The prospective trial with Venus-P valve (No. NCT02590679) was a non-randomized study that enrolled patients with severe pulmonary regurgitation after surgical repair of congenital right ventricular outflow tract obstruction with trans-annulus or RVOT patch at 5 centers in China. Patients with pulmonary stenosis were excluded. Details of the trial design and protocols were provided in the supplemental method. Procedural and short-term results were reported previously [7,8]. The initial time frame of follow-up was 1, 3, 6, and 12 months after the procedure. An extended follow-up study was conducted in our center. Thirty-eight patients were included and provided written informed consent for annually follow-up and analysis of their CT imaging data (Fig. 1). The ethics committee of our institution approved the study protocol.

According to the trial protocol, follow-up was performed together with routine transthoracic echocardiography, electrocardiograph (ECG) and chest radiography at post-implant, 1 month, 3 months, 6 months, 1 year. The routine tests were also conducted in the additional follow-up study.

A 12-lead ECG was used to detect the new-onset arrhythmia, which was defined as any atrial or ventricular arrhythmia documented on 12-road ECG during routine follow-up. Only the first event was counted. Transthoracic echocardiography (TTE) was performed with commercially available machines to assess the hemodynamic performance of implanted bioprosthesis. Images of the RVOT were obtained from the parasternal long-and short-axis views. Pulmonary regurgitation (PR) was assessed with color Doppler grading as none, trace/trivial, mild, moderate, or severe. Continuous wave and color Doppler were used to measure transvalvular velocities and gradients across the RVOT. Tricuspid regurgitation (TR), assessing in the four-chamber view, was graded as none, trace/trivial, mild, moderate, or severe. Chest radiograph with anteroposterior and lateral views was obtained for the surveillance of stent fracture. Cardiac CT scanning were performed at 3 months after the implantation for the evaluation of stent shape and further geometric analysis. Cardiac magnetic resonance (CMR) was performed at 6 months after the implantation for assessment of changes in right ventricular ejection fracture (RVEF), right ventricular end-diastolic volume index (RVEDVI) and right ventricular end-diastolic diameter (RVEDD).

Adverse events including all-cause mortality, TPV reintervention, valve dysfunction, stent fracture and endocarditis were recorded. TPV reintervention was defined as any catheter or surgical procedures performed on the implanted TPVs. Valve dysfunction was defined as an increase in pulmonary mean gradient over 10 mmHg compared with baseline or more-than-moderate regurgitation evaluated by echocardiography. Stent fracture was diagnosed by chest X-ray and classified into 3 types [11]. Type I, one strut fracture without loss of stent integrity. Type II, fracture with loss of stent integrity. Type III, fracture associated with the separation of fragments or embolization. Endocarditis was diagnosed according to the modified Duke criteria.

CT was performed on a 64-slice scanner, Discovery HD 750 (GE Healthcare, Waukesha, Wisconsin). A contrast-enhanced protocol with 80 to 120 ml of iodixanol at 5 ml/s followed by 30 ml of normal saline. The detector collimation width was 0.625 mm. Detector coverage was 40 mm with 1.25 mm section reconstruction and the slice interval was 1.25 mm. Phase and angle-dependent dose modulation was used with titration of radiation dose. Images were reconstructed at 0.6 mm slices with 0.3 mm overlap and iterative reconstruction for evaluation at 10% intervals within the 0%–90% RR range. For optimal image quality, measurements were taken in diastole at 75% of the R-R interval in the local reconstruction software (AW server 2.0, GE Healthcare, Waukesha, Wisconsin).

Deformation was analyzed according to 2 geometrical aspects, stent expansion and circularity. The circularity ratio (CR), and expansion ratio (ER) were calculated [10] at the valve level where the three leaflets coaptate and one third of the stent struts are allocated to each of the three leaflets. CR was defined as the ratio of minimum to maximum external stent diameter (Fig. 2). The circularity ratio of <0.9 was considered as non-circular TPV. TPV expansion rate was the ratio of CT-derived external valve area to the nominal external valve area, and <0.9 was labeled as an under-expansion valve. Under-expansion or non-circularity was considered as severe if the ER or CR was under 0.8, and mild if the ER or CR was between 0.8 and 0.9. Geometrical details were analyzed with repeated measurements and the agreement for the deformed TPV was satisfying (Pearson’s correlation coefficient > 0.9).

Continuous variables were presented as mean and standard deviation, and categorical parameters were presented as the frequency with percentages as appropriate. Clinical and imaging variables were compared using student-t-test. Categorical variables were compared using the χ² test. Freedom-from-event estimates and curves were from the Kaplan-Meier method and presented as rate and 95% confidential interval. A p value < 0.05 was considered statistically significant. All calculations were performed using SPSS software (SPSS version 22; IBM Corp., Armonk, NY, USA).

Table 1 lists the demographic data of the cohort. The mean age at time of valve deployment was 24.5 ± 12.8 years and 2 patients were over 50 years of age. Most patients (92%) were diagnosed with TOF and had patch repaired RVOTs. All patients were successfully implanted with Venus-P valves in the native RVOT and were discharged with the valve in the desired location. There was no procedural complication occurred. Normal pulmonary valve function was achieved in all patients. More-than-moderate TR was present in 8 patients at baseline and 3 after the implantation (21.1% vs. 7.9%, p < 0.05) (Fig. 3). There was no significant difference in transvalvular gradient before and after the procedure. New York Heart Association Class I or II at baseline accounted for 21 patients and 37 after implantation (55.3% vs. 97.4%, p < 0.05). Favorable changes of right heart overload were demonstrated by CMR in 25 patients (Supplemental Figure S1), with decreases in RVEDVI and RVEDD, from 146.9 to 101.9 ml/m², and 40.9 to 34.1 mm, respectively (p < 0.05). No significant difference was seen in RVEF.

Detail CT analysis were performed in 35 patients except for 1 death and 2 lost from follow-up at 3 months. Under-expansion valve was detected in 22 patients (63%) and non-circular stent shape was found in 16 patients (46%). Severe under-expansion or non-circularity (ER or CR < 0.8) was detected in 3 and 4 patients, respectively. Characteristics between valves with normal shapes and non-circular/under-expanded shapes were compared in the Table 2. Under-expansion is more common in patients implanted with larger valves (>30 mm) (p = 0.04).

During a median follow-up of 4.8 years (range 0.3–8.1), 1 patient died of the multiorgan failure resulting from endocarditis at 3 months, who was a 14-year female with a history of psoriasis. One patient underwent the surgical explant at 1 year for the vegetations grown around the TPV and the histology demonstrated the diagnosis of endocarditis. Recurrence of significant PR occurred in 3 patients at a mean time of 26 months. All those 3 patients were classified as more-than-moderate PR on echocardiography and recommend to the close visit without any intervention. Five-year freedom from valve dysfunction was 89.3% (95%CI 81.1–97.6) (Fig. 4). Endocarditis occurred in 3 patients at a mean time of 7 months (Table 3). Except the 2 endocarditis cases mentioned above, the third patient was admitted to the hospital at 6 months because of a low-grade fever which brought down after 3-week antibiotic treatment.

Stent fracture was diagnosed in 7 patients at a median time of 2 years. All stent fracture cases were categorized in to type I according to the risk stratification, while the type II/III SF were absent. Freedom from any SF at 5-year was 83.5% (95%CI 73.8–93.2) (Fig. 4). SFs mainly occurred at the RVOT flare in 6 patients and 1 was detected at both the RVOT flare and the distal pulmonary flare. All SFs at RVOT flare were adjacent to the sternum (Fig. 5). Patients with non-circular valves showed a higher rate of stent fracture (31% vs. 5%, p = 0.02), while the prevalence of stent fracture was similar between groups with or without valve under-expansion (18% vs. 23%, p = 0.35).

Frequent ventricular premature beats occurred in 1 patient after implanted and disappeared spontaneously 1 week later. No extra arrhythmia or sudden heart death was seen during follow-up. One patient with pre-existing atrial fibrillation underwent went transcatheter ablation at 3 years after TPV implantation.

Although this study provides important new data, it nevertheless suffers from important limitations. This is a small study with limited numbers followed to 5 years and serial CT and CMR scans were not obtained throughout the follow-up period to better understand the development of frame changes of the Venus-P valve as well as the right ventricular remodeling. It is also not clear if the Venus-P valve also performs well in patients with elevated gradients in RVOT who were initially excluded from the trial.