Biventricular Pacing in End-Stage Heart Failure ... - Springer Link

Journal of Interventional Cardiac Electrophysiology 4, 395±404, 2000 #2000 Kluwer Academic Publishers. Manufactured in The Netherlands.

Biventricular Pacing in End-Stage Heart Failure Improves Functional Capacity and Left Ventricular Function Patricia F. Bakker, MD, Huub W. Meijburg, MD, Jaap W. de Vries, MD, Morton M. Mower, MD, Andra C. Thomas, RN, Michael L. Hull, MS, Etienne O. Robles de Medina, MD and Johan J. BredeÂe, MD Heart Lung Institute (P.F.B., H.W.M., E.O.R.d.M., J.J.B.) and Institute of Anesthesiology (J.W.d.V.), University Hospital, Utrecht, The Netherlands, and Guidant=CPI (M.M.M., A.C.T., M.L.H.), St. Paul, MN

Abstract. Background Asynchronous patterns of contraction and relaxation may contribute to hemodynamic and functional impairment in heart failure. In 1993, we introduced biventricular pacing as a novel method to treat heart failure by synchronous stimulation of the right and left ventricles after an appropriate atrioventricular delay. The objectives of this study were to assess the early and long-term effects of this therapy on functional capacity and left ventricular function in patients with severe heart failure and left bundle branch block. Methods and Results Twelve patients with end-stage congestive heart failure, sinus rhythm and complete left bundle branch block were treated with biventricular stimulation at optimized atrioventricular delay. The NYHA functional class and maximal bicycle exercise capacity were assessed. Systolic and diastolic left ventricular function were studied with echocardiography and radionuclide angiography. Data was collected at various intervals during 1-year follow-up. Cumulative survival [95% CI] was 66.7% [40.0,93.4] at 1 year and 50 % [21.8, 78.2] at 2 and 3 years. Median NYHA class improved from class IV to class II at 1 year (p 0.008). After 6 weeks an increase in exercise capacity occurred, which was sustained. A less restrictive left ventricular ®lling pattern, an increase in dP=dt and left ventricular ejection fraction, and a decrease in mitral regurgitation were observed early and long-term. Conclusions Biventricular pacing at optimized atrioventricular delay results in improvement in functional capacity, which is associated with improved systolic and diastolic left ventricular function, and a decrease in mitral regurgitation during short- and long-term follow-up. Key Words. biventricular pacing, heart failure, functional capacity, hemodynamics

Electrical and mechanical nonuniformity in the heart can be associated with a signi®cant degree of asynchrony and can have a deleterious effect on cardiac function [1±3]. Decreased myocardial performance in various pathophysio-

logical conditions may be induced or aggravated by the occurrence of asynchrony [1,2,4]. Recent progress in cardiac pacing therapy has addressed the issue of improving atrioventricular, inter- and intraventricular asynchrony in order to improve cardiac performance and functional status [5]. This may be particularly important for patients with advanced stages of congestive heart failure, who are not amenable to medical therapy and not eligible for cardiac transplantation. The surface electrocardiogram in patients with severe congestive heart failure frequently shows prolongation of the PR interval and widening of the QRS complex. Several of these patients have left bundle branch block, which results in asynchronous left ventricular activation and has various deleterious effects on systolic and diastolic left ventricular function [3,6,7]. We hypothesized that mechanical atrioventricular, inter- and intraventricular asynchrony could be improved in those patients by simultaneously pacing the right and left ventricles after an appropriate atrioventricular (AV) delay. We also assumed that improvement of asynchrony would improve left ventricular function and functional capacity. Based on these hypotheses, we introduced a new pacing therapy for heart failure by implantation of a dual-chamber pacemaker with right and left ventricular electrodes [8]. The purpose of

This study was supported by Guidant=CPI, St. Paul, MN Address for correspondence: Patricia Bakker, MD. Heart Lung Institute, University Hospital, Heidelberglaan 100, 3584 CX Utrecht, The Netherlands. PO box 85500, 3508 GA, Utrecht, The Netherlands. Phone: 31 30 2506179 Fax: 31 30 2505415; E-mail: [email protected] Received 1, July 1999; accepted 10, December 1999 395

396

Bakker et al.

this study was to examine the early and longterm effects on functional capacity and systolic and diastolic left ventricular function in patients with end-stage congestive heart failure and left bundle branch block. Limited availability of donor hearts and high mortality in this patient group creates an urgent need for therapeutic alternatives [9].

Methods Patient Selection Starting with the ®rst implant in March 1993, biventricular pacing at an optimized AV delay was offered as a last resort to 12 consecutive patients with end-stage congestive heart failure who were rejected for cardiac transplantation because of contraindications according to the general guidelines and standard practice [10]. They had to ful®ll the following criteria: age 18 to 75 years, a history of chronic congestive heart failure, New York Heart Association (NYHA) functional class III or IV despite optimal medical therapy, sinus rhythm, and a wide QRS complex (>120 ms) with left bundle branch block morphology. The study was designed to only select patients with left bundle branch block to ensure the presence of inter- and intraventricular asynchrony with the aforementioned detrimental effects on hemodynamics, and to allow comparability of the ventricular conduction defect in this series. Patients with correctable causes of heart failure, atrial ¯utter, atrial ®brillation or unstable angina pectoris were excluded. The Ethics Committee of the University Hospital, Utrecht, The Netherlands approved the study protocol. Prior to enrollment in the study all patients gave written informed consent. Patients were enrolled over a 2-year period. Pacemaker Implantation Procedure Surgical procedure A dual-chamber pacemaker model Delta TRS 937 (3 patients) or Vigor DDD 950 (9 patients) with associated leads and accessories, Guidant=CPI, St. Paul, MN, was implanted under general anesthesia. To obtain stable anaesthesia for hemodynamic evaluation, anaesthesia was induced with sufentanil 2 mcg=kg and propofol 1 mg=kg, and maintained with continuous infusion of sufentanil 0.3 mcg=kg=hour and propofol 1±2 mg=kg=hour. Heating blankets were used to maintain normothermia. A unipolar active ®xation lead or tined lead (Model 4169 or 4161) was transvenously positioned in the right ventricular apex, and a bipolar active ®xation lead (Model 4269) in the right atrial appendage. Thereafter, a left anterolateral minithoracotomy ( 6 cm) was

performed in the sixth intercostal space. The pericardium was opened and a unipolar screwin electrode (Model 4320) was positioned low on the anterolateral wall of the left ventricle. Accepted acute pacing stimulation thresholds (at 0.5 ms pulse width) for passive and active ®xation leads were less than 1.0 V and 1.5 V respectively. The left ventricular electrode was tunneled to the pacemaker pocket created in the left pectoral area. To connect the two ventricular leads to the ventricular channel of the pacemaker, a unipolar Y-connector (Model 6836) or a bipolar adapter (Model 6021) was used. Hemodynamic monitoring and pacing protocol At implantation, an arterial line was introduced in the radial artery and a 7.5 F Swan-Ganz thermodilution catheter (Edwards Laboratories) was inserted through the right internal jugular vein. At the end of the surgical procedure, the following parameters were measured during the pacing protocol and are listed in the order in which they were used to select the optimal pacemaker settings: cardiac output (CO), mean pulmonary capillary wedge pressure (PCWP), pulmonary artery pressure (PAP), right ventricular ejection fraction (RVEF), systolic arterial blood pressure (SBP), pulse pressure (PP), and mean right atrial pressure (RAP). To determine the optimal AV delay, testing was performed in each patient in sinus rhythm and with AV delays of 100 and 150 ms while stimulating both ventricles simultaneously in the VDD mode. Some patients received additional testing with shorter or longer AV delays (25±250 ms). The various pacemaker settings were programmed in a random order and data was collected after 5 minutes of stabilization in each setting. Before each reprogramming, baseline rhythm was maintained for 5 minutes. In the event sinus bradycardia and deterioration of hemodynamics occurred during anesthesia, AAI pacing was applied in lieu of intravenous inotropes. In these patients, hemodynamic data collected during AAI pacing was used as baseline, and pacing was performed in the DDD mode. For de®nitive programming of the atrial sensed AV delay at the end of the procedure, 50 ms was deducted from the optimal atrial paced AV delay to compensate for the increased interatrial conduction time during atrial stimulation [11]. Pre- and Postoperative Studies For the purpose of this study, patients were observed and data was collected for 1 year postimplant. However, clinical data was collected through 3 years follow-up. Prior to pacemaker implantation the following studies were per-

Biventricular Pacing Improves Heart Failure

formed and repeated at 2 weeks, 6 weeks, 3 months, 6 months, and 1 year follow-up, unless indicated otherwise. The investigators involved with clinical follow-up and the investigators analyzing test results were unaware of each other's ®ndings. Assessments of Functional Capacity and Diuretic Dose NYHA functional class and medication dosages were noted at each follow-up visit. Heart failure medications were not adjusted postoperatively with the exception of diuretics if necessary. This clinical data continued to be collected from baseline through 3 years postimplant. Maximal bicycle exercise capacity was assessed starting with 3 minutes of unloaded pedaling, followed by 10 W increments of work rate every 2 minutes. Breath-by-breath gas exchange measurements were performed with an Oxycon Sigma (version 2.00) gas analyzer (Erich Jaeger GmbH, HoÈchberg, Germany). Ergometry data was analyzed for peak oxygen consumption (VO2 ), VO2 at anaerobic threshold (AT), and exercise duration. The V-slope method was used to determine the point at which anaerobic threshold was attained [12]. The respiratory quotient had to be 1.0 for the test to be included in the analysis. Exercise tests were not performed at 2 weeks follow-up.

Electrocardiographic and Holter Studies Resting 12-lead surface electrocardiograms were obtained and analyzed for heart rate, PR interval, QRS width, QRS morphology and electrical axis, and pacing mode. Twenty-four-hour Holter studies were obtained prior to pacemaker implantation to assess the presence or absence of atrial ®brillation and ventricular arrhythmias. During in-hospital stay, the presence of biventricular stimulation was evaluated by telemetry. Pacemaker function and pacing mode were checked predischarge and at all follow-up intervals. In patients with high pacing thresholds, additional assessments were performed as appropriate. Echocardiography A complete M-mode, two-dimensional and Doppler transthoracic echocardiographic study was performed in each patient by means of a Sonos 1000 ultrasound system (Hewlett Packard, Andover, MA), using 2.5 and 3.5 mHz transducers, with the patient in a left lateral recumbent position. M-mode measurements were performed according to the recommendations of the American Society of Echocardiography [13]. Stroke volume was calculated as the product of the left ventricular out¯ow tract (LVOT) area and velo-

397

city time integral (VTI, cm) of transaortic ¯ow. Doppler transmitral ¯ow velocity was recorded and early (E, cm=s) and late (A, cm=s) diastolic ¯ow velocities, E=A ratio, deceleration time of E and diastolic ®lling period (DFP) were obtained [14]. Isovolumic relaxation time (IVRT) was measured using pulsed Doppler [14]. The E and A velocities, E=A ratio, DTE and IVRT were used to recognize patterns of abnormal left ventricular ®lling, estimate ®lling pressures, and report severity of diastolic dysfunction as recommended [14]. Interpretations were made correcting for age [14]. From the continuous wave Doppler signal of the mitral regurgitation jet obtained with a 1.9 mHz non-imaging probe, positive dP=dt was calculated, and mitral regurgitation duration was determined [15]. Using color Doppler imaging, mitral regurgitation jet area was graded as the ratio of maximal regurgitant jet area to left atrial area, as described by Helmcke and colleagues [16]. The images were stored on VHS-videotape for off-line analysis by an experienced reader blinded to knowledge of the study protocol or patient data. Measurements of all indices were carried out on a minimum of 5 consecutive cardiac cycles, and the average value was used in the analysis. Intraobserver variability was determined from 10 randomly selected studies. The only differences were found for left ventricular end-systolic dimension and diastolic ®lling period, mean difference [95% CI] 1.1 mm [0.2,2.0] and 6.5 ms [4.0,9.0] respectively. However, the mean variability was very small as a percentage of the mean for any of the measurements tested. Technically adequate Doppler recordings of pulmonary venous ¯ow could not be obtained in the majority of patients, and this data was therefore not analyzed. Echocardiographic studies were not performed at 6 weeks follow-up. Nuclear Studies Multiple gated acquisition radionuclide angiography with technetium-99m-labeled red blood cells was performed at rest to measure left ventricular ejection fraction (LVEF) [17]. This data was not collected at 2 and 6 weeks after implantation. Statistical Analysis Atrial-triggered biventricular pacing was evaluated for both short-term and long-term effects. Descriptive statistics (mean standard deviation) were used to present data at baseline. In evaluating changes over time, mean changes with 95% con®dence interval [CI] were used to determine the magnitude of the effect and the statistical estimate. The mean and standard error were used to construct the con®dence

398

Bakker et al.

limits. One-, two-, and three-year survival is presented as a percentage with 95% CI. The Sign test was used to test changes in median NYHA class for statistical signi®cance. Changes in all continuous variables and differences in acute hemodynamic improvement between AV delays of 100 and 150 ms were tested for statistical signi®cance by the Wilcoxon sign-rank test. Linear regression analysis was used to evaluate the relation between baseline PR interval and optimal AV delay. The relations between the improvement in variables for cardiac function and functional capacity and the PR interval and the QRS width at baseline were also studied by linear regression analysis. The overall level of probability for statistical signi®cance was set to p 0.05, and from there was corrected for multiple comparisons when appropriate. Changes at two follow-up periods were analyzed for statistical signi®cance (p 0.025): 3 months representing the early effects and 1 year for long-term effects.

Results Patient Characteristics The patients (5 men, 7 women) had a mean age of 64.4 5.7 years. Heart failure occurred after myocardial infarction in 4 patients, and was due to nonischemic heart disease in 8 patients. The duration of symptoms was 4.7 1.3 years. Eight patients were in NYHA class IV and 4 in class III. Left ventricular ejection fraction measured 15 6 %. The PR interval was 217 20 ms and the QRS width 194 21 ms with left bundle branch block morphology and an electrical axis of ÿ 15 to ÿ 60 . All patients were in sinus rhythm and no episodes of atrial ®brillation were recorded on preoperative Holter registrations. However, multifocal premature ventricular contractions were recorded in 10 patients, couplets in 7, and nonsustained ventricular tachycardia episodes in 6. One patient had a history of sustained ventricular tachycardia (VT) and syncope, and 2 patients had been resuscitated from out-of-hospital cardiac arrest. Implantation of a cardioverterde®brillator was considered in the latter 3 patients, but not performed in view of their poor heart failure status without being eligible for cardiac transplantation and the unknown ef®cacy of the therapy under evaluation. The standard aim of medical treatment was a combination of angiotensin-converting enzyme inhibitors, digoxin, and diuretics. However, this combination was not possible in all patients due to development of serious side effects and resulted in angiotensin-converting enzyme inhibitors in 8 patients, other vasodilators in 9 patients, digoxin

in 11 patients, and diuretics in 12 patients. Amiodarone was indicated to control either atrial or ventricular arrrhyhtmias in 7 patients. Optimization of AV Delay At baseline, CO measured 3.7 0.6 L=min, and PCWP 16 7 mm Hg. A maximal increase in CO of 0.6 L=min [0.4,0.8] with a decrease in PCWP of 2 mm Hg [1,4] was found during biventricular stimulation. In 8 patients the largest hemodynamic improvement was measured at an AV delay of 100 ms, in 3 patients at 150 ms, and in 1 patient at 25 ms. In the latter patient, the lowest device setting for AV delay of 70 ms yielded values approximate to the hemodynamic optimum and was selected. At the conclusion of the pacing study, mean RAP was 6.8 2.6 mm Hg versus 7.0 2.8 mm Hg at baseline, mean PCWP 15.5 6.7 mm Hg versus 15.3 6.9 mm Hg, and heart rate 72.4 13.1 beats=min versus 73.4 14.3 beats=min, indicating that loading conditions and heart rate were unchanged after completion of the protocol. Differences in hemodynamic improvement between AV delays 100 and 150 ms were statistically signi®cant. However, the magnitude of these differences showed a large variation among patients. The following differences were measured: CO, 0.4 0.4 (p 0.01); PCWP, 1.5 1.7 mm Hg (p 0.01); RVEF, 3.6 2.7% (p 0.02);. SBP, 5.2 2.9 mm Hg (p 0.01); and PP, 3.5 2.1 (p 0.01). Regression analysis of baseline PR interval on optimal AV delay showed a p value of 0.10. The correlation coef®cient r was 0.50, indicating a moderate positive correlation. Outcome and Adverse Effects Follow-up data were available for all 12 study patients. Six patients died during 3-year followup. Cumulative survival [95% CI] was 66.7% [40.0, 93.4] at 1 year, and 50% [21.8, 78.2] at 2 and 3 years. Early deaths. Two patients died within 30 days. Out-of-hospital witnessed sudden death occurred at 4 weeks in 1 patient with preoperative sustained ventricular tachycardia and syncope, who otherwise responded well to the treatment. The other patient died in-hospital 30 days after pacemaker implantation. After clinical improvement during the ®rst week, heart failure was aggravated by long-lasting episodes of monomorphic VT, resistant to antiarrhythmic drugs. The same VT morphology (nonsustained) had been diagnosed on preimplant Holter registrations.


Late deaths. Late witnessed sudden death occurred at 8 months in 1 patient with nonsustained VT preimplantation. Three patients died from recurrence and progression of congestive heart failure at 11, 16, and 21 months respectively. In 2 of them hemodynamic and clinical deterioration were related to episodes of loss of pacing. This condition was caused in 1 patient by the patient's heart rate exceeding the upper rate limit, which was treated by adapting pacemaker settings. In the other patient refractory atrial ®brillation resulted in acute hemodynamic deterioration and subsequent death.

399

substantially lower from 2 weeks through 6 months (Table 1). The decrease was signi®cant at 3 months. At 2- and 3-years follow-up, the dose of furosemide had changed by 7195 mg=day [7395, 5] and 7188 mg=day [7392, 16] respectively. Table 2 summarizes the exercise test results for the entire study group. Individual patient data are depicted in Figure 1. Substantial improvements were found at all follow-up times. Peak VO2 , VO2 AT and exercise time increased in 9 out of 10 patients at 6 weeks. Increases in all 3 variables were largely maintained through 1year follow-up. At 1 year, 7 of 8 patients had a higher peak VO2 , VO2 AT, and a longer exercise time than prior to implantation. The changes in all three variables were statistically signi®cant at 3 months follow-up. At 1 year, only VO2 AT was signi®cantly increased. There was no correlation between improvement in the above variables and the PR interval and QRS width at baseline.

Pacing thresholds. Rising left ventricular stimulation thresholds resulted in repeated loss of capture of the left ventricle in 2 patients between 1 and 3 months. Acute congestive heart failure symptoms recurred during those episodes. Pacemaker reprogramming obtained reinstitution of biventricular stimulation and concomitant symptomatic improvement. After 6 months, one of these patients had continuous loss of left ventricular capture which was solved by replacement of the bipolar adapter with a unipolar Yconnector. However, the patient's clinical condition showed little improvement at that time. In 10 patients loss of capture of the left ventricle was not an issue. At 1 year, left ventricular stimulation threshold measured 2.4 V [2.0, 2.8] at 0.5 ms pulse width.

Electrocardiographic Results Resting heart rate decreased by a minimum of 8 bpm between 6 weeks and 6 months (Table 3). QRS width decreased immediately after implantation and remained lower through 6 months follow-up. The early and long-term effects were not statistically signi®cant. Echocardiographic and Nuclear Study Results Cardiac dimensions. The left atrium and left ventricle were signi®cantly enlarged at baseline. Left atrial end-systolic diameter measured 52 7 mm, left ventricular end-diastolic diameter 79 14 mm and left ventricular end-systolic diameter 71 13 mm. These dimensions remained unchanged during follow-up. Table 4 shows the echocardiographic test results at 1year follow-up.

Changes in Functional Capacity and Diuretic Dose Table 1 shows NYHA functional class for the patient group during 1-year follow-up. Median NYHA class improved by at least 1 class at all time periods. The effects at 3 months and 1 year were statistically signi®cant. At 2- and 3-years follow-up, median NYHA class was 2 (p 0.03) and 2.5 (p 0.06) respectively. The dose of furosemide required maintaining hemodynamic stability after implantation was

Diastolic left ventricular function. In 6 patients, preimplant transmitral ¯ow velocity showed a single ®lling wave pattern with extre-

Table 1. Median NYHA Class and Changes in Diuretic Dose over One Year Variable Median NYHA Furosemide (mg=day)

Preimplant

2 Weeks

6 Weeks

3 Months

6 Months

1 Year

4

3

2.5

2.25

564 706

ÿ 396 [ÿ 759,ÿ 33]

ÿ 455 [ÿ 893,ÿ 17]

3 p 0.004 ÿ 351 [ÿ 613,ÿ 134] p 0.01

2 p 0.008 ÿ 226 [ÿ 490,43] p 0.09

ÿ 267 [ÿ 468,ÿ 66]

Median NYHA class is presented with the corresponding p values for the differences compared to preimplant. Dose of furosemide at baseline is presented as mean SD. Dose changes at follow-up are mean with 95% CI [ ], with the corresponding p values for early and long-term effects.

400

Bakker et al.

Table 2. Pre- and Postoperative Ergometry Results Variable

Preimplant Mean

6 Weeks Change

3 Months Change

6 Months Change

1 Year Change

10.9 3.8

2.0 [0.5,3.5]

2.6 [ÿ 0.2,5.4]

VO2 AT

8.2 2.3

1.6 [0.4,2.8]

Exercise time (min)

7.9 3.9

3.5 [1.1,5.9]

3.1 [0.9,5.3] p 0.01 2.6 [1.2,4.0] p 0.01 3.4 [1.0,5.8] p 0.02

2.7 [0.1,5.3] p 0.1 3.1 [1.4,4.8] p 0.01 2.2 [ÿ 0.8,5.2] p 0.15

Peak VO2 (mL=min=kg)

(mL=min=kg)

3.3 [1.4,4.7] 2.0 [ÿ 2.0,6.0]

VO2 indicates oxygen uptake; AT, anaerobic threshold. Baseline values are presented as mean SD and changes as mean with 95% CI [ ] and the corresponding p values for early and long-term effects.

Fig. 1.

Changes in peak VO2 in the individual patients during 1-year follow-up. Curves a, b, c, and d indicate patients

who died. Curves c and e indicate the patients with loss of left ventricular capture and curve d the patient with loss of pacing at 11 months.

Table 3. Changes in Electrocardiographic Measurements during One Year Variable

Preimplant Mean 2 Weeks Change 6 Weeks Change 3 Months Change 6 Months Change 1 Year Change

Heart rate (bpm)

89 17

ÿ 10 [ÿ 21,1]

ÿ8 [ÿ 15,ÿ 1]

QRS width (ms)

194 21

ÿ 21 [ÿ 36,ÿ 6]

ÿ 23 [ÿ 37,ÿ 90 ]

ÿ 12 [ÿ 21,ÿ 3] p 0.03 ÿ 18 [ÿ 37,ÿ 5] p 0.05

ÿ9 [ÿ 16,ÿ 2] ÿ 21 [ÿ 35,ÿ 7]

ÿ8 [ÿ 17,1] p 0.1 ÿ 19 [ÿ 37, ÿ 1] p 0.08

Baseline values are presented as mean SD, changes are mean with 95% CI [ ] with the corresponding p values for early and longterm effects.

mely short DFP (161 54 ms), and an abnormally short IVRT 78 8 ms). These ®ndings are consistent with a severely restricted left ventricular ®lling physiology and high ®lling pressures. Transmitral ¯ow pattern changed: distinct E

and A waves appeared, indicating a less restrictive ®lling pattern with lower ®lling pressures. Simultaneously, DFP increased to 326 ms [210, 442] at maximum improvement within the ®rst 3 months.


Table 4. Changes in Echocardiographic Measurements at One-Year Variable

Preimplant Mean

1 Year Change

LAES (mm)

52 7

LVES (mm)

71 13

LVED (mm)

79 14

Peak E (cm=s)

89 10

Peak A (cm=s)

62 35

E=A

2.1 1.7

DFP (ms)

270 132

SV (mL)

37 9

ÿ 0.4[ÿ 3.4, 2.6] p 0.9 ÿ 7[ÿ 12,ÿ 2] p 0.06 ÿ 4[ÿ 8,0] p 0.2 ÿ 32[ÿ 46,ÿ 18] p 0.06 5[ÿ 31,41] p 1.0 ÿ 0.2[ÿ 1.4,1.0] p 0.8 20[ÿ 74,114] p 1.0 5[0,10] p 0.05 ÿ 0.12[ÿ 0.20, ÿ 0.03] p 0.03 243[113,373] p 0.03 ÿ 28[ÿ 111,55] p 0.6 20[ÿ 8,48] p 0.2 56[34,78] p 0.008

2

MRJA=LAA (cm )

0.41 0.16

dP=dt (mm Hg=s)

466 130

MR duration (ms)

420 168

IVRT (ms)

90 18

DTE (ms)

99 30

Baseline values are presented as mean SD, changes are mean with 95% CI [ ] and corresponding p values. LAES indicates left atrial end-systolic dimension; LVES, left ventricular end-systolic dimension; LVED, left ventricular enddiastolic dimension; E, peak early ®lling velocity; A, peak late ®lling velocity; DT E, deceleration time of E; DFP, diastolic ®lling period; IVRT, isovolumic relaxation time; SV, stroke volume; MRJA, maximal mitral regurgitant jet area; LAA, left atrial area; MR, mitral regurgitation.

In the remaining group of 6 patients, having a mitral in¯ow velocity pattern with separate E and A waves, 4 patients had a restrictive ®lling pattern at baseline (E=A ratio > 2, DTE < 120 ms),

401

and in 2 patients, pseudonormalization was probably present. This pattern changed during followup to a ®lling pattern of predominantly delayed relaxation (E=A < 2, DTE > 150 ms). In these 6 patients, DFP measured 381 74 ms at baseline and did not change during follow-up. Systolic left ventricular function. Data are summarized in Table 5. A substantial increase in dP=dt occurred in all patients through the entire follow-up. Stroke volume increased from baseline at 3 months and 1 year. An increase in LVEF was measured at 6 months and 1 year. Changes in parameters for systolic left ventricular function were not statistically signi®cant. Mitral regurgitation. Mitral regurgitation jet area decreased after implant which was statistically signi®cant at short-term follow-up (Table 6). Mitral regurgitation duration decreased by a minimum of 20 ms, which was not signi®cant. Regression analysis. No correlation of PR interval and QRS width at baseline with the above variables for cardiac function was found.

Discussion The main ®ndings of this prospective study are the improvements in functional capacity, systolic and diastolic left ventricular function, as well as the decrease in mitral regurgitation. The magnitude of these improvements was clinically relevant and sustained in the majority of patients during follow-up. Loss of biventricular pacing was associated with the development of increasing symptoms of congestive heart failure and supports the potential bene®cial effects of biventricular pacing in these patients. Repetition of such episodes had progressive detrimental effects, particularly in patients with NYHA func-

Table 5. Changes in Systolic Left Ventricular Function during One Year Variable dP=dt (mm Hg=s)

Preimplant Mean

2 Weeks Change

3 Months Change

6 Months Change

1 Year Change

466 130

200 [87,287]

179 [91,267] p 0.03 11 [0,22] p 0.05 6 [ÿ 1,13] p 0.05

210 [92,328]

243 [113,373] p 0.03 5 [0,10] p 0.05 5 [1,10] p 0.13

SV(mL)

37 9

1 [ÿ 3,5]

LVEF (%)

15 6

N=A

4 [ÿ 3,11] 8 [1,15]

SV, stroke volume; LVEF, left ventricular ejection fraction. Baseline values are presented as mean SD, changes are mean with 95% CI [ ], and p values for early and long-term effects are shown.

402

Bakker et al.

Table 6. Changes in Mitral Regurgitation during One-Year Follow-up Variable

Preimplant Mean

2 Weeks Change

3 Months Change

6 Months Change

1 Year Change

MRJA=LAA (cm2)

0.41 0.16

ÿ 0.26 [ÿ 0.26,ÿ 0.25]

ÿ 0.10 [ÿ 0.19,ÿ 0.08]

MR duration (ms)

420 168

ÿ 55 [ÿ 3,ÿ 107]

ÿ 0.14 [ÿ 0.21,ÿ 0.07] p 0.007 ÿ 20 [ÿ 87,47] p 0.8

ÿ 0.12 [ÿ 0.2,ÿ 0.03] p 0.03 ÿ 28 [ÿ 111,55] p 0.6

ÿ 20 [ÿ 103,63]

MRJA, maximal mitral regurgitant jet area; LAA, left atrial area; MR, mitral regurgitation. Baseline values are presented as mean SD, changes are mean with 95% CI [ ], and p values for early and long-term effects are shown.

tional class IV. There were no deaths related to the surgical procedure. The incidence of sudden death in our series is in accordance with that reported in the literature for comparable patient populations [9,18]. The observed improvements in functional capacity and left ventricular function supported our hypotheses. The main mechanisms that contributed in the present study to improved diastolic function and reduced ®lling pressures are presumably: less asynchronous relaxation patterns, decreased mitral regurgitation, and improved diastolic ®lling times. All of which are known to affect relaxation, chamber stiffness and or ®lling pressures [14,19]. Improvement of systolic ventricular function may be the result of a more normal activation and contraction sequence [20]. Optimized mechanical atrioventricular synchrony, probably also plays a role [19]. This is supported by the signi®cant differences in hemodynamic improvement at different AV delays. The last decade has witnessed a growing interest in the therapeutic potential of arti®cial cardiac stimulation to treat congestive heart failure [5]. As yet, a limited number of studies has been published on the acute and early effects of biventricular stimulation [20±28]. Comparison of the results is hampered by differing baseline patient characteristics in these studies with respect to NYHA functional class, degree of left ventricular dysfunction, atrioventricular conduction delay, type of ventricular conduction disturbances, and differing sites of left ventricular stimulation. Auricchio et al. and Kass and coworkers showed acute improvement of global systolic function, and resequencing of segmental wall motion of the left ventricle was demonstrated by Saxon and coworkers [20, 23±25]. These data are in agreement with our premise. The acute effects on hemodynamics as well as early effects on NYHA functional class and QRS duration observed in the present study are in agreement with the ®ndings of others [26±28]. However, PCWP (16 7 mm Hg) at baseline in our study was lower than in the other two series

(31 10 mm Hg; 27 7 mm Hg) reporting on acute changes in PCWP during biventricular stimulation [26,27]. Acute hemodynamic measurements in the present study were performed under anesthesia, whereas in the other studies measurements were performed in conscious patients. In the present study, anesthetic drugs were selected that had no cardiodepressive effect, prevented release of endogenous catecholamines by stress (as may happen in conscious patients) and provided stable anesthesia for hemodynamic measurements. These drugs decrease ®lling pressures, such that an underestimation of the magnitude of the effect of pacing on PCWP may occur. The percentage change is not affected. In one series, repetitive loss of capture of the left ventricle was reported to result in recurrence of severe heart failure symptoms [27]. These symptoms were resolved by reinstitution of biventricular stimulation, which is consistent with our observations. Preliminary results of the InSync study, evaluating the feasibility of a transvenous approach to implant leads in the cardiac veins, showed an 84% success rate of left ventricular lead implantation [28]. The leads were implanted in widely differing positions, mainly because of anatomical constraints. This may represent a limitation of the transvenous approach, since it is presumably important to be able to select a pacing site close to the area of delayed activation [25]. Although this approach is less invasive, the number of early deaths did not differ from the present study, and no surgery-related complications were observed in either studies.

Limitations This study, albeit prospective, does not involve a randomized control group. The study has a small sample size, which limits the ability to detect signi®cant differences. In most cases, changes represented clinical signi®cance due to their magnitude and general sustainability and appli-


cation across the patient population. Because of the uncontrolled nature of the study a potential placebo effect of device implantation can be present. The results were obtained in end-stage heart failure patients, such that the results may not be applicable to patients with less severe cardiac dysfunction. In addition, the QRS complex was very wide in this series, which implies a highly abnormal pattern of activation of the left ventricle. It is unknown whether biventricular stimulation in patients with a much less abnormal QRS complex will have the same bene®cial effects. During implant testing, biventricular stimulation was performed in the VDD mode except in patients with low heart rates during anaesthesia. In the latter situation, comparisons were made with AAI pacing as baseline and biventricular pacing in the DDD mode. This approach was considered preferable over administration of inotropic drug support. One pacing site in the left ventricle was used. This pacing site was selected based upon literature suggesting that it represented potential for optimal hemodynamic responses and upon the assumption that pacing in an area of abnormally delayed activation would improve synchronous contraction and relaxation [27]. It is unknown whether other or additional sites would have been more favorable in patients with highly abnormal electrical activation patterns. Adaptation of lead systems and pacemaker programming during the initial part of the study was necessary due to the fact that existing technology and devices for bradycardia pacing were used for biventricular pacing in heart failure. Underestimation of the bene®cial effects of biventricular pacing is likely as a result of deterioration of clinical condition in those patients. Conclusions and Clinical Implications These preliminary study results suggest that biventricular pacing at optimized AV delay results in substantial early and long-term improvements in functional capacity, systolic and diastolic left ventricular function, as well as decreased severity of mitral regurgitation in patients with end-stage congestive heart failure, sinus rhythm, wide QRS and left bundle branch block. The implant procedure can be performed with an acceptable risk. Important issues are 1) the prevention of loss of biventricular stimulation by adaptation of lead systems and development of pacemaker technology for pacing to treat congestive heart failure, and 2) the combination with a cardioverter-de®brillator in selected patients. Large randomized controlled studies are needed for evaluation of the future role of biventricular

403

pacing in heart failure and a better understanding how pacing affects mechanisms that play a role in congestive heart failure. This will allow optimization of the therapy and identi®cation of patients who are likely to bene®t most.

Acknowledgment We thank Johanna G. van der Bom, MD, PhD, for advice on the statistical analysis. Guidant=CPI, St. Paul, MN, supported this study.

References 1. Heyndrickx GR, Paulus WJ. Effect of asynchrony on left ventricular relaxation. Circulation 1990;81 (Suppl III):III-41±III-47. 2. Brutsaert DL. Nonuniformity: a physiologic modulator of contraction and relaxation of the normal heart. J Am Coll Cardiol 1987;9:341±348. 3. Grines CL, Bashore TM, Boudoulas H, Olson S, Shafer P, Wooley CF. Functional abnormalities in isolated left bundle branch block. The effect of interventricular asynchrony. Circulation 1989;79:845±853. 4. Burkhoff D, Oikawa RY, Sagawa K. In¯uence of pacing site on canine left ventricular contraction. Am J Physiol 1986;H428±H435. 5. Bakker PFA. Cardiac stimulation as nonpharmalogical treatment for heart failure. In: Van Hemel NM, Wittkampf FHM and Ector H, eds. The pacemaker clinic of the 90's. Dordrecht, The Netherlands: Kluwer Academic Publishers, 1995; 185±197. 6. Bramlet DA, Morris KG, Coleman RE, Albert D, Cobb FR. Effect of rate dependent left bundle branch block on global and regional left ventricular function. Circulation 1983;7:1059±1065. 7. Xiao HB, Lee CH, Gibson DG. Effect of left bundle branch block on diastolic function in dilated cardiomyopathy. Br Heart J 1991;66:443±447. 8. Bakker P, de Jonge N, KloÈpping C, Meijburg H, Barmen 't Loo C, Wittkampf F, van Mechelen R, Mower M, Thomas A. Biventricular pacing in congestive heart failure. Clinical Research 1994;42:327A. 9. Stevenson LW, Couper G, Natterson B, Fonarow G, Hamilton MA, Woo M, Creaser JW. Target heart failure populations for newer therapies. Circulation 1995;92 [suppl II]:II-174-II-181. 10. Hunt SA. Cardiac transplantation: the 24th Bethesda Conference. J Am Coll Cardiol 1993;22:1±64. 11. Reynolds DW. Hemodynamics of cardiac pacing. In: Ellenbogen KA, ed. Cardiac Pacing. Cambridge, Massachusetts: Blackwell Science; 1996:145±146. 12. Beaver WL, Wasserman K, Whipp BJ. A new method for detecting anaerobic threshold by gas exchange. J Appl Physiol 1986;60:2020±2027. 13. Sahn DJ, DeMaria A, Kisslo J, Weyman AE. Recommendations regarding quantitation in M-mode echocardiography: Results of a survey of echocardiographic measurements. Circulation 1978;58:1072±1084. 14. Rakowski H, Appleton C, Chan KL, Dumesnil JG, Honos G, Jue J, Koilpillai C, Lepage S, Martin RP, Mercier LA, O'Kelly B, Prieur T, San®lippo A, Sasson Z, Alvarez N, Pruitt R, Thompson C, Tomlinson C.

404

15.

16. 17.

18.

19.

20.

21.

Bakker et al.

Canadian consensus recommendations for the measurement and reporting of diastolic dysfunction by echocardiography. J Am Soc Echocardiogr 1996;9:736±760. Bargiggia GS, Bertucci C, Recusani F, Raisaro A, de Servi S, Valdes-Cruz LM, Sahn DJ, Tronconi L. A new method for estimating left ventricular dP=dt by continuous wave Doppler-echocardiography; Validation studies at cardiac catheterization. Circulation 1989;80:1287±1292. Helmcke F, Nauda NC, Hsiang MC. Color Doppler assessment of mitral regurgitation with orthogonal planes. Circulation 1987;75:175±183. Thrall JH, Freitas JE, Swanson D, Rogers WL, Clare JM, Brown ML, Pitt B. Clinical comparison of cardiac blood pool visualization with technetium-99m red blood cells labeled in vivo and with technetium-99m human serum albumin. J Nucl Med 1978;19:796±803. Middlekauff HF, Stevenson WG, Stevenson LW, Saxon LA. Syncope in advanced heart failure: high risk of sudden death regardless of origin of syncope. J Am Coll Cardiol 1993;21:110±116. Nishimura RA, Hayes DL, Holmes DR, Tajik AJ. Mechanisms of hemodynamic improvement by dualchamber pacing for severe left ventricular dysfunction: an acute Doppler and catheterization hemodynamic study. J Am Coll Cardiol. 1995;25:281±288. Saxon LA, Kerwin WF, Cahalan MK, Kalman JM, Olgin JE, Foster E, Schiller NB, Shinbane JS, Lesh MD, Merrick SH. Acute effects of intraoperative multisite ventricular pacing on left ventricular function and activation=contraction sequences in patients with depressed ventricular function. J Cardiovasc Electrophysiol 1998;9:13±21. Cazeau S, Ritter P, Bakdach S, Lazarus A, Limousin M, Henao L, Mundler O, Daubert JC, Mugica J. Four chamber pacing in dilated cardiomyopathy. PACE 1994;17(Pt. II):1974±1979.

22. Foster AH, Gold MR, McLaughlin JS. Acute hemodynamic effects of atrio-biventricular pacing in humans. Ann Thorac Surg 1995;59:294±300. 23. Auricchio A, Salo RW. Acute hemodynamic improvement by pacing in patients with severe congestive heart failure. PACE 1997;20 (Pt.1):313±324. 24. Kass DA, Chen C-H, Curry C, Talbot M, Berger R, Fetics B, Nevo E. Improved left ventricular mechanics from acute VDD pacing in patients with dilated cardiomyopathy and ventricular conduction delay. Circulation 1999;99:1567±1573. 25. Auricchio A, Stellbrink C, Block M, Sack S, Vogt J, Bakker P, Klein H, Kramer A, Ding J, Salo R, Tockman B, Pochet T, Spinelli J. Effect of pacing chamber and atrioventricular delay on acute systolic function of paced patients with congestive heart failure. Circulation 1999;99:2993±3001. 26. Blanc J-J, Etienne Y, Gilard M, Mansourati J, Munier S, Boschat J, Benditt D, Lurie KG. Evaluation of different ventricular pacing sites in patients with severe heart failure. Results of an acute hemodynamic study. Circulation 1997;96:3273±3277. 27. Cazeau S, Ritter P, Lazarus A, Gras D, Backdach H, Mundler O, Mugica J. Multisite pacing for end-stage heart failure: Early experience. PACE 1996;19 (Pt.II):1748±1757. 28. Gras D, Mabo P, Tang T, Oude Luttikuis, Chatoor R, Pedersen A-K, Tscheliessnigg H-H, Deharo J-C, Puglisi A, Silvestre J, Kimber S, Ross H, Ravazzi A, Paul V, Skehan D. Multisite pacing as a supplemental treatment of congestive heart failure: Preliminary results of the Medtronic Inc. InSync Study. PACE 1998;21 (Pt.II):2249±2255. 29. Lister JW, Klotz KH, Jomain SL, Stuckey JH, Hoffman BF. Effect of pacemaker site on cardiac output and ventricular activation in dogs with complete heart block. Am J Cardiol 1964;14:494±503.