Abstract
DECLARATION
None declared
DH has an Aortic Surgery Fellowship funded by Ethicon
MF is one of three thoracic aortic surgeons in Liverpool and the principal writer of this manuscript. AO and MK are the other two thoracic aortic surgeons on who's practice this work is based. They have both had an editorial role in writing. DH and MB are Aortic Fellows and have contributed significantly in literature searches, generating data summaries and in contributing writing to sections of the essay
Descending thoracic aortic aneurysms and thoracoabdominal aortic aneurysms are classified according to their longitudinal extent, at their greatest reaching from left subclavian artery to iliac vessels and involving two body cavities. 1 The natural history of thoracoabdominal aortic aneurysms is difficult to determine accurately; however, several studies have estimated survival in patients followed up after having been turned down for surgery or declining intervention.1,2 The outlook for these patients is poor.
In Crawfords original series (1986) 1 of 94 unop-erated patients, the two-year survival was 24%. A larger study by Perko et al. (1995) 2 documented similarly poor outcomes with five-year survival of 0.39 ±0.07, 0.23 ±0.06 and 0.18 ±0.05 for isolated thoracic aneurysms, thoracoabdominal aneurysms and abdominal aortic aneurysms, respectively. This has not improved despite current optimal medical therapy; however, predicting annual risk of rupture has become accurate allowing timely intervention in an elective setting.3,4 Risk of rupture is most strongly related to diameter with the so-called ‘hinge point’ being around 7 cm in the descending thoracic aorta, at which point 43% of those under follow-up will have ruptured or dissected. 5 In asymptomatic patients the indication for surgery occurs roughly when the annual risk of rupture is greater than the perioperative risk of death. 4
Intervention is a major undertaking for surgeon and patient. Open surgery has improved significantly over recent years; however, even in high-volume reference centre risks remain substantial for the largest of aneurysms (death [10%], paraplegia [7.5%] and renal failure [15.9%] 6 ). However, data from this same group 6 suggest much improved survival following surgery, with five-year survivals between 66% and 75%, depending on the severity of the aneurym, significantly better than that estimated in natural history studies. Relatively recently, endovascular approaches have been pioneered to circumvent the need for extensive high-risk surgery; however, these approaches have not been without their problems, including high cost and re-intervention rates. 7 Hybrid approaches with endovascular stenting and re-routing of visceral vessels have been published by several groups, including some in the UK, 8 but seem to offer no clear advantage. The literature remains controversial and no randomized trial has been performed comparing best medical therapy with surgery, stent or a hybrid procedure. American Heart Association guidelines 9 and international consensus statements 7 exist and provide a comprehensive comparison of these various approaches and we do not intend to reproduce the arguments here.
The picture in the UK remains particularly unclear with respect to prevalence, intervention, outcome and arrangement of services. This essay attempts to estimate the consumption of medical services by patients with this condition within the UK, attempts to understand the intervention rate and form and asks whether the UK can offer a service with adequate outcomes at acceptable costs.
What do we know about the burden of thoracic and thoracoabdominal aneurysms in the UK?
The incidence of thoracic and thoracoabdominal aneurysms, specifically within the UK, is difficult to estimate. Several international population-based studies have estimated the incidence rate of thoracoabdominal aneurysms at around six new aneurysms per 100,000 person years. 10 In the UK, we may use hospital episode statistics (HES) data, (www.hesonline) to get an indication of activity within the English National Health Service (NHS) (population 52 million) and National Statistics Service (www.ons.gov.uk) to get an indication of resulting mortality (England and Wales, population 55 million). HES ‘diagnostic’ data suggest around 1000 admissions per year related to this disease (‘thoracic’, ‘thoracoabdominal’, ‘with rupture’, ‘without rupture’). Cause of death data stated by Office of National Statistics suggest around 650 deaths per year (‘thoracic aortic aneurysm’ [ICD 171.1/2] and ‘thoracoabdominal aortic aneurysm’ [ICD 171.5/ 6], ‘with rupture’ and ‘without rupture’). The data set suggests that in 2010 there were only nine deaths from ‘ruptured thoracoabdominal aneurysms’ nationally, clearly in gross error and likely reflecting diagnostic and coding errors. For comparison, this compares with 3593 deaths (2010) from abdominal aortic aneurysm with rupture (ICD 171.3). Although crude, these data help us understand the level of consumption of services within the NHS.
What do we know about treatment in the UK?
Current intervention on thoracic and thoracoabdominal aneurysms within the UK is not well documented. It is performed by cardiac surgeons, vascular surgeons, interventional radiologists and a few cardiologists, and varies by region and local historically based arrangements. There is no single regulatory body and no single registry. Data are stored by the Society for Cardiothoracic Surgery, the Vascular Society and commercial companies including the UK Registry for Thoracic Stent Grafts. In addition, some registries are pan-European, making it difficult to extract UK-specific data. Figure 1 shows data acquired from Dr Foster describing ‘spells’ of elective and non-elective repair of thoracic or unspecified aneurysms in English hospitals. Although these figures also include intervention on ascending and arch thoracic aortic aneurysms, they give an interesting representation of the total activity and distribution of thoracic aortic activity with centres in England.
Distribution of interventional activity on all thoracic aortic aneurysms by centre within England (Dr Foster)
Table 1 describes known published outcomes for surgical, endovascular and hybrid approaches in the UK. Some of the data are historical and should be judged in that context. For open surgical approaches, data from St Mary's published in 1995 11 and 1999 12 have influenced approaches within the UK. Results in this study suggested very poor outcomes from an open approach. However, these data should be judged in their historical context with no or little use of perfusion, cell salvage or clotting products. More contemporary results presented by Thoracic Aortic Aneurysm Service at the Liverpool Aortic Symposium (2011) (www.aorticaneurysm) suggest results comparable with any international institution are achievable within the UK. In a series of 80 elective open repairs (60% Extent I and II [left subclavian to either visceral vessels or aortic bifurcation]), 30-day mortality was 6.25% and in-hospital mortality was 11.2%, with a permanent paraplegia rate of 2.5%.
Known outcomes for intervention on thoracic and thoracoabdominal aortic aneurysms in the UK
U, unreported
Overall results Crawford Extent l-V
Paper reports 9/23 (40%) ‘survivors’ (undefined) as well as a 40% 30-day mortality
Outcomes after including an additional 20 urgent patients
All done using a totally abdominal approach and a supracoeliac clamp
See ref. 8
See ref. 11
See ref. 12
See ref. 16
See ref. 14
See ref. 15
Liverpool Aortic Symposium, June 2011, www.aorticaneurysm.org.uk
See ref. 13
Contemporary results for open repair of suprarenal and Type IV thoracoabdominal aortic aneurysms. Richards et al., BJS 2009;45-9
Society for Cardiothoracic Surgery Bluebook 2003
Interestingly, within Scotland, the National Services Advisory Group nominated the Royal Infirmary of Edinburgh the sole national provider of thoracoabdominal aortic aneurysm intervention in April 2001 (www.nsd.scot.nhs.uk). The group has published outcomes for Type IV (abdominal) and what they describe as supracoe-liac aneurysms, via a totally abdominal approach with a 30-day mortality of 6% (3/53). Their government review in 2007 (www.nsd.scot.nhs.uk) does suggest a practice dominated by this group of patients (Extent IV, 60%), with patients undergoing Extent II repairs (left subclavian to aortic bifurcation) only 14%, respectively. This, together with the lack of provision for cardiopulmonary bypass, presumably reflects the lack of involvement of cardiac surgical services. Certainly, their outcomes for this particular extent are comparable with any international centre. However, the poor early outcomes from open surgery in the UK have undoubtedly driven attempts at total endovascular solutions and hybrid approaches reflecting some international practices.
Results from the Royal Liverpool University Hospital suggest acceptable early results from Extent II branched stent grafting 13 and other less complex pathologies published from London.14,15 However, UK commissioners have increasingly questioned the clinical and cost-effectiveness of endovascular approaches as international studies have shown mixed outcomes from this approach. Certainly, in our Institution, we are required to request funding for endovascular approaches on a case-by-case basis but require no such scrutiny for open surgical intervention.
Hybrid approaches with staged extra-anatomical bypass and endovascular stenting have been trialed in an attempt to reduce morbidity, mortality and cost.8,16 However, results have also been mixed and this approach has not provided the expected solution. The Society of Cardi-othoracic Surgeons has published limited data on outcomes from intervention on the descending thoracic aorta and thoracoabdominal aneurysms; however, the data are rudimentary and few conclusions can be drawn. 17
What is the solution for the UK?
Can and should the UK provide a service?
A key to answering this question is in deciding whether we believe there is evidence that intervention can alter the natural history of this disease process to provide either symptomatic or prognostic benefit. As suggested earlier, the international literature does document particularly poor survival for patients managed medically,1,2 improved significantly following intervention. 6 There are no data published on survival with and without intervention in the UK; however, unpublished survival in our own group of postoperative aortic patients in Liverpool compares favourably with published survival for medically treated patients.
At present, it is necessary to accept that international data on clinical effectiveness of intervention on thoracoabdominal aneurysms may be extrapolated to the English NHS; however, this is the subject of a recent call by the UK National Institute for Health Research. Having drawn the tentative conclusion that survival following intervention in the UK is improved, is there sufficient volume to underpin acceptable outcomes? This is uncertain, however, the Liverpool model, which predominantly accepts patients from The North West (population 7 million), has an annual activity of between 150 and 180 shared between three surgeons. Approximately 25-30 of these cases relate to intervention on thoracoabdominal aortic aneurysms, and internationally comparable results have been achieved (Table 1). Providing this model can be duplicated, it would suggest that in a population of over 50 million in England there is roughly sufficient volume for 4–5 centres. Interestingly, NHS Scotland, with a population of around five million, commissioned the Vascular Surgery Unit at the Royal Infirmary of Edinburgh to perform 25 interventions per year on the thoracoabdominal aorta from April 2001 (www.nsd.scot.nhs.uk). In their government review in 2007, their service was deemed clinically robust and offered value for money. These figures are comparable with the population in the North West and the activity through Liverpool. Based on this, we believe there is sufficient international evidence for intervention, sufficient UK volume and proof that acceptable outcomes can be achieved in the UK NHS framework.
How should the UK provide a service?
We believe there is a case for a national strategy with a few designated regional centres in a system analogous to heart and lung transplant services and paediatric cardiac services. This model certainly seems to function within the framework of NHS Scotland mentioned above. We believe the provision of services throughout the UK should be subject to a joint national review by Royal College of Surgeons, Society of Cardiothoracic Surgery and the Vascular Society with input from interventional radiology, intensive care and commissioners.
Crucial to service provision should be a team approach manifesting as a mandatory, inclusive and comprehensive multidisciplinary team (MDT) discussion on every patient. Each centre should have expertise in open surgery, endovascular approaches and hybrid procedures where indicated, including use of extracorporeal support. This should ensure that each patient is matched to appropriate treatment regimens and adequately consented. Each centre should be subject to regular audit to ensure adequate activity and outcomes. All national centres should commit to regular bench-marking exercises, exchange of processes, audit and provide training. Without being prescriptive, based on new strategic health authority clusters within England, it would be appropriate for a single centre within each of the four clusters, serving a population or around 10 million each and giving sufficient volume to drive quality outcomes.
What does the UK have to contribute?
Separate vascular and cardiac training
The UK is quite different to USA and Europe with training in cardiac surgery unlinked to vascular surgery. Vascular surgery is a separate specialty with its roots in general surgery with significant involvement in interventional radiology. Thus, training and service provision have influenced the development of services for intervention on the thoracoabdominal aorta. Effectively, this has led to a dependence on collaboration between specialties with a reliance on cardiac surgeons to provide adjuncts such as cardiopulmonary bypass, and on vascular surgeons and radiologists to provide guide wire skills and endo-vascular options. In some centres, collaboration has flourished to the benefit of patients; however, in other centres treatment has been concentrated to one form of intervention such as surgery or endovascular approaches. The benefits of this arrangement are in the fact that skills are highly developed in respective specialties and where collaboration occurs, the treatment options are impressive and outcomes are comparable with any international centre.
Strong multidisciplinary team processes
Cardiothoracic surgery practice in the UK has a strong tradition with involvement in MDTs, particularly in lung cancer surgery and oesopha-geal cancer surgery. In addition, new European guidelines on the management of ischaemic heart disease 18 place the MDT at the heart of decision-making. Currently, the only international guidelines on thoracoabdominal aortic aneurysm intervention, published by the American Heart Association, 9 make no mention of involvement of an MDT in managing cases. Liverpool has a strong MDT process ensuring intervention is personalized to patients depending on pathology and co-morbidities. Key to this is preoperative assessment by anaesthetics/intensivists and their inclusion in the whole perioperative process. The National Centre for Thoracoabdominal Intervention in Scotland advocate an MDT process and indeed of the 55 patients assessed in their 2007 report (www.nsd.scot.nhs.uk), only 22 were offered intervention suggesting a screening process.
Mandatory reporting of outcomes
The UK has mandatory reporting of all cardiac surgical outcomes. This ensures that surgeon-specific and centre-specific outcomes are reported on a three-year cycle. By so doing, quality as measured by mortality is assured.
Independent assessments
The UK has regular independent review of clinical and cost-effectiveness of intervention in the form of the National Institute of Clinical Excellence (www.nice.org). This body has reviewed approaches to thoracoabdominal aortic aneurysms and dissection recommending placing an MDT process at the centre of decision-making. 18
What is the case for the MDT?
As suggested, the MDT is central to many clinical pathways throughout the UK, however, are these any evidence for effectiveness? Certainly NICE Guidelines 18 and various society guidelines9,19 mandate its involvement and intuitively it is difficult to argue against the process. However, does it alter resection rates and survival in cancer, or reduce inappropriate PCI in treatment of ischae-mic heart disease?
In the case of thoracic aortic surgery in the UK, treatment modalities are highly specialized and originate from different specialities including cardiac surgery, vascular surgery and interventional radiology. Because of the highly specialized nature of each approach, the majority of centres, but not all, focus on a single approach, either surgery, stent or hybrid operations. This approach to service arrangement does largely ignore the fact that patient factors and anatomy of disease should ideally dictate the interventional approach. The Thoracic Aortic Aneurysm Service in Liverpool has advocated a true MDT approach, with all patients discussed in the presence of cardiac surgeons, vascular surgeons and interventional radiologists, preferably in the presence of an anaesthetist/intensivist. Specifically, the presence of a cardiac surgeon at the MDT, with knowledge of extracorporeal support, and active perioperative involvement should be mandatory. This approach ensures a robust application of the evidence and that intervention is tailored to the patient rather than visa versa.
There has been a tendency in some centres to super specialize in one form of intervention. This ‘one size fits all’ approach is not in the patient's best interests. To properly understand the risk and benefits of approaches it would seem appropriate to discuss every patient at an MDT; however, robust evidence for the process is admittedly lacking. A joint decision is particularly vital for patients relegated to best medical therapy in giving the clinicians shared decision-making and in considering patient-clinician shared ‘advanced directives’ for inevitable emergency admissions.
What is the alternative?
In the recent past thoracoabdominal aneurysms have remained untreated in the belief that intervention has exceptionally high risks and costs. Is it acceptable to relegate this disease as a ‘mode of death’? Activity, outcomes and survival within Merseyside, Cheshire and North Wales and NHS Scotland, suggest this should not be the case. It is a fallacy to suggest that this is a disease of old age and our experience suggests median age of our aortic practice is 63. It seems unthinkable that the UK cannot provide a regional and comprehensive service to treat this disease.
Conclusions
We believe there is sufficient volume and evidence of adequate outcomes for the UK to offer a comprehensive service for intervention on thoracic and thoracoabdominal aortic aneurysms. Indeed, the UK has several facets of service provision that would allow for development of an internationally unique and effective system. However, we suggest there is a need for a review of service provision by all stakeholders and a rationalization of services. A comprehensive and inclusive MDT must be central to the process.