TAVI vs. Surgical AVR: Choosing a PARTNER for relieving aortic stenosis.15 July 2011
By: Abdul Hameed
South Yorkshire Deanery
Transcatheter versus Surgical Aortic-Valve Replacement in High-Risk Patients. Placement of Aortic Transcatheter Valves (PARTNER TRIAL) N Engl J Med 2011; 364:2187-2198 June 9, http://www.nejm.org/doi/full/10.1056/NEJMoa1103510#t=abstract
For a webcast discussion with the lead author see http://www.theheart.org/article/1204701.do
The first randomised control trial (PARTNER-A) to directly compare Transcatheter- Aortic Valve replacement (TAVI) with surgical aortic valve replacement (AVR) in patients with severe aortic stenosis has reported that TAVI is as effective as an AVR in reducing mortality although at the expense of doubling the risk of stroke .
The first parallel arm of this large prospective multi-centre randomised trial was reported last year (PARTNER-B) , were TAVI demonstrated an impressive 20% absolute risk reduction in mortality as compared to standard therapy (which included balloon aortic valvuloplasty) in patients ineligible for a surgical AVR 
The prognosis of aortic stenosis following the onset of symptoms is poor and aortic valve replacement through open heart surgery (AVR) had remained the only means of improving outcomes for such patients. However given the increasing age of the population and a greater burden of co-morbidity a significant number of patients with severe AS, have been deemed to pose too-high a risk be eligible for conventional surgery.
Following the first human report of a percutaneous AVR (TAVI) in 2002 approximately 20,000 TAVIs have since been performed worldwide in high risk, non-operable patients  with outcomes similar to those estimated for AVR, as estimated by risk scoring systems. However the current PARTNER-A study is the first to directly compare the two approaches head to head. This trial compared implantation of the SAPIEN bioprosthetic heart- valve system, under general anaesthesia with conventional surgical implantation of a bioprosthetic AVR.
From an initial screen of 3105 patients, 699 patients with severe aortic stenosis who fulfilled the entry criteria below and were considered to be suitable for surgical AVR were randomised to either TAVI (n=348) or surgical AVR (n=351). Of the TAVI group 244 were assigned to the transfemoral and 104 to a trans-apical approach, based on suitability of peripheral arterial anatomy. Following a TAVI patients were treated with dual anti-platelet Rx for 6 months.
Inclusion criteria were patients with severe symptomatic degenerative calcific aortic stenosis as defined by an AVA<0.8cm2 and either a mean AVG>40mmHg or peak velocity>4m/sec. Patients with bicuspid or non-calcific aortic valve disease were excluded as were those with coronary disease requiring revascularisation. Additionally patients with LVEF<20%, severe MR/AR, a recent TIA/stroke, severe renal failure and with either a small (<18mm) or large (>25mm) were also excluded because of the profile of the catheter-mounted valve delivery system (22-24Fr).
Baseline characteristics confirmed a high risk population as defined by the following: mean age 84 years, >75% with CAD, >40% with prior CABG, 30% previous CVA, >40% with PVD, >40% COPD, 40% with pulmonary Hypertension, >40% with A.F and Logistic EuroScore 29%. Over 90% of the population were in NYHA Class III or IV but mean LVEF was 50%. Mean AVA was 0.6 -0.7cm2 with a mean AVG of 43mmHg.
Patients were followed up for a minimum of 1 year (Median 1.4). The trial was designed to test if TAVI was non-inferior to AVR.
Primary Endpoint (intention to treat principle)
There was a non-significant trend towards to reduced overall mortality in the TAVI group which fulfilled the criteria for non-inferiority. No differences in CV death were also observed. For comparison the 30day and 1year mortality rates for a virtually identical group of high risk patients comparing TAVI vs. standard therapy from the PARTNER-B trial were 5% vs. 2.8% and 30.7% vs. 50.7% respectively. It is also worthy of mention that 30day mortality rates for the patients who underwent surgical AVR were better (by 32%) then would have been estimated from the validated operative risk scoring algorithms.
Although it was an unpowered comparison the death rate at 30 days for trans-apical TAVI was higher than that for trans-femoral TAVI (3.7% vs. 8.7% in the as per treatment analysis but 3.3% vs. 3.8% in the ITT analysis). The TAVI arm had a non-significant increase in the rates of rehospitalisation at 1 year (18.2% vs. 15.5% p=0.38)
Procedural outcomes during TAVI (n=348)
During TAVI, 3 patients died during the procedure (c.f 1 in AVR). In 9 patients the valve embolised and in 2 patients this required a repeat TAVI whilst 5 underwent conversion to AVR. 3 patients had an aortic dissection (compared to 2 undergoing AVR) of note the majority of the operators at the participating centres were on an early part of the learning curve for TAVI (<2 procedures prior)
The overall stroke rate was more than doubled at both 30 days (3.1% absolute and 57% relative risk increase NNH 33) and after 1 year (5% absolute and 60% relative risk increase, NNH 20) and there was a greater frequency of major vascular complications (7.8% absolute, 70% relative risk increase, NNH 13). Unsurprisingly major bleeding and AF post Surgical AVR were significantly more frequent post AVR.
*=1) thoracic aortic dissection, 2) access site related vascular injury (dissection, perforation, rupture, pseudoaneurysm) leading to death/>3u blood Transfusion or end organ damage and 3) distal embolisation (non-cerebral) and 4) LV perforation
TAVI patients had significantly shorter ICU and overall hospital stays (8 vs. 12 days p<0.001) No differences in worsening renal function or the need for RRT were evident. A greater proportion of Patients undergoing TAVI were in NYHA class I/II by 30days compared to those undergoing a sternotomy but this was equally achieved by 1 year. Mean AVA improved significantly following either intervention and statistically greater in TAVI, although this may not be clinically significant (1.59+/-0.48cm2 vs. 1.44+/-0.47cm2 p=0.002). There was however a greater risk of paravalvular aortic regurgitation following TAVI (12.2% vs. 0.9%) at 30 days and 1 year (6.8% vs. 1.9%) although why this frequency reduced over time is not clear or discussed.
This indeed was a high risk group of patients with severe Aortic stenosis and TAVI was certainly as effective as surgical AVR with respect to improving survival rates at 1 year. It was also associated with shorter stays in ICU and hospital and significantly less bleeding events. The downsides were that it increased the rates of stroke/TIA at 30 days (1 for every 33 patients having a TAVI) and after 1year (1 event in 20 patients undergoing TAVI) and there was a greater frequency of major vascular complications (1 event in every 13 TAVI performed) Furthermore more patients were left with mod/severe aortic valve regurgitation (1 in every 20 patients), which in itself may pose medium-long term consequences.
The future of TAVI?
The Sapien valve system used in the trial was a first generation transcatheter system and current platforms are now 4th generation, which are much smaller (18Fr) and so should decrease vascular complication and stroke rates. These will be further reduced as operator experience increases and potentially with the routine use of embolic protection devices. If stroke rates are reduced then certainly TAVI will march even further forward and may well be tested in lower risk populations with aortic stenosis in whom surgery is indicated. Such optimism should be welcomed by both patients and interventionists alike, but only after rigorous evaluation of efficacy and longer term durability of TAVI have been convincingly demonstrated.
1. Smith, C.R., et al., Transcatheter versus Surgical Aortic-Valve Replacement in High-Risk Patients. New England Journal of Medicine, 2011. 364(23): p. 2187-2198.
2. Leon, M.B., et al., Transcatheter Aortic-Valve Implantation for Aortic Stenosis in Patients Who Cannot Undergo Surgery. New England Journal of Medicine, 2010. 363(17): p. 1597-1607.
3. Cribier, A., et al., Percutaneous Transcatheter Implantation of an Aortic Valve Prosthesis for Calcific Aortic Stenosis: First Human Case Description. Circulation, 2002. 106(24): p. 3006-3008.
4. Webb, J. and A. Cribier, Percutaneous transarterial aortic valve implantation: what do we know? European Heart Journal, 2011. 32(2): p. 140-147.
Number of hits: 3389 Add Comments