Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
J Vasc Surg. Author manuscript; available in PMC 2010 September 3.
Published in final edited form as:
PMCID: PMC2933104

Restenosis after renal artery angioplasty and stenting: Incidence and risk factors



Management of renal artery stenosis (RAS) with primary renal artery percutaneous angioplasty and stenting (RA-PTAS) is associated with a low risk of periprocedural death and major complications; however, restenosis develops in a subset of patients and repeat intervention may be required. We examined the incidence of restenosis after RA-PTAS and associations with clinical factors.


Consecutive patients undergoing RA-PTAS for hemodynamically significant atherosclerotic RAS associated with hypertension or ischemic nephropathy, or both, between October 2003 and September 2007 were identified from a registry. Restenosis was defined using duplex ultrasound (DUS) imaging as a renal artery postintervention peak systolic velocity (PSV) ≥180 cm/s. The incidence and temporal distribution of restenosis was analyzed using survival analysis based on treated kidneys. Associations between clinical factors and recurrent stenosis were examined using proportional hazards regression.


RA-PTAS was performed on 112 kidneys for atherosclerotic RAS during the study period. Initial postintervention renal artery DUS imaging confirming PSV <180 cm/s in 101 kidneys, which formed the basis of this analysis. Estimated restenosis-free survival was 50% at 12 months and 40% at 18 months. Decreased risk of restenosis was associated with preoperative statin use (hazard ratio [HR], 0.35; 95% confidence interval [CI], 0.16–0.74; P = .006) and increased preoperative diastolic blood pressure (DBP; HR, 0.70 per 10-mm Hg increase in preoperative DBP; 95% CI, 0.49–0.99; P = .049). No other factors assessed were associated with restenosis.


Restenosis occurs in a substantial number of patients treated with RA-PTAS. Preoperative statin medication use and increased preoperative DBP are associated with reduced risk of restenosis. In the absence of contraindications, statins should be considered standard therapy for patients with atherosclerotic renal artery stenosis.

Atherosclerotic renal artery stenosis (RAS) may cause severe hypertension or renal dysfunction, or both, and is associated with an increased risk for cardiovascular events.1 Patients with atherosclerotic RAS who have poorly controlled hypertension or progressive deterioration of renal function, or both, despite appropriate medical management, often are treated with renal artery revascularization, although the benefits associated with treatment remain controversial. Renal artery percutaneous angioplasty and stenting (RA-PTAS) has become the most common method of renal artery intervention. Unfortunately, restenosis has been reported in 17% to 44% of arteries after RA-PTAS28 and may be associated with recurrent clinical symptoms despite an initial treatment response.

Analyses to date have identified associations between restenosis after RA-PTAS and renal artery diameter,5,9 stent diameter,10 weight/body mass index,11 and smoking.12 The purpose of this study was to determine the frequency and predictors of restenosis after primary RA-PTAS in a single-center cohort of adult patients with atherosclerotic RAS.


Study population

This investigation was conducted with the approval of the Wake Forest University Health Sciences Institutional Review Board. Consecutive primary RA-PTAS procedures performed for hemodynamically significant atherosclerotic RAS were identified from a procedure registry. All treated patients had hypertension with or without renal dysfunction, and indications for RA-PTAS were determined by individual operators. The analysis excluded RA-PTAS performed for restenosis or fibromuscular dysplasia.

All procedures were performed by vascular surgeons at the Wake Forest University School of Medicine between October 2003 and September 2007. Standard preparation, procedural management, and follow-up for patients treated with RA-PTAS at our center have been described previously.13,14 Balloon-mounted stents were used in all patients and sized to match the diameter of the distal, normal-caliber renal artery as measured by angiography at the time of treatment. Renal duplex ultrasound (DUS) imaging was performed before the intervention and ≤24 hours after RA-PTAS. Thereafter, routine DUS surveillance of the renal artery was conducted at 1 and 6 months, and then at 6-month intervals.

Data collection and management

Clinical data, including patient demographics, comorbidities, and laboratory results, were retrospectively collected from the electronic medical record. Presence of left ventricular hypertrophy was determined by baseline electrocardiography. Renal DUS data, including peak systolic velocity (PSV), resistive index, acceleration time, and kidney length, were collected from a prospectively maintained clinical vascular laboratory database.

Anatomic information was retrospectively collected from angiograms performed during RA-PTAS by manual electronic caliper measurement of archived images. All measurements were performed in a nonblinded fashion by a single individual. Angiographic percentage of RAS was determined by measuring the smallest luminal diameter at the point of maximal stenosis and comparing it with the lumen of the main renal artery distal to the lesion and any poststenotic dilatation, if present. Anatomic locations of stenotic lesions were classified as described by Baumgartner et al.3

Estimated glomerular filtration rate (eGFR) was calculated from the serum creatinine level using the abbreviated Modification of Diet in Renal Disease formula.15 Renal dysfunction was categorized as severe for patients with an eGFR <30 mL/min/1.73 m2, moderate for patients with eGFR of 30 to 60 mL/min/1.73 m2, and none for patients with an eGFR >60 mL/min/1.73 m2. Renal artery resistive index was calculated from segmental velocities as [1–(end diastolic velocity/peak systolic velocity)].16

Assessment of restenosis

Hemodynamically significant recurrent RAS was identified using DUS imaging. Patients underwent routine DUS examinations immediately after treatment and then at 1, 3, and 6 months. DUS studies were performed using a 5.2-MHz curvilinear probe with Doppler color flow, with either a Philips IU22 (Philips Healthcare, Andover, Mass) or an ATL HDI 5000 (Advanced Technology Laboratories, Bothell, Wash) US system using a previously described technique.17

Restenosis was defined as renal artery PSV ≥180 cm/s in a stented artery previously documented as free of restenosis. This DUS criterion for renal artery restenosis has been used by others1820 and was internally validated at our center (Fig 1, online only). Renal artery PSV ≥180 at the time of the first postintervention DUS study was considered residual stenosis and not interpreted as restenosis. Repeat renal artery angiography or intervention, or both, was undertaken at the discretion of individual operators in the setting of hemodynamically significant restenosis on follow-up DUS imaging accompanied by deterioration of renal function after initial postintervention improvement or worsening of hypertension (eg, increased number of anti-hypertensive agents, increase in blood pressure, or hypertensive emergency) after an initial hypertension response to RA-PTAS.

Fig 1
Validation of peak systolic velocity (PSV) ≥180 cm/s as indicator of renal artery restenosis. The horizontal dashed line indicates PSV of 180 cm/s, and the vertical dashed line indicates angiography-defined stenosis of 60%. In our own institutional ...

Statistical analysis

Data on baseline patient comorbid medical conditions, blood pressure, medication use, renal function data, and demographics are reported using mean ± standard deviation for continuous variables and count (%) for categoric variables. Restenosis was assessed for the kidneys using models controlling for within-subject correlation, and data are reported using model-based mean ± standard error. The incidence and temporal distribution of recurrent RAS were analyzed using survival analysis based on treated kidneys. Associations between clinical factors and time to restenosis were examined using Cox proportional hazards regression. Both of these methods accounted for correlated observations. Cox proportional hazards regression modeling was performed using stepwise selection (P ≤ .10 for model entry). Candidate covariates for model selection are listed in Table I. Results were evaluated for significance using α = 0.05, and hazard ratios (HR) are expressed with 95% confidence intervals (CI). Statistical analyses were performed using SAS 9.1 software (SAS Institute, Cary, NC).

Table I
Candidate covariates for Cox proportional hazards modeling


Incidence of restenosis and associations with clinical factors

Primary RA-PTAS for atherosclerotic RAS was attempted in 112 kidneys during the study period. Of these, eight were excluded due to residual stenosis on initial postintervention renal artery DUS imaging, two were excluded due to lack of renal artery DUS follow-up data, and one was excluded as a technical failure related to inability to access the target renal artery. The remaining 101 kidneys in 91 patients form the basis of this analysis.

Mean preintervention angiographic RAS was 79.1% ± 12.9%. Of these stenotic lesions, 73% were categorized as ostial, 18% involved the proximal renal artery, and 9% involved the truncal renal artery. Distal renal artery balloon occlusion was used for embolic protection during 90% of procedures. Patient baseline demographics are reported in Table II. Mean patient age was 68.8 ± 10.1 years, all patients had hypertension, 53% were women, and 87% were white. According to the eGFR, moderate or severe renal insufficiency was observed in 72% of patients. The mean preintervention renal artery diameter was 5.6 ± 0.1 mm, and mean stent diameter was 5.7 ± 0.1 mm.

Table II
Demographics of the 91 study patients

There were no periprocedural deaths. Estimated restenosis-free survival adjusted for within-subject correlation was 50% at 12 months and 40% at 18 months (Fig 2). Proportional hazards regression analysis demonstrated decreased risk for restenosis associated with preoperative statin use (HR, 0.35; 95% CI, 0.16–0.74; P = .006) and preoperative diastolic blood pressure (DBP; HR, 0.70 per 10-mm Hg increase in preoperative DBP; 95% CI, 0.49–0.99; P = .049). No other covariates assessed (Table I) were associated with restenosis-free survival. Predicted restenosis-free survival stratified by statin medication use is displayed graphically in Fig 3.

Fig 2
Estimated restenosis-free survival. The estimation method accounts for the correlated data. The standard error of the survival estimate is <0.1 for the displayed postintervention interval.
Fig 3
Time to restenosis stratified by statin medication use. Estimation method accounts for correlated data. Line becomes broken when the standard error of the survival estimate is > 0.1.

Clinical manifestations and management of restenosis

Clinical manifestations and management of restenosis are summarized in Table III. At a mean postintervention interval of 5.5 months, 28 recurrent lesions were identified in 27 patients. Bilateral restenosis developed in one patient managed with staged bilateral primary RA-PTAS, whereas the remaining restenoses were unilateral. In 17 of 27 patients (63%) with restenosis, there were no associated clinical manifestations such as worsening of hypertension, need for resumption of previously discontinued antihypertensive agents, or decline in eGFR after an initial improvement. In the setting of improvement after RA-PTAS in hypertension control or renal function, or both, these patients were managed with continued medical therapy and DUS surveillance without repeat angiography. The remaining 10 patients (37%) with restenosis identified by DUS imaging had associated hypertension or a decline in eGFR, or both, prompting repeat intervention. Angiography findings in these patients confirmed the presence of ≥60% diameter-reducing instent restenosis in all arteries.

Table III
Clinical manifestations and procedural management of renal artery stenosis after renal artery percutaneous angioplasty and stenting

Among individuals who underwent repeat intervention for restenosis, procedural management initially consisted of surgical revascularization in one patient and repeat angioplasty in nine; four of the nine repeat angioplasties were performed using cutting balloons. A second restenosis developed in one patient managed with repeat angioplasty and was later treated with aortic endarterectomy plus renal artery bypass. One patient with restenosis was hospitalized for an acute hypertensive emergency associated with pulmonary edema and improved clinically with aggressive medical management. A DUS examination after discharge from the hospital revealed interval progression of the restenosis to occlusion.


This analysis represents an attempt to characterize the incidence of restenosis after RA-PTAS documented by DUS follow-up and to characterize associated risk factors. In a kidney-based analysis of primary RA-PTAS for atherosclerotic disease, estimated risk of restenosis was 50% at 12 months and 60% at 18 months. Clinical manifestations, including worsening of hypertension or decline in eGFR, or both, were observed frequently in patients with restenosis, and these patients were most often managed with repeat endovascular treatment. In multivariable proportional hazards analysis, use of statin medications and preintervention DBP were both associated with decreased risk of restenosis.

DUS imaging has demonstrated validity for identification of restenosis after RA-PTAS at a number of centers.1821 Among patients studied with renal DUS imaging before digital subtraction angiography, Bakker et al18 observed a 100% sensitivity and 74% specificity for PSV >180 cm/s in determining the presence of restenosis after RA-PTAS. Using this criterion, the incidence of restenosis reported in this study is comparable with the incidence reported by others using DUS imaging for surveillance of stented renal arteries,68 although our validation analysis (Fig 1, online only) shows PSV >180 cm/s may underestimate the true incidence of anatomic restenosis.

In the current study, restenosis was diagnosed at a relatively early interval after primary RA-PTAS of 5.5 months and was associated with hypertension or decline in eGFR, or both, in 37% of patients. The poor primary patency of RA-PTAS for atherosclerotic RAS and the frequent need for repeat intervention underscore the current controversy regarding benefits of percutaneous treatment of RAS, particularly when considering the modest hypertension and renal function benefits observed with this method.14

Previously described associations with restenosis or the need for repeat intervention after RA-PTAS include body weight and body mass index,11 renal artery diameter,5 stent diameter,9,10 and smoking.12 We were unable to specifically evaluate body mass index as a predictor of restenosis due to incomplete patient height data, but did not observe a significant relationship between weight or any of these other factors and restenosis in the present study. Post hoc power analysis confirmed ≥80% power to detect hazard ratios >3 for these variables, but power to detect smaller effect sizes was limited and may have resulted in a type II statistical error.

Statin use was associated with a significant reduction in restenosis risk, however, and patients not treated with statin medications had a nearly threefold increase in estimated risk for restenosis over time. Although an association between statin use and restenosis after RA-PTAS has not been described previously, to our knowledge, reduction in coronary artery restenosis associated with statin use has been observed,22 and similar effects on postinjury renal artery remodeling would therefore seem plausible. Potential mechanisms through which statin use may have reduced the risk of restenosis include effects on serum cholesterol or the composition and morphology of atheromatous plaques, or both. Unfortunately, lipid data within this retrospectively collected data set was incomplete, precluding meaningful interpretation of cholesterol values in relationship to statin use or restenosis.

Beneficial pleiotropic effects of statins include favorable influences on atherosclerotic plaque stability, inflammation, endothelial function, matrix metalloproteinase activity, and nitric oxide bioavailability.23,24 Decreased incidence of anatomic progression of atherosclerotic RAS was observed with statin use by Cheung et al.25 Presumably, this beneficial effect on natural disease history might also translate into protection against secondary progression of atherosclerosis after RA-PTAS.

The relatively rapid development of restenosis we observed after intervention, however, seems more consistent with neointimal hyperplasia rather than secondary progression of atherosclerotic disease as the responsible pathophysiologic mechanism. Inhibition of vascular smooth muscle cell migration and proliferation, along with induction of neointimal smooth muscle cell apoptosis, are previously described in vitro statin effects26,27 that make protection against neointimal hyperplasia after RA-PTAS a plausible hypothesis. Statin-induced plaque stabilization resulting from increased collagen and reduced lipid content has also been characterized28 and may contribute to a more favorable interface between the stent and the stenotic lesion, theoretically reducing odds of plaque rupture through stent interstices at the time of deployment.

We also observed an association between increasing preintervention DBP and freedom from restenosis. Unlike the protective effect associated with statin use, this relationship was unanticipated. Activation of the renin-angiotensin system is a known stimulus for neointimal hyperplasia.29 Hypothetically, more frequent use of angiotensin-converting enzyme (ACE) inhibitors or angiotensin receptor blockers, or both, among patients with higher baseline DBP might have resulted in a confounding protective effect against restenosis. However, no significant differences in baseline blood pressure, number of antihypertensive agents, ACE inhibitors, or angiotensin receptor blockers were noted between patients according to restenosis status (Table II), raising the possibility that this observation may represent a type I statistical error. In the absence of supporting data from other studies, we can only speculate about the unexpected relationship observed between baseline DBP and freedom from restenosis.

This study provides novel information about factors associated with restenosis after RA-PTAS; however, several additional limitations exist that deserve comment. Although restenosis was verified with angiography among all patients undergoing repeat intervention, most patients with restenosis did not have associated clinical sequelae and therefore did not undergo additional confirmatory imaging. Increased rates of positive angiography results have been observed when patients are studied in the setting of clinical suspicion,5 and the incidence of restenosis therefore may have been overestimated due to false-positive DUS results. Such a possibility is supported by the findings of Bakker et al,18 who determined renal artery DUS imaging to be 100% sensitive but only 74% specific for restenosis after RA-PTAS.

Analysis of clinical disease recurrence (ie, anatomic restenosis associated with worsening of hypertension or renal dysfunction), as an alternative outcome therefore might have increased the specificity of our findings and permitted angiographic confirmation of DUS results among all patients experiencing the event. We instead defined restenosis with DUS imaging given the reliability of this imaging method at our center and its widespread use as the primary postintervention surveillance method in current clinical practice.

Nonrandom allocation of patients to statin therapy in this retrospective study also might have resulted in bias if statin nonuse was a global indicator of suboptimal medical management, although the stratified comparisons in Table II seem to argue against this possibility. Finally, symptoms of restenosis that occurred subsequent to development of a detectable anatomic lesion might have been missed due to the relatively short follow-up in this study, which also precluded analysis of late restenosis.


Restenosis after primary RA-PTAS for atherosclerotic RAS occurs frequently and is often accompanied by physiologic manifestations. Considered in combination with the beneficial extrarenal cardiovascular effects of statins, the decreased risk of restenosis associated with use of these medications supports their routine use in patients undergoing RA-PTAS.


Dr Edwards is supported by the American Vascular Association and Lifeline Foundation Research Career Development Award. Grant support was also provided by The National Heart, Lung, and Blood Institute of the National Institutes of Health to Dr Edwards (Grant 1K23HL083981-01), Dr Geary (Grant R01 HL57557), and Dr Hansen (Grant 5K12HL083763-02).


Competition of interest: none.

Presented at the Thirty-third Annual Meeting of the Southern Association of Vascular Surgery, Tucson, Ariz, Jan 14, 2009.

Additional material for this article may be found online at


Conception and design: MC, ME, JA, JP, RG, KH

Analysis and interpretation: MC, ME, JS, KH

Data collection: MC, JP, JS

Writing the article: MC, ME, JS, KH

Critical revision of the article: MC, ME, JA, JP, RG, KH

Final approval of the article: MC, ME, JA, JP, RG, KH

Statistical analysis: JA, MC

Obtained funding: Not applicable

Overall responsibility: MC


1. Edwards MS, Craven TE, Burke GL, Dean RH, Hansen KJ. Reno-vascular disease and the risk of adverse coronary events in the elderly: a prospective, population-based study. Arch Intern Med. 2005;165:207–13. [PubMed]
2. Leertouwer TC, Gussenhoven EJ, Bosch JL, van Jaarsveld BC, van Dijk LC, Deinum J, et al. Stent placement for renal arterial stenosis: where do we stand? A meta-analysis. Radiology. 2000;216:78–85. [PubMed]
3. Baumgartner I, von Aesch K, Do DD, Triller J, Birrer M, Mahler F. Stent placement in ostial and nonostial atherosclerotic renal arterial stenoses: a prospective follow-up study. Radiology. 2000;216:498–505. [PubMed]
4. Dorros G, Jaff M, Jain A, Dufek C, Mathiak L. Follow-up of primary Palmaz-Schatz stent placement for atherosclerotic renal artery stenosis. Am J Cardiol. 1995;75:1051–5. [PubMed]
5. Lederman RJ, Mendelsohn FO, Santos R, Phillips HR, Stack RS, Crowley JJ. Primary renal artery stenting: characteristics and outcomes after 363 procedures. Am Heart J. 2001;142:314–23. [PubMed]
6. Nolan BW, Schermerhorn ML, Powell RJ, Rowell E, Fillinger MF, Rzucidlo EM, et al. Restenosis in gold-coated renal artery stents. J Vasc Surg. 2005;42:40–6. [PubMed]
7. Nolan BW, Schermerhorn ML, Rowell E, Powell RJ, Fillinger MF, Rzucidlo EM, et al. Outcomes of renal artery angioplasty and stenting using low-profile systems. J Vasc Surg. 2005;41:46–52. [PubMed]
8. Tullis MJ, Zierler RE, Glickerman DJ, Bergelin RO, Cantwell-Gab K, Strandness DE., Jr Results of percutaneous transluminal angioplasty for atherosclerotic renal artery stenosis: a follow-up study with duplex ultrasonography. J Vasc Surg. 1997;25:46–54. [PubMed]
9. Bates MC, Rashid M, Campbell JE, Stone PA, Broce M, Lavigne PS. Factors influencing the need for target vessel revascularization after renal artery stenting. J Endovasc Ther. 2006;13:569–77. [PubMed]
10. Vignali C, Bargellini I, Lazzereschi M, Cioni R, Petruzzi P, Caramella D, et al. Predictive factors of instent restenosis in renal artery stenting: a retrospective analysis. Cardiovasc Intervent Radiol. 2005;28:296–302. [PubMed]
11. Kane GC, Hambly N, Textor SC, Stanson AW, Garovic VD. Restenosis following percutaneous renal artery revascularization. Nephron Clin Pract. 2007;107:c63–9. [PubMed]
12. Shammas NW, Kapalis MJ, Dippel EJ, Jerin MJ, Lemke JH, Patel P, et al. Clinical and angiographic predictors of restenosis following renal artery stenting. J Invasive Cardiol. 2004;16:10–3. [PubMed]
13. Edwards MS, Craven BL, Stafford J, Craven TE, Sauve KJ, Ayerdi J, et al. Distal embolic protection during renal artery angioplasty and stenting. J Vasc Surg. 2006;44:128–35. [PubMed]
14. Corriere MA, Pearce JD, Edwards MS, Stafford J, Hansen KJ. Endovascular management of atherosclerotic renal artery stenosis: early results following primary intervention. J Vasc Surg. 2008;48:580–7. [PubMed]
15. Levey AS, Bosch JP, Lewis JB, Greene T, Rogers N, Roth D. A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation. Modification of Diet in Renal Disease Study Group. Ann Intern Med. 1999;130:461–70. [PubMed]
16. Radermacher J, Chavan A, Bleck J, Vitzthum A, Stoess B, Gebel MJ, et al. Use of Doppler ultrasonography to predict the outcome of therapy for renal-artery stenosis. N Engl J Med. 2001;344:410–7. [PubMed]
17. Hansen KJ, Tribble RW, Reavis SW, Canzanello VJ, Craven TE, Plonk GW, Jr, et al. Renal duplex sonography: evaluation of clinical utility. J Vasc Surg. 1990;12:227–36. [PubMed]
18. Bakker J, Beutler JJ, Elgersma OE, de Lange EE, de Kort GA, Beek FJ. Duplex ultrasonography in assessing restenosis of renal artery stents. Cardiovasc Intervent Radiol. 1999;22:475–80. [PubMed]
19. Napoli V, Pinto S, Bargellini I, Vignali C, Cioni R, Petruzzi P, et al. Duplex ultrasonographic study of the renal arteries before and after renal artery stenting. Eur Radiol. 2002;12:796–803. [PubMed]
20. Parenti GC, Palmarini D, Bilzoni M, Campioni P, Mannella P, Ginevra A. Role of color-Doppler sonography in the follow-up of renal artery stenting. Radiol Med (Torino) 2008;113:242–8. [PubMed]
21. Motew SJ, Cherr GS, Craven TE, Travis JA, Wong JM, Reavis SW, et al. Renal duplex sonography: main renal artery versus hilar analysis. J Vasc Surg. 2000;32:462–9. [PubMed]
22. Kastrati A, Schomig A, Elezi S, Schuhlen H, Dirschinger J, Hadamitzky M, et al. Predictive factors of restenosis after coronary stent placement. J Am Coll Cardiol. 1997;30:1428–36. [PubMed]
23. Davignon J. Beneficial cardiovascular pleiotropic effects of statins. Circulation. 2004;109(23 suppl 1):III39–43. [PubMed]
24. Calabro P, Yeh ET. The pleiotropic effects of statins. Curr Opin Cardiol. 2005;20:541–6. [PubMed]
25. Cheung CM, Patel A, Shaheen N, Cain S, Eddington H, Hegarty J, et al. The effects of statins on the progression of atherosclerotic renovascular disease. Nephron Clin Pract. 2007;107:c35–42. [PubMed]
26. Erl W, Hristov M, Neureuter M, Yan ZQ, Hansson GK, Weber PC. HMG-CoA reductase inhibitors induce apoptosis in neointima-derived vascular smooth muscle cells. Atherosclerosis. 2003;169:251–8. [PubMed]
27. Erl W. Atorvastatin-induced downregulation of survivin and vascular smooth muscle cell apoptosis: a causal relationship in restenosis? Cardiovasc Drugs Ther. 2007;21:141–4. [PubMed]
28. Crisby M, Nordin-Fredriksson G, Shah PK, Yano J, Zhu J, Nilsson J. Pravastatin treatment increases collagen content and decreases lipid content, inflammation, metalloproteinases, and cell death in human carotid plaques: implications for plaque stabilization. Circulation. 2001;103:926–33. [PubMed]
29. Heeneman S, Sluimer JC, Daemen MJ. Angiotensin-converting enzyme and vascular remodeling. Circ Res. 2007;101:441–54. [PubMed]