The aim of the present study was to compare the outcomes after coronary artery bypass grafting (CABG) and percutaneous coronary intervention (PCI) in patients aged ≥80 years. The present analysis included 274 patients who underwent isolated CABG and 393 patients who underwent PCI. The patients undergoing PCI had a greater prevalence of a history of cardiac surgery and recent myocardial infarction and had more frequently undergone emergency revascularization. Patients undergoing CABG had a significantly greater prevalence of 3-vessel coronary artery disease. The unadjusted 30-day mortality rate was 8.8% after CABG and 7.4% after PCI (p = 0.514). However, on multivariate analysis, CABG was associated with a significantly increased risk of 30-day mortality (odds ratio 2.246, 95% confidence interval 1.141 to 4.422). The unadjusted overall intermediate survival was significantly poorer after PCI (at 5 years, CABG 72.2% vs PCI 59.5%, p = 0.004), but this was not confirmed on multivariate analysis. PCI and CABG had similar intermediate survival rates when adjusted for propensity score (p = 0.698), a finding confirmed by the analysis of 130 propensity score-matched pairs (at 5 years, CABG 66.4% vs PCI 58.9%, p = 0.730). In conclusion, the survival of patients aged ≥80 years undergoing CABG is excellent, and the suboptimal survival after PCI seems to be related to the disproportionately greater risk of these patients compared to those undergoing CABG. When adjusted for important clinical variables, PCI and CABG achieved similar intermediate results.
Data from the United Nations have indicated that by 2050 the >80-year age group is projected to reach 379 million worldwide, about 5.5 times as many in 2000, when there were 69 million persons aged ≥80 years. Coronary revascularization procedures are becoming more common in this age group with the increasing numbers of persons aged ≥80 years, as well as because of the favorable outcomes of octogenarians after coronary artery bypass grafting (CABG) and percutaneous coronary intervention (PCI). The use of PCI in these high-risk patients is attractive because of its minimally invasive nature and its somewhat lower operative mortality. However, CABG has been shown to achieve excellent intermediate survival. Whether PCI is superior to CABG in patients aged ≥80 years is largely unknown, and we investigated this issue in the present study.
Methods
The present study included a consecutive series of 274 patients who underwent isolated CABG at the Oulu University Hospital and Turku University Hospital, Finland, from January 2001 to January 2011 and 393 consecutive patients who underwent PCI at the Turku University Hospital, Finland, from January 2002 to January 2011. All the patients were aged ≥80 years. The ethics committee of our institutions approved the study protocol. Data on the cause and date of death for all patients were retrieved from the Finnish National Registry Statistics Finland. The mean follow-up was 3.6 ± 2.6 years. The cause of late death was unknown for 9 patients (2.3%) who had undergone PCI and for 7 patients (2.6%) who had undergone CABG, and these were considered noncardiac deaths for the present analysis. The main outcome end points of the study were cardiac and all-cause mortality.
Statistical analysis was performed using PASW, version 18 (IBM SPSS, Chicago, Illinois). Fisher’s exact test, the Mann-Whitney U test, and the Kaplan-Meier test were used for univariate analysis. Multivariate analysis was performed using logistic and Cox regression analyses with backward selection by including variables with p <0.05 on univariate analysis. The treatment groups differed with respect to the pretreatment covariables. Therefore, the propensity score was calculated by logistic regression analysis with backward selection by including the clinical variables with p <0.20 on univariate analysis. This was used for 1-to-1 matching and for adjustment of the risk in the overall series. One-to-one propensity score matching between the study groups was done with a caliber width (0.06) estimated according to Austin. p Values <0.050 were considered statistically significant.
Results
The patients undergoing PCI were significantly older and more frequently had a history of cardiac surgery, had had a recent myocardial infarction, and had required emergency revascularization ( Table 1 ). Patients undergoing CABG had a significantly greater prevalence of 3-vessel coronary artery disease ( Table 1 ).
| Variable | Overall Series | Propensity Score-Matched Pairs | ||||
|---|---|---|---|---|---|---|
| CABG Group (n = 273) | PCI Group (n = 392) | p Value | CABG Group (n = 130) | PCI Group (n = 130) | p Value | |
| Age (years) | 82.0 ± 1.7 | 83.3 ± 2.5 | <0.0001 | 82.5 ± 2.0 | 82.6 ± 2.1 | 0.808 |
| Women | 107 (39%) | 193 (49%) | 0.010 | 48 (37%) | 59 (45%) | 0.166 |
| Serum creatinine (mg/dl) | 89 ± 24 | 94 ± 52 | 0.508 | 90 ± 25 | 102 ± 71 | 0.178 |
| Pulmonary disease | 34 (13%) | 39 (10%) | 0.271 | 13 (10%) | 16 (12%) | 0.555 |
| Extracardiac arteriopathy | 28 (10%) | 31 (8%) | 0.275 | 20 (15%) | 12 (9%) | 0.131 |
| Diabetes mellitus | 47 (18%) | 88 (22%) | 0.124 | 33 (25%) | 25 (19%) | 0.233 |
| Hypertension | 182 (68%) | 273 (70%) | 0.588 | 92 (71%) | 92 (71%) | 1.000 |
| Stroke | 15 (6%) | 25 (6%) | 0.663 | 10 (8%) | 12 (9%) | 0.656 |
| Neurologic dysfunction | 9 (3%) | 5 (1%) | 0.100 | 6 (5%) | 1 (1%) | 0.120 |
| Previous percutaneous coronary intervention | 23 (9%) | 42 (11%) | 0.367 | 16 (12%) | 13 (10%) | 0.541 |
| Previous cardiac surgery | 4 (2%) | 31 (8%) | <0.0001 | 4 (3%) | 3 (2%) | 1.000 |
| Coronary arteries narrowed (n) | <0.0001 | 0.488 | ||||
| 1 | 5 (2%) | 83 (21%) | 5 (4%) | 3 (2%) | ||
| 2 | 30 (11%) | 170 (43%) | 27 (21%) | 34 (26%) | ||
| 3 | 238 (87%) | 139 (36%) | 98 (75%) | 93 (72%) | ||
| Left main stenosis | 118 (43%) | 119 (43%) | 0.961 | 56 (43%) | 48 (51%) | 0.268 |
| Left ventricular ejection fraction ≤50% | 72 (26%) | 105 (27%) | 0.906 | 45 (35%) | 32 (25%) | 0.077 |
| Myocardial infarction <3 mo | 154 (56%) | 303 (77%) | <0.0001 | 91 (70%) | 84 (65%) | 0.355 |
| Emergency procedure | 26 (10%) | 125 (32%) | <0.0001 | 23 (18%) | 24 (19%) | 0.872 |
| Beating heart surgery | 106 (39%) | — | — | 3 (4%) | — | — |
| ≥1 Mammary artery graft | 224 (82%) | — | — | 99 (76%) | — | — |
| No. distal anastomoses | 3.3 ± 1.0 | — | — | 3.1 ± 1.1 | — | — |
| Vessels treated by percutaneous coronary intervention | — | 1.2 ± 0.4 | — | — | 1.2 ± 0.5 | — |
| Drug-eluting stents (n) | — | 106 (27%) | — | 35 (27%) | — | |
| Short and intermediate overall survival | 0.004 | 0.730 | ||||
| 30 day | 91.2% | 92.6% | 89.2% | 92.3% | ||
| 1 year | 87.5% | 86.0% | 83.1% | 86.2% | ||
| 3 years | 80.4% | 71.6% | 71.3% | 69.7% | ||
| 5 years | 72.2% | 59.5% | 66.4% | 58.9% | ||
The 30-day mortality rate was 8.8% after CABG and 7.4% after PCI (p = 0.514). Diabetes (p = 0.018), an emergency procedure (p <0.0001), and serum creatinine (p <0.0001) were independent predictors of 30-day mortality on logistic regression analysis. When the procedure was adjusted for the latter variables, CABG (odds ratio 2.246, 95% confidence interval 1.141 to 4.422; p = 0.019), serum creatinine (odds ratio 1.019, 95% confidence interval 1.009 to 1.029; p <0.0001), and emergency procedure (odds ratio 4.725, 95% confidence interval 2.389 to 9.346; p <0.0001) were significantly associated with an increased risk of 30-day mortality (area under the receiver operating characteristics curve 0.747, 95% confidence interval 0.671 to 0.823; p = 0.296, Hosmer-Lemeshow test).
On univariate analysis, PCI was associated with significantly poorer overall survival (p = 0.004; Table 2 ), although the freedom from fatal cardiac events was similar between the study groups (p = 0.187, Table 3 ). Cox regression analysis showed that the treatment methods did not affect either all-cause mortality or cardiac mortality ( Tables 2 and 3 ).
| Variable | Univariate Analysis (p Value) | Multivariate Analysis (HR, 95% CI) |
|---|---|---|
| Procedure type ⁎ | 0.004 | — |
| Age | <0.0001 | 1.118, 1.056–1.183 |
| Serum creatinine | <0.0001 | 1.006, 1.004–1.008 |
| Pulmonary disease | 0.008 | 1.941, 1.344–2.804 |
| Diabetes | 0.001 | 1.627, 1.193–2.218 |
| Recent myocardial infarction | <0.0001 | 1.606, 1.163–2.217 |
| Ejection fraction ≤50% | 0.004 | — |
| Neurologic dysfunction | 0.027 | 2.575, 1.254–5.289 |
| Extracardiac arteriopathy | 0.002 | 1.557, 1.045–2.319 |
| Emergency procedure | <0.0001 | 1.644, 1.202–2.249 |
