Concurrent renal artery stent during endovascular infrarenal aortic aneurysm repair confers higher risk for 30-day acute renal failure

Abstract

Objective

Concurrent renal artery angioplasty and stenting (RAAS) during endovascular aneurysm repair (EVAR) of infrarenal abdominal aortic aneurysm (AAA) has been practiced in an attempt to maintain renal perfusion. The aim of this study was to identify the current practice of RAAS during EVAR and its effect on perioperative renal outcome.

Methods

Patients with infrarenal AAA were identified from the American College of Surgeons National Surgical Quality Improvement Program (ACS NSQIP, 2011-2014) database. Baseline characteristics of patients with concurrent RAAS during EVAR were compared with those of patients who underwent EVAR only. Bivariate and multivariable logistic regression analyses controlling for patients’ demographics, comorbidities, and operative factors were used to evaluate the predictors of 30-day acute renal failure (ARF). Sensitivity analysis was done to evaluate the role of RAAS in patients with prior kidney disease.

Results

Overall, 6183 patients underwent EVAR for infrarenal AAA during the study period. Of them, 281 patients had RAAS during EVAR (4.5%). The median age of the patients was 74 years; 81.7% of the cohort was male, but a higher proportion of female patients received EVAR + RAAS compared with patients who underwent EVAR only (26.3% vs 17.9%; P < .001). There was no difference between groups in terms of comorbidities, being on dialysis, or functional status, yet the EVAR + RAAS group had a higher proportion of patients with glomerular filtration rate <60 mL/min/1.73 m² (45.2% vs 37.2%; P = .011). RAAS was associated with significantly higher odds for development of ARF (adjusted odds ratio [aOR], 4.27; 95% confidence interval [CI], 2.06-8.84; P < .001). Other highly predictive factors of 30-day ARF were glomerular filtration rate <60 (aOR, 2.92; 95% CI, 1.47-5.78; P = .002), emergency status (aOR, 2.97; 95% CI, 1.21-7.27; P = .017), and ruptured AAA as the indication for EVAR (aOR, 4.74; 95% CI, 1.80-12.50; P = .002). Patients with prior kidney disease who had EVAR + RAAS demonstrated a 12-fold higher odds for 30-day ARF (aOR, 12.37; 95% CI, 4.66-32.89; P < .001).

Conclusions

Concurrent RAAS was found to be a significant determinant of adverse renal outcomes after EVAR for infrarenal AAA. This effect was present even after controlling for patients’ risk factors that might contribute to postoperative ARF.

Endovascular aneurysm repair (EVAR) had been increasingly adopted as the modality of choice to manage abdominal aortic aneurysms (AAAs), with >70% of AAA repair done with EVAR.^1,2 Acute renal failure (ARF) continues to be a major complication after EVAR, although to a lesser extent than after open repair. Postoperative ARF after EVAR varies widely, perhaps owing to the lack of standardized definition of it.^3,4 The reported incidence rate of ARF ranges from 2% in-hospital ARF to 18% at 1 year or more after EVAR.^1,5,6 The risk is even higher (26%) in patients with ruptured AAA who survived the first 24 hours after EVAR.⁷ Of note, the 5-year follow-up of the Dutch Randomized Endovascular Aneurysm Management (DREAM) trial estimated an average decline of estimated glomerular filtration rate (eGFR) of 0.9 (63.9) mL/min/1.73 m² per year after EVAR, and this decline was similar to what is observed after open repair.⁸

Concurrent interventions during EVAR are not uncommon. A recent study using the Vascular Study Group of New England registry showed that 29% of elective EVAR was performed with one or more concurrent interventions, and renal artery angioplasty was done in 3% of all elective EVAR.⁹

Renal artery angioplasty and stenting (RAAS) was initially used in cases of severe renal artery stenosis with refractory hypertension.¹⁰ In the context of EVAR, a chimney renal artery graft is often placed in repair of pararenal, juxtarenal, or higher aneurysms to provide an additional fixation site above the renal arteries and to prevent its migration.^11,12 However, the role of RAAS in infrarenal aneurysms and its effect on renal perfusion and subsequent perioperative kidney function are not fully described.

Renal insufficiency determined by an eGFR of <60 mL/min/1.73 m² was found to be an independent risk factor for development of AAA.¹³ Furthermore, preoperative kidney dysfunction is an established predictor of higher mortality and adverse outcomes after AAA repair.^14–16

The primary objective of this study was to evaluate the predictors of perioperative ARF and to delineate the association between concurrent RAAS during EVAR and perioperative ARF. In addition, we sought to evaluate the joint effect of prior kidney disease and RAAS on ARF after EVAR.

METHODS

The Johns Hopkins Medicine Institutional Review Board determined that this study is exempt from review because it is a secondary analysis of a deidentified registry.

Data source

The American College of Surgeons National Surgical Quality Improvement Program (ACS NSQIP) procedure-targeted Participant Use Data Files were queried to identify patients who underwent EVAR from 2011 to 2014.¹⁷ The NSQIP is a nationally validated, outcome-based program aiming to improve the quality of surgical care. Hospitals voluntarily report to the NSQIP using the patient’s medical chart rather than claims data, allowing more clinically relevant details to be captured and examined. Data entry is performed by trained surgical clinical reviewers. The sampling process is systematic, yet surgical clinical reviewers ensure random case selection from each day of the week by using an 8-day cycle system. The procedure-targeted NSQIP provides detailed information pertinent to the aneurysm and the procedural events. For example, EVAR indications, aneurysm diameter, proximal and distal extent, type of access, embolization and revascularization of hypogastric artery, concomitant lower extremity revascularization, and concomitant renal artery stenting are all readily available in the database. Of note, the indication for renal artery stenting is not reported in the NSQIP. For example, a renal stent could be added adjunctively as a chimney/snorkel configuration, as part of a fenestrated graft, or to bail out thrombosis, dissection, or inadvertent coverage or for other reasons at the surgeon’s discretion. In addition, the anatomic details of the aneurysm’s neck and access vessels are not captured in this database. We merged the procedure-targeted NSQIP with the generalized 2011-2014 NSQIP by using the case identifier variable to obtain demographics, comorbidities, and 30-day postoperative outcomes.

Patients’ characteristics

Adults who underwent EVAR for infrarenal aneurysms with no concomitant renal stenting (EVAR only) were compared with patients who had EVAR with RAAS (EVAR + RAAS). We identified infrarenal aneurysms through a variable coding the proximal extent of the aneurysm. Patients’ demographic information and comorbidities, including baseline kidney function parameters, were readily available from the database. The eGFR was calculated by the Modification of Diet in Renal Disease study equation¹⁸:

$$\text{GFR}\left(\text{ml}/min/1.73{\mathrm{m}}^2\right)=175\times {\left({\mathrm{S}}_{\text{creatinine}}\right)}^{-1.154}\times {\left(\text{Age}\right)}^{-0.203}\times \left(0.742\text{if female}\right)\times \left(1.212\text{if African American}\right)\left(\text{conventional units}\right)$$

As recommended by the National Kidney Disease Education Program, we used the eGFR cutoff of 60 (mL/min/1.73 m²) to discriminate between patients with chronic kidney dysfunction and patients with acceptable GFR.¹⁹ We generated a binary variable to describe preoperative kidney disease; any patient with GFR <60 or on dialysis at time of EVAR was considered to have suboptimal prior kidney function. In addition, we generated a combined variable to assess the additional role attributed to renal artery stenting in patients with or without prior kidney disease. Therefore, four risk groups were created: patients without prior kidney disease and no RAAS, patients without prior kidney disease who had RAAS, patients with prior kidney disease and no RAAS, and patients with prior kidney disease and with RAAS.

Postoperative outcome

Perioperative, 30-day ARF was defined as a rise in serum creatinine level of >2 mg/dL from the preoperative baseline value or the need for hemodialysis, peritoneal dialysis, hemofiltration, hemodiafiltration, or ultrafiltration in a patient who pre-operatively did not require dialysis. Furthermore, we sought to evaluate the postoperative ARF effect on mortality as a secondary outcome.

Statistical analysis

Baseline characteristics were reported in count (percentages) for categorical or binary variables. Continuous variables in this study were tested for normality using the Shapiro-Wilk test and found to be nonparametric; therefore, median and interquartile ranges (IQRs) were used to report continuous variables. Wilcoxon rank sum test, Pearson χ², and Fisher exact tests were implemented when appropriate. Bivariate and multivariable logistic regression analyses were performed to evaluate the predictors of 30-day ARF. To quantify the additional role of renal artery stenting in the risk of postoperative ARF, sensitivity analysis was implemented to account for this relationship by using the aforementioned risk group variable for renal failure. We repeated the bivariate and the multivariable logistic regression analyses using the same predictors of 30-day ARF except for GFR and renal stent. The rationale behind using this sensitivity analysis is to evaluate the interplay between prior kidney disease and renal artery stenting and to identify any synergistic effect, if present, on postoperative ARF. We then assessed the effect of postoperative ARF on mortality by employing multivariable logistic regression analysis controlling for the previously reported predictors of mortality after EVAR (patient’s demographic characteristics, comorbidities, emergency repair, aneurysm diameter, and indication of repair) in addition to postoperative ARF as a potential predictor.¹⁵ To avoid overfitting while attempting to adjust for possible confounders, Hosmer-Lemeshow test was used to assess the goodness of fit of the main model as well as of the sensitivity analysis model.²⁰ Receiver operating characteristic curves were constructed to evaluate the predictive ability of the multivariable logistic models by the C statistic calculation. Statistical significance was determined as a P value of < .05. Stata version 14.0 (StataCorp, College Station, Tex) was used to perform the analysis.

RESULTS

During the study period, we identified a total of 6183 patients who underwent endovascular repair for infrarenal aneurysms. RAAS was concomitantly used during EVAR in 281 patients (4.5%). The median age at presentation was similar in both groups (median age [IQR] in EVAR + RAAS, 74 [69-80] years; in EVAR only, 74 [68-81] years; P = .99). The majority of the cohort were male (81.7%); however, female gender constituted a higher proportion in the EVAR + RAAS group than in the EVAR only group (26.3% vs 17.9%; P < .001). Race and ethnicity did not vary between groups (both P > .05), with the majority being white (91.9%) or from a non-Hispanic ethnic background (98.2%). Both groups had proportionally similar prevalence of diabetes, hypertension, history of congestive heart failure, history of chronic obstructive pulmonary diseases, smoking status, and obesity (P > .05; Table I).

Table I.

Baseline characteristics and postoperative complications of patients undergoing endovascular aneurysm repair (EVAR) alone vs patients undergoing EVAR and concomitant renal artery angioplasty and stenting (RAAS)

Variable	EVAR only (n = 5902; 95.5%)	EVAR + RAAS (n = 281; 4.5%)	P value
Demographics
Age, years	74 (68-81)	74 (69-80)	.99
Female gender	1057 (17.9)	74 (26.3)	<.001
Race			.19
White	4911 (92.0)	231 (88.9)
Black	292 (5.5)	20 (7.7)
Others	134 (2.5)	9 (3.5)
Ethnicity	95 (1.8)	5 (1.9)
Comorbidities
Current smoking	1801 (30.5)	95 (33.8)	.24
BMI ≥30	1850 (32.2)	79 (29.0)	.27
History of COPD	1055 (17.9)	58 (20.6)	.24
DM	937 (15.9)	50 (17.8)	.39
HTN	4695 (79.6)	232 (82.6)	.22
CHF in 30 days before surgery	93 (1.6)	7 (2.5)	.24
Currently on dialysis (preoperatively)	71 (1.2)	2 (0.7)	.77
ASA classification ≥3	5556 (94.3)	269 (95.7)	.300
Prior abdominal aortic surgery	1343 (24.6)	70 (26.4)	.30
Dependent vs independent functional health status	183 (3.1)	10 (3.6)	.49
GFR	67.6 (52.7-83.5)	63.1 (49.0-81.8)	.67
GFR <60	1934 (37.2)	113 (45.2)	.016
Preoperative creatinine value	1.03 (0.87-1.30)	1.09 (0.90-1.30)	.011
Preoperative BUN value	18 (14-23)	19 (14-23.5)	.17
Procedural characteristics
Emergency status	593 (10.1)	34 (12.1)	.26
Transfusion >4 units of packed RBCs in 72 hours before surgery	136 (2.3)	8 (2.9)	.56
Operation time, minutes	128 (99-169)	171 (124-237)	<.001
Aneurysm diameter, cm	5.5 (5.1-6.2)	5.6 (5.1-6.5)	.07
Indication of EVAR			.030
Asymptomatic	4702 (80.7)	215 (77.3)
Nonruptured symptomatic	392 (6.7)	14 (5.0)
Redo	166 (2.9)	16 (5.8)
Rupture	384 (6.6)	24 (8.6)
Others	184 (3.2)	9 (3.2)
Distal aneurysm extent			.002
Aortic	2419 (48.2)	100 (41.7)
Common iliac	1947 (38.8)	93 (38.8)
Internal iliac	375 (7.5)	20 (8.3)
External iliac	281 (5.6)	27 (11.3)
Complications
ICU admission	2019 (34.5)	122 (43.9)	.001
Postoperative ARF	91 (1.5)	11 (3.9)	.002
Other concomitant procedures
Hypogastric artery embolization	403 (6.8)	36 (12.8)	<.001
Hypogastric artery revascularization	261 (4.4)	41 (14.6)	<.001
Lower extremity revascularization	235 (4.4)	15 (5.8)	.29

ARF, Acute renal failure; ASA, American Society of Anesthesiologists; BMI, body mass index; BUN, blood urea nitrogen; CHF, congestive heart failure; COPD, chronic obstructive pulmonary disease; DM, diabetes mellitus; GFR, glomerular filtration rate; HTN, hypertension; ICU, intensive care unit; RBCs, red blood cells.

Categorical variables are presented as number (%). Continuous variables are presented as median (interquartile range). Boldface values indicate statistical significance (P < .05).

Preoperative blood urea and creatinine levels did not differ between groups; however, a higher percentage of patients with GFR <60 was found in the group that had concomitant RAAS compared with patients who underwent EVAR only (45.2% vs 37.2%; P .011). Median GFR was lower in the EVAR + RAAS=group (median GFR [IQR], 63.0 [49.0-81.8] vs 67.6 [52.7-83.5]; P = .016). Overall, 73 patients (1.2%) were on dialysis before surgery, with similar distribution in both groups (P = .77).

Patients who had concomitant RAAS (EVAR + RAAS) were more likely to have distal extension of the aneurysms beyond the distal aorta into the common iliac arteries compared with patients with EVAR only (distal extension to the aorta, 41.7% vs 48.2%; common iliac, 38.8% vs 38.8%; internal iliac, 8.3% vs 7.5%; external iliac, 11.3% vs 5.6%; P = .002). The median operation time was 43 minutes longer when a renal stent was introduced (median operation time [IQR]: EVAR only, 128 [99-169] minutes; EVAR + RAAS, 171 [124-237] minutes; P < .001). However, patients undergoing EVAR + RAAS did not require more transfusion. Concurrent hypogastric artery interventions occurred more often in patients with EVAR + RAAS than in patients with EVAR only, but concurrent lower extremity revascularization was similar in both groups (Table I). Overall, perioperative ARF occurred in 102 patients; 74 (72%) of them required de novo renal replacement therapy within 30 days after EVAR. Perioperative ARF was more common in the group that had RAAS (3.9% vs 1.5%; P .002). EVAR + RAAS patients were more likely to be admitted to the intensive care unit compared with patients who underwent EVAR only (43.9% vs 34.5%; P = .001).

Bivariate logistic regression analysis yielded a significant association between RAAS and postoperative ARF (odds ratio [OR], 2.60; 95% confidence interval [CI], 1.38-4.92; P = .003). GFR <60 was associated with a sixfold increase in the odds for postoperative ARF (OR, 6.73; 95% CI, 3.94-11.49; P < .001). Similarly, an increase in blood urea level, a history of congestive heart failure, being totally dependent, emergency status, redo repair, and ruptured aneurysm repair were associated with higher odds for postoperative ARF (all P < .05; Table II).

Table II.

Bivariate and multivariate logistic regression analyses for the odds of 30-day postoperative acute renal failure (ARF)

Variables	Bivariate logistic regression analysis		Multivariate logistic regression analysis
	OR (95% CI)	P value	OR (95% CI)	P value
Renal stenting	2.60 (1.38-4.92)	.003	4.27 (2.06-8.84)	<.001
Age	1.04 (1.02-1.06)	.001	1.01 (0.98-1.05)	.42
Male vs female gender	0.65 (0.41-1.02)	.06	0.71 (0.39-1.31)	.27
Race
White	Reference		Reference
Black	0.83 (0.30-2.29)	.72	0.86 (0.27-2.72)	.80
Other races	1.84 (0.67-5.11)	.24	1.04 (0.32-3.38)	.94
Hypertension	1.49 (0.86-2.58)	.16	1.72 (0.74-3.99)	.21
Diabetes	1.29 (0.79-2.11)	.31	1.29 (0.68-2.44)	.43
BMI ≥30	0.87 (0.56-1.35)	.54	0.78 (0.43-1.41)	.41
GFR <60	6.73 (3.94-11.49)	<.001	2.92 (1.47-5.78)	.002
History of COPD	1.33 (0.83-2.13)	.23	0.98 (0.51-1.88)	.95
History of CHF	3.25 (1.29-8.16)	.012	1.77 (0.56-5.61)	.33
Blood urea, mg/dL	1.05 (1.04-1.06)	<.001	1.04 (1.02-1.06)	<.001
Functional status
Independent	Reference		Reference
Partially dependent	1.49 (0.54-4.10)	.44	0.17 (0.02-1.68)	.13
Totally dependent	5.89 (1.36-25.48)	.018	1.09 (0.10-12.08)	.94
Emergency status vs elective	10.81 (7.26-16.10)	<.001	2.97 (1.21-7.27)	.017
ASA classification ≥3	1.46 (0.54-4.01)	.46	0.43 (0.13-1.46)	.18
Diameter, cm	1.41 (1.29-1.55)	<.001	1.17 (1.03-1.34)	.019
Indication
Asymptomatic	Reference		Reference
Nonruptured symptomatic	1.28 (0.45-3.60)	.64	0.75 (0.21-2.74)	.66
Redo	5.14 (2.26-11.66)	<.001	1.49 (0.41-5.46)	.55
Rupture	16.32 (10.48-25.40)	<.001	4.74 (1.80-12.51)	.002
Others	4.83 (2.13-10.96)	<.001	2.39 (0.66-8.63)	.19

ASA, American Society of Anesthesiologists; BMI, body mass index; CHF, congestive heart failure; CI, confidence interval; COPD, chronic obstructive pulmonary disease; GFR, glomerular filtration rate; OR, odds ratio.

Boldface values indicate statistical significance (P < .05).

Multivariable logistic regression analysis adjusting for all potential confounders showed a stronger association between renal artery stenting and postoperative ARF (OR, 4.27; 95% CI, 2.06-8.83; P < .001). GFR <60 increased the odds of postoperative ARF by nearly threefold (OR, 2.92; 95% CI, 1.47-5.78; P = .002). For each 1-cm increase in aneurysm diameter, there is an associated 17% increased risk for postoperative ARF (OR, 1.17; 95% CI, 1.03-1.34; P = .019). Emergent EVAR confers higher risk for postoperative ARF compared with elective cases (OR, 2.97; 95% CI, 1.21-7.27; P = .017; Table II).

Fig shows that concomitant RAAS has additional risk attributed to the use of RAAS in patients with prior kidney disease compared with patients with normal baseline kidney function without RAAS, patients who had normal kidney but had RAAS, patients with prior kidney disease without RAAS, and patients who had normal kidney but had RAAS (P < .001). Furthermore, the sensitivity analysis using those risk categories demonst rated a vast increase in the adjusted OR of 30-day ARF in patients with prior kidney disease who had EVAR + RAAS (OR, 12.37; 95% CI, 4.66-32.89; P < .001; Table III). Patients with prior kidney disease and no RAAS were not different from patients with prior kidney disease who had RAAS in terms of median eGFR (median eGFR [IQR]: 48.3 [38.7-54.4] vs 48.0 [39.4-55.2], respectively; P = .85). The risk of mortality was 22 times higher if the patient had postoperative ARF (adjusted OR, 22.30; 95% CI, 11.93-41.69; P < .001). C statistic of all multivariable logistic regression models was 0.89. Hosmer-Lemeshow test for all models proved good fitness (all P > .05).

Fig.

Proportion of patients who developed acute renal failure (ARF) postoperatively in each risk group. RAAS, Renal artery angioplasty and stenting.

Table III.

Sensitivity analysis using patients’ risk categories based on the absence or presence of prior kidney disease or renal artery angioplasty and stenting (RAAS)

Risk categories	Bivariate logistic regression analysis		Multivariate logistic regression analysisa
	OR (95% CI)	P value	OR (95% CI)	P value
Normal kidney function + no RAAS	Reference		Reference
Prior kidney disease + no RAAS	6.63 (3.75-11.73)	<.001	2.95 (1.42-6.16)	.004
Normal kidney function + RAAS	3.20 (0.72-14.14)	.13	4.60 (0.98-21.65)	.05
Prior kidney disease + RAAS	18.70 (8.00-43.72)	<.001	12.37 (4.66-32.89)	<.001

CI, Confidence interval; OR, odds ratio.

Bivariate and multivariate logistic regression analyses are shown for the odds of 30-day postoperative acute renal failure (ARF) based on risk groups. Boldface values indicate statistical significance (P < .05).

^aAdjusted for the same variables as in Table II, excluding glomerular filtration rate (GFR) and renal stent.

DISCUSSION

Our findings suggest that when a renal artery stent is introduced during EVAR for infrarenal AAA, the risk of 30-day ARF increases nearly fourfold. This effect was independent of baseline renal function and the patient’s characteristics and comorbid conditions. This result is similar to the reported odds of in-hospital renal dysfunction after elective EVAR in the study by Ultee et al, in which concomitant renal angioplasty or stenting was found to be an independent risk factor for both 30-day mortality and in-hospital renal dysfunction.⁹ This adverse role had also been observed in a prior study by Protack et al, in which the investigators demonstrated a significant rise in creatinine concentration within 1 month of intervention in patients who had EVAR + RAAS compared with patients who had RAAS for other indications.²¹ That study had an objective slightly different from ours as the authors compared two distinct populations of patients, although they shared the same atherosclerotic pathologic process; patients with AAA and patients with renal artery stenosis. The observed difference in renal outcomes could in part be attributed to the difference in the gravity of the patient’s general condition. However, there are conflicting studies that suggested the safety of RAAS during EVAR.^22,23 Yet, those studies were reflecting single-institutional experience with relatively small sample size and might be underpowered to detect a significant difference.

Other factors that were found to be associated with higher perioperative ARF risks were preoperative GFR level <60 and blood urea level. This is in line with previous studies that showed an association between preoperative renal impairment and the risk of postoperative renal dysfunction.^1,23,24 Therefore, we sought to specifically examine the risk added by RAAS in patients with prior kidney disease. Compared with patients of normal baseline kidney function who did not undergo concurrent RAAS, the risk of 30-day ARF in patients with prior kidney disease is 3-fold higher, and in patients with prior kidney disease and RAAS, there is a 12-fold increase. This indicated that RAAS contributed to the risk of 30-day ARF on a multiplicative scale. To our knowledge, this is the first study to report and to measure the association between RAAS and kidney dysfunction before EVAR. This interaction could be explained by the fact that patients with dysfunctional kidneys to start with might be particularly sensitive to renal artery manipulation and subsequent shower embolization into the parenchyma. Others hypothesized that fixation of the proximal end of the renal artery during EVAR could be compensated for by the mobility of the distal end; however, this mobility could be limited if a stent were placed.²⁵ On the other hand, Mehta et al found no effect of fixation level on postoperative renal failure.²⁴

Emergency repair increased the odds of perioperative ARF nearly threefold, whereas repair for ruptured aneurysms was associated with an almost fivefold increased risk. This can be attributed to kidney hypoperfusion in the case of ruptured AAA and lack of adequate hydration before urgent EVAR, which has been proved to be an effective measure in preventing contrast-induced nephropathy.²⁶ Aneurysm diameter was also a significant predictor of 30-day ARF, which might reflect a challenging anatomy of the aneurysm that could precipitate the overall procedural risk, operation time, and exposure to contrast material.

Another interesting finding was the association between female gender and the prevalence of concomitant renal artery stenting (26.3% in EVAR + RAAS vs 17.9% in EVAR only; P < .001). This could indicate more difficult anatomy in female patients, a finding that has been reported before.^27,28

Complex aneurysm anatomy can potentially limit the use of EVAR or require multiple additional interventions. Numerous techniques and devices have been introduced to repair aneurysms with complex anatomy that had once been deemed unsuitable for the endovascular approach. Those endografts had been designed to accommodate various aneurysmal neck morphologies and different proximal extents.²⁹ For example, the chimney (chEVAR) or snorkel configuration is used to optimize the sealing of endografts in cases in which the aneurysm’s neck is short.³⁰ Dias et al stated that renal chimney grafts increased the suitability for EVAR from 34% to 46% in patients with ruptured AAA in which the aneurysm neck was <15 mm in length.³¹ Several reports have shown that although chEVAR might overcome the anatomic risk, it is associated with higher perioperative renal dysfunction.^32–34 The chEVAR is primarily used in patients with juxtarenal or higher AAAs or when the neck of an infrarenal aneurysm is considered hostile. Thus, to make a fair comparison, we sought to examine only infrarenal AAAs in an attempt to make the anatomic risks as homogeneous as possible. Yet unfortunately, the NSQIP does not provide neck thrombus, calcification, angulation, length, or diameter or specify the maneuver by which the repair took place.

Outside the context of EVAR, RAAS is often used in the management of refractory hypertension that is caused by severe renal artery stenosis.³⁵ It showed higher efficacy in lowering blood pressure yet no added advantage in terms of reversing or halting the progression of renal disease.³⁵ However, with advanced medical therapies, the role of RAAS has diminished, and it is no longer superior to medical management.³⁶ In addition, there is certainly no role for stenting in incidentally identified renal artery stenosis.^37,38 With the angiographic imaging obtained during EVAR, we hypothesize that surgeons might accidentally identify significant renal artery stenosis and might attempt to correct it with stenting for fear of making it worse with the stent graft. It is also possible that misdeployment of the stent graft and accidental coverage of the renal artery could be the indication for some of these cases. This is more likely to occur in cases with difficult anatomy.

Finally, our study might underestimate the actual number of patients with renal impairment after EVAR. The authors are limited by the NSQIP definition of renal insufficiency, a rise of 2 mg/dL in serum creatinine concentration, which is well above what is considered the first stage of the Aneurysm Renal Injury Score. Moreover, the need for dialysis after EVAR in patients who previously did not require dialysis (72.5% of our ARF cohort) immediately places the patient at stage 4.³ Starting renal replacement therapy in a patient who previously did not require it is a devastating complication to the patients as well as to the surgeon. We have demonstrated that there is a 22-fold increase in the odds of mortality when postoperative ARF develops (OR, 22.30; 95% CI, 11.93-41.69; P < .001), which confirms results from prior studies.^7,39 Furthermore, dialysis and the extended length of stay in the hospital due to postoperative ARF add to the financial burden of EVAR on the health care system because hemodialysis is fully covered by Medicare services.

This study has several limitations that arise from its retrospective design. Selection bias is of concern in comparing two interventions based on nonrandomized observational studies. Large databases are particularly prone to inaccuracies and missing information. Data entry in the NSQIP is done by nonphysicians; however, validation studies have proved the adequate performance of the NSQIP compared with medical records.⁴⁰ Despite the availability of details pertinent to EVAR, contrast material volume is not reported in the NSQIP; the indication for renal stent introduction and details about repair configuration are not captured as well. Therefore, we were not able to adjust for the risk added by those essential factors. Another limitation to this study is the lack of aneurysm neck anatomic details. The indications for renal stent use during EVAR is not assessed in the NSQIP. These indications could explain the benefit or harm added by RAAS; thus, further research is warranted to better assess the role of RAAS.

CONCLUSIONS

Renal artery stenting confers higher risk of adverse renal outcomes after EVAR for infrarenal aortic aneurysms. This effect was independent of preoperative GFR and other comorbidities that might clinically explain deteriorating renal function. Furthermore, to place a patient on dialysis who was dialysis independent before EVAR is a devastating event that has great negative impact on the patient and the health system. We believe that manipulating the renal arteries might compromise the renal perfusion, particularly in patients with preoperative chronic kidney disease, in whom the effect of RAAS has been shown to be synergistic. We acknowledge that the decision to place a renal artery stent is dictated by factors peculiar to the patient’s anatomy and the surgeon’s best judgment; however, we strongly recommend extreme caution when intervening on renal arteries during EVAR, particularly with the lack of clear guidelines to delineate the appropriate indications. If a renal stent is ever introduced, we recommend scrutiny of the patient’s renal function by blood and serum measures. Duplex ultrasound is also helpful in evaluating renal stent patency as well as sac diameter and endoleak without the added risk of contrast-enhanced computed tomography angiography on kidney function. Further research is warranted to elucidate the biologic mechanism by which concurrent RAAS adversely affects renal function and to establish evidence-based practice guidelines to clarify the benefits and associated risks of concurrent RAAS in the setting of EVAR.

HIGHLIGHTS.

• Type of Research: Retrospective analysis of prospectively collected data of the American College of Surgeons National Surgical Quality Improvement Program (ACS NSQIP) registry

• Take Home Message: Renal artery stenting done concomitantly with 281 EVARs was associated with an almost 4-fold increase in the risk of acute kidney injury, and those with prior kidney disease had a 12-fold higher odds.

• Recommendation: The authors suggest against concomitant renal artery stenting in patients with otherwise uncomplicated EVAR.