Subcutaneous Defibrillator Implantation With or Without Defibrillation Test: The Primary Results of the Randomized PRAETORIAN-DFT Trial

Abstract

BACKGROUND:

To improve survival of patients at risk of sudden cardiac death, subcutaneous implantable cardioverter defibrillators (S-ICDs) require optimal implant positioning for effective shocks. Defibrillation (DF) testing is recommended but carries serious risks. The PRAETORIAN score predicts defibrillation outcomes on the basis of chest X-ray. The PRAETORIAN-DFT (Prospective Randomized Comparative Trial of Subcutaneous Implantable Cardioverter Defibrillator Implantation With and Without Defibrillation Testing) trial evaluated whether omission of DF testing guided by the PRAETORIAN score is noninferior for first-shock efficacy.

METHODS:

In this multinational trial, S-ICD patients from 37 centers were randomized to DF testing or no DF testing. In the No-DF testing group, the PRAETORIAN score was evaluated before discharge. The primary end point was failed first shock for spontaneous ventricular arrhythmias, as a surrogate for defibrillation success, tested for noninferiority with a 3% absolute risk margin. Secondary end points included mortality, potential DF testing–related complications, and S-ICD revisions.

RESULTS:

The included 965 patients (No-DF testing, n=483; DF testing, n=482) were followed for a median of 41 months. Failed first shock for spontaneous ventricular arrhythmia occurred in 1.7% of the No-DF testing group versus 2.3% of the DF testing group (–0.6% [95% CI, –2.6 to 1.4], P<0.001). There were no significant differences in all-cause mortality (hazard ratio [HR], 0.9 [95% CI, 0.6–1.4]) or arrhythmic death (HR, 0.4 [95% CI, 0.04–3.4]). Potential DF testing–related complications occurred in 1.7% in the DF testing group. Postoperative S-ICD revisions due to inadequate positioning were identical between groups (n=2 each).

CONCLUSIONS:

PRAETORIAN score–guided omission of DF testing after S-ICD implantation did not increase the risk of failed first shocks for spontaneous ventricular arrhythmias and reduced procedural risk without increasing S-ICD revisions.

REGISTRATION:

URL: https://www.clinicaltrials.gov; Unique identifier: NCT03495297.

Clinical Perspective.

What Is New?

• This is the first randomized trial to demonstrate that omitting defibrillation testing when subcutaneous implantable cardioverter defibrillator (S-ICD) implantation is guided by the PRAETORIAN score does not compromise long-term S-ICD efficacy and is noninferior to routine defibrillation testing.

What Are the Clinical Implications?

• The PRAETORIAN score can be used to noninvasively evaluate S-ICD implant position and improve implantation technique.

• Routine defibrillation testing during de novo S-ICD implantation may be safely omitted when guided by the PRAETORIAN score.

• Avoiding defibrillation testing may improve procedural safety and simplify S-ICD implantation.

• An individualized approach remains warranted, particularly in patients with PRAETORIAN scores ≥90.

Implantable cardioverter defibrillators (ICDs) are an established therapy for the prevention of sudden cardiac death.^1–3 Defibrillation (DF) testing was routinely performed after transvenous ICD implantation to confirm the device’s ability to terminate ventricular arrhythmia (VA), but earlier trials demonstrated that omitting DF testing after transvenous ICD implantation does not compromise long-term shock efficacy or arrhythmic mortality.^4,5 Given the procedural burden, costs, need for anesthesia, and the risk for serious complications such as stroke, prolonged resuscitation, and death, DF testing is no longer routinely recommended after transvenous ICD implantation.^3–7

The subcutaneous ICD (S-ICD) is an entirely extravascular system designed to avoid transvenous lead–related complications.⁸ Since the S-ICD is entirely subcutaneous, and effective defibrillation by the S-ICD is dependent on specific criteria for implant positioning, findings on DF testing for transvenous ICDs cannot be extrapolated to DF testing for S-ICDs. Therefore, current guidelines recommend performing DF testing after S-ICD implantation.⁷ The PRAETORIAN score was designed to assess correct S-ICD positioning on chest X-ray and predict DF testing success.⁹ This score has been validated both prospectively and retrospectively, demonstrating that a score <90 reliably predicts DF testing success.^10,11 The PRAETORIAN-DFT (Prospective Randomized Comparative Trial of Subcutaneous Implantable Cardioverter Defibrillator Implantation With and Without Defibrillation Testing) trial prospectively evaluated whether omission of DF testing, guided by the PRAETORIAN score, is noninferior to routine DF testing after S-ICD implantation with respect to shock efficacy for terminating spontaneous VAs.

Methods

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Trial Design and Oversight

The PRAETORIAN-DFT trial is an investigator-initiated, randomized, international, multicenter trial conducted in Europe, Israel, and the United States. The study protocol was approved by the local ethics committees at all participating centers and conducted in accordance with the Declaration of Helsinki. The full study design has been published previously and registered on ClinicalTrials.gov (NCT03495297).¹² This investigator-initiated study was funded by Boston Scientific, who had no role in study design, data collection, analysis, interpretation, or writing of the manuscript. Data were reviewed by an independent data and safety monitoring board at regular intervals. Principal investigators analyzed the data in accordance with the statistical analysis plan and prepared the manuscript. The authors vouch for the accuracy and completeness of the data and for the fidelity of the trial to the protocol.

Patients

Adult patients eligible for ICD therapy according to current guidelines and planning to undergo de novo S-ICD implantation were included. Implantations were performed by physicians who had performed ≥13 S-ICD implantations.¹³ Key exclusion criteria were the presence of an intracardiac thrombus, atrial fibrillation without appropriate anticoagulation therapy, or other contraindications to DF testing per physician discretion. Full eligibility criteria are provided in the Supplemental Material. All patients provided written informed consent.

Randomization, Procedures, and Follow-Up

Eligible patients were randomly assigned at a 1:1 ratio to DF or no DF testing (No-DF testing) using varying block sizes, (2, 4, and 6) stratified by center (S). In the DF testing group, management of failed testing was at physician discretion but required at least device or lead repositioning or additional DF testing, with other measures allowed as deemed appropriate. In the No-DF testing group, the PRAETORIAN score was calculated before discharge by the physician who performed the implantation. The PRAETORIAN score is a discrete ordinal score with values between 30 and 900 points (Figure S1), which stratifies the risk of conversion failure into 3 categories: low risk (<90 points), intermediate risk (90 to <150 points), and high risk (≥150 points). After confirmation of chest X-ray quality and eligibility, the score is derived through a 4-step process on the basis of subcoil fat, subgenerator fat, the position of the generator in relation to the midaxillary line, and body mass index.⁹ Further details are provided in the Supplemental Material. If the score was ≥90, physicians were instructed to perform, at minimum, DF testing.

The DF testing was considered successful when an induced VA (induced by a 50-Hz, 200-mAmp pacing burst) was terminated within 5 s after a first 65-J shock. The first test used shock in standard polarity (coil to can); if unsuccessful, a second test was performed with reversed polarity (can to coil). If a VA could not be induced after 3 attempts, a 10-J shock was delivered to evaluate high-voltage impedance. Postimplantation chest X-ray were obtained in both treatment groups.

Data were collected at follow-up visits as part of standard care and through remote monitoring. Programming was left to site discretion, but dual-zone settings (conditional, 180–200 beats per minute (bpm); unconditional, ≥250 beats per minute) and SMART-Pass activation were strongly recommended. Programming and all changes in programming were recorded and actively monitored to guarantee equality in both treatment groups.

End Points

The primary end point was failed first shock for a spontaneous VA, defined as lack of termination of ventricular tachycardia or ventricular fibrillation (VF) within 5 s of shock delivery. This end point has been used previously as a surrogate for ineffective ICD therapy and is further discussed in the Supplemental Material.^4,5 Secondary outcomes included lead or device repositioning and potential DF testing–related complications, defined as complications occurring within 24 hours after DF testing and those occurring within 30 days after DF testing that were considered potential DF testing related. In addition, a composite of all complications occurring within 30 days after S-ICD implantation was recorded in both groups. Arrhythmia and mortality end points were adjudicated by an independent clinical event committee of electrophysiologists blinded to treatment allocation. The end points are described in detail in the Supplemental Material (Table S2).

Statistical Analysis

The primary analysis assessed whether omission of DF testing, guided by the PRAETORIAN score, was noninferior to DF testing, using cumulative event rates expressed as absolute risk. Estimated S-ICD first-shock efficacy was assumed at 90%,^14,15 with a prespecified lower boundary of 75%, both corresponding to >99.9% arrhythmia termination within 5 shocks. Assuming a 20% incidence of spontaneous VA during follow-up,^14,15 the expected failed first shock rate was estimated at 2% with an acceptable rate of 5% using the lower boundary, yielding a 3% noninferiority margin. Given a 2% true event rate, a noninferiority margin of 3%, 1-sided α=0.025, and 90% power, 458 patients per treatment group were required; with 5% attrition, target enrollment was 965.¹²

Analyses for all end points followed the modified intention-to-treat principle. Patients who did not receive an S-ICD after randomization or who underwent randomization in error were excluded. The primary analysis was planned after either an overall cohort event rate of 2% or a median follow-up of 40 months, whichever came first.

The Kaplan-Meier method was used for visualization of time-to-event data, while hazard ratio (HRs), cumulative event rate, and 95% CI were estimated using Cox proportional hazards regression. Odds ratio (OR) with 95% CI was estimated using logistic regression. Data were assumed to be missing at random, and data were censored at the last known event-free date for patients with incomplete follow-up. Confidence intervals were not adjusted for multiplicity. Prespecified subgroup analyses were performed by sex, body mass index, primary versus secondary prevention, and preserved versus reduced left ventricular ejection fraction.

Sensitivity analyses for the primary end point included per-protocol analyses, adjustment for reported shocks that could not be adjudicated due to missing ECGs and differences in shocks for VA between groups, analysis based on the difference in proportions, Fine-Gray modeling to account for the competing risk of death, and censoring after Empower leadless pacemaker implantation (providing antitachycardia pacing capability). Statistical analyses were performed using R software (version 4.3.2).

Results

Patients and Treatments

Between May 2018 and January 2023, 973 patients were enrolled at 37 centers, of whom 965 were included in the modified intention-to-treat group. These patients were randomly assigned to No-DF testing (n=483) or DF testing (n=482) (Figure 1; Figure S2). Baseline characteristics were comparable between treatment groups (Table 1). Mean age was 54±14 years, 24% were women, and mean left ventricular ejection fraction was 40%. The S-ICD was implanted for secondary prevention in 34%. There was no difference in the programmed therapy zones between the 2 treatment groups at implantation or at last follow-up, with a median conditional zone of 200 beats per minute and a median unconditional zone of 250 beats per minute (Figure S3).

Figure 1.

Flowchart of patient inclusion. DF indicates defibrillation; ICD, implantable cardioverter defibrillator; IQR, interquartile range; LVAD, left ventricular assist device; and S-ICD, subcutaneous implantable cardioverter defibrillator. *See Figure S2 for additional information. †In 2 patients with a PRAETORIAN score <90, DF testing was performed per patient or physician discretion. ‡In 2 patients, DF testing was not attempted: 1 due to physician discretion and 1 due to a device-related infection before DF testing could be scheduled. §In 3 of 12 patients with 3 failed inductions, no 10-J shock was delivered. ¶Median time to loss to follow-up was 21.1 months (IQR, 3.4–37.5) for the No-DF testing group and 18.0 months (IQR, 3.5–37.5) for the DF testing group.

Table 1.

Baseline Characteristics

In the DF testing group, DF testing was attempted in 480 of 482 patients (99.6%). In 2 patients, DF testing was not attempted due to physician discretion (n=1) or device extraction for a device-related infection before DF testing was scheduled (n=1). In 14 patients, induction failure (n=12) or spontaneous conversion (n=2) occurred. In 9 patients, the first 65-J shock during DF testing was unsuccessful, and a subsequent successful 80-J shock was accepted by physician discretion, with no repeated DF testing attempts or device repositioning, which deviated from the study protocol. Among patients who underwent protocol-compliant DF testing, successful defibrillation was achieved in 445 of 457 (97.4%) patients. There was no significant difference in median time to shock between patients with successful and unsuccessful first shocks during DF testing (15.0 s versus 15.0 s; P=0.47; Table S3). Of 12 patients with failed DF testing, 5 patients underwent repositioning of the lead or generator in the same procedure, and 1 patient underwent lead repositioning the following day. Successful DF testing was achieved in all patients who underwent device repositioning, resulting in a total of 450 of 457 (98.5%) patients with successful DF testing.

In the No-DF testing group, the PRAETORIAN score could not be calculated in 3 patients due to missing chest X-ray. In 6 patients, the S-ICD was positioned outside of the indicated anatomical areas; subsequent actions are shown in Table S4. A total of 459 of 474 (96.8%) patients had a low PRAETORIAN score, whereas 14 had an intermediate score and 1 a high score. In 14 of these 15 patients, protocol-mandated DF testing was attempted; in 1 patient, DF testing was not performed per physician discretion. Details on actions in patients with an intermediate or high PRAETORIAN score are shown in Table S4. In 2 patients, VF could not be induced. All induced VAs were successfully converted, and no device revisions were performed

Intraoperative repositioning occurred numerically more frequently in the No-DF testing group than in the DF testing group (n=23, 4.7% versus n=15, 3.1%; OR, 1.55 [95% CI, 0.80–3.08]). Across both groups, repositioning occurred primarily after intraoperative imaging, whereafter the lead was repositioned in 27 patients, the generator in 9 patients, and the lead and generator in 2 patients (Table 2). Lead repositioning was performed in 4 patients because of malposition on the initial postimplantation chest X-ray: 2 in the No-DF testing group and 2 in the DF testing group despite successful DF testing (Table S4). Indications for repositioning included lead dislocation, lead placement in the mediastinum, or other inadequate lead positions.

Table 2.

Procedural Information

Follow-Up and Primary End Point

After a median follow-up of 41 months (41 [interquartile range, 31–55] No-DF testing versus 42 [interquartile range, 31–55] DF testing, P=0.59), the primary end point of failed first shock on spontaneous VA occurred in 7 patients (1.7%) in the No-DF testing group and 11 patients (2.3%) in the DF testing group (absolute difference, −0.6% [95% CI, –2.6 to 1.4]; P for noninferiority<0.001; Figure 2). Also, there was no difference between groups in the secondary end points of all-cause mortality (HR, 0.9 [95% CI, 0.6–1.4]) and arrhythmic death (n=1 No-DF testing versus n=3 DF testing; HR, 0.4 [95% CI, 0.04–3.4]; Table 3). These results were consistent across subgroups, with no significant interaction (Figure S4). Sensitivity analyses for the primary end point showed consistent results (Tables S5 and S6).

Figure 2.

Cumulative incidence of failed first shock by treatment group. DF indicates defibrillation.

Table 3.

Secondary End Points

During follow-up, 61 patients (13.8%) in the No-DF testing group and 76 (15.0%) patients in the DF testing group received shocks for spontaneous VA (155 versus 284 shocks, respectively). Among patients who received shocks for spontaneous VAs, 11.5% in the No-DF testing group and 14.5% in the DF testing group experienced a failed first shock (P=0.80). There was no difference between the treatment groups in the number of patients experiencing shocks for VA (HR, 0.79 [95% CI, 0.57–1.11]) or in first- and last-shock efficacy (Table S7). Across PRAETORIAN score categories, 57 patients in the low-risk category experienced shocks for VA, 3 patients in the intermediate/high risk category, and 1 patient for whom the PRAETORIAN score could not be calculated (Figure 3). No patient died as a direct result of S-ICD shock failure.

Figure 3.

Distribution of patients who experienced shocks for ventricular arrythmia (VA) by PRAETORIAN score category* (No-DF testing group). DF indicates defibrillation. *The PRAETORIAN score was calculated before discharge by the physician who performed the implantation.

Complications within 24 hours after DF testing occurred in 8 (1.7%) patients in the DF testing group and none in the No-DF testing group (Table 3; Table S8). These included 6 patients with refractory VF, defined as requiring more than 1 external rescue shock, of whom 1 patient needed cardiopulmonary resuscitation. In addition, 1 patient with respiratory arrest required mask ventilation, and in 1 patient, VF was converted to atrial fibrillation, resulting in an inappropriate shock. All events were ultimately terminated either by repeated external rescue or S-ICD shocks, and no deaths were attributed to defibrillation testing. The 30-day complication rate was significantly lower in the No-DF testing group compared with the DF testing group (n=8, 1.7 % versus n=23, 4.8%; HR, 0.34 [95% CI, 0.15–0.77]; Figure 4; Table 3), driven by refractory VF, unplanned intensive care unit stay, and cardiac decompensation. The overall incidence of patients with S-ICD–related complications requiring invasive intervention during follow-up was similar between groups (n=28, 5.2% No-DF testing group versus n=29, 5.9% DF testing group; HR, 0.97 [95% CI, 0.58–1.63]; Table 3; Table S9).

Figure 4.

Potential defibrillation (DF) testing–related complications after DF testing and 30-day complication composite. *All potential DF testing–related complications were also included in the 30-day complication composite.

Discussion

This is the first randomized comparative trial to demonstrate that S-ICD implantation without defibrillation testing, guided by the PRAETORIAN score, does not compromise long-term S-ICD efficacy and is noninferior to implantation with defibrillation testing. Despite concerns that DF testing omission could increase the risk of unrecognized ineffective ICD therapy, both all-cause and arrhythmic mortality appeared to be similar between groups.

The primary findings are robust, as the upper bound of the CI remained well within the prespecified noninferiority margin. These findings remained consistent in the per-protocol analysis and after accounting for differences in the number of shocks for VA between the treatment groups, underscoring the potential use of the PRAETORIAN score as an alternative for DF testing. These findings support omitting routine defibrillation testing after de novo S-ICD implantation when guided by the PRAETORIAN score.

In our trial, defibrillation testing raised important safety concerns, with 6 patients experiencing refractory VF. The DF testing group also had a significantly higher 30-day complication rate, driven by refractory VF, unplanned intensive care unit stay, and cardiac decompensation. Except for refractory VF, these events were not predefined as potentially DF testing related in the protocol but may plausibly be attributed to DF testing. Previous studies in both transvenous ICDs and S-ICDs also found DF testing to lead to (nearly) fatal events, related to either VF induction or general anaesthesia.^6,16–18 Furthermore, DF testing requires general or deep anesthesia. Omitting DF testing offers logistical advantages and greater flexibility in anesthesia management and allows alternatives such as nerve blocks or monitored anesthesia care.^19,20 Consistent with these considerations, observational data already show declining DF testing use in contemporary S-ICD implantation.^16,21,22 This study demonstrates that DF testing can be safely omitted without compromising shock efficacy for terminating spontaneous VA and for survival.

Importantly, although the PRAETORIAN score can only be calculated after the procedure, this did not result in a higher rate of repositioning in the No-DF testing group. It seems that operators more often repositioned the device or lead under fluoroscopy and performed impedance checks during implantation, likely reflecting heightened attention to implant position. The high proportion of low PRAETORIAN scores compared with other studies reflects implantation practice in experienced centers and underscores the role of the score as a guide to optimal implantation technique.^10,23–25

Prior to the current analysis, the PRAETORIAN score was validated in the DF testing group within this trial, showing a 99% chance of a successful DF testing with a low PRAETORIAN score.¹¹ Similarly, other studies have validated that a low PRAETORIAN score predicts successful DF testing and conversion of spontaneous VA.^17,26 This trial, analyzing shock efficacy for terminating spontaneous VA, confirms that, in patients with low PRAETORIAN scores, DF testing can be safely omitted. Although the proportion of patients with intermediate and high PRAETORIAN scores was too low to calculate predictive values, defibrillation testing or, preferably, device repositioning should be considered in case of scores ≥90, with consideration of reoperation risks. Although low impedance is sometimes regarded as a surrogate for a low PRAETORIAN score, it may overestimate DF efficacy, as anterior S-ICD positioning can cause energy shunting across the thoracic wall, reducing myocardial mass within the DF pathway.²⁷ For patients undergoing S-ICD generator or system replacement, an individualized assessment is required, taking into account the initial PRAETORIAN score, impedance, changes in body mass index, and sensing characteristics. Patients undergoing generator replacement were not included in the present study; however, previous data on S-ICD generator replacements have demonstrated no increase in failed DF testing among patients with low PRAETORIAN scores.²⁸

This trial is subject to several limitations. First, the proportion of participants with high PRAETORIAN scores was low, as were the numbers of reinterventions and spontaneous VAs that are available for statistical analysis in this group. Second, the protocol did not specify a course of action when chest X-ray did not meet the required criteria necessary to calculate the PRAETORIAN score. Third, although the study population is representative of a general ICD population, some subgroups were relatively small. In addition, the relatively young age and lower comorbidity burden of the study population may limit the generalizability of our findings to older and more comorbid patients. Fourth, some individuals were considered ineligible for participation in this trial due to physician discretion to perform or omit DF testing, and some patients were not treated according to their allocated treatment, introducing the possibility for selection bias. Fifth, not all DF tests were performed according to protocol. Sixth, although mortality did not differ between treatment groups, this trial was not powered to detect differences in mortality. Last, despite extensive efforts, not all electrograms could be retrieved. However, sensitivity analyses remained consistent after accounting for this.

The PRAETORIAN-DFT trial showed that, when guided by the PRAETORIAN score, omission of defibrillation testing after S-ICD implantation was safe and did not increase the risk of failed first-shock therapy for spontaneous VAs. Given that defibrillation testing is associated with potential life-threatening complications, avoiding routine defibrillation testing enhances patient safety and simplifies the procedure without compromising S-ICD effectiveness.

Article Information

Acknowledgments

The authors thank the patients who participated in this trial and the physicians and research coordinators for contributions to the conduct of this trial.

Disclosures

Dr Knops reports consultancy fees from Abbott Laboratories, Medtronic USA Inc, Boston Scientific Corporation, and AtaCor Medical and research grants from Abbott Laboratories and Boston Scientific Corporation. He also has equity in Medtronic, Kestra Medical, Abbott Laboratories, Boston Scientific Corporation, and AtaCor Medical. Dr Marquie reports consultancy fees from Boston Scientific Corporation. Dr Nordbeck reports consultancy fees from Idorsia, Boston Scientific Corporation, Cardiac Dimensions, Sangamo, Sanofi-Genzyme, Amicus Therapeutics Inc, and Takeda Pharmaceuticals. Dr Tilz reports consultancy fees from Boston Scientific Corporation and Abbott, and reports research grants from Abbott. Dr Lambiase reports research grants from Boston Scientific Corporation. Dr Boersma reports consultancy fees from Medtronic and Boston Scientific Corporation. Dr Burke is a consultant for Boston Scientific Corporation, receives honoraria and research grants from Boston Scientific Corporation, and has equity in and serves as Chief Medical Officer for AtaCor Medical. Dr Betts reports consultancy fees from Acutus Medical, Inc and Abbott Laboratories. Dr Probst reports consultancy fees from Biotronik Inc, Medtronic, Inc, and Boston Scientific Corporation. Dr Kuschyk reports consultancy fees from Boston Scientific Corporation. Dr Golovchiner reports consultancy fees from Boston Scientific Corporation. Dr Miller reports consultancy fees from Boston Scientific Corporation. Dr de Jong reports consultancy fees from Medtronic USA, Inc. Dr Lewis reports institutional relationships with Boston Scientific Corporation. Dr Marijon reports research grants from ZOLL Medical Corporation, Boston Scientific Corporation, the Abbott Fund, and Medtronic USA, Inc. Dr Defaye reports research grants from Medtronic, the Abbott Fund, and Boston Scientific Corporation. Dr Martin reports consultancy fees from Boston Scientific Corporation, Medtronic, and Abbott Vascular and research grants from Boston Scientific Corporation and Abbott Vascular. Dr Field reports consultancy fees from Boston Scientific Corporation. Dr Shaik reports consultancy fees from Boston Scientific Corporation. Dr Boveda reports consultancy fees from Medtronic and Boston Scientific Corporation. Dr Kooiman reports consultancy fees from Boston Scientific Corporation. Dr Kaiser reports other financial relationships with Boston Scientific Corporation. Dr Maass reports consultancy fees from the European Union. Dr Wilde reports consultancy fees from Solid Biosciences Inc and Rocket Pharmaceuticals. Dr El-Chami reports consultancy fees from Medtronic USA, Inc and Boston Scientific Corporation.

Supplemental Material

Tables S1–S9

Figures S1–S4

Supplementary Material

cir-154-520-s001.pdf ^{(1MB, pdf)}

Appendix

J. Vogler, M. van Gent, H. Al Hashimi, F.J. Oosterwerff, DJ. Van Doorn, A.E. Borger van der burg, D. Nemirovsky, P.P.H.M. Delnoy.

Funding Statement

The PRAETORIAN-DFT trial was an investigator-initiated trial. Boston Scientific financially supported the trial.