Abstract
Background
Few published data describe long-term survival of dialysis patients undergoing surgical versus percutaneous coronary revascularization in the era of drug-eluting stents (DES).
Methods and Results
Using United States Renal Data System data, we identified 23,033 dialysis patients who underwent coronary revascularization (6178 coronary artery bypass grafting [CABG], 5011 bare-metal stent [BMS], 11,844 DES), 2004–2009. Revascularization procedures decreased from 4347 in 2004 to 3344 in 2009. DES use decreased by 41% and BMS use increased by 85% 2006–2007. Long-term survival was estimated by the Kaplan-Meier method and independent predictors of mortality examined in a comorbidity-adjusted Cox model. In-hospital mortality for CABG patients was 8.2%; all-cause survival at 1, 2, and 5 years was 70%, 57%, and 28% respectively. In-hospital mortality for DES patients was 2.7%; 1, 2, and 5 year survival was 71%, 53%, and 24% respectively. Independent predictors of mortality were similar in both cohorts: age >65 years, white race, dialysis duration, peritoneal dialysis, and congestive heart failure, but not diabetes. Survival was significantly higher for CABG patients who received internal mammary grafts (IMG) (HR 0.83, P<0.0001). Probability of repeat revascularization accounting for the competing risk of death was 18% with BMS, 19% with DES, and 6% with CABG at 1 year.
Conclusions
Among dialysis patients undergoing coronary revascularization, in-hospital mortality was higher after CABG but long-term survival was superior with IMGs. In-hospital mortality was lower for DES patients, but probability of repeat revascularization was higher and comparable to BMS patients. Revascularization decisions for dialysis patients should be individualized.
Keywords: coronary, dialysis, drug-eluting stent, survival
Introduction
Randomized studies have compared long-term outcomes of coronary artery bypass grafting (CABG) and percutaneous coronary intervention (PCI) with bare-metal stents (BMS) in non-dialysis patients, yielding concordant results: comparable rates of death and myocardial infarction, but higher rates of repeat revascularization with BMS.1,2 Since US Food and Drug Administration approval in 2003, drug-eluting stents (DES) have been used for a majority of PCI procedures in the US owing to reduced incidence of in-stent restenosis and repeat revascularization, but they have failed to demonstrate any survival advantage relative to BMS.3 Improved outcomes associated with DES have led to their evaluation in randomized studies comparing outcomes relative to CABG in high-risk patient subsets for whom surgical revascularization was traditionally recommended.4 However, patients with end-stage renal disease (ESRD) undergoing maintenance dialysis have been excluded from any randomized evaluation of the comparative efficacy of coronary revascularization strategies.
Cardiac disease is a major cause of mortality among dialysis patients; according to recent estimates, it accounts for about 38% of all-cause deaths, and about 13% of cardiac deaths are ascribed to myocardial infarction.5 Dialysis patients constitute, arguably, one of the highest risk groups with regard to mortality associated with coronary revascularization. Herzog et al6 reported 2-year survival rates of 56% and 48% following CABG and PCI with BMS, respectively. Hemmelgarn et al7 noted 8-year mortality rates following coronary revascularization nearly 2-fold higher among dialysis than non-dialysis patients: 86% versus 45% following CABG, and 80% versus 41% following PCI. More recently, Tsai et al8 reported 30-month mortality rates of 52% for long-term dialysis patients undergoing PCI 2004–2007.
In the earliest era of PCI, percutaneous transluminal coronary angioplasty (PTCA) and CABG, 1978–1995, were compared using United States Renal Data System (USRDS) data. Despite higher in-hospital mortality rates with CABG versus PTCA (12.5% versus 5.4%), all-cause mortality rates at 1 year were comparable and 2-year survival was more favorable following surgical revascularization.9 In the era of BMS, comparative survival following coronary revascularization was readdressed using USRDS data, 1995–1998.6 Again, despite 2-fold higher in-hospital mortality with CABG (8.6% versus 4.1% for BMS), improved survival with CABG was noted within 1 year after the index procedure.
The advent of DES renewed optimism regarding the potential of improving PCI-related outcomes in dialysis patients. In a meta-analysis of seven studies, Abdel-Latif et al10 reported significant reduction in repeat revascularization and a trend toward reduced mortality with DES compared with BMS in dialysis patients. In a large cohort of US Medicare patients including 5182 dialysis patients, Tsai et al8 reported a statistically significant mortality benefit associated with DES versus BMS. Conversely, high in-hospital mortality rates after CABG continue to remain a formidable challenge11; techniques such as off-pump surgery have demonstrated only a modest impact on short-term mortality.12
Thus, the optimal option for coronary revascularization in dialysis patients remains a matter of debate. Nevis et al13 concluded from an exploratory meta-analysis of 17 studies evaluating CABG versus PCI that existing data are inadequate to make a determination regarding the optimal strategy of coronary revascularization for dialysis patients. Using USRDS data, we sought to evaluate long-term survival rates and probability of repeat coronary revascularization of dialysis patients undergoing surgical and percutaneous coronary revascularization in the contemporary era, and to identify independent predictors of mortality.
Methods
Using the USRDS database (n = 546,160 eligible dialysis patients, 2004–2009), this retrospective study identified dialysis patients who were hospitalized for their first coronary revascularization procedure after initiation of renal replacement therapy (n = 23,033). Eligible patients had received renal replacement therapy for ≥ 90 days before revascularization. Using Medicare claims, all patients undergoing CABG without concomitant valve surgery or PCI with DES or BMS placement were identified between January 1, 2004, and December 31, 2009, and followed through December 31, 2010. Using the codes listed in Table S1 (supplemental material), 6178 CABG patients, 5011 BMS patients, and 11,844 DES patients were identified in the study period. Patients undergoing PTCA alone were not included in the analysis. Patients who underwent both surgical and percutaneous intervention during the same hospital stay were also excluded.
Survival was determined from the time of revascularization to death or censoring. Patients who underwent renal transplant or were lost to follow-up before December 31, 2010, were censored. Death was identified from the USRDS database. Cause-specific mortality was determined using the Centers for Medicare & Medicaid ESRD Death Notification (form CMS-2746). Long-term survival was estimated with the Kaplan-Meier method, using the log-rank test to compare differences in survival. Comorbidity-adjusted Cox proportional hazards models were used to assess the effect of comorbid conditions on survival in each revascularization cohort. Comorbid conditions studied included prior myocardial infarction; congestive heart failure; other cardiac conditions including valvular heart disease; presence of a pacemaker, and arrhythmia; prior coronary revascularization; non-skin malignancies; peripheral vascular disease; cerebrovascular accident or transient ischemic attack; chronic obstructive pulmonary disease; gastrointestinal disease; gall bladder disease; and liver disease. Cumulative probability of repeat coronary revascularization (accounting for the competing risk of death) was calculated using unadjusted, non-parametric methods.14
The chi square test was used to detect differences between proportions. All reported P values are 2-sided. All statistical analyses were performed using the SAS system for Windows, version 9.2 (SAS institute, Inc.).
Results
Between 2004 and 2009, 23,033 dialysis patients underwent coronary revascularization procedures; 6178 underwent CABG (4521 [73%] with internal mammary graft [IMG]), 11,844 underwent PCI with DES, and 5011 underwent PCI with BMS. Median follow-up periods were 1.63 years for CABG (25th percentile, 0.55; 75th percentile, 2.96), 1.60 years for DES, (0.75, 2.96), and 0.99 years for BMS (0.47, 2.26). The trend in revascularization procedures in dialysis patients, 2004–2009, is depicted in Figure 1. During the study period, the overall number of coronary revascularization procedures decreased from 4347 in 2004 to 3344 in 2009. Annual volume of CABG procedures was similar during the study period, typically accounting for 25% to 30% of all procedures. From 2004 to 2006, DES accounted for 59% of all procedures and BMS for about 15%. However, from 2007 to 2009, a dramatic change occurred in the pattern of stent use, with a marked decrease in use of DES and a corresponding increase in use of BMS. From 2006 to 2007 alone, the number of DES procedures dropped from 2494 to 1462 (a 41% reduction), while the number of BMS procedures increased from 606 to 1120 (an 85% increase). Preliminary data from 2010 suggest a trend toward a rebound in DES use (47% of all procedures) and a corresponding decrease in BMS use (26%).
Figure 1.
Rates of revascularization procedures in dialysis patients in the United States, 2004–2009. BMS, bare-metal stent; CABG, coronary artery bypass graft surgery; DES, drug-eluting stent.
Demographic characteristics of the study group are summarized in Table 1. Notably, a greater proportion of younger patients underwent surgical revascularization procedures than percutaneous procedures (P < 0.0001). Among patients who underwent CABG with IMG, 43% were aged 45 to 64 years, 12% 75 to 79 years, and only 7% older than 80 years. Among patients who underwent PCI with DES, however, 35% were aged 45 to 64 years and 31% 65 to 74 years, and proportions in older age groups were relatively higher; 16% were aged 75 to 79 years and 14% 80 years or older. There was a statistically significant difference in dialysis modality between groups; peritoneal dialysis was more common in the CABG than in the PCI group (P = 0.03). Diabetes as cause of renal failure was similar (58.2% versus 57.1%, P = 0.13), but a greater proportion of patients in the PCI than in the CABG group were likely to have hypertension as primary cause of renal failure (P = 0.016). Proportions of patients with congestive heart failure (63.8% versus 63.5%, P = 0.67) and diabetes (76.7% versus 77.1%, P = 0.61) were comparable in the CABG with IMG and the DES groups. Among patients who underwent CABG without IMG, prevalence of congestive heart failure was slightly higher and prevalence of diabetes somewhat lower than in the other revascularization groups.
Table 1.
Baseline characteristics of dialysis patients undergoing coronary revascularization procedures in the United States, 2004–2009.
| Characteristic | CABG | PCI | ||
|---|---|---|---|---|
|
| ||||
| With IMG | Without IMG | DES | BMS | |
| All, n = 23,033 | 4521 (19.6) | 1657 (7.2) | 11,844 (51.4) | 5011 (21.8) |
| Men | 2895 (64.0) | 952 (57.5) | 6455 (54.5) | 2891 (57.7) |
| Race | ||||
| White | 2960 (65.5) | 997 (60.2) | 7601 (64.2) | 3283 (65.5) |
| Black | 1222 (27.0) | 561 (33.9) | 3456 (29.2) | 1464 (29.2) |
| Other/unknown | 339 (7.5) | 99 (6.0) | 787 (6.6) | 264 (5.3) |
| Age, yr | ||||
| < 45 | 169 (3.7) | 99 (6.0) | 493 (4.2) | 191 (3.8) |
| 45–64 | 1920 (42.5) | 579 (34.9) | 4134 (34.9) | 1613 (32.2) |
| 65–74 | 1570 (34.7) | 554 (33.4) | 3706 (31.3) | 1602 (32.0) |
| 75–79 | 536 (11.9) | 250 (15.1) | 1871 (15.8) | 742 (14.8) |
| ≥ 80 | 326 (7.2) | 175 (10.6) | 1640 (13.8) | 863 (17.2) |
| ESRD etiology | ||||
| Diabetes | 2691 (59.5) | 903 (54.5) | 6962 (58.8) | 2654 (53.0) |
| Hypertension | 1129 (25.0) | 445 (26.9) | 3085 (26.0) | 1477 (29.5) |
| Other/unknown | 701 (15.5) | 309 (18.6) | 1797 (15.2) | 880 (17.6) |
| Dialysis duration, yr | ||||
| < 2 | 1809 (40.0) | 606 (36.6) | 4921 (41.5) | 1893 (37.8) |
| 2–5 | 1876 (41.5) | 680 (41.0) | 4909 (41.4) | 2121 (42.3) |
| 6–10 | 665 (14.7) | 290 (17.5) | 1648 (13.9) | 813 (16.2) |
| ≥ 11 | 171 (3.8) | 81 (4.9) | 366 (3.1) | 184 (3.7) |
| Dialysis modality | ||||
| Hemodialysis | 4203 (93.0) | 1570 (94.7) | 11,156 (94.2) | 4721 (94.2) |
| Peritoneal dialysis | 318 (7.0) | 87 (5.3) | 688 (5.8) | 290 (5.8) |
| Comorbid conditions | ||||
| CHF | 2886 (63.8) | 1101 (66.4) | 7518 (63.5) | 3161 (63.1) |
| COPD | 1202 (26.6) | 471 (28.4) | 3226 (27.2) | 1463 (29.2) |
| CVA/TIA | 1235 (27.3) | 462 (27.9) | 2743 (23.2) | 1156 (23.1) |
| Cancer | 316 (7.0) | 107 (6.5) | 934 (7.9) | 484 (9.7) |
| Diabetes | 3468 (76.7) | 1182 (71.3) | 9130 (77.1) | 3652 (72.9) |
| Dysrhythmia | 2029 (44.9) | 792 (47.8) | 4991 (42.1) | 2286 (45.6) |
| GI disease | 491 (10.9) | 232 (14.0) | 1399 (11.8) | 708 (14.1) |
| Liver disease | 436 (9.6) | 177 (10.7) | 1150 (9.7) | 517 (10.3) |
| Other cardiac | 2380 (52.6) | 975 (58.8) | 5957 (50.3) | 2570 (51.3) |
| PVD | 2100 (46.4) | 807 (48.7) | 5595 (47.2) | 2359 (47.1) |
Note: Values are n (percent).
BMS, bare-metal stent; CABG, coronary artery bypass graft; CHF, congestive heart failure; COPD, chronic obstructive pulmonary disease; CVA/TIA, cerebrovascular accident/transient ischemic attack; DES, drug-eluting stent; ESRD, end-stage renal disease; GI, gastrointestinal; IMG, internal mammary graft; PCI, percutaneous coronoary intervention; PVD, peripheral vascular disease.
Kaplan-Meier estimates for all-cause survival for each revascularization modality were individually constructed (Figures 2A–D). Of note, 1-year survival was significantly higher in 2009 than in 2004 for CABG (P = 0.002 by log rank test) and DES (P = 0.015) patients, but not for BMS patients (63.4% in 2004 versus 62.3% in 2009, P = 0.57). Thus, there was evidence of improvement in outcomes for CABG and DES during the study period. Surgical revascularization carried a significant initial mortality hazard. In-hospital death was 8.2% for patients undergoing CABG (Figure 2A); this rate was lower for patients undergoing revascularization with an IMG (7.8%) than for patients undergoing revascularization without an IMG (9.3%, P = 0.059). Survival rates at 1 month were 90% and 87% for CABG with and without IMG, respectively (Figure 2B). At 12 months after the index revascularization, all-cause survival for CABG with IMG was 72%, versus 64% for CABG without IMG. At 2 years, patients who underwent CABG with IMG had a significant survival advantage (59.6%) compared with patients who underwent CABG without IMG (50.6%, P < 0.0001), a difference that persisted in longer-term follow-up. For patients undergoing percutaneous revascularization, in-hospital mortality was 2.7% for DES and 4.9% for BMS (P < 0.0001). Thus, in-hospital mortality was significantly higher in the context of surgical vs. percutaneous revascularization (8.2% vs. 3.4%, P < 0.0001). Survival rates at 1 month were 94% for DES patients, and subsequent all-cause survival was 71% at 1 year and 53% at 2 years (Figure 2C). Survival rates for BMS patients were 63% at 1 year and 48% at 2 years (Figure 2D).
Figure 2.
Kaplan-Meier survival curves depicting all-cause survival after (A) coronary bypass surgery ([B] with or without internal mammary graft) and percutaneous coronary intervention (using [C] drug-eluting stents and [D] bare-metal stents) in dialysis patients, 2004–2009. BMS, bare-metal stent; CABG, coronary artery bypass graft surgery; DES, drug-eluting stent.
Of the nearly 14,000 deaths identified, more than half were attributable to cardiovascular causes (Table 2; all 55%, CABG 53%, DES 56%, BMS 57%), and about 10% were due to infection (all 9.6%, CABG 12%, DES 9.1%, BMS 8.1%). Thus, infection as a cause of death was more prevalent in the CABG than in the PCI cohort. Notably, withdrawal from dialysis accounted for 7.4% of all deaths. Almost 40% of the study cohort were censored prior to death, 91% due to end of the follow-up period, 7% due to change in type of renal replacement therapy (including 2% who underwent kidney transplant), and the remaining 2% due to changes in primary payer or eligibility status.
Table 2.
Cause of death among the revascularization cohorts.
| Cause of Death | All | Cohort
|
||
|---|---|---|---|---|
| CABG | DES | BMS | ||
| Cardiac | 7038 (50.4) | 1655 (47.1) | 3772 (51.1) | 1611 (52.6) |
| Cardiovascular | 7695 (55.1) | 1846 (52.6) | 4114 (55.7) | 1735 (56.7) |
| Infection | 1339 (9.6) | 420 (12.0) | 672 (9.1) | 247 (8.1) |
| Withdrawal | 1035 (7.4) | 243 (6.9) | 559 (7.6) | 233 (7.6) |
| Other/unknown | 3892 (27.9) | 1003 (28.6) | 2043 (27.7) | 846 (27.6) |
| Total | 13,961 | 3512 | 7388 | 3061 |
Note: Values are n (%).
BMS, bare-metal stent; CABG, coronary artery bypass graft; DES, drug-eluting stent
We created separate Cox models to assess independent predictors of mortality for each revascularization modality (Table 3). Not surprisingly, increasing age, increasing dialysis duration, and comorbid conditions (in particular, congestive heart failure, chronic obstructive pulmonary disease, cerebrovascular accident/transient ischemic attack, peripheral vascular disease) were independently predictive of increased mortality following both surgical and percutaneous revascularization. Black race was associated with a significantly reduced hazard of mortality in all three groups: CABG, HR 0.88, P = 0.0009; DES, HR 0.87, P < 0.0001; BMS, HR 0.86, P = 0.0006. Interestingly, diabetes was an independent predictor of mortality in the BMS group (HR 1.18, 95% CI 1.07–1.31), but not in the CABG (HR 1.07, 95% CI 0.97–1.1) or DES (HR 1.03, 95% CI 0.96–1.10) groups. Among patients undergoing surgical revascularization, use of an IMG was independently associated with a reduced hazard of long-term mortality (HR 0.83, 95% CI 0.77–0.90, P < 0.0001).
Table 3.
Results of three separate Cox proportional hazards models evaluating independent predictors of mortality after coronary revascularization, 2004–2009.
| Variables | Cohort
|
|||||
|---|---|---|---|---|---|---|
| DES | BMS | CABG | ||||
|
| ||||||
| HR (95% CI) | P | HR (95% CI) | P | HR (95% CI) | P | |
| Female | 0.98 (0.94–1.03) | 0.4044 | 0.96 (0.89–1.03) | 0.229 | 1.05 (0.98–1.12) | 0.1789 |
| Race | ||||||
| Black | 0.87 (0.82–0.91) | < 0.0001 | 0.86 (0.79–0.94) | 0.0006 | 0.88 (0.81–0.95) | 0.0009 |
| Other/unknown | 0.78 (0.71–0.86) | < 0.0001 | 0.86 (0.73–1.02) | 0.0862 | 0.88 (0.76–1.01) | 0.0669 |
| Age, yr | ||||||
| 65–74 | 1.24 (1.17–1.32) | < 0.0001 | 1.40 (1.27–1.53) | < 0.0001 | 1.33 (1.23–1.44) | < 0.0001 |
| ≥ 75 | 1.75 (1.65–1.86) | < 0.0001 | 1.92 (1.75–2.11) | < 0.0001 | 1.75 (1.60–1.91) | < 0.0001 |
| ESRD etiology | ||||||
| Diabetes | 1.21 (1.11–1.31) | < 0.0001 | 1.32 (1.17–1.48) | < 0.0001 | 1.16 (1.04–1.30) | 0.01 |
| Hypertension | 1.07 (0.99–1.16) | 0.079 | 1.24 (1.10–1.38) | 0.0003 | 1.08 (0.97–1.20) | 0.1674 |
| Dialysis duration, yr | ||||||
| 2–5 | 1.20 (1.14–1.26) | < 0.0001 | 1.26 (1.17–1.37) | < 0.0001 | 1.20 (1.11–1.30) | < 0.0001 |
| 6–10 | 1.32 (1.22–1.41) | < 0.0001 | 1.33 (1.20–1.48) | < 0.0001 | 1.41 (1.27–1.55) | < 0.0001 |
| ≥ 11 | 1.25 (1.08–1.44) | 0.0028 | 1.39 (1.12–1.73) | 0.0026 | 1.55 (1.30–1.86) | < 0.0001 |
| Dialysis modality | ||||||
| Peritoneal dialysis | 1.20 (1.07–1.35) | 0.0019 | 1.38 (1.16–1.63) | 0.0002 | 1.42 (1.22–1.64) | < 0.0001 |
| Comorbid conditions | ||||||
| CHF | 1.39 (1.31–1.46) | < 0.0001 | 1.31 (1.20–1.43) | < 0.0001 | 1.34 (1.23–1.45) | < 0.0001 |
| COPD | 1.26 (1.20–1.33) | < 0.0001 | 1.13 (1.05–1.23) | 0.0019 | 1.28 (1.19–1.38) | < 0.0001 |
| CVA/TIA | 1.15 (1.09–1.21) | < 0.0001 | 1.08 (0.99–1.18) | 0.068 | 1.19 (1.10–1.28) | < 0.0001 |
| Cancer | 1.20 (1.01–1.19) | 0.0276 | 1.15 (1.03–1.30) | 0.0177 | 1.04 (0.91–1.18) | 0.5801 |
| Diabetes | 1.03 (0.96–1.10) | 0.4388 | 1.18 (1.07–1.31) | 0.0018 | 1.07 (0.97–1.19) | 0.1731 |
| Dysrhythmia | 1.32 (1.26–1.39) | < 0.0001 | 1.31 (1.22–1.42) | < 0.0001 | 1.31 (1.22–1.40) | < 0.0001 |
| GI disease | 1.10 (1.03–1.18) | 0.006 | 1.11 (1.00–1.22) | 0.0417 | 1.10 (0.99–1.21) | 0.0708 |
| Liver disease | 1.09 (1.01–1.18) | 0.0209 | 1.10 (0.98–1.23) | 0.1025 | 1.13 (1.02–1.26) | 0.0215 |
| Other cardiac | 1.15 (1.10–1.21) | < 0.0001 | 1.16 (1.07–1.25) | 0.0002 | 1.21 (1.13–1.30) | < 0.0001 |
| PVD | 1.18 (1.13–1.24) | < 0.0001 | 1.26 (1.17–1.35) | < 0.0001 | 1.24 (1.16–1.33) | < 0.0001 |
| Mammary graft | 0.83 (0.77–0.90) | < 0.0001 | ||||
Note: Reference groups are age < 65 years, men, white race, ESRD etiology other/unknown, dialysis duration < 2 years, hemodialysis, absence of the condition.
BMS, bare-metal stent; CABG, coronary artery bypass graft; CHF, congestive heart failure; CI, confidence interval; COPD, chronic obstructive pulmonary disorder; CVA/TIA, cerebrovascular accident/transient ischemic attack; DES, drug-eluting stent; GI gastrointestinal; HR, hazard ratio; PVD peripheral vascular disease.
We further evaluated repeat coronary revascularization (CABG or PCI) accounting for the competing risk of death among patients undergoing surgical and percutaneous revascularization. Of 6178 patients who originally underwent CABG, 773 (12.5%) underwent a repeat coronary revascularization procedure; of 11,844 patients who originally received DES, 3404 (28.7%) underwent a repeat procedure, and of 5011 who originally received BMS, 1246 (24.9%) underwent a repeat procedure. Figure 3 shows the distribution of types of repeat procedures performed, given the original procedure. The most common type of repeat procedure was DES (49% of CAB patients, 58% of DES patients, and 41% of BMS patients). The cumulative probability of repeat revascularization was 1.4% at 1 month, 3.8% at 6 months, 6.0% at 1 year, 9.9% at 2 years, 12.7% at 3 years, and 15.7% at 5 years for CABG; corresponding estimates for DES were 5.7%, 13.6%, 19.1%, 25.1%, 28.8%, and 32.5%, and for BMS 5.0%, 14.1%,18.3%, 23.2%, 26.0%, and 29.0%. Importantly, the cumulative probability of repeat revascularization was similar in DES and BMS patients at all time intervals studied.
Figure 3.
Types of repeat coronary revascularization procedures in dialysis patients who originally underwent CABG, percutaneous coronary intervention with DES or BMS, or PTCA. BMS, bare-metal stent; CABG, coronary artery bypass graft; DES, drug-eluting stent; PTCA, percutaneous transluminal coronary angioplasty.
We attempted to quantify the variation in the distribution of type of revascularization procedures performed on a procedure center level. We counted the number of revascularization claims at each center (limited to centers at which at least 10 revascularization procedures were performed in dialysis patients during the year), then calculated what percentages of claims were for CABG and for PCI. Table S2 (supplemental material) displays the percentages of CABG procedures. In general, in 2004–2005, in the middle half of centers, about 17% to 36% of revascularization procedures were CABGs. The numbers in 2006–2008 reflect changes in revascularization types (fewer DES procedures), and in the middle half of centers, about 25% to 50% of revascularization procedures were CABGs. The distribution in 2009 was similar, but appeared to be moving closer to the 2004–2005 distribution. In summary, for the middle half of centers, CABGs made up between one-fourth and one-half of revascularization procedures, and for most centers, CABGs made up between 15% and 60%. Therefore, important variations in distribution of revascularization procedures occurred, although it could be assumed that patient selection was relatively similar for most centers, with only a few outlier centers.
Discussion
These observational data attempt to bridge a hiatus in the existing literature pertaining to the choice of the optimal revascularization strategy in dialysis patients in the contemporary era. Data from a large, representative sample of dialysis patients demonstrate the trade-offs pertinent to clinical decisions regarding the optimal revascularization strategy in this high-risk population: high in-hospital mortality rates but superior long-term survival (especially with use of IMGs) with surgical revascularization, and higher short-term survival but higher probability of repeat revascularization with percutaneous revascularization using BMS and DES.
Importantly, survival rates in the surgical and DES cohorts (but not in the BMS cohort) have improved somewhat in the contemporary era. We found 2-year survival rates after CABG with IMG, DES, and BMS of 60%, 53%, and 48%, respectively, from 2004 to 2009, compared with 56% and 48% after CABG and BMS, respectively; from 1995 to 1998.6 The shift in patterns of coronary revascularization among dialysis patients is also relevant. The decrement in DES use after 2006, with a proportional increase in BMS use (Figure 1), coincides with trends toward reduced PCI volume nationally since 2004,15 and in particular reduced DES volume after emergence of concerns regarding stent thrombosis.16 Multiple contemporaneous publications addressed concerns pertaining to use of DES, including off-label use, incidence of stent thrombosis relative to BMS, and especially the unique phenomenon of “late” stent thrombosis.17,18 National trends since 2010 have subsequently demonstrated an upswing in DES use.
Previous studies have compared outcomes of CABG and DES in dialysis patients with conflicting results. Manabe et al19 studied 28 CABG and 18 DES patients and reported similar 2-year survival in the two groups but significantly lower rates of major adverse cardiac outcomes in the surgical group. Sunagawa et al20 studied a series of 29 CABG and 75 DES patients, reporting a 2-year survival advantage with CABG. In these studies, short-term mortality rates associated with CABG were significantly lower than average, suggesting that these were selected patients from single centers, thus limiting the generalizability of the results. Ashrith et al21 evaluated patients with chronic kidney disease and multivessel coronary artery disease undergoing coronary revascularization, including 87 ESRD patients (54 CABG, 33 DES). The authors reported comparable 30-day survival rates in the two groups for dialysis patients, but a trend toward improved survival with CABG for chronic kidney disease patients. In summary, the theme from prior studies has been that in dialysis patients, the initial advantage of PCI, although longer lasting with DES than with BMS6 and PTCA,9 is generally overtaken by surgical revascularization after about 18 months. As we were preparing the revised version of this article, Chang et al22 published a study using USRDS data to examine the comparative effectiveness of multi-vessel PCI versus CABG in a propensity model. The overall survival data are concordant with data we report, but we believe that the structure of the USRDS database, which includes no data on coronary angiography, precludes valid assessment of comparative effectiveness of revascularization strategies.
Significant selection biases germane to the choice of revascularization procedures in actual clinical practice are contingent upon variables that cannot be retrospectively measured or reconciled accurately using an administrative database (e.g., angiographic characteristics including target vessel diameter, lesion complexity, and coronary calcification; left-ventricular ejection fraction; estimation of bleeding risk; surgical expertise; etc.). Thus, a direct comparison of survival between CABG and DES (or a comparative effectiveness analysis) was not judged to be reasonable and could potentially be misleading using these observational data. Instead, our study describes survival data for the revascularization cohorts and provides a national benchmark pertaining to short-term and long-term outcomes for each revascularization strategy to guide the clinician.
Reasons underlying an apparent lack of long-term benefit of PCI with DES in the dialysis population compared with non-ESRD patients bear emphasis. In particular, it is notable that the probability of repeat revascularization for DES was comparable to probability for BMS at all intervals studied among dialysis patients; this finding is contrary to findings in non-ESRD patients, but consistent with studies performed in Japan reporting rates of angiographically detected restenosis in dialysis patients receiving BMS (24%–43%) or DES (22%–31%).23–26 Improved outcomes associated with DES in non-ESRD patients cannot be automatically extrapolated to the ESRD population because of qualitative differences in morphological and physiological characteristics of coronary lesions in ESRD patients. Atherosclerotic plaques in ESRD patients are characterized by severe calcification and medial thickening,27 resulting in a higher proportion of angiographically complex lesions (type B2/C) predicting lower procedural success with PCI.28 Also, the rate of neointimal hyperplasia and late lumen loss is higher in ESRD patients, resulting in an almost 4-fold higher risk of in-stent restenosis and a corresponding higher rate of adverse cardiac outcomes, including mortality.28 Importantly, the pattern of restenosis is frequently diffuse (versus focal in non-ESRD patients), which may be associated with higher risk of repeat revascularization.26 Recent data from a multi-center study in Japan using sirolimus-eluting stents emphasize the more than 2-fold higher rates of target lesion revascularization and more than 5-fold greater risk of cardiac mortality after adjustment for baseline characteristics and procedural characteristics in hemodialysis versus non-dialysis patients.29 Hemodialysis has been consistently shown to be an independent risk factor for in-stent restenosis following placement of stents eluting sirolimus28 and paclitaxel,30 both first generation DES platforms. Finally, the pathophysiological factors underlying the process of accelerated atherosclerosis also predispose to stent thrombosis, a phenomenon of particular concern with DES,16,31 which could also result in higher rates of myocardial infarction and mortality.
The relatively higher short-term and long-term mortality of dialysis patients undergoing BMS placement relative to DES and CABG surgery are underscored in this study and need to be reconciled. This observation could reflect selection bias inherent in observational data; i.e., patients chosen to undergo BMS placement may have conditions (e.g., bleeding diathesis, non-cardiac surgery, etc.) that preclude DES placement and affect survival. An alternative explanation could relate to in-stent restenosis and resultant need for repeat revascularization with BMS (however, our data on repeat revascularization do not support this idea). In a study employing a strategy of routine coronary angiography to detect in-stent restenosis in a non-dialysis population, Schühlen et al32 showed that mortality rates in 4-year follow-up were higher in patients with in-stent restenosis than in patients without in-stent stenosis. As noted above, some data indicate in-stent restenosis in as many as 40% of BMS in dialysis patients developed in-stent restenosis, and importantly, the diffuse pattern of restenosis in this population likely portends higher risk of repeat revascularization.26 A recent study by Sakakibara et al33 from a single center in Japan provided encouraging data showing that everolimus-eluting stents markedly reduced rates of angiographic in-stent restenosis with associated modest reduction in major adverse cardiac events relative to sirolimus-eluting stents, likely related to a combination of reduced inflammation associated with the platform, thinner stent struts, and more rapid endothelialization. Possibly, newer generations of DES may change the panorama of coronary revascularization in dialysis patients in the future.
The data from this study should be interpreted in the context of important limitations. These observational data have the inherent potential for selection bias and unmeasured confounders, and should therefore be considered exploratory and hypothesis generating. To derive conclusions directly applicable to clinical practice, a prospective randomized study would be the ideal method of evaluating comparative outcomes of dialysis patients undergoing PCI with DES versus CABG surgery. Because our data are derived from an administrative database, prognostically relevant clinical information pertaining to angiographic characteristics, lesion complexity, and left ventricular ejection fraction are not available, as previously described. Similarly, data regarding use of adjunctive pharmacological therapy and complications thereof, particularly any associated risk of bleeding, as well as platform of DES used, are not available. Although the probability of repeat revascularization was studied, we were unable to discern target lesion, target vessel revascularization, or revascularization in different coronary distributions.
In summary, these data illustrate a recent shift in interventional practice patterns in US dialysis patients, with a decrease in DES use and a corresponding rise in BMS use since 2006. Additionally, these data suggest that although CABG with an IMG was associated with high in-hospital mortality rates (although lower than CABG without IMG), it was also associated with improved long-term survival in dialysis patients. DES use was associated with relatively lower in-hospital mortality but significantly higher probability of repeat revascularization in the future. DES may be a reasonable consideration in dialysis patients in whom an IMG (typically used to bypass the left anterior descending coronary artery vascular territory) is not an appropriate option in the revascularization strategy, or whose overall life expectancy is judged to be limited. For some patients, PCI with DES might be preferable, as the higher peri-operative mortality (and likely morbidity) of CABG might be judged to be an inferior choice, despite potentially superior long-term survival. The findings from this study support the recently popularized notion of adopting a “heart team” approach, i.e., deriving input from interventional cardiologists and cardiovascular surgeons to determine an individualized, optimal approach for coronary revascularization in dialysis patients.
Supplementary Material
Acknowledgments
The authors thank United States Renal Data System colleague Nan Booth, MSW, MPH, ELS, for manuscript editing.
Funding Sources: This study was performed as a deliverable under Contract No. HHSN267200715003C (National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health, Bethesda, Maryland). The interpretation and reporting of these data are the responsibility of the authors and in no way should be seen as an official policy or interpretation of the US government.
Footnotes
Conflict of Interest Disclosures: None
References
- 1.Daemen J, Boersma E, Flather M, Booth J, Stables R, Rodriguez A, Rodriguez-Granillo G, Hueb WA, Lemos PA, Serruys PW. Long-term safety and efficacy of percutaneous coronary intervention with stenting and coronary artery bypass surgery for multivessel coronary artery disease: a meta-analysis with 5-year patient-level data from the ARTS, ERACI-II, MASS-II, and SoS trials. Circulation. 2008;118:1146–1154. doi: 10.1161/CIRCULATIONAHA.107.752147. [DOI] [PubMed] [Google Scholar]
- 2.Bravata DM, Gienger AL, McDonald KM, Sundaram V, Perez MV, Varghese R, Kapoor JR, Ardehali R, Owens DK, Hlatky MA. Systematic review: the comparative effectiveness of percutaneous coronary interventions and coronary artery bypass graft surgery. Ann Intern Med. 2007;147:703–716. doi: 10.7326/0003-4819-147-10-200711200-00185. [DOI] [PubMed] [Google Scholar]
- 3.Stettler C, Wandel S, Allemann S, Kastrati A, Morice MC, Schömig A, Pfisterer ME, Stone GW, Leon MB, de Lezo JS, Goy JJ, Park SJ, Sabaté M, Suttorp MJ, Kelbaek H, Spaulding C, Menichelli M, Vermeersch P, Dirksen MT, Cervinka P, Petronio AS, Nordmann AJ, Diem P, Meier B, Zwahlen M, Reichenbach S, Trelle S, Windecker S, Jüni P. Outcomes associated with drug-eluting and bare-metal stents: a collaborative network meta-analysis. Lancet. 2007;370:937–948. doi: 10.1016/S0140-6736(07)61444-5. [DOI] [PubMed] [Google Scholar]
- 4.Serruys PW, Morice MC, Kappetein AP, Colombo A, Holmes DR, Mack MJ, Ståhle E, Feldman TE, van den Brand M, Bass EJ, Van Dyck N, Leadley K, Dawkins KD, Morh FW SYNTAX Investigators. Percutaneous coronary intervention versus coronary-artery bypass grafting for severe coronary artery disease. N Engl J Med. 2009;360:961–972. doi: 10.1056/NEJMoa0804626. [DOI] [PubMed] [Google Scholar]
- 5.U.S. Renal Data System. USRDS 2011 Annual Data Report: Atlas of Chronic Kidney Disease & End-Stage Renal Disease in the United States. 2011. Bethesda, MD: National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases; 2011. [Google Scholar]
- 6.Herzog CA, Ma JZ, Collins AJ. Comparative survival of dialysis patients in the United States after coronary angioplasty, coronary artery stenting, and coronary artery bypass surgery and impact of diabetes. Circulation. 2002;106:2207–2211. doi: 10.1161/01.cir.0000035248.71165.eb. [DOI] [PubMed] [Google Scholar]
- 7.Hemmelgarn BR, Southern D, Culleton BF, Mitchell LB, Knudtson ML, Ghali WA. Survival after coronary revascularization among patients with kidney disease. Circulation. 2004;110:1890–1895. doi: 10.1161/01.CIR.0000143629.55725.D9. [DOI] [PubMed] [Google Scholar]
- 8.Tsai TT, Messenger JC, Brennan JM, Patel UD, Dai D, Piana RN, Anstrom KJ, Eisenstein EL, Dokholyan RS, Peterson ED, Douglas PS. 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;58:1859–1869. doi: 10.1016/j.jacc.2011.06.056. [DOI] [PubMed] [Google Scholar]
- 9.Herzog CA, Ma JZ, Collins AJ. Long-term outcome of dialysis patients in the United States with coronary revascularization procedures. Kidney Int. 1999;56:324–332. doi: 10.1046/j.1523-1755.1999.00540.x. [DOI] [PubMed] [Google Scholar]
- 10.Abdel-Latif A, Mukherjee D, Mesgarzadeh P, Ziada KM. Drug-eluting stents in patients with end-stage renal disease: meta-analysis and systematic review of the literature. Catheter Cardiovasc Interv. 2010;76:942–948. doi: 10.1002/ccd.22562. [DOI] [PubMed] [Google Scholar]
- 11.Cooper WA, O’Brien SM, Thourani VH, Guyton RA, Bridges CR, Szczech LA, Petersen R, Petersen ED. Impact of renal dysfunction on outcomes of coronary artery bypass surgery: results from the Society of Thoracic Surgeons National Adult Cardiac Database. Circulation. 2006;113:1063–1070. doi: 10.1161/CIRCULATIONAHA.105.580084. [DOI] [PubMed] [Google Scholar]
- 12.Shroff GR, Li S, Herzog CA. Survival of patients on dialysis having off-pump versus on-pump coronary artery bypass surgery in the United States. J Thorac Cardiovasc Surg. 2010;139:1333–1338. doi: 10.1016/j.jtcvs.2009.08.021. [DOI] [PubMed] [Google Scholar]
- 13.Nevis IF, Mathew A, Novick RJ, Parikh CR, Devereaux PJ, Natarajan MK, Iansavichus AV, Cuerden MS, Garg AX. Optimal method of coronary revascularization in patients receiving dialysis: systematic review. Clin J Am Soc Nephrol. 2009;4:369–378. doi: 10.2215/CJN.02640608. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Gooley TA, Leisenring W, Crowley J, Storer BE. Estimation of failure probabilities in the presence of competing risks: new representations of old estimators. Stat Med. 1999;18:695–706. doi: 10.1002/(sici)1097-0258(19990330)18:6<695::aid-sim60>3.0.co;2-o. [DOI] [PubMed] [Google Scholar]
- 15.Riley RF, Don CW, Powell W, Maynard C, Dean LS. Trends in coronary revascularization in the United States from 2001 to 2009: recent declines in percutaneous coronary intervention volumes. Circ Cardiovasc Qual Outcomes. 2011;4:193–197. doi: 10.1161/CIRCOUTCOMES.110.958744. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Lagerqvist B, James SK, Stenestrand U, Lindback J, Nilsson T, Wallentin L. Long-term outcomes with drug-eluting stents versus bare-metal stents in Sweden. N Engl J Med. 2007;356:1009–1019. doi: 10.1056/NEJMoa067722. [DOI] [PubMed] [Google Scholar]
- 17.Curfman GD, Morrissey S, Jarcho JA, Drazen JM. Drug-eluting coronary stents--promise and uncertainty. N Engl J Med. 2007;356:1059–1060. doi: 10.1056/NEJMe068306. [DOI] [PubMed] [Google Scholar]
- 18.Maisel WH. Unanswered questions--drug-eluting stents and the risk of late thrombosis. N Engl J Med. 2007;356:981–984. doi: 10.1056/NEJMp068305. [DOI] [PubMed] [Google Scholar]
- 19.Manabe S, Shimokawa T, Fukui T, Fumimoto KU, Ozawa N, Seki H, Takanashi S. Coronary artery bypass surgery versus percutaneous coronary artery intervention in patients on chronic hemodialysis: does a drug-eluting stent have an impact on clinical outcome? J Card Surg. 2009;24:234–239. doi: 10.1111/j.1540-8191.2008.00789.x. [DOI] [PubMed] [Google Scholar]
- 20.Sunagawa G, Komiya T, Tamura N, Sakaguchi G, Kobayashi T, Murashita T. Coronary artery bypass surgery is superior to percutaneous coronary intervention with drug-eluting stents for patients with chronic renal failure on hemodialysis. Ann Thorac Surg. 2010;89:1896–1900. doi: 10.1016/j.athoracsur.2010.02.080. [DOI] [PubMed] [Google Scholar]
- 21.Ashrith G, Lee VV, Elayda MA, Reul RM, Wilson JM. Short- and long-term outcomes of coronary artery bypass grafting or drug-eluting stent implantation for multivessel coronary artery disease in patients with chronic kidney disease. Am J Cardiol. 2010;106:348–353. doi: 10.1016/j.amjcard.2010.03.037. [DOI] [PubMed] [Google Scholar]
- 22.Chang TI, Shilane D, Kazi DS, Montez-Rath ME, Hlatky MA, Winkelmayer WC. Multivessel coronary artery bypass grafting versus percutaneous coronary intervention in ESRD. J Am Soc Nephrol. 2012;23:2042–2049. doi: 10.1681/ASN.2012060554. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Ishio N, Kobayashi Y, Takebayashi H, Iijima Y, Kanda J, Nakayama T, Kuroda N, De Gregorio J, Kouno Y, Suzuki M, Haruta S, Komuro I. Impact of drug-eluting stents on clinical and angiographic outcomes in dialysis patients. Circ J. 2007;71:1525–1529. doi: 10.1253/circj.71.1525. [DOI] [PubMed] [Google Scholar]
- 24.Aoyama T, Ishii H, Toriyama T, Takahashi H, Kasuga H, Murakami R, Amano T, Uetani T, Yasuda Y, Yuzawa Y, Maruyama S, Matsuo S, Matsubara T, Murohara T. Sirolimus-eluting stents vs bare metal stents for coronary intervention in Japanese patients with renal failure on hemodialysis. Circ J. 2008;72:56–60. doi: 10.1253/circj.72.56. [DOI] [PubMed] [Google Scholar]
- 25.Ikari Y, Tanabe K, Koyama Y, Kozuma K, Sano K, Isshiki T, Katsuki T, Kimura K, Yamane M, Takahashi N, Hibi K, Hasegawa K, Ishiwata S, Kiyooka T, Yokoi H, Uehara Y, Hara K. Sirolimus eluting coronary stent implantation in patients on maintenance hemodialysis: the OUCH study (outcome of cypher stent inhemodialysis patients) Circ J. 2012;76:1856–1863. doi: 10.1253/circj.cj-12-0046. [DOI] [PubMed] [Google Scholar]
- 26.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;51:92–97. doi: 10.1536/ihj.51.92. [DOI] [PubMed] [Google Scholar]
- 27.Schwarz U, Buzello M, Ritz E, Stein G, Raabe G, Wiest G, Mall G, Amann K. Morphology of coronary atherosclerotic lesions in patients with end-stage renal failure. Nephrol Dial Transplant. 2000;15:218–223. doi: 10.1093/ndt/15.2.218. [DOI] [PubMed] [Google Scholar]
- 28.Ota T, Umeda H, Yokota S, Miyata S, Takamura A, Sugino S, Hayashi K, Ishiki R, Takeichi Y, Iwase M, Inagaki H, Murohara T. Relationship between severity of renal impairment and 2-year outcomes after sirolimus-eluting stent implantation. Am Heart J. 2009;158:92–98. doi: 10.1016/j.ahj.2009.04.013. [DOI] [PubMed] [Google Scholar]
- 29.Otsuka Y, Ishiwata S, Inada T, Kanno H, Kyo E, Hayashi Y, Fujita H, Michishita I. Comparison of haemodialysis patients and non-haemodialysis patients with respect to clinical characteristics and 3-year clinical outcomes after sirolimus-eluting stent implantation: insights from the Japan multi-centre post-marketing surveillance registry. Eur Heart J. 2011;32:829–837. doi: 10.1093/eurheartj/ehq480. [DOI] [PubMed] [Google Scholar]
- 30.Takeuchi I, Moriguchi M, Imaki R, Fukaya H, Shinagawa H, Shimohama T, Tojo T, Fukuda N, Inomata T, Aoyama N, Soma K, Izumi T. Hemodialysis is an independent predictor of coronary in-stent restenosis after paclitaxel eluting stent implantation. Intern Med. 2010;49:2379–2384. doi: 10.2169/internalmedicine.49.3798. [DOI] [PubMed] [Google Scholar]
- 31.Palmerini T, Biondi-Zoccai G, Della Riva D, Stettler C, Sangiorgi D, D’Ascenzo F, Kimura T, Briguori C, Sabaté M, Kim HS, De Waha A, Kedhi E, Smits PC, Kaiser C, Sardella G, Marullo A, Kirtane AJ, Leon MB, Stone GW. Stent thrombosis with drug-eluting and bare-metal stents: evidence from a comprehensive network meta-analysis. Lancet. 2012;379:1393–1402. doi: 10.1016/S0140-6736(12)60324-9. [DOI] [PubMed] [Google Scholar]
- 32.Schühlen H, Kastrati A, Mehilli J, Hausleiter J, Pache J, Dirschinger J, Schömig A. Restenosis detected by routine angiographic follow-up and late mortality after coronary stent placement. Am Heart J. 2004;147:317–322. doi: 10.1016/j.ahj.2003.10.002. [DOI] [PubMed] [Google Scholar]
- 33.Sakakibara T, Ishii H, Toriyama T, Aoyama T, Takahashi H, Kamoi D, Kawamura Y, Kawashima K, Yoneda K, Amano T, Tanaka M, Yoshikawa D, Hayashi M, Matsubara T, Murohara T. Sirolimus-eluting stent vs. everolimus-eluting stent for coronary intervention in patients on chronic hemodialysis. Circ J. 2012;76:351–355. doi: 10.1253/circj.cj-11-0814. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.