On Sunday evening, CX 34 delegates heard the results of the world’s first multicentre, randomised, controlled trial that compared EVAR and open repair for ruptured abdominal aortic aneurysms. The Dutch AJAX trial results demonstrated no difference between EVAR and open repair in emergent cases. In the same session, the audience also heard an outline of the SWIFT study on the effect of transport on ruptured aneurysm treatment outcomes. While AJAX results shed some light on treatment for ruptured aneurysms, many questions still remain unanswered
Ron Balm, Amsterdam, The Netherlands, presented the AJAX results, and Regula von Allmen, London, UK, presented on the SWIFT study (Swiss ruptured aneurysm favourable transport), which investigates whether the time from diagnosis to intervention relate to operative death of ruptured abdominal aortic aneurysm.
“Some excellent results have been reported in Switzerland which is a circumscribed country where the centres performing vascular surgery are clearly identified,” she said. Von Allmen told delegates that there are great variations between centres in the country. “Reports from Zurich show that 50% of the ruptured aneurysm patients are managed by endovascular means and 50% by open repair. Zurich reports excellent results for EVAR with a 30-day mortality rate of 13.5%; open repair has a mortality rate of 32.4%. But if patients look anatomically unsuitable for EVAR, then this may not be a fair comparison between open and endovascular repair.
“However, when we look at data from Bern, there are only 4% who are treated by endovascular means and the vast majority, 96%, is treated with open repair, and the overall 30-day mortality is 15.3%. “From this we can see that there are pockets of excellence in the treatment of ruptured abdominal aortic aneurysm in Switzerland. As a consequence, it is claimed by one centre that it is unethical to carry out treatment other than EVAR for ruptured aneurysms, but there are obvious counter claims for open repair based on the data from Bern. Disparities in views often point to uncertainty and there is no proof that surgical approach is the key,” she said.
Close on the heels of this presentation came the results of AJAX, which showed that indeed there was no difference between EVAR and open repair in the treatment of ruptured abdominal aortic aneurysms. The primary endpoint of AJAX was combined death and severe complications at 30 days. Our hypothesis was that EVAR would do better than open repair with endpoint rate of 0.40 and 0.65 for open repair, β=0.20, α=0.05 in a sample size of 112 patients. Secondary endpoints were length of hospital and intensive care unit stay, intubation/ventilation and use of blood products.
“The trial area covered 1.2 million inhabitants and three trial centres and seven regional hospitals contributed data. All patients with ruptured aneurysms in the trial area were identified and followed,” said Balm. He told delegates that the preferred EVAR technique was use of an aorto-uni-iliac graft with contralateral occluder and femorofemoral crossover bypass. “Between April 2004 and February 2011, 520 patients with ruptured abdominal aortic aneurysms were enrolled in the trial and 90% (466) were enrolled in a trial centre.”
Three hundred and ninety five patients were evaluated with CTA, and 240 were found to have unfavourable anatomy for EVAR. Thirty nine patients were excluded, of these 16 were unfit for open repair, 11 were excluded for logistical reasons, seven for haemodynamic instability following CT, and five patients refused surgery. Balm said, “116 patients were randomised: 57 to EVAR and 59 to open repair. The results showed that in terms of the primary endpoint EVAR had a combined and severe complications rate of 42% (24/57) at 30 days. In the open repair group, this rate was 47% (28/59), (ARR 5.4% [95% CI -13 to +23]). These results showed, said, Balm, that the hypothesis that EVAR is better than open repair, could not be confirmed.
With regard to the secondary endpoints, ICU stay with open repair was 48 hours while it was 28 hours with EVAR (p=0.14); hospital stay was nine days with EVAR and 13 days with open repair (p=0.57); 39 patients had to use a mechanical ventilator with EVAR while 52 did so with open repair (p=0.002). Balm said, “Blood loss with EVAR was 500 ml while it was 3500 ml with open repair (p<0.001). Forty five EVAR patients needed blood during the surgery while 56 patients did so after open repair (p=0.01).
“EVAR performed a little better on the secondary endpoints,” Balm said. He told CX delegates that death with EVAR was 21% (12/57) while with open repair it was 25% (15/59). “Was there a selection of haemodynamically stable patients, asked Balm, noting that 17% of the entire cohort was haemodynamically unstable (78/466). In the randomised controlled portion, 20% was haemodynamically unstable (23/116).” Additionally, he also posed the question of whether the triallists had selecting simple anatomy, by drawing attention to the death rate following open surgery in patients with unfavourable anatomy, which was 26% in the cohort (58/222). The 30-day death rate of all consecutive patients who underwent surgery was 30% (138/454) (95% CI 26–35%).
Balm cited data from Visser P et al that was published in EJVES in 2005. The population-based analysis showed a 41% in-hospital operative mortality in The Netherlands (95% CI 40–42%) Importantly, said Balm, ‘all comers’ were consecutively enrolled in the analysis that this was a major strength of this study.
In conclusion, Balm said, “With AJAX trial results showing that EVAR vs open repair was ARR 5.4% (95% CI -11 to +23), we saw that open repair performed much better than expected with low death rates in the randomised controlled trial, but also low death rates in the entire cohort.” He said, “This could be attributable to the introduction of the trial with optimised logistics and patient care such as the pre-operative CTA and centralisation.”