Percutaneous Access for Endovascular Abdominal Aortic Aneurysm Repair: Can Selection Criteria Be Expanded?
Article Outline
Previous reports suggest that percutaneous access for endovascular abdominal aortic aneurysm repair (P-EVAR) is as safe as open access (O-EVAR) in patients with favorable femoral anatomy. Severe femoral artery calcification and obesity have been considered relative contraindications to P-EVAR, but these criteria have not been evaluated. The purpose of this study was to assess the postoperative anatomic changes associated with P-EVAR versus O-EVAR using three-dimensional (3-D) computed tomographic (CT) reconstruction and to evaluate the overall results of the two procedures in a group of patients with suboptimal femoral anatomy. During a recent 26-month period, 173 patients underwent EVAR at our institutions, including 35 P-EVARs. Of these, 22 (63%) had complete pre- and postoperative CT imaging of the femoral arteries. These subjects were compared to 22 matched controls who underwent O-EVAR during the same period. Automated 3-D reconstructions were used to measure the following anatomic femoral artery parameters before and after EVAR: arterial depth, calcification score, minimum diameter and area, and maximum diameter and area. Of the 88 study arteries, 50 underwent open access and 38 percutaneous access (Proglide, n
=11; Prostar XL, n=27). Both groups were similar regarding sheath size, number of components, operative time, blood loss, and length of stay. Significantly more O-EVAR subjects suffered groin complications (p=0.02), including five hematomas, two wound infections, two femoral thromboses, and one vessel which required patch repair. In the P-EVAR group there was only one hematoma, which was managed conservatively. There was no difference between the P-EVAR and O-EVAR groups with respect to femoral artery calcification (Agatston scores 667±719 vs. 945±1,248, p=0.37). Obesity (body mass index >30) was documented in six (27%) of both the P-EVAR and O-EVAR groups (p=nonsignificant). Pre- and postoperative CT-derived anatomic data showed a significant decrease in the minimal vessel area with O-EVAR compared to P-EVAR (p=0.02). This study demonstrates that patients with obesity or severely calcified femoral arteries can be successfully treated percutaneously with fewer minor groin complications.
Introduction
Endovascular repair of infrarenal abdominal aortic aneurysms (EVAR) has been accepted worldwide for patients with suitable aneurysm morphology. Due to the need for introduction of large sheaths, open common femoral artery (CFA) exposure has become the gold standard for EVAR access. However, improvements in arteriotomy closure devices have made percutaneous access possible, with secure closure even after the use of large sheaths. The first report of percutaneous EVAR with the “preclose” technique appeared in 1999.1 Although subsequent limited experience has been favorable, percutaneous EVAR has not been considered universally applicable. Due to potential problems with secure arteriotomy closure, the anatomic considerations of femoral artery calcification and morbid obesity have generally been considered relative contraindications to percutaneous EVAR.2, 3 However, these criteria have not been systematically evaluated. As we have gained experience with percutaneous EVAR (P-EVAR) in selected patients, we hypothesized that the technique could be applied to most patients regardless of femoral artery calcification or body mass index (BMI). The purpose of this study was to evaluate whether P-EVAR can be safely performed in patients with suboptimal femoral anatomy, with particular emphasis on anatomic changes in the CFAs using automated three-dimensional (3-D) reconstruction measurements.
Methods
We reviewed our experience with 173 consecutive patients who underwent EVAR at our institution from March 2005 to April 2007. During this period, 35 patients (20%) underwent P-EVAR. Twenty-two of these (63%) had complete pre- and postoperative computed tomographic (CT) imaging for analysis of the femoral arteries, and these subjects served as the focus of our study. Patients with inadequate imaging of the femoral arteries were excluded. Excluded studies mainly included poorly contrasted scans, presence of prosthetic hips causing artifact, and outside preoperative scans which were not available for analysis. The last preoperative CT scan was compared to the first postoperative scan, typically after 1 month (mean 58±88 days), and the last obtained CT scan obtained (mean 313±139 days). Study subjects were compared to 22 control patients who underwent open access EVAR (O-EVAR) during the same period and matched for gender, age, device type, sheath size, and surgeon. The decision to offer percutaneous versus open access for EVAR was based entirely on surgeon preference, examining each patient's anatomy, habitus, and operative plan. Each operator evaluates the approach on a case-by-case basis without adherence to strict anatomic criteria, though the extremes of morbid obesity and completely calcified vessels were typically chosen for open access.
The following data were recorded for each patient: demographics, atherosclerotic risk factors, medical comorbidities, BMI, largest sheath size, and number of component devices placed through each CFA. Outcomes for the percutaneous and open groups were then compared, including systemic and local complications, length of hospital stay, operative time, estimated blood loss (EBL), and need for transfusion.
Anatomic factors for percutaneous access were then examined and compared in percutaneous and open subjects and included CFA depth from the skin and CFA calcification. Obesity can complicate both percutaneous and open access for EVAR, making exposure and control of the femoral vessels difficult. However, some patients may distribute fat unevenly and be more or less obese in the femoral area. For this reason, we measured the depth of the CFA on the preoperative CT scan. CFA depth was measured in a vertical line from the outer, anterior wall of the artery to the skin (Fig. 1). Calcium scoring was performed using the Vitrea Vscore (Vital Images, Minnetonka, MN) protocol for coronary vessels. This program is designed for coronary artery calcium scoring, but the technique can be used to quantify the calcification in peripheral arteries. The density threshold for measuring calcification is 130 Hounsfield units, and the degree of calcification is based on the software-derived Agatston score. An Agatston score of >400 for coronary vessel calcification is used to correlate with extensive atherosclerotic plaque.4 Calcifications in the CFA on each side were quantified separately, measuring the entire length of the CFA (Fig. 2).
Fig. 1
Obesity can complicate percutaneous or open access to the CFA. Depth to the CFA was measured in a vertical line from the anterior wall of the vessel to the skin.
Fig. 2
Calcification in the CFA was quantified using a coronary calcium scoring program. In this patient with severe calcification, the Agatston scores were 2,445 for the right CFA and 1,732 for the left CFA.
Using 3-D CT reconstruction, a detailed examination of each CFA for any anatomic abnormality was performed in the percutaneous and open groups. A centerline was created along the CFA, and the automated Vitrea software package calculated the minimum and maximum diameters, measured inner wall to inner wall. In addition, the minimum and maximum areas (cm2) were measured for each slice along the CFA. This automated the measurement process and minimized investigator bias. On a few heavily calcified arteries, the automated measurement mistakenly included the calcified wall. In these instances, the diameters were manually measured.
A detailed description of the preclose technique has been published previously.5, 6 This study includes the preclose technique using two different devices, the Prostar XL (n=27) and the Perclose Proglide (n=11) (Abbott Vascular, Redwood City, CA). Surgeon preference dictated which device was used. Each surgeon exclusively used either the Prostar XL (F. R. A.) or the Perclose Proglide (C. H. T.) during this study period.
Statistical Analysis
Descriptive statistics for categorical variables are presented as relative frequencies (percents). Continuous data were tested for normal distribution by use of the Kolmogorov-Smirnov test and are presented as medians and interquartile ranges. Statistical analyses were performed using chi-squared tests for categorical variables and Wilcoxon rank sum tests for continuous variables. A p value of 0.05 was considered statistically significant. Analyses were conducted with SAS version 9.1.3 (SAS Institute, Cary, NC).
Results
P-EVAR and O-EVAR patients had similar demographics and comorbidities, as seen in Table I. Obesity (BMI >30kg/m2) was relatively common in this study, occurring in six (27%) of the percutaneous and six (27%) of the open access patients (p=0.78). There was no difference in the type of EVAR device (p=0.19) (Table I), and the median sheath size was 18F in CFAs accessed in both groups (p=0.98). The mean number of components placed through each CFA was 1.6±0.7 in the percutaneous group and 1.4±0.5 in the open access group (p=0.49). Technical success was 100% in the percutaneous patients, none of whom required conversion to open CFA access. Although there was a trend toward shorter operating times in the percutaneous group, this difference was not statistically significant (158±46 vs. 176±59min, p=0.32). There was no difference in blood loss between the percutaneous and open access patients (273±268 vs. 287±195mL, p=0.81). The less invasive percutaneous access did not translate into shorter hospital stays, with the median length of stay being 2±1 days in both groups (p=0.86).
Table I. Characteristics of patients in the percutaneous and open access groups
| Percutaneous (n=22) | Open (n=22) | p | |
|---|---|---|---|
| Mean age (years) | 72±10 | 71±8 | 0.75 |
| Sex (male) | 19 | 21 | 0.60 |
| Mean BMI (kg/m2) | 27.4±5 | 27.4±6 | 1.0 |
| BMI >30 | 6 (27%) | 6 (27%) | 1.0 |
| Active smoking | 7 (32%) | 7 (32%) | 1.0 |
| Diabetes | 5 (23%) | 1 (5%) | 0.19 |
| Hypertension | 16 (73%) | 18 (82%) | 0.72 |
| Hypercholesterolemia | 10 (46%) | 13 (59%) | 0.55 |
| CAD | 15 (68%) | 10 (45%) | 0.22 |
| CVD | 5 (23%) | 5 (23%) | 1.0 |
| PAD | 4 (18%) | 2 (9%) | 0.66 |
| EVAR device type | 0.19 | ||
| AneuRx | 13 (59%) | 7 (32%) | |
| Excluder | 5 (23%) | 9 (36%) | |
| Zenith | 4 (18%) | 6 (27%) |
There were no major groin complications such as life-threatening bleeding or hematomas requiring transfusions, delayed hospitalization, or reoperation. There were significantly fewer minor groin complications in percutaneous patients compared to the open group (2.6% vs. 20.0%, p=0.02) as seen in Table II. One percutaneous patient and five open patients developed a minor hematoma, defined as a palpable mass on the first postoperative visit that was confirmed with ultrasound. There were no wound infections in the percutaneous group; one open patient developed a localized cellulitis, and another developed a superficial infection, which were treated with antibiotics. Two CFA thromboses were discovered in the open group; there were no CFA thromboses in the percutaneous patients. Both femoral thromboses were noted at follow-up (one at 6 months and the other at 1 year) and were associated with claudication. Both patients have mild symptoms, and neither has required reintervention. One open access patient required an unexpected CFA endarterectomy and patch angioplasty. Of the 13 percutaneous patients excluded for inadequate CT imaging of the femoral arteries, one (7.7%) developed a minor hematoma; and there were no other major or minor complications.
Table II. Minor groin complications after EVAR
| Percutaneous (n=38) | Open access (n=50) | |
|---|---|---|
| Minor hematoma | 1 | 5 |
| Minor infection | 0 | 2 |
| Thrombosis | 0 | 2 |
| Patch repair | 0 | 1 |
| Total∗ | 1 (2.6%) | 10 (20.0%) |
∗p |
Anatomic Factors
In order to evaluate the anatomic criteria related to access selection, the 38 percutaneously accessed CFAs were compared with the 50 CFAs accessed by the open technique. Percutaneous access vessels had a median depth of 36±18.2mm and open vessels had a depth of 39±18.0mm (p=0.59). The median Agatston score for femoral calcification was 667±719 in the percutaneous group and 945±1,248 in the open group (p=0.37). Using a threshold value of >400, 17 (45%) of the percutaneous vessels and 20 (40%) of the open vessels had severe femoral calcification. There was no significant difference even for the most severely calcified vessels with Agatston scores >2,000 (percutaneous n=4 vs. open access n=7, p=0.70).
To assess for CFA stenosis or aneurysmal change, preoperative CT measurements were compared to the first postoperative scan, obtained at 1 month after hospital discharge (Table III). Measurements from the most recent (mean follow-up=257 days) CT were also recorded. As noted in Table III, the minimal vessel area was significantly reduced in the postoperative period for vessels accessed by open techniques; the median minimal vessel area did not change at all for the percutaneous group. Interestingly, the maximum vessel area did not change in either group. Although there was some measurable decrease in the median vessel diameter in the percutaneous group, this difference was not significant compared with the open group.
Table III. CFA size changes after EVAR according to type of repair
| Variable | Change in P-EVAR group (median, IQR) | Change in O-EVAR group (median, IQR) | p∗ |
|---|---|---|---|
| Maximum vessel area (cm2) | 0.0 (−0.7–0.03) | 0.0 (−0.1–0.05) | 0.78 |
| Minimum vessel area (cm2) | 0.0 (−0.04–0.04) | −0.4 (−0.12–0.01) | 0.04 |
| Maximum vessel diameter (mm) | −0.1 (−0.3–0.5) | 0.0 (−0.4–0.5) | 0.95 |
| Minimum vessel diameter (mm) | −0.05 (−0.5–0.4) | −0.3 (−0.9–0.2) | 0.25 |
∗Wilcoxon-Mann-Whitney test. |
Discussion
A completely percutaneous approach for EVAR is appealing. It is consistent with the minimally invasive nature of EVAR, patients like it, and it has the benefit of fewer local wound complications. The devices are used off-label for P-EVAR, and there are no large, randomized studies comparing it to O-EVAR. However, there have been increasing reports of successful use of the preclose technique.5, 6, 7, 8 There is a learning curve as with any other technique, and patient selection may play an important role. In our initial experience, patients were selected for a percutaneous approach if they did not have obesity or heavily calcified vessels. However, as the surgeons in this study had more success in difficult patients, these selection criteria were relaxed. This study examines outcomes in patients with relative contraindications to P-EVAR.
Standard O-EVAR involves oblique incisions overlying the CFA. Wound complications are not infrequent with this approach and include hematomas, lymphoceles, femoral neuropathy, and wound infection. Morasch et al.9 compared 47 patients with bilateral percutaneous access to a group of open femoral cutdown patients for EVAR. At 30 days, their access-related complications were 23% in the open group compared to 0% in the percutaneous group. These complications were considered minor, and none required therapeutic intervention. This minor complication rate was similar in the present study (20% open access vs. 2.6% percutaneous, p=0.02). In the only randomized study of percutaneous versus open access for EVAR, Torsello et al.2 found a 20% lymphocele rate in the open group compared to none in the percutaneous group, which is also similar to our findings. Although lymphoceles and small hematomas are minor complications, they may be a source of discomfort to the patient. Our standard practice is to partially reverse heparin during EVAR with protamine, and this may contribute to the number of minor hematomas in the open access group.
The detailed anatomic measurements in this study were considered to be a critical part of the evaluation of percutaneous access because other noninvasive tests such as ankle-brachial indices are not adequately sensitive to detect a subclinical stenosis. Although there have been reports of both pseudoaneurysm formation and stenosis after P-EVAR,7 there were no stenoses or aneurysmal changes in the percutaneous vessels in this study. This compares to a decreased diameter and minimum vessel area (0.3mm and 0.4cm2, p=0.04) in O-EVAR vessels. Although this incremental change in the open access vessels compared to the percutaneous vessels is statistically significant, the clinical importance of this finding is unclear. Other than the two open access patients who had femoral thrombosis at follow-up and were excluded from this anatomic analysis, no patients were symptomatic. In patients with a decrease in diameter of the CFA after open access, the change was noted on the first postoperative CT and was stable on subsequent imaging. The more important finding may be that there was no deformation of the CFA after percutaneous access and closure. Specifically, no stenosis or aneurysmal change in the percutaneous group was noted on any postoperative CT scan in the analysis.
Two techniques for P-EVAR (Prostar XL and Perclose Proglide) were used in this study. Analyzing the two percutaneous methods, there was no difference in any variable in the anatomic analysis or in the overall outcomes. Each technique has relative advantages and disadvantages. The Prostar XL technique typically requires only one device and may therefore have a cost advantage, but it requires a more extensive subcutaneous dissection and may have a theoretically higher infection rate due to the braided sutures. However, no infection after P-EVAR with the Prostar XL device has been noted in this study or in our overall institutional experience. The Perclose Proglide device is less expensive, but two are used with our technique, which also requires an additional wire and device exchange. The sutures are polypropylene and may be less prone to infection, but they are more prone to fracture during sheath exchanges. In fact, Lee et al.6 recommends serial dilations to help avoid this pitfall. We have not used serial dilators other than a 7F sheath prior to the larger sheath or the EVAR device itself and have not noted a problem with suture fracture.
Other studies have shown an overall decrease in the length of the operation with P-EVAR.6, 7, 9 In this study, there was a trend toward shorter operative time in the P-EVAR group, though the difference was not significant (158 vs. 176min, p=0.32). Similarly, EBL was lower in the percutaneous group (273 vs. 287mL, p=0.81). These EBL estimates, however, are somewhat difficult and error-prone with both P- and O-EVAR. No patient in either group required a blood transfusion. Although others7, 9 have shown decreased length of stay, in this study it was 2 days in both groups. These parameters have been used to help justify the higher cost with percutaneous access relating to the suturing device. The focus of this study was not on cost analysis, but others have shown that P-EVAR is not cost-effective.6 The main benefit of the percutaneous approach is improved patient satisfaction. Even in this study with significantly more groin complications with O-EVAR, these were managed conservatively and did not increase overall costs.
Our data suggest that P-EVAR can be safely applied to patients with obesity and severe femoral calcification. Obesity has been considered a contraindication due to difficult access.2, 6 However, 27% of the percutaneously accessed arteries were in obese patients (BMI >30), and there were no complications related to access in this group. CFA depth may be a more accurate measurement of the difficulty in accessing the vessels in an obese patient. However, excessive CFA depth did not represent a contraindication to percutaneous access in the present study. The mean femoral artery depth in the percutaneous group was 36mm, and the deepest artery treated with this technique was 107mm (Prostar XL).
Severe femoral calcification has been reported to make percutaneous access for EVAR more difficult. Nehler et al.10 reported two cases of entrapped devices caused by calcified vessels in obese patients. The needles bent and were unable to be retrieved into the barrel mechanism due to the patient's severe femoral calcification and obesity. In our study, even the most severely calcified vessels (Agatston score >2,000, n=4) could be treated percutaneously. All of these patients had circumferential calcification of the CFA, and the mean BMI was 27.4±5. It is important to note that an extensive subcutaneous dissection is performed, allowing the barrel of the Prostar XL device to seat itself on the artery and retrieve the needles as they exit through the vessel.
A number of study limitations must be acknowledged. The small size of this study makes the analysis prone to a Type II statistical error. The retrospective nature makes the study subject to selection bias. The type of access, percutaneous or open, was entirely the operator's decision and may have led to favorably selecting percutaneous patients based on anatomic criteria, comorbidities, or the complexity of the planned operation. Including only the surgeons who performed percutaneous access (F. R. A., C. H. T.), 56% of their patients were selected for P-EVAR. As detailed above, there was no difference in the examined variables between the two groups, but this source of error cannot be eliminated. In fact, the majority of operations in this series were not performed percutaneously. Finally, the experience of the operating surgeon may prove to be an important determinant of outcome in the percutaneous approach. The two surgeons whose techniques were used as the basis for this study are highly experienced with P-EVAR. This study did not occur within their respective learning curves, and the present conclusions may not apply to individuals with more limited experience.
Conclusion
This study demonstrates that patients with obesity or severely calcified femoral arteries can be successfully treated percutaneously with fewer minor groin complications. Anatomic analysis demonstrated no deformation of the CFAs after P-EVAR.
References
- . Closure of large percutaneous access sites using the Prostar XL percutaneous vascular surgery device. J Endovasc Surg. 1999;6:168–170
- . Endovascular suture versus cutdown for endovascular repair: a prospective randomized pilot study. J Vasc Surg. 2003;38:78–82
- . Percutaneous repair of aortic aneurysms: a prospective study of suture-mediated closure devices. Eur J Vasc Endovasc Surg. 2006;32:261–265
- . Quantification of coronary artery calcium using ultrafast computed tomography. J Am Coll Cardiol. 1990;15:827–832
- . Totally percutaneous aortic aneurysm repair: experience and prudence. J Vasc Surg. 2006;43:270–276
- . Total percutaneous access for endovascular aortic aneurysm repair (“Preclose” technique). J Vasc Surg. 2007;45:1095–1101
- . Percutaneous AAA repair: is it safe?. J Endovasc Ther. 2004;11:621–626
- . Percutaneous thoracic and abdominal aortic aneurysm repair: techniques and outcomes. Vascular. 2006;14:270–281
- Percutaneous repair of abdominal aortic aneurysm. J Vasc Surg. 2004;40:12–16
- . Iatrogenic vascular injuries from percutaneous vascular suturing devices. J Vasc Surg. 2001;33:943–947
PII: S0890-5096(08)00343-9
doi:10.1016/j.avsg.2008.09.002
© 2009 Annals of Vascular Surgery Inc. Published by Elsevier Inc All rights reserved.
