RESULTS
Five hundred patients have been enrolled in the SAVI-1 and SAVI-2 registries. The patient flow is displayed in ure 1. In SAVI-1, approximately 70% of centres were first-time users, and 15 (58%) new centres were included in SAVI-2. No centre enrolled more than 15% of patients. The centre list is provided in the Supplementary Material, Table S1.
Figure 1: Patient flow diagram.
Baseline and procedural characteristics are displayed in Table 1. Risk scores were slightly lower in the SAVI-1 compared to the SAVI-2 cohort (Logistic EuroSCORE-I of 22.3 ± 12.7% vs 24.7 ± 15.7%).
Table 1: Baseline and procedural characteristics
SAVI-1 SAVI-2 Total
n = 250 n = 250 n = 500
Baseline
  Age [years] 80.9 ± 6.3 80.7 ± 5.9 80.8 ± 6.1
  Male 126 (50.8)a 104 (41.6) 230 (46.2)
  Logistic EuroSCORE-I [%] 22.3 ± 12.7 24.7 ± 15.7 23.4 ± 14.3
  STS Score [%] 8.0 ± 5.9 11.9 ± 10.0 9.8 ± 8.3
Procedure
  Prior balloon valvuloplasty 240 (97.2)b 218 (87.2) 458 (92.2)
  Device sizes used
   S 84 (33.6) 68 (27.2) 152 (30.4)
   M 93 (37.2) 109 (43.6) 202 (40.4)
   L 73 (29.2) 73 (29.2) 146 (29.2)
  Deployment under rapid pacing 176 (71.3)b 9 (3.6) 185 (37.0)
  Post-dilatation 97 (38.8) 110 (44.0) 207 (41.4)
Procedure
  Successful implant 245 (98.0) 241 (96.4) 489 (97.8)
  Valve-in-valve procedure 2 (0.8) 6 (2.4) 8 (1.6)
  Conversion to surgery 3 (1.2) 0 3 (0.6)
Data are presented as mean ± SD or n (%). NA: not assessed. aunknown in two patients, bunknown in three patients.
Of all 500 procedures, in SAVI-1 and SAVI-2, implantation of the ACURATE TA valve was successful in 489 (97.8%) patients. In SAVI-1, 3 conversions to open-heart surgery occurred (1 intraprocedural event when the device was pulled into the left ventricle during delivery system withdrawal, 1 on post-procedure day 1 for severe aortic regurgitation, and another on day 7 for severe aortic regurgitation after tilted valve deployment). Two valve-in-valve procedures using another commercially available transcatheter heart valve were performed for severe aortic regurgitation. In SAVI-2, no conversion to open heart occurred, but 6 valve-in-valve procedures (of these, 1 was associated with coronary obstruction, 2 with valve migration 1 with coronary obstruction and valve migration), 1 unplanned use of cardiopulmonary bypass and 1 mitral valve apparatus damage/dysfunction. Both patients with coronary obstruction and valve-in-valve procedure were diagnosed with porcelain aorta at baseline.
Echocardiographic parameters were similar between the registries, with slightly less paravalvular leakage ≥2+ at 1 year in SAVI-2 (0.8% vs 4.1%) (Table 2, Fig. 2). The NYHA classification improved from 93.5% of patients in NYHA class III/IV at baseline to 13.0% at 1 year (Fig. 3, Supplementary Material, Table S2).
Figure 2: Echocardiographic parameters over time (SAVI-1 and SAVI-2). (A) Aortic valve haemodynamics, measured as mean gradient and effective orifice area (EOA), and (B) percentage of patients with paravalvular leakage. Values remained stable between early follow-up at discharge/30 days and 1 year.
Figure 3: New York Heart Association (NYHA) classification at baseline and follow-up (SAVI-1 and SAVI-2).
Table 2: Echocardiographic assessments
SAVI-1 SAVI-2 Total
n = 250 n =  250 n =  500
Baseline
  Effective orifice area [mm2] 0.71 ± 0.21 0.75 ± 0.31 0.73 ± 0.26
  Mean gradient [mmHg] 43.1 ± 17.4 42.2 ± 14.9 42.7 ± 16.2
n =  219 n =  203 n =  422
30 days Discharge/7-daysa
  Effective orifice area 1.44 ± 0.45 1.48 ± 0.43 1.45 ± 0.44
  Mean gradient [mmHg] 12.4 ± 5.8 12.5 ± 6.0 12.5 ± 5.9
  Paravalvular leak
   0: non/trace 159 (72.6) 141 (69.5) 300 (71.1)
   1+: mild 55 (25.1) 59 (29.1) 114 (27.0)
   2+: moderate 5 (2.3) 3 (1.5) 8 (1.9)
   3+: moderate/severe 0 0 0
   4+: severe 0 0 0
1 year n = 148 n = 126 n = 274
  Effective orifice area 1.51 ± 0.38 1.57 ± 0.40 1.54 ± 0.39
  Mean gradient 12.9 ± 5.3 11.1 ± 4.4 12.1 ± 5.0
  Paravalvular leak
   0: non/trace 111 (75.0) 80 (63.5) 191 (69.7)
   1+: mild 31 (20.9) 45 (35.7) 76 (27.7)
   2+: moderate 5 (3.4) 1 (0.8) 6 (2.2)
   3+: moderate/severe 1 (0.7) 0 1 (0.4)
   4+: severe 0 0 0
Data are presented as mean ± SD or n (%). Not all measurements were available for all patients displayed,.
a whichever was earlier.
Relevant safety outcomes are displayed in Table 3, and additional outcomes in Supplementary Material, Table S3. Notably, only SAVI-2 outcomes were categorized by VARC-2 criteria. In general, outcomes were similar between SAVI-1 and SAVI-2. Mortality at 30 days was 6.4% [95% CI: 3.3–9.4] for SAVI-1 and 7.2% [95% CI:3.9–10.4] for SAVI-2, and at 1 year 18.9% [95% CI:13.9–23.7] and 20.8% [95% CI:15.6–25.8] respectively, P = 0.64 (Fig. 4).
Figure 4: All-cause mortality per Kaplan–Meier estimate.
Table 3: Kaplan–Meier estimates of relevant clinical outcomes at 30 days and 1 year
SAVI-1a SAVI-2 Total HR
30 days
  Mortality 16 (6.4) [3.3–9.4] 18 (7.2) [3.9–10.4] 34 (6.8) [4.6–9.0] 1.12 (0.57:2.2)
   Cardiovascular 5 (2.0) [0.3–3.8] 9 (3.6) [1.3–5.9] 14 (2.8) [1.4–4.3] 1.80 (0.60:5.37)
   Non-cardiovascular 11 (4.5) [1.9–7.0] 9 (3.7) [1.3–6.0] 20 (4.1) [2.3–5.8] 0.82 (0.34:2.00)
  Strokeb 7 (2.9) [0.8–4.9] 4 (1.6) [0.0–3.2] 11 (2.2) [0.9–3.5] 0.57 (0.17:1.95)
  Myocardial infarction 1 (0.4) [0.0–2.0] 1 (0.4) [0.0–1.2] 2 (0.4) [0.0–1.0] 1.00 (0.06:15.99)
  AKI stage 3 NA 7 (2.8) [0.7–4.9] – NA
  Other TAVI-related complications post-discharge 0 0 0 NA
  New onset of atrial fibrillation/atrial flutter NA 20 (8.2) [4.7–11.5] – NA
  Permanent pacemaker 25 (10.2) [6.3–14.0] 25 (10.2) [6.3–13.9] 50 (10.2) [7.5–12.9] 1.01 (0.58:1.76)
1 year n = 241 n = 243 n = 484
  Mortality 47 (18.9) [13.9–23.7] 51 (20.8) [15.6–25.8] 98 (19.9) [16.3–23.3] 1.10 (0.74:1.64)
   Cardiovascular 15 (6.4) [3.2–9.4] 19 (8.2) [4.5–11.6] 34 (7.3) [4.9–9.6] 1.31 (0.66:2.57)
   Non-cardiovascular 32 (13.2) [8.9–17.4] 32 (13.8) [9.2–18.1] 64 (13.5) [10.4–16.5] 1.02 (0.62:1.66)
  Stroke 10 (4.2) [1.6–6.7] 7 (3.2) [0.8–5.5] 17 (3.7) [1.9–5.4] 0.72 (0.27:1.89)
  Other TAVI-related complications post-discharge 0 0 0 NA
Data are presented as n (%) [95% CI], and HR. AKI-acute kidney injury, HR-hazard ratio, calculated using the Cox proportional hazards regression analysis, NA-not assessed, TAVI-transcatheter aortic valve implantation, ano VARC classification, bnot further specified if disabling or not disabling.