OUP user menu

Management of renovascular disease: a review of renal artery stenting in ten studies

C.G. Isles, S. Robertson, D. Hill
DOI: http://dx.doi.org/10.1093/qjmed/92.3.159 159-167 First published online: 1 March 1999

Abstract

To evaluate the efficacy and safety of renal artery stents in renovascular disease, we identified 10 descriptive studies containing sufficient information for systematic evaluation. No randomized comparisons of stenting with angioplasty or with surgery were found. Overall, stents were placed in 416 renal arteries in 379 patients, mean age 64 years (range 27–84), 56% male. Of the stenoses, 97% were atheromatous (inter-study range 71–100%), 80% ostial (22–100%) and 31% bilateral (12–87%). The clinical indication for stenting was usually hypertension with or without mild renal impairment. Radiological indications for stenting were: narrowing of ≥50% (in 9/10 studies) as a result of elastic recoil (58%) or dissection (2%) at the time of angioplasty; restenosis some time after angioplasty (15%); or as a primary procedure (25%). Technical success was reported in 96–100% of procedures. Restenosis (≥50% narrowing), evaluated in 312/416 (75%) arteries, generally between 6 and 12 months, was 16% overall. Hypertension was cured by stenting (DBP ≤90 mmHg on no treatment) in 34/379 (9%) overall and in 34/207 (16%) of those whose renal function was normal initially. Six of 379 (1.6%) patients died within 30 days of stenting, but in only two (0.5%) was death judged to be procedure-related. Complications, other than those which led to dialysis, occurred in 42/379 (13%) patients, one third requiring intervention, ranging from blood transfusion to a surgical bypass procedure. Renal function as judged by serum creatinine concentration (SCC) improved in 26%, stabilized in 48% and deteriorated in 26% of patients whose renal function was impaired initially (SCC >133 μmol/l). In one study, with average baseline SCC >200 μmol/l, successful stenting slowed the rate of progression of renal failure when renal function was deteriorating beforehand. Nine of 379 (2.4%) patients, including 7/14 (50%) whose SCC was ≥400 μmol/l initially, required dialysis after stenting. Stenting should be offered by specialist centres as a secondary procedure for unsuccessful angioplasty, or restenosis following angioplasty, to patients with renovascular disease and uncontrolled hypertension, advancing renal failure or pulmonary oedema.

Introduction

Our understanding of renovascular disease has advanced considerably during the last decade. The reasons for this include a greater appeciation of the natural history of both the atherosclerotic and fibromuscular forms of the disease;13 the increasing recognition of athero-embolic disease;4,5 the widespread use of percutaneous transluminal angioplasty (PTRA)68 and the introduction of renal artery stenting;9 the advent of new and more effective techniques for surgical revascularization;10,11 and the realization that bilateral renovascular disease is an important and potentially reversible cause of acute and chronic renal failure in the elderly1215 and of recurrent pulmonary oedema16,17 in patients with poorly controlled hypertension and renal insufficiency.

Unfortunately these advances have not been mirrored by similar progress in the evaluation of the different forms of treatment. Ramsay, in a penetrating review of 10 angioplasty series in 1990, found that selection criteria varied, that blood pressure follow-up was haphazard and subject to measurement bias, and that five definitions of cure and 10 definitions of improvement were used. Cures were claimed for 19% (range 9–29%) of atheromatous and 50% (25–85%) of fibromuscular patients. He concluded that `if these data had been used to support the efficacy of an antihypertensive drug no regulatory body would have found them acceptable'.6

Eight years after Ramsay's review of angioplasty, enthusiasm for angioplasty alone in atheromatous disease has begun to wane. This is because a substantial proportion of patients either have an unsatisfactory result through elastic recoil or dissection, or re-present at a later date with restenosis.7 Radiologists now use stents if they fail to achieve a technically successful dilatation with PTRA, and some opt for a stent as a primary procedure. Ten separate reports of renal artery stenting have been published.1827 It seems timely therefore to review the stent studies to see what, if any, lessons have learned since the 1990 review of angioplasty.

Methods

Data sources

The ten studies analysed in this review were published between 1991 and 1997. Five of the ten were referenced in a Lancet editorial in 1997. The remaining five papers were found by Medline search or obtained from the references of review articles. Three additional early studies of stenting were identified but rejected. All three were published in 1991. The first reported the results of stenting in only 10 patients,28 the second in 11 patients29 and the third contained insufficient data for comparative analysis.30 Studies of both angioplasty and stenting in which it was not possible to distinguish the results of stenting separately, such as that by Pattynama and colleagues,31 were also excluded from the analysis.

Data extraction

The following data were extracted by systematic review of each paper: country of origin, period of study, number of patients including age and gender, the number of vessels stented and the proportion of stenoses that were atheromatous, ostial and bilateral. Patients with unilateral stenosis and contralateral occlusion were considered to have bilateral disease. Technical aspects of interest included radiological indications for stenting, type of stent, technical success (% residual stenosis) and restenosis during follow-up. Clinical outcomes in each paper were determined from reported measurements of blood pressure and serum creatinine before and after stenting. We also extracted data concerning duration of follow-up, complications and death. Technical aspects were generally more easily retrieved from the 10 papers than clinical outcomes, and four authors supplied missing data on request. For the purpose of this analysis, selected baseline data and outcomes of interest have been presented in tabular form, the results summed and averages obtained. Full details of the information extracted from each of the papers are available from CI.

Results

Baseline clinical details

Stents were placed in a total of 416 renal arteries in 379 patients (Table 1). The average age of the patients was 64 (range 27–84 years) and 56% were male. Ninety-seven per cent of the stenoses were atheromatous (range 71–100% in different studies), 80% were ostial (range 22–100%) and 31% were bilateral (range 12–87%). The clinical indication for stenting was usually hypertension with or without mild renal impairment. None of the studies included patients with pulmonary oedema.

View this table:
Table 1

Baseline clinical data in 10 stent studies

First author and yearnAge (range)Male (%)Atheromatous (%)Ostial (%)Bilateral (%)
NG, not given.
Rees, 1991 2866(48–80)4610010021
Hennequin, 1994 2155(27–74)52 71 3324
Raynaud, 1994 1858(31–77)56 83 2244
Dorros, 1995 7667(37–84)5310010021
van de Ven, 1995 2466(45–81)5410010087
Rundback, 1996 2070(57–84)45100 9220
Henry, 1996 5965(27–84)66 93 5312
Blum, 1997 6860(31–80)6510010018
Boisclair, 1997 3363(37–77)42100 5445
Harden, 1997 3267(49–79)NG100 7578
All 10 studies37964(27–84)56 97 8031

Technical aspects

Although four different angiographic inclusion criteria were used in the 10 studies, the minimum stenosis was 50% in nine, indicating that the great majority of patients had haemodynamically significant lesions (Table 2). Palmaz stents were used in eight studies, and Wallstents in two. The most common radiological indication for stent placement was failed PTRA. This was due to elastic recoil, usually of an ostial stenosis, at the time of the procedure in 58%. Two per cent were stented for dissections occurring during PTRA, and 15% for restenosis occurring some time after a previous PTRA. Three studies used stents as a primary procedure, accounting for 25% of the stents included in this review.

View this table:
Table 2

Technical aspects in 10 stent studies

First author, yearArteriesInclusion stenosis (%)Stent typeIndications for stentStents evaluated (% total sents)Restenosis (% stents evaluated)Average* time to evaluation (months)
RecDisResPri
Rec, elastic recoil; Dis, dissection; Res, restenosis; Pri, primary procedure. *Follow-ups are means except **medians.
Rees, 1991 28≥40Palmaz71 029 018 (64) 7 (39) 8
Hennequin, 1994 21≥70Wallstent38 062 020 (95) 4 (20)29
Raynaud, 1994 18≥60Wallstent6611 61718 (100) 2 (11)11
Dorros, 1995 92≥70Palmaz 0 0 010056 (61)14 (25) 7
van de Ven, 1995 28≥50Palmaz58 0 04223 (82) 3 (13) 6**
Rundback, 1996 24≥60Palmaz83 017 016 (67) 3 (19) 6
Henry, 1996 64≥70Palmaz66 331 054 (84) 5 (9)14
Blum, 1997 74≥50Palmaz85 114 074 (100) 8 (11)24
Boisclair, 1997 35≥60Palmaz83 017 0 9 (23) 0 (0) 8
Harden, 1997 32≥50Palmaz94 6 0 024 (75) 3 (13) 6**
All 10 studies416≥50 in 98 Palmaz58 21525312 (75)49 (16) 6–12
  studies2 Wallstent in 7 studies

Technical success, using four criteria that ranged from <50% to <10% residual stenosis, ranged from 96–100%. Restenosis, defined as ≥50% narrowing in all 10 studies, was evaluated by angiography or doppler in 312/416 (75%) arteries. The average time for re-evaluation was relatively short at 6–12 months in 7/10 studies. The overall restenosis rate was 49/312 (16%). Restenosis was 30/136 (22%) in the first five studies and 19/176 (11%) in the second five, suggesting improved stent technology and/or radiological expertise with time. Three studies reported secondary patency rates in excess of 90% at intervals of up to 5 years in a total of 148 patients after 14 had had a second procedure for restenosis. Most authors felt that restenosis was common enough to justify follow-up and re-intervention.

Blood pressure

Hypertension was present in nearly all patients, and was the main clinical reason for stenting in 58% (i.e. those with serum creatinine concentration <133 μmol/l). All ten studies categorized blood pressure after stenting as cured, improved or unchanged. The criterion for cure—diastolic pressure ≤90 mmHg on no therapy—was the same in all ten studies, but the definition of improvement varied. Blood pressure was cured by stenting in 34/379 (9%) overall and in 34/207 (16%) of those whose renal function was normal initially (Table 3). Because none of the studies described their technique of blood pressure measurement, gave a formal trial of medical therapy before stenting or reported serial measurements of blood pressure before and after stenting, we did not consider the category `improved' to be sufficiently robust for detailed analysis. Average blood pressures during follow up when patients were taking fewer antihypertensive drugs, were lower than immediately before stenting suggesting that, at the very least, stenting had reduced the need for antihypertensive medication in most patients.

View this table:
Table 3

Blood pressure at baseline and during follow up in 10 stent studies

First author and yearInitial BP (mmHg)Average drugs (initially)Follow-up BP (mmHg)Average drugs (follow-up)BP cured by stenting (%)BP cured in subgroup with normal SCC (%)
NG, not given; SCC, serum creatinine concentration. 1BP data reported as cured, improved or unchanged only; 2average blood pressure for 17/18 patients—BP not given for one patient on haemodialysis; 3BP given for 54/59 patients only—data missing in five patients; 4systolic and diastolic during follow-up not given, but mean arterial pressure fell from 132 to 112 mmHg at 12 months in this study; 5blood pressure shown is median not mean.
Rees, 19911NGNGNGNG 3/28 (11) 3/14 (21)
Hennequin, 1994182/107NG138/78NG 3/21 (14) 3/15 (20)
Raynaud, 19942179/1052.2156/881.6 1/18 (6) 1/10 (10)
Dorros, 1995168/872.3156/812.0 4/76 (6) 4/46 (9)
van de Ven, 1995200/105NG160/90NG 0/24 (0) NG
Rundback, 1996179/862.4143/781.8 0/20 (0) 0/9 (0)
Henry, 19963172/981.8144/811.410/59 (17)10/49 (20)
Blum, 19974188/1052.9NGNG11/68 (16)11/48 (23)
Boisclair, 1997165/872.6155/802.1 2/33 (6) 2/16 (13)
Harden, 19975169/951.6164/871.4 0/32 (0) 0 (0)
All 10 studiesBP and number of drugs were lower during follow-up34/379 (9)34/207 (16)

Renal function

Comparison of baseline and follow-up serum creatinine concentrations (SCC) is shown in Table 4. Overall, 42% patients were considered to have a degree of renal impairment, as judged by baseline SCC >133 μmol/l (1.5 mg%). Renal function was improved after stenting in 26% of these, deteriorated in 26% and remained unchanged in 48%. As a result the average SCC during follow-up was not significantly different from baseline. One study with a median baseline SCC of 257 μmol/l reported serial measurements of SCC before and after stenting, and showed that successful stenting can slow the rate of progression of renal failure in patients whose renal function is deteriorating beforehand.26 Only 18 (5%) patients overall had SCC >400 μmol/l initially. Four of these were dialysis-dependent at the time of stenting, and two recovered sufficient renal function after stenting to discontinue dialysis, but seven of the remaining 14 patients (50%) required haemodialysis during follow-up. In all, nine (2.4%) patients started dialysis after stenting. Partial recovery of renal function occurred in two patients, while the other seven remained dialysis-dependent. Two of the nine died during follow-up.

View this table:
Table 4

Renal function at baseline and during follow-up in 10 stent studies

First author and yearBaseline creatinineFollow-up creatinineCreatinine>133 μmol/l (%)Improved (%)Unchanged (%)Deteriorated (%)Creatinine>400 μmol/lDialysis before stentFirst dialysis after stent
ARF,acute renal failure; CRF, chronic renal failure. Baseline and follow-up creatinine in μmol/l are means except for: 1not given but <176 (Rees et al.) and <164 (Hennequin et al.) from data in papers; 2mean creatinine excluding one patient on haemodialysis; 3median creatinine excluding two patients with CRF; 4median creatinine for whole group at baseline, follow-up reported as change in slope of reciprocal creatinine. Criteria for improvement and deterioration in renal function were ≥ 10–20% change in creatinine in 7/10 studies. In two studies the change in creatinine was not specified, and in one study the change in creatinine was >18 μmol/l.
Rees 19911<176No change 14 (50) 5(36) 5(36) 4(28) 102
Hennequin 19941<164No change  6(29) 1(17) 3(49) 2(34) 000
Raynaud 1994 2137148  8 (44) 1(13) 4(50) 3(37) 111
Dorros 1995150150 30 (39) 8(28) 8(28)13(45) 000
van de Ven 19953143133 NG NG NG NG 001
Rundback 1996146153 11 (55) 2(25) 6(75) 0(0) 000
Henry 1996125132 10 (17) 2(20) 6(60) 2(20) 201
Blum 1997109108 20 (29) 0(0)20(100) 0(0) 000
Boisclair 1997143144 17 (52) 7(41) 6(35) 4(24) 101
Harden 19974257NG 32 (100)11(34)11(34) 9(28)1333
All 10 studies<200 μmol/l in 9 studiesNo significant  change148/355  (42)37(26)69(48)37(26)18(5)49(2.4)

Complications

Six of the 379 (1.6%) patients died within 30 days of stenting, but in only 2/379 (0.5%) was death judged to have been related to the procedure. Complications, other than those which led to dialysis, were reported by the authors of the 10 studies in 42/379 (11%) patients. If the need for dialysis is included, then 51/379 (13.5%) patients may have experienced a complication as a result of stenting, though it is difficult to be certain how many patients starting dialysis did so as a result of the procedure itself, or by progression of their underlying disease.

Some complications were serious, others less so. The estimates of the frequencies of some complications must necessarily be regarded as imprecise, given the unsystematic way in which they were reported. Complications for which specific interventions were required included haemorrhage requiring transfusion or drainage (4), femoral artery pseudoaneurysm requiring ultrasound guided compression (3), damage to brachial artery requiring surgical repair (2), guidewire perforation of renal artery leading to surgical bypass (1), stent thrombosis treated by urokinase (1), distal renal embolus requiring streptokinase (1), dissection of iliac artery which was treated by stenting (1) and septicaemia requiring intravenous antibiotic (1). Complications for which no specific intervention was required included worsening renal failure that stabilized or recovered without dialysis (12), distal renal embolism or occlusion that was either asymptomatic or caused lumbar pain only (5), haematoma at puncture site (5), guidewire perforation of renal artery (4), asymptomatic stent occlusion (1) and cholesterol embolism to lower limbs (1).

Discussion

The optimal management of renovascular disease remains to be determined. This is partly because this complex condition is a heterogeneous disorder, and partly because clinicians and radiologists have been reluctant to submit their patients to randomized comparisons of treatment. Lack of trial evidence, however, has not stood in the way of strongly-held opinion, which includes the recent claim that `angioplasty is now the accepted first line therapy for renal artery stenosis'.9

Two randomized trials of intervention by PTRA versus continued medical therapy, conducted in Scotland and France, have since been published. In the Scottish study, 27 patients with unilateral and 28 with bilateral atherosclerotic RAS were randomized separately to PTRA or medical therapy,32 whereas all 49 patients in the French study had unilateral atherosclerotic RAS.33 Importantly, both trials included a run-in period on a standardized antihypertensive drug regimen. The main findings were: no difference between the two treatment arms for blood pressure control or serum creatinine in patients with unilateral RAS, in either study. In the French study, small falls in clinic blood pressures were not confirmed by the more objective 24-h ambulatory blood pressure recordings. Patients with bilateral RAS undergoing PTRA in the Scottish study had lower systolic blood pressure during follow-up (p=0.02), but no overall change in serum creatinine, an unexpected result. The authors of both studies concluded that previous uncontrolled assessments of PTRA overestimated its potential for lowering blood pressure. A third comparison of PTRA and medical therapy, the Dutch Renal Artery Stenosis Co-operative (DRASTIC) Study is due to report in 1998 (F. Derkx, personal communication).

Historically, the main form of intervention in renovascular disease has been surgery, which has many advocates and for which published results are at least as good as those reported for angioplasty. Thus, it has been claimed that surgical intervention by aortorenal and more recently hepatorenal or splenorenal bypass may be an effective form of therapy for renovascular disease with poorly controlled hypertension, progressive renal failure24 or pulmonary oedema.5,6 The same methodological criticisms apply however: namely frequent failure to describe the method of blood pressure measurement; absence of a formal trial of medical therapy pre-operatively; and lack of serial measurements of blood pressure or renal function before and after surgery.

In the only published randomized comparison of PTRA and surgery, 58 low-risk patients (unilateral RAS, age <70 years, not diabetic and serum creatinine <300) were enrolled.34 Secondary patency rates (after four patients in the PTRA group had an operation and one surgical patient had PTRA) were similar (97% surgical and 90% PTRA). Using an intention-to-treat analysis, 3/29 PTRA and 5/29 surgical patients were cured of their hypertension (BP <140/90 on no treatment). The authors concluded that PTRA should be the treatment of choice for atherosclerotic renal artery stenosis causing renovascular hypertension, if combined with intensive follow up and aggressive re-intervention. They also conceded during discussion that their patients were highly selected, and that their results could not be extrapolated to the more commonly encountered complex elderly patient with bilateral disease, renal insufficiency and atherosclerotic disease at other sites.34 There has been one other randomized comparison of PTRA and surgery, at Vanderbilt in the 1980s, but this study experienced recruiting difficulties, and its findings have never been published (R.H. Dean, personal communication).

The recent advent of renal artery stenting seems likely to have reduced rather than enhanced the prospect of further randomized comparisons. None of the ten studies described in this review have compared stenting with any other treatment modality, yet all ten have concluded that stenting is technically feasible, without major complication and an effective form of therapy for ostial stenoses. Examination of the data in these studies suggests that their authors' confidence in the benefits for blood pressure and renal function may be misplaced. Cure of hypertension may be possible in 16% patients in whom the clinical indication for stenting is hypertension without renal impairment. Methodological problems mean that it is difficult to interpret the category of `improved' blood pressure control, but the fact that patients were taking fewer antihypertensive medications during follow-up, suggests that stenting does at least have a drug-sparing effect. Successful stenting may also permit the use of ACE inhibitors, provided that renal function continues to be monitored closely.22

Overall, renal function did not improve significantly after stenting in the patients with renal impairment in these ten studies, but nor did it worsen. Given the natural history of renovascular disease which is to progress from unilateral to bilateral, from stenosis to occlusion, and from normal to elevated creatinine concentration,1 these results might indicate that renal function was preserved by stenting. Only one of the ten studies reported serial measurements of SCC before and after stenting, and this strongly suggested (Figure 2) that successful stenting may slow the rate of progression of renal failure in patients whose renal function is deteriorating beforehand. These are however short-term results (median follow-up was 6 months) in a small number of patients. We have even less data on patients with more advanced renal failure. Only 18/379 (5%) patients had baseline serum creatinine >400 μmol/l. Two of the four patients who were on regular haemodialysis before stenting managed to discontinue dialysis, but seven of the remaining fourteen patients (50%) required haemodialysis during follow-up (Table 4). Clearly we need more data on the effects of stenting in patients with severe renal failure.

Figure 1.

Ten renal artery stent studies published between 1991 and 1997 showing centre of origin, duration of study and number of patients.

Figure 2.

Changes in slope of reciprocal creatinine before and after stent in 23 patients. The slope value is reciprocal creatinine×10−6 l μmol−1 day−1. The upward change in 18 patients indicates slowing in the rate of progression. Reproduced from reference 26 by permission.

The price to be paid for these benefits is that 10–15% are likely to restenose within a year of stenting, although with aggressive reintervention secondary patency rates of up to 90% at 5 years may be possible. A similar proportion of patients will experience a complication of stenting, and one-third of these will require some form of further intervention as a result. Specific therapies will range from blood transfusion, ultrasound compression of a pseudoaneurysm, thrombolytic therapy for stent thrombosis, surgical repair or stenting of a damaged vessel through to a surgical bypass procedure. Fortunately procedure-related deaths are very uncommon, at <1%.

Where do these findings leave renal artery stenting in the management of renovascular disease? The time is ripe for a carefully designed large-scale randomized comparison of percutaneous renal angioplasty with or without stenting plus best medical treatment versus best medical treatment alone, in high-risk patients with renovascular disease who are considered suitable for both forms of intervention. Such a study will shortly begin recruiting patients and will include a second randomization between angioplasty alone and angioplasty with stenting for those randomized to the intervention group.35 In the continuing absence of a randomized comparison, we recommend that stenting be offered by specialist centres as a secondary procedure for unsuccessful PTRA, or restenosis following PTRA, to patients with renovascular disease and uncontrolled hypertension, advancing renal failure or pulmonary oedema. If coupled with a greater commitment to managing patients by agreed protocols, this approach would undoubtedly yield useful descriptive data on the clinical outcome of stenting. Such a strategy is clearly a compromise, but does at least have the attraction of limiting a potentially invasive procedure to those at highest risk of an event if left untreated, on the grounds that they are likely to have the most to gain by successful treatment. It is also supported by reports indicating that worthwhile improvements in renal function3639 and pulmonary oedema16,17 are possible in selected patients.

Acknowledgements

We thank Mrs Josephine Campbell for her help in the preparation of this paper.

Footnotes

  • Author note: Since this paper was accepted, a randomized comparison of angioplasty with stent (n=40) versus angioplasty alone (n=41) in ostial atherosclerotic renovascular disease has been published (van de Ven, Lancet 1999; 353:282–6). Primary patency rates at six months were 75% for stent and 29% for angioplasty. The corresponding secondary patency rates were 80% and 51%. Despite the better angiographic result there were no differences in blood pressure and renal function between the two groups at six months. The authors concluded that primary stenting should be the preferred procedure for ostial atherosclerotic renovascular disease.

References

View Abstract