|Home | About | Journals | Submit | Contact Us | Français|
Renal insufficiency (RI) has been shown to be associated with increased major adverse cardiovascular events after percutaneous coronary intervention. We reviewed the impact of RI on the pathogenesis of coronary artery disease and outcomes after percutaneous coronary intervention in the form of drug-eluting stent (DES) implantation in these high-risk patients. We searched the English-language literature indexed in MEDLINE, Scopus, and EBSCO Host research databases from 1990 through January 2009, using as search terms coronary revascularization, drug-eluting stent, and renal insufficiency. Studies that assessed DES implantation in patients with various degrees of RI were selected for review. Most of the available data were extracted from observational studies, and data from randomized trials formed the basis of a post hoc analysis. The outcomes after coronary revascularization were less favorable in patients with RI than in those with normal renal function. In patients with RI, DES implantation yielded better outcomes than did use of bare-metal stents. Randomized trials are needed to define optimal treatment of these high-risk patients with coronary artery disease.
ADMA = asymmetric dimethylarginine; BMS = bare-metal stent; CAD = coronary artery disease; CIN = contrast agent—induced nephropathy; DES = drug-eluting stent; DM = diabetes mellitus; MACE = major adverse cardiac event; MI = myocardial infarction; PCI = percutaneous coronary intervention; PES = paclitaxel-eluting stent; RI = renal insufficiency; SES = sirolimus-eluting stent; ST = stent thrombosis; TLR = target lesion revascularization; TVR = target vessel revascularization
Renal insufficiency (RI), defined as creatinine clearance lower than or equal to 60 mL/min, is prevalent in patients with coronary artery disease (CAD).1-5 In patients with acute coronary syndrome undergoing percutaneous coronary intervention (PCI), even mild RI is an independent predictor of worse outcomes.1 Renal insufficiency is a potent risk factor for CAD and qualifies as a coronary risk equivalent. Chronic RI not only increases cardiovascular event risk but also is associated with worse outcomes if an event occurs.6 Diagnostic coronary angiography and PCI are used less frequently in patients with reduced renal function; risk of complications from these procedures is increased in the presence of RI.7,8 This increased risk has been shown to occur at even moderate levels of RI, despite the use of advanced stent technology and optimal adjunctive medical therapy.9
Best et al10 reported an RI prevalence of 49.1% in patients undergoing elective PCI. In that study, RI had a negative impact similar to that of diabetes mellitus (DM) on cardiovascular outcomes.10 One year after PCI, chronic RI was associated with a doubling of mortality in patients with mild RI, a 5-fold increase in patients with moderate RI, and a 12-fold increase in patients with severe RI.10 In primary PCI for myocardial infarction (MI), a notable incremental reduction in survival was evident for each decline of 10 mL/min in creatinine clearance.11 Unfortunately, most randomized trials of patients undergoing PCI have excluded patients with severe RI.12,13 This narrative review aims to identify the impact of a drug-eluting stent (DES) in the performance of PCI in patients with RI.
We evaluated the impact of RI on the pathogenesis of CAD and outcomes after PCI (in the form of DES implantation) in these high-risk patients. Pertinent articles indexed in MEDLINE, Scopus, and EBSCO Host research databases were searched from 1990 through January 2009, using the search terms coronary revascularization, drug-eluting stent, and renal insufficiency. Studies that assessed DES implantation in patients with RI with and without dialysis were selected for review. Studies that used a bare-metal stent (BMS) vs a DES in patients undergoing dialysis were also reviewed. Studies that had been conducted only on BMSs or that did not specify the type of stent used were excluded.
The association between major adverse cardiac events (MACEs) and the type of stent used was evaluated. Of the 381 articles reviewed, 117 (31%) described the mode of revascularization in patients with RI (62 PCI and 55 coronary artery bypass grafting). Among studies that reported the use of PCI in patients with RI, 54 studies described the impact of stent use, and 8 studies described the use of balloon angioplasty alone. The type of stent was specified in 36 articles (16 DES and 20 BMS). Not all studies used the same renal parameters to define RI, particularly in the predialysis stages. This review defines RI as a creatinine clearance of 60 mL/min or lower. Creatinine clearance was estimated using the Cockcroft-Gault formula.1-5 Patients who had end-stage renal disease and were receiving hemodialysis or peritoneal dialysis were defined as dialysis patients.
This review incorporated pooled data from 5 studies of dialysis patients who underwent stent implantation to assess long-term MACEs. P<.05 was considered statistically significant.
The increased risk associated with PCI in patients with RI is multifactorial and is related to baseline patient and lesion characteristics.
Patients with RI often were older, had DM and hypertension, more commonly had a history of MI or coronary artery bypass graft surgery, and more often presented with acute coronary syndromes, all of which contribute to worse outcomes after revascularization.9,14-16 In particular, patients with DM had substantially worse outcomes after PCI.17 In patients with acute coronary syndromes who underwent primary PCI (balloon angioplasty or BMS implantation), the 6-year rates of MACEs were significantly higher in diabetic patients with RI (47.6%), nondiabetic patients with RI (36.4%), and diabetic patients without RI (36.0%) than in nondiabetic patients without RI (28.4%).18 Whether this higher rate in diabetic patients is due to incomplete revascularization, which is more common in these patients, is unclear.17-19
Patients with RI were more likely to have anatomic findings that put them at high risk, including left main coronary artery stenosis, left anterior descending artery involvement, multivessel disease, saphenous vein graft disease, and more type C than type A and B lesions.14-16
Furthermore, the atherosclerotic plaque morphology in patients with RI may compromise the benefits of PCI seen in patients with normal renal function.16 The coronary plaques in patients with end-stage renal disease are characterized by increased media thickness and marked calcification. Patients with RI have more unfavorable lipid profiles, elevated serum homocysteine and cytokine levels, aggressive atherosclerosis, endothelial dysfunction, and increased prothrombotic activity.2,20 Furthermore, plasma asymmetric dimethylarginine (ADMA) is increased in patients with chronic RI, even at early stages. ADMA is a naturally occurring amino acid that competitively inhibits the activity of nitric oxide synthase; elevated plasma ADMA levels are associated with endothelial dysfunction.2,21 Preprocedural plasma ADMA levels may independently predict subsequent adverse cardiovascular events in patients undergoing PCI.22
There is no consensus in the current literature about whether rates of procedural success are worse in patients with chronic RI compared with patients with preserved renal function. At least 2 large studies showed no association between renal function and procedural success.1,23 However, other studies demonstrated that the rate of success for BMS placement was lower in patients with RI compared with those who had normal renal function (87.2% vs 92.0%; P=.01)11,19; this lower success rate was even more prominent in dialysis patients after DES use.24 The low success rate in patients with RI was attributable to problems with stent delivery, perhaps because of increased calcification that caused greater residual stenosis after PCI.25 Difficulty in stent deployment and less than optimal angiographic results have been found to be independent predictors of in-hospital mortality in patients with RI.9 In a recent study of 10,821 patients with various degrees of RI, Osten et al9 demonstrated that procedural success rates when rotablation was required compared with no rotablation were 78.5% vs 93.4% (P=.02) in patients with at least moderate RI and 94.1% vs 95.5% (P=.78) in those with relatively normal renal function. In that study, at least moderate RI was an independent predictor of each of the procedural variables: residual stenosis greater than 20% (P=.03), 1 or more undeliverable stents (P=.003), and smaller stent diameter (P<.001).
Patients with at least moderate RI had a greater than 4-fold increase in access site complications, most likely because of accelerated atherosclerotic process present throughout the peripheral arterial bed, with subsequent development of peripheral arterial calcium.9,26 Furthermore, the higher rate of bleeding complications observed in patients with at least moderate RI may be related to the predisposition of these patients to hemostatic abnormalities with platelet dysfunction.9,27,28
Platelet dysfunction in patients with RI is due to intrinsic platelet abnormalities and impaired platelet—vessel wall interaction. The normal platelet response to vessel wall injury with platelet activation, recruitment, adhesion, and aggregation is defective in advanced RI. Dialysis is also associated with thrombosis as a result of chronic platelet activation due to contact with artificial surfaces during dialysis.28 Reduced platelet aggregation, intrinsic platelet dysfunction, and abnormalities in platelet-endothelial interactions are found in RI and may in part account for the increased bleeding risk with PCI.27,29 However, a prothrombotic state is manifested in patients with RI by increases in fibrinogen, von Willebrand factor, and tissue factor and reductions in antithrombin III.29,30 Thus, it is unclear whether dual antiplatelet therapy with aspirin and clopidogrel is beneficial and safe in patients with RI.31 Krasopoulos et al32 demonstrated a significant relationship between resistance to aspirin and RI (P<.03). Platelet arachidonic acid metabolism has been well documented to be abnormal in patients with RI.33 This leads to altered thromboxane synthesis, a key factor for the development of resistance to aspirin and one likely related to the increased activity of phospholipase A2 in the platelets of patients with RI.32,33 Interestingly, antiplatelet therapy such as clopidogrel and tirofiban provided no benefit in aspirin-resistant patients,32 a finding that may have implications in the poststenting era in patients with RI. Further studies are warranted to confirm this hypothesis.
The use of adjunctive therapy, including glycoprotein IIb/IIIa inhibitors and direct thrombin inhibitors, with PCI may improve patient outcomes. However, patients with severe RI have been excluded from most studies that evaluated adjunctive therapy for PCI using antiplatelet and antithrombotic agents because of the potential risks of increased hemorrhage.8 From the limited data available, despite dosage adjustment of glycoprotein IIb/IIIa inhibitors, the risk of bleeding increases with decreasing renal function.34 In a small group of patients, Frilling et al35 found a nonsignificant trend of increased mortality in the RI group when abciximab was used during PCI. Post hoc analysis of the results of the ESPRIT (Enhanced Suppression of the Platelet IIb/IIIa Receptor with Integrilin Therapy) trial36 demonstrated that the beneficial effect of glycoprotein IIb/IIIa inhibitors on short-term outcomes after PCI was greater in patients with mild RI than in those with severe RI. Subanalysis of the REPLACE-2 (Randomized Evaluation in PCI Linking Angiomax to Reduced Clinical Events) trial demonstrated beneficial effects of bivalirudin comparable to those of heparin and glycoprotein IIb/IIIa inhibition, regardless of renal function.37 Prospective studies are needed to guide clinicians in the proper use of adjunctive therapy in these high-risk patients.
In patients with chronic RI, restenosis is a major clinical limitation of PCI. The restenosis rate is higher in dialysis than in nondialysis patients.38 The incidence of clinical restenosis is reduced from 60% to 81% after balloon angioplasty to 30% after BMS use in patients with chronic RI.25 The rate of restenosis in these high-risk patients is further reduced with DES use, but it remains higher than in patients with normal renal function and in dialysis patients. Some investigators hypothesized that this high restenosis rate contributes to the higher mortality rate seen in these patients.10,39 However, this concept was refuted later by Lemos et al,3 who reported that mortality risk did not decrease with the reduction of restenosis in their patients.
The higher rate of restenosis after DES placement has been related to the higher incidence of DM and diffuse and calcified lesions among dialysis patients.40 Coronary calcification may cause DES underexpansion, which is a risk for restenosis after PCI, even with DES.41 Damage to the polymer of the DES might occur while it is being delivered through a calcified coronary artery in dialysis patients.42,43 The inhibitory effects of sirolimus on neointimal growth in dialysis patients may be insufficient to prevent restenosis. Furthermore, accelerated atherosclerosis and restenosis are related to 2 important factors—granulocyte activation and oxidative stress.44 These inflammatory processes are associated with chronic RI, with or without dialysis. Oxidative stress is enhanced with chronic RI, even in patients studied before beginning dialysis.45 However, in patients undergoing dialysis, granulocytes are activated during the contact of whole blood with artificial surfaces such as the dialyzer membrane, resulting in the release of inflammatory mediators.46 Such inflammatory processes may initiate the process of neointimal growth after stent placement.45,46 In dialysis and nondialysis patients, the most important predictors of restenosis after sirolimus-eluting stent (SES) implantation were treatment with dialysis, coronary calcification, and implantation of multiple stents.38
In the TAXUS-IV trial, restenosis was almost always focal, even in patients with moderate RI.4 Reintervention in focal restenosis is associated with better outcomes than the outcomes in nonfocal restenotic lesions. From the limited data available, it appears that focal restenosis is more likely to occur than nonfocal restenosis with DES use in patients with RI.
Recent reports have raised the possibility that death, MI, and stent thrombosis (ST), especially late thrombosis, are more common with DES than BMS.47 Although ST may be theoretically reduced in these patients because of impaired platelet function, ST may be increased because of more frequent adverse baseline clinical characteristics, including high incidence of DM, lower left ventricular ejection fraction, and underuse of antiplatelet therapy.8,48-50
Table 1 demonstrates that RI was an important independent predictor of ST after DES use in 4 large studies.49-52 Moreover, Nakazawa et al45 reported acute ST complicated with MI in 2 patients with RI. Halkin et al4,53 demonstrated that ST rates were higher after DES than BMS use, regardless of patients' renal function. Subacute thrombosis was reported in 2 studies.24,38 In a recent prospective study,2 the rates of all types of ST after DES use were significantly higher in patients with creatinine clearance less than 60 mL/min than in those whose creatinine clearance was 60 mL/min or higher (subacute ST, 1.52% vs 0.14%; late ST, 2.60% vs 0.84%; very late ST, 1.96% vs 0.70%). A smaller number of studies reported no increased risk of ST in patients with RI.24,38 Resistance to or discontinuation of dual antiplatelet therapy before the recommended times for DES is associated with a much higher risk of ST compared with later discontinuation.54 A probable explanation for an increased risk of thrombosis with DES compared with BMS after discontinuation of antiplatelet therapy is incomplete neointimal coverage secondary to delayed arterial healing.55 In advanced RI, delayed arterial healing may be related to the atheromatous changes of the arterial intima, remodeling of the arterial media, and subsequent increased arterial stiffness.16
Patients with RI who are not undergoing dialysis are at serious risk of developing contrast agent—induced nephropathy (CIN) after diagnostic and therapeutic coronary interventions. After PCI (regardless of stent type and baseline renal function), the development of CIN is associated with worse in-hospital and 1-year outcomes.11,17,34,56
Major cardiovascular trials frequently have excluded patients with RI and have not provided adequate information on the renal function and the effect of PCI on these patients.13 In the current review, we tallied the outcomes of 12,189 patients with various degrees of chronic RI who underwent DES implantation in 16 studies.2-5,24,38,42,45, 53,57-63 Ten studies compared DES and BMS use3-5,38,42,53,57,60,61,63; the DES subtype was specified in 12 studies (SES3,38,42,45,57-60,62,63 and paclitaxel-eluting stent [PES]4,53). One study reported on all 3 types of stents in chronic RI.57
Six studies reported DES use in dialysis patients,38,42,53,59-61 and 5 studies compared DES with BMS use in dialysis patients38,42,53,60,61 (Table 2). In pooled data from 5 studies of 626 dialysis patients who underwent stent implantation, long-term MACEs were reported in 22% of patients implanted with a DES and in 38% of patients implanted with a BMS (P<.001).38,42,53,60,61 Das et al61 studied 89 dialysis patients who had been implanted with a DES or a BMS. Patients in the DES group received longer stents than patients in the BMS group (20.4 mm vs 17.3 mm; P=.006). The average stent diameter was smaller in the DES group (2.9 mm vs 3.1 mm; P=.008). The authors reported that DES was more effective than BMS in reducing 9-month target vessel revascularization (TVR). However, the mortality rate remained strikingly high, irrespective of the type of stent used. Similarly, large reductions in TVR have been noted in other studies comparing DES and BMS in dialysis patients.63-65
Ishio et al42 compared 54 dialysis patients (total number of lesions, 69) who were implanted with an SES with 54 dialysis patients (total number of lesions, 58) who underwent PCI using BMS implants. The investigators reported that the rate of in-stent restenosis was lower after SES than after BMS implantation (22% vs 40%; P=.04) in patients undergoing dialysis; however, rates of in-segment restenosis (31% vs 43%; P=.30) and target lesion revascularization (TLR) (14% vs 21%; P=.40) were not statistically different between lesions treated with SES and those treated with BMS. Target lesion revascularization after BMS was similar to that reported by Das et al61; however, TLR after SES use was higher than that reported by Hassani et al24 and Das et al.61 In the series reported by Ishio et al,42 the mortality rate was lower than in these 2 studies (2% vs 16%24 and 25%61). Although the superiority of DES has yet to be established in patients with RI,3-5,57 3 studies38,53,61 have shown the superiority of SES compared with BMS, and 2 studies42,60 have shown the superiority of BMS implants. Table 2 shows that DES use is associated with a nonsignificant reduction of restenosis and mortality in dialysis patients.
We reviewed 10 published studies comparing DES use in dialysis and nondialysis patients. Six studies24,45,57,58,62,63 compared DES use in nondialysis vs dialysis patients with RI; 5 of these studies used both DES and BMS (Table 3). Four studies were conducted among nondialysis patients with RI.2-5
DES implantation (an SES or PES) for de novo coronary artery lesions in patients with severe RI reduced 1-year clinical cardiac events compared with BMS, mainly by reducing TVR, without an effect on mortality.5,57,63 Mishkel et al58 demonstrated that the 18-month survival free from TVR after implantation of SES vs PES was similar, regardless of the severity of RI.
Sasao et al62 studied 170 patients who underwent SES implantation. No significant difference in MACEs was observed between the dialysis and nondialysis patients. The occurrence of TLR was higher in the dialysis than in the nondialysis patients. The clinical outcomes of the nondialysis patients were comparable to those of patients in previous studies: the incidence of any clinical cardiac event was 12.5%, and the incidence of TLR was 4.6% during the follow-up period. MACE and TLR rates after SES implantation in dialysis patients were higher than those of nondialysis patients.45,60,62
Lemos et al3 demonstrated higher 1-year mortality in patients with RI after PCI regardless of stent type compared with mortality in patients with normal renal function. However, compared with dialysis patients with RI, nondialysis patients with RI treated by DES had better clinical outcomes.24,45,62 In a recent study from China, the hazard ratio for a composite end point of mortality and MI was significantly less in patients with normal or mild RI vs moderate or severe RI after DES use.2 In a recent study from Korea of 1003 patients with an estimated glomerular filtration rate less than 60 mL/min/1.73 m2, DES use was associated with significantly lower all-cause mortality (5.4% vs 13.3%; P<.02) and incidence of MI (1.7% vs 10.0%; P<.003) than BMS use; however, no significant differences were noted in the rates of restenosis (20.6% vs 26.5%; P=.36). 66 Approximately 10% of these patients were undergoing dialysis. In the current review, most studies demonstrated that DES was superior to BMS in nondialysis patients with RI, with a lower rate of restenosis but without an effect on mortality. The outcomes in dialysis patients appear to be worse.
A common theme of all the trials in the current review is that the marked limitations in sample size often preclude the statistical significance of event rate comparisons, making any conclusion problematic. Although use of DES has improved outcomes in dialysis patients, outcomes after PCI are less favorable in patients with RI (especially those undergoing dialysis) than in those with normal renal function. These unfavorable outcomes can be explained by many factors, including patient baseline characteristics, lesion and vessel complexities, early and aggressive atherosclerotic processes, lack of guidelines for the proper use of adjunctive therapy, the frequent occurrence of CIN, and the high incidence of restenosis and ST. Prospective studies using a gradation scheme of RI to compare the outcomes of patients undergoing PCI with DES are urgently needed. The decision to place a DES vs a BMS in patients with RI requires a careful assessment of benefits and risks in randomized controlled studies.