Percutaneous Thoracic Endovascular Aortic Repair is Not Contraindicated in Obese Patients

Abstract

Objectives

There is limited data describing the preclose technique with the Perclose Proglide device in thoracic endovascular aortic repair(P-TEVAR), particularly in obese patients, where use of this technique is thought to be relatively contraindicated. The purpose of this analysis is to describe our experience with P-TEVAR and compare outcomes in patients with or without obesity.

Methods

All TEVAR procedures at a single institution from 2005–2011 were reviewed and P-TEVAR patients were stratified by body mass index(BMI: obesity = ≥ 30 kg/m²). Preoperative CT scans were analyzed for access vessel depth, calcification and morphology. Technical success was defined as the ability to achieve hemostasis and maintain limb perfusion without the need for common femoral artery exposure and/or obligate surgical repair of the vessel within a 30-day postoperative period. Generalized estimating equations and stepwise logistic regression were used to develop prediction models of preclose failure.

Results

536 patients were identified, in whom 355(66%) P-TEVAR procedures were completed [366 arteries; N= 40 (11%) bilateral]. Compared to non-obese patients(N = 264), obese patients(N = 91) were typically younger(59±16 vs. 66±16 years; P = .0004) and more likely to have renal insufficiency(28% vs. 17%; P = .05) and/or diabetes mellitus(19% vs. 9%; P = .02). Number of Perclose deployments were similar between groups(P = NS). Mean sheath size(French:25.4 vs. 25.0; P = .04), access vessel inner diameters [8.5±1.9mm vs. 7.9±2.0; P = .02)] and vessel depth(50±20mm vs. 30±13; P < .0001) were greater in obese patients. Adjunctive iliac stents were used in 7% of cases [obese, N = 10(11%) vs. non-obese 16(6%); P = .2]. Overall technical success was 92% [92% non-obese vs. 93% obese patients(P=.7)]. Three patients required subsequent operations for access complications; two obese patients(2%) and one non-obese patient (0.4%)(P = .3). Independent predictors of failure were adjunctive iliac stent(OR 9.5; 95%CI 3.3–27.8, P < .0001), >2 Perclose devices(OR 7.0; 2.3–21, P = .0005), and smaller access vessel/sheath size ratio(OR multiplies by 1.1 for each .01 decrease in ratio; 1.02–1.2, P = .007) (AUC = .75).

Conclusion

Obesity is not a contraindication to P-TEVAR. P-TEVAR can be performed safely, despite the need for larger diameter sheaths. However, patients predicted to need adjunctive stenting or possessing smaller access vessel diameter to sheath size ratios are at highest risk of preclose failure using the Perclose Proglide device, and selective use of this technique is recommended.

Introduction

Thoracic endovascular aneurysm repair(TEVAR) is increasingly performed for a variety of thoracic aortic pathologies^1–3. Thoracic endografts tend to be larger in diameter than those utilized in the abdominal aorta and require larger sheaths for delivery, some up to 27 French in outer diameter(OD). Consequently, TEVAR procedures are often performed by delivering the endograft through open femoral exposure, or creation of an aortic/iliac conduit in 20–30% of cases^{4, 5}. Due to the success of the preclose technique for aortic endograft placement^{6, 7}, our practice has evolved to implement this access strategy in the majority of TEVAR patients (P-TEVAR), despite the need for larger sheath sizes.

In addition to shorter operative times⁷, potential advantages of percutaneous access include reduced discomfort, earlier ambulation, and a lower rate of wound complications^{8, 9}. Wound complications with open femoral exposure in endovascular aortic repair have been reported in 3–5% of patients, despite efforts to reduce this risk by making limited transverse or oblique incisions¹⁰. Obesity is a known risk factor for groin-wound morbidity ^{10, 11}, and this patient population potentially stands to benefit the most from percutaneous access for endovascular aortic procedures. However, in initial reports of the preclose technique, obesity was felt to be a relative contra-indication due to concerns about access vessel depth and suture capture^{7, 12}. Currently, there are limited data analyzing P-TEVAR, and no publications specifically examine the impact of obesity on procedural safety and success. The purpose of this analysis is to describe our experience with P-TEVAR and compare outcomes in obese and non-obese patients.

Methods

Approval for this study was obtained from the University of Florida College of Medicine Institutional Review Board.

Database and subjects

All patients undergoing TEVAR for any indication at the University of Florida between 2005 and 2011 were prospectively entered into an endovascular database. This database was queried for demographics, comorbidities, indications, and postoperative complications. Confirmation of patient and procedure specific outcomes was verified with retrospective review of the electronic medical record(EMR). Comorbidities and procedure related outcomes were defined and graded utilizing SVS reporting guidelines¹³. Patients undergoing percutaneous access and closure of a common femoral artery(CFA) that received thoracic endograft delivery were identified and further reviewed. Subjects undergoing femoral exposure or open conduit placement for device insertion were excluded. Preoperative computed tomographic angiograms(CTA) were examined to determine anatomical and morphologic data that were not routinely included in the database, including access vessel depth, degree of femoral plaque burden and/or calcification.

Definitions

Patients were dichotomized as obese or non-obese and outcomes were further analyzed. Obesity was defined as a body mass index(BMI) ≥ 30 kg/m²(World Health Organization definition: www.who.int/mediacentre/factsheets)¹⁴. Technical success of the preclose technique during TEVAR was defined as the ability to achieve hemostasis and maintain limb perfusion without the need for CFA exposure and/or obligate surgical repair of the vessel for 30-days postoperatively. Any access-related complications identified in the EMR or on postoperative CT review were also considered failures. These events were tabulated even if they were conservatively managed. Complications that were categorized as a preclose technical failure included development of lower extremity emboli, surgical site infection(deep or superficial requiring antibiotics and/or surgical debridement), de novo access vessel lesions such as hematoma(which was clinically diagnosed and either treated conservatively or with surgical evacuation), flow-limiting dissections, clinically-significant stenosis(e.g., new onset claudication/limb ischemia, ≥ 50% cross-sectional diameter reduction and/or vessel occlusion), pseudoaneurysms, arteriovenous fistulas, or documented Perclose device malfunction requiring arterial repair.

Access vessel evaluation

CFA depth and cross-sectional diameter was measured from a predefined anatomic reference point. The reference point was chosen in the mid-femoral head, and measured in a straight line from the anterior vessel wall to the skin surface immediately overlying the artery. This was thought to be the most consistent anatomical marker for analysis of vessel depth and morphology, and is typically above the CFA bifurcation(Figure 1). Per our protocol, the caudal extent of all CT scans extended below the femoral head, and for each patient the entire femoral bifurcation was visualized for any anatomic variation. If a patient’s femoral artery bifurcation was located cranial relative to the femoral neck, then the vessel depth was measured at the location of the bifurcation.

Figure 1. Method of Femoral Access Vessel Assessment.

The common femoral artery was located at the mid-femoral head and a measurement from the skin to the anterior vessel wall was obtained to determine access vessel depth (A). The anterior-posterior (B), and left-right (C) inner vessel diameter, as well as a plaque morphology and calcium score was obtained for each vessel accessed with a sheath that was 20 French outer diameter, or larger.

CFA calcification and morphology was scored based on presence or absence of atherosclerotic plaque from the superficial epigastric artery to the femoral bifurcation (0: normal; 1: < 50% posterior wall plaque, no calcium; 2: plaque and/or calcium involving ≥ 50% of luminal diameter; 3: anterior wall calcification; 4: circumferential calcification). A similar scoring system has been previously described in a report predicting vascular complications of percutaneous aortic valve replacement¹⁵. Two independent observers(JK, SS) reviewed postoperative imaging and scored femoral access anatomy and morphology. Observed agreement between the reviewers was 96% with a κ of 0.92(95% confidence intervals[CI] .86–1.00).

Clinical practice

TEVAR case planning in all patients was completed on a three dimensional workstation(TeraRecon, Aquarius Inc., San Matteo, CA) and access vessel diameter, tortuosity and calcification were routinely evaluated. Surgeon discretion determined eligibility of patients for P-TEVAR and no standard criteria were used to select patients during the study interval. All TEVAR patients routinely underwent pulse examination pre- and postoperatively. Additionally, Doppler insonation was commonly utilized to evaluate pedal signals at the beginning and end of each case and throughout the patient’s hospitalization. Selective use of ankle-brachial indices occurred in patients with pre-existing peripheral arterial occlusive disease, but no standardized protocol was in place over the study period. Need for adjunctive interventions such as conduits or iliac angioplasty and/or stent placement was left to the surgeon’s discretion. In general, iliac diameters ≤ 7 millimeters and/or presence of significant atherosclerotic disease often prompted selective iliac angioplasty(±stent) or open/endovascular conduit use.

“Preclose” technique

An in depth description of the technical elements of percutaneous access in aortic endograft repair has been described previously by our group and others^{7, 9, 12}. Briefly, in the majority of cases, initial percutaneous access was achieved utilizing a 21-gage micropuncture needle placed through the anterior CFA wall with fluoroscopic guidance¹⁶ and/or selective ultrasound visualization¹⁷, depending on surgeon preference. A 0.018-inch introducer wire was passed through the access needle followed by Seldinger exchange for a 4 French introducer sheath(Cook Medical, Bloomington, IN). While some surgeons in the group use bony landmarks at the time of blood return from the micropuncture needle to determine the puncture site, others perform a hand-injection digital subtraction arteriogram with a magnified, ipsilateral oblique view to verify successful CFA cannulation.

Next, the preclose technique⁷ was completed using the Perclose Proglide^® device(Abbott Vascular, Redwood City, CA). The Proglide^® device is a suture-mediated access closure device that is designed to provide closure of a 5–21 Fr arteriotomy by deploying two small nitinol needles through the entire thickness of the anterior CFA wall, with two devices required for >8F sheaths, per the instructions for use. These needles are attached to a pre-tied, 3-O Prolene suture. Within the instructions for use, the manufacturer explicitly states that the Proglide^® system is designed for closure of only common femoral arteriotomy up to 21 French with the preclose technique, and has not been tested in patients with significant arterial calcification, obesity or need for larger sheath diameters (http://www.abbottvascular.com/static/cms_workspace/pdf/ifu/vessel_closure/eIFU_Perclose_ProGlide.pdf).

Our technique represents an “off label use” as several patients within this study had some or all of these attributes. To complete the preclose technique, we deploy two Proglide^® sutures at the beginning of the procedure, off-set by 10-degrees rotation in a medial and lateral direction. The pre-tied knot is tied down after procedure completion as previously described⁷. Surgeon choice determined the need to place additional sutures at case completion depending on impression of hemostasis. Judgment of the need to perform femoral exposure and repair due to Preclose failure was left to the operating surgeon’s discretion.

Statistical methods

A majority of patients were determined to have one preclose site per TEVAR procedure, and contralateral femoral access outcomes were not included in the analysis (typical contralateral CFA sheath diameters=6Fr) unless the contralateral sheath was >20 Fr outer diameter(OD). In those cases where the contralateral sheath was >20 Fr OD and the preclose technique was used, those vessels were analyzed as a separate vessel in the analysis. Continuous variables were analyzed using Student’s t tests and categorical variables were compared with chi-square or Fisher’s exact test. Generalized estimating equations were used to assess whether BMI(considered as a continuous variable) was a predictor of failure when accounting for covariates. Preclose failure was used as the outcome variable. The primary predictor was BMI, and patients were treated as a repeated factor to account for the double observations(bilateral CFA access) for some cases. Manual backward step-wise elimination was performed based on p-values and model predictive power. Since primary interest was in BMI, BMI was retained as a covariate in all models regardless of its estimated effect, and all interactions between BMI and the other covariates were also tested. All data processing and statistical analysis was completed using the R statistical software package(https://www.r-project.org/). A P-value < .05 was considered significant.

Results

Between March 2003 and May 2011, 536 TEVAR patients were identified, in whom 355(66%) had total percutaneous access using the preclose technique (Figure 2). There were 366 P-TEVAR arteries identified that were closed using Perclose Proglide^® devices, with a mean of 2.1±0.4 devices used per artery. There were 270 men(76%).and the overall mean(±standard deviation: SD) BMI was 26.6±5.4. There were 91 patients identified with a BMI ≥ 30 kg/m², and mean BMI in this cohort of patients was 34.8±3.8. Differences in demographics of non-obese and obese patients are highlighted in Table 1. Notably, obese patients were typically younger, more likely to be male, and have renal insufficiency and/or diabetes mellitus. However, obese patients less frequently had a history of coronary artery disease.

Figure 2. Patient and Access Vessel Selection for P-TEVAR.

A total of 536 patients were available at the time of analysis, with 365 arteries from 355 patients accessed with a sheath that was > 20Fr outer diameter. Of these, 336 were successfully closed percutaneously using the preclose technique, with 29 failures. When using a body mass index ≥ 30 as a definition of obesity, it was not associated with a higher rate of failure (P=NS vs. non-obese patients).

Table 1.

Patient characteristics and comorbidities^δ

Feature, No. (%)	Non-obese (N = 264)	Obese (BMI > 30) (N = 91)	P-value
Age, mean ± SD, years	66±16	59±16	.0004
Female, No. (%)	71 (27)	15 (17)	.06
Prior groin operation	13 (7)	2 (3)	.2
Comorbidities, No. (%)
Hypertension	220 (86)	79 (87)	.9
Dyslipidemia	105 (41)	31 (34)	.3
Smoking	91 (36)	29 (32)	.6
Coronary artery disease	67 (26)	14 (15)	.04
COPD	50 (20)	11 (12)	.1
Renal insufficiency	44 (17)	25 (28)	.05
Arrythmia	34 (13)	8 (9)	.3
Diabetes mellitus	23 (9)	17 (19)	.02
Cerebrovascular disease	28 (11)	6 (7)	.3
Peripheral arterial disease	20 (8)	4 (4)	.4
Congestive heart failure	15 (6)	4 (4)	.8

^δFischer’s exact, t-test or Chi-square test when appropriate; COPD, chronic obstructive pulmonary disease

Details of the anatomic and procedure specific variables are highlighted in Table 2. The mean access vessel depth was significantly greater in obese patients, with an average depth of 50±20mm compared to 30±13mm for non-obese patients(P < .0001) (Figure 3A). Mean sheath diameter was slightly larger among obese patients(P=.04), but average stent diameters were similar between the two groups. Obese patients were more likely to have larger inner femoral access vessel diameters compared to non-obese patients(P=.02)(Figure 3B). Pre-TEVAR deployment adjunctive iliac stents were utilized in 7%(N=27) of all cases(11% obese patients vs. 6% non-obese patients, P=.2). Greater than 2 Proglide devices were used in 8.8% (N = 32) of cases. For the 294(80%) procedures in which patients had adequate CT imaging for femoral plaque and calcium scoring, there was no difference in categorization of access vessel morphology between obese and non-obese patients(P = .3; Table 3).

Table 2.

Anatomic and procedure specific variables between non-obese and obese P-TEVAR patients (mean±SD)^δ

Feature (No. %)	Non-obese (N = 274)	Obese (BMI ≥ 30) (N = 92)	P-value
“Preclose” vessel
Bilateral access	31 (11)	9 (10)
Left femoral artery	60 (22)	17 (18)
Right femoral artery	183 (67)	66 (72)	.7
Procedural urgency
Elective	168 (61)	57 (62)
Urgent/emergent	106 (39)	35 (38)	1
Access vessel depth (mm)	30±13	50±20	<.0001
Access vessel inner diameter (mm)	7.9±2.0	8.5±1.9	.02
Largest sheath diameter (Fr)	25.0±1.5	25.4±1.4	.04
Maximum stent diameter (mm)	35.3±5.7	36.2±4.8	.2
Number Perclose used	2.1±0.4	2.1±0.4	.3
Iliac stent use	16 (6)	10 (11)	.2

^δFischer’s exact, Mann-Whitney or Chi-square test when appropriate; Fr, French

Figure 3. Body Mass Index and Access Vessel Morphology Trends.

These scatter plots demonstrate that, as expected, the access vessel depth is greater as patient BMI increases (A). Of note, the inner vessel diameter also tended to be larger with increasing BMI (B).

Table 3.

Access vessel morphology score as a function of patient body habitus^*.

Femoral artery pathology No. (%)	Non-obese (N = 215)	Obese(BMI > 30) (N = 79)
No calcium	82 (38%)	41 (52%)
Posterior wall plaque	74 (34%)	22 (28%)
>50% of vessel circumference plaque	5 (2%)	0
Anterior wall Ca+2	38 (18%)	11 (14%)
Circumferential Ca+2	16 (7%)	5 (6%)

^*P =0.3 by Fisher’s exact test; imaging to review for access vessel scoring was not available for 72 procedures, and rate of missingness did not differ between obese and non-obese patients

The overall preclose success rate was 92%: 91% for non-obese patients and 93% for obese patients(P=.7). Twenty-nine patients had documented technical failure of the preclose technique during their TEVAR procedure. No perioperative deaths were identified that were directly attributed to failed percutaneous access. No significant differences in BMI were noted between patients experiencing successful or unsuccessful percutaneous CFA closure(P=.2). Additional details of the analyzed covariates for technical success of percutaneous TEVAR are depicted in Table 4. The outcomes of the 29 patients with failed percutaneous access are demonstrated in Figure 4. Notably, a vast majority of the failures(90%; N=26) were recognized and managed at time of the index TEVAR procedure. Three patients underwent subsequent operations within 30-days of P-TEVAR for access vessel complications related to the site of the preclose study vessel; two obese patients and one non-obese patient(P=.3). The non-obese patient had a left iliofemoral occlusion with subsequent repair. The two remaining patients underwent femoral pseudoaneurysm repair(1-mycotic, 1- non-mycotic).

Table 4.

Comparison of tested covariates for success or failure of percutaneous TEVAR

Feature (No. % or SD)	Success (n=336)	Failure (n=29)	P-value*
BMI	27.5 (5)	26.4 (4)	.2
Gender
Female	78 (90)	9 (10)	.3
Age	64 (16)	70 (13)	.03
Race
White	244 (72)	23 (79)
Black	69 (21)	3 (11)
Other	23 (7)	2 (9)	.3
Prior groin operation	14 (93)	1 (7)	.9
Sheath size (Fr±SD)	25.1 (±1.5)	25.6 (±1.3)	.05
Number of perclose
≤ 2	311 (93)	21 (72)
>2	24 (7)	8 (28)	.0007
Iliac stent used	18 (5)	8 (28)	<.0001
Access vessel
Both	35 (10)	4 (14)
Right	232 (69)	17 (59)
Left	69 (21)	8 (28)	.5
Access vessel diameter (mm)	8.1 (±2.0)	7.5 (±1.5)	.08
Access vessel depth (mm)	35.5 (±17.8)	36.1 (±15.5)	.8
Femoral plaque/Ca+2 score□	1.1 (±1.3)	1.6 (±1.3)	.04
Comorbidity score	2.6 (±1.5)	3.1 (±1.8)	.09
Urgency
Elective	208 (92)	17 (8)
Urgent/Emergent	128 (91)	12 (9)	.7
Device category
Cook	144 (92)	12 (8)
Gore	164 (93)	13 (7)
Other	26 (90)	3 (10)	.9

^*p-values are the results of generalized estimating equations logistic models to account for repeated observations on some subjects;

^□femoral plaque/Ca+2 score was not used in the final model due to 20% rate of missingness

Figure 4. Preclose Failure Management.

Preclose failure was defined as the inability to achieve hemostasis and maintain limb perfusion without the need for common femoral artery exposure and/or obligate surgical repair of the vessel within a 30-day postoperative period (see methods for further details). Three hematomas were documented clinically and confirmed with duplex ultrasound and managed expectantly (no transfusion/evacuation). All of these hematomas resolved during postoperative follow-up. Of the other 26 failures, four required an iliofemoral bypass, 10 required an endarterectomy and patch angioplasty, and 12 arteriotomies were repaired primarily.

Predictors of TEVAR “Preclose” failure

Seventy-two percutaneously accessed arteries had inadequate CT imaging data for femoral plaque scoring due to lack of intravenous contrast. This was secondary to the emergent nature of their procedures, as well as the retrospective methods of the study(20% of total observations). As such, access vessel plaque scores could not be reliably obtained for these patients, and thus were not considered in the multivariable models. However, access vessel diameter and vessel depth were included. The final multivariable analysis dataset included 365 observations(29 failures and 336 successes) in 355 patients; of note, the success rate of the analyzed cohort did not differ significantly from the cohort not analyzed(92% vs. 96%, P=.23).

Probability of preclose failure during TEVAR was inversely correlated to BMI, which had significant interaction with sheath diameter(e.g., ↓ BMI + ↑ sheath size ∝ ↓ failure). Specifically, among patients with failed percutaneous access, a negative correlation(r = −0.21) was detected [18 of 29 failure patients(62%) had a BMI that was below the average BMI for the entire study cohort and their average sheath sizes were greater than the entire study cohort]. This interaction seemed greatest in patients with a BMI< 26 who had a sheath >24 French utilized during their P-TEVAR procedure(OR 2.3, 95% CI 1.03–5.3, P=.04). Additional covariates that had linear relationships with failed percutaneous access included age and sheath size. Greater patient age and increasing sheath size was correlated with higher probability of preclose failure(Figure 5). When examining the impact of access vessel diameter to sheath size ratio, probability of failure increased in a linear fashion(access vessel size: sheath ratio ↓ ∝ ↑ failure; median ratio in successful P-TEVAR, .32 [IQR .27, .37] vs. failed P-TEVAR, .29 [IQR .24, .33]; P=.02). The interaction between BMI and access vessel to sheath diameter ratio is depicted in Figure 6. The final multivariable model predicted preclose failure correctly for 74.8% of subjects, and based on this modeling, independent predictors of failure were adjunctive iliac stent(OR 9.5; 95%CI 3.3–27.8, P < .0001), >2 Proglide^® devices(OR 7.0; 2.3–21, P=.0005), and smaller access vessel/sheath size ratio(OR multiplies by 1.1 for each .01 decrease in ratio; 1.02–1.2, P=.007).

Figure 5. Probability of preclose failure during P-TEVAR.

This graph demonstrates some of the univariate associations with preclose failure during P-TEVAR. Notably, factors such as age and sheath diameter were linearly correlated with failure; however BMI had an inverse relationship (e.g. ↓ BMI∝↑ failure rate). Of the three demonstrated variables in this graph, only sheath diameter independently predicted failure in multivariable analysis. The optimal prediction model was obtained when accounting for the interaction of sheath size and access vessel inner diameter.

Figure 6. Interaction between access vessel diameter and sheath size.

This plot demonstrates the interaction between the femoral access vessel diameter and sheath size ratio as a function of patient BMI. A decreasing access vessel diameter to sheath size ratio (e.g. smaller vessel and larger sheaths) was an independent predictor of preclose failure during P-TEVAR.

Discussion

This study demonstrated that percutaneous closure of the CFA during TEVAR can be completed safely with a high degree of technical success, independent of patient BMI. This report represents one of the largest experiences with percutaneous TEVAR published to date, and is the first to demonstrate that obesity is not associated with failure of the preclose technique. The predictors of preclose failure were primarily related to iliofemoral access vessel quality, specifically access vessel diameter, sheath size and need for adjunctive iliac stenting. Interestingly, contrary to other reports of percutaneous closure of large femoral access during endovascular aortic repair, larger sheath diameters(>24Fr) were not independently associated with failure unless they were utilized in patients with lower BMI(<26).

In the initial report from our institution that demonstrated safety of percutaneous CFA access for delivery of aortic endografts, 71 of the 183 patients underwent TEVAR. Compared to infrarenal aortic devices, thoracic devices are typically larger and therefore the access-related complication rates are frequently higher, regardless of the technique chosen to deliver the device^{4, 18, 19}. As such, there is significant selection bias in any such report, as was evident in our initial series in which the preclose cohort was compared to a contemporaneous series of 154 surgically-exposed femoral arteries. Nevertheless, it was shown that smaller sheath access(12–14Fr) was statistically more likely to achieve successful closure percutaneously than large groin access(20–24Fr)(99% vs. 91.4%, P< .01)⁷. These results are consistent with other publications on P-EVAR²⁰ and potentially could curb enthusiasm for P-TEVAR.

As overall clinical practice evolved and device/sheath technology has improved with respect to both luminal profile reduction and hydrophilic materials, clinicians have become collectively more aggressive in pursuing minimally invasive access^{8, 15, 21}. In our own practice, this has translated into a reduction in the number of iliac conduits, once reported to be required in 20% of TEVAR procedures⁴. Totally percutaneous endovascular aortic repair now constitutes an overwhelming majority(>95%) of our practice. Nationally, however, it is likely that clinicians still utilize routine open femoral exposure during endovascular aortic repair for a variety of reasons, including a paucity of literature for P-TEVAR, technical unfamiliarity with the preclose technique, as well as potential financial implications of percutaneous suture closure device utilization and inability to submit a reimbursement code for femoral exposure. Patient-related factors such as age, gender, femoral artery calcification, obesity, and scarred and/or reoperative groin vessels have been reported to contribute to percutaneous failures and in some studies are used as exclusion criteria of percutaneous aortic repair^{20, 22, 23}. Of note, a relative contraindication to the preclose technique with our initial experience was patient obesity^{7, 12}. The greater vessel depth and increased interposed peri-arterial fat was thought to confer a higher risk of suture non-capture for the preclose technique. Indeed, this bias was recently corroborated by Skagius and colleagues, who documented that increased groin subcutaneous fat was associated with primary failure of P-TEVAR using the ProStar XL device²⁴ (Abbott Vascular, Redwood City, CA).

However, in the current study, our results demonstrate that obesity was not associated with lower technical success when performing percutaneous aortic repair. Interestingly, lower BMI was correlated with decreased vessel depth, but also inner vessel diameter(Figure 3B), which was more frequently associated with P-TEVAR failure when larger sheath sizes were employed during the procedures(Figure 5).

With regard to vessel morphology, the results of this analysis are consistent with previous reports on the impact of CFA disease and success of totally percutaneous aortic repair^{23, 25}. Specifically, higher femoral plaque scores were associated with failure in the univariate analysis. However, increased CFA plaque/Ca⁺² burden did not fall out in the multivariable model as an independent predictor, and this was likely related to multiple factors including the non-parametric distribution of the scores, the fact that 20% of patients had inadequate imaging to accurately discern plaque morphology(Table 3) and the selection bias for which patients were thought to be candidates for P-TEVAR.

Metcalfe et al. reported that the success rate of percutaneous aortic endovascular repair in 186 femoral arteries was 95.2%⁸. On univariate analysis, they found that vessel depth, diameter, preoperative renal failure, and operator experience predicted failure of the percutaneous technique. Of note, operators were considered “experienced” when he/she had successfully performed 20 such procedures. On multivariable analysis, only operator experience predicted failure. Although this report used a single Prostar XL device(Abbott Vascular, Redwood City, CA), the overall success rate and impact of vessel diameter was remarkably similar in our report. We did not specifically examine operator experience as all of the operators in this study were considered higher-volume endovascular aortic surgeons(>20 procedures/year).

Our current clinical practice for totally percutaneous aortic repair involves extensive preoperative planning using centerline of flow 3D-reconstruction. Access vessel depth, diameter, tortuosity, and calcification are evaluated and assessment made about the feasibility of transfemoral delivery of the endoprosthesis. Endovascular iliac conduits¹⁹ are used selectively even in conjunction with percutaneous CFA access. Clinical scenarios where we typically do not attempt percutaneous aortic repair include procedures utilizing an aorto-uni-iliac endograft in which a femorofemoral bypass is planned, extensive anterior CFA calcification, and presence of a prosthetic graft in the femoral artery. The implications of percutaneous failure are generally benign, and a sheath can be inserted for hemostasis while femoral exposure is performed. Thus, we are quite aggressive with attempting totally percutaneous repair, even in situations where we feel it may fail.

There are several important limitations to this study that are inherent to its single-center retrospective design. First, it was not possible for us to retrospectively determine the exact site of femoral artery access for our vessel analysis, and we acknowledge that we may not have evaluated the pertinent portion of the femoral artery. Further, we were unable to determine whether and to what degree the overlying skin and subcutaneous tissue was cranially displaced during access, as is often done in obese patients, which would alter the depth of the artery from the skin. We feel that these limitations were mitigated by utilizing a fixed bony landmark in all patients.

Additionally, no standardized protocol was in place that designated when patients underwent open femoral exposure or percutaneous access for TEVAR. Due to patient numbers, extensive subgroup analysis of different obesity strata was not possible, so it is conceivable that in patients with greater extremes of obesity, outcomes with P-TEVAR may be different. Further, robust predictive modeling was not feasible due to the limited number of preclose failure events. Unfortunately, the impact of femoral plaque/calcium score could not be analyzed in multivariable analysis due to missing data. Importantly, no comparison to the outcomes of standard open femoral exposure during TEVAR was presented, but we have reported this type of comparative analysis previously in a mixed EVAR/TEVAR patient population^{6, 7}. Also, routine preoperative and serial postoperative ABI measurements were not uniformly obtained, which may have provided a more sensitive assessment of any hemodynamic perturbations caused by percutaneous access. Longer follow-up is needed to determine if there are any risks of developing common femoral artery stenosis after percutaneous aortic repair. Lastly, all P-TEVAR procedures were performed by surgeons with extensive totally percutaneous aortic repair experience so the impact of operator learning curve could not be analyzed.

Conclusions

P-TEVAR appears to be safe, despite the need for larger diameter sheaths during the procedure. Obesity is not a contraindication for the preclose technique; however, patients predicted to need adjunctive stenting or possessing smaller access vessel to sheath diameter ratios are at highest risk of failure, and selective use of this technique is recommended.