Featured Article

Introduction
Renal artery stenosis (RAS) is not uncommon among patients with abdominal aortic aneurysm (AAA), and has been identified in 20–38% of patients with aortic disease studied with arteriography.^1–4 Because treatment of RAS can be performed in combination with open or endovascular repair of AAA without the need for additional exposure and/or access, the question of whether RAS should be repaired simultaneously is frequently encountered. Combined repair is considered a standard approach when symptomatic RAS (manifested by severe hypertension and/or ischemic nephropathy) and AAA warranting intervention coexist, but optimal management is less clearly defined when subcritical or asymptomatic RAS is discovered incidentally during the planning of AAA repair. The decision to undertake combined repair, therefore, must be individualized based on anatomy, operative risk, clinician experience, and potential patient benefit expected from the added intervention. This review summarizes the natural history of RAS, describes outcomes associated with combined repair of AAA and RAS, and presents suggested strategies for clinical decision making.

Natural History of Atherosclerotic Renal Artery Stenosis
The addition of renal artery intervention to primary AAA repair in the absence of severe hypertension and/or renal insufficiency has been termed “prophylactic repair.”^5,6 The conceptual basis for this approach to asymptomatic RAS is that renal revascularization provides irretrievable benefit by averting progression to symptomatic disease that would otherwise occur with medical management following AAA repair without simultaneous renal intervention. Review of the natural history of renal artery stenosis is, therefore, useful to examine the likelihood of disease progression and consider the potential benefit gained from renal artery intervention.

Selected series examining the natural history of RAS are displayed in Table 1. In a retrospective series, Tollefson and Ernst described sequential arteriogram findings from 48 patients diagnosed with RAS on abdominal aortograms performed for evaluation of aortic disease.⁷ RAS was not suspected prior to aortography in these patients, and anatomic progression was observed in 53% of vessels over a mean follow up of 4.5 years, with 9% of arteries progressing to occlusion. Rates of anatomic progression and renal artery occlusion have ranged between 11–60% and 12–16%, respectively, in both retrospective and prospective series of patients with RAS using arteriography evaluation.^4,8,9 Progression of RAS has also been characterized with serial duplex ultrasound. Zierler et al prospectively observed progression among 24% of renal arteries in hypertensive patients at 1 year,¹⁰ while Pearce et al observed a 4% incidence of progression with duplex ultrasound in a comparatively normotensive cohort.¹¹ Several authors have observed renal artery occlusion exclusively among patients with ? 60% stenosis during serial arteriogram or ultrasound surveillance.^7,10,12,13 In their prospective study using serial renal duplex scans, Caps et al identified a systolic blood pressure of 160 mm Hg, diabetes mellitus, and 60% stenosis as baseline factors associated with subsequent progression of atherosclerotic renal artery disease.¹³

Most studies have defined progression of atherosclerotic renal vascular disease (RVD) based on anatomic criteria, but the natural history of disease progression has also been described from a functional perspective. Dean et al characterized preoperative deterioration in estimated glomerular filtration rate (eGFR) among patients with RAS awaiting intervention.¹⁴ Patients in this cohort who exhibited a post-intervention reduction in the rate of functional decline exhibited greater rates of pre-intervention functional decline in contrast to patients whose eGFR did not improve following intervention. Although these findings suggest that renal artery intervention mitigates functional deterioration caused by RAS in symptomatic patients, Williamson et al observed neither a postoperative increase in serum creatinine nor reduction in survival after AAA repair without renal intervention in a cohort of 200 patients with asymptomatic RAS over a mean follow up of > 6 years.⁶

Although published data on the natural history of RVD lacks uniformity regarding the definition and incidence of disease progression, several general conclusions can be drawn from the above studies, which may inform clinical decision making when evaluating a patient with AAA and RVD. Progression of renal artery stenosis occurs in a significant minority of patients with RVD, but anatomic progression is not categorically accompanied by deterioration of renal function and/or exacerbation of hypertension. On the other hand, hypertension and/or ischemic nephropathy in the setting of hemodynamically significant RAS are presumably preceded by subcritical RAS in an asymptomatic patient that progresses anatomically. Factors that may identify patients more likely to experience progression of their RAS include hypertension, diabetes, and severity of stenosis. Renal artery occlusion during imaging surveillance appears uncommon in the setting of a low-grade (60%) stenosis.

Combined Repair of Abdominal Aortic Aneurysm and Renal Artery Stenosis
Treatment of AAA and RAS within a single procedural setting has been performed using renal artery endarterectomy, reimplantation, and/or bypass in combination with aortic reconstruction. Endovascular approaches have also been described for simultaneous management as well as adjuncts to open techniques. Also, factors affecting the decision to perform a combined repair can differ based on whether an open and endovascular approach is being undertaken.

Open repair. The addition of renal artery revascularization to open repair of infrarenal AAA increases both the duration and complexity of the operation, especially when bilateral renal artery repair is performed. Published series of combined repair have evaluated various criteria for renal revascularization, indications for aortic intervention (AAA, aortoiliac occlusive disease, thoracoabdominal and/or juxtarenal aortic aneurysms), methods of repair, proportion of patients with preoperative hypertension and/or impaired renal function, and outcomes selected for analysis.

Perioperative mortality following combined open repair of AAA and RAS has been reported in the range of 3–6.5% in most recent series.^5,15–21 Increased perioperative mortality has been observed with combined repair of AAA and RAS versus combined repair in the setting of aortoiliac occlusive disease.²² Cardiac disease is the most frequently reported cause of perioperative death,¹⁶ with other causes including bleeding, stroke, and multisystem organ failure. Clinical factors associated with perioperative mortality have varied between cohorts and have included abnormal preoperative serum creatinine, advanced age, and diabetes.^16,19 While the addition of a renal artery procedure to aortic surgery has resulted in a minimal increase in death and complication rates in several cohorts,^{5,16,17,20,23–25} a significantly higher mortality rate was observed among patients undergoing combined repair versus aortic reconstruction alone (5.3% versus 0.7%, respectively) at our center.¹⁵ This observed perioperative mortality was also greater than that observed for isolated renal artery repair (5.3% versus 1.6%), but this difference was not statistically significant. A similar increase in mortality related to renal intervention was reported by Huber et al in their analysis of operative mortality following elective infrarenal aortic reconstructions. These authors observed a nearly two-fold increase in operative mortality with the addition of a renal procedure (4.9% versus 8.9% for aortic intervention alone), but this difference was not statistically significant.²⁶ Early nonfatal complication rates following combined repair have ranged from 14–36%,^{15–20,24,25,27} with observed major complications including myocardial infarction, arrhythmia, acute deterioration in renal function, bleeding, pneumonia, respiratory failure, stroke, lower extremity atheroembolization, bowel ischemia, ileus, congestive heart failure, and infection.

Long-term survival following combined repair of AAA and RAS has been reported between 61–83% at 5 years.^{16,18,19,23,27–29} Demonstrated predictors of decreased late survival following combined repair have included diabetes and preoperative renal dysfunction.^16,19 Tsoukas et al performed a retrospective analysis of long-term survival among patients undergoing combined repair stratified based on preoperative serum creatinine and noted a significantly worse long-term survival among patients with preoperative creatinine > 2 versus < 2 (53% versus 85%, respectively).²¹ A relevant study for comparison with these findings was reported by Williamson et al,6 who examined the natural history of unrepaired, asymptomatic RAS among a cohort of 171 patients following AAA repair; over a mean follow-up > 6 years, they did not detect a significant association between high-grade RAS and long-term survival, dialysis dependence, or increase in serum creatinine. However, these authors observed significant relationships between RAS and hypertension, as well as number of antihypertensive medications required for blood pressure management. These findings seem to support a conservative strategy in the setting of asymptomatic RAS.

Evaluations of blood pressure response after renal revascularization have included a variety of criteria.^22,30 Uncontrolled preoperative hypertension has been associated with improved blood pressure response in patients undergoing combined aortic and renal repair,¹⁹ and hypertension response rates from selected series are displayed in Table 2. In general, these results are comparable to hypertension responses reported from large series of renal artery revascularization performed primarily on patients with RAS in the setting of severe hypertension and/or ischemic nephropathy without a concomitant AAA repair.^31–33 These observations support the concept of combined repair in severely hypertensive patients with RAS and AAA. The natural history of asymptomatic RAS unrepaired at the time of AAA repair has been associated with increased postoperative blood pressure and number of antihypertensive agents.⁶ However, a reduction in the postoperative incidence of hypertension through prophylactic repair in asymptomatic patients has not been demonstrated. Although Chaikof et al reported long-term freedom from hypertension among 75% of patients with asymptomatic RAS undergoing prophylactic repair during simultaneous aortic operation,⁵ Sterpetti et al observed incident hypertension among only 18% of patients with asymptomatic RAS following isolated aortic intervention.³⁴

Renal function responses to combined repair have varied widely (Table 2). At our own institution, a postoperative decline in serum creatinine 20% was observed in one third of patients with abnormal preoperative serum creatinine, while 14% of patients had a postoperative increase in serum creatinine 20%.¹⁵ In this cohort, the most favorable renal function responses were obtained in the subgroup with the worst preoperative renal function (50% response among patients with preoperative serum creatinine ? 3 mg/dL). A similar relationship between preoperative serum creatinine and renal function response to combined repair has also been observed by others,^27,35,36 and an association between renal function response and subsequent freedom from hemodialysis has also been described.²³ While patients with severe preoperative renal function impairment may have the best chance of renal function response to combined repair, they also constitute the group at greatest risk for postoperative renal failure requiring hemodialysis. Renal failure requiring hemodialysis has occurred in 2–18% of patients following combined AAA and renal artery repair and has been associated with elevated preoperative serum creatinine.^{17–19,21,24,35} Postoperative renal function decline has occurred more frequently than improvement following combined repair in some series,^17,21 and renal function decline following combined repair has been observed in approximately 9% of patients undergoing repair for hypertension in the setting of normal or mildly impaired preoperative renal function.¹⁹ Such observations suggest that prophylactic renal artery repair to prevent deterioration of renal function appears to be of uncertain value, especially given that patients with normal preoperative renal function risk postoperative functional deterioration in the absence of potential for improvement.

Several technical factors related to the conduct of renal artery revascularization with AAA repair have been evaluated for their association with mortality, as well as early and late excretory renal function outcomes. Outcomes reported following renal endarterectomy versus bypass have been similar, and selection biases favoring one approach over the other appear related to anatomy and surgical expertise. More frequent use of endarterectomy in the setting of bilateral RAS has been described (presumably for reasons related to operative time and technical preference),¹⁶ while differences in preoperative risk factors between patients undergoing endarterectomy versus bypass have also been noted within cohorts utilizing both approaches.²⁴ Increased prevalence of comorbid cardiovascular conditions and preoperative renal function impairment have been noted in patients with bilateral RAS,²⁸ and an association between perioperative mortality and bilateral versus unilateral renal artery revascularization has been described.^21,28 However, this relationship has not been observed in select cohorts defined by relatively normal preoperative renal function.^5,19 In addition to mortality, bilateral versus unilateral renal artery repair has also been associated with improved postoperative renal function in several series.^27,28,35 Allen et al also observed a significant association between postoperative renal function and use of intraoperative renal hypothermia.²⁷

Endovascular repair. Renal artery angioplasty and stenting has been utilized for treatment of RAS in combination with both open and endovascular AAA repair. Renal artery angioplasty and stenting has also been employed with fenestrated endografts, but renal artery intervention in this scenario has primarily been undertaken to prevent renal artery occlusion by the aortic endograft rather than for native renal artery occlusive disease.

Staged management of RAS with AAA using renal artery angioplasty and stenting prior to open AAA repair has been described,^37,38 citing perceived reductions in operative blood loss, length of intensive care unit and/or hospital stay, and both morbidity and mortality. Ballard et al compared patient characteristics and procedural results between patients undergoing combined open repair of AAA and RAS and patients managed with renal artery angioplasty and stenting prior to staged open AAA repair.³⁷ These authors did not observe a significant difference in operating time, blood loss, intensive care unit days, total hospital days, or functional outcomes between treatment groups. Technical failure of renal angioplasty and stenting occurred in 4 patients but did not adversely affect subsequent open renal artery revascularization. Therre et al reported technically successful renal artery angioplasty and stenting prior to open aortic operation in 15/16 patients.³⁸

Renal angioplasty and stenting, in combination with endovascular AAA repair, has also been described, either simultaneously or in a staged fashion.^39,40 Baril et al reported a series of 56 patients managed in this fashion. In relation to aortic repair, renal artery angioplasty and stenting was performed preoperatively in 66%, intraoperatively in 13%, and postoperatively in 21% of patients. All patients who underwent simultaneous combined repair in this series had flow-limiting renal artery lesions resulting from procedure-related atheroembolization and/or coverage by the aortic endograft, and there were no perioperative mortalities or major complications. These authors noted a mean postoperative decrease in creatinine clearance of 4.2 mL/min, with 12.5% of patients experiencing a decrease > 30% of baseline. Of note, 11 patients in this cohort had asymptomatic RAS.

Experience with combined endovascular treatment of RAS and AAA is currently limited. While perioperative mortality and morbidity compare favorably to combined open repair, combined endovascular repair has thus far yielded inferior renal function outcomes while hypertension response and rates of recurrent RAS are unknown. Given the potential need for nephrotoxic contrast administration both intraoperatively and during postoperative imaging follow up, the combined endovascular approach has distinct disadvantages for patients with severe preoperative renal insufficiency. Although prophylactic renal artery angioplasty and stenting of asymptomatic, high-grade RAS has been proposed as a protective measure against atheroembolism during endovascular AAA repair,40 this rationale must be questioned given recent data demonstrating atheroembolization as a routine consequence of endovascular renal artery intervention.^41,42

Furthermore, renal artery angioplasty and stenting has been performed following endovascular AAA repair with high rates of technical success, even in the setting of suprarenal endograft fixation,⁴³ suggesting that observation of asymptomatic RAS at the time of aneurysm repair does not preclude subsequent endovascular renal artery intervention if RAS progresses to symptomatic disease. We, therefore, favor conservative management of asymptomatic RAS at the time of endovascular AAA repair, combined open management of concomitant RAS and AAA when these lesions both warrant repair based on their individual merits, and combined endovascular AAA repair with renal artery angioplasty and stenting as a technique for management of complications resulting from endograft deployment, or in patients for whom open repair offers prohibitive risk.

Conclusion
Management of the patient with AAA and RAS involves consideration of renal and aortic anatomy, operative risk, and clinical manifestations of disease. While renal artery stenosis exhibits anatomic progression in a minority of patients, this progression is not uniformly associated with development of severe hypertension and/or renal function impairment and presents as renal artery occlusion in a small subset of patients, most of whom have high-grade lesions. We favor conservative management of asymptomatic RAS in the setting of AAA, and combined open repair of AAA and RAS when AAA is associated with severe renal artery stenosis and hypertension or ischemic nephropathy. Comparable results from combined repair can be obtained with either renal artery bypass or endarterectomy, and method of repair should be selected based on the anatomic pattern of disease and surgical experience. Perioperative mortality associated with combined repair is 3–6%, with hypertension responses observed in 50–70% of patients and renal function responses observed in 18–60% patients in recent series. While operative morbidity and mortality are much lower with combined endovascular repair of AAA and RAS, experience with this approach is limited, and benefits in terms of hypertension and/or renal function have not yet been demonstrated. Renal artery angioplasty and stenting prior to open surgical repair of AAA is an alternative management option for select patients with severe comorbid conditions or at centers with limited experience with open renal artery revascularization, although evidence supporting this approach over combined open repair is scant. We have, therefore, rarely utilized renal artery angioplasty and stenting prior to open aortic repair, and have also limited combined endovascular management of RAS and AAA to use as an alternative to open conversion for renal artery complications, resulting from endovascular AAA repair or in patients who are high-risk candidates for open repair.