Author + information
- Received April 5, 2018
- Revision received July 23, 2018
- Accepted July 25, 2018
- Published online November 28, 2018.
- Hope Caughron, BAa,
- Dennis Kim, MD, PhDa,b,
- Norihiko Kamioka, MDa,
- Stamatios Lerakis, MDa,
- Altayyeb Yousef, MDa,
- Aneesha Mainia,
- Shawn Reginaulda,
- Anurag Sahu, MDa,
- Subhadra Shashidharan, MDc,
- Maan Jokhadar, MDa,
- Fred H. Rodriguez III, MDa,
- Wendy M. Book, MDa,
- Michael McConnell, MDa,
- Peter C. Block, MDa and
- Vasilis Babaliaros, MDa,∗ ()
- aDivision of Cardiology, Emory University School of Medicine, Atlanta, Georgia
- bDivision of Cardiology, Children’s Healthcare of Atlanta, Atlanta, Georgia
- cDivision of Cardiothoracic Surgery, Emory University School of Medicine, Atlanta, Georgia
- ↵∗Address for correspondence:
Dr. Vasilis Babaliaros, Emory University Hospital F606, 1364 Clifton Road, Atlanta, Georgia 30322.
Objectives This study compares 30-day, 1-year, and 3-year echocardiographic findings and clinical outcomes of transcatheter pulmonary valve-in-valve replacement (TPVR) and repeat surgical pulmonary valve replacement (SPVR).
Background In patients with adult congenital heart disease and previous pulmonary valve replacement (PVR) who require redo PVR, it is unclear whether TPVR or repeat SPVR is the preferred strategy.
Methods We retrospectively identified 66 patients (TPVR, n = 36; SPVR, n = 30) with bioprosthetic pulmonary valves (PVs) who underwent either TPVR or repeat SPVR at Emory Healthcare from January 2007 to August 2017.
Results The TPVR cohort had fewer men and more patients with baseline New York Heart Association (NYHA) functional class III or IV. There was no difference in mortality, cardiovascular readmission, or post-procedural PV reintervention at 30 days, 1 year, or 3 years. Post-procedural echocardiographic findings showed no difference in mean PV gradients between the TPVR and SPVR groups at 30 days, 1 year, or 3 years. In the TPVR cohort, there was less right ventricular dysfunction at 30 days (2.9% vs. 46.7%; p < 0.01), despite higher baseline NYHA functional class in the SPVR cohort.
Conclusions In patients with bioprosthetic PV dysfunction who underwent either TPVR or SPVR, there was no difference in mortality, cardiovascular readmission, or repeat PV intervention at 30 days, 1 year, or 3 years. Additionally, TPVR and SPVR had similar intermediate-term PV longevity, with no difference in PV gradients or PVR. The TPVR cohort also had less right ventricular dysfunction at 30 days despite a higher baseline NYHA functional classification. These intermediate-term results suggest that TPVR may be an attractive alternative to SPVR in patients with previous bioprosthetic surgical PVs.
Transcatheter pulmonary valve-in-valve replacement (TPVR) is an option for many congenital heart disease patients who require redo pulmonary valve replacement (PVR) (1,2). However, there are minimal data regarding intermediate- and long-term follow-up for such patients and no reports comparing current intermediate-term outcomes of TPVR with surgical PVR (SPVR) (3–9). This study reports 30-day, 1-year, and 3-year echocardiographic findings and clinical outcomes in patients who had surgical bioprosthetic pulmonary valve (PV) dysfunction and underwent either TPVR and or repeat SPVR.
We retrospectively identified 256 patients who had previous surgical bioprosthetic PVs and underwent either TPVR (67 patients) or repeat SPVR (189 patients) at Emory Healthcare from January 2007 to August 2017. A multidisciplinary heart team determined the interventional approach for each patient according to procedural history, patient risk, imaging, and anatomic findings. Patients with right ventricle (RV)-to-pulmonary artery (PA) conduits were included if they had a previous surgical bioprosthetic valve in the conduit. We excluded patients who needed concomitant surgical procedures on the left heart (e.g., aortic valve replacement, mitral valve replacement or repair, aortic root reconstruction), patients with active endocarditis (antibiotics at the time of procedure), and patients with previous mechanical, transcatheter, and homograft PVs. Patients who received both a repeat SPVR and a TPVR during the study period were only included in the cohort associated with their first eligible PVR. A total of 66 patients (TPVR, n = 36; SPVR, n = 30) met the inclusion criteria for this study.
Echocardiographic parameters were reported according to the guidelines of the American Society of Echocardiography definition (10). Echocardiographic findings reported at 30 days and 1 year were collected within 6 months of the stated time point. Echocardiographic findings at 3 years were collected within 1 year of the stated time point. RV systolic pressure was echocardiographically measured based on a tricuspid regurgitation (TR) jet and an assumed central venous pressure. Readmission data, mortality status, and repeat PV procedure dates were collected through follow-up appointments, hospital readmission records, and direct phone contact.
All TPVRs were performed with balloon expandable transcatheter heart valves via transfemoral or transjugular access (n = 1). Valve size was determined based on previous surgical valve size, echocardiographic findings, and multidetector computed tomography. Selected patients had pre-implantation stenting of the conduit or RV outflow tract. All surgical PVRs were performed via standard sternotomy with cardiopulmonary bypass. Surgical patients had their valves replaced within a new conduit according to the discretion of the surgeon.
Continuous variables were reported as median (interquartile [IQR]). The values were tested using Wilcoxon rank sum test. The categorical variables were evaluated with Fisher exact test. Composite survival and cardiovascular readmission curves were analyzed by Kaplan-Meier method and compared with the log-rank test. All tests of hypotheses were 2-sided and conducted at a 0.05 level of significance. All statistical analyses were performed using SPSS statistics version 24 (IBM Corporation, Armonk, New York).
We identified 66 patients with previous bioprosthetic PVs who met the aforementioned inclusion criteria. Of these patients, 36 underwent TPVR and 30 underwent SPVR. Baseline characteristics including lung disease, liver disease, diabetes, and previous arrhythmia were similar between the 2 groups (Table 1). There were fewer men in the TPVR cohort compared with the SPVR cohort (36.1% vs. 63.3%; p = 0.05) and lower baseline creatine in the TPVR group (0.8 [IQR: 0.6 to 0.9] vs. 1.2 [IQR: 0.8 to 1.0]; p < 0.01). The TPVR group had more patients with New York Heart Association functional class III or IV (75.0% vs. 33.3%; p < 0.01). The median age was 30.0 years (IQR: 24.0 to 38.5 years; p = 0.89). Both groups had a similar median number of prior sternotomies (TPVR 3.0 [range 2.0 to 3.0] vs. SPVR 2.0 [IQR: 2.0 to 3.0]; p = 0.18) with similar median years since the previous SPVR (TPVR 13.0 [IQR: 9.3 to 19.0] vs. SPVR 14.5 [IQR: 9.0 to 19.5]; p = 0.52).
The TPVR and SPVR groups also had similar previous PVR procedural characteristics. Both groups had an average previous surgical PV size of approximately 26 to 27 mm and a similar number of patients with previous RV-to-PA conduits (TPVR n = 9 vs. SPVR n = 11; p = 0.42).
More patients in the SPVR cohort received a procedure for combined pulmonary stenosis and regurgitation (TPVR 47.2% vs. SPVR 70.0%; p = 0.03), whereas more patients in the TPVR cohort received a procedure for isolated pulmonary regurgitation (TPVR 47.2% vs. SPVR 16.7%; p = 0.01). Figures 1A and 1B demonstrate the scope of TPVR and SPVR, respectively.
The procedural details are listed in Table 2. One-third of the transcatheter patients had their procedures performed under conscious sedation, with only 2 of these patients requiring conversion to intubation (1 patient for mild hemoptysis and the second after respiratory arrest) (11,12). All of the transcatheter patients received a balloon expandable valve; 16 SAPIEN XT or SAPIEN 3 valves (Edwards Lifesciences, Irvine, California) and 20 Melody valves (Medtronic, Minneapolis, Minnesota). Before valve implantation, 17 of the patients had at least 1 stent placed in the RV outflow tract or in the RV-to-PA conduit. In a subgroup analysis, there was no difference in regurgitation or PV gradients when comparing the TPVR group that had at least 1 stent placed and the TPVR group that had no pre-stenting.
All of the surgical procedures were performed through a traditional midline sternotomy. Most of the patients had a bioprosthetic PV implanted; only 1 patient received a PV homograft. Of the surgical procedures, 15 cases had a simultaneous conduit replacement with either a Contegra valve (Medtronic) (5 cases) or a bioprosthetic valve sewn into a new conduit (10 cases). Eight patients had concomitant procedures including tricuspid valve replacements, tricuspid valve repair, and ventricular septal defect closure.
Median follow-up was 776.0 (IQR: 367.5 to 1393.5) days. The short-term procedural outcomes are listed in Table 3. As expected, the TPVR cohort had fewer intensive care unit hours (0.0 [IQR: 0.0 to 0.0] h vs. 49.1 [IQR: 41.9 to 97.7] h; p < 0.01) and fewer post-procedural hospitalization days (1.0 [IQR: 1.0 to 1.0] days vs. 6.0 [IQR: 5.0 to 8.0] days; p < 0.01) compared with the SPVR cohort. Only 1 (2.8%) patient in the TPVR group had a minor bleeding complication compared with 11 (36.7%) patients in the SPVR group (7 minor and 4 major or life-threatening). Between the TPVR and SPVR cohorts, there were similar rates of postoperative infection, new atrial fibrillation or flutter, and new pacemaker implantation though the numbers are small. In the TPVR group, there were no repeat procedural interventions at 2-weeks post-procedure compared with 2 (6.7%) repeat surgical interventions in the SPVR group. The new PV internal diameter to body surface area ratio was the same in both the TPVR and SPVR groups (12.9 [IQR: 11.2 to 14.1] mm/m2 vs. 12.0 [IQR: 11.3 to 12.7] mm/m2; p = 0.09).
As listed in Table 4, intermediate-term outcomes were similar in both groups. The TPVR cohort had 1 patient death at 30 days compared with 0 in the SPVR cohort (p = 1.00). There were no additional mortalities in either cohort at 1 year or 3 years. The death in the TPVR cohort occurred after intraprocedural respiratory arrest with anoxic brain injury (difficult airway for emergent intubation).
There was no difference in the number of patients with cardiovascular readmission at 30 days (0.0% vs. 6.7%; p = 0.10), 1 year (13.9% vs. 13.3%; p = 1.00), or 3 years (19.4% vs. 16.7%; p = 1.00). There were no patients who required rehospitalization due to recurrence of heart failure within 1 year. At 3 years, there were 2 patients in the TPVR cohort (n = 2 of 36, 5.5%) and 1 patient in the SPVR cohort (n = 1 of 30, 3.3%) who required rehospitalization due to heart failure. The majority of patients who required a repeat cardiovascular admission had a primary diagnosis of arrhythmia or pacemaker placement (n = 13 at 3 years). A small number of patients were also readmitted for deep vein thrombosis (n = 1), vascular stent placement (n = 2), RV-to-PA conduit thrombus (n = 1), endarterectomy (n = 1), and cerebrovascular accidents (n = 1).
Rates of repeat PV intervention were similar between the 2 groups, with only 1 patient in the SPVR group requiring repeat PV intervention at 1 year. There were no cases of valvular stent fracture reported via echocardiography in the TPVR cohort. There were no cases of endocarditis at 30 days, 1 year, or 3 years in either the TPVR or SPVR cohort. Notably, in the patients with prior endocarditis (4 TPVR and 5 SPVR) there was also no evidence of recurrent infection. Figure 1C shows the Kaplan-Meier curve for composite outcomes.
All baseline echocardiographic findings were similar between the 2 groups (Table 5), with the exception of a lower average peak PV gradient in the TPVR cohort (38.5 [IQR: 23.4 to 54.7] mm Hg vs. 55.0 [IQR: 39.0, 73.0] mm Hg; p = 0.05). The number of patients with moderate or severe pulmonary regurgitation was the same between the TPVR and SPVR groups (86.1% vs. 86.7%; p = 1.00) and the prevalence of significant concomitant valve disease was similar between the 2 groups.
Echo findings were grossly similar between the 2 groups at 30 days, 1 year, and 3 years (Table 6). There was no difference in mean PV gradients between the TPVR and SPVR groups at 30 days (12.0 [IQR: 8.0 to 16.2] vs. 13.5 [IQR: 8.0 to 18.5]; p = 0.44), 1 year (14.0 [range 12.0, 15.0] vs. 19.0 [IQR: 8.0 to 22.0]; p = 0.30), and 3 years (12.0 [IQR: 9.0 to 18.3] vs. 21.0 [IQR: 13.0 to 39.6]; p = 0.30). Additionally, there was no difference in the prevalence of moderate or severe pulmonary regurgitation at 30 days (0.0% vs. 0.0%), 1 year (0% vs. 20%; p = 0.11), or 3 years (11.8% vs. 18.8%; p = 0.66). The rates of greater than or equal to moderate aortic regurgitation and mitral regurgitation were the same in both groups at all 3 time points.
In the TPVR group, there were fewer patients with greater than or equal to moderate RV dysfunction (RVD) at 30 days (2.9% vs. 46.7%; p < 0.01) and 1 year (10.5% vs. 40.0%; p = 0.07) compared with the SPVR group, but there was no statistically significant difference in the prevalence of RVD at 3 years (17.6% vs. 43.8%; p = 0.14). Both groups had a lower incidence of greater than or equal to moderate TR at 30 days post-procedure compared with pre-procedure. Seven patients in the SPVR group received a concomitant tricuspid intervention (1 tricuspid repair and 5 tricuspid replacements), while no patients in the TPVR cohort received a tricuspid intervention.
This study demonstrates that: 1) the mortality, cardiovascular readmission rate, and incidence of repeat valve procedures were the same in both the TPVR and SPVR cohorts at 30 days, 1 year, and 3 years; and 2) echocardiographic indicators of PV function and PV failure were similar in TPVR and SPVR cohorts at 30 days, 1 year, and 3 years. These findings suggest that TPVR may be an alternative to SPVR in patients with previous bioprosthetic PVR. The results of this study are limited to intermediate-term follow-up and should be confirmed with larger cohorts and long-term follow-up to evaluate the efficacy and longevity of TPVR as compared with SPVR.
As in similar studies of congenital heart disease populations, our patients were generally younger with fewer comorbidities when compared with other patients receiving valve replacements (13,14). Our results showed that patients undergoing TPVR had fewer short-term clinical complications compared with SPVR, but there was no difference in 30-day mortality, cardiovascular readmission, or repeat PV procedures between the 2 cohorts. In these younger patients, early discharge and ease of recovery may be more desirable, as it allows patients to return to work, school, and other daily activities. Additionally, there may be unmeasured psychological merits of TPVR in congenital heart disease patients due to shorter hospitalizations and easier recoveries in patients with a history of significant hospital exposure (15).
There were no cases of endocarditis at 30 days, 1 year, or 3 years in either the TPVR or SPVR cohort. Importantly, both TPVR and SPVR patients with a history of prior endocarditis did not have recurrence in the 3-year follow-up period. This may indicate that TPVR is a safe option for patients with a history of treated endocarditis. These patients will require close long-term follow-up.
Our study also found there is no difference in the mortality, cardiovascular readmission rate, or PV reintervention rate between TPVR and SPVR at 1 year or 3 years. The single death in the TPVR cohort was partially attributed to difficulty obtaining an airway. Due to the complexity of this procedure, having an additional anesthetist or intensivist may be beneficial. In our study, patients required an average of >2.5 procedures before their repeat PVR and thus a minimally invasive strategy of TPVR may be a preferred treatment option for degenerated pulmonary bioprosthesis. Overall, these results further support the utilization of TPVR as an alternative to SPVR to reduce the lifetime surgical burden in these congenital patients (16,17).
Through the baseline peak PV gradient was higher in the SPVR cohort, it has been suggested that an isolated peak PV gradient has lower validity in determining the degree of PV stenosis than a mean PV gradient (18). Both the peak and mean PV gradients were similar between the TPVR and SPVR cohort at 30-day, 1-year, and 3-year follow-up. This suggests that short- and intermediate-term bioprosthetic valve longevity and function are similar for both transcatheter valves and surgical valves in the pulmonary position (8,19,20).
In addition to concerns about valve durability after TPVR, it is suspected that transcatheter valves may result in a smaller effective orifice area when compared with surgical valves (21). In our population, the average post-operative predicted PV internal diameter to body surface area ratio was similar for the TPVR and SPVR cohorts. Thus, our results suggest that the difference in the final effective orifice area may not be substantial and further supports the use of TPVR as an alternative to SPVR.
In our study, the number of patients with moderate or severe TR post-procedure was higher in the TPVR group due to the concomitant tricuspid repair and replacement in the SPVR cohort. Importantly, in both the TPVR cohort (all without tricuspid intervention) and the SPVR cohort, there appears to be an improvement in the severity of TR after PVR alone. These data further support recent studies that suggest a sustained improvement in TR after both TPVR (22) and isolated SPVR (23–25). The appropriate management of TR in patients with additional valvular disease continues to be a clinical dilemma and therefore patients with significant concomitant TR may be better served with SPVR until transcatheter strategies become available to address the tricuspid valve (24–26).
Last, this study showed improvement in RVD in the TPVR cohort and a lower prevalence of RVD greater than or equal to moderate in the TPVR group compared with the SPVR group at both 30 days and 1 year. The difference in prevalence of RVD is no longer statistically significant at 3 years post-procedure due to an increase in the number of patients with RVD greater than or equal to moderate in the TPVR cohort. It remains unclear as to why patients with RVD who undergo TPVR demonstrate an improvement in RV function immediately post-procedure and then a regression at 3 years. This finding requires further investigation with a larger cohort of patients.
This was a nonrandomized, single-center retrospective study. The decision for TPVR versus SPVR was made by the heart team. The scope of TPVR and SPVR differs slightly, as SPVR allows for conduit exchange for severe conduit dysfunction and also allows for tricuspid repair or replacement for severe TR. Measurements for the internal diameter to body surface area ratio were calculated based on the predicted inner diameter instead of the measured diameter due to a lack of postoperative computed tomography scans in both cohorts. Due to the complicated nature of these congenital cardiac patients, it is challenging to make direct comparisons between the cohorts, despite similar baseline clinical and procedural characteristics. Additionally, our results cannot be extended to certain patient populations as we excluded patients with concomitant surgical procedures on their left heart, previous mechanical PVs, previous homograft valves, and active endocarditis. Additional long-term validation cohorts will be needed beneficial to confirm the results of this study.
In patients with bioprosthetic PV dysfunction who underwent either TPVR or SPVR, there was no difference in mortality, cardiovascular readmission, or repeat PV intervention at 30 days, 1 year, or 3 years. Additionally, TPVR and SPVR had similar intermediate-term PV longevity, with no difference in PV gradients or PVR. The TPVR cohort also had less RVD at 30 days despite a higher baseline New York Heart Association functional classification. These intermediate-term results suggest that TPVR may be an alternative to SPVR in patients with previous pulmonary bioprostheses.
WHAT IS KNOWN? TPVR is an option for many congenital heart disease patients who require a redo PVR. However, there are minimal data regarding intermediate- and long-term follow-up for such patients and no reports comparing current intermediate-term outcomes of TPVR to SPVR.
WHAT IS NEW? There was no difference in mortality, cardiovascular readmission, or repeat PV intervention at 30 days, 1 year, or 3 years between the TPVR and SPVR patients. In addition, TPVR and SPVR had similar intermediate-term PV longevity, with no difference in PV gradients or PVR. These intermediate-term results suggest that TPVR may be an alternative to SPVR in patients with previous pulmonary bioprostheses.
WHAT IS NEXT? Further studies with larger numbers of patients will be necessary to assess long-term patient outcomes and valve longevity.
Dr. Lerakis is a consultant for Edwards Lifesciences and Abbott Vascular. Dr. Babaliaros is a consultant for and received research grant support from Abbott Vascular and Edwards Lifesciences. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- interquartile range
- pulmonary artery
- pulmonary valve
- pulmonary valve replacement
- right ventricle/ventricular
- right ventricular dysfunction
- surgical pulmonary valve replacement
- transcatheter pulmonary valve-in-valve replacement
- tricuspid regurgitation
- Received April 5, 2018.
- Revision received July 23, 2018.
- Accepted July 25, 2018.
- 2018 American College of Cardiology Foundation
- Cabalka A.K.,
- Hellenbrand W.E.,
- Eicken A.,
- et al.
- Chatterjee A.,
- Bajaj N.S.,
- McMahon W.S.,
- et al.
- Meadows J.J.,
- Moore P.M.,
- Berman D.P.,
- et al.
- Haas N.A.,
- Carere R.G.,
- Kretschmar O.,
- et al.
- Gillespie M.J.,
- McElhinney D.B.,
- Kreutzer J.,
- et al.
- Malekzadeh-Milani S.,
- Ladouceur M.,
- Cohen S.,
- Iserin L.,
- Boudjemline Y.
- Babaliaros V.,
- Devireddy C.,
- Lerakis S.,
- et al.
- Jensen H.A.,
- Condado J.F.,
- Devireddy C.,
- et al.
- Frigiola A.,
- Hughes M.,
- Turner M.,
- et al.
- Holland J.E.,
- Cassidy A.R.,
- Stopp C.,
- et al.
- Jones T.K.,
- Rome J.J.,
- Armstrong A.K.,
- et al.
- Kurkluoglu M.,
- John A.S.,
- Cross R.,
- et al.
- Kogon B.,
- Mori M.,
- Alsoufi B.,
- Kanter K.,
- Oster M.
- Antunes M.J.,
- Rodriguez-Palomares J.,
- Prendergast B.,
- et al.