DAVID
PubMed • Full text
Clinical Question
In patients with an indication for an implantable defibrillator, does obligate dual-chamber pacing improve mortality or hospitalization for heart failure compared to ventricular backup pacing?
Bottom Line
In patients with an indication for an implantable defibrillator, ventricular backup pacing is superior to dual-chamber pacing in terms of hospitalization rate and possibly overall survival.
Major Points
By the early 2000s, on the strength of RCTs such as MADIT-II, SCD-HeFT and AVID, the indications for implantable defibrillators (ICDs) had expanded to include patients both with a history of significant ventricular arrhythmias as well as patients at high risk for developing them due to severe cardiomyopathy. In most of these studies, the effectiveness of ICD therapy in reducing mortality was established using devices with a single lead in the right ventricle. Since these single-chamber ventricular devices cannot sense atrial activity, pacing with these devices can result in simultaneous atrial and ventricular contraction leading to bothersome symptoms associated with backflow of blood from the atria (the pacemaker syndrome). As a result, these devices were usually set to a "backup only" mode whereby pacing would only activate at very low heart rates. Furthermore, out of fear of causing pacemaker syndrome by eliciting bradycardia, patients with CHF often failed to receive optimal doses of beta-blocker and/or antiarrhythmic therapy. As a result, there was a belief that routine placement of dual-chamber ICDs with the capability to provide synchronous atrial and ventricular pacing at a normal heart rate may allow these patients to tolerate more optimal medical therapy for HF with no risk of pacemaker syndrome.
In the 2002 Dual chamber and VVI Implantable Defibrillator (DAVID) trial, 506 patients with indications for ICD therapy but no firm indication for antibradycardia therapy underwent dual-chamber ICD placement. Most patients had severe cardiomyopathy with mean LVEF 27% and were undergoing ICD placement for secondary prevention due to a history of documented VT/VF. Patients were randomized 1:1 to either a dual-chamber pacing mode (DDDR) with a baseline pacing rate of 70 bpm or a single-chamber ventricular pacing mode (VVI) with backup pacing at 40 bpm. At one year, the primary outcome of death or hospitalization favored backup VVI pacing only by a margin of 10% over DDDR pacing. Although not statistically significant, there was also a 4% reduction in mortality with VVI pacing.
The results of DAVID were surprising in that DDDR pacing was expected to produce a benefit in terms of HF morbidity by allowing HF medication dosing to be maximized, and with the support of more physiologic AV sequential pacing. Instead, patients undergoing DDDR pacing appeared to do significantly worse, an effect thought to be caused by the large asymmetry in exposure to RV apical pacing among the treatment arms (3% in VVI arm and 56% in DDDR arm). This was explored further with a post-hoc analysis demonstrating a trend toward a higher rate of the composite outcome when RV pacing exceeded 40%. This was the first trial to provide strong evidence that any form of RV pacing can be detrimental to LV systolic function by induction of electromechanical dyssynchrony with a resultant increase in heart failure incidence. Similar findings would later be discovered in re-analyses of the MOST trial. In addition to motivating later investigations of the harms of RV pacing in studies such as BLOCK-HF, the results of DAVID strongly suggest that all forms of external ventricular pacing should be minimized in patients who have the ability to utilize the intrinsic ventricular conduction system.
Guidelines
As of July 2016, no guidelines have been published that reflect the results of this trial.
Design
- Multicenter, randomized, single-blind, parallel-arm
- N=506 patients with an indication for ICD placement
- Dual-chamber pacing (n=250)
- Single-chamber ventricular pacing (n=256)
- Setting: Multiple centers in the US
- Period: October 2000 - September 2002
- Median follow-up: 8.4 months
- Analysis: Intention-to-treat
- Primary outcome: Freedom from death or hospitalization for heart failure
Population
Inclusion Criteria
Any one of the following:
- VF and LVEF ≤40%
- Syncopal sustained VT and LVEF ≤40%
- Nonsyncopal VT and LVEF ≤40% and 1 of the following:
- Sustained VT, systolic BP <80 mmHg, or significant cardiac symptoms
- NSVT with significant symptoms and EPS-inducible sustained VT or VF
- NSVT (minimal or no symptoms) and EPS-inducible sustained VT or VF
- Out-of-hospital unexplained syncope, heart disease, and EPS-inducible sustained VT or VF and LVEF ≤40%
- EPS-inducible VT or VF within 6 weeks prior to randomization and LVEF ≤40%
Exclusion Criteria
- Permanent pacemaker
- Preexisting endocardial pacing leads
- CABG, PCI, cardiac, or other arrhythmia surgery planned but not yet performed
- Symptomatic bradycardia or second- or third-degree AV block
- Disqualifying atrial fibrillation
- Atrial fibrillation of unknown duration
- Atrial fibrillation of more than 6 months duration
- Atrial fibrillation at the time of randomization
- Need for electrical or chemical cardioversion in the last month
- Frequent, uncontrolled atrial tachyarrythmia
- Awaiting cardiac transplantation
- Condition likely to limit cooperation
- Geographically inaccessible
- Enrolled in a conflicting study
- Prisoner or ward of the state
- Unable to give informed consent
- Life expectancy <1 year
Baseline Characteristics
From the ventricular pacing group.
- Demographics: Mean age 66 years, male 81%
- Arrhythmia: VF 4%, VT 8%, AF 12%
- Co-morbidities: MI 67%, HF 56%, HTN 64%, DM 28%, HLD 58%, CAD 84%, dilated CM 16%
- Cardiac function: Mean LVEF 28%, NYHA I-II 89%
- ICD Indication: VF 19%, syncopal VT 7%, symptomatic sustained VT 16%, symptomatic NSVT with positive EPS 9%, asymptomatic NSVT with positive EPS 25%, unexplained syncope with positive EPS 15%, hemodynamically stable VT 10%
- Medications: ACE/ARB 63%, beta-blocker 55%, CCB 13%, antiarrhythmic 9%
Interventions
- All patients received dual-chamber ICDs
- Patients then randomized 1:1 to dual-chamber pacing at 70bpm (DDDR-70) or ventricular backup pacing at 40bpm (VVI-40)
- Randomization stratified by site, history of HF, and history of AF
- Patients but not investigators were blinded to pacing mode
- All patients received guideline-directed HF therapy with neurohormonal blockade, beta blockade, and aldosterone antagonism
- Hospitalization for heart failure defined by clinical criteria and initiation of inpatient heart failure-directed therapy
- Follow-up occurred every 3 months
Outcomes
Comparisons are VVI vs. DDDR pacing
Primary Outcome
- Freedom from death or hospitalization for heart failure (at 1 year)
- 83.9% vs. 73.3% (HR 1.61; 95% CI 1.06-2.44; P<0.03)
Secondary Outcomes
- HF hospitalization
- 30 (13.3%) vs. 43 (22.6%) [HR 0.65; 95% CI 0.41-1.03; P=0.07]
- Death
- 15 (6.5%) vs. 23 (10.1%) [HR 0.62; 95% CI 0.32-1.19; P=0.15]
Adverse Events
- Early death (30 days post-implant): 0.8% vs. 2.2%
Criticisms
- Trial stopped early due to a trend towards worse outcome with DDDR-70 pacing, limiting further quantification of the effects of DDDR pacing in comparison to backup VVI pacing.
- Vast majority of patients included qualified for ICD based on secondary prevention indications (previous VT/VF). It is unclear whether these results necessarily extend to patients undergoing ICD placement for primary prevention.
- Conflation of pacing mode with pacing amount makes it difficult to draw firm conclusions as to the driver of increased HF morbidity with DDDR pacing (ie, is it DDDR pacing or 70bpm pacing that is harmful?), although corroborative evidence suggests it is the amount of RV pacing that is most critical.
Funding
- Study sponsor St. Jude Medical (manufacturer of dual-chamber ICDs) reviewed manuscript prior to publication