Author + information
- Received December 19, 2016
- Revision received April 17, 2017
- Accepted May 21, 2017
- Published online September 4, 2017.
- Lisa Bergersen, MD, MPHa,∗ (, )
- Lee N. Benson, MDb,
- Matthew J. Gillespie, MDc,
- Sharon L. Cheatham, PhDd,
- Andrew M. Crean, MDe,f,
- Kan N. Hor, MDd,
- Eric M. Horlick, MDe,
- Te-Hsin Lung, PhDg,
- Brian T. McHenry, MSh,
- Mark D. Osten, MDe,
- Andrew J. Powell, MDa and
- John P. Cheatham, MDd
- aDepartment of Cardiology, Boston Children’s Hospital, Boston, Massachusetts
- bDivision of Cardiology, Labatt Family Heart Center, Hospital for Sick Children, Toronto, Ontario, Canada
- cDepartment of Cardiology, Children’s Hospital of Philadelphia, Philadelphia, Pennsylvania
- dDepartment of Cardiology, Nationwide Children’s Hospital, Columbus, Ohio
- ePeter Munk Cardiac Centre, University Health Network, Toronto General Hospital, Toronto, Ontario, Canada
- fJoint Department of Medical Imaging, Toronto General Hospital, Toronto, Ontario, Canada
- gCoronary and Structural Heart Clinical Department, Medtronic, Santa Rosa, California
- hCoronary and Structural Heart Research and Innovation Department, Medtronic, Mounds View, Minnesota
- ↵∗Address for correspondence:
Dr. Lisa Bergersen, Harvard Medical School, Department of Cardiology, Boston Children’s Hospital, Mail Code BCH 3215, 300 Longwood Avenue, Boston, Massachusetts 02115.
Objectives This study sought to obtain in vivo data to confirm assumptions on device loading conditions and assess procedural feasibility, safety, and valve performance.
Background The Harmony transcatheter pulmonary valve (Medtronic, Minneapolis, Minnesota) was designed for patients with severe pulmonary regurgitation who require pulmonary valve replacement.
Methods Three sites participated in this first Food and Drug Administration–approved early feasibility study using an innovative device design to accommodate the complex anatomy of the right ventricular outflow tract. Potentially eligible patients underwent review by a screening committee to determine implant eligibility. Six-month outcomes are reported.
Results Between May 2013 and May 2015, 66 subjects were enrolled, and 21 were approved for implant and underwent catheterization; 20 were implanted. Catheterized patients had a median age of 25 years, were predominantly diagnosed with tetralogy of Fallot (95%), had severe pulmonary regurgitation (95%), and had trivial or mild stenosis. The device was delivered in the desired location in 19 of 20 (95%) patients. Proximal migration occurred in 1 patient during delivery system removal. Two devices were surgically explanted. Premature ventricular contractions related to the procedure were reported in 3 patients; 2 were resolved without treatment. One patient had ventricular arrhythmias that required treatment and later were resolved. At 1 month, echocardiography revealed none or trivial pulmonary regurgitation in all and a mean right ventricular outflow tract gradient of 16 ± 8 mm Hg (range 6 to 31 mm Hg).
Conclusions In this feasibility study of the Harmony transcatheter pulmonary valve device, there was high procedural success and safety, and favorable acute device performance.
Patients with congenital heart defects such as tetralogy of Fallot (TOF) often are left with pulmonary regurgitation (PR) after their initial surgical repair. This may progress over time, requiring additional interventions to establish the pulmonary competence necessary for improved long-term quality of life (1). Transcatheter pulmonary valve replacement (TPVR) is a new, less invasive alternative to surgery for pulmonary valve replacement (2). Recently the U.S. Food and Drug Administration approved 2 transcatheter pulmonary valves (TPVs) for TPVR in dysfunctional right ventricle-to-pulmonary artery (RV-to-PA) conduits: the Melody TPV (Medtronic, Minneapolis, Minnesota) and the Edwards SAPIEN XT transcatheter heart valve (Edwards Lifesciences, Irvine, California). However, these devices have limited application in off-label use within nonconduit, native right ventricular outflow tracts (RVOTs) due to the heterogeneity in anatomic shapes and sizes, leading to a need for TPV options with novel designs (3,4). The Harmony TPV (Medtronic) is a self-expandable device designed to accommodate the larger RVOTs typical in patients with native RVOTs. This device (currently available in 1 size) is being evaluated in the Study of the Native Outflow Tract Transcatheter Pulmonary Valve (NCT01762124), a Food and Drug Administration–approved early feasibility study (5) with limited enrollment. This article reports 6-month outcomes from this feasibility study.
This nonrandomized, prospective, multicenter, early feasibility study was conducted at 3 centers (Boston Children’s Hospital, Boston, Massachusetts; Nationwide Children’s Hospital, Columbus, Ohio; and the Hospital for Sick Children, Toronto, Canada). The primary goal was to evaluate in vivo loading conditions, as demonstrated by radial compression, linear compression, axial compression, bending, and torsion, using post-implant computed tomography (CT) angiography assessment. Additionally, clinical assessments were collected at baseline and at follow-up periods of 1 month, 3 months, and 6 months, and will be collected annually to 5 years.
Patients with native or patch-repaired RVOTs, with PR (classified as severe either by echocardiography or with a PR fraction ≥30% as measured by cardiac magnetic resonance [CMR] imaging), and who were either symptomatic secondary to the PR or who had a RV end-diastolic volume index ≥150 ml/m2 were screened for anatomic suitability. Those who met all inclusion criteria and no exclusion criteria (Online Table 1) were consented and proceeded to the next steps of the screening process (6).
Potentially eligible subjects underwent cardiac CT angiography to evaluate the anatomic fit and to create a stereolithographic model of the RVOT in systole and diastole. A screening committee met bimonthly with the study investigators to discuss patient suitability for implantation. Sixty-six patients provided informed consent (1 patient was enrolled into the study twice, as the patient was exited following initial screening, and subsequently re-enrolled over 1 year later for rescreening), and 21 were approved for implantation of the Harmony TPV device.
The Harmony TPV is a porcine pericardial tissue valve mounted on a self-expanding nitinol frame. The device has an outer diameter of 23.5 mm at the valved section and is approximately 55 mm in length. The TPV is treated with an alpha amino oleic acid antimineralization process to mitigate leaflet calcification, and a 0.2% glutaraldehyde sterilant (Figure 1A). The delivery system is a 25-F coil-loading catheter with an integrated sheath. The loading funnel collapses the valve to facilitate mounting on the delivery system, and the retractable sheath helps to control self-expansion of the TPV during deployment (Figure 1B).
Clinical data collection
Patients will be followed for 5 years after implantation. Data are collected by the primary investigator at each site and include transthoracic echocardiography, fluoroscopy, and CMR imaging assessments. These data are entered into an online database managed by Medtronic, and all procedure-related and follow-up imaging data are mailed to the sponsor for review. A detailed schedule for pre-implant data collection and follow-up visits can be found in Online Table 2.
All selected patients underwent cardiac catheterization with general anesthesia using access from the femoral vein and in most cases, an additional femoral vein or right internal jugular vein for angiographic imaging during implantation. Angiography of the RV and RVOT was performed, as was coronary artery imaging (aortography or selective) if required.
Early in the implant experience, it was noted that cannulating the left PA was the most stable approach for delivery system placement and valve deployment. Once a stable wire position was secured in the distal left lower lobe PA with a Lunderquist wire (Cook Medical, Bloomington, Indiana), the delivery system was advanced to the PA. Deployment of the valve began in the proximal left PA by retracting the sheath and gradually exposing each strut until the valve was fully deployed in the main PA, whereas the most proximal strut was deployed into the distal RVOT. Once fully unsheathed in the RVOT, the valve was released from the delivery system with counterclockwise rotation of the delivery handle, allowing the coil loading system to release the frame. The delivery system was then withdrawn carefully to minimize the potential for entanglement with and dislodgement of the implant. This process was more difficult than originally expected, which will be discussed more thoroughly in the Discussion section.
For 2-dimensional imaging of the Harmony TPV, images were obtained from the parasternal long and short axis views of the RVOT or, if not feasible, the apical or subcostal positions. PR was assessed with color Doppler as none, trace or trivial, mild, moderate, or severe. Continuous wave and color Doppler were used to measure Doppler velocities across the RVOT and tricuspid regurgitation, which was graded as none, trace or trivial, mild, moderate, or severe (Online Table 3). In 1 implanting center, intracardiac echocardiography was used after implant for assessment of PR and the presence of a paravalvular leak.
All CMR studies were performed as part of routine clinical care using standard CMR imaging protocol for TOF patients, which includes balanced steady-state free precession cine images, phase contrast of the aorta and PA, and post-contrast 3-dimensional magnetic resonance angiography images. The end-diastolic volume, end-systolic volume, and ejection fraction were obtained from cine images. Aortic and pulmonary regurgitant fractions were calculated from phase contrast flow curves. Pre-implant cross-sectional measurements were obtained from 3-dimensional post-contrast magnetic resonance angiography images at multiple levels (RVOT, subvalve, midvalve, supravalve, midtrunk and pre-bifurcation) as well as length of the RVOT (from the RVOT to the pre-bifurcation) to determine enrollment eligibility and before performance of CT angiography scans.
All baseline CT images were obtained using a dual-source, multidetector CT scanner (SOMATOM Definition Flash, Siemens, Malvern, Pennsylvania). Pre-implant scans were used to characterize anatomy and create stereolithographic models (6). A CT scan following the same protocol was performed within 4 days after implant to assess device loading conditions (bending, torsion, and radial, linear, and axial compression) for primary endpoint assessment of device performance. The loading conditions data will be used as inputs to further testing of the current Harmony TPV, and for further device development.
Biplane cinefluoroscopy was used as the imaging modality in follow-up. Satisfactory views of the valve included those at implant and a “down-the-barrel” view (from the proximal opening to the distal opening of the implant). Radiographic assessment of potential stent fractures were classified as type I, II, or III following the Nordmeyer et al. (7) classification system used for the Melody TPV.
Clinical events committee and data safety and monitoring board
A joint Clinical Events Committee and Data Safety and Monitoring Board consisted of 2 cardiologists and 1 cardiothoracic surgeon, independent of Medtronic, and the study investigators. The Clinical Events Committee and Data Safety and Monitoring Board reviewed serious adverse events and conducted periodic reviews of data related to safety, data integrity, and the overall conduct of the trial.
Categorical variables are summarized by frequency, and continuous variables are presented as mean ± SD. Outcome results are reported only for the patients who underwent implantation.
Sixty-six subjects underwent CT angiography evaluation and 21 patients were approved for implant of the Harmony TPV device. Of 21 patients, 52% were men with a mean age of 28 ± 14 years. Most were diagnosed with TOF (n = 20) and had augmented RVOTs or transannular patch repairs (n = 19). Twenty patients had severe PR by echocardiography with minimal RVOT obstruction (mean RVOT gradient 11 ± 5 mm Hg). Baseline CMR images showed all patients with reported values had a pulmonary regurgitant fraction >30%; the mean indexed RV end-diastolic volume was 159 ± 34 ml/m2 (range 107 to 241 ml/m2). Table 1 summarizes baseline characteristics for catheterized patients. Figure 2 shows a study overview flow diagram. Figure 3 shows the dynamic behavior of the RVOT anatomy before implant.
Twenty of the 21 patients who underwent catheterization received a Harmony TPV. One patient, a 50-year-old patient with TOF and absent left PA, received a Melody TPV in the right PA (off-label use of the Melody TPV) due to investigator concerns about severe PA hypertension observed at catheterization. This patient’s PA systolic pressure was 94 mm Hg compared with a mean of 33 ± 7 mm Hg for the other 20 patients. In the 20 implanted patients, the Harmony TPV was implanted in the intended location within the RVOT. However, in 1 patient the device migrated proximally toward the RV after release and during delivery system removal. The device position was lower than desired but functioned adequately (no significant PR per the primary investigator) and was left in place. One patient had a concomitant left PA stent redilation. Before implant, angiographic data demonstrated a mean pulmonary annulus diameter of 25 ± 4 mm and a mean length (RV to branch PA) of 48 ± 9 mm (Table 2). Two patients were noted to have paravalvular leak (1 mild, 1 moderate) by intracardiac echocardiography that could not be seen on the discharge echocardiogram. Hemodynamic data before and after implantation are summarized in Table 3 and Figure 4. As expected, there was an increase in PA diastolic and mean pressure after valve placement. The mean procedural time was 129 ± 46 min.
All patients were free of device explant at 24 h, but in 1 patient the device migrated distally within 24 h. In this patient, the TPV was surgically explanted during the same hospitalization and within 48 h of the catheterization, and the patient was discharged 2 days after the explant surgery. Of the remaining patients, 17 were discharged the day after the procedure, and 1 was discharged on day 3 with episodes of nonsustained ventricular tachycardia that began during the catheterization procedure and continued after the implant. This patient was transferred to the step-down unit, and an implantable loop recorder was placed on day 2. Ventricular tachycardia was not present in follow-up. This patient did have a history of ventricular ectopy before referral. In the remaining patients only minor events related to the procedure were noted during the implant hospitalization, including bleeding controlled with manual compression, nausea, and musculoskeletal discomfort.
Device assessment after implant by CT
After implant, CT data were collected in 19 patients and demonstrated device conformation to the RVOT anatomy, with good apposition at the distal and proximal ends of the device, providing sealing and preventing migration. The central valve-housing region was generally unconstrained and not compressed by the anatomy. There was minimal elongation or shortening of the device, with all devices having length change less than 3%. TPV cyclic deformation (change in dimensions over cardiac cycle) showed wide patient-to-patient variation, with dimensional changes from 1% to 37%. Deformation of the TPV inflow, positioned near the native pulmonary annulus, was roughly twice the deformation seen in the TPV outflow and valve-housing sections, which are typically within the main PA. Compared to preclinical animal evaluations, the device cyclic deformation was on average much lower in human subjects. The cyclic deformation of the implanted TPV was also significantly lower than the native anatomy deformation. Example images showing the device and its interaction with the anatomy are shown in Figure 5.
After discharge and before the 1-month evaluation, 2 patients were evaluated for mildly elevated temperatures. Both patients had normal blood work and negative blood cultures and did not require readmission.
At the 1-month follow-up visit, 18 of the remaining 19 patients were asymptomatic. One patient complained of persistent fatigue and exercise intolerance. Transthoracic echocardiography indicated an increased RVOT gradient, and fluoroscopy demonstrated a type II stent fracture with associated partial frame collapse. These findings were confirmed at cardiac catheterization (Figure 6). The device was surgically explanted during the same visit, and the patient recovered without complications. Of the remaining 18 patients at the 1-month follow-up, PR was either not present or trace or trivial, and the average peak or mean RVOT gradient was 28 ± 14 mm Hg or 16 ± 8 mm Hg, respectively. There was 1 case of mild paravalvular leak, but the remaining patients had trace to no paravalvular leak. All patients had trivial or mild tricuspid regurgitation, with the exception of 1 patient with moderate tricuspid regurgitation that was stable from pre-implant findings and then graded as mild at 3- and 6-month assessment. Follow-up fluoroscopy was performed on all 18 patients. Two patients had a type I fracture (stent fracture without loss of stent integrity) identified on screening fluoroscopic imaging.
All remaining 18 patients returned for both the 3- and 6-month follow-up assessments. Echocardiographic data remained consistent with those observed at the 1-month visit. Of seventeen patients with evaluable PR data, 16 continued to have none to trace or trivial PR, and 1 patient had mild PR at the 6-month visit (Figure 7). At 6 months, mean RVOT gradient was 15 ± 6 mm Hg (Figure 8). There were 2 patients with mild paravalvular leak. Subsequent 3- and 6-month fluoroscopy demonstrated 1 new case of type I stent fracture, detected on a 6-month fluoroscopy exam. There was no change in stent fracture severity for the patient identified with a type I device fracture at the 1-month visit. Follow-up echocardiography and fluoroscopy data are summarized in Table 4. Based on the data collected during the 6-month visit, overall device integrity and functionality appeared to be well maintained in the 18 patients.
Explants and data monitoring
There were 2 explants, 1 after the implant procedure during the same hospitalization stay for nonemergent removal secondary to device migration and the other due to a type II stent fracture with loss of device structural integrity and RVOT obstruction at 1-month follow-up. The second event was classified as a serious unanticipated device-related event, leading to a halt in enrollment during Clinical Events Committee and Data Safety and Monitoring Board review. Bench testing and root cause analysis were conducted and submitted to the Food and Drug Administration for review. Screening committee evaluations thereafter included more detailed analysis and consideration of device position. However, a primary causative etiology was not determined, the device frame was not altered, and no new inclusion criteria were introduced.
The significance of pulmonary valve replacement procedures was not recognized until recent years, as PR is often well tolerated early after TOF repairs (8). However, severe RV dilation and PR can develop slowly over time, leading to further complications later in patients’ lives (9). With this understanding, the prevalence of surgical pulmonary valve replacement has increased over the last decade, coincident with the development of alternative, less invasive options such as TPVR. Currently a few patients receive off-label application of various TPVs for use in native RVOTs. Unfortunately, heterogeneity in the morphologic characteristics of the RVOT leads to a gap between patient needs and available percutaneous technologies.
Given the need for such devices, development of TPVR products is a clinical necessity. At present the patient who requires PVR but has a large native RVOT that is not anatomically appropriate for existing commercially available percutaneous valves has limited options aside from surgery. The Harmony TPV was developed to accommodate the TPVR needs of this patient population. This is the only TPV device designed specifically for the native RVOT that has undergone an early feasibility clinical trial in the United States and Canada.
The early feasibility pathway is a new and unique opportunity to bring the Harmony TPV and other design concepts into clinical trial sooner. Outside the United States, similar devices are being studied in ongoing clinical trials in China and Europe and have shown promising short-term valve performance (10).
The early clinical outcomes of the Harmony TPV demonstrate promising device performance and preservation of stent integrity in the majority of cases. Compared with baseline, patients had significant improvements in PR. By the 6-month follow-up, there were minimal changes in paravalvular leak incidence, mean RVOT gradient, or tricuspid regurgitation, demonstrating preserved device function. Two patients had the Harmony TPV explanted, 1 because of device migration within 24 h post-delivery and another because of a type II stent fracture at 1-month follow-up. In the first patient, the perimeter plot analysis suggested a “borderline fit” with a small amount of interference contact in the RVOT during systole only, similar to the interference contact in the distal PA during diastole only. The Screening Committee and the implanters agreed that this anatomy would provide a “learning experience.” After embolization, the importance of a larger interference contact area of the frame in both systole and diastole was recognized and was incorporated into the screening process. In the second patient, the frame fracture and infolding was unanticipated by the earlier animal studies and bench testing. After a stop in implanting, the Data Safety and Monitoring Board, Clinical Events Committee, and engineers analyzed the frame and recommended a larger “circumferential” contact area in the dynamic RVOT. An additional screening assessment was added going forward. Two additional patients exhibited a type I stent fracture with no significant associated symptoms. Because of observations during removal of the delivery system after deployment of the TPV device, a design change addressing the length and configuration of the distal “carrot” tip should be considered to allow for easier removal. Furthermore, the frame of the current device has suboptimal visibility on fluoroscopy, making precise positioning more difficult.
The results of the Harmony TPV study will be analyzed for product improvement and development of additional sizes to address the broad range of RVOT anatomies and of ways to maintain or improve device integrity. In this regard, Medtronic has begun a pivotal study to further examine the safety and efficacy of the Harmony TPV (NCT02979587).
This early feasibility study was limited by its small patient cohort size. A rigorous, multidisciplinary screening process was used to determine device-patient suitability, and only a small percentage of screened patients were ultimately selected for implant. This limits the generalizability of the results to a wider patient population. Furthermore, it is not feasible for all patients to be screened with CT angiography and the creation of a stereolithographic 3-dimensional model, as this process is time consuming, costly, and increases overall radiation exposure. Additionally, patient enrollment was limited to 3 sites, each with an experienced catheterization cardiologist performing the procedure. This might have skewed results and led to an unrealistically high procedural success rate that may not be reflective of procedural success within the general physician population.
Despite these limitations, this study presents the current application of the Harmony TPV for the native RVOT after transannular patch repair of TOF, demonstrating high rates of procedural success and promising early clinical outcomes. Continual follow-up and a larger patient population will be needed to assess longer-term durability, function, and safety of this device design to determine its feasibility as a novel transcatheter therapy for the dysfunctional native RVOT.
This is the first report presenting early clinical outcomes of the early feasibility study for the Medtronic Harmony TPV, and results indicate preserved valvular function at 6 months. The device was limited to a small group of patients due to anatomical constraints; however, with the development of more implant sizes, the Harmony TPV has the potential to serve a larger patient population that currently lacks TPVR technology.
WHAT IS KNOWN? RVOT surgery is common in babies with congenital heart disease, with nearly 75% of patients left with severe residual PR. A specifically designed nonsurgical, transcatheter option to restore pulmonary valve competence is needed to offer patients an alternative to surgical pulmonary valve replacement.
WHAT IS NEW? This first Food and Drug Administration–approved early feasibility study allowed an innovative valve design to be tested in these patients with complex right ventricular outflow tract anatomy and succeeded in restoring pulmonary valve function.
WHAT IS NEXT? The important data derived from the engineering analysis, perimeter plots, and stereolithographic models in these 66 patients will allow more valves and frame designs to be manufactured to treat the majority of patients with severe PR after surgical repair of complex congenital heart disease.
Overall study management for the Native Outflow Tract TPV Research Clinical Study was provided by Kristin Smith, MBA, and Kristin Boulware. Jessica Dries-Devlin, PhD, CMPP, assisted with tables and figures and ensured technical accuracy of the manuscript. All are employees of the sponsor, Medtronic.
For supplemental tables, please see the online version of this article.
The study was designed and funded by the sponsor, Medtronic (Minneapolis, Minnesota). Dr. Bergersen has served as a consultant for 480 Biomedical. Drs. Benson and Gillespie have served as consultants for Medtronic. Dr. S.L. Cheatham reports that her spouse is a consultant, proctor, and principal investigator for Medtronic. Dr. Hor has served as a consultant for Pfizer (no honoraria received in 2016), Myocardial Solution (honoraria received in 2016), Marathon Pharma (no honoraria received in 2016), and Bristol-Myers Squibb (no honoraria received in 2016). Dr. Horlick has served as a proctor and consultant for Medtronic; has served on the North American advisory board for Medtronic; and has served as a consultant for Edwards Lifesciences and St. Jude Medical. Dr. Lung and Mr. McHenry are employees and shareholders of Medtronic. Dr. Osten has served as a proctor for Medtronic. Dr. J.P. Cheatham has served as a consultant, proctor, and principal investigator for Medtronic. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- cardiac magnetic resonance
- computed tomography
- pulmonary artery
- pulmonary regurgitation
- right ventricle/ventricular
- right ventricular outflow tract
- tetralogy of Fallot
- transcatheter pulmonary valve
- transcatheter pulmonary valve replacement
- Received December 19, 2016.
- Revision received April 17, 2017.
- Accepted May 21, 2017.
- 2017 American College of Cardiology Foundation
- Rosengart T.K.,
- Feldman T.,
- Borger M.A.,
- et al.
- Malekzadeh-Milani S.,
- Ladouceur M.,
- Cohen S.,
- Iserin L.,
- Boudjemline Y.
- Meadows J.J.,
- Moore P.M.,
- Berman D.P.,
- et al.
- ↵Farb A, Abel D. Investigational Device Exemptions (IDEs) for Early Feasibility Medical Device Clinical Studies, Including Certain First in Human (FIH) Studies: Guidance for Industry and Food and Drug Administration Staff. 2013. Available at: https://www.fda.gov/downloads/medicaldevices/deviceregulationandguidance/guidancedocuments/ucm279103. Accessed June 28, 2016.
- Gillespie M.J.,
- Benson L.N.,
- Bergersen L.,
- et al.
- Nordmeyer J.,
- Khambadkone S.,
- Coats L.,
- et al.