Aortic Stenosis Decision Paper

Risk stratification of patients

28. 
    Patients with severe AS may be broadly categorised into operable or inoperable patients. Inoperable patients are patients who are considered too high-risk for surgery, who without aortic valve replacement would receive limited intervention for symptomatic relief (palliative care and medical management). High-risk patients are eligible for surgery but carry a high risk of mortality.
    
29. 
    Operable patients may be further stratified by their preoperative risk of mortality from surgical (or transcatheter) aortic valve replacement. There are standard preoperative algorithms for stratifying patients into low, moderate or high risk of operative mortality. The best of these algorithms in cardiac surgery, the Society of Thoracic Surgeons Predicted Risk of Mortality Score (STS-PROM), stratifies patients risk using about 30 variables (including among other factors presence of significant comorbidities, age, ethnicity, and prior cardiac surgery). Low risk is defined as an operative risk of death of less than 4% (STS-PROM <4%). Moderate or intermediate risk and high risk are defined as STS 4%≤ 8%, and STS>8%, respectively. The algorithm has shown a high degree of accuracy in international and New Zealand specific studies. Analysis of the STS database, comprising several hundred thousands of patients, has shown about 5% of AVR patients may be classified as high-risk, 15% as moderate-risk, and 80% as low-risk (see National Health Committee Transcatheter Aortic Valve Implantation Tier 3 report).
    
30. 
    Further detail is provided on TAVI, sutureless AVR, medical management and palliative care below. Figure presents a schematic estimating the number of patients with aortic stenosis for each treatment pathway. The number of patients with a primary hospital diagnosis of AS in 2012/13, 860, is used as a starting point and then the evidence gathered through the development of the Tier 2 and Tier 3 assessment process is used to calculate the number of patients at each decision-making or risk stratification point. A sensitivity analysis for the schematic is provided in the Tier 2 AS assessment. Note:
    

• 
  The high-risk patient pool is very small, with less than 30 patients per annum. TAVI, sutureless AVR and conventional surgical AVR are alternative options for high-risk patients.
  
• 
  Conventional surgical AVR remains the gold standard for moderate-risk patients, though emerging evidence is accruing for TAVI and sutureless AVR in this relatively small population.
  
• 
  Conventional AVR is the gold standard treatment for low-risk patients, comprising about 80% of patients eligible for valve replacement.
  
• 
  About a third to two-fifths of patients with severe symptomatic AS may be ineligible for surgical AVR. Most of these patients are ineligible due to significant comorbidities, meaning invasive intervention is likely to be futile.
  
• 
  In less than a third of inoperable patients TAVI may be feasible, and of these patients most will have significant comorbidities, meaning intervention may be futile. A small group of patients, less than 30 patients per annum, who are inoperable for technical reasons may benefit from TAVI with a reasonable prognosis. These patients have anatomical features that make them poor candidates for surgery but have fewer comorbidities. The evidence for this patient population is limited to a non-pre-specified secondary analysis of a randomised controlled trial and a Canadian TAVI registry.
  

• 
  Balloon valvuloplasty is not specifically identified in Figure . The procedure is very infrequently used in New Zealand. It may be used for palliation, as a bridge to TAVI or surgical AVR, or as a preoperative test for TAVI.
  

Figure : Estimation of operable and inoperable population and their possible treatment – 2012/13

Source: NHC analysis, adapted from Osnabrugge et al 2013 ⁽¹⁹⁾ and Makkar 2014. ⁽⁵³⁾ 95% confidence intervals are in brackets. Confidence intervals are approximated for the technically inoperable group.

Transcatheter aortic valve Implantation (TAVI)

31. 
    TAVI is a novel technology that is changing treatment choices for aortic stenosis internationally. It allows for the replacement of the aortic valve, without the need for a median sternotomy (incision that separates the sternum to open the chest cavity) or cardiopulmonary bypass (a machine that takes over the role of the heart and lungs) that accompanies conventional surgical valve replacement. The introduction of TAVI has caused much controversy amongst policy makers and medical practitioners, due to the high cost of the valve and the difficulty in defining its precise role within current models of care for aortic stenosis. The large volume of published research is such that it has been difficult for governing agencies to position the technology in the sector without recommendations becoming obsolete.
    
32. 
    In 2010/11, at the request of the National Health Board, Auckland DHB’s Clinical Practice Committee (Auckland CPC) provided an independent evaluation of TAVI. At that time, the National Service and Technology Review Advisory Committee (NSTR) had been discontinued and a reconfigured NHC not yet established. Based on the evidence available at the time, Auckland CPC recommended interim approval for high-risk surgical candidates, although it felt cost utility was heavily dependent on complication rates. Auckland CPC did not recommend TAVI for non-surgical candidates because of unfavourable cost utility, although it considered there was a high level of evidence of both safety and efficacy for inoperable patients. The Auckland CPC also indicated that it would be necessary to further review TAVI in 12 to 18 months' time, as more evidence became available.
    
33. 
    TAVI was added to the NHC's reactive work programme in August 2012, following discussions between the NHC, the manager of electives within the National Health Board and the chairs of the New Zealand Cardiac Network (NZCN) and the National Cardiac Surgery Clinical Network. The NHC published a briefing report on TAVI in September 2012. A literature review was then undertaken and a draft review presented to NHC in May 2013, alongside a proposed engagement process with key stakeholders. This report highlighted the need for a stakeholder workshop to address issues around TAVI's cost-effectiveness and financial sustainability.
    
34. 
    Stakeholder workshops were undertaken in 2014 (discussed below) concurrent to and informing the NHCs aortic stenosis assessment work.
    

Current Government policy for funding TAVI

35. 
    Ministerial approval for public funding for TAVI was provided in April 2011, with approval granted for high-risk surgical candidates, subject to nationally consistent criteria and patient prioritisation, including regional multidisciplinary patient review committees. Approval was not granted for inoperable patients, moderate-risk patients or low-risk patients and has not subsequently been granted. High-risk surgical cases were to be identified by having one or more of the following high-risk features:
    

• 
  Advanced age
  
• 
  Previous coronary artery bypass graft (CABG)
  
• 
  Heavily calcified aorta
  
• 
  High risk Logistic EuroSCORE or STS score
  
• 
  Previous chest irradiation
  
• 
  Previous coronary artery bypass graft (CABG)
  
• 
  Sum of comorbidities (such as renal, pulmonary, hepatic and cerebrovascular), severe pulmonary hypertension, previous chest radiation.
  

36. 
    TAVI patients were/are expected to benefit from the procedure in terms of quality of life improvement, with an estimated life expectancy greater than two years. The guidance stated that eligibility creep, using TAVI for non-high-risk patients, should be avoided. Here it was recommended that clear documentation of the patient selection processes and outcomes followed by regular audit were undertaken. A pathway of care for TAVI patients (Figure ) was documented highlighting four important steps:
    

1. 
   A clinical decision that the patient would benefit from a surgical valve replacement.
   
2. 
   A clinical review by two cardiac surgeons to evaluate risk benefit for AVR versus TAVI.
   
3. 
   Selection for TAVI is made by a TAVI selection group.
   
4. 
   Prioritised for access and informed patient consent.
   

Figure : Current TAVI selection pathway

Source: National Health Board Communication to Chief Executive Officers 7 October 2011

37. 
    Selection for TAVI is made by a TAVI selection group comprising at least one interventional cardiologist, non-interventional cardiologist, cardiac anaesthetist (or cardiac intensivist) and a non-cardiac physician, alongside two cardiac surgeons. As an alternative to conventional surgical aortic valve replacement, TAVI is included within the cardiac surgery targets.
    

NHC assessment of TAVI

38. 
    The Tier 2 assessment of AS recognises that TAVI needs to be considered as one intervention in the overall model of care for AS (refer next section). TAVI has a role for high-risk patients provided this population can be clearly and transparently defined and the application of TAVI carefully monitored. There is also potential benefit for technically inoperable patients, who are not currently eligible for public funding.
    
39. 
    The Tier 3 assessment of TAVI synthesises current evidence relating to the NHC's four assessment domains: clinical safety and effectiveness, economic, societal and ethical, and feasibility of adoption.
    
40. 
    Internationally, there are many different TAVI devices available and in development. The two most commonly used valves are the Medtronic CoreValve and Edwards Lifesciences Sapien valve. TAVI can be delivered via percutaneous access (transfemoral delivery), or via minimally invasive surgical access (transapical, transaortic/direct aortic or subclavian delivery).⁽²⁾ Different access routes may be appropriate for different patients, but the two main access routes are the transfemoral (TF) retrograde (against normal blood flow) approach and the transapical (TA) antegrade (in the direction of normal blood flow) approach. The transfemoral approach is the dominant approach used in New Zealand.
    
41. 
    Internationally, TAVI is supported by health technology assessment agencies and professional bodies for inoperable patients or patients at high risk of mortality from surgical AVR – where surgical AVR remains the gold standard treatment for low and moderate-risk patients. That includes current professional guidance from the American College of Cardiology/American Heart Association (2014), and the Joint Task Force on the Management of Valvular Heart Disease of the European Society of Cardiology (ESC) and the European Association for Cardio-Thoracic Surgery (EACTS) (2012). In practice, however, it appears that TAVI is already being performed in moderate and low-risk populations in some countries, particularly Germany.
    

Clinical safety and effectiveness

42. 
    Compared with surgical AVR in high-risk patients, TAVI appears to have similar survival and lower rates of major bleeding events; but is associated with higher rates of aortic regurgitation and major vascular complications. TAVI is also associated with higher rates of pacemaker implantation with the CoreValve but not the Edwards Sapien Valve. Permanent pacemaker insertion is three to four times greater using the CoreValve than the Edwards valve. Renal complications appear to be similar between TAVI and surgical AVR, but elevated using the transapical approach.
    
43. 
    For inoperable patients, compared with medical management, TAVI is associated with superior survival and reduced hospital readmissions, but higher rates of aortic regurgitation, vascular complications, stroke, and major bleeding events. All-cause mortality for surgically inoperable TAVI patients is still very high in randomised evidence, where it was 43% at two years in the PARTNER randomised controlled trial.
    

Economic

44. 
    Aortic valve replacement is a high-cost procedure, whether undertaken with conventional surgical AVR or TAVI. The lifetime cost of TAVI ranges from $84,000 to $98,000, for high-risk and technically inoperable patients respectively. The lifetime cost of surgery ranges from $90,000 to $101,000, while medical management is relatively inexpensive, costing about $15,000 per patient over their remaining lifetime.
    
45. 
    TAVI appear to be highly cost-effective but only in a very small group of patients, namely those patients who are at high risk from surgery. The high-risk AVR patient population in New Zealand (STS>8%) is relatively small, probably less than 30 patients per annum based on current AVR volumes. For these patients, TAVI appears to be cost-saving compared with surgery, largely due to the reduced length of stay associated with percutaneous intervention.
    
46. 
    Our cost-saving finding in high-risk patients is consistent with a recent UK analysis sponsored by the British Heart Foundation which found TAVI had an incremental saving of $3,000 (£1,300) per patient compared with surgical AVR in high-risk patients. A 2013 economic assessment by the Ontario Health Technology Assessment Committee (OHTAC) also found TAVI had an incremental cost saving of $5,000 (C$4,600) per procedure compared with surgery in high-risk patients.
    
47. 
    For the inoperable patient population, TAVI does not appear to be cost-effective, where the cost per quality adjusted life years is about $74,000. In the subgroup of ‘technically inoperable’ patients, patients who have fewer comorbidities and tend to be younger, the cost per QALY is $40,000, due to vastly better life expectancy. The evidence for this patient population is, however, limited to a secondary analysis of a single randomised controlled trial and registry. Sensitivity analysis shows that cost-effectiveness is most sensitive to the relative efficacy of TAVI compared with surgical AVR and medical management.
    

Budgetary impact

48. 
    The projected potential budgetary impact of TAVI depends on the population for whom TAVI is funded. On the basis of population growth of people over 65 years of age, TAVI volumes are expected to expand from 66 procedures in 2012/13 to about 84 procedures in 2019/20. For this base case scenario we project a cost profile of about $4.4 million (Confidence interval: $3.5m, $5.3m) in 2015/16, increasing to $6.7 million ($5.3m, $8.0m) in 2019/20, accounting for both the index admission cost and attributable follow-up costs.
    
49. 
    If TAVI is expanded to inoperable patients, including patients with significant comorbidities, intervention rates could grow to 210 procedures in 2019/20. For this scenario the additional cost (beyond the base case) would be $6.2 million ($5.0m, $7.5m) in 2015/16 increasing to $7.5 million ($6.2m, $9.3m) in 2019/20. That is, the total cost of TAVI would be approximately $14.2 million in 2019/20 if expanded to inoperable patients.
    
50. 
    If TAVI is expanded to ‘technically inoperable’ patients only, the intervention rate could grow to 113 procedures by 2019/20. Under this scenario, the additional cost (compared with our base case) would be about $1.5 million ($1.2m, $1.8m) increasing to $2.0 million ($1.6m,$2.4m) over the same time period.
    
51. 
    These scenarios assume an expansion of TAVI volumes. There is, however, little doubt that current volumes are well in excess of what would be expected if only high-risk patients were receiving TAVI in New Zealand. Sixty-six TAVI procedures were undertaken in 2012/13, whereas, on the basis of current AVR volumes, high-risk volumes should account for no more than 20 to 30 patients annually in New Zealand. This finding is corroborated by a recent retrospective study of all TAVI undertaken in New Zealand between 2008 and 2014 that found TAVI had been performed in moderate and high-risk patients. Two substitution scenarios are thus suggested: either TAVI should be limited to high-risk patients only, with moderate-risk patients converting to surgical AVR; or TAVI should be limited to high-risk patients with moderate-risk patients converted to AVR, and remaining TAVI volumes taken up by technically inoperable patients.
    
52. 
    Under the first scenario, TAVI volumes would be reduced to 20 to 30 patients per annum. With a proficiency requirement of at least 20 TAVI operations per annum, the scenario implicitly assumes a reduction of three TAVI centres to one. As surgical AVR does not appear to be significantly less costly compared with TAVI in moderate-risk patients, we do not project any savings through this scenario. The second substitution scenario assumes maintenance of current TAVI volumes (and allowance for population growth) with any shift to technically inoperable cases consequent to moderate-risk TAVI volumes being converted to surgical AVR. Under this scenario, we project an additional cost of $1 million ($0.8m, $1.2m) increasing to $1.7 million ($1.4m, $2.1m) over five years. This is less than the prior scenario, where technically inoperable volumes were in addition to base volumes, but not significantly less due to the implied conversion of moderate risk TAVI patients to surgical AVR.
    
53. 
    This represents our favoured scenario as it maintains capability across centres whilst avoiding further indication creep into lower-risk population groups. Base case volumes are still one-third in excess of forecast demand from high-risk and technically inoperable patients. Pragmatically, we are not suggesting a reduction in base case volumes, as the apparent excess volume provides some allowance for uncertainty in the estimation of demand. As volumes are not expected to expand significantly in the foreseeable future, TAVI centres should be limited to the current three: Auckland, Waikato and Canterbury.
    
54. 
    Under our preferred scenario, national TAVI volumes correspond to a rate of 10.7 patients per 100,000 population aged 65 years or older. This assumes a base population of 615,000 people aged 65 years or older in 2012/13, with a national publicly funded intervention rate of 66 TAVI in the same year. The budgetary impact from the preceding scenarios is presented in Table .
    

55. 
    
    

Table Future cost of TAVI under various expansion and substitution scenarios

Scenario		2015/16	2016/17	2017/18	2018/19	2019/20
Base case	Volume	71	74	77	80	84
	Total Cost	$4.4m ($3.5m,$5.3m)	$5.0m ($4.0m,$6.0m)	$5.6m ($4.5m,$6.7m)	$6.1m ($4.9m, $7.4m)	$6.7m ($5.3m,$8.0m)
Expansion
Inoperable		108 (86,131)	113 (90,136)	117, (93,142)	122 (97,147)	127 (101,153)
	Additional cost	$6.2m ($5.0m, $7.5m)	$6.5m ($5.2m,$7.8m)	$6.9m ($5.5m,$8.3m)	$7.2m ($5.8m,$8.8m)	$7.5m ($6.2m,$9.3m)
Technically inoperable		25 (12,37)	26 (13,39)	27 (13,40)	28 (14,42)	29 (15,44)
	Additional cost	$1.5m ($1.2m,$1.8m)	$1.6m ($1.3m,$1.9m)	$1.7m ($1.4m,$2.1m)	$1.9m ($1.5m,$2.2m)	$2.0m ($1.6m,$2.4m)
Substitution
High-risk only	Volume	23 (22,24)	24 (23,25)	25 (23,26)	26 (24,27)	27 (25,28)
	Additional cost	$0.0m	$0.0m	$0.0m	$0.0m	$0.0m
High-risk and technically inoperable	Volume	71	74	77	80	84
	Additional cost	$1.0m ($0.8m,$1.2m)	$1.2m ($1.0m,$1.5m)	$1.4m ($1.1m,$1.7m)	$1.6m ($1.3m,$1.9m)	$1.7m ($1.4m, $2.1m)

Note: Additional cost is additional to the base case, more detail on the estimated budgetary impact is contained in the Tier 3 TAVI assessment.

Societal and ethical

56. 
    The Midland region appears to have a higher TAVI intervention rate than other regions (by DHB of domicile), most likely owing to the high volume of TAVI undertaken in Waikato DHB. Patients with severe aortic stenosis tend to be elderly and can have significant morbidities. Mid-term outcomes, even with a successful procedure, can be very poor for these patients. Hence, appropriate patient selection is imperative to avoid futile outcomes. In technically inoperable patients, TAVI is a substitution for medical management rather than surgical AVR and thus raises issues of equity of access that would need to be worked through with the sector.
    
57. 
    Feasibility of adoption issues are covered in the model of care section below.