Rationale and design of a randomized trial to assess the safety and efficacy of MultiPoint Pacing (MPP) in cardiac resynchronization therapy: The MPP Trial

Abstract

Background

Although the majority of Class III congestive heart failure (HF) patients treated with cardiac resynchronization therapy (CRT) show a clinical benefit, up to 40% of patients do not respond to CRT. This paper reports the design of the MultiPoint Pacing (MPP) trial, a prospective, randomized, double‐blind, controlled study to evaluate the safety and efficacy of CRT using MPP compared to standard biventricular (Bi‐V) pacing.

Methods

A maximum of 506 patients with a standard CRT‐D indication will be enrolled at up to 50 US centers. All patients will be implanted with a CRT‐D system (Quartet LV lead Model 1458Q with a Quadra CRT‐D, Abbott) that can deliver both MPP and Bi‐V pacing. Standard Bi‐V pacing will be activated at implant. At 3 months postimplant, patients in whom the echocardiographic parameters during MPP are equal or better than during Bi‐V pacing are randomized (1:1) to either an MPP or Bi‐V arm.

Results

The primary safety endpoint is freedom from system‐related complications at 9 months. Each patient's response to CRT will be evaluated using a heart‐failure clinical composite score, consisting of a change in NYHA functional class, patient global assessment score, HF events, and cardiovascular death. The primary efficacy endpoint is the proportion of responders in the MPP arm compared with the Bi‐V arm between 3 and 9 months.

Conclusion

This trial seeks to evaluate whether MPP via a single quadripolar LV lead improves hemodynamic and clinical responses to CRT, both in clinical responders and nonresponders.

1. INTRODUCTION

Cardiac resynchronization therapy (CRT) is a well‐established therapy for heart failure (HF) and has been shown to produce significant clinical benefits, including reduced mortality and HF hospitalizations, and improved symptoms and quality of life (Abraham et al., 2002; Anand et al., 2009; Bristow et al., 2004; Cazeau et al., 2001; Cleland et al., 2005; Moss et al., 2009). However, the proportion of patients who fail to respond to CRT remains significant (Chung et al., 2008). The mean responder rate from the 15 largest contemporary CRT studies has been approximately 67% and even lower, 57% based upon echocardiographic parameters (Bax & Gorcsan, 2009). In the MIRACLE study, 34% of patients did not demonstrate an improvement in a HF clinical composite score (CCS), which was a combination outcome measurement of all‐cause mortality, HF‐related hospitalization, NYHA functional class and patient global assessment (PGA) score into an outcome measure (Young et al., 2003).

The cause of lack of a response to CRT is probably multifactorial, complex, and not completely understood; however, there is a general consensus that suboptimal LV lead placement is one of the more common reasons (Becker et al., 2007). Although the use of two LV unipolar or bipolar leads has been proposed to improve LV synchrony and consequently the response to CRT, an alternative approach is the use of a single quadripolar lead (Figure 1) with four electrodes that can stimulate the LV from two of 10 possible vectors (Table 1). By using the quadripolar CRT‐D system (Abbott), MultiPoint™ Pacing (MPP)—LV pacing using two different vectors—can then be selected (Table 1). Additionally, a programmable delay (5–80 ms) can be introduced between the two LV pacing pulses, and the two LV pulses can be delivered either before or immediately after the right ventricular pacing pulse.

Figure 1.

Diagram of the Quartet 1458Q quadripolar lead showing electrode location and nomenclature

Table 1.

The 10 possible LV pacing vectors with the Quartet 1458Q lead

Vector number	Vector
1	D1—M2
2	D1—P4
3	D1—RVCoil
4	M2—P4
5	M2—RVCoil
6	M3—M2
7	M3—P4
8	M3—RVCoil
9	P4—M2
10	P4—RVCoil

D1, M2, M3, and P4 are the quadripolar electrode terminals shown in Figure 1.

Small prospective studies have shown CRT with MPP can result in acute improvements in contractility, hemodynamics, and dyssynchrony compared with standard biventricular (Bi‐V) pacing (Gutleben et al., 2013; Pappone et al., 2014; Rinaldi et al., 2013, 2014; Thibault et al., 2013). Additionally, in a recent study by Pappone et al. (2015a), MPP was shown to result in both mid‐term (3 months) and long‐term (12 months) improvements in LV reverse remodeling and clinical response compared to standard Bi‐V pacing (Pappone et al., 2015).

However, there is a need for larger trials to confirm both the safety and longer term efficacy of CRT with MPP. Furthermore, no CRT trial has been designed to specifically evaluate both MPP and Bi‐V pacing in a group of clinical nonresponders. Therefore, the objective of the present report is to describe the study rationale and design of the MPP trial. This trial is a prospective, randomized, double‐blind, controlled study that will compare chronic safety and efficacy of MPP to standard Bi‐V pacing over 9 months of follow‐up in patients implanted with a quadripolar CRT‐D system.

2. METHODS

2.1. Study design and oversight

The MPP trial is a prospective, randomized, double‐blind, multicenter clinical study sponsored by the manufacturer of the quadripolar CRT‐D system (Abbott) and approved by the Food and Drug Administration and Institutional Review Board at each of the participating centers. A steering committee with the participation of the sponsor will be responsible for the design and conduct of the study and reporting of the findings. Clinical events will be adjudicated by separate blinded events committee. Monitoring and collection of the data and data analyses will be performed by the sponsor in partnership with the steering committee. The primary data analysis will be done by a statistician at Abbott.

2.2. Patients

A maximum of 506 patients with a standard CRT‐D indication will be enrolled at up to 50 US centers. Enrollment in the MPP trial is expected to require approximately 24 months, and the study duration is anticipated to be approximately 36 months.

The study will enroll eligible patients who have a standard clinical indication for implantation of a CRT‐D system for the treatment of HF and life‐threatening ventricular tachyarrhythmias, not limited to patients with symptomatic Class III HF (Epstein et al., 2008; Russo et al., 2013) and will include patients with both new CRT‐D implantation or who have an upgrade from an existing ICD or pacemaker with no prior LV lead placement. Patients will be excluded if they meet any of the following criteria: have an existing Class I recalled lead, are anticipated to need heart transplantation within the next 9 months, have undergone cardiac transplantation within 40 days of enrollment, have had a cerebrovascular accident or transient ischemic attack within 3 months of enrollment, have had a recent myocardial infarction or unstable angina within 40 days or cardiac revascularization within 3 months of implant, have permanent atrial fibrillation, are hypersensitive to a single 1.0 mg dose of dexamethasone sodium phosphate, are less than 18 years of age, are currently participating in a clinical investigation that includes an active treatment arm, are pregnant or planning to become pregnant during the duration of the study, or have a life expectancy of <9 months. All patients will give informed consent before enrollment and the study will be approved by the Investigational Review Board of the participating centers.

All patients enrolled will have cardiac performance (two‐dimensional echocardiography) and other clinical and demographic variables assessed within 30 days prior to implant (Table 2). A preimplant occlusive coronary venous angiogram in at least two views (LAO 20° to 40° and RAO 20° to 40°) followed by postimplant fluoroscopic images in the same views will be collected at implant (Table 2). All patients will have a quadripolar CRT‐D system implanted (Quartet LV lead Model 1458Q with a Quadra CRT‐D, Abbott), and Bi‐V pacing will be activated at implant. The LV pacing vector, atrioventricular delay and interventricular (VV) delay programmed at implant will be left to the discretion of the physician.

Table 2.

Schedule of follow‐up procedures summary

Evaluation	Enrollment	Implant	Predischarge visit	3‐Month visit	6‐Month visit	9‐Month visit
Inclusion/exclusion evaluation and informed consent	X
NYHA classification	X			X	X	X
Patient Global Assessment score				X	X	X
MLWHF questionnaire	X			X	X	X
Intrinsic QRS width and morphology	X
2D echocardiography	X			X	X	X
Cardiac medications list	X			X	X	X
Venograms and fluoroscopic images in at least two matching views (LAO 20° to 40° and RAO 20° to 40°)		X
Chest x‐ray (PA and lateral)			X
Randomization				X
Prior heart‐failure events	X
All‐cause hospitalizations				X	X	X
LV lead capture threshold testing and pacing lead impedance		X	X	X	X	X
Bipolar capture threshold testing, signal amplitude and pacing lead impedance for RA and RV leads		X	X	X	X	X
Device session records		X	X	X	X	X

NYHA, New York Heart Association MLWHF, Minnesota Living with Heart Failure; PA, posteroanterior; LAO, left anterior oblique; RAO, right anterior oblique; LV, left ventricular; RA, right atrial; RV, right ventricular.

This study is double‐blinded to reduce the effect of bias. The study center's staff (e.g., electrophysiologist and EP nurses) performed all tests requiring viewing of the programmed pacing mode or the patient's ECG. Separate authorized site personnel (e.g., blinded assessors) assessing NYHA functional class and PGA as part of an HF CCS (Packer, 2001) and the patients both remain blinded. Neither the patient nor the designated blinded assessors have knowledge of the pacing mode assigned. Patients carry a card identifying them as MPP study patients to minimize the risk of becoming unblinded by other health care providers. It is the intent of the study design and center management to minimize potential study bias through these efforts. All efforts are made to maintain the blind until after the 9‐month follow‐up visit has been completed.

2.3. Postimplant screening and randomization

At 3 months postimplant, patients will be screened to determine whether they have responded to Bi‐V CRT. Each patient's response to CRT will be evaluated using an HF CCS, (Packer, 2001) which includes both objective and subjective measures of clinical status. The CCS includes NYHA functional class, PGA score, HF events, and cardiovascular death. The PGA score is assessed as a change in each patient's self‐assessed condition compared with preimplant and ranges from 1 (markedly better) to 5 (markedly worse). A HF event is defined as hospitalization for HF of ≥24 hr, or any inpatient or outpatient treatment requiring the intravenous administration of diuretics, inotropes, and/or vasodilators. Patients who have at least an improvement in NYHA functional class of ≥1 or a PGA score of 2 (better) or 1 (markedly better) in combination with no HF events and no cardiovascular death will be categorized as “Improved.” Patients who have worsened NYHA functional class or a PGA score of 4 (worse) or 5 (markedly worse) or have a HF event or cardiovascular death will be categorized as “Worsened.” Patients who are not “Improved” or “Worsened” will be categorized as “Unchanged.” Patients who are “Improved” will be classified as CRT responders, and those who are “Worsened” or “Unchanged” will be grouped together as nonresponders.

In addition to screening the patients for their response to CRT at 3 months postimplant, patients will also have acute echocardiographic measurements performed (velocity‐time integral of the transmitral flow (EA VTI); Jansen et al., 2006) at various MPP combinations (Table 3), and these will be compared with Bi‐V pacing. During the acute hemodynamic testing, any vector combinations and timing delays for both MPP and Bi‐V can be programmed per physician's discretion. Only patients in whom the echocardiographic measurements during MPP is equal or better than Bi‐V pacing will be randomized (1:1) to either a Bi‐V arm or an MPP arm (Figure 2). The randomized patients will include both CRT responders and nonresponders based upon the CCS. Patients in whom none of the MPP combinations tested yields an equal or better acute EA VTI hemodynamic response compared with the Bi‐V configuration tested will be assigned to an Observational arm and followed with Bi‐V pacing (Figure 2). For patients randomized to the MPP arm, MPP will be programmed by the physician to any pair of vectors, each of which yields equal or better acute cardiac performance compared with Bi‐V pacing. For patients randomized to the Bi‐V arm, the final Bi‐V programming will be left to the physician's discretion.

Table 3.

MultiPoint Pacing vector combinations

Group	Vector combination	LV1		LV2
		Cathode	Anode	Cathode	Anode
A	1	D1	M2	P4	M2
A	2	D1	M2	P4	RVc
A	3	D1	P4	P4	M2
A	4	D1	P4	P4	RVc
A	5	D1	RVc	P4	M2
A	6	D1	RVc	P4	RVc
B	7	D1	M2	M3	M2
B	8	D1	M2	M3	P4
B	9	D1	M2	M3	RVc
B	10	D1	P4	M3	M2
B	11	D1	P4	M3	P4
B	12	D1	P4	M3	RVc
B	13	D1	RVc	M3	M2
B	14	D1	RVc	M3	P4
B	15	D1	RVc	M3	RVc
C	16	M2	P4	P4	M2
C	17	M2	P4	P4	RVc
C	18	M2	RVc	P4	M2
C	19	M2	RVc	P4	RVc
D	20	D1	M2	M2	P4
D	21	D1	M2	M2	RVc
D	22	D1	P4	M2	P4
D	23	D1	P4	M2	RVc
D	24	D1	RVc	M2	P4
D	25	D1	RVc	M2	RVc
E	26	M3	M2	P4	M2
E	27	M3	M2	P4	RVc
E	28	M3	P4	P4	M2
E	29	M3	P4	P4	RVc
E	30	M3	RVc	P4	M2
E	31	M3	RVc	P4	RVc
F	32	M2	P4	M3	M2
F	33	M2	P4	M3	P4
F	34	M2	P4	M3	RVc
F	35	M2	RVc	M3	M2
F	36	M2	RVc	M3	P4
F	37	M2	RVc	M3	RVc

Figure 2.

Flowchart of trial, from patient enrollment to end of follow‐up

2.4. Follow‐up and Endpoints

All patients will have follow‐up visits at 6 and 9 months postimplant (Figure 2). All data collected at each time point in the study are summarized in Table 2. Authorized site personnel (e.g., Blinded Assessors) who assess the NYHA functional class and the PGA score as part of the CCS will be blinded to randomization assignment. At the 6‐month visit in the MPP arm, physicians will have the option to repeat additional acute hemodynamic testing similar to that conducted at the 3‐month visit in order to tailor MPP programming.

At the 9‐month follow‐up visit, CRT response status will be evaluated again using the CCS. The NYHA class and PGA score will be compared with the 3‐month visit for the evaluation of the primary efficacy endpoint. For patients who were responders at the 3‐month visit, those who are “Improved” and “Unchanged” between 3 and 9 months will be classified as responders, whereas those who are “Worsened” will be grouped together as nonresponders. For patients who were nonresponders at the 3‐month visit, those who are “Improved” will be classified as responders, whereas those who are “Worsened” and “Unchanged” will be grouped together as nonresponders.

For the primary efficacy endpoint, a noninferiority test will be used to compare the percentage of nonresponders between the two arms of the study between 3 and 9 months of follow‐up. A noninferiority margin of 15% will be used. The hypothesis is that between 3 and 9 months of follow‐up, the nonresponse rate to CRT in the MPP arm will not be inferior to that in the Bi‐V arm. In the multicenter FREEDOM study, the proportion of nonresponders (as defined by the CCS in this study) between 3 and 9 months was 44%. Therefore, assuming a binomial distribution with a 44% probability for nonresponse between 3 and 9 months in both study arms, the sample size required for 85% power to reject the null hypothesis at the 5% significance level is 394. To adjust for a potential net crossover of 15% (assuming a 15% crossover rate, which is likely overestimated by design), the effective sample size will be increased from 394 patients to 404 patients. An overall attrition rate of 20% is assumed, and since the primary efficacy endpoint requires 404 patients, the total number of patients required to be enrolled in this study is 506.

The primary safety endpoint, freedom from system‐related events through 9 months of follow‐up, will be evaluated by an objective performance criterion (OPC). All patients who have an attempted implant or a successful quadripolar CRT‐D system implant will be included in the analysis of this endpoint. The survival time for patients who experience a system‐related complication will be calculated as the number of days from the date of implant to the date the complication is first identified. For patients who do not experience a system‐related complication and withdraw from the study before the time of analysis, the survival time will be censored on the date of their withdrawal. For patients who do not have a system‐related complication by the time of analysis, the survival time will be censored on the date of data cut‐off. This endpoint will be evaluated using the Kaplan–Meier method to estimate the freedom from system‐related complications from implant to 9 months, and the 97.5% lower confidence bound (LCB) will be calculated for freedom from system‐related complications on the log‐survival scale. The null hypothesis will be rejected if the 97.5% LCB is greater than the OPC of 75%.

3. DISCUSSION

Conventional Bi‐V pacing has failed to improve CRT outcomes in a substantial percentage of patients. As a result, newer pacing therapies and strategies have been developed. One such therapy is the use of multiple LV pacing leads in an attempt to activate larger areas of the myocardium (Ginks et al., 2012; Leclercq et al., 2008; Rinaldi et al., 2015; Shetty et al., 2014; Yoshida et al., 2007). Previous animal and clinical work has shown that simultaneously exciting a larger mass or volume of cardiac tissue results in faster depolarization velocity and shorter left ventricular transventricular conduction times (Lloyd, Heeke, Lerakis, & Langberg, 2007; Theis, Bavikati, Langberg, & Lloyd, 2009). In addition, by capturing a larger volume of cardiac muscle, the site of latest intrinsic activation within the left ventricle may be more likely to be depolarized early, resulting in better synchronization and maximizing cardiac output. Multisite LV pacing using multiple leads has been evaluated in numerous small trials and the results have been both promising and inconsistent (Ginks et al., 2012; Leclercq et al., 2008; Rinaldi et al., 2015; Shetty et al., 2014; Yoshida et al., 2007). In addition, the need to use multiple LV leads significantly increases the complexity, duration, and risk of the procedure. For this reason, MPP using a single quadripolar lead is a particularly attractive alternative.

Compared to traditional Bi‐V pacing, several small studies have shown that MPP delivered through a quadripolar LV lead provides acute benefit as measured by LV dP/dtMax (maximum rate of rise of LV pressure; Thibault et al., 2013) LV dyssynchrony, (Rinaldi et al., 2013) LV peak radial strain, (Rinaldi et al., 2014) LV pressure‐volume loop parameters, (Pappone et al., 2014), and LV electrical activation (Shetty et al., 2014). A recent study by Pappone et al. (2015b) demonstrated that MPP offers effective stimulation of the LV and achieves mid‐term and long‐term improvements in LV reverse remodeling and LV function compared to standard Bi‐V pacing. Additionally, a recent study utilizing pressure wire measurements by Zanon et al. (2015) in 29 patients showed that MPP yielded a consistent increase in hemodynamic response compared with Bi‐V pacing. In addition, the MPP‐induced improvement in contractility was associated with significantly greater narrowing of the QRS complex compared with standard Bi‐V pacing (Zanon et al., 2015). However, these previous evaluations of MPP have been limited to small feasibility studies and have not provided long‐term clinical benefits of MPP in a large multicenter trial setting. Therefore, there is a definite need for additional safe and effective MPP LV pacing strategies in the treatment of congestive HF patients.

The MPP trial was designed to provide evidence of both safety and efficacy in the treatment of both CRT responders and nonresponders in a double‐blinded fashion, and the trial design incorporates a number of unique aspects: (i) the ability to LV pace at two different sites using a single lead; (ii) only patients in whom MPP provides equal or better acute cardiac function compared with Bi‐V at 3 months will be randomized to an MPP arm or Bi‐V arm. (iii) the option for nonresponders to potentially benefit by receiving a second form of LV pacing (MPP) or by continuation of Bi‐V pacing. Since acute echocardiographic measurements have been shown to predict the chronic response to CRT (Duckett et al., 2011; Oguz et al., 2002; Tournoux et al., 2007) the acute improvements in cardiac performance with MPP at 3 months may well translate into a sustained benefit that will be reflected in a difference between the response rates to CRT between the MPP and Bi‐V arms at 9 months. The information gained from this trial design will include the conversion rate from nonresponder to responder, conversion rate from responder to super‐responder, long‐term reverse remodeling data, and correlation between electrode location and long‐term clinical outcome. MPP may become an important new method to optimize CRT especially in patients not responsive to BiV pacing.

DISCLOSURES

Gery Tomassoni—Advisor, Speaker, Medical Device Board: Abbott, Biosense Webster, Medtronic, Boston Scientific, Biotronics, Siemens, STXS, Topera, Atricure & Pfizer; James Baker II—Medical device advisory board: Abbott; Raffaele Corbisiero—Consulting: Boston Scientific, Abbott; Charles Love—Consulting: Medtronic, Abbott, Spectranetics, Convatec; David Martin—Consulting: Biotronik, Abbott; Robert Sheppard—Nothing to disclose; Seth Worley—Royalties: Pressure Products, Merit Medical, Teaching Honoraria: Medtronic, Abbott; Kwangdeok Lee—Employee at Abbott; Imran Niazi—Consulting: Abbott.

Tomassoni G, Baker J II, Corbisiero R, et al. Rationale and design of a randomized trial to assess the safety and efficacy of MultiPoint Pacing (MPP) in cardiac resynchronization therapy: The MPP Trial. Ann Noninvasive Electrocardiol. 2017;22:e12448 10.1111/anec.12448