Revised duplex criteria and outcomes for renal stents and stent grafts following endovascular repair of juxtarenal and thoracoabdominal aneurysms
Article Outline
- Abstract
- Materials and methods
- Results
- Discussion
- Author contributions
- Supplementary data
- References
- Copyright
Objectives
To assess outcomes and develop duplex scan criteria that will reliably determine the luminal status of covered and uncovered renal stents following fenestrated and branched endovascular repair.
Methods
A prospective database of patients treated with fenestrated and branched endografts between 2001 and 2006 was reviewed. All patients with evidence of renal artery pathology including duplex scan assessed peak systolic velocity (PSV) <50 or >200 cm/s, renal aortic ratio (RAR) >3.5, elevation of the serum creatinine >30%, computed tomography (CT) evidence of renal stenosis underwent further analyses including medical chart review, and a review of CT and duplex scan imaging data. Correlations of ultrasound scan, CT, angiographic, and clinical outcomes were conducted and receiver operator curve (ROC) analysis was performed. Freedom from stenosis or occlusion was determined by Kaplan-Meier analysis with differences assessed by log rank tests.
Results
A total of 518 renal arteries were treated with uncovered or covered renal stents (287 patients). Mean follow-up was 25 months. The estimated freedom from stenosis at 12, 24, and 36 months were 95% (95% confidence interval [CI] 93-98), 92% (89-96), and 89% (85-93) for uncovered stents, and 98% (96-100), 97% (95-100), and 95% (91-100) for covered stents (log rank P = .04). Secondary interventions were performed in 20% of the patients who developed stenoses. Only one of the detected stenoses that was not treated with a secondary intervention progressed to occlusion. Duplex scan criteria derived from ROC analysis correlating with curved planar reconstruction (CPR) from axial imaging data calculated a 60-99% in-stent stenosis to be associated with a PSV >280 cm/s or RAR >4.5. Occlusions were best identified by a mid renal artery PSV <57 cm/s in conjunction with an RAR <1.2.
Conclusion
Revised ultrasound scan criteria have been developed to improve the sensitivity and specificity of non-invasive interrogation of renal stents following endovascular aneurysm repair (EVAR). Covered renal stents are associated with a lower incidence of in-stent stenosis and are thus recommended over uncovered stents for use in fenestrated or branched endografts.
Renal stenting has been advocated as a method to treat renal arterial disease in the setting of deteriorating renal function or poorly controlled renovascular hypertension.1, 2, 3, 4 Balloon-expandable stents are primarily utilized to treat ostial lesions, while angioplasty alone has been advocated for more distal lesions.5 Recently, pure endovascular treatment of juxtarenal and thoracoabdominal aneurysms has been described.6, 7, 8, 9, 10 Specific renal outcomes have been previously reported with an initial series of patients,11 yet scientifically-derived criteria for assessing renal artery pathology following simple renal stenting or fenestrated/branched grafting is absent. The choice of stent type, device sizing, and the use of covered stents is similarly founded by bias rather than data. Some groups have postulated higher rates of in-stent thrombosis with the bulkiness associated with covered stents12 whilst others consider the potential for neointimal hyperplasia between stent struts to support the use of drug-coated stents. No studies exist where multiple imaging modalities are used to interrogate the study population through extended follow-up periods.
Duplex ultrasonography scan is considered the primary test utilized to assess for de novo renal stenoses, and criteria in the absence of intervention have been suggested.13 The techniques used to preserve renal flow in juxtarenal aneurysms, and bridge aortic graft material to the renal arteries in thoracoabdominal aneurysms, most commonly involves the use of renal stents. In a manner akin to renal stenting for renal ostial stenoses, stents are placed into the main renal trunk and intended to extend 1-3 mm into the aorta. Following renal stent implantation or endografting in conjunction with renal stenting, ultrasonographic criteria used to indicate stenoses within or remote from the stented artery are poorly defined. Some groups have continued to use criteria applied to unstented renal arteries14, 15 (peak systolic velocity [PSV] >200 cm/s or renal aortic ratio [RAR] >3.5), while others have suggested modified criteria.16, 17 None have suggested revised criteria to be used in the setting of an aortic endovascular graft. The protrusion of a renal stent into the aorta results in turbulence, typically with higher PSVs; while the loss of vascular compliance imposed by an endovascular graft alters aortic velocities. The combination of these two factors has the potential to markedly distort the renal to aortic ratio, resulting in confusion regarding the status of a renal artery after treatment. Therefore, in the absence of follow-up imaging studies, supplementing duplex ultrasound scan data are of little use without a method of validating the duplex scan results. The intention of this study was to assess the methods by which the covered and uncovered renal stents used in conjunction with an aortic graft are assessed, and to evaluate the outcomes in the context of clinical relevance.
Materials and methods
Between August 2001 and December 2006, 295 patients underwent endovascular repair of juxtarenal (166) or thoracoabdominal (129) aortic aneurysms with fenestrated or branched devices. Procedures were performed under a sponsored physician investigational device exemption study. All patients signed an informed consent form that had been approved by our institutional review board. Patients were considered high risk for open surgery, based upon physiologic and anatomic characteristics.8 The techniques used to implant these prostheses have also been previously described.8 The presence of a renal stenosis prior to aneurysm treatment was not a contraindication for treatment, and all patients underwent renal angiography at the time of the aneurysm repair. Patients were included in this analysis only if follow-up duplex scan and CT data were available (Table I), they were not on dialysis prior to the aneurysm repair, and the device utilized required the stenting of one or more renal arteries. A total of 8 patients were excluded from the analysis because no renal arteries were involved in their repair, or directional branches for the renal arteries were utilized. The procedural records and imaging studies were reviewed for the remaining 287 patients. Analysis was performed per renal artery treated for a total of 518 renal arteries.
Table I. Patient accountability through 24 months
| Number of patients (%) | ||||
|---|---|---|---|---|
| Eligible for visit | Followed | CT | US | |
| 1 years | 235 | 205 (87) | 193(82) | 169(72) |
| 2 years | 126 | 107(85) | 107(85) | 101(80) |
Techniques
Our experience began with treatment of juxtarenal aortic aneurysms with fenestrated endografts using uncovered balloon-expandable renal stents in 2001.8 Covered stents were later employed in the setting of thoracoabdominal aneurysms and selectively in juxtarenal aneurysms.18 Both JOMED (Abbott Vascular, Redwood City, Calif) and iCast (Atrium Medical Corp, Hudson, NH) have been used. Since 2006, covered stents have been used for both thoracoabdominal as well as juxtarenal aneurysms. The stresses and angulations imposed on the renal stents may differ between some thoracoabdominal aneurysms and juxtarenal aneurysms.
In brief, access from the contralateral femoral artery is utilized to cannulate fenestrations from within the partially deployed prosthesis. Guiding sheaths are placed within the fenestrations and the proximal fixation system is completely released.
The renal arteries are imaged through the guiding sheaths and a balloon catheter is selected to match the renal artery diameter. The stent is then delivered on the balloon and positioned such that the majority of the stent is within the renal artery and 2-4 mm project into the aorta. The aortic portion of the stent is then flared with a 10-12 mm angioplasty balloon, allowing it to function as a rivet, apposing the aortic graft material against the renal orifice. The remainder of the procedure is carried out in a manner similar to a conventional bifurcated Zenith (Cook Inc, Bloomington, Ind) abdominal aortic aneurysm (AAA) endograft.
Technical success was defined as the placement of a patent endoprosthesis with all vessels initially intended to be incorporated into the desired fenestration remaining patent, in addition to the absence of a type I or III endoleak at the completion angiogram.
Evidence of renal arterial issues
A conglomerate endpoint was established intended to capture all renal stenoses or occlusion during follow-up. To accomplish this, given the lack of appropriate guidelines to identify patients with renal artery stenosis or occlusion following aortic repair, the criteria were intentionally broad in an effort to include all patients with potential renal artery compromise. All renal ultrasound scans were classified in one of six ways: normal, 60-99% renal artery stenosis (PSV >200 cm/s or RAR >3.5), >80% renal artery stenosis (end-diastolic velocity [EDV] >140 cm/s), potential renal artery stenosis (any value of a PSV >200 cm/sec or <50 sm/sec), renal artery occlusion or other (no flow, not visualized, not determined). Additionally, patients with any renal stent stenosis, occlusion, or fracture noted on any follow-up CT scan were selected for further review. Hence all patients with an identifiable anatomic defect were included in this review.
Clinical events were considered to have occurred under the following circumstances: sustained decrease in the calculated Glomerular Filtration Rate (GFR) (using the Modification of Diet in Renal Disease [MDRD] formula,19 Fig 1) of 30% or more, a significant decrease in renal size (>10 mm reduction in renal length), any renal-directed secondary intervention, temporary or permanent dialysis, new-onset or worsening of hypertension. All patients with any of the above clinical events were also selected for further review.
Fig 1.
Modification of Diet in Renal Disease (MDRD) formula. Levey AS, Coresh J, Balk E, Kausz AT, Levin A, Steffes MW, et al. National Kidney Foundation practice guidelines for chronic kidney disease: evaluation, classification, and stratification. Ann Intern Med 2003;139(2):137-47. GFR, Glomerular Filtration Rate; Pcr, polymerase chain reaction.
CT scans were reconstructed by using thin-slice algorithms (with z-plane resolution typically between 0.75 mm and 1 mm) and were assessed by multiplanar reconstruction (MPR), three-dimensional (3D), and centerline of flow (CLF) techniques using a 3D workstation (Aquarius, Terarecon, San Mateo, Calif). Following the creation of a centerline through each renal artery, a curved planar reconstruction (CPR) image was generated for each renal artery that allowed optimal visualization of each treated vessel. CPR is a technique for reformatting axial CT data using vessel-centering algorithms, to display the entire length of a tubular structure within a single image and has been previously validated.20 Renal size was measured on 3D reconstructions by taking the greatest cranial to caudal measurement from the upper pole to the lower pole. Renal stenoses were considered relevant on CT image processing if any of the following were observed: a luminal stenosis of >50% was visible on the CPR; renal stent kinked or crushed on stent template reconstructions; lack of renal parenchymal enhancement on 3D reconstructions associated with a reduction in renal size suggested stent occlusion; lack of luminal opacification on axial, MPR, and CPRs of post-contrast scans confirmed renal stent occlusion. Following identification of potential renal arterial compromise from any of the aforementioned criteria (duplex scan, clinical, or CT), a detailed analysis of the renal data available was conducted for each patient. All patient factors were charted in a longitudinal manner.
Curve planar reconstructions (CPR) of the renal artery correlated well with both PSV and RAR. Additional correlation was noted as during the longitudinal follow-up of patients with renal lesions with or without secondary interventions. Binary values were ascribed to specific renal arteries as stenotic (>50% luminal narrowing) or non-stenotic (luminal narrowing <50%) based on CPR assessment. Then, based upon the constructed ROC curves, we determined the optimum point on each ROC curve to assess for a threshold value for PSV, EDV, and RAR that had the highest combination of sensitivity and specificity.
Follow-up protocols
Throughout the study period, patients have been maintained on single antiplatelet therapy with aspirin (acetylsalicylic acid [ASA]), unless other indications for more aggressive antiplatelet or anticoagulant were present. Follow-up studies were conducted at discharge, 1 month, 6 months, 12 months, and yearly thereafter and included routine laboratory studies, renal and mesenteric duplex scans, and CT scans. Secondary interventions were performed for renal stent stenoses that were associated with a deterioration of hypertension control, deterioration in renal function, progressive on serial follow-up, or considered high grade on duplex ultrasound scans and computed tomographic angiography (CTA) reconstructions.
Statistical analysis
Data were stored in a database using Oracle Clinical software S-Plus 7.0 (Insightful, Wash) and NCSS 2007 (NCSS, Kaysville, Utah) were used for all statistical analyses. Categorical variables were summarized as number and percentage, continuous variables as mean and standard deviation (SD). Comparisons between categorical and continuous outcomes were performed with the χ2 test and t test, respectively. Freedom from stenosis, freedom from occlusion and mortality were evaluated with life table analysis using the Kaplan-Meier method. Differences in outcomes between renal arteries with covered and uncovered stents were evaluated with the log rank test. The outcomes of renal arteries with covered stents vs uncovered stents were compared using Cox proportional hazards models, and strength with the hazard ratio (HR) and 95% confidence intervals (CI).
We further evaluated the discriminative ability of PSV, EDV, and RAR for the diagnosis of stenosis and occlusion by CT scan by using ROC curve analysis. The c-statistic, equivalent to the area under the ROC curve and a measure of discrimination, ranges from 0.5 (no discrimination) to 1 (perfect discrimination). An ROC curve was built for PSV, EDV, and RAR, and the best combination of sensitivity and specificity is given by the point in the curve that is closer to the top left angle. These combinations are related to values of PSV, EDV, and RAR, which were regarded as the best diagnostic thresholds. The respective positive predictive values (PPV) and negative predictive values (NPV) for each threshold were calculated.
Results
A total of 518 renal arteries were treated with uncovered stents (287) or covered stents (231) in 287 patients. The mean follow-up was 26 months (33 and 15 months, respectively, for uncovered and covered stents) with a range of 9 months to 6 years. The baseline characteristics of the two groups are presented in Table II. All patients who had uncovered stents placed had a juxtarenal aneurysm. Of the patients who had a covered stent placed, 49% had a juxtarenal, and 51% had a thoracoabdominal aneurysm. Technical success was achieved in 98.1% of cases. Successful access into all but three of the 518 targeted renal arteries was achieved at the time of the initial procedure. The 30-day mortality rate was 1.9% for juxtarenal aneurysms and 6.9% for thoracoabdominal aneurysms (types I-IV) and 4.2% for the entire group.
Table II. Baseline characteristics comparing patients treated with covered renal artery stents to those with uncovered renal artery stents
| Uncovered stents | Covered stents | |
|---|---|---|
| Number of patients | 158 | 129 |
| Number of renal stents | 287 | 231 |
| Age at inclusion (years) | 75.7*±7.1 | 75.8*±7.9 |
| Gender/male, n (%) | 132(84%) | 104(81%) |
| Systolic BP (mm Hg) | 135*±22 | 130*±21 |
| Diastolic BP (mm Hg) | 74*±10 | 72*±10 |
| Heart rate (bpm) | 71*±11 | 72*±13 |
| Diabetes, n (%) | 27(17%) | 19(15%) |
| Current smoker, n (%) | 26(18%) | 14(18%) |
| Renal status (Cr) | 1.2*±0.4 | 1.4*±0.8 |
| Hemoglobin(g/dL) | 13.5*±1.8 | 13.6*±1.7 |
Renal stent occlusion
A total of 13 of 287 (4.5%) uncovered renal stents occluded on follow-up imaging. The mean time to occlusion was 3.7 months (SD 3.3, range, 0-9). There were 6 patients who underwent angiography and 2 of these occlusions were recanalized and remain patent to date. The mean stent diameter for this group was 6.5 mm (SD .85, range, 5-7) which did not differ from patients without occlusions. Of the 13 patients with renal stent occlusions, 12 had a >30% deterioration in GFR at the time of occlusion. This drop in GFR was sustained in only seven of these cases and the other patients recovered GFR.
A total of five of 231 (2.2%) covered renal stents occluded on follow-up imaging. The mean time to occlusion was 3.2 months (SD 2.5, range, 2-7). Two patients with occlusions underwent angiography but neither was successfully recanalized. The mean stent diameter for this group was 6 mm (SD 1.2, range, 4-7) which also did not differ from the patent stent group diameter. Four out of five had a >30% deterioration in GFR at the time of occlusion, all of which have been sustained throughout the follow-up period. Freedom from renal stent occlusion in the two groups is presented in Fig 2.
Fig 2.
Kaplan-Meier curves for freedom from occlusion in months. The solid line represents those treated with a covered stent and the dashed line represents those treated with an uncovered stent. The development of renal stent occlusion between the two groups was not statistically different. Hazard ratio (HR) for occlusion (covered vs uncovered): 0.5 (0.2-1.4), P = 0.2.
Timing and etiology of renal stent occlusions
Most (11 of the 18) renal stent occlusions occurred prior to the 1-month CT scan. The remaining seven occlusions occurred between the 1-month and 6-month studies. There were no late renal stent occlusions.
Of the 11 patients in whom renal stent occlusion was apparent at the first postoperative visit, 4 were related to procedural dissections, 3 had multiple overlapping stents placed into a tapered luminal diameter renal artery, 3 had renal stents that were compressed by angulation between the fenestration and renal artery origin, and in 1 patient no etiology was discovered.
Of the 7 patients in whom renal stent occlusions developed between the 1-month and 6-month visits, five occurred on the right side and two on the left side. Three patients had kinks visible between the right renal stent and the artery distal to the stent, 2 had multiple overlapping stents placed into a smaller lumen in the distal renal artery, 1 had periprocedural embolization to the kidney, and in 1 patient no etiology was discovered (Table III).
Table III. Etiology of renal artery stent occlusion
| Renal artery stent occlusions – etiology | Number | % |
|---|---|---|
| Procedural dissection | 6 | 33 |
| Kink – at proximal renal stent | 3 | 17 |
| Kink – at distal stent edge | 3 | 17 |
| Small renal artery (<4 mm, accessory renal artery) | 3 | 17 |
| Procedural embolization | 1 | 6 |
| Unknown | 2 | 10 |
| TOTAL | 18 | 100 |
Revised duplex scan criteria for occlusions
Occlusions are suggested by a lack of flow seen in the renal artery stent or PSV <57 cm/s and RAR <1.2. If the renal artery cannot be visualized on an ultrasound scan, this should also raise concern for a possible occlusion and correlation made to the CT scan (Table IV, Fig 3).
Table IV. Revised ultrasound criteria for the detection and grading of renal artery stent stenosis and occlusion in patients with fenestrated or branched endografts
| Renal stent stenosis | Revised ultrasound criteria | CT criteria |
|---|---|---|
| <60% stenosis | PSV 57-280 cm/s | As visualized from centerline of flow and curved planar reconstructions |
| 60-99% stenosis | PSV >280 cm/s and/or RAR >4.5 | As visualized from centerline of flow and curved planar reconstructions |
| Occlusion | No flow seen in renal artery stent OR PSV <57 cm/s and RAR <1.2 (or unable to determine RAR because proximal RA not visualized) | Lack of contrast enhancement through the lumen of the stent on centerline and curved planar reconstruction and marked difference in renal parenchymal opacification in 3D reconstructions |
Fig 3.
Parametric receiver operator curve (ROC) analysis for the detection of renal stent occlusion using peak systolic velocity (PSV) and renal aortic ratio (RAR). For PSV the area under the ROC curve is 0.91 (95% confidence interval [CI] 0.88-0.94) indicating good discrimination between those with stent occlusion and no occlusion. For RAR the area under the ROC curve is 0.88 (95% CI 0.84-0.91) indicating good discrimination between those with stent stenosis and no stenosis.
Renal stent stenosis
A total of 30 out of 287 (10%) uncovered stents developed a stenosis in follow-up. The mean time to detection of stenosis was 16 months (SD 14.4, range, 1-48) and none were present during the immediate postprocedure period. There was no change in renal size pre- and post-stenosis. Six of these 30 stenoses were treated by secondary intervention for deteriorating renal function or progression of the stenosis with repeated stenting and all six remain patent and free of stenosis.
A total of six of 231 (2.5%) covered stents developed a stenosis in follow-up. The mean time to detection of stenosis was 10 months (SD 12.4, range, 1-36) and none were present during the immediate postprocedure period. There was no change in renal size pre- and post-stenosis. One patient who had a stenosis with an associated renal stent fracture underwent secondary intervention with repeated stenting and remains patent and free of stenosis.
Freedom from renal stent stenosis in the two groups is presented in Fig 4. The HR for stenosis when comparing covered vs uncovered stents is 0.4 (95% CI 0.2 to 0.9, P = .04) favoring the use of a covered stent. GFR pre-procedure, at the time of stenosis detection and at 1-year post-detection for patients with ultrasound scan-detected renal stenosis did not differ significantly. Covered stent stenosis occurred only at the distal stent edge whilst uncovered stent stenosis occurred at both the proximal and distal portions of the stent.
Fig 4.
Kaplan-Meier curves for freedom from stenosis in months. The solid line represents those treated with a covered stent and the dashed line represents those treated with an uncovered stent. The development of renal stent stenosis between the two groups was statistically different. Hazard ratio (HR) for stenosis (covered vs uncovered): 0.4 (0.2-0.9), P = .04.
Revised duplex scan criteria for stenoses
Using the revised criteria, a 60-99% renal artery stent stenosis in a fenestrated endograft is best detected using thresholds for PSV >280 cm/s (sensitivity [Sn] 93%, specificity [Sp] 100%, PPV 99%, NPV 99%) or RAR >4.5 (Sn 83%, Sp 89%, PPV 42%, NPV 98%) (Table IV and Fig 4).
Physiologic renal dysfunction
Overall, 20% (31) of the patients treated with uncovered renal stents had a sustained >30% deterioration in GFR (Fig 5). Of these 31 patients, renal arterial defects were noted by duplex scan studies in 15 patients (8 stenoses and 7 occlusions). The remaining 16 of 31 patients who developed renal dysfunction had deficits that were attributed to periprocedural renal embolization and/or contrast nephropathy. Renal infarcts were observed in the postoperative CT scans in 50% (8/16) of this subgroup. A total of 7 patients required hemodialysis during follow-up (3 temporary and 4 permanent). Two new-onset dialysis patients had renal occlusions, none had renal stenoses, and 5 had their renal insufficiency attributed to embolization or contrast nephropathy.
Fig 5.
Parametric receiver operator curve (ROC) analysis for the detection of renal stent stenosis using peak systolic velocity (PSV) and renal aortic ratio (RAR). For PSV the area under the ROC curve is 0.99 (95% confidence interval [CI] 0.99-1.00) indicating optimal discrimination between those with stent stenosis and no stenosis. For RAR, the area under the ROC curve is 0.91 (95% CI 0.82-0.96) indicating good discrimination between those with stent stenosis and no stenosis.
Covered renal stents were used in 129/287 patients, of which 16% (21) patients had a sustained >30% deterioration in GFR (Fig 6). Of these 21 patients, no duplex scan-detected renal stent stenoses were found. Renal occlusions were noted in 4 patients during follow-up. Renal dysfunction progressed to require dialysis in 2 of the 4 patients with renal occlusions (1 temporary and 1 permanent dialysis). The remaining 17 of 21 patients who developed worsening renal function had their deficit attributed to periprocedural renal embolization and/or contrast nephropathy. Renal infarcts were observed in the postoperative CT scan in 40% (8/17) of this subgroup, and of these, 4/17 required dialysis during the course of follow-up (2 temporary and 2 permanent).
Fig 6.
Flow diagram demonstrating renal events during follow-up (F/U) of all patients treated using uncovered renal stents. GFR, Glomerular Filtration Rate.
Discussion
The vast majority of renal stents in patients treated with fenestrated and branched devices for aneurysmal disease remain stable during follow-up. There was a low risk of severe renal complications (only 2% cumulative incidence of new permanent dialysis requirement) despite the prevalence of renal insufficiency of the patient population prior to treatment of aneurysms that involve or abut the renal arteries. These results compare favorably with surgical series for both juxtarenal aneurysms21 and thoracoabdominal aneurysms.22, 23 However, when stringent assessment criteria were applied to our patient population, 20% of patients were noted to develop some level of physiologic renal dysfunction, 7% developed renal stenoses, and 3% developed occlusions. Much like the results of contemporary renal stenting trials,24 the minority of renal functional deficits in this series can be attributed to renal artery pathology. The most common causes of renal dysfunction included embolization of atheromatous debris or the development of a nephropathy that may be associated with contrast administration. However, there was a significant subset of patients who developed renal dysfunction as a result of arterial lesions created by the procedure or device. Therefore, it remains critical for clinicians to have non-invasive methods allowing for the accurate assessment of renal arteries following treatment of these complex aneurysms. The absence of defined duplex scan criteria, the lack of correlation between CT imaging studies and other studies, and the inability to define the need or benefit from intended secondary interventions hamper follow-up paradigms after treatment with these devices.
The detection of native renal artery stenoses is well established. The sensitivity and specificity of duplex ultrasound scan, when done in an accredited vascular lab, has been touted to be greater than 98% for de novo renal lesions.13 Given the effect of a stent protruding from the renal artery into the aorta on turbulence and arterial velocity, there remains debate as to the optimal ultrasound scan criteria indicative of renal stenosis.14, 15, 16, 17 In the setting of a renal stent in conjunction with an aortic stent graft, where the aortic velocities are also altered, conventional duplex scan criteria indicative of stenosis must be carefully questioned.
The absence of a non-invasive gold standard to define renal artery pathology resulted in the comparison of two imaging datasets (CT and duplex scan) which have both been advocated as a means to assess renal arterial pathology. The strong correlation between the two results (ROC r values 0.88 to 0.99; Fig 1, Fig 2) adds credence to the concept that renal artery lesions can be detected non-invasively using either modality. The resulting ROCs also provide a means to refine duplex scan criteria, minimizing the risk of false positive diagnoses, and improve the sensitivity and specificity of the technique (Table III). A total of 18 patients with uncovered stents and 11 patients with covered stents had a false positive elevated velocity on duplex scan. This exemplifies the need for specific training of ultrasound scan technicians to properly interrogate a branched or fenestrated graft. Similarly, the construct of curved planar reconstructions through renal artery stents also required a defined skill set. Both techniques are useful during the follow-up of such patients, and should be developed at each center employing these endovascular therapies. Given that the methods correlate well with each other, they may serve as a means for validation or one method may be used in the absence of other data.
The management of patients with diagnosed renal artery pathology evolved throughout the course of the trial. The initial aggressive approach was likely not warranted given the somewhat benign natural history of most renal lesions in this setting as evidenced by the lack of progression of disease. It is difficult to determine the exact criteria which prompted intervention as the study progressed, given that our understanding of the need for treatment evolved. However, in general, any clinical evidence of renal dysfunction in conjunction with evidence of renal arterial pathology resulted in further investigation. Renal stenoses underwent confirmatory angiography with angioplasty and stenting in the setting of significant stenoses (>50% angiographic stenosis postprocedure). However, the majority of renal stenoses were not treated, particularly in the setting of preserved renal function, or when patients were considered unfit for secondary procedures. Overall, in the setting of preserved renal function, 17% of patients with uncovered stents, and 5% of patients with covered stents developed renal stenoses. This compares to 39% of patients who had a pre-existing >75% angiographic renal artery stenosis prior to stent graft placement. We could not correlate the presence of preoperative renal artery stenosis with any postoperative outcome. Of the 36 patients with documented renal stenoses diagnosed at a mean time of 15 months following device implantation, 28 had no evidence of renal function deterioration during follow-up of 24 months after the detection of the stenosis. Patients treated with uncovered stents who developed renal dysfunction only had a detectable renal artery lesion 50% of the time. Similarly, patients treated with covered stents who developed renal dysfunction had detectable lesions only 20% of the time. Our current practice dictates that in the setting of renal dysfunction, worsening stenosis or occlusion, secondary interventions should be considered. However, in the absence of renal dysfunction, asymptomatic stable renal artery lesions are observed.
Renal occlusions behaved differently when contrasted with renal stenoses. Most of the renal occlusions were attributed to technical issues encountered during the implantation procedure. This was evidenced by the early appearance of the renal occlusions relative to renal stenoses (mean time to occlusion diagnosis was 4 and 3 months for uncovered and covered stents, in contrast to stenosis diagnosis at 17 and 10 months, respectively). The most common causes of renal occlusions included dissection and excessive tortuosity, two factors that are extremely difficult to rectify with secondary procedures. Two renal occlusions occurred when accessory renals, that were quite small, were incorporated into the repair. However, most renals that occluded were not small, given that the mean balloon size used to deploy stents into renal arteries that ultimately occluded was 6.3 mm. A subset of patients with renal occlusions were selected for attempted endovascular recanalization (no open surgical revascularization procedures were performed). Although two attempted recanalizations were successful, with patent renal arteries noted on follow-up studies thereafter, renal function measurably improved in only 1 of these patients.
Observed differences in the development of stenoses in patients treated with uncovered and covered stents merits discussion (Figure 7). Most renal stenoses associated with uncovered stents were located at the proximal aspect of the stent, while stenoses observed following placement of covered stents occurred at the distal end of the stent. When placing fenestrated or branched grafts, the proximal aspect of renal stents and stent grafts were flared in an attempt to rivet the graft material to the renal artery or seal against the reinforced nitinol ring. Over dilation such as this has been associated with intimal/medial injuries, and a hyperplastic response potentially resulting in late restenosis.25, 26 Coverage of the injured segment with graft material likely impedes the ingrowth of tissue by simply forming a physical barrier, or rendering the arterial wall ischemic. Distal renal arterial lesions likely have an alternative explanation. Inspiration and expiration result in considerable renal motion, which appears to have the greatest effect at the distal edge of the renal stent (see web movie, online only). Additionally, any tortuosity in the proximal renal artery is shifted distally following placement of a balloon expandable stent or stent grafts. This problem was compounded by the fact that covered stents were more commonly placed deeper into the renal vasculature than uncovered stents (to establish firm fixation and sealing for thoracoabdominal aneurysms), resulting in exaggerated distal renal artery kinking in some circumstances. Interestingly, this was observed in conjunction with a renal stent occlusion more often on the right renal artery rather than the left. We hypothesize that the added anterior-posterior angulation of the right renal artery as it dives under the inferior vena cava (Figure 8) had a detrimental effect on the patency of the vessel, as over 70% of all late occlusions occurred on the right. Today, we are likely to extend the balloon-expandable stent graft with a self-expanding stent in an effort to taper the imposed stiffness, creating a more smooth transition to a tortuous distal renal segment. Given the overall benefit of covered stents, one might question whether such devices should be used for de novo renal artery stenoses as well.
Fig 7.
Flow diagram demonstrating renal events during follow-up (F/U) of all patients treated using covered renal stents. GFR, Glomerular Filtration Rate.
Fig 8.
Shown are postoperative computed tomography (CT) scan reconstructions of a patient in whom a right renal artery kink (A, B) was recognized intra-operatively and treated with a self-expanding stent at the distal renal stent edge to renal artery junction (C, D) (arrow).
The results reported in this analysis compare favorably with other reports of experiences with similar devices. Pooled literature results (including our earlier report)27 estimate a 6% (95% CI 1.7-15%) incidence of renal artery occlusion,28 and our observed incidence of 2% certainly falls within this range. Although there are other reports of experiences with fenestrated and branched devices,10, 29, 30, 31 most do not have sufficient information regarding detailed renal artery and renal function analysis, or have limited enrollment to assess the incidence of renal issues. Other investigators have suggested that patients should be maintained on antiplatelet therapy (including both aspirin and clopidogrel) indefinitely following stent implantation,12 however we had not adopted this policy, and our patients were maintained only on aspirin indefinitely, unless otherwise indicated.
There exist several limitations with this series. First and foremost, the choice of stents was not randomized, and thus there exists fundamental differences (anatomic and physiologic) in the patient population. This remains as a confounding variable in determining the potential for improved efficacy of covered stents in the patient not requiring concomitant aortic stenting. Further issues include the absence of defined method of non-invasively assessing the renal artery status after device implantation. Therefore, although we assimilated two different methods for validation, the potential for both to reflect artifact or have fundamental errors exists. Finally, temporal disparities in enrollment relegate to two treatment groups to different durations of follow-up. Will we see greater incidence of late renal artery problems in the stent graft group as time goes on? These issues serve as points of discussion for the design of future trials intended to assess renal stenting, fenestrated stent grafting, and complications following complex aneurysm repairs.
The results from this study of 518 stented renal arteries with a mean follow-up of over 2 years demonstrate that there exists a low risk of posttreatment renal failure progressing to require hemodialysis (2%) and a cumulative incidence of renal occlusions in 3% of stented arteries. Stenoses were more commonly noted when uncovered stents were employed, although the natural history of untreated renal artery stenosis following endovascular treatment of complex aneurysms is rarely one of progression to occlusion or worsening renal function. The revised duplex scan criteria provided in conjunction with available high resolution CT data will detect most renal stenoses, allowing clinicians to then determine whether further intervention will be warranted on clinical grounds.
Author contributions
Supplementary data
Video 1.
References
- . Renal artery stenting slows the rate of renal function decline. J Vasc Surg. 2007;45:726–731
- . Is renal artery stenting the correct treatment of renal artery stenosis? (The case for renal artery stenting for treatment of renal artery stenosis). Circulation. 2007;115:263–269
- The management of renal artery atherosclerosis for renal salvage: does stenting help?. J Vasc Surg. 2007;45:101–108
- Diabetes mellitus does not preclude stabilization or improvement of renal function after stent revascularization in patients with kidney insufficiency and renal artery stenosis. Catheter Cardiovasc Interv. 2007;69:902–907
- . Catheter-based therapy for atherosclerotic renal artery stenosis. Circulation. 2006;113:1464–1473
- . Endoluminal aortic grafting with renal and superior mesenteric artery incorporation by graft fenestration. J Endovasc Ther. 2001;8:3–15
- . Branched stent-grafts for endovascular repair of aortic and iliac aneurysms. Tech Vasc Interv Radiol. 2005;8:56–60
- Endovascular management of juxtarenal aneurysms with fenestrated endovascular grafting. J Vasc Surg. 2004;39:279–287
- Mid-term results of endovascular aneurysm repair with branched and fenestrated endografts. J Vasc Surg. 2006;44:9–15
- . Outcomes of fenestrated endografts in the treatment of abdominal aortic aneurysm in Western Australia (1997-2004). J Endovasc Ther. 2006;13:320–329
- Fenestrated endovascular grafting: the renal side of the story. J Vasc Surg. 2005;41:181–190
- . Fenestrated endovascular grafts for juxtarenal aortic aneurysms: a step by step technical approach. Catheter Cardiovasc Interv. 2007;69:554–571
- . The utility of duplex ultrasound scanning of the renal arteries for diagnosing significant renal artery stenosis. Ann Intern Med. 1995;122:833–838
- Doppler ultrasound (pre- and post-contrast enhancement) for detection of recurrent stenosis in stented renal arteries: preliminary results. Australas Radiol. 2000;44:36–40
- Duplex ultrasound for follow-up examination after stent-angioplasty of ostial renal artery stenoses. Ultraschall Med. 2002;23:315–319
- . Duplex ultrasonography in assessing restenosis of renal artery stents. Cardiovasc Intervent Radiol. 1999;22:475–480
- . Enhanced flow velocity after stenting of renal arteries is associated with decreased renal function. Nephron. 2007;105:84–89
- Beyond the aortic bifurcation: branched endovascular grafts for thoracoabdominal and aortoiliac aneurysms. J Vasc Surg. 2006;43:879–886
- National Kidney Foundation practice guidelines for chronic kidney disease: evaluation, classification, and stratification. Ann Intern Med. 2003;139:137–147
- . Multidetector CT angiography in the follow-up of patients treated with renal artery stents: value of different reformation techniques compared with axial source images. J Endovasc Ther. 2007;14:387–394
- Open repair of juxtarenal aortic aneurysms (JAA) remains a safe option in the era of fenestrated endografts. J Vasc Surg. 2008;47:695–701
- . Open surgical repair of 2286 thoracoabdominal aortic aneurysms. Ann Thorac Surg. 2007;83:S862–S864
- . Surgical treatment of aneurysms of the thoraco-abdominal aorta. La Revue du practicien. 2002;52:1078–1083
- Angioplasty and STent for Renal Artery Lesions (ASTRAL trial): rationale, methods and results so far. J Hum Hypertens. 2007;21:507–508
- . Restenosis after balloon angioplasty and/or stent insertion - origin and prevention. Acta Radiol. 2002;43:442–454
- . Myointimal hyperplasia: basic science and clinical considerations. Semin Vasc Surg. 1998;11:142–148
- . A prospective analysis of fenestrated endovascular grafting: intermediate-term outcomes. Eur J Vasc Endovasc Surg. 2006;32:115–123
- . Fenestrated and branched endovascular techniques for thoraco-abdominal aneurysm repair. J Cardiovasc Surg (Torino). 2005;46:131–140
- . Short to midterm outcomes of fenestrated endovascular grafts in the treatment of abdominal aortic aneurysms: a systematic review. J Endovasc Ther. 2006;13:747–753
- . Fenestrated endografting for aortic aneurysm repair: a 7-year experience. J Endovasc Ther. 2007;14:609–618
- . Endovascular treatment of thoracoabdominal aortic aneurysms. J Vasc Surg. 2008;47:6–16
Competition of interest: none.
Additional material for this article may be found online at www.jvascsurg.org.
PII: S0741-5214(08)01953-8
doi:10.1016/j.jvs.2008.11.024
© 2009 The Society for Vascular Surgery. Published by Elsevier Inc. All rights reserved.
