Is Evar Feasible in Challenging Aortic Neck Anatomies? A Technical Review and Ethical Discussion

Background: Endovascular aneurysm repair (EVAR) has become an accepted alternative to open repair (OR) for the treatment of abdominal aortic aneurysm (AAA) despite “hostile” anatomies that may reduce its effectiveness. Guidelines suggest refraining from EVAR in such circumstances, but in clinical practice, up to 44% of EVAR procedures are performed using stent grafts outside their instruction for use (IFU), with acceptable outcomes. Starting from this “inconsistency” between clinical practice and guidelines, the aim of this contribution is to report the technical results of the use of EVAR in challenging anatomies as well as the ethical aspects to identify the criteria by which the “best interest” of the patient can be set. Materials and Methods: A literature review on currently available evidence on standard EVAR using commercially available endografts in patients with hostile aortic neck anatomies was conducted. Medline using the PubMed interface and The Cochrane Library databases were searched from 1 January 2000 to 6 May 2021, considering the following outcomes: technical success; need for additional procedures; conversion to OR; reintervention; migration; the presence of type I endoleaks; AAA-related mortality rate. Results: A total of 52 publications were selected by the investigators for a detailed review. All studies were either prospective or retrospective observational studies reporting the immediate, 30-day, and/or follow-up outcomes of standard EVAR procedures in patients with challenging neck anatomies. No randomized trials were identified. Fourteen different endo-grafts systems were used in the selected studies. A total of 45 studies reported a technical success rate ranging from 93 to 100%, and 42 the need for additional procedures (mean value of 9.04%). Results at 30 days: the incidence rate of type Ia endoleak was reported by 37 studies with a mean value of 2.65%; 31 studies reported a null migration rate and 32 a null conversion rate to OR; in 31 of the 35 studies that reported AAA-related mortality, the incidence was null. Mid-term follow-up: the incidence rate of type Ia endoleak was reported by 48 studies with a mean value of 6.65%; 30 studies reported a null migration rate, 33 a null conversion rate to OR, and 28 of the 45 studies reported that the AAA-related mortality incidence was null. Conclusions: Based on the present analysis, EVAR appears to be a safe and effective procedure—and therefore recommendable—even in the presence of hostile anatomies, in patients deemed unfit for OR. However, in order to identify and pursue the patient’s best interest, particular attention must be paid to the management of the patient’s informed consent process, which—in addition to being an essential ethical-legal requirement to legitimize the medical act—ensures that clinical data can be integrated with the patient’s personal preferences and background, beyond the therapeutic potential of the proposed procedures and what is generically stated in the guidelines.


Introduction
Endovascular aneurysm repair (EVAR) has become an accepted alternative to open repair (OR) for the treatment of abdominal aortic aneurysm (AAA), with >75% of AAA repairs performed endovascularly [1]. Undoubtedly, EVAR is associated with lower 30-day mortality and morbidity, faster discharge, and fewer complications than OR [2], but some concerns remain about durability, need for long-term follow-up, and reinterventions [1][2][3].
Based on preclinical engineering assessments and clinical study results, particular anatomical characteristics, specifically aortic neck diameter, length, angle, and shape, are recommended to guide patient selection for EVAR [4]. Indeed, unfavorable anatomy, documented in 40-60% of treated AAAs, seems to be the factor most related to negative outcomes [5][6][7]. Despite this, in "real-world" clinical practice, up to 44% of EVAR cases are performed using stent grafts outside their instruction for use (IFU) due to the presence of a "hostile" aortic neck anatomy, with acceptable short-and mid-term outcomes [8][9][10]. In order to overcome this problem, several different technical solutions, such as parallel grafts and fenestrated or branched devices, have been developed as an alternative to standard EVAR grafts in patients with a "challenging neck". However, all those solutions present a relevant risk of reintervention due to gutter-related endoleaks and target vessel instability, and higher cost compared to standard EVAR; lastly, custom-made devices are not suitable for urgent/emergent cases [7].
Nevertheless, because EVAR durability is related to the maintenance of a seal between the stent graft and the aortic neck, some authors suggested that challenging the neck could affect long-term outcomes, increasing the risk of type Ia endoleak, reintervention, and aneurysm-related mortality rates [11][12][13][14][15]. As a result of these long-term results, current guidelines suggest limiting or even refraining from adopting EVAR in patients with challenging aortic necks [16][17][18][19].
Moving from this inconsistency between "real world practice" and "best practice" suggested by current guidelines [20], the purpose of this contribution is to report technical and clinical results of the use of EVARs in challenging anatomies and to analyze and discuss the ethical implications of implementing these procedures, to identify criteria by which the patient's "best interest" should be defined in these specific circumstances.

Search Strategy
A literature review on currently available evidence on standard EVAR using commercially available endografts in patients with hostile aortic neck anatomies was conducted by three investigators (F.A., S.C., P.S.), and an eligibility assessment of studies for inclusion in this review was performed in a non-blinded standardized manner by the same investigators. Disagreements were resolved by discussion and consensus.
Medline using the PubMed interface and Cochrane Library databases were searched from 1 January 2000 to 6  A further search was undertaken including a manual screen of the reference lists of selected articles identified through the electronic search; only English papers were considered for the present review.

Study Selection: Study Design and Data Extraction
Studies included in the analysis met the following criteria: (a) randomized controlled studies, non-randomized studies, and observational studies; (b) published as original research from 2000 to 2021; (c) clearly reported the results of the patient treated in the presence of "hostile" neck anatomy; (d) described the results of standard EVAR procedures in the immediate postoperative period, and/or at 30 days, and/or at mid-or long-term follow-up.
Papers regarding endovascular abdominal sealing (EVAS) procedures were excluded due to the completely peculiar sealing mechanism of the graft and the unsatisfactory results leading to market withdrawal [22]. Case reports, letters to the editor, and conference abstracts papers were all excluded from the present analysis.
The methodological quality of studies was evaluated independently by the same three Investigators with the Newcastle-Ottawa scale, which was used to assess the quality of non-randomized studies [23].
Data were extracted into a standard Microsoft Excel file by three independent authors (F.A., S.C., P.S.) as follows: first author, year of publication, country, study design, data source/institution, duration of the study, age, number of patients, criteria, and outcomes.

Outcomes
Outcomes considered for the present analysis were EVAR technical success, need for unplanned adjunctive procedures, conversion to open repair reintervention, stent graft migration, type I endoleak occurrence at 30 days, and at mid/long-term follow-up were recorded, as well as AAA-related mortality at same time intervals.

Statistical Analysis
The data are reported as mean and standard deviation (SD) or as absolute frequencies and percentages (%). All analyses were calculated using SPSS version 25 (IBM Corp, Armonk, NY, USA).

Results
Cumulatively, a total of 172 publications were identified and their titles and abstracts were reviewed. Of these, 52 publications were selected by the investigators for a detailed review before the face-to-face panel meeting and selected for the current report ( Figure 1).

Study Characteristics
All studies were either prospective or retrospective observational studies reporting the immediate, 30-day, and/or follow-up outcomes of standard EVAR procedures in patients with challenging neck anatomies. No randomized trials were identified. The study population ranged from 12 to 1189 patients, and the period during which selected studies were published extended from 2003 to 2021 (Table 1).  Fourteen different endograft systems were used in selected studies (Table 1) Almost all the patients included in this review underwent a surveillance imaging protocol consisting of computed tomographic angiography (CTA). The great majority of the selected studies referred to guidelines suggested criteria to define a challenging neck anatomy [16][17][18][19].

Technical Success and Adjunctive Procedures
Forty-five studies reported the technical success rate, defined as the successful introduction and deployment of the stent grafts in the absence of surgical conversion, type I or III, or renal artery coverage. Technical success rates ranged from 89.2% to 100%; in two studies failure was due to unintentional renal artery coverage and no-to high-flow endoleak [53,55].
Forty-two studies reported the need for adjunctive procedure rate during EVAR procedures to achieve proximal seal: those procedures were either represented by needs for repeated proximal fixation site ballooning, aortic cuff or endoanchor implantation, and rescue chimney procedures. The necessity for adjunctive procedures ranged between 0 and 51%, with a mean value of 9.04%.

Thirty-Day Results
Regarding 30-day results, the type Ia endoleak incidence rate was reported by thirtyseven studies and ranged between 0 and 27.3% (mean value 2.65%); 23 studies reported an incidence <2%, and all but one an occurrence lower than 9% [10]. No endograft migration was reported, according to data available in 31 studies out of the 52 included (Table 2).
Reintervention due to type Ia endoleak occurred in 0-15.2% of cases (mean value 1.66%) with 30 out 37 studies recording an incidence <2%. Regarding conversion to open repair, data are reported by thirty-four studies: in 32 published experiences, no conversions were performed at a 30-day follow-up interval, while two studies [53,59] reported approximately a 2% rate (Table 2).
Lastly, data about AAA-related death rate were reported by thirty-five studies: in the great majority of them, the incidence was null, while 4 studies reported an AAArelated death rate ranging between 0.5 and 1.3%, in absence of aneurysm sac rupture after EVAR [24,32,37,44]. It is noteworthy that in the experience of Troisi et al. [32], mortality was observed only in patients initially treated for ruptured AAA (Table 2).

Mid-Term Follow-Up Results
All but three studies [43,60,70] reported a follow-up period longer than 30 days. Follow-up duration was evaluated by polling the data of the considered study: mean follow-up was 27.38 months (range 1-120, SD ± 23,12; Figure 2).
The type Ia endoleak incidence rate was reported by forty-eight studies and ranged between 0 and 14.8% (mean value 6.65%); half of the studies reported an incidence <2%, and seven an occurrence greater than 10% [10,25,33,40,51,58,69]. Endograft migration rate was reported with a mean value of 1.86% by 44 studies, 30 of them not reporting cases of migration. However, three studies observed an incidence >10% [25,47,69] (Table 2).    Reintervention due to type Ia endoleak occurred in 0-24.1% of cases (mean value 4.38%) with 25 studies recording an incidence <2%. Regarding conversion to open repair, data are reported by forty-one studies: in 33 published experiences, no conversions were performed at the last follow-up visits, while two studies [58,68] reported a conversion rate greater than 5% (Table 2).
Lastly, data about AAA-related death rates were reported by forty-five studies: in 28 of them, the incidence was null, while 17 studies reported an AAA-related death rate ranging between 0.6 and 11% with only Abbruzzese and coworkers reporting an incidence >10% [29] (Table 2).

Discussion
After a careful and extensive analysis of the existing scientific literature, the main result of the present contribution is that despite the above-mentioned inconsistency between "real world practice" and "best practice" suggested by current guidelines, standard EVAR in patients presenting a so-called hostile or challenging proximal aortic neck anatomy could be safely performed. Consistent with those findings, a recently published systematic review confirmed a significant reduction in perioperative mortality for EVAR compared to open repair without showing differences in AAA-related mortality at mid-term followup [71]. However, the risk of reinterventions still represents a major issue in those kinds of procedures and should be properly assessed in the decision-making process [71].
Of course, not all patients could be a candidate for this type of treatment, and open repair still could be considered the standard of care in patients fit for surgery [16][17][18][19]. However, from a technical point of view, standard EVAR could be safely and effectively performed (and consequently proposed) to those patients judged unfit for open surgery.
Nevertheless, a word of caution is needed regarding patients with the concurrence of multiple anatomical characteristics affecting the EVAR feasibility [7], and those presenting a wide aortic neck diameter (<30 mm) [10,51,69].

Ethical Considerations
The technical feasibility of standard EVAR by itself is not sufficient to exclude all ethical implications related to performing an elective procedure outside the IFU, even in fragile patients unfit for open surgery. Indeed, even if the present review shows various scientific evidence confirming that standard EVAR can be proposed even outside the IFU to patients ineligible for open surgery, identifying the most appropriate therapeutic procedure for every single patient remains a challenging issue requiring careful case-by-case evaluation for patient's best interests.
Although this type of evaluation should always be considered an essential part of good clinical practice, both close attention to ethical requirements and acquiring a patient's proper informed consent play a crucial role in approaching a so fragile subgroup of patients, due to the aforementioned "inconsistency" between clinical practice and guidelines [72]. In this scenario, correctly informing the patient, in addition to being an essential ethical-legal requirement to legitimize the medical act, allows integration in the decision-making process those results that are important for the patient, beyond the theoretical advantages of each proposed procedure and guideline statement [73].
As is known, the traditional "paternalistic" approach, in which the doctor was considered autonomously capable to decide for the patient's best interest, has given way to a more holistic approach in which the patient's autonomy and self-determination are completely integrated into the decision-making process. This need to develop more patientcentered healthcare has led to a redefinition of the concept of patient's best interest, in which physician and patient are both equally involved as decision-makers. This shared decision-making approach requires integrating the best available medical evidence with the patient's values, beliefs, and preferences, in an ongoing dialogic process, that promotes high-quality health care decisions from both an objective (physician), and a subjective (patient) perspective [74]. Moreover, preliminarily identifying patient priorities is essential to ensure that these aspects could be evaluated in clinical practice [75].
However, recognizing the value of patient preferences and their relevance to medical decision-making can be difficult, especially when they differ from classical clinical outcomes. In other words, faster discharge times, or the absence of postoperative discomfort could be considered extremely important by the patient (and therefore have a greater value in the decision-making process) than the crude mortality rate [76]. Notably, even studies specifically designed to address the patient's perspective (usually based on "quality of life" as a quantitative equivalent) are essentially based on items defined by health professionals, which may not reflect the patient's perspective. For example, these studies may not capture the patient's "concerns about symptoms", "the impact of possible outcomes/complications", as well as issues related to "self-control and decision-making" [73].
Consequently, to enable patients to make decisions that are fully consistent with their individual preferences, physicians should strive to properly inform about the different treatment options, and the risks and benefits associated with each alternative [77], not forgetting the patients' perspective [74].
All the above is necessary and, at the same time, particularly complex to obtain in fragile AAA patients unfit for open surgery and presenting with hostile aortic anatomies. Approaching such a patient, vascular surgeons are requested to discuss all treatment options (F/B-EVAR, Ch-EVAR, off-label use of EVAR, and even non-intervention), and their relative risks and benefits, personalizing information for every single patient [78,79].
Moreover, the informed consent process requires not only that the physician inform the patient, but also that the patient fully understands the information provided. Therefore, the information must be presented consistently with the understanding of each individual patient, based on her/his education level, age, and psychological and emotional status [20]. Physicians should not rush the patient's decision: the patient's informed consent process is, in fact, a "process" and not a simple "act". The physician-patient relationship must be established and strengthened through a dialogue that requires commitment and time. Physicians should encourage the patient to reflect on her/his preferences, values, and goals, ask for more information, express her/his doubts, discuss with relatives, and seek a second opinion in case of uncertainty [74].
Lastly, any deficiency in the informed consent acquiring process undermines the legitimacy of the consent itself, breaks the relationship of trust between patient and physician, and potentially leads to litigations [80]. On the contrary, the more detailed the information is, the more actively the patients are involved, and the more likely they are satisfied with their decisions and their expectations are met [74,81].
Despite all these considerations, patients' information needs are not always satisfied in everyday clinical practice: patients, especially those presenting AAA, complain about a lack of information on the treatment option and relative risks and benefits [82].
With the aim of overcoming this vulnus in the patient-physician relationship, several interventions were implemented to improve the quality of informed consent and to foster patient understanding of treatment options and outcomes [83][84][85][86].
Physicians should consider what patients really want to know, and what information is truly useful for them to make a decision. It is not always necessary to report everything, especially when it comes to complex clinical or statistical data. However, the physician must recognize that some patients want to be thoroughly informed about treatment-different options, and risks and benefits associated with each available option-some others prefer to receive less information, and others choose to receive no information, exercising their "right not to be informed" [72,87].

Study Limitations
First of all, the present study is a narrative review and not a systematic review, consequently, the statistical power of here presented data should be carefully evaluated. Moreover, not all included studies reported results on standard EVAR performed in standard anatomies, therefore, a proper comparison was not made between patients presenting with standard and hostile anatomies treated by the same operators, in the same centuries, during the same period.

Conclusions
In conclusion, it is not possible to establish a priori what is in the best interest of the patient; it is not possible when there is reliable scientific evidence, and it is even less so in case of inconsistency between "real world practice" and "best practice" suggested by current guidelines such as in the reported clinical scenario. In those complex cases, physician experience, available data from reviews, guidelines' recommendations, and patients' preferences should all be considered and carefully evaluated to reach a joint decision and to choose the right tailored approach for every single case.