Abstract
In another analysis of the Heli-FX EndoAnchor System (Medtronic Vascular, Santa Rosa, CA, USA) from the stalwarts in the Netherlands, Goudeketting et al 1 refreshingly move away from the derivative analyses of the ANCHOR registry.2,3 While midterm results have been reported in similar scenarios, typically interrogating hostile neck anatomy 4 in abdominal aortic aneurysms (AAAs), this time the authors take it slightly further to include the chimney technique (8 in this series of 51 patients), given the bench test reassurances that gutters are reduced in chimney endovascular aneurysm repair (EVAR). 5
There is plenty of rhetoric on nonadherence to instructions for use (IFUs) and adverse events, yet if one examines each and every endograft IFU, they are in fact fairly nonspecific when it comes to the neck characteristics (Table 1), referring only to “adequate morphology.” Nevertheless, hostile neck anatomy has become more relevant to the loss of seal and fixation as the EVAR boundaries are being pushed further and further. The current IFU for the Endurant device (Medtronic Vascular) is now licensed for use in AAAs with a 5-mm neck using supplementary EndoAnchor fixation.
Comparison of Instructions for Use (IFU) for Aortic Endografts, Including a Newer-Generation Device (Incraft), to Demonstrate the Lack of Detail When Referring to Neck Characteristics in Particular.
Endurant II/IIs Stent-Graft System (Medtronic Vascular, Santa Rosa, CA, USA).
Excluder Endoprosthesis (W.L. Gore & Associates, Flagstaff, AZ, USA).
Zenith Alpha Abdominal Endovascular Graft (Cook Medical, Bloomington, IN, USA).
Incraft AAA Stent-Graft System (Cordis, a Cardinal Health company, Baar, Switzerland),
It is understood that more EndoAnchors are needed the larger the neck diameter. Late neck dilatation, which occurs with all devices that use radial force to achieve a seal, is recognized as a risk factor for device migration, 6 and there is now a clear suggestion that endostapling may arrest this process. 7 Interestingly, factors such as smoking and hypertension have been implicated as well. 8 Suprarenal fixation, or any primary fixation method in any endograft, is simply not foolproof, with clear reports of migration 9 (typically in the 2 to 3 years postimplantation and beyond 8 ) or even device disconnection.
If we scrutinize the methodology in the Dutch report, 1 then a retrospective study of 51 patients may seem small in the general scheme of things. However, the truth is that EndoAnchor use has had slow uptake. New studies beyond the now somewhat aging registries will come from small groups of enthusiasts such as the authors; in that light, 51 patients is a respectable number.
Moving from broad generalities to focus on the specifics of the study, we are informed of the nature of the devices used and the appropriate numbers of EndoAnchors deployed for each typical case, with only the odd exception. There are 2 groups according to usage, a primary (prophylactic, n=31) cohort and a secondary (rescue/therapeutic, n=20) group, as one sees in usual practice with EndoAnchors. Overall, 39% of this cohort ended up with secondary EndoAnchor deployments (with 94% overall having hostile neck criteria), which begs the question: Why did these patients not have primary EndoAnchor deployments to supplement fixation? 10
Much is made of the issue of reinterventions after EVAR, but few actually pay heed to the old proverb “A stitch in time saves nine,” which in this case translates to adopting a primary endostapling approach for all patients with hostile neck anatomy undergoing EVAR. The authors reflect on this somewhat as they look at the superior results with primary endostapling. 11
With regards to the 31-patient primary group, 12 (39%) had necks >29 mm. This is an appropriate index in this context, given that the threshold for moving from 4 to 6 EndoAnchor deployments occurs at >29 mm, still the remaining patients had around 6 EndoAnchors implanted. This is really typical of modern practice as it provides more symmetrical circumferential deployments all around the neck. It may therefore well be time to forget using 4 EndoAnchors for the “smaller” necks (some of which will still dilate eventually) and go for a default of 6, which is typical in my practice now. The neck length in this group was an average 11 mm, thus out of the IFU for all the devices used without EndoAnchors, except, in hindsight, the Endurant.
I tend to deploy at the 9 and 3 o’clock positions and then opt for 4 more, either in 2/4/8/10 or 1:30/4:30/7:30/10:30 clock face positions. Simply deploying 2 at the 9/3 clock face and then 2 more in a <29-mm neck is simply not symmetrical enough; otherwise one has to skip the 9/3 positions (which is immediately available in a standard anteroposterior projection) and opt for the other 4 positions as highlighted above. The 12/6 o’clock deployments seem necessary only when there is angulation in a sagittal plane (Figure 1). We do not have enough data regarding small necks as to whether “more is better,” but it certainly seems no worse. Only long-term data paralleling the results from other studies 7 can answer that question.
(A) A hyperangulated conical neck (sagittal view), (B) endograft deployment with (C) additional Heli-FX EndoAnchor deployments at 12 and 6 o’clock to counter the sagittal migration forces.
The secondary group typically had more EndoAnchors deployed; this is also reflected in the 18-patient (90%) subgroup here that had an AAA neck >29 mm. The neck length in this group was an average 7.5 mm (this may be primary or reflect loss of neck length). This again raises the question as to why these necks did not have primary endostapling!
The authors take a novel geometrical stance with deployments. I am of course delighted to see neck geometry described as “conical” or “tapered” without mention of the “reverse taper” that has surreptitiously crept into the literature, a bit like saying the “road ahead has reverse narrowed,” which underpins the redundancy of such terminologies. The authors took a quadrant-based approach to the neck, which is fine in itself, except for the obvious quandary that the 9 and 3 o’clock deployments then straddle 2 quadrants each, which muddies the analysis. However, they also used the circumferential approach, which meant that 9 patients had 180° EndoAnchor coverage (adjacent quadrants), but in the 8 who had 2 opposing quadrants covered, the circumferential descriptor becomes difficult, similar to the 9/3 situation with quadrants above. The value of looking at the deployments in this fashion may be to develop a template for future articles wherein both parameters will likely be needed to describe the amount of circumferential fixation obtained. Simply put, a 4-EndoAnchor approach at 1:30/4:30/7:30/10:30 clock face positions achieves true quadrantic symmetry with a 90° distributional gap.
The authors realistically inform us that things can go wrong as well, including EndoAnchor fracture and loss requiring snaring and retrieval. These events have certainly occurred in my own practice, and I have used an additional “sandwich” technique with an aortic cuff to exclude the culprit EndoAnchor from the circulation. Device explantation with EndoAnchors in situ also poses a new challenge, though a pragmatic “neo-neck” approach can be used as the authors have indeed done.
Neck dilatation was acceptable over almost 24 months (≅0.1 mm/month), but the bottom line is that there was no further migration at all. Looking to complications that have a direct bearing on EndoAnchor usage, 6 patients still had a type Ia endoleak (we are not informed if this is in the primary or secondary group though). This may raise questions about selection of device [alternative endograft with different sealing method, eg, Ovation (Endologix, Irvine, CA, USA)] and/or procedure (should they have been selected for fenestrated/branched EVAR instead?).
Besides highlighting the textbook limitations regarding such retrospective studies, the authors do not inform us as to which devices were more likely to need reinterventions. There is a hodgepodge of devices, for example, cuffs in AFX bodies, which is of course an acceptable out-of-the-box approach for secondary interventions. This would have been a useful opportunity to glean some insights into whether the new IFU for the Endurant with EndoAnchors actually has some substance. We are also no wiser as to how much the secondary interventions cost overall, notwithstanding the base cost of the EndoAnchors themselves, which typically run £4200 (US$5385) per case.
What does this article tell us then? First, that there is a “failure rate” even with the use of EndoAnchors given the declining freedom from type Ia endoleak at 2 years (87.3%). Second, all EndoAnchor deployment should be ideally planned in advance to avoid deployment through calcified areas, with the attendant failure risks. Third, EndoAnchor deployment seems to result in more “benign” type I endo-leaks, which is also our experience, notwithstanding that not each and every such endoleak actually needs treating. 12 Fourth and last, the authors failed to give us a glimpse into how well the Endurant device behaves in conjunction with EndoAnchors.
In summary, there are clear missed opportunities in terms of what the article tells us, but overall, at least in the midterm, EndoAnchors work. I would state they would need to be used both more often and also more efficiently/effectively so that long-term data can be accumulated from routine clinical practice. For that the Heli-FX EndoAnchor System needs to become a truly standard component of the endovascular interventionist’s repertoire and not just as an afterthought.
