Ischemic nephropathy and concomitant aortic disease: A ten-year experience

Ischemic nephropathy and concomitant aortic disease: A ten-year experience

Ischemic nephropathy and concomitant aortic disease: A ten-year experience Elliot L. Chaikof, MD, PhD, Robert B. Smith III, MD, Atef A. Salam, MD, Tho...

1MB Sizes 0 Downloads 22 Views

Ischemic nephropathy and concomitant aortic disease: A ten-year experience Elliot L. Chaikof, MD, PhD, Robert B. Smith III, MD, Atef A. Salam, MD, Thomas F. Dodson, MD, Alan B. Lumsden, MD, Andrzej S. Kosinski, PhD, Kellie A. Coyle, MD, and Robert C. Allen, MD, Atlanta, Ga. Purpose: The durability of renal preservation after surgical intervention has not been well defined, particularly in patients with associated aortic disease. A review of all patients at the Emory University Hospital with renal insufficiency (creatinine level ~ 1.8) and concomitant atherosclerotic aortic and renovascular disease was undertaken. Methods: Fifty patients underwent both renal revascu1arization (71 kidneys) and the repair of aneurysmal or symptomatic aortic occlusive disease between 1982 and 1992. Hypertension was present in 96% of patients and diabetes was present in 10%. The preoperative estimated glomerular filtration rate (EGFR) was 25.18 ± 8.29 ml/min (creatinine level 3.1 ± 1.5 mg/dl). Operative management included bilateral renal artery repair (n = 21), unilateral repair alone (n = 17), and unilateral repair with contralateral nephrectomy (n = 12). The relative percent change in the postoperative EGFR (~7 days after operation) increased by at least 20% in 42% of the patients, had decreased by 20% or more in only 4%, and was otherwise categorized as unchanged in the remaining 54% of the study group. Results: The 30-dayoperative mortality rate was 2.0% (1 of 50). Forty-five of the surviving 49 patients (91.8%) were available for follow-up (median 49 months). During this period nine patients (18.4%) eventually required dialysis, four within 6 months of operation, and 19 patients died. Neither subgroup experienced a retrieval of renal function after operation. Five-year survival rate was 61 %, and a trend was noted between the risk ofdeath and the relative change in EGFR after operation (p = 0.13). The likelihood of eventually requiring long-term dialysis was highest among those patients with low preoperative functional renal reserve as measured by preoperative creatinine level of 3 mg/dl or greater (p < 0.0001), or preoperative EGFR less than 20 ml/min (p = 0.0001). Blood pressure was cured or improved in 50% at late follow-up. Conclusions: Early improvement of renal function may be observed in nearly one half of patients subjected to combined aortic and renal revascu1arization. Nonetheless, renal preservation may not be sustainable in patients with compromised preoperative function. Intervention before marked functional decline remains the best option for minimizing the risk of eventual dialysis. (J VASe SURG 1994;19:135-48.)

Despite our best efforts the incidence of end-stage renal disease (ESRD) in the United States continues to rise at a rate of more than 5% per year. l The potential impact on our health care system, particuFrom the Joseph B. Whitehead Department of Surgery, Emory University School ofMedicine and Division of Biostatistics (Dr. Kosinski), Emory University School of Public Health, Atlanta. Presented at the Forty-first Scientific Meeting of the International Society of Cardiovascular Surgery, North American Chapter, Washington, D.C., June 7-8, 1993. Reprint requests: Elliot L. Chaikof, MD, PhD, Division of Vascular Surgery, Emory University Hospital, Box M-ll, 1364 Clifton Rd., N.E., Atlanta, GA 30322. Copyright © 1994 by The Society for Vascular Surgery and International Society for Cardiovascular Surgery, North American Chapter. 0741-5214/94/$1.00 + .10 24/6/50762

larly in these times of budgetary constraints, is self-evident. In part this figure may well represent the persistent failure to adequately exclude potentially treatable causes of progressive renal insufficiency, most notably, chronic ischemia. 2-s Although ischemic kidney disease may be produced by fibromuscular dysplasia, aortic dissection, or vasculitis, atherosclerotic occlusive disease remains the most common underlying cause. The efficacy of revascularization for the preservation or retrieval of renal function is a currently accepted tenet among vascular surgeons. 6 ,7 Numerous case reports and retrospective analyses from the past 3 decades have documented the reversal of azotemia by both surgical and transcatheter 135

136 Chaikof et at.

techniques. 4 ,8-17 At times, improvement in renal function, with the removal of an occasional patient from hemodialysis, has been nothing short of dramatic. Nonetheless, these reports have been limited by the absence of long-term follow-up and the inclusion of an often heterogeneous patient population. Study groups have been comprised of patients with both fibromuscular dysplasia and atherosclerotic occlusive disease and have included those whose renal insufficiency has ranged from very mild levels of dysfunction to patients undergoing long-term hemodialysis. Furthermore, in many reports there has been a failure to distinguish between patients who demonstrated a relatively acute deterioration of renal function over a period of days or weeks and those who manifest long-standing, yet steadily worsening, chronic renal insufficiency. We have reviewed our experience with surgical intervention for chronic renal insufficiency in patients with concomitant atherosclerotic aortic and renovascular disease over a 10-year period. Characteristically, the potential benefit of renal revascularization is tempered in this population by limited life expectancy and the increased operative risk of an additional surgical procedure. Particular emphasis in this analysis was directed at characterizing the effect of renal revascularization on clinically meaningful endpoints - death and dialysis. MATERIAL AND METHODS Patient population. A retrospective analysis of all patients between January 1982 and March 1992 who underwent combined aortic reconstruction and renal revascularization at the Emory University Hospital was performed. Selection criteria included admission serum creatinine level of l.8 mg/dl or greater and the presence ofeither an abdominal aortic aneurysm or symptomatic aortoiliac occlusive disease. Patients were not included if their primary procedure was the repair of a thoracoabdominal aneurysm. Data was collected with respect to (1) patient demographics, including cardiovascular risk factors, severity of kidney disease, and angiographic findings; (2) operative management, complications, and early functional results; and (3) late functional status, including need for subsequent intervention or dialysis, as well as blood pressure response and survival. Long-term data were obtained as of December 1992 from clinic or hospital records or by telephone interview with the patient, the family, or primary physician. Definitions. The definitions of renal function, blood pressure response, and bilaterality of renovas-

JOURNAL OF VASCULAR SURGERY January 1994

cular disease as described by Dean et al. 4 were used. Serum creatinine levels were converted to estimated glomerular filtration rate (EGFR) with the Rolin18 modification of the Cockroft and Gault19 formula. This accounts for age, sex, weight, and body surface area. Early postoperative renal function (~7 days after operation, but before discharge) was categorized as improved if the EGFR increased by 20% or greater and was categorized as worsened if there was a decrease in EGFR of at least 20% or if dialysis was required. Otherwise, renal function after surgery was classified as unchanged. Thus our study population of sUJgically treated patients could be stratified into responders (improved) or nonresponders (unchanged or worsened) for subgroup analysis directed at identifying potential clinical markers of success or failure. Blood pressure and antihypertensive medication requirements were determined at the time of last follow-up, at least 6 months after operative intelWention. Patients were considered cured on the basis of the absence of antihypertensive medication and a diastolic blood pressure less than or equal to 95 mm Hg. Improvement was defined by either (1) a drop in diastolic pressure greater than or equal to 20 rom Hg and no additional medication; (2) postoperative control of diastolic pressure with less than a 20 mm Hg drop in pressure and reduction of at least one antihypertensive medication; or (3) a reduction of at least two medications in those patients who were normotensive before operation. Patients were otherwise classified as treatment failures. Renovascular disease was categorized as bilateral if there was either angiographically significant ( > 60% ) bilateral occlusive disease or unilateral occlusive disease in a true solitary kidney. Furthermore, patients who underwent unilateral revascularization and contralateral nephrectomy were considered to have bilateral involvement. Otherwise, patients were categorized as having unilateral disease. Statistical analysis. A Wilcoxon signed rank test was used to determine whether the change in postoperative renal function, as measured by the relative change in EGFR, was statistically significant. Wilcoxon rank sum tests were used to screen for the association of this relative change in EGFR after operation with (1) patient demographic features (sex and categorized age); (2) risk factors for atherosclerosis (tobacco use, diabetes mellitus, and hypertension); (3) cardiovascular disease; (4) type of aortic disease (aneurysmal or occlusive); (5) anatomic sites of atherosclerosis (cerebrovascular, coronary artery, and peripheral vascular disease); and (6) the severity of renovascular disease (categorized

JOURNAL OF VASCULAR SURGERY Volume 19, Number 1

Chamof et at.

Table I. Clinical characteristics of patient population Age (years) Range Median Mean Sex Male Female Smoking Diabetes mellitus Hypertension ;2: 3 medications < 3 medications Preop ACE inhibitor ASHD Prior CABGjPTCA Remote myocardial infarction Positive coronary angiogram

PVD Claudication Rest pain Mesenteric insufficiency CVD Prior CEA

137

Table II. Angiographic data

50-79 67 65.9 37 13 40 5 48 25 23 20 38 14 15

(74%) (26%) (90%) (10%) (96%) (52%) (48%) (42%) (76%) (28%) (30%)

13 25 25 2

(26%) (50%) (50%) (4%)

o

14 (28%) 7 (14%)

ACE, Angiotensin-converting enzyme; ASHD, arteriosclerotic heart disease; CABG, coronary artery bypass grafting; PTCA, percutaneous transluminal coronary angioplasry; PVD, peripheral vascular disease; CVD, cerebrovascular disease; CEA, carotid

endarterectomy.

preoperative serum creatinine level or EGFR, and bilateral or unilateral involvement). The most important factors influencing late survival and need for eventual dialysis was determined by multivariate Cox regression modeling. This approach allows the consideration of both continuous (e.g., age, preoperative serum creatinine level) and categorical variables (e.g., sex, hypertension). The probabilities of death or dialysis were estimated in relevant subgroups with Kaplan-Meier analyses. Differences between Kaplan-Meier functions were assessed with log rank tests. Two-sample t tests were used to detect associations between blood pressure control (improved or cured vs failed) and continuous variables. Significance was reevaluated with a Wilcoxon rank sum test when necessary. The significance for categoric variables was analyzed with a Fisher exact test. RESULTS Preoperative patient profile. The clinical characteristics and angiographic data of our patient population are summarized in Tables I and II, respectively. Of note, hypertension was present for less than 1 year in six patients (13%), 1 to 5 years in 10 (21%), and greater than 5 years in the remaining 34 patients (66%). The severity of kidney disease

Abdominal aortic aneurysm Aortoiliac occlusive disease Combined aortic aneurysmal and occlusive disease

25 19 6

Isolated SMA stenosis or occlusion Isolated celiac artery stenosis or occlusion Combined SMA and celiac occlusive disease

2 1 1

Unilateral renal artery disease Bilateral renal artery disease Bilateral renal artery occlusive disease Unilateral renal artery occlusive disease in a solitary renal artery

10 40 21 19*

SMA, Superior mesenteric artery. *Includes two patients who had a previous nephrectomy.

among the group was characterized by a mean preoperative serum creatinine level of 3.1 ± 1.5 mg/dl and EGFR of25.18 ± 8.29 ml/min. Normal values at our institution for isothalamate clearance glomerular filtration rate (GFR) are approximately 100 ml/min. Ten percent of our patients had undergone previous percutaneous renal artery angioplasty. Operative management and early results. Methods of aortic and renal artery reconstruction and postoperative complications are presented in Tables III and IV. The preferred method of renal artery reconstruction has been bypass grafting with a Dacron prosthesis. Intraoperative confirmation of the adequacy of surgical reconstruction was achieved with a continuous-wave Doppler flow probe. Contralateral nephrectomy was performed in those patients with severe hypertension and a shrunken kidney not visualized on angiographic examination. The most frequent postoperative complication and the sole postoperative death was related to underlying heart disease. Mild temporary worsening of renal function was noted in two patients. No patients required dialysis in the immediate postoperative period. The incidence of postoperative complications was not significantly higher among those patients undergoing bilateral procedures. The 30-day operative mortality rate was 2% (1 of 50). The relationship between preoperative EGFR and the relative percent change in postoperative EGFR is presented in Fig. 1. After operation EGFR increased by at least 20% in 21 patients (42%), worsened by 20% or more in 4% of patients, and was otherwise unchanged in the remaining 54% of the study group. The median change was + 18%, which differs significantly from the 0% change (p < 0.0001, Wilcoxon signed rank test). Stratification with respect to site of disease (bilateral vs

JOURNAL OF VASCULAR SURGERY January 1994

138 Chaikof et al.

Table III. Operative treatment Methods of aortic reconstruction Aortobifemoral graft Aortobiiliac graft Aortic tube graft

Table IV. Postoperative complications 10 33 7

Methods of renal artery reconstruction Graft replacement of aortic aneurysm Bilateral renal artery bypass with reimplantation of the lMA Unilateral renal artery bypass with contralateral nephrectomy with contralateral endarterectomy with contralateral reimplantation with reimplantation of the lMA to a solitary kidney

25 7 1 5 7 1 1 2 1

Graft replacement of aortic occlusive disease Bilateral renal artery bypass with reimplantation of the lMA Unilateral renal artery bypass to a solitary kidney with contralateral nephrectomy

19

Graft replacement of aortic aneurysm with occlusive disease Bilateral renal artery bypass Bilateral reimplant Unilateral revascularization with contralateral nephrectomy

8 3 3 1 4 6 1 1 3 1

IM, Inferior mesenteric artery.

unilateral) did not reveal a propensity for greater success in one group rather than another (p = 0.67) (Table V). Early functional success was otherwise not associated with any of the screened clinical variables. Late functional outcome. Forty-five of the surviving 49 patients (9l.8%) were available for follow-up (median 49 months). Nine patients eventually required dialysis (18.3%), four within 6 months of operation. Of note, only one of 12 patients who at the time of surgery underwent contralateral nephrectomy eventually required dialysis. In addition, among five patients who had undergone previous renal artery angioplasty, ESRD developed in only one patient. Overall, the likelihood of eventual long-term dialysis, despite operation, was higher in those patients with lower preoperative functional renal reserve. The relative risk (RR) of subsequent kidney failure, obtained from the Cox analysis, was 3.0 (p < 0.0001, 95% confidence interval (CI): l.76, 5.11) for each l.0 mg/dl increment in preoperative serum creatinine level. Similarly, the relative risk per unit increase of EGFR was l.24 (p = 0.0001,95% CI: l.10, 1.40). In other words, a decrease in the preoperative EGFR by 5 mljmin was associated with nearly tripling the risk of eventual renal shutdown (RR 2.97; 95% CI: l.64, 5.43).

Mild transient worsening of renal function Myocardial infarction Congestive heart failure without infarction Cardiac arrythmia Reexploration for bleeding Incidental splenectomy Colonic ischemia (nontransmural) Hepatitis Ileus Major depression VVoundinfection Clostridium difficile colitis

2 4 1 2 1 1 1 1 2 1 2 1

Additional Cox modeling also revealed a nonstatistically significant trend (p = 0.15) between operative outcome and subsequent dialysis. For each 1% improvement in postoperative EGFR, the risk of hemodialysis declined by 0.98 (95% CI: 0.95, l.01), or equivalently a 20% improvement in the EGFR decreased this risk by nearly one third (RR 0.67; 95% CI 0.37, l.22). A statistically significant relationship between anatomic extent of renovascular disease and late ESRD was not detected. These relationships were further evaluated by estimation of the long-term probabilities of avoiding dialysis in various subgroups with a Kaplan-Meier analysis. (Fig. 2). Similar to those results generated by Cox modeling, the likelihood of eventually requiring dialysis was higher among those patients with low preoperative functional renal reserve as measured by preoperative creatinine levels 3 mg/dl or greater or EGFR less than 20 mljmin (data not shown). Furthermore, when patients were stratified with respect to early renal functional response, the probability of avoiding dialysis was greater among those who had achieved at least a 20% increase in EGFR ("improved" subgroup) compared with those whose renal function after operation was classified as unchanged or worsened (p = 0.0001). These results suggest that functional status after surgical intervention may be a potential indicator of later clinical outcome. Functional deterioration was examined by plotting EGFR as a function of time and expressing the rate of change, as (mljmin)mos- 1• Data were available for 17 patients in the preoperative period and 32 patients after operation (Table VI). Among those patients who did not undergo dialysis a statistically significant reduction in the rate of functional decline was noted after intervention (-2.48 ± 3.51 vs -0.12 ± 0.28 (mljmin)mos-l,p = 0.002). Blood pressure was cured or improved in 20 (50%) of the 40 patients with previous hypertension who were available for follow-up. Failure to achieve

JOURNAL OF VASCULAR SURGERY

Chaikof et at.

Volume 19, Number 1

l





100-

139

a:

La.



fa



t

• ••

!

50-

.6

•• • • • 20 - - - - - - - - - - - - - -. - - -. _.,. - - - . . . . -.. - - -~ - - ,:.

I J! J

••

• 'It

•• • -- -------- -----------

0- • -20

..

. .. • •

••



JI_ - - - - - - - - - - - - - - -



• I

o

10

15

20

25

30

35

Preoperative EGFR (mVmin) Fig. 1. Relationship between preoperative and postoperative EGFR.

blood pressure control was directly related to compromised preoperative renal reserve, as measured by the serum creatinine level (p = 0.005, Wilcoxon rank sum test). Univariate associations were not found between blood pressure control and categoric variables with a Fisher exact test. Renal angiograms and scans were not obtained on a routine basis during the study period. As a consequence, such data were available in only one third of these patients. There were three known late graft occlusions and two anastomotic stenoses, one of which was successfully managed with percutaneous angioplasty. None of these patients had worsening of EGFR after operation nor did they require eventual dialysis. Late survival. There were 19 late deaths (median follow-up time 49 months), the causes of which were heart disease (four), ESRD (five), malignancy (two), or otherwise unknown (eight). Of the patients whose cause of death was unknown only one was ascertained to have a serum creatinine level greater than 3.0 mg/dl at the time of last follow-up. Kaplan-Meier survival functions are presented in Fig. 3. When patients were stratified with respect to postoperative renal function, there is a trend toward better survival rates for patients with improved relative measures of operative success. This was noted in both subgroups defined by either a relative postoperative increase ( ~ 20%) in EGFR or decrease (~15%) in serum

Table V. Site of disease and early functional response Change in EGFR Improved No change Worsened Total

Unilateral disease

Bilateral disease

No.

No.

%

18

45% 50% 5%

3 7 0

10

% 30% 70%

20 2 40

creatinine level (data not shown). When analyzed with Cox regression with continuous covariates, the risk of death is estimated to increase 1.02 times (p = 0.017; 95% CI: 1.002, 1.046) for a 1% decrease in EGFR and 1.02 times (p = 0.020; 95% CI: 1.01,1.04) for 1% increase in the relative change in postoperative creatinine levels. Nonetheless, the results of the above Cox analysis depend heavily on a few patients and should be treated with caution. More patients are needed to definitely establish the survival impact of renal functional response to surgery. Notably, survival did not differ (p = 0.90) when compared across preoperative levels of serum creatinine or EGFR. Dialysis-free survival. The dialysis-free survival is a measure of our ability to sustain the patient's quality of life (Fig. 4). The data suggest that despite operation the quality of life in patients with poor

JOURNAL OF VASCULAR SURGERY January 1994

140 Chaikof et al.

1.0

I

\ -E

0 '0 CD

Preoperative creatinine < 3

0.8

!

I

o.

>-. :!:>=Ci Sl-

0.6

Preoperative creatinine ~ 3

ca'O

-

SlE

eo a,.r..

'0 CD

'ti

0.4 0.2

Log-rank test

E

iw

P-value

0.0 0

20

A

40

60

=0.0001 100

80

120

Time from operation (months)

1.0

E

Improved postoperative EGFR

0

'D

-I»

!

0.8

o . >-;;

:>- 0.6

Unchanged or worse postoperative EGFR

-as as'D

.0-

.oE 0.4 eo A..t:

'D

•is I»

0.2

Log-rank test

E

P-value

W

0.0 0

B

20

40

60

80

=0.0001

100

120

Time from operation (months)

Fig. 2. Kaplan-Meier estimated probability of avoiding dialysis in subgroups categorized with respect to (A) preoperative serum creatinine level less than 3.0 mg/dl versus serum creatinine level of 3.0 mg/dl or greater and (B) postoperative EGFR improved (2: 20% + ) versus worse or unchanged.

preoperative reserve (serum creatinine level 2: 3 mg/dl or EGFR ::;; 20 ml/min; EGFR stratification not shown) remains compromised. Furthermore, dialysis-free survival rate is also reduced among

patients who did not have early functional improvement. However, the absence of a control population precludes any meaningful comment with regards to the overall impact of renal revascularization.

JOURNAL OF VASCULAR SURGERY Volume 19, Number 1

Chamo! et aI.

141

Table VI. Rate of deterioration of renal function EGFR (mljmin)mos- l

Preoperative Late dialysis Yes No Total

Postoperative

Mean ± SE

Median

N

Mean ± SE

Median

N

p Value

-0.61 ± 1.73 -2.48 ± 3.51 -1.93 ± 3.16

-0.95 -0.78 -0.79

5 12 17

-4.45 ± 5.63 -0.12 ± 0.28 -1.06 ± 3.08

-2.58 -0.16 -0.21

7 25 32

NS 0.002 NS

DISCUSSION The premise that occlusive disease of the renal artery is an important and underestimated cause of renal insufficiency is not disputed within the vascular surgical community. Nonetheless, therapeutic strategies remain somewhat ill defined because of the lack of information regarding the natural history of ischemic nephropathy and specifically regarding our ability to alter the long-term course of progressive renal insufficiency by surgical intervention. The risk/benefit ratio is particularly blurred in the patient with concomitant aneurysmal or occlusive disease of the aorta. Should patients be subjected to combined aortic and renal artery reconstruction without a clear-cut test of who will benefit from renal artery reconstruction? In the absence of an accurate predictor of outcome, is there a level of renal insufficiency or anatomic configuration that warrants surgical intervention, particularly if hypertension is well controlled? In the final analysis, few studies have adequately addressed, even in a retrospective fashion, how many patients will be spared dialysis and for how long. Nor do we know whether intervention will improve overall survival rates. The objective of this analysis is to better define dialysis-free survival after combined aortic and renal artery reconstruction. Although the presence of aortic disease is an indicator of advanced atherosclerosis, we believe that the data from this review are applicable to many patients with ischemic nephropathy. Assessment of renal revascularization in human subjects has been limited in most studies to the measurement ofserum creatinine levels as a substitute for GFR determinations. Changes in muscle mass, meat intake, tubular secretion, and extrarenal metabolism of creatinine are all confounding factors in its use. 20 Stenosis of the main renal artery of 70% or greater or a gradient of at least 40 mm Hg is typically associated with a reduction of GFR. 2 The failure in a number of studies to demonstrate a direct correlation between serum creatinine levels and progression of renal artery stenosis, in part, reflects these limitations. 21 ,22 Such criticisms apply equally to the use of creatinine clearance, inverse creatinine slopes, and the

Cockcroft and Gault19 EGFR. We agree, however, with Dean et al. 4 that the EGFR may be a useful approximation when applied to an entire group. Revascularization had a short-term beneficial effect in nearly one half of our patients when considered in terms of improvement in the EGFR. In contrast to an earlier report from our institution, benefit was observed in patients with both unilateral and bilateral renal artery disease. 23 Nor could we detect a statistically significant relationship between early or late outcome and whether a bilateral or unilateral procedure was performed. Dean et al. 4 has failed to observe a statistically significant level of improvement in patients with unilateral disease when analyzed as a group. Nonetheless, they noted an increase in EGFR of20% or greater in 33% of these patients (4 of 12). The small number of patients that fall into the category of unilateral renal artery involvement prevent hard and fast rules at this time. Otherwise there were no clinical or anatomic variables that clearly predicted short-term outcome including the level of preoperative renal function. This finding is in contrast to other reports 14,23,24 in which early functional responses were most notable in patients with higher preoperative serum creatinine levels. As reported by Dean et al.,4 patients with a slow decline in GFR often demonstrate little, if any, improvement after revascularization. Perhaps our patients were biased towards this category as a consequence of intervention that was more often triggered by the presence of an aneurysm or aortoiliac occlusive symptoms and not by accelerated hypertension or renal insufficiency. A median decline in preoperative EGFRofO.79 (ml/min)mos- 1 suggests that this may well be the case. Although the serum creatinine level can provide some insight into renal function, the ability to draw reliable conclusions in the absence of serial, isotopically determined GFRs ultimately depends on an analysis of clinically meaningful endpoints - death or dialysis.20 The reported incidence oflate dialysis after revascularization has received comment in only a few reports, reflecting somewhat limited study periods. In general, ESRD despite surgery has ranged from

JOURNAL OF VASCULAR SURGERY January 1994

142 Chaikof et al.

-

C!

Total pellent population II! 0

1iI

'15 ~ ~

I

~

Ijj

N

0

o

A

20

40

60

80

100

120

Time from operation (months)

q

....

J -

EGFR improved by at least 20%

c:q 0

~ II)

0

>.

:=

:sCIS eCo


0

EGFR unchanged or worse

,Q

"n;

"'It

ci

II)

E

TA

w

C\I

ci

Log-rank test P-value = 0.15

0

ci

0

B

20

40

60

80

100

120

Time from operation (months)

Fig. 3. Kaplan-Meier life-table analysis. A, Total patient population. B, Postoperative EGFR improved (~20% + ) versus worse or unchanged.

4.4% to 14%.4,15,25 In this review, the need for eventual dialysis was greatest among patients with an elevated preoperative creatinine level or depressed EGFR and was not predicted by other clinical

variables, including the pattern of renal artery occlusive disease. In fact, the relative risk for eventual dialysis was 3.0 for each 1.0 mg/dl increment in preoperative serum creatinine level. Among patients

JOURNAL OF VASCULAR SURGERY Volume 19, Number 1

Chailwf et at.

143

q

.,...

Preoperative creatinine <3

co

o

Preoperative creatinine ~ 3

N

o log-rank test P-value = 0.0001

q

o

o

20

40

A

60

80

120

100

Time (months)

....q EGFR improved by at least 20% cq

o

EGFR unchanged or worse

log-rank test P-value = 0.054

q

o

o

B

20

40

60

80

100

120

Time (months)

Fig. 4. Kaplan-Meier estimated probability of dialysis-free survival. A, Preoperative serum creatinine level less than 3.0 mg/dl versus serum creatinine level of 3.0 mg/dl or greater. B, Postoperative EGFR improved (~20% + ) versus worse or unchanged.

whose preoperative creatinine level was 3.0 mgldl or greater, the estimated 3-year probability of ESRD was 32% (95% CI: 4%, 51%). In contrast, the incidence of late kidney failure was low among our

patients with only a moderate degree of preoperative azotemia (serum creatinine level: 1.8 to 3.0 mgldl). Bredenberg et ai. 25 have recently noted that five of six patients who had serum creatinine levels above 5.5

144

Chaikof et at.

mg/dl underwent dialysis, despite proven graft patency. The ability to maintain normal blood pressure with minimal or no medication may well affect the rate of progressive renal insufficiency. Untreated malignant or accelerated hypertension can certainly cause ESRD; however, the frequency with which mild to moderate essential hypertension leads to kidney failure is unclear.5 A recent European randomized, prospective study found no change in creatinine clearance in elderly patients with untreated hypertension over a 5-year period offollow-up.26 In contrast, a decline in creatinine clearance was observed in those receiving antihypertensive medication, suggesting that treatment and not blood pressure per se is the cause of worsening renal function. Hypertension was cured or improved in approximately 50% of our patients evaluated at late follow-up. This result is in the range of 53% to 78% noted in other reviews of combined aortic and renal artery reconstruction. 23 ,24,27,28 Poor late blood pressure control was statistically related to high preoperative serum creatinine levels. Indeed, we cannot with certainty account for the contribution of antihypertensive medication either to the level of renal function noted before surgery or to its decline at late follow-up. The potential benefits of combined renal artery and aortic surgery are often weighed against the advanced age and associated medical problems that are characteristic of this population. The overall 5-year survival rate was 62% (95% CI: 0.48,0.81), similar to 5-year survival rates of approximately 65% reported from both the Cleveland and Mayo Clinics in 1987. 15 ,29 However, we were somewhat surprised that long-term survival remains compromised among those patients with preoperative serum creatinine levels less than 3.0 mg/dl, despite their lower probability of eventual dialysis. Accelerated mortality rates among patients with renovascular disease undergoing dialysis have been reported. 3,30 In particular, Mailloux3 has observed a 5-year actuarial survival rate of 12%. Although dialysis may significantly compromise the life expectancy of a young or middle-aged adult, in the elderly patient with associated comorbid conditions the effect of dialysis on late survival may be marginal. It is noteworthy that the ability to impact on the quality of life, as measured by the variable dialysis-free survival rate, is significantly worse among patients with lower preoperative reserve. In many, but not all patients, markedly compromised preoperative renal function

JOURNAL OF VASCULAR SURGERY January 1994

is a probable marker for likely poor outcome. Thus, in the current absence of prospective data intervention prior to marked deterioration of renal function remains the best option for preserving the patient's quality of life by reducing the decline in renal function and decreasing the need for subsequent dialysis. Thus, we recommend repairing all angiographically significant (> 60%) stenoses in patients with associated aortic disease who have a preoperative serum creatinine level in the range of 1.8 to 3.0 mg/dl. In those who have higher levels of serum creatinine, advocacy for conservative, nonoperative management in the patient with aortic disease would be premature at this time. In our experience, an occasional patient in this category benefits from revascularization, particularly in the face of recent accelerated functional deterioration. The low operative mortality rate at our center is comparable to an earlier review of our experience. 23 Likewise, a number of recent series of combined aortic and renal artery reconstruction have documented operative mortality rates of approximately 3% or lesS. 24,31,32 Higher incidences of 10% or greater may reflect sample sizes and heterogeneity with regard to the complexity of operation and underlying medical problems. Although our preference has been to perform bilateral revascularization when needed and not "stage" the procedures, treatment recommendations need to be tempered by the condition of the individual patient. A 14% incidence in this report of nonfatal cardiac morbidity reinforces the inherent perioperative hazards. It is remarkable that, although the pathophysiologic condition of ischemia-induced acute kidney failure has been extensively investigated, the mechanisms that contribute to progressive renal dysfunction in the presence of chronic or intermittent ischemia have been little studied or defined. Dean et al. 4 has speculated that there are two mechanisms that may lead to renal dysfunction: a reversible form of chronic ischemia that is related to the presence of a flow-limiting lesion and an irreversible type related to renal microemboli produced by disease in the main renal artery. He has postulated that, left untreated, the former is associated with rapid deterioration but at the same time offers the best opportunity for retrieval of function. The latter is associated with a slower decline in function and may be less affected by surgical intervention. These postulates remain untested, but both are based on the presumption that the natural history of ischemic nephropathy will be altered by improved renal perfusion.

JOURNAL OF VASCULAR SURGERY Volume 19, Number 1

In many forms of kidney disease, progressive deterioration may occur even in the absence of the initial cause of injury. Apart from continued smallvessel disease, nonhemodynamic mechanisms, including genetic, metabolic, lipid, and coagulation factors, may act synergistically in potentiating and promoting continued glomerular damage. 33 Perhaps ischemia of an undefined duration is an initiating event. Three hypotheses seem to be plausible. Molitoris 34 has speculated that chronic ischemia may lead to proximal tubular damage, which could reduce GFR by nephron obstruction. Alternatively, changes in oxygen tension or wall shear rates produced by a flow-limiting lesion may activate or directly injure glomerular endothelium, leading to the release of procoagulant, mitogenic, or chemotactic peptides. Potentially, local activation of monocytes or platelets would further accelerate this cascade of events terminating in glomerular scarring. Finally, ischemia induced injury of glomerular mesangial or epithelial cells may lead to enhanced extracellular matrix synthesis, producing renal fibrosis, as in glomerulosclerosis. 35 Once these events are in play, GFR may continue to decline despite the correction of renal ischemia. As we begin to understand the basic pathophysiologic condition of chronic ischemia, additional therapeutic options may provide a means of altering the path to kidney failure that many follow despite surgery. Until then, our sole option remains early intervention to minimize the duration of the ischemic insult. REFERENCES 1. U.S. Renal datal system, in USRDS 1989AnnualDataReport. Bethesda, National Institute of Health, National Institute of Diabetes and Digestive Kidney Diseases, 1989. 2. Jacobson HR. Ischemic renal disease: an overlooked clinical entity? Kid Int 1988;34:729-743. 3. Mailloux LU, Bellucci AG, Mossey RT, et al. Predictors of survival in patients undergoing dialysis. Am J Med 1988;84: 855-62. 4. Dean RH, Tribble RW, Hansen KJ, O'Neil E, Craven TE, Redding JF II. Evolution of renal insufficiency in ischemic nephropathy. Ann Surg 1991;213:446-55. 5. Weisstuch JM, Dworkin LD. Does essential hypertension cause end-stage renal disease? Kid Int 1992;41:S33-S37. 6. Bergentz SE, Bergqvist D, Weibull H. Changing concepts in renovascular surgery. Br J Surg 1989;76:429-30. 7. Scobie JE, Hamilton G. Atherosclerotic renovascular disease: remediable causes of renal failure in the elderly. Br Med J 1990;300: 1670-1. 8. Morriss GC, DeBakey ME, Cooley DA. Surgical treatment of renal failure of renovascular origin. JAMA 1962; 182 :609-12. 9. Stoney RJ, Skioldebrand CG, Qvarfordt PG, Reilly LM, Ehrenfeld WK. Juxtarenal aortic atherosclerosis. Surgical experience and functional result. Ann Sur.g 1984;200:345-54.

Chaikof et al.

145

10. Whitehouse WM Jr, Kazmers A, Zelenock GB, et al. Chronic total renal artery occlusion: effects of treatment on secondary hypertension and renal function. Surgery 1981;89:753-63. 11. Madias NE, Kwon OJ, Millan VG. Percutaneous transluminal angioplasty: a potentially effective treatment for preservation of renal function. Arch Intern Med 1982;142:693-7. 12. Ying CY, Tifft CP, Gavras H, Chobanian AV. Renal vascularization in the azotemic hypertensive patient resistant to therapy. N Engl J Med 1984;311:1070-5. 13. Weibull H, Tornqvist C, Bergqvist D, et al. Reversible renal insufficiency after percutaneous transluminal angioplasty (PTA) of renal artery stenosis. Acta Chir Scand 1984;150: 295-300. 14. Dean RH, Englund R, Dupont WD, et al. Retrieval of renal function by revascularization: study of preoperative outcome predictors. Ann Surg 1985;202:367-75. 15. Hallett JW, Fowl R, O'Brien PC, et al. Renovascular operations in patients with chronic renal insufficiency: do the benefits justify the risks? J VASe SURG 1987;5:622-7. 16. NovickAC, Ziegelbaum M, Vidt DG, Gifford RW, Pohl MA, Goormastic M. Trends in surgical revascularization for renal artery disease. JAMA 1987;257:498-501. 17. Elmore JR, Ray FS, Dillihunt RC, Herbert WE. Renal failure and advanced atherosclerotic lesions. Salvage by vascular reconstruction. Arch Surg 1988;123:610-3. 18. Rolin HA, Hall PM, Wei R. Inaccuracy of estimated creatinine clearance for prediction of iothalamate glomerular filtration rate. Am J Kidney Dis 1984;4:48-54. 19. Cockroft DW, Gault MH. Prediction of creatinine clearance from serum creatinine. Nephron 1976;16:31-41. 20. Perrone RD. Means of clinical evaluation of renal disease progression. Kidney Int 1992;41:S26-S32. 21. Schreiber MJ, Pohl MA, Novick AC. The natural history of atherosclerotic and fibrous renal artery disease. U rol Clin North Am 1984;11:383-92. 22. Tollefson DF, Ernst CB. Natural history of atherosclerotic renal artery stenosis associated with aortic disease. J VAse SURG 1991;14:327-31. 23. Stewart MT, Smith RB III, Fulenwider JT, Perdue GD, Wells JO. Concomitant renal revascularization in patients undergoing aortic surgery. J VAse SURG 1985;2:400-5. 24. O'Mara CS, Maples MD, Kilgore TL Jr, et al. Simultaneous aortic reconstruction and bilateral renal revascularization: is this a safe and effective procedure? J VAse SURG 1988;8;35766. 25. Bredenberg CE, Sampson LN, Ray FS, Cormier RA, Heintz S, Eldrup JJ. Changing patterns in surgery for chronic renal artery occlusive diseases. J VAse SURG 1992;15:1018-23. 26. De Leeuw PW. Renal function in the elderly: results from the European working party on high blood pressure in the elderly trial. Am J Med 1991;90(Suppl 3A):45S-49S. 27. Brewster DC, Buth J, Darling RC, Austen WG. Combined aortic and renal artery reconstruction. Am J Surg 1976;131: 457-63. 28. Shahian DM, Najafi H, Javid H, Hunter JA, Goldin MD, Monson DO. Simultaneous aortic and renal artery reconstruction. Arch Surg 1980;115:1491-7. 29. Tarazi RY, Hertzer NR, Beven EG, O'Hara PJ, Anton GE, Krajewski LP. Simultaneous aortic reconstruction and renal revascularization: risk factors and results in eighty-nine patients. J VAse SURG 1987;5:707-714. 30. Novick AC, Textor SC, Bodie B, Khauli RE. Revasculariza-

JOURNAL OF VASCULAR SURGERY

146 Chaikof et al.

January 1994

33. Bourgoignie JJ. Progression of renal disease: current concepts and therapeutic approaches. Kidney International 1992; 41(5uppl. 36):561-565. 34. Molitoris BA. The potential role of ischemia in renal disease progression. Kidney Int 1992;41(5uppl 36):521-525. 35. Meguid El Nahas A. Growth factors and glomerular sclerosis. Kidney Int 1992;41(5uppl 36):515-520.

tion to preserve renal function in patients with atherosclerotic renovascular disease. Urol Clin North Am 1984;11;47790. 31. Barral :x, Delonne JM, Favre JP, et al. Usions arterielles renales ei aneurysmes de l'aorte abdominale sous-renale. In: Keiffer E, ed. Les Aneurysmes de I'Aorte Abdominale 50us-Renale. Paris: AERCV, 1990:337-48. 32. Branchereau A, Espinoza H, Magnan PE, Rosset E, Castro M. 5imultaneous reconstruction of infrarenal abdominal aorta and renal arteries. Ann Vasc 5urg 1992;6:232-8.

5ubmitted June 16, 1993; accepted Aug. 10, 1993.

DISCUSSION Dr. Ronald J. Stoney (San Francisco, Calif.). Dr. Chaikof et al. retrospectively analyzed the outcome of 50 patients treated in the past decade for renal functional impairment associated with perirenal aortic atherosclerosis and occlusive or aneurysmal disease of the infrarenal aorta. Their fundamental question in this study was as follows: To what degree can surgical intervention prevent the need for subsequent dialysis and for how long? Overall, 49 patients survived operation and 45 were available for follow-up to assess the effect of renal revascularization on renal function and hypertension. This study assumes that obstructive renal artery lesions greater than 60% impaired parenchymal perfusion and decreased GFR, and after correction at the time of the aortic repair, the durability of the renal functional response could be then characterized. A number of findings of this study, however, question the validity of this assumption: • 70% of repairs of unilateral renal artery stenosis failed to improve the GFR. • 50% of repairs of bilateral lesions failed to improve the GFR. • Duration of operative renal ischemia was unknown for these revascularizations. • Proof of a patent renal artery revascularization was not required in this study. • Five of 45 patients, 11%, had failed repairs, yet these did not cause a deterioration of the GFR. • Nine survivors required dialysis, yet the status of their renal revascularizations were not described. The outcome of patients with azotemia varied in this follow-up. As expected, those with the lowest preoperative functional reserve were more likely to die or require dialysis. If the GFR decreased by 5 ml/min after revascularization, the risks of eventual kidney failure tripled. Other factors must be analyzed in considering these results. • Arteriographic findings. The presence of infrarenal small vessel lesions could represent nephrosclerotic changes or atheroembolic occlusions secondary to a proximal aortic source. Neither of these would necessarily be influenced by the renal bypass. • Blood pressure control. We know that preoperative renal impairment may worsen as a function of aggressive

antihypertensive therapy. Postoperative care or improvement of blood pressure as observed in half of their survivors could also have had a similar effect on renal function. These two factors, operating either independendy or together, may provide an explanation for some of the parenchymal deterioration that occurred and is at odds with others' experiences. In a report from our institution 5 years ago, which the authors cited, 32 of 40 patients undergoing combined aortic and renal revascularization for renal impairment had stable renal function or improved renal function over a 4-year follow-up period, whereas only five patients had mild deterioration to their preoperative level. And none of those patients underwent dialysis. The value of intraoperative assessment of complex aortic branch repair has been well established in several centers. It is simple and safe to use. Duplex scanning can identify defects often seen in up to one third of patients, although most of these are minor, but it can identify defects that are likely to go on to occlude the repair and can be identified and corrected with operation. This will allow the ability to confirm patency of the repair by predischarge aortography in a higher percentage of patients. What, besides the renal occlusive disease that you repaired, do you believe contributed to the renal functional impairment in your patients? What methods of intraoperative and postoperative assessment do you now recommend for aortic branch reconstruction? What role does the relief or improvement in hypertension after renal revascularization play in the renal deterioration seen in your patients? Dr. Louis M. Messina (Ann Arbor, Mich.). Improvement in renal function was achieved in 40% of the patients in the early postoperative period. Unfortunately this early improvement in renal function was not sustained. Five patients 'died ofESRD, 18% required dialysis. One half of these patients required dialysis within 6 months of operation. What was the mechanism of the loss of renal function during long-term follow-up? The tide of your presentation implies that all of these patients had ischemic nephropathy. I share Dr. Stoney's concern about what the mechanism for the elevated

JOURNAL OF VASCULAR SURGERY Volume 19, Number 1

creatinine level was in these patients before operation. One third of the patients had unilateral renal artery disease and only underwent a unilateral reconstruction. Thus it is difficult to ascribe in these patients their poor renal function for it to be caused by ischemic nephropathy. In addition, it seems that most patients who underwent operation were not referred primarily for symptoms related to renal artery occlusive disease but for symptoms related to aortic occlusive disease or the presence of an aortic aneurysm, with the renal artery occlusive disease found incidentally. Thus the primary indication for renal revascularization seemed to be preservation of renal function rather than complications caused by ischemic nephropathy. Would you clarify how many patients were evaluated primarily for symptomatic renal artery occlusive disease? I raise this issue because at the University of Michigan, we have noted a distinct difference in the outcome for these two subgroups of patients. Two years ago we reported our results on surgical therapy for patients with symptomatic bilateral renal artery occlusive disease, characterized by recurrent pulmonary edema and poorly controlled hypertension. In these patients, the mean preoperative creatinine level was 3.8 mg/d!. At long-term follow-up the mean creatinine level was 1.7 mg/d! and nearly identical to what it was during the early postoperative period. Two of the patients who underwent preoperative dialysis had discontinuation of dialysis, and only one of the 25 patients required dialysis during long-term follow-up. Thus we have also noted a less predictable outcome in patients whose primary indication for renal revascularization is preservation of renal function, and we believe this distinction is important in terms of anticipating outcome after surgery. Have you had a similar experience? When patients had deterioration of renal function during the late follow-up, in how many was the patency of the renal revascularization determined? Were any studies performed to document the presence of other primary kidney diseases that may have been present? Dr. Brent T. Allen (St. Louis, Mo.). We recently reviewed our experience at Barnes Hospital with 102 patients undergoing combined aortic and renovascular surgery. Twenty-nine patients, 28% of the entire group, had preoperative chronic renal insufficiency, as defined by a creatinine level greater than 1.8 mg/d!. The condition of these patients was clearly demonstrated by the significantly reduced 5-year actuarial survival rate of 64.4%, compared with an 83.2% survival rate in patients with normal preoperative renal function. However, preoperative renal insufficiency was not a determinant of the reduction in postoperative renal function, with renal function being stable or improved 3 months after operation in 85.2% of these patients. An important factor in the preservation of renal function was the use of renal hypothermia. Multivariate analysis of the entire group of patients indicated that renal hypothermia induced with cold heparinized saline solution renal artery perfusion significantly improved postoperative renal function with an absolute risk reduction of 4.8.

Chaikof et at.

147

What measures did you and your colleagues use to protect the kidneys from ischemic injury during renal reconstruction? What are your recommendations for monitoring these patients after operation to identify an impending decline in renal function? Do you have a schedule of objective tests to evaluate the status of the renal reconstruction at regular intervals? Dr. Carl E. Bredenberg (Portland, Maine). A mortality rate of2% for concomitant aortic reconstruction and renal artery reconstruction is admirable over a 10-year interval. In 1991 we discussed a series of patients at the meeting of the New England Society for Vascular Surgery, of whom 50 had reconstruction for atherosclerosis threatening renal function. We agree with the authors that there are two things going on: You are revascularizing occluded arteries or stenosed arteries to kidneys that have intrinsic disease, presumably because of hypertension and atherosclerosis, and one is not going to return these patients to normal function simply by opening the artery. We approached these with a little more conservative approach: Only one third of the patients had aortic replacements; most of the rest underwent reconstructions with an artery other than the aorta, usually splenic, hepatic, or iliac. We had a somewhat similar response, although our group was less severely ill than your group; seven patients (15 % ) underwent dialysis within 2 years of operation. Only one of those was related to a late occlusion of a graft, therefore successful reconstruction had relatively substantial numbers of patients undergoing dialysis. In trying to predict outcome, we did as you did, and in preoperative creatinine levels less than 2.5 mg/d!, all of those patients did not undergo dialysis. Those creatinine levels between 2.5 and 4 mg/d!, about 15% continued or started to undergo dialysis, but in those patients with preoperative creatinine levels more than 5 mg/d!, only one of the six patients did not undergo dialysis. Of those patients undergoing dialysis, four of the seven died within 12 months of undergoing dialysis. Is there a threshold of preoperative creatinine level above which operative intervention may be considered fruitless? . Dr. David C. Brewster (Boston, Mass.). The experience detailed by Dr. Chaikof and his coworkers from Emory certainly is illustrative of changing trends in renovascular surgery that are evident to all of us; the focus of operation is increasingly on retrieval of renal function. Operation is more often undertaken in older patients and patients who have concomitant aortic disease. One of the most remarkable findings in this series is an operative mortality rate of only 1.9% for the combined procedures; this is especially notable when we recall that nearly 44% underwent simultaneous bilateral renal artery revascularization. Our experience, and that of virtually all other reported series, is that extension to include renal artery grafting with aortic reconstruction invariably leads to increased morbidity and mortality rates. The decision to add renal revascu-

148 Chaikof et ai.

larization must be carefully considered, and the problem is therefore one of patient selection. Would you elaborate on your current indications about when to add renal revascularization, particularly when the focus or goal is retrieval of renal function? Are there any preoperative predictors that are helpful, such as degree of stenosis, renal size, evidence of functional activity of the involved kidney on preoperative renal scan, or use of intraoperative renal biopsy? Such criteria would be helpful to the clinical surgeon for making this often difficult decision. Dr. Malcolm O. Perry (Lubbock, Texas). If one excludes the technical requirements imposed on the surgeons in these operations, we are seeing more operations performed for renal salvage and retrieval of function than the treatment of renovascular hypertension. How do we determine which patients are suitable for such procedures? It has been my impression that preservation of glomerular mass is the critical instant, not the size of the kidney. The methods of determining whether the glomerular mass is sufficiently preserved to allow retrieval, however, remains imperfect. Dr. Dean has talked about levels of creatinine that may be used as a predictor. As Dr. Brewster has indicated, it has been a terrible dilemma for us to decide which arteries to revascularize, and despite your excellent results, there is a considerable increase in morbidity and mortality rates when one combines these operations. In our 60 cases that were combined, we have a 5% mortality rate reported some years ago, but we do all have the dilemma of deciding when we should intervene. Is preservation of glomerular mass important? Is there a way to determine that before operation before we embark on these procedures? Dr. Elliott L. Chaikof. The questions raised by all the discussants could well comprise the table of contents for a symposium on ischemic renal disease. There has certainly been an interest in revascularization for preservation of renal function that now dates back more than 30 years. However, many of the questions surrounding this topic only recently have been approached with some rigor. As has been recognized, this study does not answer the most difficult questions associated with the management of ischemic nephropathy. We have merely stratified our patients into responders and nonresponders with the goal of estimating the magnitude of our ability to preserve clinically meaningful renal function. Ultimately, all of us recognize that whether a patient dies with a serum creatinine level of 2.5 or 5.5 is irrelevant. What is relevant is whether a patient dies after a 1- or 2-year history of dialysis. In other words, to what degree and in· which patient population will revascularization reduce the need for later dialysis and have a significant impact on the quality

JOURNAL OF VASCULAR SURGERY January 1994

of life? In this regard, we all need to analyze our data in terms of dialysis-free survival. We would like to believe that surgery benefits all our patients. Nonetheless, the observation of compromised dialysis-free survival in patients with preoperative serum creatinine levels greater than 3.0 mgjdl is sobering and requires further study. Dr. Stoney emphasized our lack of comprehensive patency data, which is admittedly the main limitation of this study. Only one third of our patients underwent angiography or renal scanning during the follow-up period. Surprisingly, none of those who had known graft occlusions or stenosis were dialysis dependent. Certainly, this does not exclude technical failures as a cause of later kidney failure in at least some of our patients. Further, as emphasized by several of the discussants, a number of our patients may well have had hypertensive nephrosclerosis or some other cause of progressive renal insufficiency. Additionally, Dr. Stoney has asked about the role of hypertension in later renal deterioration. Uncontrolled hypertension, as well as many known antihypertensives may very well adversely effect renal function. We could not relate adverse outcome to the use of either preoperative or postoperative angiotensin-converting enzyme inhibitors. As we begin to understand more about the pathophysiology of ischemic nephropathy and the role of chronic ischemia, blood pressure, and various medications in the cascade, which leads to kidney failure, our ability to impact on long-term outcome will surely improve. Dr. Messina has emphasized the need to identify those patients who have severe hypertension and pulmonary edema in association with renovascular disease. None of our patients had a history of pulmonary edema, which could not be attributed to heart disease. As the Michigan group have emphasized, these are important clinical markers of ischemic nephropathy and such a population clearly may benefit from revascularization. Dr. Allen has asked about our intraoperative techniques for renal preservation and methods to evaluate the reconstruction at late intervals. Mter the transection of the renal artery, we typically perfuse the kidney with several hundred milliliters of chilled, heparinized saline solution. Duplex imaging is being used with increasing frequency to monitor the late status of our reconstructions. Finally, Drs. Bredenburg, Brewster, and Perry have asked about indications for repair. Currently, we believe that renal revascularization is unequivocally indicated for the patient with a renal artery stenosis greater than 60% and renal insufficiency with a range in preoperative serum creatinine level between 1.8 and 3.0 mgjdl. Intervention for the patient with either an asymptomatic lesion or occlusive disease in association with a greater degree of functional compromise needs to be individualized.