An Institutional Experience With Arterial Atheroembolism

An Institutional Experience With Arterial Atheroembolism Dirk S. Baumann, MD, Daniel McGraw, MD, Brian G. Rubin, MD, Brent T. Allen, MD, Charles B. Anderson, MD, and Gregorio A. Sicard, MD, St. Louis, Missouri

Seemingly minor blue-toe lesions resulting from atheroemboli are associated with unstable atherosclerotic plaques, which are at risk for causing recurrent emboli, tissue loss, and potentially death. At Washington University Medical Center, 62 patients (31 males and 31 females), ranging in age from 38 to 89 years (mean 62.8 _ 11.7 years), were treated for cutaneous manifestations of atheroembolic disease. Most patients (62%) had spontaneous bouts of atheroembolism, but 13 (21%) had recently undergone an inciting invasive radiologic study, 10 (16%) were on anticoagulation therapy, and one (2%) experienced abdominal trauma. In addition to the cutaneous manifestations, 18 patients (29%) also developed coincidental deterioration in renal function and four (6%) had intestinal infarction from atheroemboli. Arteriography in nearly all patients (97%) implicated the aorta and iliac arteries most commonly (80%), with the femoral (13%), popliteal (3%), and subclavian (3%) arteries less frequently incriminated. Forty-two patients underwent bypass grafting procedures (36 anatomic and six extra-anatomic) after exclusion of the native diseased artery, 20 patients had endarterectomies (six with additional bypass grafts), and five patients had no corrective vascular procedures. The 30-day operative mortality rate was 5% in this series. Nineteen patients (31%) required minor amputations, whereas two required major leg amputations. Thus limb salvage was possible in 86 of 88 (98%) limbs. No further episodes of atheroembolism occurred in the involved limbs during follow-up (1 to 53 months, mean 20.2 months). We advocate urgent arteriography and surgical correction or bypass with exclusion of the offending lesion. This aggressive approach results in maximal limb salvage, low operative mortality, and excellent long-term relief of embolization. (Ann Vasc Surg 1994;8:258-265.)

Dislodged atheromatous material from carotid artery bifurcation plaques is frequently the cause of cerebral vascular events. However, despite the common occurrence of ulceration of atheromatous plaques in the aorta and its large branches, atheroembolization to the viscera, kidneys, and lower extremities is rarely recognized clinically. Panum ~ initially described this condition in 1862, but it was 'not until Flory's 2 report in 1945 that this entity became widely known. He reviewed 267 autopsies with moderate-to-advanced erosion From the Department of Surgery, Division of Vascular Surgery, Washington University School of Medicine, St. Louis, 13110. Reprint requests: Gregorio Sicard, MD, 5103 Queeny Tower, Barnes Hospital, 1 Barnes Hospital Plaza, St. Louis, MO 63110.

258

o f atherosclerotic plaques within the aorta and demonstrated the presence of cholesterol crystalshaped clefts in the small arteries of these patients' kidneys, pancreases, and spleens. In addition, Flory 2 experimentally reproduced the histologic lesions by injecting atheromatous debris from a fresh h u m a n aortic specimen into rabbits. Other investigators have subsequently confirmed his hypothesis. 36 The clinical manifestations of atheroembolism are dictated by the level of the offending plaque and can be quite variable. Atheroemboli have been implicated in digital ischemic skin changes, 7 renal failure, s-I2 perioperative myocardial infarction, 13-'~ hemorrhage, ulceration, and necrosis of the stomach and bowel wall, ~6'~7 ophthalmic disease, ~ pulmonary dysfunction, ~9 and renal allograft failure. 2~ When multiple organ systems

Vol. 8, No. 3 1994

Institutional experience with arterial atheroembolism

259

are involved, atheroembolism can mimic systemic diseases such as polyarteritis nodosa 23 or systemic necrotizing vasculitis. 24 Atheroembolism has also been confused with anticoagulation-induced necrosisf ~ In fact, anticoagulation has been proposed as an etiologic factor in the development of this syndrome. 26-2s Because of the diverse presentation of atheroembolism, it is often misdiagnosed, leading to delays in treatment and increased morbidity. The diagnostic and therapeutic challenges presented by this syndrome, in conjunction with a recent marked increase in the recognition of this problem at Barnes Hospital/Washington University Medical Center, have prompted this review of our experience with atheroembolism.

however, 62 patients were seen with this problem. During this 4 8 - m o n t h period, 2230 aortic reconstructions and lower extremity revascularizations were performed. These latter 62 patients are the subject of this report. All patients had cutaneous manifestations of atheroembolic disease in the absence of a cardiac source of embolization. The presence of adequate proximal arterial inflow was determined by physical examination and/or arterial Doppler studies, and all patients who underwent arteriography had documented arterial lesions proximal to the involved site of embolization. Patients with perioperative iatrogenic atheroembolism were excluded, as were patients whose skin changes could be explained by other disease processes.

MATERIAL AND METHODS

Clinical P r e s e n t a t i o n

Clinical M a t e r i a l Between December 1985 and December 1987, only three patients were initially seen in the vascular service at Barnes Hospital with findings consistent with atheroembolism. During this time a total of 500 aortic reconstructive and lower extremity revascularization procedures were performed. From J an u a r y 1988 until December 1991,

There were 3 i male and 31 female patients ranging in age from 38 to 89 years (mean 62.8 _+ 11.7 years). Most patients had several risk factors for the development of atherosclerosis as outlined in Table I. In addition to these risk factors, m a n y of the patients had evidence of atherosclerotic involvement of the coronary, cerebral, and renal arteries (Table II). Twenty-six patients (42%)

T a b l e I. Risk factors for atherosclerosis in patients with atheroembolism Series

No. of patients

Smoking (%)

Hypertension (%)

Hyperlipidemia (%)

Diabetes meilitus (%)

221 72

100

61 69

50

11 20

22 15 8 62

68 100 88 78

82

23

75 71

38 38

Fine et al. 3~ Kazmier 3~

Dahlberg et al." Kwaan and Connolly45 Colt et al.46 Current series

27 0 25 31

Table II. Associated vascular diseases in patients with atheroembolism

Series Fine et al. 3~ Kazmier 3~

Dahlberg et al.44 Kwaan and Connolly*~ Colt et al..6 Current series

No. of patients

Angina (%)

221 72

20

Myocardial infarction (%)

Congestive heart failure (%)

Cerebrovascular disease (%)

Renal dysfunction (%)

24 56*

21

22 46

34 NA

NA

55 20

86 NA

13 7

63 32

22 15

NA

45* 0

8 62

19

100" 31

NA = not ava2abIe. *Combined coronary artery disease.

13

260

Annals of Vascular Surgery

B a u m a n n et aI.

had characteristic bilateral lower extremity skin changes (Fig. 1), 34 (56%) had skin manifestations confined to one lower extremity, and two (3%) had evidence of atheroembolization to an upper extremity. Eighteen patients (29%1 had deteriorating renal function coincident with the onset of lower extremity skin changes, suggesting renal atheroembolization. In addition to bilateral lower extremity involvement, four patients 16%) had cutaneous changes in the anterior abdominal walls, whereas four patients (6%) had intestinal infarction secondary to atheroemboli. Palpable arterial pulses were noted at the femoral level in 82 (98%) of the 84 affected lower extremities and in the brachial artery of the two affected upper extremities. Ankle/arm indexes were available for 58 of the 84 affected limbs and were 0.84 __ 0.24. In the two patients lacking pulses, the diagnosis was made on the basis of

classic cutaneous manifestations; in one of these patients the diagnosis was subsequently confirmed on pathologic examination of the amputated limb. Presumably these patients either had thrombosis of the iliac/femoral arteries at the site of the ulcerated plaque after embolization or more proximal lesions had embolization to the lower extremities via collateral pathways. Because patients with perioperative embolization were excluded from this study, most patients had a spontaneous onset o f embolization without any antecedent event. However, 13 patients (21%) had recently Iwithin 3 months} undergone an arteriographic procedure, and 10 patients (16%) were receiving anticoagulation therapy (coumadin in eight and heparin in two l at the time of embolization. One additional patient developed atheroembolic skin changes from the abdominal aorta to the lower extremities following blunt abdominal trauma.

Arteriography Arteriography was used to delineate the diseased segment of artery in order to generate a plan for operative correction of the vessel with a lesion. This procedure was performed in 60 of the 62 patients (97%). Of the two patients not undergoing arteriography, one had severe renal failure precluding arteriography and one refused further workup to determine the source of her emboli. The following vessels were identified by arteriography to be likely sources of atheroemboli (Table III): the aorta alone in 17 patients (28%), the aorta and iliac vessels in 13 (22%), the aorta, iliac, and femoral vessels in eight (13%), an isolated iliac artery in 10 (17%), a femoral vessel in eight (13%), a popliteal artery in two (3%), and a subclavian artery in two (3%). Of the cases involving the aorta, 29 (48% of total cases) involved only the inffarenal aorta, whereas six (10%) involved the suprarenal abdominal aorta as well. Three patients (5%) had extensive thoracic and abdominal aortic disease. Ulcerated plaques

Table III. Sources of atheroemboli

Fig. 1. Lower extremity skin changes in a patient with atheroemboli. Note characteristic mottled appearance of the fifth toe and plantar surface of the foot (arrow). These ischemic areas are typically tender and have poor capillary refiI1.

Aorta Aorta and iliac arteries Aorta, iliac, femoral arteries Iliac arteries Femoral arteries Popliteal arteries Subclavian arteries

28% 22% 13% 17% 13% 3% 3%

Vol. 8, No. 3 t994

(Fig. 2) were noted on arteriography in the majority of patients (50/60 or 83%). Aneurysmal disease of the aorta was found coincidentally in 12 patients (19%), but there was no evidence of aneurysmal disease at any site other than the aorta in this group. Treatment

Except w h e n precluded by the patient's moribund condition or the patient's refusal to undergo an operation, sources of atheroemboli were either bypassed and excluded from the vascular system or treated by endarterectomy. Sixteen aortobiiliac, six aortobifemoral, 12 femoral-popliteal, and two carotid-subclavian bypass grafts were constructed after the exclusion of the offending lesion. Endarterectomies of the aorta alone (seven), aorta and iliac arteries (two), aorta, iliac, and femoral arteries (one), iliac arteries alone (six), iliac and femoral arteries (two), femoral arteries alone (one), and popliteal arteries (one) were performed to eliminate the origin of embolization. Six patients underwent both an endarterectomy procedure and placement of a bypass graft. A thrombectomy was performed to remove athero-

Institutional experience with arterial atheroembolism

261

matous debris in the one patient who had embolization following blunt abdominal trauma. Six patients were deemed medically unfit for major vascular reconstructive surgery and instead underwent extra-anatomic bypass procedures with axillofemoral and femorofemoral grafts and ligation of the proximal common femoral arteries. Two of these patients had proximal iliac artery sources, and the remainder had aortic ulcerations. Of these, three already had renal failure and were on hemodialysis. Nineteen patients (31%) required minor amputations at the transmetatarsal level or distally. One patient (2%) required below-knee amputation and one (2%) above-knee amputation. All amputations were performed for the treatment of tissue infarction from atheroemboli occurring prior to surgical intervention. Five patients did not undergo corrective vascular procedures. Three patients, one of w h o m was comatose on presentation, were deemed extremely poor surgical candidates and did not undergo reconstructive procedures. A fourth patient had embolization after being placed on warfarin therapy following coronary artery bypass grafting and aortic valve replacement. She suffered no subsequent embolization and refused further evaluation. The fifth patient had a marked affective disorder several months after coronary artery bypass surgery and likewise requested no definitive procedure be performed. These patients underwent only minor amputations (three toe amputations and one transmetatarsal amputation) or confirmatory skin and muscle biopsies (one). Of these patients, two died of recurrent embolization during their initial hospitalization or shortly thereafter. One patient already on hemodialysis did not have clinical evidence of further embolization during the subsequent 38 months. Follow-up data were not available for the remaining two patients. RESULTS

Fig. 2. Aortic arteriogram demonstrating diffuse ulceration of the abdominal aorta in a patient with bilateral lower extremity skin changes.

The 30-day perioperative mortality rate was 5% (3/62). One additional patient died 74 days postoperatively but during the same hospitalization. All four of the patients who died had renal involvement on presentation to the hospital. The cause of death in these patients was cardiac failure in three and overwhelming sepsis from visceral infarction in one. There were no episodes of embolizafion intraoperatively or postoperatively in patients whose sources of emboli were excluded from the vascular system. Operative

262

Annals of Vascular Surgery

B a u m a n n et al.

100%

52

48

\is

80%

,> -f

E

60% 40%

o 13_

20% 0% 0

i

i

i

i

i

10

20

30

40

50

Months Fig. 3. Survival curve of 62 patients treated for atheroembolism at Washington University Medical Center from January 1988 to December 1991.

morbidity, other than the preoperative extremity morbidity requiring selective amputation, included reversible renal dysfunction from acute tubular necrosis in three patients (5%), one of w h o m required temporary dialysis, transient spinal shock, which completely resolved in one patient (2%), a deep venous thrombosis in one patient (2%), and a cerebrovascular accident in one patient (2%) postoperatively. Histologic examination was performed on specimens from nine patients. One was a premortern skin biopsy taken from a patient who presented in extremis. It showed evidence of cholesterol embolization with characteristic intraluminal clefts in 100 to 200 p,m vessels surrounded by endothelial and myofibroblastic proliferation, with occasional round cells and multinucleate foreign body giant cells. Histologic examination of two nephrectomy specimens also demonstrated cholesterol emboli. The remaining six examinations were performed on lower extremity amputation specimens, five (83%) of which were positive for cholesterol emboli. At follow-up (mean = 20.2 months, range 1 to 53 months), none of the patients was found to have further episodes of atheroembolization or additional tissue loss from the distribution of the excluded arterial lesion. One patient initially treated for upper extremity atheroembolization from a subclavian lesion required iliac bypass grafting for subsequent atheroembolization to the right lower extremity from an ulcerated iliac plaque. Of the six patients who underwent extraanatomic bypass, one died of cardiac failure during the initial hospitalization. There were no further episodes of peripheral or visceral embo-

lization in these patients, and none had further deterioration of renal function (three were already on hemodialysis) during follow-up (mean 12 months). For all patients with atheroembolization, the survival rate was 83% at 1 year, 79% at 2 years, and 75% at 3 years (Fig. 3). Late death was most commonly caused by myocardial events, with cerebral vascular accidents accounting for the remainder of the deaths for which a cause was known.

DISCUSSION Atheroembolism can present in a wide variety of ways, often making its diagnosis difficult.29 Kazmier 3~has proposed subdividing disseminated cholesterol embolization into three major clinical presentations: (1) a peripheral syndrome, (2) a renal syndrome, and (3) a visceral syndrome. These syndromes most commonly present separately, but some patients m a y have a combination of two or three of them. In their review of 221 cases of histologically proven atheroembolism in the English literature from 1965 to 1985, Fine et al? 1 found the incidence of cutaneous manifestations to be 34%. Peripheral cholesterol embolization is characterized by livedo reticularis, painful toe ulceration, and eventual infarction which, although bilateral w h e n the source is aortic, is generally asymmetric. Because of the small particulate nature of the emboli, proximal extremity pulses are typically palpable, as they were in our patients. In the absence of proximal arterial occlusive disease, atheroemboli must be distinguished from other causes of distal ischemia such as acrocyanosis, connective tissue diseases, and frostbite from previous exposure to cold weather. Typically, at the time of presentation the effects of the emboli are permanent. Therefore treatment of peripheral embolization can only be directed toward prevention of further embolization. Elimination of ongoing embolization will allow the development of unaffected collateral vessels and eventual healing in the presence of sufficient inflow. The kidney is the visceral organ most often affected by cholesterol emboli, which were present in 75% of the postmortem specimens in the collective review of Fine et al. 31 In Kealy's32 review of 2126 autopsies, he found 16 cases of atheroembolization. Of these, all but three had renal involvement. Renal biopsies in living patients with u n e x p l a i n e d deterioration of renal function demonstrate an incidence of cholesterol embolization of 1%.33 In our series, 18 of 62

Vol. 8, No. 3 1994

patients (29%) had renal involvement. In cases of renal cholesterol embolization, renal function usually declines in a progressive stepwise manner, although a history of this type of deterioration cannot always be elicited. Hypertension is a frequent but not universal finding in patients w h o have had atheroemboli travel to their kidneys. This association between atheroembolism and humoral renovascular hypertension was first made by Handler 34 in 1956. Subsequently Dalakos et al. 35 documented elevated renin levels in a patient with malignant hypertension subsequently found to have renal atheroembolization at autopsy. Although establishing the diagnosis following an antecedent event such as cardiac catheterization, arteriography, or surgery may be straightforward, it is often clinically impossible to distinguish hyperreninemic hypertension induced by atheroemboli from that caused by renal artery occlusive disease. Indeed, in our series, of the 18 patients with renal involvement, only eight (44%) had a clear antecedent inciting event. The remaining patients had renal dysfunction as a result of spontaneous atheroembolic disease. Likewise laboratory studies, although they are advocated by some, are frequently not diagnostic. Urinalysis is not characteristic?' Eosinophilia on a peripheral blood smear is commonly seen with renal atheroembolism (in up to 80% of patients) but is transient, lasting only a few days. m In the current series, 4 of 18 (22%) patients with renal deterioration had mildly elevated blood eosinophil counts ( > 3% of total leukocytes). Patients with embolization to the kidneys have a much poorer prognosis than those with only embolization to the extremities. In this series all perioperative deaths occurred in patients w h o presented with renal dysfunction as a result of atheroembolic disease. McGowan and Greenburg 8 suggested that the prognosis of cholesterol atheroembolic renal disease is better w h e n it arises after anticoagulation or arteriography rather than de novo. Cessation of anticoagulation therapy has been reported to allow improvement in renal function. However, as noted by these investigators, 8 this observation m a y only reflect a selection bias. Mild cases of atheroembolization are more likely to be detected w h e n they occur in temporal proximity to such procedures than w h e n they occur spontaneously away from medical supervision. Visceral embolization is the rarest syndrome associated with atheroembolic disease. In a collective review by Fine et al., 3~ 31% of patients had

Institutional experience with arterial atheroembolism

26 3

histologic evidence of gastrointestinal tract atheroemboli, although only 10% had clinical evidence of gastrointestinal involvement. In our report, four patients (6%) presented with intestinal ischemia/infarction from embolic events. Visceral embolization in one of these patients resulted in terminal septic shock. That atheroembolization may occur after crossclamping of rigid atherosclerotic vessels has been well recognized since the beginnings of vascular surgery. = It has long been appreciated as w e l t that arteriography can incite cholesterol embolization. 36'37 Similarly, as noted in one patient in this series, abdominal trauma can also mechanically disrupt atherosclerotic plaques, releasing cholesterol crystals into the bloodstream. In the review by Fine et al., 31 nearly 20% of patients had recently had a vascular radiologic procedure prior to the onset of their atheroembolic events. At the Mayo Clinic, Kazmier 3~has reported that approximately 25% of their patients are seen following an arterial catheterization procedure. Thirteen patients (21%) in our series had recently undergone arteriography. This represents 0.06% of the approximately 23,200 arteriographic procedures performed at Barnes Hospital during the study period. In 1967, Moldveen-Geronimus and Merriam 26 first suggested that anticoagulation does not halt the embolization process and may in fact exacerbate it. This premise has been substantiated by other investigators. 25"27"28However, the exact role anticoagulation plays in the pathogenesis of atheroembolism remains unclear. Moldveen-Geronimus and Merriam 26 proposed that cholesterol emboli arise from ulcerated plaques lacking an overlying thrombus covering. Therefore, by preventing the formation of a protective thrombus, anticoagulants could aggravate or actually initiate the atheroembolic process. An alternative hypothesis is that atheroembolization occurs during the disruption of a complex plaque caused by hemorrhage into the plaque. Anticoagulation could potentiate this process by exacerbating the hemorrhage. The relative risk posed by anticoagulation therapy is difficult to determine because of the widespread use of heparin and warfarin in modern medical practice. Certainly the 10 patients in our series who were seen while receiving anticoagulation therapy represent only a small fraction of the patients treated with anticoagulation at Washington University Medical Center. In patients presenting with symptoms suggestive of atheroembolization, other potential causes of these symptoms must be excluded. A detailed

264

Annals of Vascular Surgery

Baumann et al.

history and complete physical examination can intimate other causes such as coagulopathy, endocarditis, Raynaud's disease, or vasculitis. Echocardiography will identify mural thrombi and valvular lesions, which can be responsible for similar symptom complexes. Anticoagulation therapy should be discontinued whenever feasible? 7 Cessation of anticoagulation may improve renal function 27 and allow improvement in the cutaneous manifestations of the syndrome. 2s Treatment of the cutaneous, renal, and visceral manifestations of the atheroembolism syndrome is primarily surgical. No satisfactory medical therapy exists for true cholesterol embolization. Controversy in the management of unilateral "blue-digit syndrome" has been spawned in recent years by several reports in the radiology literature. 3s'39 Kumpe et al? s have recommended immediate percutaneous transluminal angioplasty, whereas Brewer et al? 9 favor angioplasty following interim (6 to 12 weeks) anticoagulant and/or antiplatelet therapy for the treatment of unilateral blue-toe syndrome. However, both groups were treating patients with focal, unilateral stenoses or occlusions of arteries distal to the aortic bifurcation. These types of lesions act as niduses for the formation of adherent fibrinoplatelet debris, which can subsequently embolize. Therefore, as stated by Kumpe et al., 3s these groups were in fact treating fibrinoplatelet emboli arising from stenotic arterial lesions. It is important to distinguish this type of embolus arising from stenotic lesions from that emanating from ulcerative lesions. Cholesterol emboli created by the latter should not be treated with anticoagulation or angioplasty because of the risk of further embolization. The surgical treatment of cholesterol embolization is predicated on the identification of the offending lesion by biplanar arteriography. 4~ The ulcerative plaques giving rise to these emboli are often found along the posterior wall of the vessel, thus necessitating lateral or oblique views to ensure their identification. In patients with unilateral lower, extremity symptoms, access to the arterial tree should be obtained via the contralateral lower extremity so as to limit the risk of embolization from the arteriographic procedure. In patients with a presumed aortic source, catheterization via a brachial artery approach will result in a decreased incidence of catheter-related embolization. Following arteriography and identification of the source of the emboli, the patient should undergo surgery to remove or exclude the offending lesion from the vascular tree. For lesions of

the aorta, most patients should undergo aortoiliac or aortofemoral reconstruction with exclusion of the diseased vessel. If dictated by the patient's condition, extra-anatomic bypass with exclusion may be an acceptable alternative. 4a For lesions distal to the aorta, bypass or endarterectomy has proved satisfactory in our hands. Mehigan and Stoney 42 first emphasized the ability of atheromatous plaques of the iliac, popliteal, and especially the femoral arteries to serve as sources of emboli and the efficacy of endarterectomy in their treatment. Patients with multiple suspicious lesions at different levels should be treated with the least morbid procedure initially, because it is impossible to accurately ascertain the exact source of the emboli in such circumstances. In the absence of flank gangrene, toe lesions may be treated expectantly following revascularization, inasmuch as the majority will improve. 43 Likewise, following surgical treatment, renal dysfunction may be ameliorated. 8 Segmental bowel resection may be required in patients with visceral embolization.

CONCLUSION In summary, the syndrome of cholesterol embolization must be differentiated from other systemic diseases that it may mimic, as well as from other embolic phenomena. All anticoagulation therapy should be discontinued and arteriography undertaken in a timely manner. On identification of the offending lesion, prompt surgical correction as outlined above should be performed. Proper correction will minimize the morbidity and mortality of this increasingly common manifestation of atherosclerosis, as demonstrated by the results of our analysis. REFERENCES 1. Panum PL. Experimentelle Beitrage zur Lehre yon der Emboll Virchows Arch Pathol Anat 1862;25:308-310. 2. Flory CM. Arterial occlusions produced by emboli from eroded aortic atheromatous plaques. Am J Pathol 1945;21: 549-565. 3. Otken LB. Experimental production of atheromatous embollzation. AMA Arch Pathol 1959;68:685-689. 4. Snyder HE, Shapiro JL. A correlative study of atheromatous embolism in h u m a n beings and experimental animals. Surgery 1961;49:195-204. 5. Gore L McCombs HL, Lindquist RL. Observations on the fate of cholesterol emboli. J Atheroscler Res 1964;4:527-535. 6. Warren BA, Vales O. The ultrastructure of the reaction of arterial walls to cholesterol crystals in atheroembolism. Br J Exp Pathol 1976;57:67-77. 7. Karmody AM, Powers SR, Monaco VJ, et al. "Blue toe" syndrome. Arch Surg 1976;111:1263-1268.

Vol. 8, No. 3 1994

8. McGowan JA, Greenburg A. Cholesterol atheroembolic renal disease. A m J Nephrol 1986;6:135-139. 9. M a s h i a h A, Pasik S, Hurwitz N. Massive a t h e r o m a t o u s emboll to both kidneys: A fatal complication following aortic surgery. J Cardiovasc Surg 1988;29:60-62. 10. K a s i n a t h BS, Lewis EJ. Eosinophilia as a clue to the diagnosis of atheroembolic renal disease. Arch Intern Med 1987;147: 1384-1385. 11. Zatuchni J, Patel HK, C h i e m c h a n y a S. The "blue toe" syndrome with renal atheroembolism and failure. Angiology 1985;36:209-214. 12. Kassirer JP. Atheroembolic renal disease. N Engl J Med 1969;280:812-818. 13. Heggtveit HA. Coronary atheroembolism. N Engl J Med 1984;310:722. 14. Fitzgibbon GM, Keon WJ. Atheroembolic perioperative infarction during repeat coronary bypass surgery: Angiographic d o c u m e n t a t i o n in a survivor. A n n Thorac Surg 1987;43:218219. 15. Keon WJ, Heggtveit HA, Leduc J. Perioperative myocardial infarction caused by atheroembolism. J Thorac Cardiovasc Surg 1982;84:849-855. 16. Forouhar FA, Mohit M, Gardner P, et al. Cholesterol embolism causing bleeding gastric ulcers. A n n Clin Lab Sci 1988; 18:260-265. 17. Anderson WR, Richards AM, Weiss L. Hemorrhage a n d necrosis of the s t o m a c h and bowel due to atheroembolism. A m J Clin Pathol 1967;48:30-38. 18. Coppeto JR, Lessell S, Lessell IM, et al. Diffuse disseminated atheroembolism. Arch Ophthalmol 1984;102:225-228. 19. Weigent CE. Pulmonary atheroembolism complicating repair of an atherosclerotic abdominal aneurysm. M i n n Med 1978; 61:15-16. 20. Gibbons GW, Madras PN, Wheelock FC, et al. Aortofliac reconstruction following renal transplantation. Surgery 1982; 91:435-437. 21. Aujila ND, Greenberg A, Banner BE et al. Atheroembolic involvement of renal allografts. A m J Kidney Dis 1989;I3: 329-332. 22. Thurlbeck WM, C a s t l e m a n B. Atheromatous emboli to the kidneys after aortic surgery. N Engl J Med 1957;257:442-447. 23. Richards AM, Eliot RS, K a n j u h VI, et al. Cholesterol embolism. A m J Cardiol 1965;15:696-707. 24. Young DK, Burton ME, H e r m a n JH. Multiple cholesterol emboli syndrome simulating systemic necrotizing vasculitis. J Rheumatol 1986;13:423-426. 25. Zaytsev P, Miller K, Pellettiere EV. Cutaneous cholesterol emboli with infarction clinically mimicking heparin necros i s - a case report. Angiology 1986;37:471-476. 26. Moldveen-Geronimus M, Merriam JC. Cholesterol embolization: From pathological curiosity to clinical entity. Circulation 1967;35:946-953. 27. Bruns F J, Segel DP, Adler S. Control of cholesterol embolization by discontinuation of anticoagulant therapy. A m J Med Sci 1978;275:105-108.

Institutional experience with arterial atheroembolism

265

28. H y m a n BT, Landas SK, A s h m a n RF, et al. Warfarincelated purple toes syndrome a n d cholesterol microembolization. A m J Med 1987;82:1233-1237. 29. Scully RE, Mark E J, McNeely B. Case records of the Massachusetts General Hospital (case 30-1986). N Engl J Med 1986;315:308-315. 30. Kazmier FJ. Shaggy aorta syndrome and disseminated a t h e r o m a t o u s embolization. In Bergan 3J, Yao JST, eds. Aortic Surgery. Philadelphia: WB Satmders, 1989, pp 189194. 31. Fine MJ, Kapoor W, Falanga V. Cholesterol crystal embolization: A review of 221 cases in the English literature. Angiology 1987;38:769-784. 32. Kealy WE. Atheroembolism. J Clin Pathol 1978;31:984-989. 33. Jones DB, Iannaccone PM. Atheromatous emboli in renal biopsies. A m J Pathol 1975;78:261-271. 34. Handler FP. Clinical a n d pathologic significance of atherom a t o u s embolization, with e m p h a s i s on an etiology of renal hypertension. A m J Med 1956;20:366-373. 35. Dalakos TG, Streeten DHP, Jones D, et al. " M a l i g n a n t " hypertension resulting from atheromatous embolization predominantly of one kidney. A m 3 Med 1974;57:135-138. 36. Pollitt J, Lee BM. Renal failure from atheroembolism after t r a n s l u m b a r aortography. 3 A m Geriatr Soc 1971;19:989-995. 37. Ramirez G, O'Neill WM, Lambert R, et al. Cholesterol e m b o lization: A complication of angiography. Arch Intern Med 1978;138:1430-1432. 38. K u m p e DA, Zwerdlinger S, Griffin D J. Blue digit syndrome: Treatment with percutaneous transluminal angioplasty. Radiology 1988;166:37-44. 39. Brewer ML, Kinnison ML, Perler BA, et al. Blue toe syndrome: Treatment with anticoagulants a n d delayed percutaneous transluminal angioplasty. Radiology 1988;166:31-35. 40. Kempczinski RF. Lower-extremity arterial emboli from ulcerating atherosclerotic plaques. JAM_& 1979;24k807-810. 41. K a u f m a n JL, Stark K, Brolin RE. Disseminated atheroembolism from extensive degenerative atherosclerosis of the aorta. Surgery 1987;102:63-69. 42. M e h i g a n JT, Stoney RJ. Lower extremity a t h e r o m a t o u s embolization. A m J Surg 1976;132:163-167. 43. Fisher DF, Clagett GP, Brigham RA, et al. Dilemmas in dealing with the blue toe syndrome: Aortic versus peripheral source. A m d Surg 1984;148:836-839. 44. Dahlberg P J, Frecentese DF, et al. Cholesterol embolism: Experience with 22 histologically proven cases. Surgery 1989; 105:737-746. F 45. K w a n n JHM, Connolly JE. Peripheral embolism, a n enigma. .~rch Surg 1977;i12:987-990. 46. Colt HG, Begg RJ, Saporito J J, et al. Cholesterol emboli after cardiac catheterization: Eight cases and a review of the literature. Medicine 1988;67:389-400.