Cardiac Electrophysiology Review 2003;7:468472


C 2004 Kluwer Academic Publishers. Manufactured in The Netherlands.
The Dual Chamber and VVI Implantable Defibrillator
(DAVID) Trial: Rationale, Design, Results, Clinical
Implications and Lessons for Future Trials
Abstract. The Dual Chamber and VVI Implantable De-
fibrillator (DAVID) trial randomized 506 patients and
tested the hypothesis that the dual-chamber pacing
mode would produce improved hemodynamics and
would in turn reduce congestive heart failure, heart
failure hospitalizations, heart failure deaths, atrial fib-
rillation, strokes, ventricular arrhythmias, and total
mortality compared to backup ventricular pacing in
patients indicated for implantable defibrillator ther-
apy. Patients had either primary prevention indica-
tions (47%) or secondary prevention indications (53%)
for implantable defibrillator therapy but had no in-
dications for bradycardia pacemaker support. All the
patients had moderate to severe left ventricular dys-
function with a left ventricular ejection fraction of 40%
or less (mean = 27%) and were consistently treated
with angiotensin converting enzyme inhibitors or an-
giotensin II receptor blockers (86%) and beta adrener-
gic blocking agents (85%). The primary combined end-
point of hospitalization for congestive heart failure or
death was paradoxically increased and statistically sig-
nificant ( p= 0.03) at one year in the patients paced in
the dual chamber mode (22.6%) compared to patients
randomized to ventricular backup pacing (13.3%). Both
heart failure hospitalization and mortality contributed
outcome.
Another perspective would consider this a random-
ized controlled study of presence or absence of pace-
maker therapy in patients with left ventricular dysfunc-
tion and indications for implantable defibrillator ther-
apy. Ventricular backup pacing produced less than 3%
ventricular and no atrial pacing, while dual chamber
pacing produced approximately 60% atrial and ventric-
ular paced heart beats. The poor outcome in the dual
chamber paced group correlated with the percentage
of right ventricular pacing and suggests that right ven-
tricular pacing caused ventricular dyssynchrony. The
poor outcome associated with right ventricular pacing
compared to intrinsic activation in the control group
of the DAVID trial is reminiscent of the poor outcome
associated with prolonged intraventricular conduction
activation in the control groups compared to biventric-
ular pacing in the intervention groups of the cardiac
resynchronization trials.
The direct conclusion from these results are that pa-
tients with indications for implantable defibrillators
and no indication for pacing should not be paced in
the dual chamber pacing mode. It is not appropriate
to conclude that only single chamber implantable de-
fibrillators should be implanted. There are other po-
tential advantages to having an implanted atrial lead
468
Bruce L. Wilkoff and the DAVID Trial Investigators
The Cleveland Clinic Foundation, Cleveland, OH 44195, USA
including improved secondary outcomes. However the
DAVID trial results suggest that the dual chamber paced
mode was not associated with improved quality of life
or decreased frequency of hospitalization, inappropri-
ate shocks from the defibrillator or atrial fibrillation.
The more important question is what is the optimal
pacing mode in these patients? The AAIR mode is un-
der investigation in the DAVID II study in an attempt to
identify a pacing mode that preserves atrio-ventricular
synchrony, normal atrio-ventricular timing, prevents
bradycardia and also prevents right ventricular stim-
ulation.
Caution should be taken to not directly apply these
results to patients with either an indication for pace-
maker therapy or to patients with an indication for car-
diac resynchronization therapy since patients from nei-
ther population were included. However, considering
the large magnitude of the deleterious effects associ-
ated with dual chamber pacing in the DAVID trial future
studies should explore the possibility that left ventricu-
lar stimulation may be the only pacing mode capable of
preventing bradycardia without increasing death and
congestive heart failure.
Key Words. DAVID, implantable defibrillator, pace-
maker, heart failure
The Dual Chamber and VVI Implantable
Defibrillator (DAVID) Trial:
A Randomized Trial
The implantable cardioverter defibrillator (ICD) im-
proves survival for many patients with either the
history of or significant risk for life-threatening ven-
tricular arrhythmias [1]. However the randomized
clinical trials designed to test the efficacy of ICD ther-
apy were almost exclusively conducted before dual
chamber pacing was integrated into the devices. Be-
cause few of these patients had indications for brady-
cardia pacing, virtually all of the patients had their
devices programmed to backup ventricular (VVI at a
slow rate) pacing. Consequently the impact of atrial
or atrial-ventricular pacing had not been established
in this patient population.
Address corrrespondence to: Bruce L. Wilkoff, M.D., The
Cleveland Clinic Foundation, 9500 Euclid Avenue Desk F-15,
Cleveland, OH 44195, USA. E-mail: wilkofb@ccf.org
CEPR 2003; Vol. 7, No. 4
Before the Dual Chamber and VVI Implantable
Defibrillator (DAVID) trial it was recognized that
ventricular stimulation with pacemakers could ad-
versely affect ventricular contraction, but the fre-
quency and magnitude adverse effects had not
been quantified. However, pacemakers often are im-
planted in patients with modest degrees of sinus node
dysfunction and first degree AV block when nega-
tive chronotropic and dromotropic medications are
prescribed for coronary artery disease, hypertension,
left ventricular dysfunction or atrial and ventricu-
lar tachyarrhythmias. The DAVID trial tested the
hypothesis that dual-chamber pacing would permit
optimal drug therapy and improve hemodynamics
and would in turn reduce CHF, heart failure hospi-
talizations, heart failure deaths, atrial fibrillation,
strokes, ventricular arrhythmias, and total mortal-
ity compared to backup ventricular pacing (VVI) in
ICD patients [2]. The investigators thought that the
advantages of avoiding bradycardia during aggres-
sive pharmacologic treatment of an ICD population
with significant left ventricular dysfunction and no
overt indications for pacemaker therapy would out-
weigh the potential disadvantages associated with
right ventricular stimulation. The results of the trial
proved the hypothesis incorrect.
Methods
Technically, the DAVID Trial was a multicenter, ran-
domized, single-blinded, parallel study of patients
with ICDs, comparing VVI and DDDR paced modes
[3]. The DAVID trial was multicenter, randomized,
parallel, controlled evaluation of pacemaker ther-
apy in patients with left ventricular dysfunction and
indications for ICD therapy. Enrollment began in
October 2000 and concluded at the recommendation
of the data safety monitoring board on September 30,
2002.
Participants and ICD implantation
All patients had a standard indication for ICD im-
plantation, including a history of either spontaneous
or inducible sustained ventricular tachyarrhythmias
but without an indication for antibradycardia pacing.
Transvenous dual-chamber pacemaker ICDs (Pho-
ton, Photon Micro, or Photon Atlas, St Jude Medical
Inc, Sylmar, Calif) were implanted with a minimal
defibrillation safety margin of 10 J. A high risk pop-
ulation of ICD patients was included by requiring
every patient to have an LVEF of 40%.
Randomization
After ICD implantation, patients were randomly as-
signed to the VVI mode with a lower rate of 40/min
(VVI-40) or to the DDDR mode with a lower rate of
70/min (DDDR-70). Only the DDDR-70 group had
activation of supraventricular tachycardia detection
DAVID Trial 469
enhancements. This single-blind randomization (the
patient was blinded to pacing mode) was stratified by
site, history of CHF, and history of atrial fibrillation.
Programming
The prescribed programmed parameters are de-
scribed in Table 1. The investigators had freedom to
program all other parameters as clinically indicated.
Heart failure drug therapy
Pharmacologic therapy for left ventricular dysfunc-
tion and heart failure consisted of digoxin, diuretics,
ACE inhibitors, and -blockers in the doses described
in Table 2.
Objectives and outcome measures
The combined primary end point was freedom from
death or hospitalization for heart failure.
Statistical methods
The DAVID Trial used a 2-sided = .05 level test of
the null hypothesis with monitoring for early rejec-
tion of the null hypothesis. The null hypothesis was
that the time to death or hospitalization for treat-
ment of heart failure would be similar in both treat-
ment groups. Comparison was based on intention-
to-treat. The log-rank statistic was the test statistic.
The anticipated sample size was 800 subjects.
Results
Baseline characteristics
A total of 506 patients were enrolled and random-
ized in the trial. The baseline characteristics of the
randomized patients are similar to ICD recipients
in previous trials. The mean LVEF was 27% (pa-
tients with LVEF of >40% were excluded). Nearly
one half of the patients were NYHA functional class
I, 39% were NYHA class II, and only approximately
12% were functional class IIIIV. The mean intrin-
sic QRS duration was 120 ms and 30.8% of patients
had a QRS duration of at least 130 ms. Right and
left bundle-branch pattern was present in 11.0% and
16.5% respectively. The baseline characteristics were
well balanced in the 2 randomized groups.
Pharmacologic therapy for left ventricular dys-
function was balanced between groups at initial hos-
pital discharge after randomization and throughout
the study [2]. Eighty-five per cent of both randomized
groups received beta adrenergic blockade.
Follow-up data
Median follow-up was 8.4 months (range, 0
23.6 months). The cumulative percentage of ventric-
ular stimulation over time is displayed in Table 3.
470 Wilkoff CEPR 2003; Vol. 7, No. 4

Table 1. Requisite programming

Mode assignment VVI DDDR

Pacing rate 40 bpm 70 bpm

AV delay NA 100240 ms
Mode switch NA On /180 bpm
Tachycardia detection 102150 bpm Same
For patients without VT,
set to 150 bpm with tachycardia
therapies off (monitor zone)
and MTD off
Fibrillation detection 200240 bpm Same
SVT upper limit 200 bpm Same
Arrhythmia sensing mode Ventricular only Dual chamber
Sudden onset Passive (100 ms) On (100 ms)
Interval stability On (80 ms) On (80 ms) with AVA (60 ms)
Morphology On (5 of 8; 60%) Same
Tachycardia diagnosis criteria Any Same
Automatic template update On /1 day Same
MTD timer 2 min for patients Same
without VT = off
Stored ECG A and V bipolar + markers Same
pre-trigger 16 sec, maximum
duration 2 min
DFT testing 2 successful VF defibrillations Same
at 20 Joules, or 3 successful
VF defibrillations
at 25 Joules
DFT safety margin 10 Joules Same

bpm = beats per minute, ECG = electrocardiogram, NA = not applicable, AV = atrioventricular, ms = milliseconds, MTD = maximum tachycardia
duration, AVA = atrial ventricular association, DFT = defibrillation threshold, A = atrial, V = ventricular.

Table 2. Protocol-defined drug therapy in the DAVID trial Table 3. Percent beats ventricular paced

Generic name Initial dose (mg) Target (mg) VVI (%) DDDR (%) p-value

Enalapril 2.5 Bid 10 Bid 3 months 1.5 57.6 0.001

Captopril 12.5 tid 50 tid 6 months 0.7 60.7 0.001
Lisinopril 5 qd 20 qd 12 months 2.9 61.0 0.001
Benazepril 5 qd 20 qd
Fosinopril 5 qd 20 qd
Ramipril 2.5 qd 10 qd
Quinapril 5 bid 20 bid Insights and implications
Perindopril 1 qd 4 qd The Dual Chamber or VVI Implantable Defibril-
Trandolapril 0.5 qd 2 qd lator Trial could have been subtitled, A Placebo-
Moexipril 3.75 qd 12 qd Controlled Pacemaker Trial because the differen-
Carvedilol 3.125 bid 25 bid
tial outcomes are the result of the pacing therapy
Metoprolol 6.25 bid 100 bid
Bisoprolol 1.25 qd 20 qd
and mode. Consequently these results have impli-
cations for all patient populations for which atrial
and ventricular stimulation are employed. In the ad-
dition to the primary endpoint data summarized in
this article, secondary endpoint data were presented
Endpoint at NASPE 2003 in Washington DC, including quality
The event rates of the primary end point and its sub of life, hospitalization, inappropriate ICD shocks and
components are shown in Figure 1 [2]. ATP events, and atrial fibrillation. None of these re-
Contrary to original expectations, the DDDR mode sults support use of DDDR pacing in patients with-
was associated with an increased incidence in the out pacemaker indications and these findings also
composite endpoint of death or hospitalization for question the role of conventional DDDR devices for
new or worsened congestive heart failure. Both sub bradycardia pacing. VVI pacing at 40 bpm seems to
components of the primary end point trended consis- have an advantage over DDDR pacing in patients
tently with the composite endpoint. with ICD indications.
CEPR 2003; Vol. 7, No. 4
Fig. 1. Kaplan Meyer analysis of the event rates of (A) Death or first hospitalization for new or worsened CHF, (B) First
hospitalization for new or worsened CHF or (C) Death comparing patients paced in the DDDR mode compared to ventricular
backup pacing (intrinsic rhythm).
The obvious direct conclusion from this investi-
gation is that ICD patients with moderate to se-
vere left ventricular dysfunction but without symp-
tomatic bradycardia, should have the pacing mode
set to ventricular backup pacing. However the real
questions have not been definitively answered: What
kind of ICD should be implanted in patients with left
ventricular dysfunction? What is the optimal pac-
ing mode? Are there other unidentified advantages
to having an atrial lead or intrinsic to a device capa-
ble of dual chamber pacing? Finally, what is the true
cost of the atrial lead and device?
There are multiple other possible pacemaker
modes to choose in these patients depending on the
mechanism of the deleterious effects of the DDDR
mode. AAI pacing at 70 bpm, is the subject of in-
vestigation in the DAVID II trial. The reason for this
choice is that the most likely explanation for the poor
outcome in patients programmed to the DDDR mode
is ventricular desynchronization therapy, due to RV
stimulation. Alternative choices are the DDI or DDIR
modes with long programmed AV delays. Other pos-
sible mechanisms for the harm demonstrated dur-
ing the DAVID trial include the increased heart rate,
delayed intra-atrial conduction and altered AV con-
duction and filling times intrinsic to atrial pacing.
If these alternative mechanisms are the dominant
physiologies then the AAI mode will not prove to be
a safe mode. Estimates from the DAVID data sug-
gest that if the AVID, CIDS or CASH trials had been
conducted with DDDR pacing ICDs we would still be
debating if amiodarone was as effective as ICDs for
these patients [46].
DAVID Trial 471
Why was the adverse effect of DDDR pacing not
known in the past? Perhaps, despite the evidence for
ventricular dyssynchrony with RV pacing, the lack
of a control group in prior trials prevented detection
and quantification of the adverse effect. Most of the
previous trials of pacing modes, such as MOST, PASE
and CTOPP enrolled patients that needed pacing and
therefore there was similar frequency of RV pacing
in both treatment arms [79]. Alternatively it is pos-
sible that this effect is much larger in ICD patients
with left ventricular dysfunction. The DAVID trial
was limited to patients with LVEF of 40% (mean
27%) but excluded patients with persistent atrial fib-
rillation. However, the association between DDDR
pacing and adverse outcome is not likely to be iso-
lated to the population studied.
Should we be implanting only biventricular ICDs?
Should we be implanting VVI ICDs in patients with-
out pacing indications and biventricular ICDs in all
patients with pacing indications? We dont know the
answers to those questions . . . yet. In addition, I must
insist that the patients studied in the DAVID trial
are not similar to patients treated in the cardiac
resynchronization therapy trials. The patients in the
DAVID trial had not yet manifest clinical evidence of
congestive heart failure, but should we make them
biventricular ICD candidates with DDDR pacing?
Given the results of the DAVID and cardiac resyn-
chronization trials, it is my practice to be extremely
selective about the patients that receive an atrial
lead with the ICD. Dual chamber ICDs cost more
and with the expansion of indications for ICDs it is
clear that the overall dollars spent on ICD care in
472 Wilkoff
the US is going to increase. Although it is firmly es-
tablished that ICD therapy saves lives it is not es-
tablished that addition of ventricular backup pac-
ing saves lives. Other unproven features include rate
adaptive sensor driven pacing and, from the DAVID
trial, DDDR pacing. In fact, despite the encouraging
news from the cardiac resynchronization trials, espe-
cially the early reports from the COMPANION trial,
even atrial stimulation with biventricular stimula-
tion has not been evaluated adequately or critically.
Buyers beware, it is precisely the patients with
the weakest pacing indications who did the worst in
this trial. Patients with relative sinus bradycardia
and a first degree AV block paced more frequently
in the atrium and the ventricle when programmed
to the DDDR mode. Also it is not established that
AAI, DDI, AV search hysteresis or some new mode
switching mechanism will prevent the profound ad-
verse effect that DDDR pacing demonstrated in the
DAVID trial. It is possible that increased heart rate
and alterations in intra-atrial conduction and AV
conduction contribute significantly to the adverse ef-
fects. Accordingly, the DAVID II trial is comparing
the safety of AAI pacing at a rate of 70 bpm with
ventricular backup pacing.
Significant advances are being made in the tech-
niques of cardiac resynchronization therapy. Al-
though not considered appropriate in the DAVID
population at this time, left ventricular dysfunction
and congestive heart failure are progressive and de-
teriorating conditions. It is my current policy to use
as little hardware as possible until the patients con-
dition and clinical trials suggest that more compli-
cated therapy is appropriate. Leads commonly fail
over time. Implanting hardware increases surgical
time and increases the risk of infection. Finally, the
addition of an atrial lead often compromises venous
access, which then produces the need for lead extrac-
tion prior to subsequent addition of a left ventricular
lead.
Having considered all of the above situations, a
significant number of dual chamber ICDs should still
be implanted. These devices are still clearly indi-
cated in patients with AV block, less than NYHA
functional class III heart failure symptoms and in
patients with SYMPTOMATIC sinus node dysfunc-
tion. In addition, the diagnostics associated with the
atrial lead are extremely useful at the time of device
follow-up. However it is questionable if this is com-
pelling enough to justify the medical or financial cost.
Finally, patients should continue to be enrolled in tri-
als designed to establish the appropriate indications
for dual chamber ICD implantation and for selection
of the appropriate pacing mode. A more troublesome
question, not pursued in this discussion, is how the
CEPR 2003; Vol. 7, No. 4
results of the DAVID trial should affect our use of
pacemakers in patients without ICD indications.
References
1. Gregoratos G, Abrams J, Epstein AE, Freedman RA, Hayes
DL, Hlatky MA, Kerber RE, Naccarelli GV, Schoenfeld
MH, Silka MJ, Winters SL, Gibbons RJ, Antman EM,
Alpert JS, Hiratzka LF, Faxon DP, Jacobs AK, Fuster V,
Smith SC, Jr. ACC/AHA/NASPE 2002 guideline update
for implantation of cardiac pacemakers and antiarrhyth-
mia devices: Summary article: A report of the American
College of Cardiology/American Heart Association Task
Force on Practice Guidelines (ACC/AHA/NASPE Commit-
tee to Update the 1998 Pacemaker Guidelines). Circulation
2002;106(16):21452161.
2. Wilkoff BL, Cook JR, Epstein AE, Greene HL, Hallstrom
AP, Hsia H, Kutalek SP, Sharma A. The DAVID Trial In-
vestigators. Dual-chamber pacing or ventricular backup
pacing in patients with an implantable defibrillator. JAMA
2002;288(24):31153123.
3. Wilkoff BL. Should all patients receive dual chamber pac-
ing ICDs? The rationale for the DAVID trial. Curr Control
Trials Cardiovasc Med 2001;2:215217.
4. Antiarrhythmics Versus Implantable Defibrillators (AVID)
Investigators. A comparison of antiarrhythmic-drug ther-
apy with implantable defibrillators in patients resusci-
tated from near-fatal ventricular arrhythmias. N Engl J
Med 1997;337:15761583.
5. Connolly SJ, Gent M, Roberts RS, Dorian P, Roy D, Sheldon
RS, Mitchell LB, Green MS, Klein GJ, OBrien B. Canadian
implantable defibrillator study (CIDS): A randomized trial
of the implantable cardioverter defibrillator against amio-
darone. Circulation 2000;101(11):12971302.
6. Kuck KH, Cappato R, Siebels J, Ruppel R. Random-
ized comparison of antiarrhythmic drug therapy with im-
plantable defibrillators in patients resuscitated from car-
diac arrest. The Cardiac Arrest Study Hamburg (CASH).
Circulation 2000;102(7):748754.
7. Lamas GA, Lee KL, Sweeney MO, Silverman R, Leon
A, Yee R, Marinchak RA, Flaker G, Schron E, Orav EJ,
Hellkamp AS, Greer S, McAnulty J, Ellenbogen K, Ehlert
F, Freedman RA, Estes NA 3rd, Greenspon A, Goldman L.
Ventricular pacing or dual-chamber pacing for sinus-node
dysfunction. N Engl J Med 2002;346(24):18541862.
8. Lamas GA, Orav EJ, Stambler BS, Ellenbogen KA,
Sgarbossa EB, Huang SK, Marinchak RA, Estes, NA
3rd, Mitchell GF, Lieberman EH, Mangione CM Goldman
L, Quality of life and clinical outcomes in elderly pa-
tients treated with ventricular pacing as compared with
dual-chamber pacing. Pacemaker Selection in the El-
derly Investigators. N Engl J Med 1998;338(16):1097
1104.
9. Connolly SJ, Kerr CR, Gent M, Roberts RS, Yusuf S,
Gillis AM, Sami MH, Talajic M, Tang AS, Klein GJ,
Lau C, Newman DM. Effects of physiologic pacing versus
ventricular pacing on the risk of stroke and death due
to cardiovascular causes. Canadian Trial of Physiologic
Pacing Investigators. N Engl J Med 2000;342(19):1385
1391.