www.pace

ricd.c

om


63. Samet P. Hemodynamic sequelae of cardiac arrhythmias. Circu-lation 1973; 47:399–408.

64. Smith M, Kinugawa T, Dibner-Dunlap ME. Reflex control of sym-pathetic activity during ventricular tachycardia in dogs. Circula-tion 1996; 93:1033–1042.

65. Smith ML, Joglar JA, Wasmund SL, et al. Reflex control of sym-pathetic activity during simulated ventricular tachycardia in hu-mans. Circulation 1999; 100:628–634.

66. Hamdan MH, Joglar JA, Page RL, et al. Baroreflex gain predictsblood pressure recovery during simulated ventricular tachycar-dia in humans. Circulation 1999; 100:381–386.

67. Steinbach KK, Merl O, Frohner K, et al. Hemodynamics duringventricular tachyarrhythmias. Am Heart J 1994; 127:1102–1106.

68. Huikuri HV, Zaman L, Castellanos A, et al. Changes in sponta-neous sinus node rate as an estimate of cardiac autonomic toneduring stable and unstable ventricular tachycardia. J Am Coll Car-diol 1989; 13:646–656.

69. Hamer AWF, Rubin SA, Peter T, et al. Factors that predict syn-cope during ventricular tachycardia in patients. Am Heart J 1984;107:997–1005.

70. Saksena S, Ciccone JM, Craelius W, et al. Studies on left ventric-ular function during sustained ventricular tachycardia. J Am CollCardiol 1984; 4:501–508.

71. Kou WH, Calkins H, Lewis RR, et al. Incidence of lossof consciousness during automatic implantable cardioverter-defibrillator shocks. Ann Intern Med 1991; 115:942–945.

72. Weber M, Block M, Bansch D, et al. Antitachycardia pacing forrapid VT during ICD charging: A method to prevent ICD shocks.PACE 2001; 24:345–351.

73. Brown ML, Sarkar S. Discrimination of VT from VF by stabil-ity of ventricular electrogram morphology. (Abstract) PACE 2003;26:988.

74. Peinado R, Almendral J, Rius T, et al. Randomized, prospec-tive comparison of four burst pacing algorithms for sponta-neous ventricular tachycardia. Am J Cardiol 1998; 82:1422–1425.

75. Prystowsky EN, Klein GJ. Cardiac Arrhythmias. An integratedapproach for the clinician. New York: McGraw-Hill, Inc,1994.

76. Stevenson WG. Ventricular tachycardia after myocardial infarc-tion. J Cardiovasc Electrophysiol 1995; 6:942–950.

77. Sweeney MO, Guy ML, McGovern B, et al. Success of implantablecardioverter-defibrillator antitachycardia pacing algorithms forspontaneous sustained ventricular tachycardia is not dependenton local bipolar electrogram morphology. (abstract) Circulation1996; 94:I-321.


antitachycardia pacing for ventricular tachycardia using ...pacericd.com/documents/articles/atp for...

Documents