The EUROpean collaborators on Stent/graft Techniques for aortic Aneurysm Repair (EUROSTAR) have responded to the Belgian Health Technology Assessment (HTA) Group’s report published in November 2005 (and reported in issue 29 of Vascular News) that concluded endovascular aneurysm repair (EVAR) was “a failed experiment” and was “not cost effective”. However, the EUROSTAR report demonstrates that the evidence for “the rejection of EVAR for routine clinical practice cannot be justified” and that “there are currently no reliable data upon which to base calculations of cost-effectiveness.”
The latest report entitled, “EUROSTAR Report On Belgian Endovascular Abdominal Aortic Aneurysm Elective Repair Between 2001 And 2005 – EUROSTAR Report On The RIZIV Convention,” shows that outcome of Belgian patients following endovascular repair of AAA compare “very well” with standard for Europe as a whole as determined by the EUROSTAR registry. It is interesting to note that the research by the Belgian group only used the results from only the first four years from the EUROSTAR database (until June 2004), while the data collection in the latest report from the EUROSTAR contributors report cites data up to September 2005. This latest report compares data at baseline and outcomes of Belgian patients compared with other EUROSTAR patients.
The complete EUROSTAR data shows that the number of participating Belgian hospitals increased from 26 in April 2001 to 75 in October 2005. During that study period 2,068 patients were recruited on to the database, with an annual recruitment rate of approximately 500 patients. The data shows that 55% of the hospitals treated 20 or fewer cases during the study period, for an average of less than five operations per year. Over 40 patients were treated in 14 hospitals (19%) and over 80 in four hospitals (5%). Importantly, the report notes that there is no correlation between the 30-day or in-hospital mortality and case volume of the institution, for thresholds of 20 or 40 patients treated during the study period.
Between 2001 and 2005 the Belgian patients constituted 37% of the overall number of patients enrolled onto the database. The vast majority of the Belgian patients were male (94%) and the mean age was 73 years. No relevant differences were observed in age distribution between the Belgian and other EUROSTAR patients.
The operative risk profile represented by ASA-class 3 and 4 combined was significantly lower in the Belgian cohort (31%) than in the overall EUROSTAR group (57%). In contrast, hyperlipedemia, carotid artery and pulmonary disease were significantly more frequent in Belgian patients.
The mean aneurysm diameter of Belgian patients was 2.5mm smaller than in the other EUROSTAR population. According to the report authors, this difference is because all Belgian patients were included in the analysis while patients with aneurysms of <40mm diameter were excluded from the cohort of non-Belgian patients. However, it is important to note that 170 of the 291 patients with a diameter <50mm had a common iliac artery aneurysm (diameter larger than 25mm) and Iliac aneurysms may have been the predominant indication for the stent-graft procedure in this group. Overall, 28% of all patients with a recorded iliac diameter had an aneurysmal iliac artery. The report notes this measurement was often missing from the data forms. Another indication of the high prevalence of iliac aneurysms was the 20% incidence of patients with class D and E AAA morphology i.e with associated iliac disease. The investigators report that further assessment of the prevalence of small aortic aneurysm diameters demonstrated no correlation with low or high volume hospitals. The prevalence of any “problem” during the operation was significantly lower in Belgian patients (11% as opposed to 17%). Not all problems were of serious nature and type II endoleaks occurred relatively frequently (in 12.5%). The report notes that the overall 30-day or in-hospital mortality was slightly higher in the Belgian cohort (3.1%) than in other EUROSTAR patients (2.3%). There were significant correlations between mortality rate, age, fitness of the patient, ASA-class, and to a lesser extent, with the size of aneurysm. The all-cause mortality was 8.9% and 15.2% at one and two years of the follow-up, respectively, while aneurysm-related mortality was 3.7% and 4% over the same time periods. No differences were observed between these mortality rates and those recorded in other European patients. Again, no significant correlations were found between mortality rates and hospital volumes. Post-EVAR rupture of the aneurysm occurred at a very low rate (0.9% at four years of follow-up) in Belgian patients, which was comparable to the rate of rupture in the other EUROSTAR patients (1.7%), and the incidence of late conversion to open surgery was similar in the two comparative groups (4.3% vs. 4.2% in the other EUROSTAR patients). Therefore, the authors noted that procedural complication rates including 30-day mortality were statistically identical, and at follow-up all-causes mortality, aneurysm-related mortality, complication and secondary intervention rates were also the same. There were some differences between practice in Belgium and the rest of Europe in respect of the indications for treatment, with the aneurysms in the Belgian cohort slight smaller with a mean diameter of 56.8mms compared to 59.3mms for the others. Additionally, the Belgian patients were slightly fitter; 31% being classified within ASA Classes 3 and 4 compared to 57% of the others. The report confirms that these results raise some questions about the appropriate of the indications for endovascular aneurysm repair that were followed in Belgium. Indeed, the HTA researchers proposed increasing the size of the aneurysm diameter from 5cm to 5.5cm that would qualify for EVAR (ruling out half of the aneurysms treated, thereby halving costs). However, the EUROSTAR reports states, “In the absence of any data relating specifically to endovascular repair it is generally agreed that the same criteria for intervention that apply to open surgery should be exercised also in respect of EVAR (UK Small Aneurysm Trial demonstrated that in men there is no benefit associated with early surgery rather than ultra-sound surveillance for aneurysms of less than 5.5cms diameter).” The report states that another reason to operate on small aortic aneurysms is the co-existence of iliac artery aneurysms that require treatment – it is usually advisable to deal with the aortic aneurysm at the same time. There are indications that the Belgian cohort of patients included a relatively large proportion of patients with concomitant iliac artery disease compared to the patients from the rest of Europe and this, rather than inappropriate patient selection, probably accounts for the smaller mean diameter of the aneurysms treated. The report comments: “The treatment of patients after successful EVAR in Belgium and elsewhere needs to focus on a more aggressively management of co-morbid factors. Although these co-morbidities are not related to the aneurysm or the technique of repair, their treatment will contribute significantly to the survival of patients in the long-term.” The HTA report also recommended that EVAR should only be carried out by physicians who perform a significant number of cases, based on the argument that low-volume centres have more complications than high-volume centers (excluding all low-volume centres will drop the costs by approximately one third). In fact analysis of the EUROSTAR data shows no evidence of worse outcomes from low volume centres in Belgium compared to the high volume centres. The EUROSTAR report also notes that 30-day mortality rate for the Belgian cohort of patients in the EUROSTAR study was slightly higher than that in the EVAR-1 Trial. This is because some higher risk patients were included whereas there were no high-risk patients in the EVAR Trial: “However, in all other respects the outcomes in the Belgian patients were either equal or superior to those in the EVAR-1 Trial. Therefore, it can be anticipated that the lessons from the EVAR-1 Trial apply equally in Belgium as they do in the UK. On the basis of clinical results there is no doubt that EVAR is superior to open repair for at least four years in appropriately selected patients.” The issue of cost-effectiveness also appears prominently in the Belgian report, which rejects EVAR for routine clinical practice on cost-effectiveness grounds. However, the EUROSTAR reports claims that “there are currently no reliable data upon which to base calculations of cost-effectiveness.” Whilst the report acknowledges that the EVAR-1 Trial showed that EVAR was slightly more expensive than open repair (the difference being £3,311), the main determinant of the increased cost was the higher rate of secondary interventions associated with EVAR with a cumulative rate of 20% after four years. However, in the Belgian cohort, the cumulative rate of secondary intervention was 5% at one year, 7% at two years, 9% at three years and 15% at four years – significantly lower than those recorded in the EVAR-1 Trial. Additionally, the Belgian patients had a shorter mean in-patient stay than the EVAR-1 Trial patients and this will similarly impact upon costs in favour of EVAR. According to the Belgian HTA report the second most important determinant for the cost difference between EVAR and open repair was CT surveillance. Again, the EUROSTAR authors note that CT scans are being replaced by much less expensive Ultra Sound Scans and Plain X-rays and it is highly likely that these trends will apply in Belgium. Regarding the cost-effectiveness issue the report concludes: “Currently there are no valid data available upon which to base a decision regarding the relative cost-efficiency of EVAR in comparison to open repair. In the meantime unequivocal Level-1 evidence of clinical benefit associated with the EVAR in good-risk patients together with evidence from EUROSTAR registry of the excellent outcomes achieved within the Belgian Health Care system justifies the provision of continued access to EVAR for appropriately selected patients.”