Cholesterol Embolism Evaluated by Polarized Light Microscopy ... - JVIR

Cholesterol Embolism Evaluated by Polarized Light Microscopy after Primary Renal Artery Stent Placement with Filter Protection José Urbano, MD, Felix Manzarbetia, MD, and Carlos Caramelo, MD

PURPOSE: Cholesterol microembolization may explain some treatment failures after renal artery stent placement. The identification of cholesterol crystals may provide significant help in diagnosing the real frequency and severity of this complication. The aim of the present study was to examine the efficacy of polarized light imaging in the detection of cholesterol emboli trapped in a protection device. MATERIALS AND METHODS: During a period of 18 months, 15 significant atherosclerotic stenoses of the ostium of the main renal artery were treated with primary stent placement with embolic protection. The filter device used was made of polyurethane, with a pore size of 115 ␮m. The device was mounted over a 0.014-inch guide wire. For pathologic analysis, the recaptured filter basket was compressed between two slides and examined in a microscope under polarized light. RESULTS: All the stenoses were successfully treated without clinical complications. All the filters were deployed and recaptured without difficulty. Cholesterol crystals were detected in 12 filters and no cholesterol was found in three. In one case, trouble with filter manipulation precluded pathologic analysis. No worsening of renal function was detected in any patient during follow-up. CONCLUSIONS: Microscopic analysis with polarized light easily detects the cholesterol crystals trapped in the filter device. This provides evidence that renal cholesterol microembolism is highly prevalent during renal artery stent placement. J Vasc Interv Radiol 2008; 19:189 –194

RENAL artery stenosis is a progressive illness with a worsening evolution of 20% per year, leading to renal atrophy at a rate of 10% per year and renal artery thrombosis at a rate of 5% per year. Two to 20 percent of the population with chronic renal insufficiency has renal artery stenosis, as do 5%–15% of patients who begin chronic

From the Departments of Vascular and Interventional Radiology (J.U.) Pathology (F.M.), and Nephrology (C.C.), Fundación Jiménez Díaz Hospital, Universidad Autónoma de Madrid, Avenida Reyes Católicos n°2, 28080 Madrid, Spain. Received May 20, 2007; final revision received October 2, 2007; accepted October 8, 2007. Address correspondence to J.U.; E-mail: [email protected] None of the authors have identified a conflict of interest. © SIR, 2008 DOI: 10.1016/j.jvir.2007.10.006

dialysis treatment each year (1,2). Treatment of main renal artery stenoses with balloon-expandable stents is a widely established and safe treatment with evident hemodynamic benefits. These benefits include more efficient control of hypertension and an improvement, or at least slowing, of the progression of chronic renal failure (3,4). Recently, it has also been proven that renal artery stenosis is a significant independent predictor of cardiovascular events and increases left ventricle hypertrophy and congestive heart failure. The main cause of mortality in these patients is not chronic renal insufficiency, but myocardial infarction or stroke (5,6). Nevertheless, there is some controversy concerning the factors that limit the clinical usefulness of renal artery stent placement (7), and the effect of percutaneous re-

nal revascularization on preservation of renal function is still a matter of debate. In recent years, technical improvements of endovascular tools (eg, primary stent placement, guiding catheters, low-profile stents with rapidexchange monorail systems, protection devices, and drug-eluting stents) have led to more widespread use of renal stent placement and extension of the indications of this type of therapy. All these changes in aggressive adjuvant medical therapy seem to have improved clinical results (8 –10). However, despite technical improvements, better indications, and low complications rates, it is frustrating that nearly 20% of patients with chronic renal failure have disease progression (8,11). Moreover, in some cases, worsening of renal function seems to be directly and

189

190

•

Light Microscopy of Cholesterol Embolism after Renal Stent Placement

closely related (in terms of time) to stent placement, even despite apparent technical success of the intervention and an absence of complications (12). In some patients, transient or progressive worsening in renal function occurs even with normal findings on control angiography after stent placement and with the use of no more than 50 mL of iodixanol contrast agent during the procedure. Contrast agent–induced nephropathy and progression of renal disease before stent placement are the more frequent explanations of this bad outcome, but our data support the possibility that some of these cases involved cholesterol embolization that could not be diagnosed or that coexisted with the other negative factors (13–15). Cholesterol embolization damages the tissues. A greater proportion of this effect is caused by the toxic effect triggered in the target organs than by its embolic effect, which is minimal (13,16). The skin, central nervous system, retina, gut, and kidneys are the most affected organs. No angiographic signs aid in diagnosis; in fact, all the renal arteries are patent, with normal findings on control angiography after stent placement. Eosinophilia, although not definitive, is a source of suspicion; more specific clinical symptoms will depend on the embolized organ (14). Herein we describe our use of polarized light to identify the cholesterol crystals released during renal stent implantation with an embolic protection device. The real prevalence of renal cholesterol embolization during endovascular interventions is unknown.

MATERIALS AND METHODS Patients During a period of 18 months, in 11 selected patients, 15 significant atherosclerotic stenoses in the main renal artery were treated by primary stent placement with adjuvant filter protection. All patients had atherosclerotic peripheral vascular disease. Fourteen kidneys were native and one was transplanted. All kidneys were at least 9 cm in length. Median subject age was 67 years (range, 49 –78 y), and 10 subjects were men. All patients had mild or moderate chronic renal insuf-

ficiency, with a median preintervention serum creatinine level of 2.29 mg/dL (range, 1.4 –5.3 mg/dL). Eight patients also had uncontrolled medical hypertension, with acute hypertensive crisis in three. At admission, two patients were experiencing flash pulmonary edema. A recent and significant serum creatinine increase was detected in four individuals. In three patients, treatment was performed a number of days before infrarenal aortic aneurysm intervention. Three of four patients with bilateral stenosis were treated during a single procedure. In view of the poor clinical condition of one patient, unilateral stent implantation was performed initially and the other kidney was treated 1 week later. Embolic Protection Device The Accunet embolic protection device (Guidant, Santa Clara, Calif) consists of a polyurethane membrane and a nitinol basket with a diameter ranging from 4.5 to 7.5 mm, mounted on a 0.014-inch rapid-exchange wire. The membrane contains microscopic pores 115 ␮m in diameter that maintain blood flow but trap emboli. The wire can be rotated independently from the basket and the basket deploys concentrically along the wire. The doublestrut filter basket has four radiopaque platinum markers. The system is compatible with a 6-F sheath and 8-F guiding catheter. Stent Placement and Filter Technique Technical success was defined by the delivery and recovery of the Accunet embolic protection device and renal stent deployment through the Accunet wire. Stents were implanted via a right femoral (n ⫽ 12) or transbrachial (n ⫽ 3) approach. After arterial puncture, the patients received anticoagulation with an intravenous bolus of heparin (50 U/kg). The contrast agent employed was 1 mol/L gadolinium (50 mL; Gadovist, Schering, Berlin, Germany) in two patients and iodixanol (Visipaque, Amersham, Piscataway, NJ) in the rest. An intrarenal bolus of 50 ␮g glyceryl trinitrate and 1 mL 1% lidocaine was injected just before stent dilation. A Perclose device (Abbott Vascular, Redwood, Calif) was

February 2008

JVIR

used successfully in all cases. All patients were receiving aspirin before stent placement, and clopidogrel was added immediately after stent placement and maintained for 1 month. Primary lesion crossing was achieved with a 0.035-inch soft angled hydrophilic guide wire (Terumo, Tokyo, Japan) and a 4-F Cobra Glidecath (Terumo). The hydrophilic guide wire was exchanged for a 0.014-inch steel Spartacore guide wire (Guidant) and, over this wire, an 8-F Veripath renal curved guiding catheter (Guidant) was advanced inside the renal trunk. No patients required dilation before stent deployment. The Spartacore wire was then exchanged for the Accunet embolic protection device. We used diameters ranging from 4.5 to 6.5 mm depending on the distal renal artery size. All the stents were stainless-steel balloon-expandable Herculink Plus stents (Guidant) deployed at nominal pressure (6 ⫻ 18 mm, n ⫽ 8; 7 ⫻ 18 mm, n ⫽ 6; 6.5 ⫻ 12 mm, n ⫽ 1). Postdeployment dilation of the aortic edge of the stent at 18 –20 atm was performed in 12 arteries. In the remaining three cases, no postprocedural dilation was performed as a result of aortic aneurysm. A final angiogram was obtained with manual injection of 6 – 8 mL of contrast medium to assess the results and detect complications. Pathologic Analysis of the Filter All microscopic analyses were performed by a senior pathologist (F.M.). The filter basket was excised from the steel guide wire and carefully set in a vase with the top of the basket down. The device was not washed, and saline solution was spread on the bottom of the vase without covering the filter to maintain humidity. The vase was then closed. Because the cholesterol crystals are dissolved by alcohol, alcohol contact with the basket was prevented. In the pathology laboratory, the recaptured filter basket was compressed between two slides (Fig 1) and examined with a microscope (Eclipse E-400; Nikon, Yokohama, Japan) under polarized light with a magnification of 20 – 400. Our endpoint was establishment of the presence or absence of cholesterol crystals. No quantitative measurement of cholesterol was per-

Volume 19

Number 2

Urbano et al

•

191

Figure 1. A filter has been compressed between two slides for pathologic analysis. Some macroscopic debris is visible.

Figure 2. Detail of the size relationship between crystals (arrow) and the pore (magnification, ⫻200).

formed. Although some macroscopic plaque emboli and other microscopic debris were visible in some cases, they were not the focus of the study. We studied the relationship between crystal size and filter membrane and pore size (Fig 2). Statistical Analysis Numeric data are expressed as means ⫾ SD. The paired Student t test was used for comparison. A P value less than .05 was considered significant.

RESULTS Technical success was achieved in 100% of cases. The 15 renal stents were deployed without problems through the Accunet system. The devices remained stable inside the renal arteries, with satisfactory blood flow through the filters in all angiographic controls (Fig 3). The filter devices did not induce any dissection, spasm, or thrombosis. The Accunet platinum markers were sufficient for detection with conventional fluoroscopy (Allura; Phillips, Best, The Netherlands). In two cases, we had difficulties with the advancement of the Accunet recovery catheter into the renal artery. In one of them, we curled the recovery catheter tip slightly and, after several attempts, we were able to recapture the filter. In the other case, advancement of the Accunet recovery catheter

Figure 3. Angiographic sample of how the filter device works. It is located in the main renal artery before the branch bifurcations. It is fully opened and there is an optimal seal between the vessel and the filter basket. Blood flow through the filter membrane is normal.

was not possible, but it was recaptured without further difficulty with a 5-F conventional Cobra catheter (Cordis) through the renal artery. With polarized light microscopy, we established the presence of cholesterol crystals, with their typical appearance in polarized light, in 12 of the 15 filters. Cholesterol crystals depicted were clear, pale, and polygonal, with

different sizes and forms (Fig 4). When there were superimposed crystals, or they were very large, a halo appeared as a consequence of light diffraction. With the 115-␮m diameter of the filter membrane pore used as a reference, we found a great variability in crystal size. The total amount of entrapped cholesterol in the filter was very variable as well. Usually, we find crystals

192

•

February 2008


JVIR

Figure 4. Two different cases of cholesterol microcrystals entrapped in the filters (magnification, ⫻100). (a) The tip of this filter is full of cholesterol crystals. (b) There are only a few crystals inside the filter membrane. We can see the typical appearance of a cholesterol crystal with polarized light microscopy. Changes in Serum Creatinine Level 30 Days after Stent Placement Serum Creatinine Level (mg/dL) Patient No.

Before Stent Placement

After Stent Placement

1 2 3 4 5 6 7 8 9 10 11

1.7 1.4 2 1.5 3.2 2.5 1.4 1.6 1.9 5.3 1.6

1 1.1 1.9 1.5 1.3 2.4 1.4 1.2 1.4 3.3 1.2

Figure 5. A great variability in crystal sizes can be appreciated. Whitish crystals, some of them smaller than the pore of the membrane (arrows), have been trapped by the system (magnification, ⫻200).

that are much smaller than the pore size inside the filter basket (Fig 5). Our work was not quantitative, so we could not determine the most common size of cholesterol emboli. No cholesterol was found in three of the filters. In one case, the stenosis had occurred in a kidney transplanted 4 months earlier. Although this patient had peripheral vascular disease, the cause of the stenosis seemed to be fibrous rather than atherosclerotic. In another case, a problem with filter manipulation prevented pathologic analysis. After stent placement, no acute re-

nal function impairment (defined by an increase of 25% in basal serum creatinine level) or hypertensive crisis was seen. The mean follow-up was 14 months (range, 9 –27 months) for nine patients (12 arteries), and two patients were lost to follow-up. Clinical data from the first month are available from all patients. Serum creatinine measurements before and 30 days after the procedure are shown in the Table. Mean serum creatinine levels were 2.31 mg/dL ⫾ 1.26 periprocedurally and 1.59 mg/dL ⫾ 0.67 after the procedure. A P value less than .05 indi-

cates a significant change in serum creatinine level before versus after stent placement.

DISCUSSION Ex vivo studies have demonstrated that cholesterol microembolization during renal stent placement triggers a vasculitis-like reaction. This reaction can cause substantial damage to the renal parenchyma. Although progressive worsening of renal function in the atherosclerotic kidney is complex and multifactorial (6,8), we believe renal cholesterol embolism is more frequent than commonly described and is therefore underdiagnosed. Vascular

Volume 19

Number 2

protection devices, which have been successfully used in other arterial vessels, could be effectively applied in the renal artery. Such application not only works to trap macroscopic thrombus and debris, but also to avoid potential damage to the renal parenchyma caused by cholesterol crystals not detected angiographically (10,17–19). When atheroembolism occurs in previously damaged kidneys with a low functional reserve, the clinical effects are more dramatic and the risk of irreversible injury is higher (8,20). It has been proven, mainly in the carotid artery, that distal antiembolic devices protect against thrombus and debris produced during angioplasty and stent implantation (21). However, little evidence is available regarding the effect of these devices against microcrystal cholesterol emboli. We have applied a simple technique to demonstrate that embolic protection devices not only protect against thrombus and debris, but also against cholesterol microcrystals. Polarization microscopy is an easily applicable, inexpensive, and readily available tool to detect cholesterol crystals trapped in embolic protection systems (22,23). The present study reveals that cholesterol microembolization of the renal parenchyma during renal artery stent placement in atheromatous high-grade stenosis is very common. Accordingly, it can be assumed that renal cholesterol embolization is underdiagnosed, and that its real prevalence during endovascular interventions is greater than the reported incidence of 1%–2% (15,18,24). Cholesterol crystals released during spontaneous or endovascular procedure–induced rupture of atheroma plaque go through the bloodstream, reaching the distal vessels and causing an inflammatory reaction akin to vasculitis (13,16). These microcrystals reach the arterioles and the glomeruli and cause patchy or diffuse renal damage derived from the amount of crystals released by the rupture of the plaque. Macrophages and T lymphocytes surround the cholesterol crystals and trigger an inflammatory reaction that may cause tubular atrophy and interstitial fibrosis with transient or permanent renal insufficiency (25). The clinical diagnosis of cholesterol embolization is difficult, with the exception of massive cases. In the acute presentation, there may be treatment-

Urbano et al

resistant hypertension, and oliguria/ anuria may appear within several days after endovascular intervention. The subacute/chronic presentation develops as a slow and progressive worsening in renal function, with intercalated periods of acute renal insufficiency. At present, there is no effective treatment for the cholesterol embolization syndrome and prevention is the best option (13,14). It is not the aim of the present study to show that renal stent implantation with protection improves clinical results, but the hypothesis can be raised that at least a percentage of the suboptimal responses to renal stent implantation could be explained by cholesterol embolism. With the use of device-protected stents, some authors have obtained better clinical results regarding renal function stabilization and/or improvement than those achieved with the conventional technique (8,20). No filter embolic protection device has been specifically designed for renal application (8,10). Our series demonstrates that the carotid artery Accunet device can be used in renal artery interventions; however, our sample is too small to completely confirm the safety of this device. The Accunet filter is qualified to protect the renal parenchyma from cholesterol crystals, and we have had no complications with its use. However, there are some disadvantages that should be solved by the industry before the development of a renal artery–specific device. The Accunet guide wire tip is not designed for primary crossing of a renal stenosis, and an ancillary guide wire would be needed to cross the stenosis before the system can be advanced into the renal artery. When we deploy a renal stent with the Accunet system, we have to increase the guiding catheter profile by 2 F (to 8 F). In addition, the Accunet recovery catheter is not suitable for renal application; in other words, to be an optimal tool, it should be precurved and more flexible. At this moment, the first drawback of all available filters is the long distance between the filter basket and the tip of the wire. The main renal artery is not always long enough to accommodate the balloon, the stent, and the filter. Another unsolved problem is the size of the pore, because cholesterol crystals can be as small as 10 –20 ␮m

•

193

and the average filter pore size is 100 – 120 ␮m (18). We have demonstrated crystals smaller than the pore size inside the basket. It will be interesting to know how crystals smaller than the pore size were captured in some of these cases; the existence of physicochemical interactions between the cholesterol crystals and the filter’s membrane may offer an explanation for this rather puzzling finding. These small particles can occlude the most distal renal arterioles, as demonstrated in experimental work with synthetic microspheres (26). In the case of cholesterol particles, the inflammatory reaction triggered is probably worse than the embolic occlusion. For this reason, not only the size of the crystals is relevant. It is also necessary to take into account the total load of cholesterol released. Another question pertains to whether the crystals go through any fragmentation inside the filter basket that affects the filter efficacy. To our knowledge, this has not been mentioned in any previous study, and it is also an unanswered question for us. A caveat should be also raised emphasizing the fact that the available devices may not achieve complete protection against cholesterol embolism. A suboptimal seal may occur between the vessel wall and the filter basket (27). A protection system based on temporary arterial balloon occlusion and blood aspiration should capture all kind of particles of any size, but some particles may remain unreachable below the balloon and will therefore be impossible to aspirate. Use of this system is also associated with a high risk of acute ischemic damage to the renal parenchyma (10,18). Manipulations in an atheromatous aorta while gaining access to the renal artery may be responsible for some additional cholesterol embolization, and obviously the antiembolic device will not prevent this. Because of the increased wire exchanges required for Accunet device deployment, the risk of cholesterol crystal release is higher during the first steps of the procedure. Some of the extra manipulations needed to deploy the Accunet system are no longer necessary with the new commercially available carotid protection devices, which are made with a lower profile and a shapeable tip that allows the primary crossing of the stenosis without an ancillary wire. The

194

•


higher amount of cholesterol crystals will be detached at the plaque fracture during stent deployment and stenosis dilation, and at this point the embolic protection device is already working. After stent implantation, we usually perform dilation of the aortic edge of the renal stent to greater than the nominal stent deployment pressure except in patients with abdominal aneurysm. This additional flaring of the renal ostium could release more cholesterol crystals, but if the filter device is in place, the risk of embolism will be under control. Postdilation could also contribute to increase the captured emboli in our patients. In conclusion, release of cholesterol crystals during renal angioplasty is probably a more frequent phenomenon than previously thought. Its clinical significance may be decisive for some patients. Herein we have shown a simple technique to prove the presence of cholesterol crystals inside the filter devices used during renal percutaneous transluminal angioplasty and stent placement.

6.

7.

8.

9.

10.

11.

12. References 1. Zieler RE, Bergelin RO, Isaacson JA, Strandness DE. Natural history of atherosclerotic renal artery stenosis: a prospective study with duplex ultrasonography. J Vasc Surg 1994; 19:250 – 257. 2. Caps MT, Zierler RE, Polissar NL, et al. Risk of atrophy in kidneys with atherosclerotic renal artery stenosis. Kidney Int 1998; 53:735–742. 3. Rao RK, Hood DB, Weaver FA. Current endovascular management of atherosclerotic renal artery stenosis. Surg Clin North Am. 2004; 84:1353– 1364. 4. Zeller T. Renal artery stenosis: epidemiology, clinical manifestation, and percutaneous endovascular therapy. J Interv Cardiol 2005; 8:497–506. 5. Kennedy DJ, Colyer WR, Brewster PS, et al. Renal insufficiency as predictor

13. 14.

15.

16.

of adverse events and mortality after renal artery stent placement. Am J Kidney Dis 2003; 42:926 –935. Zeller T, Frank U, Mu¨ller C, et al. Predictors of improved renal function after percutaneous stent-supported angioplasty of severe atherosclerotic ostial renal artery stenosis. Circulation 2003; 108:2244 –2249. Van Jaavrsveld BC, Krijnen P, Pieterman H, et al. The effect of balloon angioplasty on hypertension in the atherosclerotic renal artery stenosis. N Engl J Med 2000; 342:1007–1014. Holden A, Hill A. Renal artery revascularization with embolic protection in patients with ischemic nephropathy. Kidney Int 2006; 70:948 –955. Sapoval M, Zahringer M, Pattynama P, et al. Low-profile stent system for treatment of atherosclerotic renal artery stenosis: the GREAT trial. J Vasc Interv Radiol 2005; 16:1195–1202. Henry M, Henry I, Polydorou A, Polydorou AD, Hugel M. Recent advances in renal artery stenting. J Cardiovasc Surg 2007; 48:411– 442. Haller C. Arteriosclerotic renal artery stenosis: conservative versus interventional management. Heart 2002; 88: 193–197. Rivolta R, Bazzi C, Stradioti P, Paparella M. Stenting of renal artery stenosis: is it beneficial in chronic renal failure? J Nephrol 2005; 18:749 –754. Meyrier A. Cholesterol crystal embolism: diagnosis and treatment. Kidney Int 2006; 69:1308 –1312. Hirano Y, Ishikawa K. Cholesterol embolization syndrome: how to recognize and prevent this potentially catastrophic iatrogenic disease. Intern Med 2005; 44:209 –210. Fukumoto Y, Tsutsui H, Tsuchihashi M, Masumoto A, Takeshita A. The incidence and risk factors of cholesterol embolization syndrome, a complication of cardiac catheterization: a prospective study. J Am Coll Cardiol 2003; 42:211–216. Sijpkens Y, Westendrop R, Van Kemenade F, Van Duinem S, Breedveld F. Vasculitis due to cholesterol embolism. Am J Med 1997; 202:302–303.

February 2008

JVIR

17. Rajesh MD. Renal artery stenting with embolic protection. Endovasc Today 2006; 5:43–52. 18. Edwards MS, Corriere MA, Craven TE, et al. Atheroembolism during percutaneous renal artery revascularization. J Vasc Surg 2007; 46:55– 61. 19. Holden A, Hill AJ. Renal angioplasty and stenting with distal protection of the main renal artery in ischemic nephropathy: early experience. J Vasc Surg 2003; 38:962–968. 20. Henry M, Henry I, Klonaris C, et al. Renal angioplasty and stenting: long term results and the potential role of protection devices. Expert Rec Cardiovasc Ther 2005; 3:321–334. 21. Yadav JS, Wholey MH, Kuntz RE, et al Protected carotid-artery stenting versus endarterectomy in high-risk patients. N Engl J Med 2004; 351:1493– 1501. 22. Waugh DA, Small DM. Identification and detection of in situ cellular and regional differences of lipid composition and class in lipid-rich tissue using hot stage polarizing light microscopy. Lab Invest 1984; 51:702–714. 23. Harvey PR, Upadhya GA. A rapid, simple high capacity cholesterol crystal growth assay. J Lipid Res 1995; 36: 2054 –2058. 24. Beek FJA, Kaatee R, Beutler JJ, Van der Ven PJ, Mali W. Complications during renal artery stent placement for atherosclerotic ostial stenosis. Cardiovasc Intervent Radiol 1997; 20:184 –190. 25. Abela GS, Azik K. Cholesterol crystals cause mechanical damage to biological membranes: a proposed mechanism of plaque rupture and erosion leading to arterial thrombosis. Clin Cardiol 2005; 28:413– 420. 26. Kimura M, Suzuki T, Hishida A. A rat model of progressive chronic renal failure produced by microembolism. Am J Pathol 1999; 7:33–38. 27. Hagspiel KD, Stone JR, Leung DA. Renal angioplasty and stent placement with distal protection: preliminary experience with the Filter Wire EX. J Vasc Interv Radiol 2005; 16:125–131.