Durable EVAR depends on a persistent seal between the proximal end of the stent graft and a non-dilated segment of infrarenal aorta (the neck). This study, like others before it4-11
demonstrated a high incidence of neck dilatation following EVAR with a self-expanding stent graft. We observed some degree of neck dilatation at the 4-year follow-up in all
cases. This rate is higher than the 20%8
to 30% seen in previous studies, probably because the follow-up is longer.
Since stent graft function depends on contact between the outside of the stent graft and the inside of the neck, one might expect that the universal neck dilatation would ultimately result in universal failure. Yet large studies have shown that secondary endoleak remains a relatively rare complication for many years after EVAR, if only with certain devices.13, 14
The current study helps to explain this apparent disparity between the incidence of neck dilatation and the incidence of long-term failure. Following repair with the Zenith stent graft, the diameter of the neck increased rapidly at early time points, and slowly at later time points. At no point did the diameter of the neck exceed the actual
diameter of the stent graft. The diameter of the neck exceeded the nominal diameter of the stent graft but only because the nominal diameter underestimated the actual
We speculate that when the stent graft, or part of the stent graft, reaches maximum diameter, its inelastic woven polyester covering loses the capacity to expand and contract with the phases of the cardiac cycle, and ceases to transmit force to the aortic wall. This protective effect is likely to be device-specific and dependent upon active barb-mediated attachment. While the loss of a stent graft’s potential for outward expansion probably enhances the stability of the neck, the corresponding reduction in friction would undermine the stability of any stent graft lacking a suprarenal stent for attachment in the pararenal aorta, or barbs for attachment in the neck.8, 9
If, as we believe, the fully expanded stent graft imposes a limit on neck diameter, it is not surprising that the potential for dilatation of the neck depends on the potential for dilatation of the stent graft within the neck; hence, the correlation between oversizing and neck dilatation. A greatly oversized stent graft can expand a lot before reaching its limit, whereas a barely oversized stent graft can expand very little.
The findings of the current study are limited by the small size of the study cohort, which results in part from the combined effects of patient demographics and the requirement for long follow-up. In 1998, when the study started, the Zenith stent graft was not approved for general use in the United States. At the time, our IDE protocol limited EVAR to high-risk patients with short life-expectancy. Many of these patients did not live long enough to qualify for inclusion in this study. It is important to note, however, that these patients did not die from aneurysm-related causes. As of 2007, when the entire underlying cohort was last examined in detail,13
there were no instances of secondary type I endoleak and no conversions to open surgery for migration. Had that not been the case, the current study would suffer from selection bias through the elimination of patients who had serious complications of neck dilatation. We also recognize that the current data apply only to patients treated with the Zenith stent graft. Other patient groups and other stent grafts may have different rates of neck dilatation, migration, and endoleak.