Draft NICE aortic guidelines cause endovascular controversy

Roger Greenhalgh speaking at the 2018 Vascular Societies’ Annual Scientific Meeting

For the vascular community in the UK, discourse on guidelines in 2018 increasingly came to focus on the proposed guidelines document from the National Institute for Health and Care Excellence (NICE), which in a draft published in May outlined the contentious recommendation for open repair of aortic aneurysms, rather than endovascular aneurysm repair (EVAR). The final guidelines, which were initially scheduled for publication in November 2018, remain unpublished at the time of writing, with the publication date “to be concluded” according to an online statement by NICE.

The document states that patients with unruptured infrarenal abdominal aortic aneurysms should not be offered EVAR if open surgical repair is suitable—nor should EVAR or complex EVAR be offered if open repair is unsuitable due to the patient’s anaesthetic and medical condition. The guidelines committee bases its recommendations largely on the EVAR 1 and EVAR 2 trials, which on 15-year follow-up found a lower mortality for open repair than for EVAR at 15 years.

Roger Greenhalgh of Imperial College in London, UK, was the principal investigator of the EVAR 1 and EVAR 2 trials. Speaking at the Vascular Societies’ Annual Scientific Meeting (VSASM; 27–29 November, Glasgow, UK), Greenhalgh made a case against the draft guidelines, arguing the data from the trials had been misinterpreted and that their application lacked appropriate context. He outlined limitations to the use of mortality outcomes, and presented new data to illustrate reasons for these mortality rates, which could be caused by suboptimal follow-up protocols following the procedure, Greenhalgh argued.

Greenhalgh’s talk, titled “Optimising post-operative surveillance and intervention following endovascular repair of abdominal aortic aneurysms”, was a paper put forward to NICE by Greenhalgh and colleagues in the EVAR trials post-operative surveillance group—an international multi-disciplinary collaboration between triallists, health economists and statisticians—and published in the British Journal of Surgery in 2018.

“Most of you are aware of the EVAR trials, and the problem that presented itself in the 15-year results,” Greenhalgh said at VSASM, referring to the fact that at 15 years, the EVAR trial data showed higher rates of aneurysm-related and total mortality during the last eight years of follow-up for patients who had EVAR than those who had open repair. “It was found that the aneurysm-related mortality was mainly caused by secondary rupture, and the all-cause mortality significant difference—to everyone’s surprise—was largely due to aneurysm-related mortality and some cancer deaths.”

Having established that the separation of mortality rates for open repair and EVAR in the EVAR trials were caused by aneurysm-related deaths due to secondary rupture in the EVAR arm, Greenhalgh moved to investigate how this problem might be mitigated. Referring to a “brilliant analysis” showing the trajectories of sac diameters over time by type of event in the EVAR trials, Greenhalgh explained: “If you measure the diameter of the sac around the endovascular device over time, you will see … if you perform a rupture-preventing intervention, you convert the shape of this trajectory towards a safer outcome which does not lead to rupture.” Corroborating this data with follow-up statistics provided by Maarit Venarmo and her vascular team in Helsinki, Finland, which showed similar trajectories on analysis.

Greenhalgh further elaborated that rupture following early EVAR procedures may have been more prevalent, as reintervention techniques and technology continue to improve over time. However, rupture-preventing interventions are also largely dependant on maintained monitoring of aneurysm sac growth—an area upon which Greenhalgh could further shed some light. Regarding the EVAR trials, he reported: “The issue was that in the first six months there was reasonable follow-up, but after a year follow-up began to drop consistently year by year, and by the nine-year follow-up only 10% of the patients were being followed.”

This is a finding that again was corroborated by the Helsinki cohort, leading the the group to establish this was not merely a “UK-only problem”—as Venarmo observed the 90% follow-up rate in the first six months dropping to 40% at five years. Greenhalgh commented that “while that is not perfect, it is better than the UK”, but maintained that follow-up rate drop-off over time remains “a widespread problem”. The clinical significance of this issue, Greenhalgh said, is that “current EVAR follow-up protocols are indeed suboptimal, and the EVAR procedure requires lifelong annual surveillance.”

“From repeated annual sac diameter measurements,” Greenhalgh continued, “we can predict that 40% of patients are classified as low risk, and no hospital follow-up is required. However, 85% of patients need rupture-preventing reintervention within two years.” Having established the threshold for criteria of vascular centre referral, Greenhalgh pointed out that the improved structure for referral and follow-up protocol, which was not yet applied at the time of the EVAR trials, directly impact the aneurysm-related mortality rate while also improving cost-efficiency post-EVAR.

If the EVAR trials were repeated today

Also speaking at the VSASM, Andrew Bradbury, who chaired the scientific guidelines committee for NICE, explained the process of the highly structured and methodical process of the institute’s guidelines production. Bradbury emphasised the reasoning behind the draft guidelines, including the prioritisation of randomised controlled trial data and the use of health economics in the UK National Health Service (NHS). At the same time, Bradbury acknowledged that a considerable amount of input from the vascular community had been considered in the months following the draft publication, and suggested there may be some changes to the final version.

Ultimately however, Bradbury maintained at that the EVAR trials, “if they were repeated today, they would likely show more or less the same outcomes,” and added that such a trial might be ethically questionable, considering the mortality outcomes of EVAR 1. The NICE draft guidelines nevertheless state that complex EVAR should not be offered to patients with an unruptured AAA if open repair is a suitable option, “except as part of a randomised controlled trial comparing complex EVAR with open surgical repair.”

To this point, Greenhalgh presented data from a discrete event simulation—a predictive model based on the real data from patients in the EVAR trials, “constructed to estimate the incidence of complications, ruptures and deaths that would have occurred under alternative surveillance protocols”. The data, which is currently under review for publication, highlights the cost per quality-adjusted life year which shows favourable outcomes for EVAR when implementing an updated follow-up protocol post-EVAR, compared to the recommendations favouring open repair as outlined in the NICE draft guidelines.

The use of this simulation, Greenhalgh explained, serves to update the trial data according to current contexts and provides valuable insight quickly, as “patients need benefit now—not after another 15 years at high costs of a repeat trial.”

The reason for delay in publication of the final guidelines has not been disclosed, but may have been caused by new evidence submitted prior to publication including the analysis presented by Greenhalgh, along with the public outcry at the draft which was a central talking point not only at the VSASM but also at several international meetings including VETHsymposium (13–17 November, New York, USA), as the NICE scientific guidelines committee currently continues to review the considerable input given by the vascular and endovascular community.


Please enter your comment!
Please enter your name here