Impact of acute coronary syndromes on survival of dialysis patients following surgical or percutaneous coronary revascularization in the United States

Abstract

Aims:

We sought to evaluate survival of dialysis (chronic kidney disease (CKD) stage 5D) patients undergoing coronary revascularization procedures in the context of acute coronary syndrome (ACS) compared with absence of ACS.

Methods and results:

CKD 5D patients undergoing coronary revascularization, 2004–2009 (n = 23,033), were identified from the United States Renal Data System. Long-term survival was estimated by the Kaplan–Meier method and independent predictors of mortality using a comorbidity-adjusted Cox proportional hazards model. Among ACS patients (n = 12,473; 54%), revascularization procedures were coronary artery bypass grafting (CABG, n = 2910), drug-eluting stents (DESs, n = 6566), and bare metal stents (BMSs, n = 2997). All-cause survival rates following these procedures, respectively, were: in-hospital 90%, 96%, 93%; one-year: 66%, 67%, 58%; two-year: 53%, 48%, 43%. Among non-ACS patients (n = 10,560; 46%), procedures were CABG (n = 3268), DESs (n = 5278), and BMSs (n = 2014). Survival rates following these procedures, respectively, were: in-hospital 94%, 99%, 98%; one year: 73%, 77%, 70%; two year: 61%, 59%, 55%. DESs (versus CABG) independently predicted mortality among ACS (hazard ratio 1.08; 95% confidence interval 1.02–1.15) but not non-ACS patients (1.01, 0.95–1.07); BMSs (versus CABG) independently predicted mortality among ACS (1.30, 1.21–1.38) and non-ACS (1.13, 1.05–1.22) patients.

Conclusions:

Among CKD 5D patients, survival was lower for ACS versus non-ACS indications following all revascularization strategies. CABG (versus DESs) was associated with higher long-term survival in the context of ACS; in the absence of ACS, long-term survival was similar after CABG or DESs. BMSs were consistently associated with worse outcomes.

Keywords

Chronic kidney disease coronary artery bypass surgery coronary disease percutaneous coronary intervention stents

Introduction

Decisions regarding coronary revascularization among patients with end-stage renal disease (ESRD) are extremely complex; reported mortality rates following revascularization are among the highest for this population. One important factor contributing to this complexity is lack of randomized data in this high-risk group. Contemporary observational data have attempted to define the optimal revascularization strategy for these patients,^1–4 highlighting the clinical tradeoffs in the choice of revascularization modalities: higher in-hospital mortality but improved long-term survival following coronary artery bypass grafting (CABG) surgery (especially using internal mammary grafts) versus lower in-hospital mortality but higher risk of future revascularization after percutaneous coronary intervention (PCI) using drug-eluting stents (DESs). These observational studies, however, have not accounted for the impact of the acuity of the clinical presentation on long-term survival. Distinguishing long-term outcomes of dialysis (chronic kidney disease (CKD) stage 5D) patients based on the acuity of the clinical context in which revascularization occurs is extremely pertinent to the clinician’s choice of revascularization modality. Several factors contribute to important differences in revascularization outcomes among acute coronary syndrome (ACS) and non-ACS patients. Patients with ACS are predisposed to higher mortality rates than patients with stable coronary artery disease (CAD) owing to the acute thrombotic milieu and higher rates of arrhythmic, mechanical, and bleeding complications. Patients with ACS are also more likely to undergo PCI in routine clinical practice, skewing mortality hazards against PCI.

Patients with advanced CKD differ in important ways from the general population. They are significantly more likely to have an index presentation with acute myocardial infarction (AMI) than with stable angina.⁵ Further, mortality rates following AMI among CKD 5D patients are extremely high,^6,7 an observation corroborated by large registries in the contemporary era.^8,9 The high mortality rates are undoubtedly multifactorial, but the impact of coronary revascularization in this context is not well studied. Using the Acute Coronary Treatment and Intervention Outcomes Network (ACTION) registry, Fox and colleagues described eight-fold higher odds of mortality following ST-segment elevation myocardial infarction (STEMI), and four-fold higher odds following non-STEMI (NSTEMI), among patients with stage 5 CKD (including stage 5D) compared with non-CKD patients.⁸ Paradoxically, patients with stage 5 CKD were also least likely to undergo reperfusion for STEMI or revascularization for NSTEMI or to receive appropriate evidence-based medication therapy. Using a large Swedish registry, Szummer and colleagues proposed that in these patients, the survival benefit of early invasive therapy was diminished relative to non-CKD patients in the context of NSTEMI.⁹ Preliminary reports suggest that survival following AMI has modestly improved among CKD 5D patients in the past two decades, apparent only among STEMI and not among NSTEMI patients,¹⁰ and is associated with higher rates of early coronary reperfusion therapies, particularly PCI.¹¹ Mortality rates of ESRD patients in the context of stable CAD have not been well described.

Using data from the United States Renal Data System (USRDS), we sought to evaluate short-term and long-term survival of CKD 5D patients undergoing coronary revascularization procedures in the context of ACS compared with absence of ACS.

Methods

Using USRDS data, we identified CKD 5D patients hospitalized for first coronary revascularization procedure, 2004–2009 (n = 23,033 of 546,160 patients). Patients were eligible for the study only if they had been receiving renal replacement therapy for at least 90 days before the procedure. In the surgical arm, only patients undergoing CABG in the absence of any concomitant valve surgery were included. Patients undergoing both surgical and percutaneous revascularization in the same hospitalization were excluded. Patients undergoing PCI with both DESs and bare metal stents (BMSs) during the same hospitalization were included in the DES group only. Patients undergoing percutaneous transluminal coronary angioplasty alone were excluded.

Using these criteria, 6178 CABG, 11,844 DES, and 5011 BMS patients were identified. Subsequently, patients undergoing revascularization for ACS (n = 12,473, 54.2%) were identified using International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) codes 410.x (excluding 410.x2), 411.1, 411.81, and 411.89: CABG, 2910; DES, 6566; BMS, 2997 patients. All other patients were considered non-ACS (n = 10,560; CABG, 3268; DES, 5278; BMS, 2014). Long-term survival was estimated by the Kaplan–Meier method and independent predictors of mortality using a comorbidity-adjusted Cox-proportional hazards model. Survival was determined from revascularization to death or censoring. All patients were followed up to 31 December 2010. Censoring was performed for patients undergoing renal transplant or lost to follow-up before 31 December 2010. Death was identified from the USRDS database. Comorbid conditions studied in the Cox model included prior myocardial infarction, congestive heart failure, and other cardiac conditions, as well as chronic conditions including non-skin malignancies, peripheral vascular disease, stroke, gastrointestinal disease, and liver disease, as previously described.⁴ The chi square test was used to detect differences between proportions. All reported p values are two-sided. Statistical analyses were performed using the SAS system for Windows, version 9.2 (SAS Institute, Inc.).

Results

Trends in Revascularization procedures among CKD 5D patients

Figure 1 outlines the trends in coronary revascularization procedures in the USA, 2004–2009, among CKD 5D patients. As previously shown,⁴ numbers of DES procedures declined substantially in the USA in 2006 and 2007, with a corresponding increase in numbers of BMS procedures. The total number of revascularization procedures declined only modestly for ACS patients (Figure 1(a)) but much more substantially for non-ACS patients (Figure 1(b)). Interestingly, the proportion of patients undergoing revascularization for non-ACS indications (both percutaneous and surgical) declined during the middle of the study period, with corresponding relative increases in the proportions undergoing revascularization for ACS indications.

CKD 5D patients with ACS

Among ACS patients (n = 12,473, 54%), 8258 procedures (66%) occurred in the context of AMI (ICD-9-CM codes 410.x excluding 410.x2) and 4215 (34%) in the context of unstable angina (ICD-9-CM code 411). Demographic characteristics of ACS patients are shown in Table 1. A higher proportion of patients aged older than 75 years underwent PCI with DES versus CABG (33.6% vs. 23.4%). Higher proportions of men underwent revascularization with CABG versus PCI with DES (60.5% vs. 52.3%). White race (about 65%) and hemodialysis as the dialysis modality (93%−94%) were predominant. Congestive heart failure (CABG, 69.4%; DES, 66.6%; BMS, 64.2%) and diabetes (CABG, 74.5%; DES, 75.7%; BMS, 72.3%) were the predominant comorbid conditions.

Table 1.

Baseline demographic characteristics of chronic kidney disease stage 5D patients with acute coronary syndrome undergoing coronary revascularization procedures in the United States, 2004–2009.

Total, n	Coronary revascularization procedure
	CABG		BMS		DES
	2910		2997		6566
Age, years, mean (SD)	66.04 (10.7)		68.41 (12.04)		67.78 (11.94)
	n	%	n	%	n	%
Age, years
20–44	89	3.1	98	3.3	258	3.9
45–64	1120	38.5	931	31.1	2120	32.3
65–74	1019	35.0	920	30.7	1984	30.2
⩾ 75	682	23.4	1048	35.0	2204	33.6
Sex
Men	1762	60.5	1690	56.4	3435	52.3
Women	1148	39.5	1307	43.6	3131	47.7
Race
White	1888	64.9	1990	66.4	4273	65.1
Black	841	28.9	852	28.4	1854	28.2
Other	181	6.2	155	5.2	439	6.7
Dialysis modality
Hemodialysis	2716	93.3	2808	93.7	6169	94.0
Peritoneal dialysis	194	6.7	189	6.3	397	6.0
CHF	2019	69.4	1925	64.2	4371	66.6
COPD	854	29.3	855	28.5	1888	28.8
CVA	839	28.8	702	23.4	1558	23.7
Cancer	208	7.1	276	9.2	563	8.6
Diabetes	2169	74.5	2167	72.3	4969	75.7
Dysrhythmia	1451	49.9	1432	47.8	2929	44.6
GI bleeding	343	11.8	443	14.8	809	12.3
Liver disease	292	10.0	303	10.1	637	9.7
Other cardiac	1615	55.5	1536	51.3	3357	51.1
PAD	1393	47.9	1402	46.8	3159	48.1

BMS: bare metal stent; CABG: coronary artery bypass graft; CHF: congestive heart failure; COPD: chronic obstructive pulmonary disease; CVA: cerebrovascular accident; DES: drug-eluting stent; GI: gastrointestinal; PAD: peripheral artery disease; SD: standard deviation

Figure 2 (panels (a)–(c)) shows Kaplan–Meier survival estimates for each revascularization modality for ACS patients. In-hospital mortality rates were highest for CABG (10.4%), intermediate for BMS (6.6%) and lowest for DES (4.0%). All-cause survival for ACS patients undergoing CABG was 66% at one year and 53% at two years; corresponding proportions for DES patients were 67% and 48%, and for BMS patients, 58% and 43%. One-year and two-year survival rates for ACS patients who underwent CABG with an internal mammary graft (IMG; n = 2262, 77.7%) and without, respectively, were 68% vs. 57.3% and 55.1% vs. 43.9%. Using a Cox proportional hazards model (Table 2), independent predictors of mortality among ACS patients were: ages 65–74 years (hazard ratio (HR) 1.30, 95% confidence interval (CI) 1.23–1.37) and older than 75 years (1.77, 1.67–1.87), dialysis duration 6–10 years (1.33, 1.24–1.42) and longer than 11 years (1.50, 1.32–1.71), peritoneal dialysis (1.38, 1.25–1.52), and congestive heart failure (1.26, 1.20–1.33). DES use (versus CABG) was independently associated with an increased hazard for mortality (HR 1.08, 95% CI 1.02–1.15). BMS use was also an independent predictor of mortality compared with CABG (HR 1.30, 95% CI 1.21–1.38) and DES (1.20, 1.13–1.26).

Table 2.

Cox-proportional hazards model evaluating independent predictors of mortality after coronary revascularization for patients with acute coronary syndrome, 2004–2009.

Variable	HR (95% CI)	p
DES (versus CABG)	1.08 (1.02–1.15)	0.0058
BMS (versus CABG)	1.30 (1.21–1.38)	< 0.0001
BMS (versus DES)	1.20 (1.13–1.26)	< 0.0001
Women	0.99 (0.94–1.03)	0.5001
Black race	0.83 (0.79–0.88)	< 0.0001
Other/unknown race	0.83 (0.76–0.92)	0.0002
Age, years
65–74	1.30 (1.23–1.37)	< 0.0001
⩾ 75	1.77 (1.67–1.87)	< 0.0001
Diabetes	1.31 (1.21–1.41)	< 0.0001
Hypertension	1.14 (1.06–1.23)	0.0002
Dialysis duration, years
2–5	1.22 (1.16–1.28)	< 0.0001
6–10	1.33 (1.24–1.42)	< 0.0001
⩾ 11	1.50 (1.32–1.71)	< 0.0001
Peritoneal dialysis	1.38 (1.25–1.52)	< 0.0001
CHF	1.26 (1.20–1.33)	< 0.0001
COPD	1.20 (1.14–1.25)	< 0.0001
CVA	1.13 (1.07–1.18)	< 0.0001
Cancer	1.07 (0.99–1.16)	0.0846
Diabetes	1.07 (1.01–1.15)	0.0359
Dysrhythmia	1.30 (1.24–1.36)	< 0.0001
GI bleeding	1.05 (0.99–1.12)	0.1107
Liver disease	1.10 (1.02–1.18)	0.0126
Other cardiac disease	1.15 (1.10–1.21)	< 0.0001
PVD	1.23 (1.17–1.29)	< 0.0001

BMS: bare metal stent; CABG: coronary artery bypass graft; CHF: congestive heart failure; CI: confidence interval; COPD: chronic obstructive pulmonary disease; CVA: cerebrovascular accident; DES: drug-eluting stent; GI: gastrointestinal; HR: hazard ratio; PVD: peripheral vascular disease

CKD 5D patients without ACS

Patterns of demographic characteristics for non-ACS patients (Table 3; n = 10,560, 46%), were similar to patterns for ACS patients. PCI with DES was more frequent than CABG among non-ACS patients aged older than 75 years (24.8% vs. 18.5%). Higher proportions of men underwent revascularization with CABG than with PCI with DES (63.8% vs. 57.2%). White race (63%) and hemodialysis (94%−95%) were predominant. Proportions of patients with diabetes were similar to proportions in the ACS group, but congestive heart failure was less frequent (CABG, 60.2%; DES, 59.6%; BMS, 61.4%).

Table 3.

Baseline demographic characteristics of chronic kidney disease stage 5D patients without acute coronary syndrome undergoing coronary revascularization procedures in the United States, 2004–2009.

Total, n	Coronary revascularization procedure
	CABG		BMS		DES
	3268		2014		5278
Age, years, mean (SD)	64.09 (11.2)		66.55 (11.9)		65.72 (11.5)
	n	%	n	%	n	%
Age, years
20–44	179	5.5	93	4.6	235	4.5
45–64	1379	42.2	682	33.9	2014	38.2
65–74	1105	33.8	682	33.9	1722	32.6
⩾ 75	605	18.5	557	27.7	1307	24.8
Sex
Men	2085	63.8	1201	59.6	3020	57.2
Women	1183	36.2	813	40.4	2258	42.8
Race
White	2069	63.3	1293	64.2	3328	63.1
Black	942	28.8	612	30.4	1602	30.4
Other	257	7.9	109	5.4	348	6.6
Dialysis modality
Hemodialysis	3057	93.5	1913	95.0	4987	94.5
Peritoneal dialysis	211	6.5	101	5.0	291	5.5
CHF	1968	60.2	1236	61.4	3147	59.6
COPD	819	25.1	608	30.2	1338	25.4
CVA	858	26.3	454	22.5	1185	22.5
Cancer	215	6.6	208	10.3	371	7.0
Diabetes	2481	75.9	1485	73.7	4161	78.8
Dysrhythmia	1370	41.9	854	42.4	2062	39.1
GI bleeding	380	11.6	265	13.2	590	11.2
Liver disease	321	9.8	214	10.6	513	9.7
Other cardiac	1740	53.2	1034	51.3	2600	49.3
PAD	1514	46.3	957	47.5	2436	46.2

Figure 3 (panels (a)–(c)) shows Kaplan–Meier survival estimates for each revascularization modality for non-ACS patients. In-hospital mortality rates were considerably lower than in the ACS group, but were highest among CABG patients (6.2% vs. 1.2% DES and 2.4% BMS). All-cause survival for non-ACS CABG patients was 73% at one year and 61% at two years; corresponding rates for DES patients were 77% and 59% and for BMS patients 70% and 55%. One- and two-year survival rates for non-ACS patients who underwent CABG with an IMG (n = 2259, 69%) and without, respectively, were 75.5% vs. 67.8% and 64.1% vs. 54.9%. Using Cox proportional hazards model (Table 4), independent predictors of mortality among ACS patients were: age 65–74 years (HR 1.31, 95% CI 1.23–1.39) and older than 75 years (1.78, 1.66–1.90), dialysis duration 6–10 years (1.34, 1.24–1.44) and longer than 11 years (1.21, 1.04–1.41), congestive heart failure (1.45, 1.36-–1.54), and chronic obstructive pulmonary disease (1.30, 1.22–1.37). DES use (versus CABG) was not independently associated with higher mortality (HR 1.01, 95% CI 0.95–1.07). However, BMS use was consistently associated with an independent hazard of mortality compared with CABG (HR 1.13, 95% CI 1.05–1.22) and DES (1.12, 1.05–1.20).

Table 4.

Cox-proportional hazards model evaluating independent predictors of mortality after coronary revascularization for patients without acute coronary syndrome, 2004–2009.

Variable	HR (95% CI)	p
DES (versus CABG)	1.01 (0.95–1.07)	0.8616
BMS (versus CABG)	1.13 (1.05–1.22)	0.0018
BMS (versus DES)	1.12 (1.05–1.20)	0.0010
Women	0.98 (0.93–1.03)	0.3658
Black race	0.93 (0.88–0.99)	0.0142
Other/unknown race	0.80 (0.71–0.90)	0.0001
Age, years
65–74	1.31 (1.23–1.39)	< 0.0001
⩾ 75	1.78 (1.66–1.90)	< 0.0001
Diabetes	1.13 (1.03–1.23)	0.0094
Hypertension	1.05 (0.97–1.15)	0.2469
Dialysis duration, years
2–5	1.18 (1.11–1.25)	< 0.0001
6–10	1.34 (1.24–1.44)	< 0.0001
⩾ 11	1.21 (1.04–1.41)	0.0127
Peritoneal dialysis	1.13 (0.99–1.29)	0.0800
CHF	1.45 (1.36–1.54)	< 0.0001
COPD	1.30 (1.22–1.37)	< 0.0001
CVA	1.17 (1.11–1.25)	< 0.0001
Cancer	1.14 (1.04–1.25)	0.0065
Diabetes	1.09 (1.01–1.18)	0.0343
Dysrhythmia	1.32 (1.25–1.39)	< 0.0001
GI bleeding	1.17 (1.09–1.26)	< 0.0001
Liver disease	1.11 (1.03–1.21)	0.0100
Other cardiac disease	1.22 (1.15–1.29)	< 0.0001
PVD	1.20 (1.14–1.27)	< 0.0001

It is noteworthy that for all intervals reported, survival for non-ACS patients was significantly higher than for ACS counterparts (p < 0.0001). Similarly, survival was significantly higher with than without IMG use among CABG patients (log-rank p < 0.0001 for both ACS and non-ACS patients).

Discussion

To the best of our knowledge, this is the first study to factor in the impact of the acuity of the clinical presentation on survival of CKD 5D patients following coronary revascularization. Some important observations are apparent. Survival following revascularization in the context of ACS was consistently lower than in the non-ACS setting across all revascularization modalities. Survival rates following surgical and percutaneous revascularization with DES were similar for stable CAD (non-ACS) patients. However, in the context of unstable CAD (ACS), PCI with DES was associated with a higher long-term mortality hazard relative to CABG. Revascularization following PCI with BMS was consistently associated with poor long-term survival regardless of the acuity of the clinical presentation. Also, a higher proportion of CKD 5D patients undergo revascularization for ACS (54%) versus non-ACS indications, unlike in the general population.

These findings urge a re-examination of best revascularization strategy following ACS among CKD 5D patients, who probably constitute one of the highest-risk patient groups undergoing revascularization in the contemporary era. Higher long-term survival after CABG in CKD 5D patients after ACS aligns thematically with reports from randomized controlled trials demonstrating that CABG is associated with a survival advantage in other high-risk subsets (e.g., diabetic patients with multivessel disease). The reasons underlying this observation, although speculative, could be related to difficulties associated with PCI in this unique population. CKD 5D patients exhibit multiple pathobiological differences in coronary involvement, including the occurrence of more proximal coronary thrombosis in AMI^12,13 and substantially higher rates of coronary calcification and complex lesions,¹⁴ impacting periprocedural success and complication rates following PCI. Interestingly, as we have previously shown, the probability of repeat revascularization is similar following DES or BMS placement among CKD 5D patients at all time intervals following PCI (unlike in the general population) and significantly higher compared with CABG,⁴ probably also impacting long-term outcomes.

CKD 5D patients also sustain among the highest bleeding rates of all patients with AMI,⁸ likely offsetting the “net” clinical benefit following PCI, and, importantly, also representing an independent predictor of mortality.¹⁵ Other potentially interesting hypotheses are plausible, and may deserve future prospective evaluation. Among CKD 5D patients with multivessel and complex CAD, more than one vulnerable plaque or co-culprit lesion may exist in the thrombotic milieu of ACS,¹⁶ such that completeness of revascularization with CABG may offer a long-term survival advantage and lead to the desirable corollary of obviating need for obligatory long-term dual antiplatelet therapy, in turn reducing bleeding rates. The diminished therapeutic efficacy of medical therapies that typically accompany PCI, for example, statins, may also play a contributory role.

An important caveat pertains to the above discussion: although long-term survival rates are higher following CABG than following DES among CKD 5D patients with ACS, short-term mortality rates are often extremely important considerations in complex revascularization decisions. In that regard, these data indicate that in-hospital mortality was significantly lower in the DES group relative to the CABG group in the context of ACS. In fact, one-year survival rates are also comparable in this extremely high-risk group of patients; differences in survival become apparent after 18–24 months following the index revascularization procedure. Given the high baseline mortality rates of CKD 5D patients with ACS, PCI with DES may be a reasonable consideration in appropriately selected clinical scenarios.

Coronary revascularization in CKD 5D patients in the absence of ACS constituted 46% of revascularization procedures. Long-term outcomes were comparable for these patients with stable CAD following PCI with DES versus CABG in this observational study (HR 1.01; p = 0.86), unlike for ACS patients. The lower number of revascularization procedures for stable CAD among CKD 5D patients relative to the general population supports findings reported by Go et al.,⁵ suggesting that a presentation with stable angina is significantly less likely than AMI among CKD 5D patients. Many coronary revascularization procedures in the absence of ACS likely occur in the context of renal transplant evaluation, possibly even in the absence of a clinical syndrome attributable to CAD. In this regard, the CKD 5D population is unique because of aggressive CAD screening performed during the renal transplant work-up and the institutional preference of several transplant centers to proceed directly with coronary angiography without preceding ischemia evaluation,¹⁷ possibly contributing to higher revascularization rates. CKD 5D patients with stable CAD, that is, absence of cardiac biomarker elevation, represent a much lower-risk population, since any biomarker elevation portends worse survival.^18,19

CKD 5D patients with stable CAD (absence of ACS) should be clinically identified as a distinct subgroup, since long-term survival (Figures 2 and 3) is consistently higher than in ACS patients regardless of revascularization modality and at all intervals following coronary revascularization. Also, in these patients, in-hospital mortality following revascularization was much lower than in ACS patients, and long-term survival was comparable following PCI with DES and CABG. These findings suggest that CKD 5D patients in future studies must not all be grouped in the same analytic category for assessment of long-term outcomes. This approach in the current study therefore represents a departure from that taken by multiple past studies,^1–4,20,21 and helps to identify a group of patients for whom clinicians may feel equally comfortable recommending either surgical or percutaneous revascularization based on coronary anatomy. Long-term outcomes in the important subset of ESRD patients with stable CAD undergoing no revascularization (medical therapy only) would be extremely pertinent to this discussion. These patients should be a target of future studies, but unfortunately could not be evaluated in the current study due to the study design.

Mortality rates are consistently higher for both the ACS and non-ACS patients undergoing PCI with BMS relative to other revascularization modalities throughout the follow-up period; this increase in mortality cannot be explained simply on the basis of higher repeat revascularization rates.⁴ Interestingly, rates of PCI with BMS increased markedly in the USA in 2006–2007, coincident with a sharp decline in rates of PCI with DES (Figure 1). Clinicians are urged to recognize and reconcile the consistent association of BMS with worse survival rates compared with PCI with DES or CABG in decision-making regarding revascularization of CKD 5D patients. We believe that patients undergoing PCI with BMS represent an inherently unique patient population in whom revascularization using other modalities is clinically precluded (higher risk of bleeding or ongoing bleeding diathesis, significant comorbid conditions, upcoming major surgery, etc.) Thus, poor long-term outcomes in this cohort likely reflect substantial selection bias related to choice of stent in much higher-risk patients, who are likely not candidates for either PCI with DES or CABG.

Despite this study’s strengths, several limitations should be acknowledged. Specifically, the contribution of significant selection biases and unmeasured confounders inherent in the choice of revascularization strategy cannot be eliminated with an observational study design. Clinical variables that routinely impact decisions pertaining to revascularization cannot be identified with these administrative data, particularly left ventricular ejection fraction, details and complexity of coronary anatomy, etc. Importantly, the possibility of selection bias contributing to the choice of revascularization modality between CABG versus PCI with DES in the context of ACS cannot be discerned except by performing a randomized controlled trial. Although higher-risk patients may have been turned down for surgical intervention (and therefore underwent high-risk PCI procedures, thus adversely impacting outcomes), it is also likely in clinical practice that higher-risk patients with more extensive and proximal CAD burden preferentially undergo surgical revascularization.

In conclusion, this observational study reports long-term survival based on the acuity of the clinical presentation or stability of CAD. This represents a paradigm shift in the approach to the study of revascularization outcomes not previously described in this high-risk population, and helps demarcate two distinct patient groups. Survival was consistently worse for CKD 5D patients with ACS than for those without ACS at all time intervals following revascularization; surgical revascularization was associated with improved long-term outcomes relative to PCI with DES, albeit with higher in-hospital and comparable one-year mortality rates. Among CKD 5D patients without ACS (a relatively lower-risk group) long-term survival was similar with CABG or PCI with DES. PCI with BMS was consistently associated with significantly worse long-term survival relative to other modalities and may represent an inferior choice as a stand-alone revascularization strategy if other options are feasible. This study provides additional and more granular information to guide clinicians in making appropriate individualized revascularization decisions for their patients. We continue to advocate for a “heart-team” approach due to the complexity of decision making regarding revascularization in this extremely high-risk patient group.

Footnotes

Acknowledgements

The authors thank USRDS colleagues Delaney Berrini, BS, for manuscript preparation, and Nan Booth, MSW, MPH, ELS, for manuscript editing. This study was performed by the Cardiovascular Special Studies Center of the United States Renal Data System. 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.

Conflict of interest

The authors declare that there is no conflict of interest.

Funding

This work was supported by the National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health, Bethesda, Maryland (Contract No. HHSN267200715003C).

References

Herzog

Collins

. Long-term outcome of dialysis patients in the United States with coronary revascularization procedures. Kidney Int 1999; 56: 324–332.

Chang

Shilane

Kazi

. Multivessel coronary artery bypass grafting versus percutaneous coronary intervention in ESRD. J Am Soc Nephrol 2012; 23: 2042–2049.

Herzog

Collins

. 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.

Shroff

Solid

Herzog

. Long-term survival and repeat coronary revascularization in dialysis patients after surgical and percutaneous coronary revascularization with drug-eluting and bare metal stents in the United States. Circulation 2013; 127: 1861–1869.

Bansal

Chandra

.; ADVANCE Study Investigators. Chronic kidney disease and risk for presenting with acute myocardial infarction versus stable exertional angina in adults with coronary heart disease. J Am Coll Cardiol 2011; 58: 1600–1607.

Herzog

Collins

. Poor long-term survival after acute myocardial infarction among patients on long-term dialysis. N Engl J Med 1998; 339: 799–805.

Iseki

Fukiyama

. Long-term prognosis and incidence of acute myocardial infarction in patients on chronic hemodialysis. Am J Kidney Dis 2000; 36: 820–825.

Fox

Muntner

Chen

.; Acute Coronary Treatment and Intervention Outcomes Network registry. Use of evidence-based therapies in short-term outcomes of ST-segment elevation myocardial infarction and non-ST-segment elevation myocardial infarction in patients with chronic kidney disease: A report from the National Cardiovascular Data Acute Coronary Treatment and Intervention Outcomes Network registry. Circulation 2010; 121: 357–365.

Szummer

Lundman

Jacobson

.; SWEDE-HEART. Influence of renal function on the effects of early revascularization in non-ST-elevation myocardial infarction: Data from the Swedish Web-System for Enhancement and Development of Evidence-Based Care in Heart Disease Evaluated According to Recommended Therapies (SWEDEHEART). Circulation 2009; 120: 851–858.

10.

Shroff

Herzog

. Abstract 11138: Survival of US dialysis patients after acute myocardial infarction: Is it finally getting better in the modern treatment era? Circulation 2012; 126(21 Suppl.): A11138.

11.

Shroff

Herzog

. Abstract 10621: Trends of early coronary revascularization with acute myocardial infarction among patients on long-term dialysis over 15 years in the United States. Circulation 2013; 128(22 Suppl.): A10621.

12.

Charytan

Kuntz

Garshick

. Location of acute coronary artery thromboses in patients with and without chronic kidney disease. Kidney Int 2009; 75: 80–87.

13.

Hanna

Chen

Roe

. Characteristics and in-hospital outcomes of patients with non-ST-segment elevation myocardial infarction and chronic kidney disease undergoing percutaneous coronary intervention. JACC: Cardiovasc Interv 2011; 4: 1002–1008.

14.

Schwarz

Buzello

Ritz

. Morphology of coronary atherosclerotic lesions in patients with end-stage renal failure. Nephrol Dial Transplant 2000; 15: 218–223.

15.

Mehran

Pocock

Nikolsky

. A risk score to predict bleeding in patients with acute coronary syndromes. J Am Coll Cardiol 2010; 55: 2556–2566.

16.

Wald

Morris

Wald

.; PRAMI Investigators. Randomized trial of preventive angioplasty in myocardial infarction. N Engl J Med 2013; 369: 1115–1123.

17.

De Lima

Sabbaga

Vieira

. Coronary angiography is the best predictor of events in renal transplant candidates compared with noninvasive testing. Hypertension 2003; 42: 263–268.

18.

Apple

Murakami

Pearce

. Predictive value of cardiac troponin I and T for subsequent death in end-stage renal disease. Circulation 2002; 106: 2941–2945.

19.

Khan

Hemmelgarn

Tonelli

. Prognostic value of troponin T and I among asymptomatic patients with end-stage renal disease: A meta-analysis. Circulation 2005; 112: 3088–3096.

20.

Hemmelgarn

Southern

Culleton

. Survival after coronary revascularization among patients with kidney disease. Circulation 2004; 110: 1890–1895.

21.

Nevis

Mathew

Novick

. Optimal method of coronary revascularization in patients receiving dialysis: Systematic review. Clin J Am Soc Nephrol 2009; 4: 369–378.