IMPROVE trial morphology analysis: The answer lies in the neck


An analysis of aneurysm morphology in the IMPROVE trial presented at the British Society of Endovascular Therapy (BSET) Annual Meeting on 26 June 2014 showed that for long aneurysm necks, mortality after endovascular aneurysm repair (EVAR) and open repair was very similar. Surprisingly, the investigators said, there was “a very strong inverse association between neck length and mortality after open repair”.

The analysis of the IMPROVE (Immediate management of the patient with rupture: open versus endovascular repair) trial on aortic neck length and survival after repair of ruptured abdominal aortic aneurysm was presented by Janet Powell and Rob Hinchliffe on behalf of the trial investigators.

Aneurysm morphology indicates whether a patient with ruptured abdominal aortic aneurysm is eligible for EVAR and may influence the outcome of both EVAR and open surgical repair. In the emergency setting, the morphological criteria for EVAR (IFU, instructions for use) may be relaxed, but by how much?,” they wrote in an abstract.

Patients with a proven diagnosis of ruptured abdominal aortic aneurysm, who underwent repair and had their admission CT scan submitted to the core laboratory, were included in this analysis of 30-day mortality and re-interventions, according to a pre-specified plan, focusing on liberal IFU (neck diameter ≤32mm, neck length ≥10mm and proximal neck angle <60°) and six morphological variables: maximum aortic diameter, neck diameter(s), neck length, proximal neck angle, neck conicality and maximum common iliac diameter.

Four hundred and fifty eight patients (364 men), mean age 76 years were included, with EVAR commenced in 177 and open repair in 281 cases, with 155 deaths and 88 re-interventions. The mean maximum aortic diameter was 8.6cm. There were no important correlations between the six morphological variables.

Patients within liberal IFU (58%) had lower mortality, although this narrowly failed to achieve statistical significance, p=0.054. Only aortic neck length (mean[SD] 23.2[16]mm) was associated significantly (inversely) with 30-day mortality both for open repair (p<0.001) and overall (p-0.007). A short aneurysm neck was the commonest reason for a patient being unsuitable for EVAR. For re-interventions, only iliac diameter showed a borderline association but this will be reassessed after 12 months of follow-up.

In conclusion, the investigators said the results showed that for long aneurysm necks mortality after EVAR and open repair is very similar “and explain the results of the Dutch and French trials, which only recruited patients suitable for endovascular repair”. They added that there was a very strong inverse association between neck length and mortality after open repair. There were very few patients undergoing EVAR with aneurysm neck lengths between 10 and 14mm, so that it is difficult to assess whether EVAR might offer a clear advantage for these neck lengths.

“We hope that our agreed collaboration with the Dutch and French trials might enable us to provide some further information by the end of the year. Observational studies, which ‘cherry pick’ long necked aneurysms for EVAR leaving all the short necked aneurysms for open repair, always show that mortality is lower after EVAR. Such observational studies are comparing apples and oranges,” the investigators said.

The pre-specified analysis plan and more information about the trial can be found at