Transcatheter aortic valve implantation for aortic stenosis

A randomised controlled trial (RCT) of 358 patients (PARTNER 1B) for whom surgical aortic valve replacement (SAVR) was unsuitable compared transcatheter aortic valve implantation (TAVI; n=179) with medical management (n=179). Patients who had TAVI had significantly lower all-cause mortality and cardiovascular mortality compared with medical management at a follow-up of 1, 2 and 5 years (31% compared with 51% at 1 year, 43% compared with 68% at 2 years and 72% compared with 94% at 5 years for all-cause mortality and 21% compared with 45% at 1 year, 31% compared with 62% at 2 years and 58% compared with 86% at 5 years for cardiovascular mortality).

In an RCT of 795 patients for whom SAVR was suitable but high risk (the CoreValve trial), a Kaplan–Meier cumulative probability analysis for all-cause mortality at 3‑year follow-up was 33% for TAVI compared with 39% for SAVR (p=0.068). In another RCT of 699 patients for whom SAVR was suitable but high risk (the PARTNER 1A trial), a Kaplan–Meier probability analysis for all-cause mortality at up to 5 years of follow-up was 68% for TAVI compared with 62% for SAVR (p=0.76). When data were pooled for both RCTs (based on an intention-to-treat [ITT] analysis), the hazard ratios did not show statistically significant differences between TAVI and SAVR for hazard of death (pooled estimates were risk ratio [RR] 0.89; 95% confidence interval [CI] 0.73 to 1.09, p=0.26 at 1 year and RR 0.95; 95% CI 0.79 to 1.13, p=0.55 at 2 years). There were no significant differences for cardiovascular mortality at 1 year (RR 1.05; 95% CI 0.79 to 1.39, p=0.73) and 2 years (RR 0.92; 95% CI 0.67 to 1.28, p=0.79).

In an RCT of 2,032 patients for whom SAVR was suitable but intermediate risk (the PARTNER 2A trial) there were no significant differences between TAVI and SAVR at 1- and 2‑year follow-up for all-cause mortality and cardiovascular mortality (all-cause mortality: 12% compared with 13%, p=0.69, at 1 year and 17% compared with 18%, p=0.45, at 2 years; cardiovascular mortality: 7% compared with 8%, p=0.4, at 1 year and 10% compared with 11%, p=0.38, at 2 years). In an RCT of 276 patients for whom SAVR was suitable but low to intermediate risk (the NOTION study) there were no significant differences between TAVI and SAVR at 1- and 2‑year follow-up for all-cause mortality and cardiovascular mortality (all-cause mortality: 5% compared with 8%, p=0.38, at 1 year and 8% compared with 10%, p=0.54, at 2 years; cardiovascular mortality: 4% compared with 8%, p=0.25, at 1 year and 7% compared with 9%, p=0.40, at 2 years). A systematic review including 2 RCTs and 6 observational studies of 16,638 patients for whom SAVR was suitable and not high risk (comprising 6,875 patients in an analysis) showed a non-significant difference in all-cause mortality for TAVI compared with SAVR (odds ratio [OR] 0.67; 95% CI 0.42 to 1.07, p=0.08, at 30 days; OR 0.91; 95% CI 0.67 to 1.23 at 1 year; OR 1.06; 95% CI 0.59 to 1.91 at long-term follow-up [more than 1 year]). In a systematic review of patients at low and intermediate risk from surgery including 4 RCTs (n=3,179 patients, including the CoreValve trial), in which patients had a mean Society of Thoracic Surgeons' (STS) risk score of 7%, TAVI was associated with a lower hazard of death at 2 years than SAVR when done by the transfemoral but not by the transapical route (transfemoral route: hazard ratio [HR] 0.79; 95% CI 0.66 to 0.94, [risk difference −3.0, 95% CI −0.8 to −4.9]; transapical route: HR 1.34; 95% CI 0.91 to 1.97).

The RCT of 358 patients (PARTNER 1B) for whom SAVR was unsuitable compared TAVI (n=179) with medical management (n=179). More patients were asymptomatic or had mild symptoms (New York Heart Association [NYHA] class I or II) in the TAVI group than those in the medical management group (at 2 years: 83% [79/95] compared with 42% [17/40], p<0.0001; at 3 years: 70% [49/70] compared with 50% [7/14], p=0.245; and at 5 years: 86% [42/49] compared with 60% [3/5], p=0.531; NYHA class was not significantly different at baseline among these groups).

In the RCT of 795 patients for whom SAVR was suitable but high risk (the CoreValve trial), a greater proportion of patients were in NYHA class I or II in the TAVI group than in the SAVR group at 1 month (83% compared with 73%, p<0.001) and at 6 months (84% compared with 79%, p=0.04). At 12 months, there were no statistically significant differences between the TAVI and SAVR groups (79% compared with 72%, p=0.10). In the other RCT of 699 patients for whom SAVR was suitable but high risk (PARTNER 1A), the proportion of patients in NYHA class I or II was the same (64%) for TAVI and SAVR at 12 months.

In the RCT of 2,032 patients for whom SAVR was suitable but intermediate risk (PARTNER 2A) there were no significant differences between TAVI and SAVR in the proportion of patients in NYHA class I or II at 1- and 2‑year follow-up. In the RCT of 276 patients for whom SAVR was suitable but low to intermediate risk (NOTION) there were no significant differences between TAVI and SAVR in NYHA class at 3‑month and 2‑year follow-up. The systematic review (4 studies; n=2,146) of patients for whom SAVR was suitable but low to intermediate risk found that TAVI was associated with an increased risk of heart failure symptoms (NYHA class II or more: OR 1.29; 95% CI 1.08 to 1.55) at 2‑year follow-up compared with SAVR.

The RCT of 358 patients (PARTNER 1B) for whom SAVR was unsuitable compared TAVI (n=179) with medical management (n=179). There was a significantly higher mean aortic valve area in the TAVI group than in the medical management group at 1‑year follow-up (1.6 cm² [standard deviation; SD 0.5] compared with 0.7 cm² [SD 0.3], significance level not given). Mean pressure gradient improved from 44.7 mmHg (SD 15.4) at baseline to 13.2 mmHg (SD 11.2) for TAVI and changed from 43.2 mmHg (SD 15.4) to 44.3 mmHg (SD 16.1) at 1 year for medical management (p values not reported). Left ventricular ejection fraction (LVEF) improved from 53.9% (SD 13.1) at baseline to 57.2% (SD 10.6) for TAVI and from 51.2% (SD 14.3) to 56.9% (SD 10.3) at 1 year for medical management.

In the RCT of 699 patients for whom SAVR was suitable but high risk (PARTNER 1A, TAVI [n=348] compared with SAVR [n=351]), there was a significantly higher mean aortic valve area in the TAVI group than in the SAVR group at 30‑day, 6‑month and 1‑year follow-up respectively (1.7 cm² [SD 0.5] compared with 1.5 cm² [SD 0.4], p=0.001, at 30 days; 1.7 cm² [SD 0.5] compared with 1.5 cm² [SD 0.5], p=0.01, at 6 months; and 1.6 cm² [SD 0.5] compared with 1.4 cm² [SD 0.5], p=0.002, at 1 year). Mean pressure gradients improved but were not statistically significant (9.9 mmHg [SD 4.8] compared with 10.8 mmHg [SD 5.0], p=0.04, at 30 days; 10.2 mmHg [SD 4.3] compared with 10.8 mmHg [SD 4.8], p=0.16, at 6 months; and 10.2 mmHg [SD 4.3] compared with 11.5 mmHg [SD 5.4], p=0.008, at 1 year). LVEFs improved but not statistically significantly (55.5 [SD 11.4] compared with 56.0 [SD 11.4], p=0.63, at 30 days; 56.2 [SD 10.8] compared with 56.8 [SD 9.9], p=0.56, at 6 months; and 56.6 [SD 10.5] compared with 57.1 [SD 10.3], p=0.64, at 1 year). Baseline values were not significantly different for all the outcomes. In the other RCT of 795 patients for whom SAVR was suitable but high risk (CoreValve trial, TAVI [n=394] compared with SAVR [n=401]), there was a significantly higher mean aortic valve area in the TAVI group compared with the SAVR group (1.70 cm² [SD 0.49] compared with 1.55 cm² [SD 0.51], p<0.001, at 1 year; and 1.79 cm² [SD 0.48] compared with 1.53 cm² [SD 0.52], p<0.0001, at 3 years). Mean pressure gradients also improved significantly (8.90 mmHg [SD 3.73] compared with 12.17 mmHg [SD 7.10], p<0.0001, at 1 year; and 7.62 mmHg [SD 3.57] compared with 11.40 mmHg [SD 6.8], p<0.0001, at 3 years). Baseline values were not significantly different for all the outcomes.

In the RCT of 2,032 patients for whom SAVR was suitable but intermediate risk (PARTNER 2A) the mean aortic valve area was significantly higher in the TAVI group than in the SAVR group at 30 days (1.7 cm² [SD 0.5] compared with 1.5 cm² [SD 0.4], p<0.001), and this persisted at 1 year (1.6 cm² [SD 0.4] compared with 1.4 cm² [SD 0.4], p<0.001) and at 2 years (1.5 cm² [SD 0.4] compared with 1.4 cm² [SD 0.4], p<0.001). Mean pressure gradients also improved significantly (9.7 mmHg [SD 3.5] compared with 10.9 mmHg [SD 4.3], p<0.001) at 30 days and this persisted at 1 year (10.7 mmHg [SD 4.5] compared with 11.5 mmHg [SD 4.4], p=0.001) and 2 years (10.8 mmHg [SD 4.6] compared with 11.7 mmHg [SD 4.8], p<0.001). LVEF was higher for the TAVI group than for the SAVR group (56.9% [SD 10.2] compared with 55.0% [SD 11.0], p=0.04) at 30 days but this was reversed at 1 year (55.9% [SD 11.2] compared with 57.2% [SD 9.9], p=0.04) and at 2 years (54.9% [SD 11.2] compared with 57.2% [SD 9.7], p=0.005). In the RCT of 276 patients for whom SAVR was suitable but low to intermediate risk (NOTION) there were significantly greater improvements in mean valve area in the TAVI group than in the SAVR group (1.7 cm² compared with 1.3 cm², p<0.001, at 3 months; 1.7 cm² compared with 1.3 cm², p<0.001, at 1 year; 1.6 cm² compared with 1.3 cm², p<0.001, at 2 years). There were no significant differences from baseline for mean pressure gradient.

In the RCT of 358 patients for whom SAVR was unsuitable (PARTNER 1B, TAVI compared with medical management) there were significant improvements in self-reported quality of life in patients in the TAVI group compared with those in the medical management group (Kansas City Cardiomyopathy Questionnaire [KCCQ] quality-of-life summary score for heart failure: mean difference [MD] 14.8; 95% CI 8.6 to 21.0, p<0.001, at 1 month and greater benefits at 6 months [MD 24.2; 95% CI 17.4 to 31.0, p<0.001] and at 12 months [MD 30.5; 95% CI 22.3 to 38.7, p<0.001]). At 12 months, the TAVI group also had higher SF‑12 physical and mental health scores (MD 5.7 and 6.4 respectively, p<0.001).

In both the PARTNER 1A and CoreValve trials (patients for whom SAVR was suitable but high risk), patients having TAVI using the transfemoral route reported a greater improvement in quality of life (measured using EQ-5D, where 0 equals dead and 1 perfect health-related quality of life) than those having SAVR (PARTNER 1A: average change of 0.08 [SD 0.25] compared with 0.02 [SD 0.25] at 1 month, 0.1 [SD 0.3] compared with 0.09 [SD 0.27] at 6 months and 0.09 [SD 0.23] compared with 0.08 [SD 0.08] at 1 year; CoreValve study: average change of 0.055 [SD 0.23] compared with −0.073 [SD 0.26] at 1 month; 0.053 [SD 0.22] compared with 0.04 [SD 0.17] at 6 months and 0.043 [SD 0.2] compared with 0.0003 [SD 0.02] at 1 year). When data from these 2 trials were pooled for the transfemoral route, the overall estimates for EQ‑5D showed statistically significant differences between the TAVI and SAVR groups at 1 month in favour of TAVI (RR 0.09; 95% CI 0.03 to 0.16; p=0.006). However, the differences were not significant at 6 months (RR 0.01; 95% CI −0.02 to 0.05, p=0.47) and at 1 year (RR 0.03; 95% CI 0.00 to 0.06, p=0.09). When data were pooled for transapical TAVI compared with SAVR (from PARTNER 1A) and non-transfemoral TAVI compared with SAVR (from the CoreValve trial), the overall estimates for EQ‑5D showed no statistically significant differences between the TAVI and SAVR groups (RR −0.03; 95% CI −0.09 to 0.04, p=0.44 at 1 month, RR −0.02; 95% CI −0.10 to 0.07, p=0.66 at 6 months and RR −0.02; 95% CI −0.09 to 0.05, p=0.58 at 1 year). There was a greater improvement in SF‑12 scores (both physical and mental) in the TAVI group than in the SAVR group at 1‑month follow-up (MD for physical summary scores 2.0; 95% CI 0.1 to 3.9, p=0.4 in PARTNER 1A and MD 4.9; 95% CI 3.1 to 6.7, p<0.001 in the CoreValve trial; MD for mental summary scores 5.4; 95% CI 3.1 to 7.7, p<0.001 in PARTNER 1A and 6.1; 95% CI 3.8 to 8.5, p<0.001 in the CoreValve trial). There were no statistically significant differences between TAVI using either the transfemoral or non-transfemoral route and SAVR at 12 months for both physical and mental scores. Statistically significant differences in favour of TAVI were reported for KCCQ quality-of-life summary score at 1‑month follow-up in both PARTNER 1A (MD 9.8; 95% CI 4.0 to 15.6, p=0.001) and in the CoreValve study (19.0; 95% CI 13.7 to 24.3, p<0.001) but did not persist at 6- and 12‑month follow-up. There were no statistically significant differences in KCCQ quality-of-life scores using either the transapical route in PARTNER 1A or the non-transfemoral routes in the CoreValve study.

In the systematic review of patients for whom SAVR was suitable but low to intermediate risk (n=2,146, including data from 795 patients in 1 study [CoreValve study] with a follow-up of 2 years), there was a non-significant difference in KCCQ quality-of-life summary score (22.2 for TAVI compared with 18.7 for SAVR, MD 3.5; 95% CI 1.9 to 8.9).

The RCT of 358 patients for whom SAVR was unsuitable (PARTNER 1B) compared TAVI (n=179) with medical management (n=179). TAVI had a statistically significantly lower hazard ratio for repeat hospitalisation because of aortic stenosis (including complications because of TAVI) than medical management at 2 year- (HR 0.41; 95% CI 0.30 to 0.58, p<0.001), 3 year- (p<0.0001) and 5‑year follow-up (p<0.001).

In the RCT of 699 patients for whom SAVR was suitable but high risk (PARTNER 1A, TAVI compared with SAVR) there was a non-significant difference in re-hospitalisation rates (59 [19%] compared with 45 [16%], p=0.38, at 1 year; 74 [25%] compared with 60 [22%], p=0.41, at 2 years; and 108 [42%] compared with 81 [34%], p=0.17, at 5 years). In the RCT of 795 patients (CoreValve trial; TAVI, n=394 compared with SAVR, n=401) for whom SAVR was suitable but high risk, there was no significant difference in re-hospitalisation rates (95 [27%] for TAVI compared with 64 [21.9%] for SAVR, p=0.087) at 3 years. In the RCT of 2,032 patients for whom SAVR was suitable but intermediate risk (PARTNER 2A) there were no significant differences in re-hospitalisation rates between TAVI and SAVR.

The specialist advisers listed key efficacy outcomes as procedural success, satisfactory device positioning, shorter length of hospital stay, haemodynamic improvement, improvement in left ventricular function, improved quality of life, improved exercise capacity, symptom relief, prolonged survival, reduced mortality and morbidity and reduced re-hospitalisation.