Natural history of atherosclerotic renal artery stenosis: A prospective study with duplex ultrasonography

Transcription

1 Natural history of atherosclerotic renal artery stenosis: A prospective study with duplex ultrasonography R. Eugene Zierler, MD, Robert o. Bergelin, MS, Janette A. Isaacson, RVT, and D. Eugene Strandness, Jr., MD, Seattle, Wash. Purpose: Although the prevalence of renal artery stenosis in patients with peripheral arterial disease is in the range of 30% to 40%, the role of renal revascularization in patients without severe hypertension or kidney failure is controversial. Duplex scanning is a noninvasive technique that is ideally suited for screening and follow-up of renal artery disease. The purpose of this study was to document the natural history of renal artery stenosis in patients who were not candidates for immediate renal revascularization. Methods: Eighty-four patients with at least one abnormal renal artery detected by duplex scanning were recruited from patients being screened for renal artery stenosis. Of the 168 renal artery/kidney sides, 29 were excluded (15 prior interventions, 6 nondiagnostic duplex scans, 8 presumed nonatherosclerotic lesions), leaving 80 patients with 139 sides for the follow-up protocol. Renal arteries were classified as normal, less than 60% stenosis, 60% or greater stenosis, or occluded by use of previously validated criteria. Results: The study group included 36 men and 44 women with a mean age of 66 years who were monitored for a mean interval of 12.7 months. The initial status of the 139 renal arteries was normal in 36, less than 60% stenosis in 35, 60% or greater stenosis in 63, and occluded in 5. Although none of the initially normal renal arteries showed disease progression, the cumulative incidence of progression from less than 60% to 60% or greater renal artery stenosis was 23% ± 9% at 1 year and 42% ± 14% at 2 years. All four renal arteries that progressed to occlusion had 60% or greater stenoses at the initial visit, and for those sides with a 60% or greater stenosis, the cumulative incidence of progression to occlusion was 5% ± 3% at 1 year and 11% ± 6% at 2 years. The mean decrease in kidney length associated with progression of renal artery stenosis to occlusion was 1.8 cm. Conclusions: Progression of renal artery stenosis, as defined in this study, occurs at a rate of approximately 20% per year. Progression to occlusion is associated with a marked decrease in kidney length. Whether this natural history can be improved by earlier intervention for renal artery stenosis remains to be determined. (J VAse SURG 1994;19:250-8.) Renal artery stenosis is the most common cause of secondary hypertension. The true prevalence of renovascular hypertension is still controversial, but estimates range from 1 % to 6% in unselected adult patients with hypertension. 1 Atherosclerosis is the most frequent renal artery lesion in patients over the age of 50 years, with fibromuscular hyperplasia more From the Department of Surgery, Division of Vascular Surgery, University of Washington, Seattle. Presented at the Forty-seventh Annual Meeting of the Society for Vascular Surgery, Washington, D.C., June 8-9, Reprint requests: R. Eugene Zierler, MD, Department of Surgery, RF-25, University of Washington, Seattle, WA Copyright 1994 by The Society for Vascular Surgery and International Society for Cardiovascular Surgery, North American Chapter /94/$ /6/ common in patients under the age of 40. Renovascular disease is also recognized as an important cause of ischemic kidney failure, with or without hypertension. 2 Atherosclerotic renal artery stenoses are especially prevalent in patients with evidence of atherosclerosis involving other segments of the arterial circulation. In a study of 395 consecutive patients undergoing arteriography, a renal artery stenosis of greater than 50% was found in 38% of patients with abdominal aortic aneurysms, 33% of patients with aortoiliac occlusive disease, and 39% of patients with lower extremity occlusive disease. 3 Although some patients with atherosclerotic renal artery disease have clinical signs or symptoms that suggest a renovascular problem, many patients who are found to have renal artery stenoses do not have

2 JOURNAL OF VASCULAR SURGERY Volume 19, Number 2 Zierler et al. 251 either severe hypertension or impaired renal function. The principal issue in the management of these uncomplicated renal artery stenoses is whether the long-term outcome can be improved by early renal revascularization. To make rational therapeutic decisions, it is essential to have accurate information on the natural history of renal artery stenosis. Such data are essential for evaluating the efficacy of therapeutic interventions, particularly in cases where the underlying condition is largely asymptomatic. Noninvasive testing methods are ideally suited for the screening and follow-up examinations that are necessary in natural history studies. In most segments of the arterial system both indirect and direct noninvasive diagnostic tests can be used. However, because none of the indirect techniques are capable of detecting renal artery stenosis, it was not until the development of abdominal duplex scanning that such lesions could be reliably identified by noninvasive means.4-7 The clinical applications of renal duplex scanning include screening of selected patients with hypertension, screening of selected patients with kidney failure, follow-up of renal revascularization (percutaneous transluminal angioplasty or bypass), evaluation of kidney transplants, and natural history studies The purpose of this study was to document the natural history of renal artery stenosis in patients who were not considered candidates for immediate renal revascularization. PATIENTS AND METHODS Patient recruitment and exclusions. During the period from January 1990 through December 1992, 84 patients were recruited for a prospective research study on the natural history of renal artery stenosis. These patients were obtained from referrals to the Vascular Diagnostic Service at the University of Washington Medical Center. The most common reasons for renal artery screening were hypertension and decreased renal function. Patients who had at least one abnormal renal artery detected by duplex scanning and who were not considered candidates for immediate surgical or radiologic intervention were eligible for inclusion. As noted below, the duplex scanning criterion for minimal renal artery disease is a peak systolic velocity (PSV) of at least 180 cm/sec. Because the research protocol required evaluation of both renal arteries in each patient, the recruitment of patients with unilateral renal artery disease resulted in the follow-up of some normal renal arteries. Informed consent was obtained with approval from the Human Subjects Review Committee at the U niversity of Washington. Because this was strictly a natural history study, renal arteries that had undergone prior surgical or radiologic interventions were excluded. Of the 168 potential renal artery sides in the recruited patients, 15 were excluded because of prior interventions, and six sides were excluded for nondiagnostic duplex scanning results. In addition, eight sides in four patients were excluded because the presumed pathologic lesion responsible for renal artery stenosis was fibromuscular dysplasia and not atherosclerosis. This left 80 patients with 139 renal arteries for inclusion in the serial follow-up protocol. All eligible patients who consented to participate in the study underwent a baseline duplex scan in the research laboratory to confirm their renal artery disease status. Subsequent evaluations were performed at 6-month intervals for patients with at least one high-grade (;::: 60% diameter reduction) renal artery stenosis and at 12-month intervals for patients with less severe renal artery lesions. Clinical questionnaire and examination. At each baseline and follow-up visit, patients were interviewed with a standardized questionnaire. Clinical information obtained during the interview included patient demographic data, risk factors for atherosclerosis (history of smoking, diabetes, hypercholesterolemia), symptoms or signs of atherosclerotic disease at other sites (coronary, carotid, and lower extremity arteries), and any interventions for treatment of arterial disease. The duration and treatment of hypertension, including medication changes, were also documented. Blood samples for blood urea nitrogen (BUN) and creatinine determinations were drawn at the baseline visit and at yearly intervals thereafter. Mter the patient interview, blood pressures were recorded in both arms with the patient in the supine position, and the higher of the two arm blood pressure measurements was used in the analysis. Finally, renal artery duplex scanning was performed. Technique of renal duplex scanning. Renal duplex scanning was scheduled in the morning after an overnight fast to minimize attenuation of ultrasound transmission by bowel gas. All examinations were performed by the same technologist (J.A.I.) using an ATL Ultramark 9 duplex ultrasound scanner (Advanced Technology Laboratories, Bothell, Wash.). Although the use of a single technologist could introduce some bias, it eliminates any intertechnologist variability in the performance of the scans. The technique of renal artery duplex scanning has been described in detail previously.4-6 In summary, patients were examined in the supine position

3 252 Zierter et at. JOURNAL OF VASCULAR SURGERY February 1994 Table I. Diagnostic criteria for classification renal artery disease by duplex scanning Renal artery diameter reduction Renal artery PSV RAR Normal <60% ~60% Occlusion < 180 em/sec 2: 180 em/sec < or ~ 180 em/sec No signal <3.5 <3.5 ~3.5 No signal RAR, Renal to aortic ratio (ratio of maximum peak systolic velocity in the renal artery to the peak systolic velocity in the aorta). If aortic peak systolic velocity is less than 40 em/sec, the RAR cannot be used, and identification of a 60% or greater renal artery stenosis is based on the finding of a focal high-velocity jet and poststenotic turbulence. with the 2.25 MHz phased array and 3 MHz mechanical transducers. Initially, the aorta was examined in longitudinal view to identify the presence of either aneurysmal or occlusive disease. The aortic PSV was measured at the level of the superior mesenteric artery with a large sample volume. The right renal artery origin was then identified in a cross-sectional view of the aorta, and velocities were recorded from the origin, proximal, middle, and distal segments of the artery. Similarly, the origin of the left renal artery was identified, and velocities were recorded from the origin, proximal, middle, and distal segments. The angle of insonation for all renal artery velocity measurements was 60 degrees or less. Maximum kidney length was measured with a horizontal kidney image obtained from either a flank or prone approach. The specific scan head position and approach used were documented so that the same view could be used for length measurement at each follow-up visit. To further reduce the variability of kidney length measurements, three length measurements were averaged. Interpretation of renal duplex scans. The degree of renal artery stenosis was classified by duplex scanning according to previously validated criteria based on the maximum renal artery peak systolic velocity (PSV) and the renal-to-aortic ratio (RAR), which is defined as the ratio of the maximum PSV in the renal artery to the PSV in the adjacent abdominal aorta. 6 These criteria permit classification of renal artery narrowing into the four categories listed in Table I: normal; less than 60% diameter stenosis; 60% or greater diameter stenosis; and occluded. Although renal arteriography was not required, the approach to renal duplex scanning in this study was identical to that used in previous validation studies performed in our laboratory.4.6 For the purposes of this study, the most severe disease found among the origin, proximal, middle, or distal segments of a renal artery established the disease category for that side. Disease progression was defined as an increase in stenosis severity to 60% or greater diameter reduction or the development of total renal artery occlusion, as documented by serial duplex scanning. Although the renal duplex criteria include normal and less than 60% stenosis categories, only disease progression to high-grade stenosis or occlusion was considered significant in this study. Renal arteries that were classified as normal or less than 60% stenosis at the baseline visit showed disease progression if they were found to have 60% or greater stenosis or occlusion on subsequent evaluations. Those arteries with 60% or greater stenosis at baseline could only progress to occlusion. If a renal artery was occluded at the baseline visit, no further disease progression was possible. To confirm progression of a renal artery lesion by duplex scanning, evidence of progression on two consecutive follow-up evaluations was ordinarily required. However, as noted in the Results section, because of the long time interval between evaluations, this was not possible in all cases. To assess changes in kidney length, the length measurement at the last follow-up visit was compared with the measurement at the baseline visit. Changes in kidney length of greater than 1. 0 cm were considered to be significant. This threshold is based on a variability study performed in our laboratory in which the length of a one-tailed 95% confidence interval in a sample of normal subjects was 0.7 cm. Statistical methods. Data for the renal followup study were entered into the Paradox microcomputer database system (Paradox Version 3.0; Borland International, Scotts Valley, Calif.) and analyzed with the SPSS (SPSS Version 4.0 for IBM PC; SPSS, Inc., Chicago, Ill.) and BMDP (BMDP Statistical Software, Inc., Los Angeles, Calif.) statistics packages. Estimates for the cumulative incidence of renal artery disease progression were calculated by lifetable analysis. Mean values of risk factors were compared with Student's t test. Comparison of proportions in two-way tables was by chi-squared test or by Fisher's exact test if the cell frequencies were small. Stepwise logistic regression was used to test the

4 JOURNAL OF VASCULAR SURGERY Volume 19, Number 2 Zierler et al. 253 Table II. Cumulative incidence of renal artery disease progression Follow-up interval Baseline renal duplex* o Months 6 Months Normal arteries No. entering interval CI 0 0 < 60% stenosis No. entering interval CI 0 3 ± 3 ;;, 60% stenosis No. entering interval CI 0 0 Total arteries Months 18 Months 24 Months ± 9 29 ± ± ± 3 5 ± 3 11 ± CI, Cumulative incidence given as percent ± standard error. *Five baseline occlusions omitted. predictive value of risk factors for renal artery disease progression. RESULTS Characteristics of the patient population. The 80 patients who provide the basis for this report include 36 men with a mean age of 66 years (range 46 to 81 years) and 44 women with a mean age of 65 years (range 40 to 81 years). Patients were monitored for a mean interval of 12.7 months, with a maximum follow-up period of 26 months. At the baseline visit, 42% of the patients had a history of either myocardial infarction or coronary artery bypass surgery, 41% had lower extremity arterial occlusive disease as indicated by an ankle-arm pressure index ofless than 0.9,38% had extracranial carotid artery disease, and 10% reported having diabetes mellitus. A history of cigarette smoking for more than 10 years was reported by 50 patients. The mean baseline BUN for the entire study group was 23 mg/dl (range 9 to 90 mg/dl), and the mean serum creatinine was 1.4 mg/dl (range of 0.5 to 7.7 mg/dl). Baseline mean systolic and diastolic blood pressures were 168 mm Hg (range no to 226 mm Hg) and 85 mm Hg (range 70 to no mm Hg), respectively. Fifty-eight of the 80 patients were taking at least one antihypertensive medication at the baseline visit, and 43 patients were taking two or more medications. On the basis of a threshold of 140/90 mm Hg, only 6% of the patients would be considered as having adequate blood pressure control at the baseline evaluation. Clinical outcome and disease progression. The disease status of the 139 eligible renal arteries, as determined by baseline duplex scanning, was normal in 36, less than 60% stenosis in 35, 60% or greater stenosis in 63, and occluded in 5. Most of the stenoses were located at the origin or within the proximal segment of the renal artery. Eleven renal artery lesions in nine patients have required therapeutic interventions during the follow-up period: nine arteries have undergone a percutaneous transluminal angioplasty, one patient has undergone a renal artery endarterectomy, and one patient has undergone a nephrectomy. In addition, two patients have begun hemodialysis for progressive kidney failure. Nine patients have died, two patients have dropped out of the study because of severe illness, and two patients have been lost to follow-up. Renal artery disease progression, as defined in this study, was assessed by life-table analysis in 95 arteries monitored for 6 months, 70 arteries monitored for 12 months, 48 arteries monitored for 18 months, and 19 arteries monitored for 24 months (Table II and Fig. 1). None of the 36 renal arteries that were normal on baseline duplex scanning showed disease progression to either 60% or greater stenosis or occlusion. Seven of the 35 renal arteries having less than 60% stenosis at the baseline visit progressed to 60% or greater stenosis during the follow-up period, resulting in a cumulative incidence of progression to 60% or greater stenosis of 23% ± 9% (mean ± standard error) at 12 months and 42% ± 14% at 24 months. Of these seven renal arteries, three underwent serial follow-up evaluations that confirmed disease progression, one had an intervention for progressive renal artery stenosis, and three showed progression of stenosis on their last follow-up visit. All four of the renal arteries that progressed to occlusion had 60% or greater stenoses at the baseline visit, and among those renal arteries with 60% or greater stenoses, the cumulative incidence of progression to occlusion was 5% ± 3% at 12 months and n % ± 6% at 24 months.

5 254 Zierler et al. JOURNAL OF VASCULAR SURGERY February 1994 Cumulative Incidence (%) Normal -. - <60% - - >=60%.- -' _ It"*" o.-. ~.. =_-... =-- -: ) months 12 months 18 months 24 months Follow-up Interval Fig. 1. Cumulative incidence of disease progression for three categories of patent renal arteries (normal, <60% stenosis, and 2:60% stenosis). Data are given in Table II. Eleven kidneys were observed to decrease in length by more than 1.0 cm between the baseline and last follow-up evaluations. The ipsilateral renal artery was classified as less than 60% stenosis in two of these cases and 60% or greater stenosis in the remaining nine cases. The mean length decrease associated with progression of high-grade renal artery stenosis to occlusion was 1.8 cm (range 0.9 to 2.7 cm). No consistent relationship was found in this patient population between progression of renal artery disease and the individual risk factors of age, systolic blood pressure level, or differences in renal function as assessed by serum BUN and creatinine (Table III). However, a significant difference in diastolic blood pressure was observed between those patients who showed renal artery disease progression and those who did not (p = 0.03). When the nine risk factors listed in Table III were included in a stepwise logistic regression to predict progression of renal artery disease, none were found to be statistically significant (p > 0.05). DISCUSSION Although the value of renal revascularization for renovas!=ular hypertension and ischemic kidney failure is generally accepted, the role of intervention for renal artery stenoses that are not associated with uncontrollable hypertension or markedly decreased renal function has not been established. Prophylactic intervention for asymptomatic renal artery stenosis can only be justified on the basis of data that include the rate of disease progression in the renal arteries and the resulting effect on clinical outcome. Duplex scanning is a noninvasive test that has been successfully used to document the natural history of asymptomatic carotid artery disease. In one study, correlation of stenosis severity with neurologic events showed a significantly' higher incidence of ipsilateral transient ischemic attack, stroke, and carotid artery occlusion in patients with internal carotid artery stenosis of 80% or greater diameter reduction. 12 Although the initial applications of duplex scanning were limited to superficial vessels such as the carotid artery, advances in ultrasound technology have made it possible to reliably examine the more deeply located vessels in the abdomen and lower extremities. In spite of a few unfavorable reports,13,14 duplex scanning is rapidly becoming established as an accurate noninvasive method for identifying renal artery disease, both as a screening test and for assessing the results of renal revascularization procedures. 4-1o Studies comparing the results of renal duplex scanning with the results of arteriography have shown that the overall accuracy for identifying renal artery disease is in the range of80% to 96%.5-7 However, renal duplex scanning requires considerable skill on the part of the examiner and a time commitment of at least 1 hour per examination. Even in the most experienced laboratories, technical difficulties may prevent satisfactory completion of renal duplex scanning in from 4% to 12% of examinations. 5,7 One problem with the current technique is failure to reliably detect accessory renal arteries. 6 Another limitation is the inability to further classify

6 JOURNAL OF VASCULAR SURGERY Volume 19, Number 2 Zierler et al. 255 Table III. Correlation of individual risk factors with presence or absence of renal artery disease progression Disease progression Baseline risk factors No Yes p Value Age Systolic blood pressure Diastolic blood pressure BUN Creatinine Years of cigarette smoking Women Diabetes mellitus Other arterial disease 65.7 ± ± ± ± ± ± % 10% 70% 66.4 ± ± ± ± ± ± % 9% 73% Values given as mean ± standard error or percent. renal artery stenoses within the 60% or greater category. Although this would be of theoretical interest, all stenoses that reduce the diameter of the renal artery by 60% or more are considered to be clinically significant because such lesions produce a significant drop in pressure and flow in the ipsilateral kidney. IS Data on the overall prevalence of renal artery stenosis have been based primarily on autopsy studies and reviews of selected patients undergoing arteriography. In an autopsy study of 295 patients, moderate or severe renal artery stenosis was found in 49% of 256 patients who had normal blood pressure and 77% of 39 patients with a history of hypertension. 16 Among 500 patients undergoing arteriography for a variety of vascular problems, renal artery abnormalities were observed in 32% of 304 patients with normal blood pressure and 62 % of 196 patients with hypertension. 17 Most reports on the natural history of renal artery stenosis include only patients who have had multiple arteriograms. In a review of 39 patients with atherosclerotic renal artery lesions who underwent serial arteriography at intervals ranging from 6 months to 10 years, an increase in stenosis severity was noted in 10 and renal artery thrombosis occurred in four, giving an overall progression rate of 36%.18 A similar study that monitored 30 patients for a mean period of 42 months found progression in 50% of the renal artery lesions. 19 Arteriographic follow-up of 85 patients for a mean period of 52 months revealed progressive renal artery stenosis in 37 patients or 44%, including 14 patients in whom renal artery occlusion developed. The risk of progression to occlusion was particularly high in renal arteries with more than 75% stenosis. 20 In the latter two studies, deterioration of renal function and a decrease in kidney size were both more common in patients showing progression of renal artery disease. A more recent study reviewed 48 patients with atherosclerotic renal artery stenoses who did not have repair of the renal artery lesions.21 All patients had a minimum of two arteriograms at least 1 year apart and were monitored for a mean period of 7.3 years. Of the 79 renal artery lesions, progression of stenosis was noted in 42 or 53%. Seven lesions progressed to occlusion, and the average renal artery stenosis on the arteriogram immediately before occlusion was 80% diameter reduction. No renal artery stenoses of less than 60% diameter reduction progressed to occlusion. A variety of factors, including a history of smoking, diabetes, serum cholesterol, changes in blood pressure, and serum creatinine level did not correlate with renal artery disease progression. Although kidney size varied inversely with stenosis severity, it did not serve as a significant predictor for progression to occlusion. Because the preceding studies are all retrospective and based on highly selected patient groups, they may not accurately represent the true progression rate of renal artery stenosis in the general population. These data can only be obtained by prospective follow-up of a less-restricted population sample. In a small prospective study, duplex scanning was used to monitor 15 patients with 19 renal artery stenoses of 60% or greater diameter reduction for a mean of 13 months. 8 Although all 19 stenotic renal arteries remained patent during the follow-up period, kidney length decreased by a mean of 1.0 cm. There was no significant change in kidney length for those contralateral kidneys with less than 60% renal artery stenosis. In this study, patients were selected for the follow-up protocol if they were not considered

7 256 Zierler et at. JOURNAL OF VASCULAR SURGERY February 1994 candidates for renal revascularization at the time of their baseline evaluation. In spite of this, nine of the 11 renal artery lesions that required an intervention during the study had the procedure within the first 12 months of the follow-up period. Although this selection process may have introduced some bias, it would include those patients for which the role of early renal revascularization is particularly uncertain. Thus, this study emphasizes the particular patient group that is of greatest clinical interest. Based on the characteristics of the patients and the duplex scan findings, the renal artery lesions in this study are assumed to be atherosclerotic, and data from this study cannot be applied to patients with fibromuscular dysplasia. Retrospective studies indicate that progression of renal artery stenosis occurs in 30% to 50% of patients monitored for up to 10 years.!8.2! In this prospective study, overall progression rates among the 134 initially patent renal arteries ranged from 8% at 12 months to 15% at 24 months. For progression from a less than 60% to a 60% or greater renal artery stenosis, the rate was approximately 20% per year. Although the incidence of progression to renal artery occlusion was relatively low in the entire study population, the rate of occlusion among renal arteries with a previously documented 60% or greater stenosis was approximately 5% per year. In addition, those kidneys associated with a renal artery occlusion showed a mean decrease in length of almost 2 cm. Because serum BUN and creatinine reflect the combined functional capacity of both kidneys, it was not possible in this study to correlate these measurements of renal function with disease progression in individual renal arteries or changes in kidney length. It is possible that methods designed to detect differential changes in renal function might allow such correlations to be done. Similarly, because most patients in this study were being treated for hypertension, the data do not allow any definitive conclusions about the relationship between blood pressure and renal artery disease progression or changes in kidney length. Finally, it is noteworthy that 11 of the 80 recruited 'patients (14%) had indications for renal intervention after entry into the follow-up protocol, including two patients who started hemodialysis for progressive kidney failure. Because this study involved a relatively small number of patients and a short mean follow-up interval, the results should be considered preliminary. However, these data support both serial follow-up of patients with less than 60% renal artery stenosis to detect disease progression and a trial of earlier renal revascularization for patients with 60% or greater renal artery stenosis to reduce the risk of occlusion and loss of renal mass. Because the chances of benefiting from intervention after renal artery occlusion has occurred are quite small, the rationale for early renal revascularization is especially compelling in patients with severe bilateral renal artery stenoses or a stenosis involving the artery to a solitary kidney. Further natural history data will be necessary to identify more specific clinical markers for predicting which kidneys are most likely to show renal artery disease progression and decreases in length. REFERENCES 1. Simon N, Franklin SS, Bleifer KH, Maxwell MH. Clinical characteristics of renovascular hypertension. JAMA 1972; 220: Jacobson HR. Ischemic renal disease: an overlooked clinical entity. Kidney Int 1988;34: Olin JW, Melia M, Young JR, Graor RA, Risius B. Prevalence of atherosclerotic renal artery stenosis in patients with atherosclerosis elsewhere. Am J Med 1990;88:46N-51N. 4. Kohler TR, Zierler RE, Martin RL, et al. Noninvasive diagnosis of renal artery stenosis by ultrasonic duplex scanning. J VAse SURG 1986;4: Taylor DC, Kettler MD, Moneta GL, et al. Duplex ultrasound in the diagnosis of renal artety stenosis - a prospective evaluation. J VAse SURG 1988;7: Hoffmann U, Edwards JM, Carter S, et al. Role of duplex scanning for the detection of atherosclerotic renal artery disease. Kidney Int 1991;39: Hansen KJ, Tribble RW, Reavis S, et al. Renal duplex sonography: evaluation of clinical utility. J VAse SURG 1990;12: Taylor DC, Moneta GL, Strandness DE J r. Follow-up of renal artery stenosis by duplex ultrasound. J VAse SURG 1989;9: Edwards JM, Zaccardi MJ, Strandness DE Jr. A preliminary study of the role of duplex scanning in defining the adequacy of treatment of patients with renal artety fibromuscular dysplasia. J VAse SURG 1992;15: Eidt JF, Fry RE, Clagett P, Fisher DF Jr, Alway CA, Fry WJ. Postoperative follow-up of renal artery reconstruction with duplex ultrasound. J VAse SURG 1988;8: Johnson CP, Foley WD, Gallagher-Lepak S, Roza AM, Adams MB. Evaluation of renal transplant dysfunction using color Doppler sonography. Surg Gynecol Obstet 1991;173: Roederer GO, Langlois YE, Jager KA, et al. The natural history of carotid arterial disease in asymptomatic patients with cervical bruits. Stroke 1984;15: Berland LL, Koslin DB, Routh WD, Keller FS. Renal artery stenosis: prospective evaluation of diagnosis with color duplex US compared with angiography. Radiology 1990;174: Desberg AL, Paushter DM, Lammert GK, et al. Renal artery stenosis: evaluation with color Doppler flow imaging. Radiology 1990;177: Haimovici H, Zinicola N. Experimental renal artery stenosis:

8 JOURNAL OF VASCULAR SURGERY Volume 19, Number 2 Zierler et at. 257 diagnostic significance of arterial hemodynamics. J Cardiovasc Surg 1962;3: Holley KE, Hunt JC, Brown AL, Kincaid OW, Sheps SG. Renal artery stenosis-a clinical-pathologic study in normotensive and hypertensive patients. Am J Med 1964;37: EylerWR, ClarkMD, GarmanJE, Rian RL, Meininger DEL Angiography of the renal areas including a comparative study of renal arterial stenoses in patients with and without hypertension. Radiology 1962;78: Meaney TF, Dustan HP, McCormack LJ. Natural history of renal arterial disease. Radiology 1968;91: Wollenweber J, Sheps SG, Davis GD. Clinical course of atherosclerotic renovascular disease. Am J Cardiol 1968;21: 60-7l. 20. Schreiber MJ, Pohl MA, Novick AC. The natural history of atherosclerotic and fibrous renal artery disease. Urol Clin North Am 1984;11: Tollefson DF, Ernst CB. Natural history of atherosclerotic renal artery stenosis associated with aortic disease. J VAse SURG 1991;14: Submirted June 10, 1993; accepted Aug. 25, DISCUSSION Dr. Calvin B. Ernst (Detroit, Mich.). We reported on the natural history of atherosclerotic renal arterial lesions in 79 vessels in 48 patients whom we monitored with serial arteriographic studies. 21 We noted that disease progressed in 53% of these vessels, with 9% progressing to occlusion. Of note is that all seven that occluded had stenoses that averaged 80%, ranging from 61% to 94%, on the arteriogram obtained immediately before the one that showed occlusion. At that time we suggested, "It is anticipated that refinements of noninvasive imaging methods will prove useful for precisely defining the natural history of asymptomatic renal arterial stenoses and thereby solve the dilemma of when to repair them." It appears from this report that we have made a major step in determining the natural history of arteriosclerotic renal arterial occlusive disease. Strandness et al. 6 has previously documented a 93% sensitivity for renal arterial duplex imaging. This is an enviable accomplishment. Dr. Zierler and his associates noted that only 3.5% of study results were indeterminate. Could you tell us what percent of studies were indeterminate beyond the initial study; that is, what was the cumulative technical inadequacy rate? This information is important when evaluating the efficacy of serial duplex imaging if it is to become a practical method for monitoring renal arterial stenoses at risk for occluding. It should be noted that of the 98 arteries with stenoses, only 12 were followed up for 24 months, and 10 of these had stenoses greater than 60%, that is, the high-risk lesion. We are interested in this high-risk lesion because five of the seven arteries that occluded were narrowed more than 80%. Unfortunately,,neither we nor you were able to identify clinical markers or risk factors for arteries prone to occlusion. And in your study, you have the problem of not determining precisely the stenosis beyond 60% narrowing. In Dr. Zierler's analysis, for stenosis 60% or greater, the cumulative incidence of progression to occlusion was 5% at 12 months and 11% at 24 months. And although this follow-up period is short, it is alarming that this rate of occlusion occurred; it suggests that, among patients with a reasonable life expectancy, the risk of renal arterial occlusion is significant, and consideration should be given to repair of renal arterial stenoses greater than 60%. On the other hand, there are no data in the literature to suggest that repair of an asymptomatic renal arterial stenosis will prolong life or improve renal function. In our study, we attempted to determine what coincidental renal arterial stenoses should be repaired at the time of aortic reconstruction. The data reported by Dr. Zierler reinforces our view at Henry Ford Hospital that adjunctive renal revascularization should be offered to patients with stenotic lesions identified on arteriography that measure greater than 80% and even a few of those stenoses greater than 60%. Therefore would you venture a recommendation of when renal revascularization should be done for asymptomatic stenoses, because the chances of benefiting from intervention after renal occlusion has occurred are quite small. This repott raises some very interesting questions relative to the repair of asymptomatic renal arterial lesions. Consequently, this report begs a prospective randomized study to identify the benefits or lack thereof of renal revascularization for asymptomatic renal atterial stenoses that now can be easily identified and followed up with duplex imaging. Dr. RichardH. Dean (Winston-Salem, N.C.). Would you repeat the selection criteria of the patients that were entered into the study; were they randomly selected from a subgroup? What was the built-in bias relative to selection of patients? If a kidney has a creatinine of 1, and if this renal artery progresses to total occlusion and collateral vessels are present, what impact does that have? Is there any correlation between progression of the renal artery stenosis in your study and deterioration and some parameter of renal function? Dr. William Bendick (Royal Oak, Mich.). Was your technologist blinded in follow-up studies with regard to the

9 258 Zierler et al. JOURNAL OF VASCULAR SURGERY February 1994 results of a previous study, because this can sometimes bias results 1 Did you notice any regression in any of these lesions that might be caused by technical errors in the examination itselp. Dr. R. Eugene Zierler (Seattle, Wash.). In our clinical vascular laboratory, the occurrence of indeterminate or nondiagnostic examination results is about 10%. In our research study it's lower than that because we have one technologist who is dedicated to the project. Regarding when we should revascularize asymptomatic or minimally symptomatic renal artery stenoses, there's still no definitive answer to that question. This study suggests that we should have a lower threshold for fixing some of these high-grade lesions in patients who require aortic operation for aneurysmal or occlusive disease as well as patients with bilateral high-grade lesions or lesions to solitary kidneys. Unfortunately, we can't subclassify any further within the 60% to 99% category by duplex scanning alone. We are evaluating some other parameters that may be useful in this regard, such as the PSV in the renal parenchyma or the cortical systolic velocity. We have observed that kidneys with high-grade renal artery lesions that also have very low cortical velocities are most likely to decrease in size, so this may be another way of predicting which kidneys are likely to get into trouble. Dr. Dean asked about patient selection bias, and that is a potential criticism of this study. The patients were simply recruited from those who were referred to the vascular laboratory for renal artery stenosis screening. Consequently, this was a selected group of patients in whom somebody was suspicious of a renovascular problem. As you can see by the number of patients who had hypertension that was not adequately controlled, these were patients who had particularly severe hypertension. But nevertheless, they were the kinds of patients in whom we have to make decisions about revascularizing renal arteries. Our technologist was not blinded to the prior duplex scanning result, and that obviously could lead to some bias. However, one advantage of our study is that the same technologist did all these studies so the interobserver variability in performing the examinations was not an issue. Regarding possible regression of lesions, we did observe some borderline lesions that tended to alternate between categories or be right on the borderline of categories; to classify an artery as progressed, we prefer to see two examination results that definitively show disease progression.