Chimney EVAR demands strategy for dealing with compromised seal

David Minion

By David Minion



In recent years, few techniques have generated greater interest and controversy in the field of endovascular aneurysm repair (EVAR) than parallel endografts. The technique refers to the placement of two or more endografts alongside each other to preserve flow to covered branch vessels. It is commonly referred to as the “snorkel” or “chimney” procedure (ch-EVAR).

Although still considered “experimental” by many, there is now fairly extensive published data on outcomes using variations of this configuration. In terms of reno-visceral chimneys, series have been published from such centres of excellence as Orebro University, Malmö University, University of California San Francisco, University of Florida, Henri Mondor University, Münster and Zurich Universities, Stanford University, University Medical Center Utrecht, and Manchester Royal Infirmary. Combined, these series represent 378 reno-visceral chimneys in 185 patients, including 13 with ruptured aneurysms, with median follow-up for the most part in the range of 12 months.

Using this combined cohort, it appears that the 30-day mortality for chimney EVAR is approximately 2.3% in elective cases and 15.4% for patients with rupture. By comparison, the American College of Surgeons National Surgical Quality Improvement Program reported a 30-day mortality of 5.1% for elective repair of aortic aneurysms involving renal or visceral vessels and 25% in cases of rupture for the year of 2011, which represents the most recent available data.  Further, the 30-day mortality for fenestrated endovascular aneurysm repair in Globalstar was 4.1%, with all cases being elective.

In terms of branch vessel preservation, the overall incidence of target vessel loss was 1.6% in the combined series for chimney EVAR, matching that of fenestrated EVAR as reported in Globalstar. The rate was 2.5% in the FDA trial for the Cook Zenith Fenestrated device.

However, the combined rate of persistent type Ia endoleak for chimney EVAR was a concerning 7.6%, with sac growth identified in approximately 11% of the patients, and postoperative rupture reported in three cases (1.6%). By comparison, Globalstar reported immediate type Ia endoleaks in 4.6% of cases (plus an additional 1.7% type III rate), but has yet to report any long-term follow-up. In the Cook FDA trial, there were no type Ia endoleaks, cases of sac growth, or ruptures. In a cursory literature review, I could find only one case of rupture after fenestrated EVAR in 10 reported series with a combined 931 patients.

These data highlight the Achilles’ heel of chimney EVAR. That is, the parallel configuration can interfere with the conformability of the endografts to the aortic wall, leading to gutters and potential continued pressurisation of the sac. Therefore, the seal achieved with parallel endografts cannot be equated to the “true” seal attained with standard EVAR. Rather, it should be conceptualised as “gutter” seal.

However, before we condemn parallel endografts, 1.6% is a low rupture rate and still leaves chimney EVAR with an overall aneurysm-related mortality less than that of open repair in a cohort considered too high risk to undergo that procedure. In addition, most of the chimney EVAR patients were not anatomically suitable for fenestrated EVAR, which may invalidate direct comparisons between the two techniques. As case in point, the average infrarenal neck length approached 10mm in the Cook FDA trial.

Further, of the three reported postoperative chimney EVAR ruptures, one was likely from a type Ib.  Another appears to have occurred in a patient with approximately 12mm of “gutter” seal in a 34mm diameter suprarenal aorta and no additional “true” seal in the infrarenal aorta. The third appears to have occurred in a patient with approximately 10mm of “gutter” seal in a 35mm diameter suprarenal aorta with an additional 5mm of “true” seal in a 29mm diameter infrarenal aorta. Even if the renal arteries had been covered in these cases, it is not clear that enough “true” seal would have been achieved for a durable result. In other words, although chimneys can be life-saving in the appropriate situation, they cannot defy the basic principles of seal. 

In EVAR, seal is actually a complex process determined by multiple factors, including the “health” of the sealing vessel, the sealing ability of the endograft, the outflow of potential endoleaks, and the length of the seal zone. How much seal is necessary for standard EVAR? Despite the FDA approving some grafts for use in shorter necks based on their 12 month performance, almost all long-term data has shown that 15mm of seal is needed for durability. For example, data from Eurostar on over 4,000 EVARs using grafts with transrenal fixation showed that, despite early success, the late endoleak rate at 48 months was 9–11% when the length of seal was 15mm or less. Further, data on over 10,000 EVAR cases in the M2S database revealed a late sac growth incidence of 43% in patients with short necks. In more recent years, the rate was an alarming 62%, a grim indication that we have become much too cavalier with respect to acceptable seal lengths.

Parallel endografts represent a minimally invasive option that can increase seal length while preserving branch vessels—all of which are inherently good and justify their use in appropriate circumstances. There are now fairly robust published data showing that chimney EVAR is comparable to alternative strategies, even though the technique has been utilised for the most part in severely disadvantaged anatomy.

However, parallel endografts are by no means a fail-proof solution for every patient that is not a candidate for standard EVAR. Despite some enthusiasts recommending this approach with only 15mm of “gutter” seal or less, the evolving data show that this threshold can lead to compromised durability. Many centres are now using techniques such as “sandwich” grafts to achieve a minimum of 5cm of “gutter” seal in ultra-challenging anatomy. Other techniques are being developed to eliminate the gutters altogether. At any rate, parallel endografts should only be performed in centres experienced in treating juxtarenal anatomy, with a well thought out strategy for dealing with the compromised seal inherent to the configuration.


David Minion is professor of Surgery, University of Kentucky Medical Center, Lexington, USA