Hybrid coronary revascularization (HCR) combines minimally invasive left internal mammary artery–to–left anterior descending coronary artery grafting with percutaneous coronary intervention of non–left anterior descending coronary arteries. The safety and efficacy of HCR in patients ≥65 years of age is unknown. In this study, patients aged ≥65 years were included who underwent HCR at an academic center from October 2003 to September 2013. These patients were matched 1:4 to similar patients treated with coronary artery bypass grafting (CABG) using a propensity-score matching algorithm. Conditional logistic regression and Cox regression stratified on matched pairs were performed to evaluate the association between HCR and CABG, and 30-day major adverse cardiovascular and cerebrovascular events (a composite of mortality, myocardial infarction, and stroke), periprocedural complications, and 3-year all-cause mortality. Of 715 patients (143 of whom underwent HCR and 572 CABG) in the propensity score–matched cohort, rates of 30-day major adverse cardiovascular and cerebrovascular events were comparable after HCR and CABG (5.6% vs 3.8%, odds ratio 1.46, 95% confidence interval 0.65 to 3.27, p = 0.36). Compared with CABG, HCR resulted in fewer procedural complications (9.1% vs 18.2%, p = 0.018), fewer blood transfusions (28.0% vs 53.3%, p <0.0001), less chest tube drainage (838 ± 484 vs 1,100 ± 579 cm 3 , p <0.001), and shorter lengths of stay (<5 days: 45.5% vs 27.4%, p = 0.001). Over a 3-year follow-up period, mortality rates were similar after HCR and CABG (13.2% vs 16.6%, hazard ratio 0.81, 95% confidence interval 0.46 to 1.43, p = 0.47). Subgroup analyses in high-risk patients (Charlson index ≥6, age ≥75 years) rendered similar results. In conclusion, although the present data are limited, we found that in older patients, the use of HCR is safe, has fewer procedural complications, entails less blood product use, and results in faster recovery with similar longitudinal outcomes relative to conventional CABG.
Coronary artery bypass grafting (CABG) is increasingly performed in older patient populations at risk for in-hospital mortality or major morbidity. Age-related diastolic dysfunction as a result of increased afterload due to arterial stiffness, as well as reduced functional reserve capacity, play a role in the worse outcomes in older patients. The use of robotic or thoracoscopic off-pump left internal mammary artery (LIMA)–to–left anterior descending coronary artery (LAD) grafting may offer a less invasive alternative to conventional CABG, particularly when integrated with percutaneous coronary intervention (PCI) for complete revascularization. However, the safety and efficacy of this integrated revascularization strategy, or hybrid coronary revascularization (HCR), have not been evaluated in older patients. To address this information gap, we conducted a study in which we assessed procedural complications, as well as 30-day and longitudinal clinical outcomes, in patients who underwent HCR versus conventional CABG.
Methods
The starting population for this analysis included the Emory University institutional Society of Thoracic Surgeons Adult Cardiac Surgery Database for all eligible cases from October 2003 to September 2013 (n = 9,902). Emory uses a custom data field to define patients on an intent-to-treat basis, in which HCR procedures involved planned LIMA-to-LAD bypass with the use of less invasive surgical techniques, combined with PCI of the remaining lesions, either performed in 1 setting or as 2 staged procedures. From this population, we excluded patients who were <65 years of age and those with clinical presentations of resuscitation or cardiogenic shock, histories of cancer (<5 years), chronic illicit substance abuse, previous cardiac surgery, concomitant noncoronary surgery, single-vessel coronary artery disease, and no LIMA use. From the remaining 4,140 patients, we then matched “as treated” HCR cases to patients who underwent elective or urgent conventional CABG with or without the use of cardiopulmonary bypass. A subgroup analysis was performed in high-risk patients using Charlson co-morbidity index ≥6 and age ≥75 years, which are indirect indicators or surrogates for frailty status. We also performed a sensitivity analysis, in which we matched all HCR “intention-to-treat” cases with control patients who underwent CABG.
Consideration for HCR at Emory hospitals, as well as the timing and sequence of the surgical and percutaneous components, is based on recommendations from a multidisciplinary heart team, as well as discussions with referring cardiologists and patients. Detailed information on indications and procedural information of the HCR program at Emory has been published previously. In summary, angiographic indications for HCR include the presence of significant proximal LAD disease or left main equivalent that is amenable to LIMA-to-LAD bypass and non-LAD lesions that allow PCI. Relative clinical contraindications for HCR include hemodynamic instability, previous cardiac or thoracic surgery, severe lung disease with the inability to tolerate single-lung ventilation, and severe morbid obesity. The sequence or stages of HCR depend on various factors; the default strategy is to perform LIMA-to-LAD revascularization first, because this approach allows surgery to be performed without dual-antiplatelet therapy. However, PCI is preferred first in patients with critical non-LAD disease, in which LIMA-to-LAD grafting is performed while patients are receiving clopidogrel. In a small number of patients, HCR is performed as a “simultaneous” 1-stage procedure, in which LIMA-to-LAD grafting is performed directly followed by PCI of non-LAD lesions, after confirmation of LIMA-to-LAD graft patency and the administration of a 600-mg loading dose of clopidogrel. At Emory, the surgical component of HCR was performed with an endoscopic atraumatic coronary artery bypass approach until 2009; thereafter, LIMA harvest was performed using robotic assistance (da Vinci Surgical System; Intuitive Surgical, Sunnyvale, California). After identification of the optimal target site on the LAD, the LIMA-to-LAD anastomosis is subsequently performed through a non-rib-spreading, 3- to 4-cm minithoracotomy, using a minimally invasive stabilizer (Nuvo; Medtronic, Inc., Minneapolis, Minnesota). The whole procedure is performed without the use of cardiopulmonary bypass. The PCI component of HCR was performed using standardized methods and techniques. In most cases, coronary stent placement involved either first-generation (sirolimus, paclitaxel) or newer generation (everolimus, zotarolimus) drug-eluting stents.
The primary outcomes of interest included the composite of death, myocardial infarction (MI) and stroke at 30 days. Secondary in-hospital outcomes were procedural complications, bleeding events, and recovery parameters, including postprocedural length of hospital stay and discharge location. Procedural complications included renal failure, prolonged ventilation, access-site infection, and reoperation. Bleeding events included chest tube drainage, need for blood transfusion, and CABG-related bleeding. Last, we compared long-term all-cause mortality between the 2 groups. Follow-up for all-cause mortality was obtained by querying the Social Security Death Index to determine dates of death for all patients in the study. Additionally, follow-up by telephone was performed up to September 2013. In-hospital events were documented in the hospital chart as part of routine medical care; this information was subsequently captured and entered into a computerized cardiac surgical database using the data fields and definitions of the Society of Thoracic Surgeons Adult Cardiac Surgery Database. The definition of CABG-related bleeding was based on the Bleeding Academic Research Consortium criteria.
After applying the aforementioned exclusion criteria, we performed propensity score matching to reduce the effects of treatment selection bias. The propensity score (or the probability of assignment to HCR or CABG) was constructed using multivariate logistic regression from available demographics and clinical and angiographic characteristics (see Table 1 ). Patients who underwent HCR were matched in a 4:1 ratio with patients who underwent CABG, using the nearest neighbor matching algorithm without replacement on the logit of the propensity score using a caliper of width equal to 0.30 SDs of the logit of propensity score. After propensity score matching, we tested whether the balance on the covariates was achieved through the matching procedures using standardized differences (an absolute standardized difference of <10% was considered a negligible imbalance) and assessment of global imbalance (p = 0.99 after matching). Comparisons of in-hospital outcomes and the 30-day composite end point were performed using conditional logistic regression analyses accounting for the matched nature of the propensity score matched sample. The Kaplan-Meier method was used to create survival curves and Cox proportional-hazards regression analysis stratified on matched pairs was used to examine differences in mortality up to 3 years after the procedure. The proportional-hazards assumptions were evaluated by visual assessment by log minus log survival plots and were found to be valid. The comparative effectiveness of HCR in the high-risk subgroups was assessed using interaction terms for treatment. Statistical analyses were performed using IBM SPSS Statistics version 20.0 (IBM Corporation, Armonk, New York). Propensity score matching was performed using an SPSS-R plug-in for R packages (Matchit, Ritools, and cem).
| Unadjusted | After PS Matching | |||||
|---|---|---|---|---|---|---|
| HCR (n = 144) | CABG (n = 3,979) | Stand Diff | HCR (n = 143) | CABG (n = 572) | Stand Diff | |
| Age (years) | 74.3 ± 6.6 | 72.6 ± 5.7 | 26.4% | 74.2 ± 6.5 | 73.9 ± 6.0 | 4.9% |
| Female | 34.0% | 32.6% | 3.1% | 33.6% | 35.9% | 4.9% |
| White | 84.0% | 80.1% | 10.6% | 83.9% | 84.1% | 0.5% |
| Body mass index (kg/m 2 ) | 26.9 ± 4.2 | 28.2 ± 5.8 | 31.3% | 26.9 ± 4.2 | 27.2 ± 5.1 | 7.0% |
| Diabetes | 36.1% | 41.4% | 10.9% | 36.4% | 38.1% | 3.5% |
| Hypertension | 91.7% | 91.0% | 2.4% | 91.6% | 91.8% | 0.6% |
| Cerebrovascular disease | 25.0% | 24.2% | 4.1% | 24.5% | 24.5% | 0.0% |
| Peripheral arterial disease | 16.7% | 19.3% | 6.9% | 16.8% | 16.6% | 0.5% |
| Chronic lung disease | 4.9% | 7.3% | 11.5% | 4.9% | 6.5% | 7.3% |
| Renal failure | 6.3% | 6.1% | 0.6% | 6.3% | 6.5% | 0.7% |
| Heart failure ∗ | 13.9% | 24.5% | 30.5% | 14.0% | 13.6% | 1.0% |
| Prior myocardial infarction | 55.6% | 48.9% | 13.3% | 55.9% | 56.8% | 1.8% |
| Left ventricular ejection fraction (%) | 54.5 ± 10.0 | 52.5 ± 12.2 | 20.3% | 54.5 ± 10.0 | 54.8 ± 11.1 | 2.6% |
| Number of diseased vessels † | 2.4 ± 0.5 | 2.7 ± 0.4 | 64.4% | 2.4 ± 0.5 | 2.4 ± 0.5 | 2.0% |
| Urgent status | 36.8% | 45.5% | 17.9% | 37.1% | 35.4% | 3.5% |
∗ Physician documentation or report of heart failure within the 2 weeks prior to admission.
† Number of diseased vessels = number of diseased major native coronary vessel systems with ≥50% narrowing of any vessel preoperatively.
Results
From October 2003 to August 2013, 161 patients aged ≥65 years underwent HCR on an intent-to-treat basis. Of those who were scheduled for HCR, 17 patients (10.6%) did not undergo 1 or both stages of HCR. Thirteen of these patients underwent minimally invasive LIMA-to-LAD grafting, but the other lesions were medically managed instead of treated with PCI. In 4 patients, HCR was converted to conventional CABG. The remaining 144 patients all underwent HCR, either as a 1-stage procedure (n = 12 [8.3%]) or as a staged procedure, in which PCI was performed first in 22 patients (16.7%) and CABG was performed first in 110 patients (83.2%). The median time interval between the 2 stages was 3 days (interquartile range 2 to 7 days).
A total of 144 older patients who underwent HCR with minimally invasive LIMA-to-LAD bypass and PCI of non-LAD coronary vessels were matched 4:1 with conventional CABG patients on the basis of 15 preoperative variables. Demographics and clinical characteristics ( Table 1 ) were well balanced after propensity score matching (standardized difference <10%). One patient who underwent HCR could not be matched and was therefore excluded. Compared with matched patients who underwent HCR, the use of cardiopulmonary bypass (0% vs 12.9%, p <0.001), the number of vein graft anastomosis (0 [0 to 0] vs 2 [1 to 2], p <0.001), as well as the total number of anastomosis (1 [1 to 1] vs 3 [2 to 4], p <0.001) was larger in patients who underwent CABG. Bilateral internal mammary artery use was 0% in the HCR group and 6.6% in the CABG group (p = 0.002). Drug-eluting stents were used in 80.8% of patients in the HCR group and involved left main coronary artery stenting in 18.3%, circumflex coronary artery stenting in 23.3%, right coronary artery stenting in 36.7%, combined circumflex coronary artery and right coronary artery stenting in 10.0%, and other combinations of stented vessels in 11.7% of patients. Complete revascularization was obtained in 137 patients (95.8%) after HCR and 529 patients (92.5%) after CABG. Median surgical operating time was shorter after HCR compared with CABG (279 minutes [interquartile range 239 to 324] vs 306 minutes [interquartile range 255 to 357], p = 0.003).
The incidence of the composite of death, MI, and stroke at 30 days was comparable between HCR and CABG (5.6% vs 3.8%, odds ratio [OR] 1.46, interquartile range 0.65 to 3.27, p = 0.36). Death occurred in 2.8% (n = 4) and 2.6% (n = 15) patients after HCR and CABG, respectively. MI was reported in 2 patients after HCR (1.4%) and in none of the CABG patients. Stroke occurred in 1.4% of patients in the 2 groups (n = 2 and n = 8). Procedural complications were significantly lower after HCR compared with CABG (9.1% vs 18.2%, OR 0.50, interquartile range 0.28 to 0.89, p = 0.018; Table 2 ). This difference was due mainly to a lower incidence of prolonged ventilation (5.6% vs 14.5%, p = 0.010), but other complications were also numerically lower in the HCR group. The secondary end point was bleeding outcomes. The incidence of CABG-related bleeding (Bleeding Academic Research Consortium type 4) was numerically (but not significantly) lower in the HCR group compared with the CABG group (7.0% vs 11.2%). Other bleeding parameters, such as the use of blood transfusion and 24-hour chest tube drainage, were significantly lower after HCR compared with CABG. Recovery time, as measured by postprocedural length of hospital stay, was significantly shorter in patients after HCR. The median follow-up period was 3.3 years in the matched cohort (interquartile range 1.4 to 5.5 years). As illustrated in Figure 1 , all-cause mortality was similar after HCR and CABG over 3 years of follow-up.
