Drug-Eluting Stents Versus Bare-Metal Stents in Patients With Decreased GFR: A Meta-analysis

Original Investigation Drug-Eluting Stents Versus Bare-Metal Stents in Patients With Decreased GFR: A Meta-analysis Zhi Jian Wang, MD, PhD,1 Kishore J. Harjai, MD,2 Chetan Shenoy, MBBS,2 Fei Gao, MD,1 Dong Mei Shi, MD,1 Yu Yang Liu, MD,1 Ying Xin Zhao, MD,1 and Yu Jie Zhou, MD, PhD1 Background: Decreased estimated glomerular filtration rate (eGFR) is a strong predictor of both mortality and subsequent cardiac events after percutaneous coronary intervention. The safety and efficacy of drugeluting (DESs) versus bare-metal stents (BMSs) in this population have not been evaluated adequately. Study Design: A systematic review and meta-analysis. Setting & Population: Studies involving the comparison of clinical outcomes between DESs and BMSs in patients with eGFR ⬍60 mL/min/1.73 m2. Studies exclusively involving patients with ST-segment elevation myocardial infarction were excluded. Selection Criteria for Studies: MEDLINE (on Ovid), EMBASE, and the Cochrane Library databases from 2002-2013 were searched for studies comparing DESs with BMSs in patients with eGFR ⬍60 mL/min/1.73 m2. Intervention: DES versus BMS implantation. Outcomes: Mortality, repeat revascularization, myocardial infarction, and stent thrombosis. Results: Data from 26 comparative studies with 66,840 patients were included. Compared with BMSs, DESs were associated with significant reductions in repeat revascularization (OR, 0.61; 95% CI, 0.50-0.74; P ⬍ 0.001) and myocardial infarction (OR, 0.85; 95% CI, 0.79-0.92; P ⬍ 0.001), with no detectable difference in stent thrombosis (OR, 0.72; 95% CI, 0.46-1.12; P ⫽ 0.1). The superiority of DESs over BMSs in decreasing mortality also was documented (OR, 0.77; 95% CI, 0.65-0.90; P ⫽ 0.01). This survival benefit of DESs over BMSs was attenuated in randomized controlled trials or adjusted observational studies versus unadjusted observational studies. Limitations: Most studies were observational studies. Meta-analysis was not performed on individual patient data. Conclusions: DES use in patients with eGFR ⬍60 mL/min/1.73 m2 is associated with a reduced rate of repeat revascularization and myocardial infarction without increased risk of stent thrombosis. The true effect of DESs versus BMSs on mortality needs to be confirmed by randomized controlled trials. Am J Kidney Dis. 62(4):711-721. © 2013 by the National Kidney Foundation, Inc. INDEX WORDS: Kidney; coronary disease; meta-analysis; stents.

P

atients with decreased glomerular filtration rate (GFR) of varying severity constitute an increasing percentage of the population undergoing percutaneous coronary intervention. Multiple studies have firmly established the high-risk nature of this population. Even patients with estimated GFR (eGFR) ⬍60 mL/min/1.73 m2 are at increased risk of mortality and cardiovascular events after percutaneous coronary intervention compared with individuals with preserved kidney function.1-9 The introduction of drug-eluting stents (DESs) has been shown to trigger less restenosis than bare-metal stents (BMSs), especially in specific patient and stenosis subsets.10 However, whether the benefit of DESs in the general population extends to patients with decreased GFR is uncertain due to limited evidence. Although several post hoc analyses and registries have compared DES and BMS efficacy in this highrisk population, these studies have had conflicting results and small numbers of patients.3-9,11-29 Prospective well-designed randomized controlled trials (RCTs), which have been considered to have more validity, have consistently excluded patients with seAm J Kidney Dis. 2013;62(4):711-721

verely decreased GFR. Recent concerns about higher risk of stent thrombosis in patients with both decreased GFR and proteinuria after implantation of a DES raised further questions about DES safety in this population.30 To shed further light on this issue, we performed a meta-analysis of the current literature to assess the clinical outcomes of DES versus BMS use in patients with decreased GFR.

From the 1Department of Cardiology, Anzhen Hospital, Capital Medical University, Beijing, China; and 2Guthrie Clinic, Sayre, PA. Received December 11, 2012. Accepted in revised form April 23, 2013. Originally published online June 17, 2013. Address correspondence to Yu Jie Zhou, MD, PhD, Department of Cardiology, Anzhen Hospital, Capital Medical University, Anzhen Avenue #2, Chaoyang district, Beijing, 100029 China. E-mail: [email protected] © 2013 by the National Kidney Foundation, Inc. 0272-6386/$36.00 http://dx.doi.org/10.1053/j.ajkd.2013.04.014 711

Wang et al

METHODS Inclusion/Exclusion Criteria We did a systematic review of the published work according to the guidelines of the Meta-analysis of Observational Studies in Epidemiology (MOOSE) group.31 Studies were considered eligible for this review if they involved the comparison of clinical outcomes between DESs and BMSs in patients with eGFR ⬍60 mL/min/1.73 m2. Exclusion criteria were defined as studies exclusively involving patients with ST-segment elevation myocardial infarction (MI) and studies not reporting relevant clinical outcomes.

Study Search Strategies A search was carefully conducted for published reports in MEDLINE (on Ovid), EMBASE, and the Cochrane Library databases (Cochrane Central Register of Controlled Trials) from January 2002 to February 2013 and then in conference proceedings of the Scientific Sessions of the American College of Cardiology, American Heart Association, European Society of Cardiology, Transcatheter Cardiovascular Therapeutics, and EuroPCR, as well as their websites. The search terms “stents,” “sirolimus,” “paclitaxel,” and their synonyms were combined with the terms “kidney,” “hemodialysis,” and “dialysis” (search strategy details are shown in Item S1, available as online supplementary material). Reference lists from identified reports and review articles were screened manually to identify any other relevant studies. No language restrictions were applied. The decision to include or exclude studies was hierarchical, initially based on the study title, followed by the abstract and finally the complete body text. The literature search, data extraction, and quality assessment were undertaken independently by 2 reviewers (Z.J.W. and F.G.) with a standardized approach. Disagreement was resolved by discussion and consulting a third reviewer (Y.J.Z.).

Data Extraction and Quality Assessment Full reports of these potentially eligible studies were retrieved by 2 reviewers for detailed assessment, and standard information was extracted into a spreadsheet. The data sought included first author, journal, publication year, study design and location, base-

line clinical characteristics, number of patients treated by BMS or DES implantation, stent type, duration of follow-up, and clinical outcomes of interest. If the required data could not be located in the published reports, the authors of the individual reports were asked to provide the missing data of interest. Quality assessment for meta-analyses of observational studies is controversial and lacks general tools. We systematically evaluated several key points of study quality according to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement32 and a systematic review by Sanderson et al.33 These key points included clear declaration of source population and inclusion/exclusion criteria, clear definition of clinical outcomes, appropriate measurement methods for clinical outcomes, appropriate methods to deal with design-specific sources of bias, appropriate methods to adjust for confounding, appropriate statistics for primary analysis of effect, and clear declaration of conflict of interest or funding sources. Two reviewers independently evaluated each report included in our analysis. Any disagreement in abstracted data was adjudicated by a third reviewer (Y.J.Z.). We collected data for all-cause mortality, repeat revascularization, MI, and stent thrombosis. When multiple categories of repeat revascularization were reported by a single primary study, targetvessel revascularization was preferred.

Statistical Analysis Individual study odds ratios (ORs) and 95% confidence intervals (CIs) were calculated for each article. When possible, we aimed to pool adjusted ORs from the primary studies; otherwise, we used raw outcome data to yield unadjusted ORs. All data were pooled with the generic inverse variance method (log of the risk estimate and its standard error) and random-effects model as a conservative approach. Statistical heterogeneity was assessed using Cochrane’s Q by ␹2 test and further quantified with I2 test. We considered P ⬍ 0.1 and I2 ⬎ 50% to be evidence of substantial heterogeneity. To explore sources of heterogeneity and robustness of results, sensitivity analyses, subgroup analyses, and metaregression were performed. Analyses were performed using Stata, version 11.0 (Stata Corp LP). Statistical level of significance for the summary treatment effect estimate was 2-tailed P ⬍ 0.05.

Articles identified from electronic database searches (n = 784) Additionally identified articles from citations/references (n = 1) Total identified articles (n = 785) Excluded (n = 673) ·Irrelevant (n = 588) ·Review articles, letters, comments, editorial, case reports (n = 85) Full-text articles assessed for eligibility (n = 112) Excluded (n =86) ·Only laboratory and IVUS data were available (n = 2) ·Comparing CKD with normal kidney function (n = 35) ·Stents in renal artery stenosis (n = 9) ·Stents in renal transplant (n = 2) ·Comparing DES with bypass surgery (n = 4) ·Comparing different DES in CKD (n = 27) ·Exclusively enrolled patients with STEMI (n = 7) Articles included in meta-analysis (n =26) ·Pooled analysis of RCTs (n = 1) ·Post-hoc analysis of RCT (n = 1) ·Observational studies (n =24)

712

Figure 1. Identification of relevant studies for inclusion in the meta-analysis. Abbreviations: CKD, chronic kidney disease; DES, drugeluting stent; IVUS, intravenous ultrasound; RCT, randomized controlled trial; STEMI, STsegment elevation myocardial infarction. Am J Kidney Dis. 2013;62(4):711-721

DES in Patients With Decreased GFR Table 1. Characteristics of the Studies Meta-analyzed Study

Inclusion Criteria of Patients

Interventions

Design; Country of Origin

Median Follow-up

No. of Patients

Halkin et al8 (2005)

eGFR ⬍60 mL/min (CG)

PES vs BMS

Post hoc analysis of RCT (TAXUS IV); 123 PES, 100 United States BMS

1y

Lemos et al9 (2005)

eGFR ⬍60 mL/min (CG)

SES vs BMS

Retrospective single-center observational study; the Netherlands

94 SES, 92 BMS

1y

Halkin et al23 (2006)

On dialysis

SES vs BMS

Prospective multicenter observational study, multivariate adjusted; United States, Canada

33 DES, 41 BMS

1y

Zhang et al21 (2006)

eGFR ⬍60 mL/min (CG)

DES vs BMS

Retrospective single-center observational study; China

264 DES, 146 BMS

17 moa

Das et al22 (2006)

On dialysis

DES vs BMS

Retrospective multicenter observational study; United States

24 DES, 65 BMS

9 mo

Kuchulakanti et al20 (2006)

Presence of previously documented MI and/or baseline SCr ⬎2 mg/dL

SES vs BMS

Retrospective single-center observational study; United States

68 SES, 120 BMS

6 mo

Ishio et al19 (2007)

On hemodialysis

DES vs BMS

Retrospective multicenter observational study; Japan

54 SES, 54 BMS

9 mo

Jeong et al16 (2008)

eGFR ⬍60 mL/min (CG)

SES or PES vs Retrospective single-center BMS observational study; Korea

76 SES, 28 PES, 1 y 50 BMS

Rosenblum et al13 eGFR ⬍60 mL/min (CG) (2009)

DES vs BMS

Retrospective single-center observational study; United States

1,291 DES, 682 BMS

1y

Aoyama et al18 (2008)

On hemodialysis

SES vs BMS

Retrospective single-center observational study; Japan

88 SES, 78 BMS

1y

Okada et al17 (2008)

On dialysis

SES vs BMS

Retrospective single-center observational study; Japan

80 SES, 124 BMS

1y

Yachi et al14 (2009)

On hemodialysis

SES vs BMS

Retrospective single-center observational study; Japan

56 SES, 67 BMS

9 mo

Kim et al15 (2009)

On dialysis

SES vs BMS

Prospective multicenter observational study; Korea

54 SES, 51 BMS

917 da

Appleby et al6 (2009)

eGFR ⬍60 mL/min (CG)

DES vs BMS

Retrospective single-center observational study, multivariate adjusted; Canada

749 DES, 2,321 BMS

4y

Shenoy et al12 (2010)

eGFR ⬍60 mL/min (CG)

SES vs BMS

Prospective single-center observational study, multivariate adjusted; United States

222 DES, 214 BMS

3.39 y

Garg et al7 (2010)

eGFR ⬍60 mL/min (CG); SCr ⱕ3.0 mg/dL

SES vs BMS

Pooled analysis of RCTs (SIRIUS, C-SIRIUS, E-SIRIUS); United States, Germany, Canada

109 SES, 119 BMS

5y

Ichimoto et al24 (2010)

On hemodialysis

SES vs BMS

Retrospective single-center observational study; Japan

63 SES, 45 BMS

SES: 22 mo; BMS: 32 mo

Oshima et al26 (2011)

On dialysis

DES vs BMS

Retrospective multicenter observational study; Japan

217 SES, 204 BMS

1y

Green et al3 (2011)

10 ⱕ eGFR ⬍60 mL/min/1.73 m2 DES vs BMS (MDRD)

Retrospective multicenter observational study, multivariate adjusted; United States

763 DES, 345 BMS

1y

Simsek et al4 (2012)

CCr ⬍60 mL/min (CG)

SES or PES vs Prospective single-center BMS observational study, multivariate adjusted; the Netherlands

85 SES, 90 PES, 6 y 72 BMS

Charytan et al11 (2011)

Age ⱖ18 y, SCr ⬎2.0 mg/dL or dialysis dependence

DES vs BMS

Prospective multicenter observational study, propensity score matched; United States

431 DES, 431 BMS

2y

Tsai et al5 (2011)

CCr ⬍60 mL/min (CG)

DES vs BMS

Retrospective multicenter observational study, propensity score matched; United States

27,567 DES, 27,567 BMS

30 mo

Barthelemy et al25 CCr ⬍60 mL/min (CG) (2011)

DES vs BMS

Prospective single-center observational study, France

126 DES, 224 BMS

1y

(Continued)

Am J Kidney Dis. 2013;62(4):711-721

713

Wang et al Table 1 (Cont’d). Characteristics of the Studies Meta-analyzed Study

Inclusion Criteria of Patients

Ishii et al27 (2012)

On hemodialysis

Kersting et al28 (2012) Resmini et al29 (2012)

Interventions DES vs BMS

Design; Country of Origin

Median Follow-up

301 DES, 204 BMS

DES: 30 mo; BMS: 42 mo

10 ⱕ eGFR ⬍60 mL/min/1.73 m2 SES or PES vs Retrospective single-center (MDRD) BMS observational study, multivariate adjusted; Germany

117 DES, 63 BMS

2.8 y

CCr ⬍60 mL/min (CG)

55 DES, 164 BMS

48 mo

DES vs BMS

Retrospective single-center observational study; Japan

No. of Patients

Retrospective single-center observational study; Italy

Note: Conversion factor for SCr in mg/dL to ␮mol/L, ⫻88.4. Abbreviations and definitions: BMS, bare-metal stent; CCr, creatinine clearance; CG, Cockcroft-Gault formula; C-SIRIUS, Canadian SIRIUS; DES, drug-eluting stent, sirolimus-eluting stent (SES) and paclitaxel-eluting stent (PES); eGFR, estimated glomerular filtration rate; E-SIRIUS, European SIRIUS; MDRD, Modification of Diet in Renal Disease Study equation; MI, myocardial infarction; RCT, randomized controlled trial; SCr, serum creatinine; SIRIUS, Study of Sirolimus-Coated BX VELOCITY Balloon-Expandable Stent in Treatment of de Novo Native Coronary Artery Lesions. a Mean.

RESULTS Literature Search From 785 potentially relevant citations identified from electronic databases and search of reference lists, 26 studies met the inclusion criteria for the analysis, including one pooled analysis of RCTs (Sirolimus-Eluting Stent in De Novo Native Coronary Lesions [SIRIUS] trial, CanadianSIRIUS, and European-SIRIUS),7 one post hoc analysis of an RCT (TAXUS-IV trial),8 and 24 observational studies (Fig 1).3-6,9,11-29 Overall, there were 66,840 patients involved in our meta-analysis, of whom 33,225 were treated with DESs and 33,615 were treated with BMSs. Characteristics of all included studies are presented in Table 1. Sample sizes ranged from 74-55,134 participants. Patient characteristics are listed in Table 2. The mean or median follow-up period ranged from 6 months to 6 years. Six observational studies were prospective analyses4,11,12,15,23,25 and the other 18 were retrospective.3,5,6,12-14,16-22,24,26-29 All studies involved direct comparison between DESs and BMSs in patients with decreased GFR. Of them, 14 studies used eGFR ⬍60 mL/min (as calculated by the Cockcroft-Gault formula) or ⬍60 mL/min/1.73 m2 (by the MDRD [Modification of Diet in Renal Disease] Study equation),3-9,12,13,16,21,25,28,29 10 studies analyzed only patients on dialysis therapy,14,15,17-19,22,23,24,26,27 and 2 studies used serum creatinine level ⬎2.0 mg/dL or dialysis dependence,11,20 which meets the criteria of eGFR ⬍60 mL/min/1.73 m2 in our analysis when either formula is used. The percentage of men in these studies ranged from 17%-80%. Mean age of study participants ranged from 58-73 years. One study included only patients 65 years or older.5 All studies had clear declaration of source population, appropriate use of statistics for primary analysis of effect, and clear definition of clinical outcomes (Table S1). A propensity score method for case matching was used in 2 studies.5,11 Multivariate-adjusted 714

estimates for mortality were available in another 5 studies.3,4,6,12,28 Mortality Data for mortality were available from 24 studies3-9,11,12,14-25,27-29 including 64,446 participants (31,717 treated by DESs; 32,729 by BMSs). For the 7 primary studies using propensity or multivariate adjustments, adjusted estimates were drawn directly from 6 studies.3,5,6,11,12,28 In the study by Simsek et al,4 adjusted effect estimates were only available for each single DES type. Therefore, we used the raw outcome data in overall summary estimate. Overall, DES versus BMS use was associated with a 23% reduction in mortality (OR, 0.77; 95% CI, 0.65-0.90; P ⫽ 0.001; Fig 2). We noticed some signs of heterogeneity in the magnitude of the effect across included studies (I2 ⫽ 49.7%; P ⫽ 0.05). In subsequent subgroup analysis (Fig 3), we found greater effect size in unadjusted and single-center observational studies, which was attenuated somewhat in RCTs/adjusted studies (P ⫽ 0.02) and multicenter studies (P ⫽ 0.02). On univariate metaregression, DESs also were found to produce more survival benefits in unadjusted studies versus RCTs or adjusted studies (P ⫽ 0.05). Repeat Revascularization All studies reported incidences of repeat revascularization, including a total of 66,840 participants. Three studies reported both target-vessel and target-lesion revascularization rates,8,15,20 10 studies reported data for only target-vessel revascularization,4,7,9,11,12,16,21,22,28,29 and 9 studies reported data for only target-lesion revascularization.13,14,17-19,24-27 In 4 studies, only the incidence of general repeat revascularization was described without further clarification.3,5,6,23 Am J Kidney Dis. 2013;62(4):711-721

Counts Age (y) DES

Male Sex

ACS

Previous MI

Diabetes

Hypertension

BMS

DES

BMS

DES

BMS

DES

BMS

DES

BMS

DES

BMS

All-Cause Mortality

MI

Repeat Revascularization

Stent Thrombosis

Halkin et al8 (2005)

NA

NA

NA

NA

NA

NA

NA

NA

NA

NA

NA

NA

8

12

TVR: 23; TLR: 16

Lemos et al9 (2005)

72 ⫾ 9

72 ⫾ 8

50

48

62

45

31

39

20

20

50

49

6

NA

TVR: 18

Halkin et al23 (2006)

65 ⫾ 11

63 ⫾ 10

NA

NA

NA

NA

NA

NA

NA

NA

NA

NA

21

5

Any: 12

NA

Zhang et al21 (2006)

NA

NA

NA

NA

NA

NA

NA

NA

NA

NA

NA

NA

30

4

TVR: 22

NA

2 NA

Das et al22 (2006)

58 ⫾ 13

64 ⫾ 10

75

75

75

74

58

48

71

83

100

100

28

5

TVR: 18

NA

Kuchulakanti et al20 (2006)

68 ⫾ 11

69 ⫾ 11

56

60

46

44

54

43

59

56

93

92

27

40

TVR: 29, TLR: 23

NA

Ishio et al19 (2007)

64 ⫾ 9

62 ⫾ 10

80

65

22

43

46

35

61

65

87

85

4

1

TLR: 19

Jeong et al16 (2008)

64 ⫾ 10

60 ⫾ 9

65

68

68

86

19

10

62

58

74

78

14

7

TVR: 10

1

NA

NA

53

54

47

47

20

22

43

44

92

91

NA

NA

TLR: 120

NA

Rosenblum et al13 (2009)

1

Aoyama et al18 (2008)

65 ⫾ 10

64 ⫾ 1

64

71

NA

NA

6

5

56

62

56

55

5

5

TLR: 31

0

Okada et al17 (2008)

67 ⫾ 10

67 ⫾ 10

74

59

16

20

38

27

73

61

79

68

37

2

TLR: 55

3

Yachi et al14 (2009)

65 ⫾ 9

66 ⫾ 8

66

72

32

22

13

13

51

49

84

85

12

0

TLR: 28

2

Kim et al15 (2009)

62 ⫾ 10

60 ⫾ 9

67

62

54

65

4

2

76

57

91

73

26

3

TVR: 17; TLR: 15

6

Appleby et al6 (2009)

NA

NA

NA

NA

NA

NA

NA

NA

NA

NA

NA

NA

NA

NA

NA

Shenoy et al12 (2010)

69 ⫾ 11

73 ⫾ 10

17

23

36

51

15

22

26

15

69

67

60

32

TVR: 43

13

NA

NA

NA

NA

NA

NA

NA

NA

NA

NA

NA

NA

40

16

TVR: 53

15

Ichimoto et al24 (2010)

66 ⫾ 8

63 ⫾ 8

76

80

10

2

37

38

63

62

90

93

24

4

TLR: 30

3

Oshima et al26 (2011)

NA

NA

NA

NA

NA

NA

NA

NA

NA

NA

NA

NA

NA

4

TLR: 135

NA

Green et al3 (2011)

NA

NA

53

52

55

64

28

31

48

41

88

87

72

63

Any: 118

NA

Simsek et al4 (2012)

NA

NA

NA

NA

NA

NA

NA

NA

NA

NA

NA

NA

93

NA

TVR: 32

5

Charytan et al11 (2011)

NA

NA

NA

NA

NA

NA

NA

NA

NA

NA

NA

NA

331

151

TVR: 132

NA

Any: 5,658

NA

Garg et al7 (2010)

Tsai et al5 (2011)

NA

NA

NA

NA

NA

67

67

28

28

38

38

85

85

6,392

2,377

Barthelemy et al25 (2011)

72 ⫾ 11

75 ⫾ 11

69

66

41

30

21

19

32

25

78

65

67

NA

TLR: 31

4

Ishii et al27 (2012)

66 ⫾ 10

66 ⫾ 9

70

68

NA

NA

6

7

56

63

66

65

197

NA

TLR: 209

11

Kersting et al28 (2012)

71 ⫾ 10

74 ⫾ 10

73

71

53

59

NA

NA

40

40

83

76

51

28

TVR: 43

24

Resmini et al29 (2012)

68 ⫾ 10

75 ⫾ 10

76

77

71

76

29

30

49

44

87

90

70

22

TVR: 34

5

Note: Age given as mean ⫾ standard deviation; values for categorical variables are given as percentages. Abbreviations: ACS, acute coronary syndrome; BMS, bare-metal stent; DES, drug-eluting stent; MI, myocardial infarction; NA, not available; TLR, target-lesion revascularization; TVR, target-vessel revascularization.

DES in Patients With Decreased GFR

Am J Kidney Dis. 2013;62(4):711-721

Table 2. Participant Characteristics of the Studies Meta-analyzed

715

Wang et al Odds ratio (95% CI)

Study ID

1.34 (0.22, 0.85 (0.69, 0.29 (0.15, 1.09 (0.83, 0.65 (0.22, 1.26 (0.64, 1.05 (0.60, 1.37 (0.32, 0.40 (0.15, 1.00 (0.40, 0.80 (0.56, 0.32 (0.03, 0.16 (0.05, 0.97 (0.48, 1.14 (0.47, 1.04 (0.45, 0.98 (0.19, 0.81 (0.39, 0.26 (0.12, 0.48 (0.25, 1.19 (0.67, 0.86 (0.81, 0.21 (0.04, 0.46 (0.22, 0.77 (0.65,

Aoyama 2008 Appleby 2009 Barthelemy 2012 Charytan 2011 Das 2006 Garg 2010 Green 2011 Halkin 2005 Halkin 2006 Ichimoto 2010 Ishii 2012 Ishio 2007 Jeong 2008 Kersting 2012 Kim 2009 Kuchulakanti 2006 Lemos 2005 Okada 2008 Resmini 2012 shenoy 2010 Simsek 2011 Tsai 2011 Yachi 2009 Zhang 2006 Overall (I-squared = 49.7%, p = 0.003)

Favors DES

.1

.2

.5

1

2

5

10

8.25) 1.05) 0.57) 1.44) 1.88) 2.49) 1.85) 5.88) 1.05) 2.51) 1.15) 3.19) 0.54) 1.95) 2.77) 2.43) 4.98) 1.70) 0.57) 0.92) 2.12) 0.91) 1.01) 0.96) 0.90)

Weight (%) 0.75 11.66 4.10 10.17 1.99 4.05 5.22 1.13 2.35 2.54 8.39 0.48 1.57 3.90 2.69 2.90 0.92 3.58 3.28 4.33 5.08 14.32 0.98 3.61 p = 0.001

Favors BMS

Figure 2. Odds ratio of mortality associated with drug-eluting (DES) and bare-metal stent (BMS) use. Abbreviation: CI, confidence interval.

Overall, for the total of 26 studies, DES use produced a 39% reduction in repeat revascularization compared with BMS use (OR, 0.61; 95% CI, 0.50-0.74; P ⬍ 0.001; Fig 4). There also was a high level of heterogeneity of results (I2 ⫽ 73.4%; P ⬍ Subgroup

0.001), which was attributed mostly to the large study by Tsai et al,5 a study that specifically included patients 65 years and older. Sensitivity analysis excluding this study resulted in a similar estimate of effect of 39% OR reduction (OR, 0.61; 95%

Odds ra
o, 95% CI

Studies included

Renal func
on stra
fica
on 5, 14,15,17-19,22,23,24,27 Dialysis dependent Non-dialysis dependent 3-9,11-16,20,21,25,28,29 Pa
ents number ≥400 3,5,6,11,12,21,27 <400 4,7-9,14-20,22-24,25,28,29 Follow-up 3,8,9,14,16-25 <2y 5,11,12,15,27-29 ≥2 y

0.89 [0.78, 1.03]

p value for heterogeneity

0·2

0.75 [0.61, 0.93] 0.86 [0.75, 0.97] 0.67 [0.48, 0.94]

0·2

0.61 [0.43, 0.86] 0.86 [0.74, 1.01]

0·07

0.86 [0.81, 0.91] 0.73 [0.60, 0.90]

0·1

4,6-12,14-25,27-29

RCT Observaonal studies

7,8 3-6,9,11,12,14-25,27-29

1.28 [0.69, 2.37] 0.76 [0.64, 0.89]

0·1

Adjusted or RCTs Unadjusted

3,5-8,11,12,28 4,9,14-25,27,29

0.88 [0.80, 0.97] 0.60 [0.44, 0.82]

0·02

Prospecve Retrospecve

4,7,8,11,12,15,23,25 3,5,6,12,14,16-22,24,27-29

0.84 [0.57, 1.26]

Mulcenter Single-center

3,5,7,8,11,15,19,22,23 4,6,9,12,14,16-18, 20,21,24,25, 27-29

0.89 [0.80, 0.99] 0.64 [0.49, 0.84]

Popula
on Older paents General populaon

5

Study type

0.2 0.5 1 Favours DES

716

0.74 [0.62, 0.89]

2 5 Favours BMS

0·6

0·02

Figure 3. Subgroup analyses for the effects of drug-eluting (DES) versus baremetal stent (BMS) use on mortality. Abbreviations: CI, confidence interval; RCT, randomized controlled trial. Am J Kidney Dis. 2013;62(4):711-721

DES in Patients With Decreased GFR Odds ratio (95% CI)

Study ID

Figure 4. Odds ratio of repeat revascularization associated with drug-eluting (DES) and bare-metal stent (BMS) use. Abbreviation: CI, confidence interval.

Aoyama 2008 Appleby 2009 Barthelemy 2012 Charytan 2011 Das 2006 Garg 2010 Green 2011 Halkin 2005 Halkin 2006 Ichimoto 2010 Ishii 2012 Ishio 2007 Jeong 2008 Kersting 2012 Kim 2009 Kuchulakanti 2006 Lemos 2005 Okada 2008 Oshima 2011 Resmini 2012 Rosenblum 2008 shenoy 2010 Simsek 2011 Tsai 2011 Yachi 2009 Zhang 2006 Overall (I-squared = 73.4%, p = 0.000)

Favors DES

CI, 0.51-0.74; P ⬍ 0.001) with a much reduced I2 value of 16% (P for heterogeneity ⫽ 0.2). Subgroup analysis showed that the benefit of DESs over BMSs on repeat revascularization exists in all subgroups, but was more prominent in RCTs (P ⫽ 0.006), studies with smaller (⬍400) numbers of

Subgroup

Studies included

Renal func
on stra
fica
on Dialysis dependent 5, 14,15,17-19,22,23,24,26,27 Non-dialysis dependent 3-9,11-13,16,20,21,25,28,29

.1

0.80 (0.37, 0.68 (0.53, 0.71 (0.32, 0.69 (0.48, 0.12 (0.01, 0.24 (0.12, 0.69 (0.45, 0.39 (0.16, 0.36 (0.09, 0.75 (0.32, 0.48 (0.33, 0.68 (0.25, 0.70 (0.19, 1.15 (0.55, 0.33 (0.11, 0.24 (0.08, 0.23 (0.08, 0.61 (0.32, 0.66 (0.44, 0.77 (0.32, 0.65 (0.45, 0.50 (0.27, 0.55 (0.25, 1.03 (0.98, 0.87 (0.37, 0.97 (0.40, 0.61 (0.50,

.2

.5

Odds ra
o, 95% CI

3,5,6,11-13,21,26,27 4,7-9,14-20,22-24,25,28,29

0.69 [0.55, 0.88] 0.54 [0.41, 0.71]

0·01

3,8,9,13,14,16-26 4-7,11,12,15,27-29

0.63 [0.57, 0.75] 0.63 [0.47, 0.84}

0·9

5

1.03 [0.98, 1.08] 0.61 [0.51, 0.74]

<0·001

4,6-29

Revasculariza
on type TVR TLR Any revascularizaon

4,7-9,11,12,15,16,20-22,28,29 8,13,14,15,17-20,24-27 3,5,6,23

0.50 [0.37, 0.68} 0.57 [0.47, 0.70} 0.78 [0.55, 1.09}

0·2

Study type RCT Observaonal studies

7,8 3-6,9,11-29

0.29 [0.17, 0.50] 0.65 [0.54, 0.78]

0·006

Adjusted or RCTs Unadjusted

3,5-8,11,12,28 4,9,13-27,29

0.65 [0.48, 0.89] 0.61 [0.51, 0.74]

0·7

Prospecve Retrospecve

0.49 [0.37, 0.66] 4,7,8,11,12,15,23,25 0.68 [0.55, 0.84] 3,5,6,12-14,16-22,24,2629 3,5,7,8,11,15,19,22,23,26 0.56 [0.39, 0.79] 4,6,9,12-14,16-18, 0.64 [0.56, 0.74] 20,21,24,25, 27-29 0.1 0.2 0.5 1 2 5 10 Favours DES Favours BMS

Mulcenter Single-center

Am J Kidney Dis. 2013;62(4):711-721

5

10

Favors BMS

p value for heterogeneity

0·6

Popula
on Older paents General paents

2

3.52 6.81 3.41 6.02 0.80 3.95 5.65 2.92 1.61 3.17 6.11 2.59 1.74 3.75 2.19 2.32 2.46 4.23 5.82 3.11 6.08 4.32 3.54 7.67 3.19 3.02 p < 0.001

participants (P ⫽ 0.001), and studies involving the general population versus older patients (P ⬍ 0.001; Fig 5). On univariate metaregression, DESs produced more benefits on repeat revascularization in RCTs (P ⫽ 0.02) or studies involving a general population versus older patients (P ⬍ 0.001; Table 3).

0.69 [0.49, 0.96] 0.61 [0.49, 0.77]

Pa
ents number ≥400 <400 Follow-up <2y ≥2 y

1

1.74) 0.88) 1.58) 1.01) 0.98) 0.48) 1.06) 0.97) 1.44) 1.76) 0.69) 1.85) 2.62) 2.39) 1.02) 0.71) 0.65) 1.16) 0.99) 1.83) 0.94) 0.94) 1.19) 1.08) 2.03) 2.36) 0.74)

Weight (%)

0·08

0·5

Figure 5. Subgroup analyses for the effects of drug-eluting (DES) versus baremetal stent (BMS) use on repeat revascularization. Abbreviations: CI, confidence interval; RCT, randomized controlled trial; TLR, target-lesion revascularization; TVR, targetlesion revascularization. 717

Wang et al Table 3. Univariate Metaregression Analysis of Possible Sources of Heterogeneity Across Studies Possible Source of Heterogeneity

Changea in ORs

95% CI

P

Mortality Stages of kidney function (per each additional stage) No. of patients (ⱖ400 vs ⬍400) Duration of follow-up (per 1-y greater) Population (older patients vs general population) RCT vs non-RCT Adjusted/RCT vs nonadjusted Prospective vs retrospective Multicenter vs single center

1.17 1.18 1.10 1.39 1.80 1.39 0.99 1.42

0.85-1.60 0.75-1.86 0.95-1.27 0.89-2.17 0.72-4.49 0.91-2.12 0.60-1.64 0.93-2.18

0.3 0.5 0.2 0.1 0.2 0.1 0.9 0.1

Repeat revascularization Stages of kidney function (per each additional stage) No. of patients (ⱖ400 vs ⬍400) Duration of follow-up (per 1-y greater) Population (older patients vs general population) RCTs vs non-RCTs Adjusted/RCT vs nonadjusted Prospective vs retrospective Multicenter vs single center

1.14 1.29 0.97 1.69 0.44 1.14 0.83 0.98

0.91-1.42 0.93-1.81 0.86-1.10 1.46-1.94 0.22-0.86 0.82-1.59 0.55-1.25 0.68-1.39

0.2 0.1 0.7 ⬍0.001 0.02 0.4 0.4 0.9

Abbreviations: CI, confidential interval; OR, odds ratio; RCT, randomized controlled trial. Proportional change.

a

MI and Stent Thrombosis A total of 59,970 patients were included in 18 studies reporting MI rates.3,5,7,8,11,12,14-23 One study was excluded because the MI rates in both groups were zero.14 Overall, DES versus BMS use was associated with a 15% reduction in MI (OR, 0.85; 95% CI, 0.79-0.92; P ⬍ 0.001), with no evidence of heterogeneity (Fig 6, upper portion). Data for stent thrombosis were available from 14 studies including 3,159 patients in whom 82 events were recorded.4,7,8,12,14-17,19,24-27,29 There was no significant difference between DES and BMS use (OR, 0.72; 95% CI, 0.46-1.12; P ⫽ 0.1). No evidence of statistical heterogeneity was identified (Fig 6, lower portion).

DISCUSSION This meta-analysis demonstrates that compared with BMSs, the use of DESs in patients with eGFR ⬍60 mL/min/1.73 m2 is associated with significantly decreased incidences of repeat revascularization and MI without increasing the risk of stent thrombosis. Furthermore, a strong trend toward a lower mortality rate was documented in DES-treated patients. GFR decline is highly prevalent and represents an important high-risk subset in patients receiving revascularization.1-9 Even patients with eGFR ⬍60 mL/min/ 1.73 m2 are at increased risk of mortality, MI, restenosis, and contrast-induced nephropathy, which is also an important independent risk factor for adverse outcomes after percutaneous coronary intervention.1-9,34 In addition to the greater frequency of traditional risk 718

factors such as diabetes and hypertension, patients with decreased GFR are more likely to have metabolic and biochemical abnormalities, including systemic inflammation and increased levels of atherogenic markers, which might predispose to accelerated cardiovascular disease.35-37 DESs have been shown to produce greater benefits in patients with high risk of restenosis, such as those with diabetes mellitus, long lesions, and small vessels.10 Patients with decreased GFR, clearly representing a high-risk population, may have the potential to derive greater benefit from DESs. Unfortunately, few studies have directly addressed treatment options on DES versus BMS use specifically in patients with decreased GFR. Because clear evidence of optimal stent choice in patients with decreased GFR is lacking, the best stent choice has been the subject of much debate for a long time and continues to be. Several recent large observational studies using propensity matching did not find significant benefit of DESs over BMSs on repeat revascularization,3,5,11 making the efficacy of DESs in this population questionable. Also, the attenuated benefit of clopidogrel and increased risk of stent thrombosis in patients with decreased GFR versus those with normal kidney function raised the concern of DES safety in this population.30,38,39 In this meta-analysis, superiority of the DES over the BMS in decreasing mortality was documented with a 23% reduction in OR. However, this survival benefit of DESs over BMSs should be interpreted with caution due to the nonrandomized nature of data sources and heterogeneity across studies. The survival Am J Kidney Dis. 2013;62(4):711-721

DES in Patients With Decreased GFR Odds ratio (95% CI)

Study ID Aoyama 2008

0.59 (0.10, 3.45)

0.17

Charytan 2011

0.84 (0.63, 1.13)

6.19

Das 2006

1.81 (0.32, 10.14)

0.18

Garg 2010

1.09 (0.42, 2.81)

0.60

Green 2011

0.88 (0.50, 1.54)

1.73

Halkin 2005

1.14 (0.37, 3.48)

0.44

Halkin 2006

1.95 (0.31, 12.43)

0.16

Ichimoto 2010

0.71 (0.10, 4.88)

0.15

Ishio 2007

0.33 (0.01, 8.00)

0.05

Jeong 2008

0.61 (0.14, 2.63)

0.25

Kersting 2012

0.97 (0.48, 1.97)

1.08

Kim 2009

1.89 (0.18, 20.22)

0.10

Kuchulakanti 2006

1.30 (0.75, 2.26)

1.77

Okada 2008

7.72 (0.38, 158.58)

0.06

Resmini 2012

0.41 (0.13, 1.34)

0.39

shenoy 2010

0.96 (0.49, 1.88)

1.20

Tsai 2011

0.84 (0.78, 0.91)

85.34

Yachi 2009

0.55 (0.08, 3.89)

0.14

Overall (I-squared = 0.0%, p = 0.922)

0.85 (0.79, 0.92)

p < 0.001

Favors DES

.1 .2

.5 1

2

5 10

Weight (%)

Barthelemy 2012

0.59 (0.06, 5.64)

3.82

Garg 2010

1.09 (0.07, 17.24)

2.55

Halkin 2005

0.16 (0.01, 3.36)

2.09

Ichimoto 2010

1.44 (0.13, 16.37)

3.29

Ishii 2012

0.39 (0.11, 1.31)

13.13

Ishio 2007

0.33 (0.01, 8.21)

1.88

Jeong 2008

0.16 (0.00, 8.21)

1.25

Kersting 2012

0.64 (0.30, 1.34)

35.22

Kim 2009

0.94 (0.20, 4.47)

8.04

Okada 2008

3.10 (0.29, 33.63)

3.41

Resmini 2012

0.65 (0.07, 5.71)

4.13

shenoy 2010

1.12 (0.38, 3.29)

16.82

Simsek 2011

4.68 (0.26, 85.69)

2.30

Yachi 2009

0.23 (0.01, 4.93)

2.07

Overall (I-squared = 0.0%, p = 0.868)

0.72 (0.46, 1.12)

p = 0.1

Favors DES

benefit of DESs over BMSs was attenuated in RCTs and adjusted observational studies versus unadjusted observational studies. Our results are similar to a meta-analysis of studies in the general population, which found a significantly lower mortality rate of DESs versus BMSs in observational studies, but not in RCTs.40 Despite larger sample size and more power to detect differences in low-frequency events compared with RCTs, observational analyses, which constituted most of our data sources, are subject to confounding in regard to the nonrandomized selection of either DESs or BMSs. Multivariable adjustment or propensity score matching can be used to diminish the effect of measured confounders on the effect estimate within individual studies. As such, the observed attenuation of the overall summary estimate of mortality Am J Kidney Dis. 2013;62(4):711-721

Favors BMS Odds ratio (95% CI)

Study ID

Figure 6. Odds ratios of myocardial infarction (MI; upper portion) and stent thrombosis (lower portion) associated with drug-eluting (DES) and bare-metal stent (BMS) use. Abbreviation: CI, confidence interval.

Weight(%)

.1 .2

.5 1

2

5 10

Favors BMS

favoring DESs versus BMSs in RCTs or adjusted studies compared with unadjusted studies is notable. Similar to our findings, Charytan et al11 also found a reduced crude mortality risk of DESs compared with BMSs that disappeared after propensity score matching. Therefore, the mortality benefit of DESs over BMSs found in this analysis still needs to be verified in larger RCTs. With respect to DES efficacy, this meta-analysis provided conclusive evidence that DES use decreased the occurrence of repeat revascularization in patients with chronic kidney disease, with a 39% reduction in OR estimate. Despite the variation in categories for repeat revascularization across included studies, all individual estimates consistently favor DESs over BMSs, except the study by Tsai et al,5 which exclusively involved an 719

Wang et al

elderly population. A potential explanation is that for elderly patients, the possible lower risk of repeat revascularization is more prone to be masked by longer survival time driven by DES use. Late stent thrombosis after DES implantation is an increasingly important issue. The risk of stent thrombosis could be even higher in patients with decreased GFRs. In a post hoc analysis of the CREDO (Clopidogrel for the Reduction of Events During Observation) trial,38 patients with eGFR ⬍90 mL/min/1.73 m2 were found to derive much less benefit from clopidogrel therapy compared with patients with normal kidney function. Attenuated clopidogrel responsiveness in patients with chronic kidney disease also has been reported by Park et al.39 In a recent study, eGFR ⬍60 mL/min/1.73 m2 was found to be an important independent risk factor for stent thrombosis after percutaneous coronary intervention.24 In this metaanalysis, reported rates of stent thrombosis for both DESs and BMSs were low (DESs, 2.4%; BMSs, 2.8%). We did not find a significant difference in stent thrombosis between DES and BMS use in this highrisk cohort. Although to our knowledge our data represent the largest reported sample size regarding stent thrombosis with DESs versus BMSs in patients with decreased eGFR, this issue needs to be addressed further by RCTs. We believe our analysis adds to the current evidence of optimal stent choice in the population with decreased eGFR. However, as with any meta-analysis, conclusions based on such data are subject to the limitations of the primary studies. Our results are based mainly on observational studies and post hoc analyses of RCTs, which are susceptible to selection bias and residual confounding. Although several studies attempted to adjust the potential confounders, crude event rates were the only available data for most of the primary studies to draw unadjusted estimates. Therefore, results from such data should be interpreted cautiously. Finally, because data were collected from many studies, follow-up length varied and baseline data were heterogeneous. Use of the randomeffects model, careful assessment of quality and heterogeneity, sensitivity analysis, subgroup analysis, and metaregression should help overcome some of these limitations. In summary, this meta-analysis provides substantial evidence that DESs significantly decrease the occurrence of repeat revascularization and MI without an increase in the incidence of mortality or stent thrombosis in patients with eGFR ⬍60 mL/min/ 1.73 m2. The true effect of DESs compared with BMSs with respect to rates of mortality needs to be confirmed by RCTs. 720

ACKNOWLEDGEMENTS Support: None. Financial Disclosure: The authors declare that they have no relevant financial interests.

SUPPLEMENTARY MATERIAL Table S1: Quality assessment of included studies. Item S1: Search strategy. Note: The supplementary material accompanying this article (http://dx.doi.org/10.1053/j.ajkd.2013.04.014) is available at www.ajkd.org.

REFERENCES 1. Blackman DJ, Pinto R, Ross JR, et al. Impact of renal insufficiency on outcome after contemporary percutaneous coronary intervention. Am Heart J. 2006;151(1):146-152. 2. Best PJ, Lennon R, Ting HH, et al. The impact of renal insufficiency on clinical outcomes in patients undergoing percutaneous coronary interventions. J Am Coll Cardiol. 2002;39(7):11131119. 3. Green SM, Selzer F, Mulukutla SR, et al. Comparison of bare-metal and drug-eluting stents in patients with chronic kidney disease (from the NHLBI Dynamic Registry). Am J Cardiol. 2011;108(11):1658-1664. 4. Simsek C, Magro M, Boersma E, et al; Interventional Cardiologists of the Thoraxcenter. Impact of renal insufficiency on safety and efficacy of drug-eluting stents compared to bare-metal stents at 6-years. Catheter Cardiovasc Interv. 2012;80(1):18-26. 5. Tsai TT, Messenger JC, Brennan JM, et al. Safety and efficacy of drug-eluting stents in older patients with chronic kidney disease: a report from the linked CathPCI Registry-CMS Claims Database. J Am Coll Cardiol. 2011;58(18):1859-1869. 6. Appleby CE, Ivanov J, Lavi S, et al. The adverse long-term impact of renal impairment in patients undergoing percutaneous coronary intervention in the drug-eluting stent era. Circ Cardiovasc Interv. 2009;2(4):309-316. 7. Garg P, Charytan DM, Novack L, et al. Impact of moderate renal insufficiency on restenosis and adverse clinical events after sirolimus-eluting and bare metal stent implantation (from the SIRIUS trials). Am J Cardiol. 2010;106(10):1436-1442. 8. Halkin A, Mehran R, Casey CW, et al. Impact of moderate renal insufficiency on restenosis and adverse clinical events after paclitaxel-eluting and bare metal stent implantation: results from the TAXUS-IV Trial. Am Heart J. 2005;150(6):1163-1170. 9. Lemos PA, Arampatzis CA, Hoye A, et al. Impact of baseline renal function on mortality after percutaneous coronary intervention with sirolimus-eluting stents or bare metal stents. Am J Cardiol. 2005;95(2):167-172. 10. Tu JV, Bowen J, Chiu M, et al. Effectiveness and safety of drug-eluting stents in Ontario. N Engl J Med. 2007;357(14):13931402. 11. Charytan DM, Varma MR, Silbaugh TS, Lovett AF, Normand SL, Mauri L. Long-term clinical outcomes following drugeluting or bare-metal stent placement in patients with severely reduced GFR: results of the Massachusetts Data Analysis Center (Mass-DAC) State Registry. Am J Kidney Dis. 2011;57(2):202211. 12. Shenoy C, Boura J, Orshaw P, Harjai KJ. Drug-eluting stents in patients with chronic kidney disease: a prospective registry study. PLoS One. 2010;5(11):e15070. 13. Rosenblum MA, Robbins MJ, Farkouh ME, Winston JA, Kim MC. Diminished benefits of drug-eluting stents versus bare metal stents in patients with severe renal insufficiency. Nephron Clin Pract. 2009;113(3):c198-c202. Am J Kidney Dis. 2013;62(4):711-721

DES in Patients With Decreased GFR 14. Yachi S, Tanabe K, Tanimoto S, et al. Clinical and angiographic outcomes following percutaneous coronary intervention with sirolimus-eluting stents versus bare-metal stents in hemodialysis patients. Am J Kidney Dis. 2009;54(2):299-306. 15. Kim BK, Oh S, Jeon DW, et al; Korean Multicenter Angioplasty Team (KOMATE) Investigators. Long-term clinical outcomes and stent thrombosis of sirolimus-eluting versus bare metal stents in patients with end-stage renal disease: results of Korean Multicenter Angioplasty Team (KOMATE) Registry. J Interv Cardiol. 2009;22(5):411-419. 16. Jeong YH, Hong MK, Lee CW, et al. Impact of significant chronic kidney disease on long-term clinical outcomes after drugeluting stent versus bare metal stent implantation. Int J Cardiol. 2008;125(1):36-40. 17. Okada T, Hayashi Y, Toyofuku M, et al. One-year clinical outcomes of dialysis patients after implantation with sirolimuseluting coronary stents. Circ J. 2008;72(9):1430-1435. 18. Aoyama T, Ishii H, Toriyama T, et al. Sirolimus-eluting stents vs bare metal stents for coronary intervention in Japanese patients with renal failure on hemodialysis. Circ J. 2008;72(1): 56-60. 19. Ishio N, Kobayashi Y, Takebayashi H, et al. Impact of drug-eluting stents on clinical and angiographic outcomes in dialysis patients. Circ J. 2007;71(10):1525-1529. 20. Kuchulakanti PK, Torguson R, Chu WW, et al. Impact of chronic renal insufficiency on clinical outcomes in patients undergoing percutaneous coronary intervention with sirolimus-eluting stents versus bare metal stents. Am J Cardiol. 2006;97(6):792-797. 21. Zhang RY, Ni JW, Zhang JS, et al. Long term clinical outcomes in patients with moderate renal insufficiency undergoing stent based percutaneous coronary intervention. Chin Med J (Engl). 2006;119(14):1176-1181. 22. Das P, Moliterno DJ, Charnigo R, et al. Impact of drugeluting stents on outcomes of patients with end-stage renal disease undergoing percutaneous coronary revascularization. J Invasive Cardiol. 2006;18(9):405-408. 23. Halkin A, Selzer F, Marroquin O, Laskey W, Detre K, Cohen H. Clinical outcomes following percutaneous coronary intervention with drug-eluting vs bare-metal stents in dialysis patients. J Invasive Cardiol. 2006;18(12):577-583. 24. Ichimoto E, Kobayashi Y, Iijima Y, Kuroda N, Kohno Y, Komuro I. Long-term clinical outcomes after sirolimus-eluting stent implantation in dialysis patients. Int Heart J. 2010;51(2): 92-97. 25. Barthelemy O, Helft G, Silvain J, et al. One-year clinical outcomes in patients with chronic renal failure treated by percutaneous coronary intervention with drug-eluting stent. Arch Cardiovasc Dis. 2011;104(12):604-610. 26. Oshima S, Noda K, Ogawa H, et al. Drug-eluting stents have no dramatic effects on dialysis patients. Eur Heart J. 2011; 32(suppl 1):754.

Am J Kidney Dis. 2013;62(4):711-721

27. Ishii H, Toriyama T, Aoyama T, et al. Percutaneous coronary intervention with bare metal stent vs drug-eluting stent in hemodialysis patients. Circ J. 2012;76(7):1609-1615. 28. Kersting S, Grumann T, Hummel J, Hauschke D, Bode C, Hehrlein C. Impact of chronic kidney disease on long-term clinical outcomes after percutaneous coronary intervention with drug-eluting or bare-metal stents. Crit Pathways Cardiol. 2012;11(3):152-159. 29. Resmini C, Di Cuia M, Ballocca F, et al. Short and long term outcome of percutaneous coronary intervention with drug eluting stent and bare metal stent in patients with chronic kidney disease. Minerva Cardioangiol. 2012;60(6):573-580. 30. Lambert ND, Sacrinty MT, Ketch TR, et al. Chronic kidney disease and dipstick proteinuria are risk factors for stent thrombosis in patients with myocardial infarction. Am Heart J. 2009;157(4): 688-694. 31. Stroup DF, Berlin JA, Morton SC, et al. Meta-analysis of Observational Studies in Epidemiology: a proposal for reporting. JAMA. 2000;283(15):2008-2012. 32. Vandenbroucke JP, von Elm E, Altman DG, et al; for the STROBE Initiative. Strengthening the Reporting of Observational Studies in Epidemiology (STROBE): explanation and elaboration. Ann Intern Med. 2007;147(8):573-577. 33. Sanderson S, Tatt ID, Higgins JP. Tools for assessing quality and susceptibility to bias in observational studies in epidemiology: a systematic review and annotated bibliography. Int J Epidemiol. 2007;36(3):666-676. 34. Brown JR, DeVries JT, Piper WD, et al. Serious renal dysfunction after percutaneous coronary interventions can be predicted. Am Heart J. 2008;155(2):260-266. 35. Cheung AK, Sarnak MJ, Yan G, et al. Atherosclerotic cardiovascular disease risks in chronic hemodialysis patients. Kidney Int. 2000;58(1):353-362. 36. Schiffrin EL, Lipman ML, Mann JF. Chronic kidney disease: effects on the cardiovascular system. Circulation. 2007;116(1): 85-97. 37. Hage FG, Venkataraman R, Zoghbi GJ, Perry GJ, DeMattos AM, Iskandrian AE. The scope of coronary heart disease in patients with chronic kidney disease. J Am Coll Cardiol. 2009; 53(23):2129-2140. 38. Best PJ, Steinhubl SR, Berger PB, et al; CREDO Investigators. The efficacy and safety of short- and long-term dual antiplatelet therapy in patients with mild or moderate chronic kidney disease: results from the Clopidogrel for the Reduction of Events During Observation (CREDO) trial. Am Heart J. 2008;155(4):687693. 39. Park SH, Kim W, Park CS, Kang WY, Hwang SH, Kim W. A comparison of clopidogrel responsiveness in patients with versus without chronic renal failure. Am J Cardiol. 2009;104(9):12921295. 40. Kirtane AJ, Gupta A, Iyengar S, et al. Safety and efficacy of drug-eluting and bare metal stents: comprehensive meta-analysis of randomized trials and observational studies. Circulation. 2009; 119(25):3198-3206.

721