The Role of Embolic Protection Devices in Renal Angioplasty and Stenting

Introduction

Atherosclerotic renovascular disease is increasingly recognized thanks to technical improvements in duplex ultrasound, magnetic resonance angiography, CT scan, routine renal angiography during cardiac catheterization, coronary procedures and particularly in hypertensive or multivascular diseased patients. It represents an important public health problem. A renal artery stenosis (RAS) is usually caused by atherosclerosis (80% of cases in patients over 40 years), and in rare cases is due to fibromuscular dysplasia (10% of cases and more often in young patients), arteritis (Takayasu’s disease), neurofibromatosis and radiation. It can also be diagnosed in a renal transplant or in a renal bypass graft.

The prevalence of RAS is high. Rihal found RAS greater than 50% in 19.2% of patients during cardiac catheterization of 297 hypertensive patients.¹ The prevalence of RAS is 35–45% in patients with peripheral vascular disease, 14–24% in patients with cerebrovascular disease and 7–30% in patients with coronary heart disease.^2–4 In patients with renal insufficiency, the incidence of unsuspected RAS is as high as 24%.⁵ RAS greater than 60% has been reported to be 6.8% in patients older than 65 years of age.⁶ The natural history of RAS depends on its etiology. Atherosclerotic RAS, as well as renal artery stenosis due to Takayasu’s Arteritis, have a high tendency to progress with time, resulting in renal artery occlusion, loss of renal mass and a subsequent decrease in kidney function. In the vessels with the most severe atherosclerotic stenoses, the result is total occlusion in 16% of the patients.^7–11 On the other hand, RAS due to fibromuscular dysplasia rarely tends to progress to occlusion.^12,13 Atherosclerotic RAS can lead to different clinical situation:

• A renovascular hypertension (secondary hypertension), which accounts for 1–5% of all cases of hypertension.^14,15

• A renal insufficiency. A rise in serum creatinine following initiation of antihypertensive therapy with ACE inhibitors may lead to the diagnosis of RAS. A RAS may be severe enough to cause ischemia and tissue damage, as is often shown by asymmetry in renal size.¹³

• A flash pulmonary edema is often the first clinical symptom of bilateral RAS.¹⁶

Renovascular hypertension can also be caused by non-atherosclerotic RAS such as fibromuscular dysplasia, which occurs in younger patients (17,18 and has a higher prevalence in females. Atherosclerotic renovascular disease represents an important public health problem. It has been demonstrated that it increases cardiovascular and all-cause mortality.

Aggressive treatment of a severe RAS (defined as a diameter stenosis of at least 70%) is recommended for patients with uncontrolled hypertension, renal insufficiency, congestive heart failure, unstable angina and for patients with a solitary or a single-functioning kidney. The treatment of RAS without hypertension or renal insufficiency is debatable but could be considered with an intention of preserving renal function and renal artery patency. The treatment options for a RAS include medical therapy, balloon angioplasty (with and without stenting), and surgery. Surgery carries a significant risk with a 2–7% perioperative mortality rate and a 17–31% morbidity rate. In addition, deterioration in renal function occurs in 11–31% of patients, reocclusion and restenosis in 5–18%.^24–26 Indications for surgery should be limited to failed percutaneous approach, hostile aorta, or infrarenal total occlusion (in association with aortic surgery).

Percutaneous transluminal renal angioplasty (PTRA) has become the cornerstone of therapy for addressing RAS and is now the first treatment to be proposed. Balloon angioplasty alone was initially proposed and is still the first line of therapy for fibrodysplasic RAS. However, several authors have reported the successful use of endovascular stents for treating suboptimal angioplasty results, and as a primary intervention for atherosclerotic lesions (particularly ostial lesions) with better immediate and long-term results than with PTA alone.^27–32 Renal angioplasty stenting can be performed using the femoral approach in most cases. The brachial or radial approach can also be used. The technique benefits from the improvements of coronary technique are monorail systems for balloon and stents, low profile devices, 0.014'' or 0.018'' guidewires, allowing direct stenting in 80–90% of the procedures. Procedural success is excellent (98–100%), with a low complication rate, a low restenosis rate and a good long-term patency rate of 85–98%.^21–26 The benefit of PTRA includes complete cure (7–19%), or at least easier management of hypertension (52–74%) in addition to preservation, or improvement of renal function.^27,28,31

However, post-procedural deterioration of the renal function occurs in a subset of patients after PTRA.^32–34 We hypothesize that atheroembolism during the procedure is a precipitating factor for this complication and a major preventable factor limiting the benefits derived from PTRA. This hypothesis is supported by recent studies.³⁵ In order to eliminate or reduce the risk of atheroembolic material being carried into the renal parenchyma, we applied the technique of distal embolic protection, using balloon or filters positioned distal to the lesion, a technique currently developed and approved for use in the coronary and cerebral circulations.^36–38 Use of such devices during PTRA may increase the rate of beneficial clinical responses after revascularization. We will report a series of 124 RAS done under protection in 105 patients. All patients provided written informed consent.

Effects on Renal Function: Rationale for Protection Devices

The effects of renal angioplasty and stenting on renal function are controversial. There is no randomized study comparing either balloon angioplasty or stenting with medical therapy. The meta-analysis reported by Nordmann et al³⁹ demonstrated no consistent difference in changes in the renal function. According to recent American Heart Association Guidelines,⁴⁰ a slowed decline in renal function is sufficient to support the claim that renal artery angioplasty is beneficial. Recent studies regarding the effects of peripheral renal angioplasty or stenting on renal function showed that a large percentage of patients seem to benefit from the procedure with a stabilization or improvement in renal function

Renal stenting in selected patients could slow the progression of renovascular renal failure.^{41–45,50–58} In patients with a normal baseline creatinine level, renal function was mostly preserved.^31,43,50,54 In a non-randomized study, PTRA improved renal function in 41–43% of patients.⁵⁹ Blum et al found no significant change in creatinine independent of baseline renal function.³⁰ White et al,⁵⁴ in his series of 100 patients, found no significant change in the creatinine level. However, in 9 of 44 patients with renal insufficiency, creatinine normalized. The effect of PTRA on renal function was demonstrated by La Batatide-Alanore et al.⁶⁰ After following 32 patients who underwent PTRA for a mean of 6 months, split renal function was completed using Captopril renal scintigraphy. The renal function of the treated kidney was found to improve significantly after PTRA. This suggests that PTRA improves the renal function in patients with unilateral RAS.

There are few data on patients with severe renal dysfunction. In Zeller’s series,50 renal function improved in 71% of patients with severe renal insufficiency, 21% were unchanged, and 8% deteriorated. In 32%, planed chronic hemodialysis was deferred. In contrast, Nama et al⁶¹ showed no improvement 1 year after the procedure in 40 azotemic patients with 61 RAS. In 60% of these patients, renal function was improved or preserved, and in 40% it was deteriorated. There was no difference found between bilateral and unilateral disease. Despite these generally favorable results in many published series, a decline in renal function was noted in a subset of patients even after successful initial technical results and a good long-term patency of the renal artery. Deterioration in renal function may be seen in 20–30% of the patients or even more. In most of the series this depends on the renal function and the creatinine level at baseline.^32,34,42,44 Dorros et al³² reported a deterioration of the renal function in 47% of the patients with a creatinine level of over 2 mg/dl. Subramanian et al⁶² demonstrated a worsening in renal function in 24% of non-diabetic patients and 27% of diabetic patients with renal insufficiency and Guerrero et al⁴⁷ reported similar results in 31% of patients with renal insufficiency. Zeller et al⁵⁰ recently reported the long-term impact of stent-supported angioplasty on renal function in a series of 340 hypertensive patients. During a mean follow-up period of 34 ± 20 months, serum creatinine significantly decreased from 1.45 ± 0.87 to 1.39 ± 0.73 mg/dl (p = 0.048). The renal function was improved in 34% of the patients, remained unchanged in 39% but deteriorated in 27%.

Baseline serum creatinine, bilateral intervention, percent diameter stenosis and triple vessels coronary diseases were independent predictors of improved renal function during follow-up. The serum creatinine decrease was not significant in patients with diabetes mellitus. Other reports^41,43,55 suggest that only treatment of bilateral, severe RAS is beneficial to renal function. In contrast to other series, the highest proportion (36%) of Zeller’s patients with deteriorated renal function was found in the subgroup with normal baseline serum creatinine. However, in 90% of these patients, the increase in serum creatinine concentration was within the normal range. The more severe the renal dysfunction was at baseline, the more the patients benefited from the intervention. Improved, or unchanged, mean serum creatinine concentrations were seen in 64% of patients with normal renal function, 82% of patients with moderate renal impairment and 92% of those with severely impaired renal function. Improvement in renal function was also seen in 37% of the patients with nephrosclerosis and a resistance index of more than 0.8.

Watson et al⁵⁸ assessed the effect of renal artery stenting on renal function in patients with chronic renal insufficiency (creatinine > 1.5 mg/dl) by comparing the slopes of the regression lines derived from the reciprocal of serum creatinine versus time plotted before and after stent deployment. Before stent deployment, all patients exhibited a negative slope indicating progressive renal insufficiency. After stent deployment the slopes were positive in 18 patients (72%), indicating improvement in renal function and less negative in 7 patients (28%) indicating a persistence of decline in renal function deterioration, albeit at a lower rate than before the procedure in these 7 patients. Many factors may be responsible for this functional deterioration: contrast media induced nephrotoxicity, progression of concomitant nephrosclerosis, lesion recurrence, hyperperfusion syndrome, and glomerular injury. However, atheromatous embolism seems to play an important role and is an increasingly recognized cause of renal function deterioration.

It is demonstrated that atherosclerotic debris commonly embolize from lesions in many vascular territories during percutaneous intervention.⁶³ Evidence of distal embolization as a frequent complication was first seen in saphenous vein graft intervention.⁶⁴ Atheroembolism is also proven during catheter treatment of certain native coronary lesions,⁶⁵ and during carotid and renal stenting.^{36–38,66,67} Distal protection devices have been able to recover embolic debris in all these territories and significantly reduce its incidence of these complications. During a renal angioplasty, cholesterol atheromatous embolism is caused by the release of microscopic plaque fragments and cholesterol crystals from the renal artery lesion or the atherosclerotic aorta into parenchymal renal vasculature during the procedure. Instrument manipulation in the aorta and renal arteries can result in detachment and embolism of atheromatous debris from ulcerated plaques. The large size of the devices used, increased length or specific difficulties of the procedure may be contributory.

Walker et al⁶⁷ recently demonstrated the great potential for embolic debris during the placement of the guiding catheter, sheath or diagnostic catheter. He performed an aggressive aspiration of the guiding catheter or sheath before any contrast injection. Large particles of atherosclerotic debris (1–3 mm) were discovered in 41.7% of the patients. He proposed that careful aspiration of catheters before injections or interventions should routinely be performed. Patients with severe atheromatous disease of the aorta and its branches, ulcerated plaques, associated lesions, such as aneurysm, and dissection are candidates for these complications. Hiramoto et al³⁵ demonstrated that angioplasty and stenting of ex-vivo aortorenal atheroma specimens using 0.018-inch guidewire system was associated with thousand of atheroemboli. Each manipulation of the specimens including simply advancing the guidewire through the lesion released thousand of fragments. The number of fragments in each size category increased with decreasing particle size. Positioning and deploying the stent released an additional bolus of fragments similar to that released after balloon angioplasty. All these fragments were of sufficient size to create vascular occlusions and initiate significant renal parenchyma damage.

This author concludes that the results of angioplasty procedures could be improved by placing distal protection devices to prevent atheroembolization. Even with modern techniques, use of a low profile system cannot prevent atheroembolism and renal function deterioration as was recently reported by Nolan et al,⁶⁸ who published a series describing 82 procedures with a renal function deterioration in 24% of the patients at 1 year. The true incidence of atheroembolism is uncertain. Many patients can have a silent course because of the large functional kidney reserve, which allows normal serum creatinine values despite a significant decline in total glomerular filtration capacity. Therefore, only the most severe cases may be detected, especially in patients with preprocedural renal dysfunction and limited functional reserve. Abnormal serum creatinine may only be observed if 50% of the nephron population is destroyed. Most patients reach a peak serum creatinine level over 3 to 8 weeks, but onset can also be sooner.

Few studies have addressed the problem of atheroembolism following RAS. Boisclaire et al reported 4 cases in 33 procedures.³⁴ Van de Ven et al²⁷ reported 2 cases in 24 procedures inducing renal insufficiency. Commeau et al⁶⁹ reported acute deterioration of renal function due to cholesterol embolism necessitating hemodialysis. Morice et al⁷⁰ also published 2 cases of partial renal infarctions in a series of 80 patients despite the use of a new low profile device and direct stenting. In the Aspire 2 Study,⁷¹ major embolic events occurred in 6.3% of the procedures. The clinical manifestation of atheroembolism may be non-specific.

• Thadhani et al⁷² reported a series of 52 patients with both renal failure and histologically proven atheroembolism after angiography in cardiovascular surgery. Within 30 days of their procedure, 50% of the patients had cutaneous signs, 14% blood eosinophilia.

• Most patients reach a peak serum creatinine level over 3 to 8 weeks⁷³ but onset may be sooner.³¹

• Haggie et al⁷⁴ reported 4 patients with proteinuria and nephritic syndrome. Atheroembolism can lead to different degrees of renal impairment:

-Moderate less of renal function;

- Severe renal failure requiring dialysis;

- Abrupt and sudden onset of renal failure;

- More frequently a progressive loss of renal function over 3 to 8 weeks;

- Chronic stable and asymptomatic renal insufficiency.

The diagnosis of atheroembolism is difficult. There is no specific test. Renal biopsy is the only definitive diagnostic tool, but its routine application is problematic. For these reasons, atheroembolism after renal intervention is often misdiagnosed as dye-induced nephrotoxicity or the progression of nephrosclerosis. However nephrotoxicity due to contrast media generally appears within 1 or 2 days or weeks. A prevention is essential in patients with renal insufficiency or in high risk patients (diabetes, elderly patients), the use of gadolinium, CoÇ angiography may be useful to reduce this risk. Renal atheroembolism not only poses a risk of renal function deterioration but also seems to decrease survival in patients undergoing endovascular procedures for RAS. Krishnamurthi et al⁷⁴ evaluated its impact on survival in 44 patients who had surgery for atherosclerotic RAS and concomitant intraoperative renal biopsy for detection of atheroemboli. Atheroembolic disease was identified in the biopsy specimens in 16 (36%) patients and correlated significantly with decreased survival (54% 5-year survival in this group versus 85% in patients without atheroembolism, p = 0.011). Boero et al⁷⁵ recently pointed out the bad prognosis of renal atheroembolism. In a series of 22 patients with an onset of symptoms from a few hours to 60 days, 11 patients (50%) were put on dialysis with a partial functional recovery in 4, while 11 patients (50%) died. Thus, it can be concluded that cholesterol embolism is a frequent cause of renal failure and is associated with a high mortality rate. The increasing number of such patients, the cost due to renal function deterioration and subsequent end stage renal disease requiring dialysis represents a significant long-term problem. No specific treatment can be suggested for renal atheroembolism. Therefore, the main aim should be the prevention during renal interventions. The selection of the patients may limit the risk, but more and more elderly, high-risk patients with advanced atheromatous diseases need treatment, and it is difficult to refuse these patients the benefits of the procedure. Certain technical points are very important and need to be mentioned. The procedures should be as atraumatic as possible with use of small devices and the adaptation of coronary angioplasty techniques. Direct stenting is not sufficient to avoid embolism.

As mentioned, Walker et al proposed the careful aspiration of the catheters. The “no touch” technique was also proposed to minimize atheroembolization.⁷⁶ Beyond these technical considerations to circumvent atheroembolisms, we applied the concept of protected renal angioplasty and stenting.^36–38 The rationale for distal embolic protection is similar to that of brain protection during angioplasty and stenting of the carotid arteries. Several studies have shown that protection devices with occlusion balloon or filter are effective in reducing the risk of embolization to the brain^77,78 and that these techniques are now mandatory in this field and represent the standard of care.⁷⁹ We postulated that the same technique could be suitable during renal angioplasty and stenting to reduce the risk of atheroembolism and deterioration of the RF.

Impact of Protection Devices during Angioplasty of RAS Techniques. The technique has been previously described.^36–38 All procedures are performed under local anesthesia and intravenous sedation. A 6–8 Fr guiding catheter is placed at the ostium of the renal artery, most of the time via a percutaneous femoral approach. Two types of protection devices can be used:

• Occlusion balloon. The Guardwire system (Medtronic, Minneapolis, Minnesota) is the same as that used for carotid protection.

• Filters. Current filters available on the market can be placed in the renal artery (same filters as for carotid angioplasty).

With the guiding catheter appropriately positioned, the protection device is carefully advanced across the lesion and positioned 2 or 3 cm beyond the target site. In the majority of the cases, crossing the lesion with a protection device is easy and is the same technique as with any guidewire. In a few cases (angulated artery, tight, calcified stenosis), it could be helpful to first place a more supportive “buddy wire” in the renal artery before attempting to cross the lesion with the protection device. In case of very tight stenosis, a predilatation of the lesion with a coronary balloon may also be indispensable before crossing the lesion. With a Percusurge device, the Microseal adapter is attached and the occlusion balloon inflated to occlude the renal artery. On detaching the adapter, the occlusion balloon remains inflated. With a filter, it is deployed by carefully removing the delivery sheath. Contrast injection confirms the accurate positioning of the protection device.

Angioplasty and stenting of the renal artery can be performed under protection using low profile devices, monorail technique. We perform a predilatation of the stenosis only in case of tight or calcified lesions. A direct stenting is done in the majority of the cases. Different types of stents were implanted. The diameter of the stent is chosen equal to the diameter of the artery. The stent is carefully positioned so that it protrudes into the aortic lumen by 1 or 2 mm. After stent placement, the protection device is retrieved:

• If a Percusurge device is used, after stent deployment the aspiration catheter is advanced over the wire to the level of the lesion and positioned adjacent to the protection balloon. Any debris is removed using a 20 ml syringe connected to the proximal end of the catheter. At least 2 aspirations are performed. After removing the aspiration catheter, the Microseal adapter is reattached to the Guardwire and the occlusion balloon deflated allowing normal vessel flow. If the angiographic result is satisfactory, the device is removed.

• With filters, angiographic control is performed after stent deployment. If the result is satisfactory, the filter is withdrawn with the removal sheath. To facilitate protection device withdrawal and prevent it from getting caught on the stent, the guiding catheter is routinely advanced into the stent up its distal end. The aspirated blood or the filters are sent to the laboratory for analysis.

In some cases with filters, we can observe a slow or no-flow after stent deployment due to a clogged filter. In that case, there is a stagnant column of blood proximal to the filter. Simply recapturing the filter will result in embolization to the kidney. It is indispensable to perform an aspiration proximal to the filter prior to recapturing the device, either with an aspiration catheter or simply with the guiding catheter. Some aspiration catheters that can be used include the Export Catheter (Medtronic), Pronto (Vascular Solutions, Inc, Minneapolis, Minnesota), or the Diver (Invatec, Roncadello, Italy). This technique of distal protection has some limitations we will discuss later.

To overcome some problems encountered with current filters, a new filter was recently proposed and is being investigated. This filter, the FiberNet® (Lumen Biomedical Inc, Plymouth, Minnesota), consists primarily of polyester fibers located coaxially around the distal tip of a guidewire assembly. This filter is soft, conformable and when actuated, expands radially to fill the vessel providing excellent apposition to the vessel wall.

Contained and captured emboli are recovered/removed both by aspiration through the retrieval catheter, and also by retention within the filter fibers when the filter is closed and retracted into the retrieval catheter. Aspiration is achieved through the retrieval catheter using the vacuum syringes to provide suction. This filter enables capture emboli as small as 30/40 microns without compromising the flow through the filter, and the number of particles removed appeared much higher than with other filters.

The possibility of suction through the retrieval catheter during device removal is probably one of the major improvements with this device, allowing cleaning of the dilated area and of the inner part of the stent. This technique allows aspiration of the debris, which can protrude through the struts of the stents after stent placement and dilatation and which could embolize after the procedure. The maximum diameter of this device is 7 mm, which is adequate in most of the cases. The landing zone required to place this filter is shorter than with other devices, about 1 cm. We performed the first human study in renal arteries with this new protection device and the results will be published soon.

Medications and Patient Surveillance

In all of these stenting procedures, patients were given ticlopidine (500 mg/day) or clopidogrel (75 mg/day) and aspirin (100 mg/day) before the procedure. During the procedure, an intravenous bolus of 5000 to 10,000 units of unfractioned heparin was routinely administrated at the beginning of the procedure to have an activated clotting time around 250–300 sec. The post-procedural drug regimen included aspirin (100 mg/day) indefinitely and ticlopidine (250 mg/day) or clopidogrel (75 mg/day) for one month. Patients remained in the hospital for 48 hours to monitor serum creatinine levels and adjust blood pressure medication. Renal duplex scanning was scheduled the day after the procedure, at 6 and 12 months and then annually. Angiography was performed when a restenosis was suspected on the basis of positive clinical and duplex scan findings. Serum creatinine values are measured before and after the procedure (day 1) and at 1 and 6 months, with biannual measurements thereafter. For our last patients we evaluated the Glomerular Filtration Reserve (GFR), which seems a more reliable parameter to appreciate the renal function.

Results

Only single-center series on renal angioplasty and stenting under protection have been reported.

Personal series. From January 1999 to July 2006, 124 RAS were treated in 105 patients with poorly controlled hypertension (72 males and 33 females), mean age was 64.5 ± 11.7 years (22–87), with percutaneous angioplasty and stenting under distal protection. All of these patients were diagnosed to have atherosclerotic RAS by renal duplex scanning and angiography. The indication for endovascular treatment was a stenosis >70%. Written informed consent was obtained from all patients. One bilateral procedure was performed in 18 patients and we treated 2 renal arteries on the same side in one patient. Nine patients had a solitary or single functioning kidney (1 transplant renal artery). Twenty-four patients had moderate renal insufficiency (serum creatinine 1.5–1.9 mg / dl) and 15 had severe renal dysfunction (serum creatinine greater than or equal to 2 mg / dl). The RAS was located at the ostium in 108 cases. Mean percentage stenosis was 85.9 ± 8.3% (70–99). Mean lesion length was 11.3 ± 3.0 mm (8–29). The diameter of the artery was estimated at 6 mm in 87 arteries and 5 mm in 37 arteries. A total of 59 patients had diffuse severe atherosclerosis of the abdominal aorta. Of the patients, 32 had diabetes mellitus, 73 were current smokers, 66 had hyperlipidemia, and 73 had associated coronary disease. Cerebrovascular disease was found in 20 and lower extremity peripheral artery disease in 37. A 6–8 Fr guiding catheter was placed at the ostium of the renal artery via a percutaneous femoral approach in all patients except one who presented with total occlusion of both iliac arteries necessitating a brachial approach. Different protection devices were used:

• Occlusion balloon (Percusurge): 46 procedures • Filters: 76 procedures - EPI (Boston Scientific, Natick, Massachusetts): 59 - Angioguard (Cordis, Warren, New Jersey): 7 - Emboshield (Abbott Vascular, Redwood City, California): 6 - FiberNet (Lumen Biomedical, Plymouth, Minnesota): 4 - Accunet (Guidant, Indianapolis, Indiana): 2 Stenoses were easily crossed with the protection device in 123 cases due to their low profile and flexibility. One predilatation was required to cross a subocclusive calcified stenosis. We had no difficulty in deflating the occlusion balloons or in removing the protection devices. A direct stenting was performed in 96 cases, and 127 different stent models were used, including Boston Express (n = 27), Cordis Genesis (n = 29), Medtronic (n = 15), Guidant Herculink (n = 14), Cordis Corinthian (n = 11), Carbostent Sorin (n = 7), Cordis P 154 (n = 6), Abbott (n = 6), Nir (n = 5), Stentec (n = 4), Cordis M3 (n = 2), Biotronik (n = 1). Three patients required 2 stents in the same artery to treat a long lesion. Technical success was obtained for all arteries (100%), with good stent deployment, no significant residual stenoses and complete covering of the lesion. Mean renal artery occlusion time with the Percusurge technique was 6.46 ± 2.42 mm. The mean time in situ for filters was 4.22 ± 1.18 mm, which is shorter than with the occlusion balloon. Two patients developed an arterial spasm (one with Percusurge, one with EPI filter) at the site of protection device, which immediately responded to local vasodilatator therapy. No dissection of the renal artery due to a protection device was seen.

Particulate analysis.

• With the Percusurge technique and the aspiration catheter, we aspirated visible debris in all patients (as with the carotid procedures). The aspirated blood samples were analyzed and studied by microscopy and scanning electron microscopy. Different particles were isolated and identified. Their number varied from 13 to 208 per procedure (mean ± SD 98.1 ± 60.0), and diameter ranged from 38 to 6206 µm (mean ± SD 201.2 ± 76.2 µm). The particles were atheromatous plaques, cholesterol crystals, necrotic cores, fibrin, fresh thrombi, organized thrombi, platelets, and macrophage foam cells, thrombogenic lipoid masses. Only fresh thrombi should be drug sensitive but these emboli were detected in only 10% of cases. The other particles are likely to be manageable only using mechanical means.

• We removed, with filters, visible debris in 80% of the cases. Two filters were totally blocked by large particles, with flow being totally interrupted, and 2 other filters were almost totally blocked with low flow. The flow was restored after aspiration with the guiding catheter and removal of the filter. One patient was 26 years old.

• With the new filter FiberNet^® visible debris were removed in all cases and 5 times more than with other filters. Minimum length 28 microns, average length 105 microns, maximum length 6.8 mm. Of the particles, 70% were less than 100 microns. Average area for FiberNet^® debris captured: 117.1 mm² (25.1–208). Atheromatous material was found in both aspirate and filter samples (65% in aspirate samples).

Follow up. The mean follow-up period was 18.3 ± 9 months (2–89 months). Four patients died from myocardial infarction, one at day 3 (patient who underwent coronary angioplasty shortly before renal stenting), and one at 6 months and one at 1 year of follow-up. One multivascular patient with severe renal insufficiency developed an acute renal failure after the procedure probably due to contrast nephropathy necessitating dialysis, and died one week later from multiorgan failure. Two patients were lost to follow up after one year, and 7 patients presented an in-stent restenosis successfully treated by a new balloon angioplasty. We have not seen device-related late vascular lesions in the 38 angiographic controls performed during the follow up.

Hypertension. During the follow up we observed no difference with other published series of patients treated without protection. Systolic and diastolic blood pressure decreased significantly. The number of medications also declined significantly (2.8 vs. 1.2). Nineteen patients were cured (18%), 61 improved (59%), and 25 remained unchanged (22%).

Renal function. As mentioned, we observed one acute function deterioration. There is no statistically significant difference during the follow-up. At 6 months, (91 patients), we observed only one deterioration of the RF in a patient with moderate renal insufficiency at baseline, 21 improvements in patients with renal insufficiency, 69 stabilizations. At 2 years (75 patients), we observed only 2 deteriorations of the RF (3%):

• One in a patient with normal RF before the procedure, normal RF at 6 month follow-up and who developed a renal insufficiency at 2 years. This patient was treated for bilateral stenosis, one side without protection.

• One in a patient with moderate renal insufficiency. 96% of the patients remained stabilized (n = 54) or improved (n = 19). At 3 years (37 patients), 93% of the patients remained stabilized (n = 26) or improved (n = 9). Twenty patients have RF deterioration.

Other Published Series

Few reports have been published in the literature. Holden et al reported a first series of 46 procedures in 37 patients with preprocedural renal impairment performed with the Angioguard filter.⁶⁶ They found the same results: renal function stabilized or improved in 95% of cases. Only 5% of the patients demonstrated an unchanged decline. No patients experienced acute post procedural deterioration. These results are better than in most reports in the literature and also better than in a group of similar patients with ischemic nephropathy who underwent stent revascularization without distal protection at the same institution.⁵⁵ The improved results are thought to be due to the prevention of cholesterol embolization during the procedure by distal filter baskets. A total of 65% of the filters contained embolic material, including fresh thrombus, chronic thrombus, atheromatous fragments and cholesterol clefts.

More recently, Holden et al⁸⁰ published a larger series of 106 RAS treated under protection with filters in 90 patients with ischemic nephropathy. He reported one acute deterioration of RF. At a mean follow up of 18.2 months, RF was improved in 36% of cases, stabilized in 55%. He reported only 8% of progressive decline of RF. Eggebrecht et al published a case of renal angioplasty performed under distal protection with the Percusurge device for an in-stent restenosis.⁸¹ Histologic examination of blood retrieved from the distally occluded vessel showed foam cells and an amorphous lipoid substance as markers of atherosclerotic plaque debris. The authors concluded that the Guardwire could be included in routine angioplasty maneuvers. Li and colleagues also published a case of successful renal angioplasty and stenting under distal protection with the Percusurge Guardwire.⁸² Guerkens and colleagues used the Angioguard filters in six patients and retrieved macroscopic debris in every case.⁸³ Edwards et al⁸⁴ more recently used balloon occlusion and an aspiration catheter (Percusurge device) to treat 32 RAS. Renal insufficiency was experienced by 92% of the patients. RF response at 4–6 week follow-up was improved in 50% of the patients and unchanged in 50%. In no patient or after any procedure was RF observed to be worsened, and 54% of the patients with RF deterioration were improved.

For the authors, the results with this technique of RAS under protection represent a marked improvement in short-term RF response rates compared with previously published experiences and approximate the short-term results reported after open surgical revascularization, and they concluded that these data suggest that this technique may prevent RF harm during RAS as a result of atheroembolism and warrant further investigation.

Discussion

RAS is increasingly diagnosed in patients suffering from hypertension and renal insufficiency and in multivascular diseased patients. PTRA is now the first treatment to be proposed with a high technical success rate, a low complications rate, a low restenosis rate, and good long-term anatomical results. A large percentage of patients seem to benefit from the procedure concerning hypertension and renal insufficiency, with stabilization or improvement in RF, and in selected patients, RAS could slow the progression of renovascular renal failure and may delay the need for renal replacement therapy.^{27–31,41–60} However, as we have noticed in many published series, a deterioration RF may be detected in at least 20–30% of the patients, not only in patients with renal insufficiency, but also in patients with normal RF at baseline and even after successful initial technical results and good long-term patency.^{32–34,41–45,50}

Many factors may account for this functional deterioration. But atheroembolism during the procedure could play an important role. As we have mentioned its diagnosis is difficult and it is crucial to prevent it. The risk can be reduced with a good selection of patients and some technical points, but the use of protection devices seems the best technique to overcome this problem. For this reason, we have applied the concept of protected renal angioplasty and stenting to circumvent atheroembolism. The rationale for distal embolic protection is similar to that of brain protection during carotid angioplasty and stenting of the carotid arteries. These techniques are now mandatory in this field and represent the standard of care^.79 We have postulated that the same technique could be of value in the management of RAS in mitigating the risk of atheroembolism. Our results show only one acute deterioration of RF probably due to contrast nephropathy in a multivascular patient in poor condition and renal insufficiency at baseline. At 6-month follow-up, 98% of the patients were stabilized or improved and only two instances of deterioration (7%) appeared. Of the patients, 93% remained improved or stabilized. These results seem to be favorable when compared with available published data and may well be attributed to the use of a protection device during the procedure.

It is noteworthy that visible atherosclerotic debris was extracted in all cases with balloon occlusion techniques and in 80% of cases with current filters. But with the new filter FiberNet®, visible debris was removed in all cases and at least five times more than with other filters. Four filters were totally or almost totally blocked by debris, thereby highlighting the role of protection devices. Similar results were also recently published by Holden et al⁸⁰ and Edwards et al.⁸⁴ Both authors insist on the role of protection devices to reduce atheroembolism and the risk of RF deterioration. However, this technique has some limitations as we have already mentioned.

• Distal protection devices do not prevent emboli from reaching the kidney during initial catheter manipulation or angiography or while crossing the lesion. We discussed some technical points to reduce the risk of atheroembolism.

• The use of distal embolic protection devices may be limited by the renal anatomy and the lack of devices currently available on the market and dedicated for this application.

• In the case of large vessels, we have to carefully select the device and choose a device with a diameter at least equal to the diameter of the renal artery.

• In case of an early renal artery bifurcation, it is not possible to protect all arteries. The protection device could be placed in the main branch. We could use two protection devices, one protection device is placed in each branch, but this technique could be limited by technical, anatomical problems and by the cost. In fact, in daily practice, 90% of the renal arteries can be protected with current protection devices.

• To place the protection device we need a landing zone of at least 1.5 to 2 mm, which may be a problem with long stenosis or non-ostial stenosis. Current protection devices are not dedicated to renal arteries and have to be modified and improved. The new Fibernet device does not require a long landing zone.

• The occlusion balloon has its own limitations and disadvantages: - It may deflate or may be non occlusive during the procedure;

     - Some particles may be too large for suction (very rare);

     - Below the balloon there is a shadow zone where some particles may remain blocked and are difficult or impossible to aspirate with the aspiration catheter. These particles may migrate to the kidney when the balloon is deflated;

     - The occlusion balloon could lead to transient nephron ischemia. In our series the mean occlusion time was short 6.46 mm. We do not think that this transient ischemia could cause any damage to the kidney. This occlusion time is less than that of the campling required during a surgical procedure. Edwards et al⁸⁴ also used this technique without any complication.

• Filters also have limitations and disadvantages:

     - Bad wall apposition with the possibility of migration of some particles around the filters;

     - A filter may plug up with suspended particles that will embolize if we retrieve the filter. In this case before retrieving the filter we have to very carefully aspirate the blood below the filter (as we did in some cases);

     - A filter may thrombose: we need an effective anticoagulation with an ACT > 250 to 300 seconds;

     - Closure and retrieval of filters can dislodge their content collected during the procedure;

     - Some difficulties in retrieving a filter may be encountered. The filter could get caught on the struts of the stent during retrieval;

     - But the major limitation of the filters is the pore size, which in general is greater than 100 microns. This is larger than the size of microcholesterol crystals⁸⁵ and allows small particles going to the kidney. Atheroemboli typically occlude the medium-sized arterioles (150 to 200 microns in diameter) and glomerular capillaries.

The pathogenesis of renal failure may be due entirely to occlusion of these vessels. But reactive inflammation surrounding the cholesterol crystals may play a significant role in causing the luminal occlusion and subsequent renal failure.⁸⁶ As we have mentioned, the new generation of filters like FiberNet® should reduce the risk of atheroembolism and of RF deterioration. Particles of 30–40 microns can be removed with this device and the possibility of suction with the retrieval catheter is an important improvement.

Although no device-related complications occurred in this small series of patients, adding another instrument to the procedure while trying to prevent complications could create new problems. The potential for renal artery thrombosis during protection is extremely small, even with occlusion balloons, because the patients are on heparin and antiplatelet therapy. Moreover, the duration of occlusion is usually short, less than the time required to perform the distal anastomosis of a conventional aortorenal bypass. The risk of dissection with a protection device is negligible, but one has to consider the possibility of spasm (2 cases in our series), which is, however, usually treatable medically. The indications for this technique of renal protection are debatable. Is the technique indicated for all patients? The indications for protection in patients with normal renal function need to be considered. The incidence of procedure-related decline in the renal function is considered to be low in this patient group. Many interventionists think that the extra cost of a renal protection system for these patients is not justified. However, I would like to point out the series recently published by Zeller et al⁵⁰ where the highest proportion (36%) of patients with worsened renal function were found in the subgroup with normal baseline creatinine. In our series, 66 hypertensive patients were treated with protection despite a normal serum creatinine. Only 1 patient experienced a renal function deterioration detected at 2-year follow-up. This technique of including renal protection may therefore become the standard of care in the future. At the present time, selective indications should at least be pointed out:

• Patients with renal insufficiency and a creatinine level of more than 1.4 mg/dl, or possibly better, and a GFR of less than 50 ml/min;

• Elderly patients;

• Patients with ischemic nephropathy;

• Bilateral RAS;

• Solitary or single functioning kidney;

• Patients with diseased aorta and renal ostia;

• Possibly diabetes.

As has been mentioned by Sos et al,⁸⁵ physicians performing renal artery interventions must understand the non-linear relationship of renal function reserve (RFR) and serum creatinine (SCr) and its implication for the margin of safety. Severe iatrogenic renal parenchymal damage due to interventional or diagnostic procedures can be masked in patients with normal pre-intervention global SCr values. Thus, 50% of total renal mass can be destroyed without any change in renal function. Patients with elevated SCr whose renal function is at the “knee” of this curve have very diminished renal reserve and are at much greater risk. An additional 10% loss of renal parenchyma can put such a patient on dialysis. There can be extensive damage to the kidneys that, in many patients with normal preintervention renal function, may not be apparent during or after renal intervention. It is difficult to know exactly which patient needs protection. For example, a patient with 60% renal function reserve (normal SCr) before the procedure could have renal insufficiency after the procedure if cholesterol embolism destroyed 20% of the nephrons. A good evaluation of the GFR is necessary before the procedure and particularly in patients with limited renal function, elderly patients and patients with extensive atheromatous disease.

The choice of which protection device to use merits discussion. We have seen no difference between balloon occlusion and filter use in carotid angioplasty, but our series is limited. For Sos et al,⁸⁵ balloon occlusion protection devices seem more applicable, because the filters have a pore size of 100 microns or more, allowing small particles and cholesterol crystals into the kidney. For Holden et al, filters are the ideal renal protection system.⁸⁰ New protection devices have to be developed and dedicated to the renal system. The new Fibernet device seems promising, allowing capture of smaller particles, with a good wall opposition and because only a short landing zone is required, it should enlarge the possibilities of renal protection.

Conclusion

Renal artery angioplasty and stenting is largely performed with very good technical success, good anatomical results, and a low complication rate and with a good long-term patency rate. The effects on blood pressure are encouraging. However, the deterioration of renal function after the procedure, which may occur in 20–40% of cases may limit the immediate benefits of this technique. Atheroembolism seems to play an important role. Physicians dedicated to this field should be aware of the risks of atheroembolism and deterioration of renal function after renal interventional procedures.