High Post-Procedural Transvalvular Gradient or Delayed Mean Gradient Increase after Transcatheter Aortic Valve Implantation: Incidence, Prognosis and Associated Variables. The FRANCE-2 Registry

Mean Gradient (MG) elevation can be detected immediately after transcatheter aortic valve implantation (TAVI) or secondarily during follow-up. Comparisons and interactions between these two parameters and their impact on outcomes have not previously been investigated. This study aimed to identify incidence, influence on prognosis, and parameters associated with immediate high post-procedural mean transvalvular gradient (PPMG) and delayed mean gradient increase (6 to 12 months after TAVI, DMGI) in the FRANCE 2 (French Aortic National CoreValve and Edwards 2) registry. The registry includes all consecutive symptomatic patients with severe aortic stenosis who have undergone TAVI. Three groups were analyzed: (1) PPMG < 20 mmHg without DMGI > 10 mmHg (control); (2) PPMG < 20 mmHg with DMGI > 10 mmHg (Group 1); and (3) PPMG ≥ 20 mmHg (Group 2). From January 2010 to January 2012, 4201 consecutive patients were prospectively enrolled in the registry. Controls comprised 2078 patients. In Group 1(n = 131 patients), DMGI exceeded 10 mmHg in 5.6%, and was not associated with greater 4-years mortality than in controls (32.6% vs. 40.1%, p = 0.27). In Group 2 (n = 144 patients), PPMG was at least 20 mmHg in 6.1% and was associated with higher 4-year mortality (48.7% versus 40.1%, p = 0.005). A total of two-thirds of the patients with PPMG ≥ 20 mmHg had MG < 20 mmHg at 1 year, with mortality similar to the controls (39.2% vs. 40.1%, p = 0.73). Patients with PPMG > 20 mmHg 1 year post-TAVI had higher 4-years mortality than the general population of the registry, unlike patients with MG normalization.


Introduction
Transcatheter aortic valve implantation (TAVI) is now a well-established alternative to conventional surgical aortic valve replacement (SAVR) in prohibitive, high-risk, and intermediate risk patients with symptomatic aortic valve stenosis [1][2][3][4][5][6]. The extension of indications for TAVI to patients at lower risk is, however, still a matter of debate in patients younger than 75 years of age, despite the last two low-risk randomized studies showing very promising results with the Sapien 3 and Evolute valves [7,8].
In order to standardize the definitions of valve-and patient-oriented durability outcomes and to enable the objective evaluation of existing and novel TAVI prostheses and to compare efficacy versus SAVR, a consensus statement was published by the European Association of Percutaneous Cardiovascular Interventions [9]. Their recommendations to define hemodynamic structural valve deterioration are based on the transprosthetic mean gradient (MG) and aortic regurgitation (AR) severity assessed by echocardiography.
MG elevation can be detected immediately after the procedure or secondarily during the echocardiographic follow-up. The delayed mean transvalvular gradient increase (DMGI) could be more related to valve deterioration, while the immediate post-procedural transprosthetic mean gradient (PPMG) generally represents valve under-expansion, prosthesis patient mismatch, or pressure recovery and high flow. Several studies analyzed incidence and variables associated with immediate PPMG elevation, particularly in valvein-valve procedures [10,11]. A few also studied long-term gradient progression and the impact on prognosis [12,13]. However, to our knowledge, these two different patterns of elevated gradients after TAVI have not been compared previously.
The aim of the present study was to identify the frequency, determinants, and influence on the prognosis of immediate high PPMG with DMGI in the FRANCE 2 (French Aortic National CoreValve and Edwards 2) registry.

Population
The design of the FRANCE 2 registry was previously described in the 1-, 3-and 5-year follow-up reports [12,14,15]. Briefly, the registry included all consecutive symptomatic patients (New York Heart Association class >II) with severe aortic stenosis (defined as valve area ≤ 0.8 cm 2 , mean valve gradient ≥ 40 mm Hg, or peak aortic jet velocity ≥ 4.0 m/s) ineligible for SAVR on heart team evaluation due to coexisting risk features. A total of 34 centers (all 33 French centers and 1 in Monaco) prospectively enrolled all patients undergoing TAVI between January 2010 and January 2012. All patients provided written informed consent for the anonymous processing of their data, and the institutional review board of the French Ministry of Health approved the registry. All patients received either a self-expandable device (CoreValve ReValving System, Medtronic, Minneapolis, MN, USA) or a balloonexpandable device (Edwards SAPIEN or SAPIEN XT prosthesis, Edwards Lifesciences, Irvine, CA, USA). The choice of prosthesis, approach (transfemoral, transapical, or subclavian), and anesthesia (general or local) was at the operator's discretion. Antithrombotic therapy was left to each individual patient's team to decide In the present study, patients without discharge transthoracic echocardiography (TTE) or without 6 or 12 months TTE were excluded from the present analysis. A total of three groups of patients were analyzed: (1) patients with PPMG < 20 mmHg without DMGI > 10 mmHg at 6 or 12 months (control group); (2) patients with PPMG < 20 mmHg and with DMGI > 10 mmHg at 6 or 12 months (Group 1); and finally, (3) patients with PPMG ≥ 20 mmHg at discharge (Group 2).

Transthoracic Echocardiography (TTE) Evaluation
TTE was performed on the same day as the follow-up visits: before hospital discharge, at 30 days, 6 months, 12 months, and then annually. Valve function was assessed in terms of mean gradient, orifice area, and presence and severity of aortic regurgitation (graded from 0 to 4, with higher grades indicating greater severity). The transprosthetic mean gradient was calculated using the modified Bernoulli formula (∆p = 4 × V 2 with v: velocity through aortic valve; mean gradient was calculated by averaging the instantaneous gradients over the ejection period, using the traced velocity curve, and was done by the software directly), and bioprosthesis surface was calculated using the continuity equation (AVA = (A LVOT × VTI LVOT )/VTI AS ; with AVA: Aortic Valve Area; A LVOT : Area at left ventricular outflow tract; VTI LVOT : velocity time integral of flow at left ventricular outflow tract; VTI AS : velocity time integral of flow through aortic valve). We chose a mean transprosthetic gradient cut-off at a ≥20 mmHg and ≥10 mmHg change from the post-procedural echocardiography, which corresponded to at least moderate hemodynamic structural valve deterioration according to the consensus statement by the European Association of Percutaneous Cardiovascular Interventions [9]. AR and valve area were not used as a component to define groups 1 and 2.

Follow-Up and Data Management
According to protocol, visits recording clinical status, events, and echocardiography were planned at 1 month, 6 months, and 1, 2, 3, 4 and 5 years. Data regarding clinical status, complications, and echocardiography were recorded. All adverse events, including mortality, were defined according to VARC (Valve Academic Research Consortium) criteria and were adjudicated by an independent committee. Data were recorded on a standardized electronic case-report form and was sent over the internet to a central database (Axonal). Database quality control was performed by checking data against source documents for 10% of patients in randomly selected centers. All fields were examined for missing data or outliers, and teams were asked to complete or correct data wherever possible. Outlying data were checked and excluded if erroneous; exclusion concerned less than 1% of the data.

Statistical Analysis
Continuous variables were expressed as mean ± standard deviation or median ± interquartile range according to distribution. Comparison between groups of associated variables used the Student's t-test, ANOVA or nonparametric tests for continuous variables, and the χ 2 test or Fisher's exact test for categorical variables. Cox proportional univariate analysis was used to identify variables associated with high PPMG or DMGI. Variables with p < 0.10 were selected for multivariate analysis. p-values ≤ 0.05 were considered to indicate statistical significance. Kaplan-Meier survival analysis was used to analyze all-cause mortality and a logrank test was used to compare mortality between the three groups. Comparisons between groups 1 and 2 and the control group were performed using the same Cox model. In the present analysis, due to the very low number of patients remaining in the groups 1 and 2 beyond 4 years, only the 4-year follow-up data were analyzed. All analyses used SAS software, version 9.2 (SAS Institute, Cary, NC, USA).

Results
From January 2010 to January 2012, 4201 consecutive patients underwent TAVI and were prospectively enrolled in the FRANCE 2 registry. Before hospital discharge, TTE was performed on 3478 patients. Of them, 2353 patients underwent TTE between the 6th and 12th months following intervention. The majority of patients (2209) had a PPMG < 20 mmHg at discharge. The control group consisted of 2078 patients with PPMG < 20 mmHg without DMGI > 10 mmHg at 12 months. Group 1consisted of 131 patients (5.6%) with PPMG < 20 mmHg and DMGI > 10 mmHg, and finally, Group 2 consisted of 144 patients (6.1%) with a PPMG ≥ 20 mmHg at discharge (Figure 1). 12th months following intervention. The majority of patients (2209) had a PPMG < 20 mmHg at discharge. The control group consisted of 2078 patients with PPMG < 20 mmHg without DMGI > 10 mmHg at 12 months. Group 1consisted of 131 patients (5.6%) with PPMG < 20 mmHg and DMGI > 10 mmHg, and finally, Group 2 consisted of 144 patients (6.1%) with a PPMG ≥ 20 mmHg at discharge (Figure 1). Overall clinical, procedural, and echocardiographic characteristics according to group are summarized in Table 1.  Overall clinical, procedural, and echocardiographic characteristics according to group are summarized in Table 1.

DMGI in Patients with PPMG < 20 mmHg (Group 1 versus Control Group)
In group 1, the mean gradient increased from 8.7 ± 3.2 mmHg to 19.5 ± 8.1 mmHg during the first year of follow-up ( Figure 2). A total of ninety-two patients (70.2%) had a mean gradient between 20 and 30 mmHg, 19 (14.5%) between 30 and 40 mmHg, and 1 patient had mean gradient ≥ 40 mmHg. Clinical, procedural, and echocardiographic characteristics of the patients in Group 1versus the control group are summarized in Table 1. In comparison to patients without increased mean gradient > 10 mmHg (control group), patients with DMGI had, at baseline, less severe symptoms (p = 0.006), higher blood pressure (p = 0.018), higher left ventricle Clinical, procedural, and echocardiographic characteristics of the patients in Group 1versus the control group are summarized in Table 1. In comparison to patients without increased mean gradient > 10 mmHg (control group), patients with DMGI had, at baseline, less severe symptoms (p = 0.006), higher blood pressure (p = 0.018), higher left ventricle ejection fraction (LVEF) (p = 0.032), lower mean gradient (p< 0.0001), and more frequent valve-in-valve procedures (p < 0.0001). There was no significant difference in incidence of post-procedural AR between Group 1versus the control (12.1% vs. 14.5%; p = 0.38). At 4 years, incidence of stroke (6.1% vs. 3.5%, p = 0.12) and acute heart failure (23.7% vs. 20.7%, p = 0.42) did not differ between Group 1 and controls. At the 4-year follow-up, there was no significant difference in all-cause mortality according to presence or absence of DMGI in patients with PPMG < 20 mmHg (p = 0.27 with the control group) (Figure 3). Multivariate analysis of factors associated with DMGI is presented in Table 2. NYHA class I or II (p = 0.0029), absence of high blood pressure (p = 0.029), valve-in-valve procedures (p < 0.0001), valve ≤ 23 mm (p = 0.0019), absence of pre-procedural aspirin treatment (p = 0.04), and lower PPMG (p < 0.0001) were independently associated with the occurrence of DMGI during the first year of follow-up after TAVI.  Multivariate analysis of factors associated with DMGI is presented in Table 2. NYHA class I or II (p = 0.0029), absence of high blood pressure (p = 0.029), valve-in-valve procedures (p < 0.0001), valve ≤ 23 mm (p = 0.0019), absence of pre-procedural aspirin treatment (p = 0.04), and lower PPMG (p < 0.0001) were independently associated with the occurrence of DMGI during the first year of follow-up after TAVI.

Patients with PPMG ≥ 20 mmHg (Group 2 versus Control Group)
Of the patients with elevated PPMG ≥ 20 mmHg (n = 144: 6.1% of the total population with TTE between the 1st and 12th month following intervention), 126 underwent TTE during the following year. Echocardiographic data were missing for 18 patients (lost to follow-up or death).
Interestingly, at 1 year, in the initial 144 patients of Group 2, 83 showed a decrease in MG below 20 mmHg, while 43 patients still showed MG ≥ 20 mmHg. Moreover, at 4 years, as shown in Figure 4, patients with a spontaneous reduction in MG (reaching MG < 20 mmHg during the first year) had lower mortality than those who still showed a mean gradient ≥ 20 mmHg (p = 0.025), with mortality similar to controls (39.2% vs. 40.1%, respectively; p = 0.73). On the other hand, only patients still showing MG > 20 mmHg at 1 year after TAVI had higher 4-year mortality than controls (54.3% vs. 40.1%; p = 0.007). Table 4 summarizes the peri-procedural characteristics and echocardiographic findings at discharge, 6 months, and 1 year for patients with a high PPMG according to MG progression during the first year of follow-up (MG decrease to <20 mmHg or persistence of MG ≥ 20 mmHg). Notably, at discharge, the indexed aortic valve area was smaller in patients with persisting MG ≥ 20 mmHg than in patients with a decreased gradient (0.67 ± 0.2 vs. 0.90 ± 0.28; p < 0.001).  Table 4 summarizes the peri-procedural characteristics and echocardiographic findings at discharge, 6 months, and 1 year for patients with a high PPMG according to MG progression during the first year of follow-up (MG decrease to <20 mmHg or persistence of MG ≥ 20 mmHg). Notably, at discharge, the indexed aortic valve area was smaller in patients with persisting MG ≥ 20 mmHg than in patients with a decreased gradient (0.67 ± 0.2 vs. 0.90 ± 0.28; p < 0.001). able 4. Baseline, procedural, and echocardiographic characteristics of patients with high post-procedural mean gradient Group 2), according to the evolution of the mean gradient over the first year of follow-up (decrease < 20 mmHg or stability 20 mmHg).

Decrease of the MG over
Year < 20 mmHg n = 83   Table 4. Baseline, procedural, and echocardiographic characteristics of patients with high post-procedural mean gradient (Group 2), according to the evolution of the mean gradient over the first year of follow-up (decrease < 20 mmHg or stability ≥ 20 mmHg).

Discussion
The present study reports mid-term clinical and one-year echocardiographic outcomes of patients prospectively included in the FRANCE-2 registry with initial or secondary increase in PPMG, representing the largest cohort of consecutive TAVI patients with available echocardiographic follow-up. The main findings were: (1) DMGI > 10 mmHg was found after discharge at 6 or 12 months in 5.6% of patients and was not associated with higher mortality at 4 years; (2) six variables were associated with DMGI: three clinical conditions (absence of high blood pressure, NYHA class I or II, absence of pre-procedural aspirin treatment), two procedural characteristics (smaller valve size, valve-in-valve procedures), and one echocardiographic parameter (lower PPMG); (3) PPMG ≥ 20 mmHg at discharge was identified in 6.1% of the total registry population and was associated with higher all-cause mortality at 4 years (48.7% versus 40.1%; p = 0.005); (4) two-thirds of patients with PPMG ≥ 20 mmHg at discharge showed a less than 20 mmHg decrease at 1 year and had all-cause mortality consistent with the control population as a whole; (5) four clinical variables (younger age, BMI ≥ 35 kg/m 2 , dyslipidemia, lower EuroSCORE), two procedural variables (prosthesis size ≤ 23 mm, valve-in-valve procedure) and one echocardiographic parameter (higher pre-procedural mean gradient) were independently associated with a higher PPGM; (6) incidence of AR ≥ 2 was similar in all 3 groups.

DMGI in Patients with PPMG < 20 mmHg
These patients had good post-procedural TAVI results with a secondary increase in MG, possibly indicating rapid structural deterioration of the valve. In line with previous reports, this secondary increase in MG after TAVI did not seem to be associated with excess early or medium-term mortality [16,17]. Moreover, no difference was found in terms of the incidence of stroke or heart failure. Consistent with the literature, smaller valve and valve-in-valve procedures were associated with early structural valve deterioration [17][18][19]. Interestingly, in the present analysis, pre-procedural aspirin treatment was associated with lower incidence of DMGI, while clopidogrel and anticoagulation treatment were not.

Patients with PPMG ≥ 20 mmHg
Two-thirds of patients with PPMG ≥ 20 mmHg after TAVI (66%) recovered MG < 20 mmHg at 1 year. Only those with persistent MG > 20 mmHg at 1 year after TAVI had higher 4-year mortality than the controls (54.3% vs. 40.1%; p = 0.007), while patients with MG normalization had similar mortality to the controls (39.2% vs. 40.1%; p = 0.73). It could be argued that this sub-group of patients with a decrease of MG < 20 mmHg over the first year did not have valve under-expansion but rather transient hyperflow through the aortic valve due to left ventricular hypertrophy, which spontaneously regressed at 6 months after potential remodeling of the left ventricle. Brian et al. suggest that early regression of left ventricular hypertrophy can be observed up to 6 months after TAVI [20]. In contrast, the patients with persistent MG ≥ 20 mmHg at 1 year probably have poor valve expansion (highly calcified aortic valve) or limited expansion (more valve-in-valve procedures; 30.2% for patients with persistent MG ≥ 20 mmHg vs. 1.2% for patients with a decrease in MG < 20 mmHg, p < 0.001). Nevertheless, in daily practice, reaching optimal valve expansion at the end of a procedure is a major factor in reducing mortality; however, MG > 20 mmHg should not lead to systematic balloon post-dilatation since it does not necessarily correspond to the under-expansion of the valve, but possibly to transient hyperflow. Inadequate implant apposition can also lead to a mean gradient greater than 20 mmHg and can be reduced by the appropriate choice of prosthesis and the use of preand post-dilation to decrease the final gradient. During in-hospital ultrasound evaluation, in the case of MG > 20 mmHg, a CT-scan could be proposed to analyze valve deployment in greater depth, with repeated transthoracic ultrasound evaluation at 6 months and 1 year to assess MG progression. If under-expansion is identified on the CT-scan, additional post-dilation can be discussed to improve the mean valve gradient, which has an impact on the patient's long-term prognosis.

Limitations
This study had limitations that need to be taken into consideration. First, the low number of patients remaining in Groups 1 and 2 after 4 years did not allow statistical analysis after that date. However, the FRANCE-2 population tends to have particularly high-risk baseline profiles, and the number of surviving patients falls rapidly over time. Second, the lack of systematic post-procedural CT-scans to evaluate percutaneous prosthetic valve expansion could also constitute a limitation, but this additional examination is not part of standard of care. A CT-scan could have also been interesting to search for the presence of subclinical leaflet thrombosis after implantation. In fact, the mean gradient in general is higher in patients with post-operative thrombosis and may be a risk for accelerated valve degeneration, and therefore an indication of an early gradient increase. Third, the lack of systematic core laboratory evaluation for echocardiographic assessment was another limitation, notably with no formal identification of the mechanism of transient hyperflow through the aortic valve for patients in Group 2 with an MG decrease below 20 mmHg. Finally, the present analysis includes only the earlier generation of percutaneous aortic valves, and it would be interesting to validate the findings with the latest generation of valves.

Conclusions
DMGI during the year after TAVI in patients with initial PPMG < 20 mmHg was mostly seen in small percutaneous valve and valve-in-valve procedures and was not associated with a significant increase in 4-year mortality. In contrast, patients with persistent PPMG > 20 mmHg at 1 year after TAVI had higher 4-year mortality than the control FRANCE 2 registry population in contrast to patients with MG normalization.
Author Contributions: All authors listed have made a substantial, direct, and intellectual contribution to the work and approved it for publication. All authors have read and agreed to the published version of the manuscript.
Funding: Edwards Lifesciences and Medtronic partly funded the FRANCE TAVR registry. Edwards Lifesciences and Medtronic had no role in data management, data analysis, or writing of the manuscript.

Institutional Review Board Statement:
The study was conducted according to the guidelines of the Declaration of Helsinki and approved by the French Ministry of Health.
Informed Consent Statement: Informed consent was obtained from all subjects involved in the study. Written informed consent has been obtained from the patients to publish this paper.
Data Availability Statement: The data will be not available online.

Conflicts of Interest:
The authors declare no conflict of interest.