The aim was to examine timing, causes, and predictors of death during long-term follow-up after contemporary percutaneous coronary intervention (PCI) using a large multicenter Australian registry. The cohort consisted of 10,682 consecutive patients from the Melbourne Interventional Group registry undergoing PCI (February 2004 through November 2009). For the first time in Australia, long-term mortality rates of a PCI cohort were defined by linkage to the National Death Index database. The cohort (mean age 64 ± 12 years) comprised 75% men, 24% diabetics, 59% with multivessel disease, 4.4% with renal failure, 25% with ST-elevation myocardial infarction (STEMI), 2.5% with cardiogenic shock, and 5.1% with heart failure. Drug-eluting stents (DES) were used in 43% of cases. Mean follow-up was 3.2 ± 0.5 years. In-hospital, 30-day, 12-month, and long-term (3.2 ± 0.5 years) mortalities were 1.6% (80% cardiac), 2.1% (79%), 3.9% (61%), and 8.2% (50%), respectively. Independent predictors of long-term mortality included age (hazard ratio 1.05, 95% confidence interval 1.04 to 1.06), cardiogenic shock (4.58, 3.60 to 5.83), renal failure (3.14, 2.58 to 3.82), previous heart failure (1.97, 1.60 to 2.41), STEMI (1.79, 1.47 to 2.18), peripheral vascular disease (1.72, 1.4 to 2.11), non-STEMI (1.58, 1.32 to 1.90), multivessel disease (1.47, 1.24 to 1.74), current smoking (1.39, 1.12 to 1.71), diabetes (1.36, 1.16 to 1.59), and cerebrovascular disease (1.33, 1.06 to 1.60, p <0.01 for all comparisons). DES deployment appeared protective against late mortality (hazard ratio 0.85, 0.73 to 0.99, p = 0.04); however, after 30 days, there was no difference in mortality rates between those who received a bare metal stent and those who received a DES. In conclusion, different clinical variables such as renal and heart failure predicted long-term mortality after PCI, whereas DES use in this large registry was not associated with late mortality risk.
Predictors of mortality after percutaneous coronary intervention (PCI) at 30-day and 1-year follow-up are well established, and various scoring models exist to define risk after PCI. Fewer data exist for long-term outcomes, which is pertinent because of potential concerns of late thrombosis in an era of widespread drug-eluting stent (DES) use. The Melbourne Interventional Group (MIG) is an established voluntary collaboration of interventional cardiologists practicing at 6 tertiary referral hospitals in Melbourne, Australia. Data for consecutive patients undergoing PCI are submitted to a central registry and linkage of this registry to the Australian National Death Index has provided meaningful long-term mortality data after PCI. The aim of our study was to examine the timing, causes, and predictors of death during long-term follow-up after contemporary PCI using this large multicenter Australian registry.
Methods
The cohort consisted of all consecutive patients from the MIG registry undergoing PCI over a 6-year period for whom prospectively collected data were available. We report complete demographic, clinical, and procedural characteristics of the total PCI population enrolled to date. Standard techniques were used for all PCIs. Choice of anticoagulation, use of glycoprotein IIb/IIIa antagonists, intravascular ultrasound, and other procedural devices were left to the discretion of the operator. Oral antiplatelet therapy followed internationally accepted guidelines at the time, which recommended a combination of aspirin and clopidogrel for a minimum of 4 weeks for patients with bare-metal stents (BMS) and ≥6 months for patients with DES.
We specifically focused on comparing patients according to stent type. Patients included in the DES group had ≥1 DES implanted, whereas those in the BMS group had only BMS implanted. Use of BMS or DES was guided by funding-linked criteria from the Victorian State Health Department. Criteria for use of DES were consistent at all participating hospitals and were based on clinical guidelines developed in 2003 by the Victorian Department of Human Services. Criteria for use of DES included ≥1 of the following: (1) diabetes mellitus, (2) target vessel diameter ≤2.5 mm, (3) target lesion length ≥20 mm, (4) bifurcation lesion, (5) ostial lesion, (6) in-stent restenosis, and (7) long-term total occlusions.
Data were recorded prospectively on a standard electronic case-report form that includes standardized definitions for all fields. Data elements were incorporated from several current interventional databases including the American College of Cardiology–National Cardiovascular Data Registry. The registry is coordinated by the Monash Centre for Cardiovascular Research and Education in Therapeutics at Monash University in Melbourne, and data collection is overseen by a research coordinator at each hospital. The case-report form was developed using TeleForm 9 (Cardiff, Vista, California). Completed forms are electronically uploaded into the central database. An independent audit was conducted at all enrolling sites by an investigator not affiliated with that institution. Overall data accuracy was 96.6%, which compares favorably with other large multicenter registries.
Although we routinely perform 30-day and 12-month follow-up on all patients, the focus of this study was long-term mortality rates of the MIG cohort as defined by linkage to the Australian National Death Index. The Australian National Death Index is a database housed at the Australian Institute of Health and Welfare, which contains records of all deaths occurring in Australia since 1980. Data are obtained from the registries of births, deaths, and marriages in each state and territory. The index is designed to facilitate the conduct of epidemiologic studies and its use is strictly confined to medical research. The following variables for each deceased subject were identified: name, date of birth (or estimated year of birth), age at death, gender, date of death, state/territory of registration, and registration number. Cause of death (cardiac cause of death or not) was also analyzed. This information was derived by linking the Australian National Death Index registration numbers for deaths with the Australian National Mortality database. The index therefore relies on accurate completion of death certificates by treating physicians and is not adjudicated. Timing, cause, and predictors of mortality were determined.
Continuous data are expressed as mean ± SD and were compared by analysis of variance. Categorical variables are presented as frequency and percentage and were compared using chi-square test. A survival curve was generated by the Kaplan–Meier method. A multivariable Cox proportional hazards regression model was created using baseline clinical and angiographic characteristics and procedure-related variables to identify independent predictors of survival (presented as hazard ratio [HR], 95% confidence interval [CI], and p value). Variables entered into the model included age, gender, diabetes, hypertension, hypercholesterolemia, smoking history, previous PCI, previous cardiac bypass surgery, previous congestive heart failure, left ventricular ejection fraction, renal failure (defined as creatine >0.2 mmol/L), peripheral vascular disease, chronic obstructive pulmonary disease, multivessel disease, presentation with acute coronary syndrome (ST-segment elevation myocardial infarction [STEMI] or non-STEMI), type of stent (DES), left anterior descending or left main coronary artery disease, restenotic lesions, reference vessel <2.5 mm, total stent length, complex lesions (American College of Cardiology/America Heart Association type B2/C), ostial or bifurcation lesion, long-term total occlusion, and use of glycoprotein IIb/IIIa inhibitors. Statistical analyses were performed using STATA 10.1 (STATACORP, College Station, Texas).
Results
The cohort consisted of 10,682 consecutive patients from the MIG registry undergoing PCI (from February 2004 through November 2009) with a mean follow-up of 3.2 ± 0.5 years. Baseline demographic, lesion-related, and procedural characteristics are presented for the total cohort and then analyzed according to stent type (DES vs BMS; Tables 1 and 2 ).
| Overall | DES | BMS | p Value | |
|---|---|---|---|---|
| Number of procedures | 10,682 | 4,622 (43%) | 6,060 (57%) | |
| Age (years), mean ± SD | 64.4 ± 12.0 | 64.3 ± 11.8 | 64.3 ± 12.2 | 0.02 |
| Men | 8,016 (75%) | 3,468 (75%) | 4,548 (75%) | 0.98 |
| Diabetes mellitus | 2,601 (24%) | 1,447 (31%) | 1,154 (19%) | <0.01 |
| Hypertension (blood pressure >140/80 mm Hg or on therapy) | 6,888 (64%) | 3,040 (65%) | 3,848 (63%) | 0.01 |
| Hypercholesterolemia (>5.0 mmol/L or on therapy) | 7,644 (72%) | 3,379 (73%) | 4,265 (70%) | <0.01 |
| Former or current smoker | 7,162 (67%) | 2,979 (65%) | 4,183 (69%) | <0.01 |
| Body mass index (kg/m 2 ), mean ± SD | 28.2 ± 5.0 | 28.2 ± 5.0 | 28.2 ± 5.1 | 0.74 |
| Left ventricular ejection fraction (%), mean ± SD | 53 ± 11% | 54 ± 11% | 53 ± 11% | <0.01 |
| Previous myocardial infarction | 3,273 (30%) | 1,584 (34%) | 1,689 (27%) | <0.01 |
| Previous percutaneous coronary intervention | 2,703 (25%) | 1,428 (30%) | 1,275 (21%) | <0.01 |
| Previous coronary bypass | 967 (9%) | 519 (11%) | 448 (7%) | <0.01 |
| Heart failure | 546 (5%) | 217 (4%) | 329 (5%) | 0.11 |
| Cerebrovascular disease | 657 (6%) | 274 (5%) | 383 (6%) | 0.40 |
| Chronic obstructive pulmonary disease | 505 (4.8%) | 175 (3.8%) | 330 (5.5%) | <0.01 |
| Peripheral vascular disease | 733 (6.9%) | 336 (7.3%) | 397 (6.6%) | 0.14 |
| Renal impairment (creatine >0.2 mmol/L) | 468 (4.4%) | 218 (4.7%) | 250 (4.1%) | 0.14 |
| Clinical presentation | ||||
| ST-segment elevation myocardial infarction | 2,692 (25%) | 830 (17%) | 1,862 (30%) | <0.01 |
| Non–ST-segment elevation myocardial infarction | 2,681 (25%) | 1,165 (25%) | 1,516 (25%) | 0.82 |
| Cardiogenic shock | 271 (2.5%) | 64 (1.4%) | 207 (3.4%) | <0.01 |
| Elective (i.e., non urgent) | 4,516 (42%) | 2,297 (49%) | 2,219 (36%) | <0.01 |
| Overall | DES | BMS | p Value | |
|---|---|---|---|---|
| Number of lesions | 12,747 | 5,820 (46%) | 6,927 (54%) | |
| Type B2/C lesion | 6,320 (49%) | 3,095 (52%) | 3,225 (46%) | <0.01 |
| De novo lesion | 12,009 (94%) | 5,344 (91%) | 6,665 (96%) | <0.01 |
| In-stent restenosis | 738 (5.8%) | 476 (8.2%) | 262 (3.8%) | <0.01 |
| Target vessel | ||||
| Left main coronary artery | 122 (1.0%) | 73 (1.3%) | 49 (0.7%) | <0.01 |
| Left anterior descending coronary artery | 4,277 (33%) | 2,108 (36%) | 2,169 (31%) | <0.01 |
| Right coronary artery | 4,085 (31%) | 1,487 (25%) | 2,598 (37%) | <0.01 |
| Left circumflex coronary artery | 1,752 (13%) | 763 (13%) | 989 (14%) | <0.01 |
| Bypass graft | 369 (2.9%) | 214 (3.7%) | 155 (2.2%) | <0.01 |
| Multivessel coronary disease | 5,974 (59%) | 2,705 (63%) | 3,269 (56%) | <0.01 |
| Mean total stent length (mm), mean ± SD | 16.9 ± 5.4 | 18.2 ± 6.1 | 15.7 ± 4.2 | <0.01 |
| Stent length ≥20 mm | 3,789 (29%) | 2,348 (40%) | 1,441 (20%) | <0.01 |
| Mean stent diameter (mm), mean ± SD | 2.9 ± 0.5 | 2.7 ± 0.4 | 3.1 ± 0.5 | <0.01 |
| Vessel diameter ≤2.5 mm | 3,275 (25%) | 2,211 (37%) | 1,064 (15%) | <0.01 |
| Glycoprotein IIb/IIIa use | 3,124 (29%) | 1,246 (27%) | 1,878 (31%) | <0.01 |
| Drug eluting stent use | 10,682 (100%) | 4,622 (43.3%) | — | — |
| Intra-aortic balloon pump | 245 (2.3%) | 64 (1.4%) | 181 (3.0%) | <0.01 |
| Procedural success | 10,244 (95%) | 4,573 (98%) | 5,671 (93%) | <0.01 |
The cohort was relatively high risk with 24% diabetics, 63% presenting with an acute coronary syndrome, 49% with complex type B2/C lesions, and 59% having multivessel disease. DES use was 43% and overall procedural success was 95%. DES were used more frequently in diabetics, patients with previous MI and previous PCI, and more so in elective cases. DES were also preferentially used in type B2/C lesions, in-stent restenotic lesions, longer lesions, and smaller-diameter vessels, which reflects the selective criteria for their use. In contrast, BMS were used more frequently in patients with STEMI and cardiogenic shock.
In-hospital, 30-day, 1-year, and long-term (3.2 ± 0.5 years) mortalities for the total cohort were 1.6%, 2.1%, 3.9%, and 8.2%, respectively ( Table 3 ). Mortality rates at these intervals were further analyzed according to patient stent group. Cause of early death in all groups was predominantly cardiac; however, the proportion of cardiac deaths decreased to approximately 50% at long-term follow-up, with no significant difference between those who received a BMS and those who received a DES ( Table 3 ). Mean time to death was 2.4 ± 0.5 years. In patients alive at 1 year, mortality rate at long-term follow-up was 4.6%.
| Mortality rates | Overall | DES | BMS | p Value |
|---|---|---|---|---|
| (n = 10,682) | (n = 4,622) | (n = 6,060) | ||
| In hospital | 1.6% | 1.0% | 2.1% | <0.001 |
| (% cardiac) | (79) | (82) | (87) | |
| 30 days | 2.1% | 1.3% | 2.6% | <0.001 |
| (% cardiac) | (79) | (77) | (82) | |
| 12 months | 3.9% | 3.1% | 4.5% | <0.001 |
| (% cardiac) | (61) | (58) | (66) | |
| Long-term : Australian National Death Index (median 3.2 years) | 8.2% | 7.7% | 8.7% | 0.072 |
| (% cardiac) | (52) | (50) | (54) |
Characteristics of nonsurvivors compared to survivors are presented in Table 4 (univariate analysis).
| Variable | Total | Nonsurvivors | Survivors | p Value |
|---|---|---|---|---|
| (n = 10,682) | (n = 876) | (n = 9,806) | ||
| Mean age (years) | 64.1 | 70.6 | 64.3 | <0.001 |
| Mean body mass index (kg/m 2 ) | 28.1 | 26.7 | 28.2 | 0.01 |
| Mean left ventricular ejection fraction (%) | 53.7 | 37.8 | 53.8 | <0.001 |
| Women | 24% | 29% | 24% | 0.004 |
| Hypertension | 64% | 74% | 63% | <0.001 |
| Diabetes | 24% | 35% | 23% | <0.001 |
| Current smoking | 23% | 16% | 23% | <0.001 |
| Renal failure | 4.4% | 18% | 3.2% | <0.001 |
| Peripheral vascular disease | 6.9% | 17% | 5.9% | <0.001 |
| Previous stroke | 6.2% | 12% | 5.6% | <0.001 |
| Recent heart failure | 5.1% | 24% | 4.3% | <0.001 |
| Previous myocardial infarction | 30% | 39% | 29% | <0.001 |
| Previous coronary bypass | 9.1% | 15% | 8.5% | <0.001 |
| Multivessel coronary disease | 59% | 74% | 57% | <0.001 |
| Acute coronary syndrome | 62% | 73% | 61% | <0.001 |
| ST-segment elevation myocardial infarction | 25% | 33% | 24% | <0.001 |
| Cardiogenic shock | 2.5% | 15% | 1.4% | <0.001 |
| Left anterior descending artery stent | 36% | 40% | 36% | 0.02 |
| Left main stent | 1.1% | 3.2% | 0.9% | <0.001 |
| Complex lesion (typeB2/C) | 52% | 58% | 51% | <0.001 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
