|Home | About | Journals | Submit | Contact Us | Français|
Less-invasive options are available for surgical treatment of multivessel coronary artery disease. We hypothesized that stenting combined with grafting of the left anterior descending artery with the left internal thoracic artery through a minithoracotomy (hybrid procedure) would provide the best outcome.
Patients with equivalent numbers of coronary lesions (2.8 ± 0.4) underwent either hybrid (n = 15) or off-pump coronary artery bypass through a sternotomy (n = 30). Early and 1-year outcomes were compared. Blood drawn from the aorta and coronary sinus immediately postoperatively was analyzed for activation of coagulation (prothrombin fragment 1.2 and activated Factor XII), myocardial injury (myoglobin), and inflammation (interleukin 8) by using an enzyme-linked immunosorbent assay. Target-vessel patency was determined by means of computed tomographic angiographic analysis.
The hybrid procedure was associated with significantly shorter lengths of intubation and stays in the intensive care unit and hospital and perioperative morbidity (P < .05). Intraoperative costs were increased but postoperative costs were reduced for the hybrid procedure compared with off-pump coronary artery bypass through a sternotomy. As a result, overall total costs were not significantly different between the groups. After adjusting for potential confounders, assignment to the hybrid group was an independent predictor of shortened time to return to work (t = −2.12, P = .04). Patient satisfaction after the hybrid procedure, as judged on a 6-point scale, was greater versus that after off-pump coronary artery bypass through a sternotomy. Finally, the hybrid procedure showed significantly reduced transcardiac gradients of markers of coagulation, myocardial injury, and inflammation and a trend toward significant improvement in target-vessel patency.
Perhaps because of reduced myocardial injury, inflammation, and activation of coagulation, patients undergoing the hybrid procedure had better perioperative outcomes and satisfaction, with excellent patency at 1 year’s follow-up. These promising preliminary findings warrant further investigation of this procedure.
Despite major improvements in stent technology, the left internal thoracic artery (LITA) bypass graft remains the superior long-term option for treating a stenosis of the left anterior descending coronary artery (LAD).1,2 Compared with a stent, the LITA graft is resistant to thrombosis and atherosclerosis and provides protection from progression of proximal coronary artery disease (CAD). A growing list of less-invasive options has become available that exploit the benefit of the LITA, including off-pump coronary artery bypass grafting (CABG) through a sternotomy (OPCAB) or multivessel revascularization through a small thoratomy.3
A third alternative, percutaneous coronary intervention (PCI)/stenting combined with surgical LITA to LAD grafting through a minithoracotomy (the hybrid procedure), has theoretic advantages. Stents substitute for the saphenous vein graft (SVG) as a bypass conduit, and LITA grafting through a minimally invasive approach minimizes surgical morbidity. This hybrid approach has not been widely adopted because of a number of practical concerns: the need for close cooperation of surgical and interventional groups, the logistic concerns of timing and sequencing of the procedures, and the use of aggressive anticoagulation in the surgical patient.
As a result of these challenges, the status quo for the surgical treatment of multivessel CAD is to perform a sternotomy for bypass grafting of a single LITA and multiple SVGs. At our institution, the surgical and interventional portions of the hybrid procedure have been completed simultaneously in a single operative suite. The purpose of this study was to compare the perioperative and 1-year outcomes of this state-of-the-art approach to the hybrid procedure compared with those of standard OPCAB.
Fifteen consecutive patients underwent the simultaneous hybrid procedure at our institution from January 2005 through December 2006. Using a prospective case-controlled study design, we matched a parallel control group of 30 patients who underwent OPCAB according to demographics, risk factors, comorbidities, coronary anatomy, medical therapy, and operative surgeon (RP). These matching criteria included known risk markers for outcomes with surgical revascularization (Table 1). Inclusion criteria for the hybrid procedure were the presence of multivessel CAD that involved greater than 70% LAD obstruction judged a suitable surgical target and the presence of a non-LAD coronary lesion (or lesions) suitable for PCI, as adjudicated by 2 interventionalists (BR and DZ) and 1 surgeon (RP). Hemodynamic instability, acute coronary syndromes, or situations in which complete revascularization was not possible served as exclusion criteria for the hybrid procedure. Patients with chronic renal insufficiency (creatinine value, >2.0 mg/dL) and allergy to radiographic contrast were also excluded from enrollment.
Patients were followed up daily during hospitalization and then at 1 year as an outpatient to assess mortality, target-vessel patency, and other outcomes. Demographics, preoperative risk factors and medications, and intraoperative and postoperative data were prospectively recorded into a relational database. All patients provided informed consent to be enrolled in the study (UMB IRB no. 25350).
After median sternotomy, the LITA was procured and the saphenous vein was harvested by using an endoscopic approach (VasoView6; Guidant Systems, Inc, Minneapolis, Minn). Proximal anastomoses were performed with a partial occluding aortic clamp. All distal anastomoses were performed on the beating heart facilitated by suction-based exposure and stabilizing devices (Octopus 4.3; Medtronic, Inc, Minneapolis, Minn).
Patients were placed in a supine position with the left chest slightly elevated and intubated with a double-lumen endotracheal tube to allow for collapse of the left lung. An 8- to 10-cm anterior lateral thoracotomy in the fourth intercostal space was created to expose the in situ LITA with the assistance of the LITA lift retractor (Genzyme Cardiovascular, Cambridge, Mass). The conduit was then harvested as a pedicle. LITA-to-LAD CABG was performed through a small thoracotomy without the use of cardiopulmonary bypass. Selective use of an intracoronary shunt was used based on changes, which were suggestive of anterior myocardial ischemia. A stabilizer (Octopus 4.3; Medtronic, Inc) was then secured over the LAD, and the anastomosis was performed on the beating heart. Additional targets were revascularized by using PCI performed immediately after completion of LITA grafting in a specially designed operating suite outfitted with fluoroscopic equipment. In most cases the thorax was closed before PCI. However, for those patients with higher levels of bleeding, the thorax was left open until the completion of PCI to allow a second evaluation of hemostasis. Access was achieved through the femoral artery by using 6F guiding catheters. Guidewire and stent selection, along with predilation and postdilation, were left to the discretion of the operator. Drug-eluting stents were implanted in all patients. Both the Cypher Sirolimus-eluting stent (Cordis Corp, Miami Lakes, Fla) and the Taxus Paclitaxel–eluting stent (Boston Scientific, Inc, Natick, Mass) were used.
In both groups unfractionated heparin was administered intraoperatively to obtain a kaolin-based activated clotting time (ACT) of greater than 300 seconds and a heparin level of greater than 2 IU/mL, according to the heparin–protamine titration assay (HMS Heparin Assay Cartridges; Medtronic, Inc). Also, aspirin (325 mg by mouth daily) was given to both groups preoperatively and within 6 hours postoperatively. Heparin was reversed with half the recommended dose of protamine for OPCAB. For patients undergoing the hybrid procedure, heparin was not reversed, and a loading dose of 300 mg of clopidogrel was administered through a nasogastric tube on arrival to the intensive care unit (ICU), followed by 75 mg daily thereafter. GPIIb/IIIa antagonists were not used.
During hospitalization, major adverse cardiac events (MACEs) were monitored by determining mortality, perioperative myocardial infarction on the basis of new Q-waves or troponin I levels of greater than 5 times normal values, clinically evident stroke, or the need for coronary artery reintervention. Intraoperative blood loss was quantified by using a Cell Saver device. Chart review was conducted to assess intraoperative packed red blood cell transfusions, intraoperative cardiac index and central venous pressure, postoperative length of intubation, daily serum creatinine level, length of ICU and hospital stay, and peak pain score (scale, 0–10).
At 1 year after the operation, patients were interviewed, and their medical records were reviewed to determine the following outcomes: (1) mortality, (2) MACEs at 1 year, (3) New York Heart Association (NYHA) angina classification, (4) duration of pain from surgical incision persisting after the operation, (5) length of time to return to work or normal activities, and (6) overall satisfaction with the procedure based semiquantitatively on a score of 1 to 6, with 1 being dissatisfied and 6 being completely satisfied. Total costs were obtained for each group by using the hospital’s database.
Differences in regional hypercoagulability, ischemia, and inflammation were measured by collecting arterial (“aortic”) and coronary sinus (CS) blood samples 30 minutes after heparin reversal into tubes containing 3.2% citrate. CS blood was obtained through a heparin-bonded catheter (Cook, Inc) placed into the CS through trans-jugular access for patients undergoing the hybrid procedure or by means of direct external puncture of the CS with an 18-gauge butterfly needle for patients undergoing OPCAB. Platelet-poor plasma was obtained by means of rapid centrifugation (2000g) and stored at −80°C. Thrombin formation was assayed on the basis of the level of prothrombin fragment 1.2 (F1.2) by means of enzyme-linked immunosorbent assay (ELISA; Dade Behring, Marburg, Germany), and the activity of the contact activation pathway was evaluated by assaying activated Factor XII (American Diagnostica, Inc, Stamford, Conn). Cardiac myoglobin and interleukin 8 releases were determined by comparing aortic versus CS levels in the blood samples by means of ELISA (Life Diagnostics, Inc, West Chester, Pa, and Bender MedSystems, Vienna, Austria, respectively). Comparison of these markers in the CS (F1.2CS) to a simultaneously obtained aortic (F1.2Ao) sample allowed for calculation of the percentage of transcardiac change, as follows: (F1.2CS−F1.2Ao)/F1.2Ao×100. Additionally, systemic blood samples were obtained on postoperative day 1 to assess troponin I levels by means of ELISA (Life Diagnostics, Inc).
In addition, target-vessel patency was assessed by using 16-channel CTA (420-ms rotation with 100–150 mL of contrast agent administered intravenously at 5 mL/s and retrospective electrocardiographic gating; Philips MX8000, Cleveland, Ohio) at both 5 days (postoperative) and 1 year (follow-up) after the operation. Patency at both time points was defined as any flow through the entire graft or stent, regardless of the presence of stenosis. The graft was classified as nonpatent if a stump was seen or if there was no contrast in an area known by operative report to contain a graft, as previously described.4 The diagnosis of stent thrombosis was based on screening computed tomographic (CT) angiographic analysis, clinical signs/symptoms of ischemia confirmed as a result of stent closure by means of conventional angiographic analysis, or both.
The primary end point of this trial was a comparison of postoperative morbidity between the 2 groups. We performed univariate analysis to compare potential confounders of the relationship between group assignment and time to return to work/normal activities using the Student t test and the Fisher exact test for continuous and dichotomous variables, respectively. Variables with a P value of less than .1 between groups were included in a stepwise logistic regression model, with the dependent variable being categorical (time to return to work of >1 or <1 month) and independent variables that were both continuous (age, body mass index, and preoperative creatinine value) and categorical (NYHA classification, diabetes, and number of diseased vessels). Statistical analyses were performed by a bio-statistician (AJ).
Baseline patient characteristics were similar between the 2 groups in all assessed preoperative and intraoperative variables (Table 1). All patients received a LITA graft, and additional targets were addressed per group assignment. The hybrid group received 22 stents (11 Sirolimus-eluting stents and 11 Taxus Paclitaxel–eluting stents), and the OPCAB group received 40 SVGs, 6 right internal thoracic artery grafts, and 6 radial artery grafts. At 1 year, 80% of patients in each group successfully completed the follow-up, and 100% returned to assess target-vessel patency.
There were no mortalities and no readmissions during the postoperative period in either group. Postoperative MACEs developed in no patients in the hybrid group and in 7 patients in the OPCAB group because of 6 myocardial infarctions and 1 stroke (0% vs 23%, P = .05). Compared with the OPCAB group, patients in the hybrid group maintained better intraoperative cardiac indices (35% vs 23% decrease from baseline, P = .08) and less of an increase in central venous pressure (50% vs 17% increase from baseline, P = .01). Patients in the hybrid group also required less red blood cell transfusions (0.2 ± 0.4 vs 1.4 ± 1.4 U, P ≤ .0001), shorter intubation times (1.3 ± 3.4 vs 20.6 ± 25.7 hours, P ≤ .001), and shorter lengths of stay in the ICU (0.98 ± 0.42 vs 2.42 ± 1.57 days, P ≤ .0001) and hospital (3.7 ± 1.4 vs 6.4 ± 2.2 days, P ≤ .0001). Postoperative renal insufficiency, defined as an increase of serum creatinine values more than 25% above baseline values, was noted in 3 (10%) patients after OPCAB but was not seen after the hybrid procedure.
These differences resulted in a significant reduction in costs for the patients undergoing the hybrid procedure in the postoperative period. In contrast, OPCAB showed a significant reduction in intraoperative costs, largely because of shorter operative times and the use of autologous grafts rather than stents ($9819 ± $2229 vs $14,691 ± $2967, OPCAB vs hybrid procedure; P < .001). As a result of this difference in intraoperative versus postoperative costs, there was no significant difference in the overall hospital costs between the groups (Figure 1). Maximum pain scores, ranging from none (0) to the most severe ever experienced (10), were higher after minithoracotomy (hybrid procedure) than sternotomy (OPCAB; 8.6 ± 1.8 vs 6.8 ± 2.7, P = .01).
As the result of different protocols for heparin reversal, there was a significant difference in the ACT measured immediately at the completion of the case for patients in the hybrid group compared with those in the OPCAB group (235 ± 56 vs 132 ± 23 seconds, P < .001).
At 1 year’s follow-up, there was no mortality in either group. MACEs were noted in 1 (7%) of the patients in the hybrid group and in 7 (23%) of the patients in the OPCAB group as a result of 1 reintervention required for a patient in the hybrid group for stent thrombosis. This patient was also the only patient in either group with angina at 1 year (mean NYHA angina classification: 0.2 vs 0, hybrid procedure vs OPCAB). The duration of the time that it took for pain to completely resolve was shorter for the hybrid procedure versus OPCAB (10.3 ± 10.9 vs 45.5 ± 33.6 days, P = .004). Overall satisfaction scores were also higher after the hybrid procedure (Figure 2), with significantly more patients reporting that they were completely satisfied (83% vs 42%, hybrid procedure vs OPCAB; P < .001).
Patients returned to work or normal activities quicker after the hybrid procedure versus OPCAB (Figure 3). The average time to return to work in patients with a sternotomy was 4.4 ± 3.1 months, which is significantly greater than the 1.75 ± 1.0 months for patients with a minithoracotomy (t = 3.68, P = .0008). There was a significant relationship between group assignment (ie, hybrid vs OPCAB) and returning to work before 1 month (χ2 statistic = 7.08, P = .008). The odds of returning to work at less than 1 month were significantly better for the hybrid procedure versus OPCAB after adjusting for potential confounders (odds ratio, 7.60; 95% confidence interval, 1.61–35.91; P = .01). On multivariate regression analysis, the choice of procedure was a significant predictor of the time to return to work (t = −2.12; standard error = 0.96; P = .04; 95% confidence interval, −4.04 to −0.11). None of the other variables analyzed were significant predictors.
There was significantly less intraoperative blood loss during the hybrid procedure versus OPCAB (579 ± 406 vs 1091 ± 601 mL, P = .004). Despite better hemostasis, the transcardiac gradients of markers of thrombosis (F1.2 and activated Factor XII), ischemia (myoglobin), and inflammation (interleukin 8) were all significantly reduced after the hybrid procedure versus OPCAB (Figure 4). Troponin I levels on day 1 were significantly less in patients undergoing the hybrid procedure versus those undergoing OPCAB (1.4 ± 0.7 vs 2.8 ± 2.6 ng/mL, P < .03).
Conventional angiography performed during each case intra-operatively and CT angiography performed before hospital discharge confirmed target-vessel patency for all patients in the hybrid group. Predischarge CT angiographic results obtained in the OPCAB group documented patency in 77 (94%) of 82 grafts, with early failure noted in 1 right internal thoracic artery graft, 1 radial artery graft, and 3 SVGs. At 1 year, a single stent failed in the hybrid group versus 7 additional SVG failures in the OPCAB group (97% vs 85% overall patency, Figure 5).
OPCAB was initially touted as a less-invasive alternative to the traditional on-pump technique, yet direct comparisons of outcomes between these methods have been surprisingly similar. The hybrid procedure has been established in prior reports as a viable alternative to open-sternum CABG for selected patients with multivessel disease.5–7 We performed stenting and bypass simultaneously using a specially outfitted hybrid operating suite at our center. A variety of better outcomes were noted in patients treated with this state-of-the-art approach compared with OPCAB, including reduced lengths of intubation and ICU stay, less transfusions, quicker resolution of pain, earlier return to work/normal activities, and greater overall satisfaction with the surgical experience. Although target-vessel patency at 1 year’s follow-up was excellent in both groups, cardiac release of hypercoagulability markers was reduced after the hybrid procedure, suggesting a decreased risk for coronary thrombotic events in these patients. Our experience suggests that an encouraging array of advantages might follow the adoption of a less-traumatic technique, such as the hybrid procedure, that have not yet materialized by merely avoiding cardiopulmonary bypass.
Finding reduced postoperative costs for our hybrid group was notable given that OPCAB represents the best available evidence-based approach for cost savings.8,9 It is possible that enthusiasm for the hybrid procedure biased the management of variables that are often based on subjective judgments and also are important drivers of hospital costs, such as time for extubation, length of stay, and red blood cell transfusion. We believe the risk of this type of bias was minimized by the fact that our study was performed in the context of decreasing reimbursements for CABG, which creates a strong incentive to limit costs regardless of the approach. The differences in postoperative costs were offset by higher intraoperative costs for the hybrid procedure because of longer operative times and the costs of coated stents. Further reductions in operative times because of the learning curve of minimally invasive surgery and the recent transition to less-costly bare-metal stents to avoid delayed thrombosis might help improve total costs in our ongoing analysis.
Despite higher pain scores after minithoracotomy versus sternotomy, the odds of patients in the hybrid group returning to work within the first month were 7-fold better than in the OPCAB group, even after adjusting for potential confounders. Multivariate techniques cannot adjust for all the factors that influence a subjective end point such as the appropriate time to return to work. Although not detected in interviews, patients, their providers, or both could have had the preconceived notion that the hybrid procedure is more likely than OPCAB to allow a quicker return to normalcy. However, early return to work was the stated goal of a vast majority of the patients in this cohort. Recovery time defined in this broader context provides an important benchmark for programmatic success from the standpoint of the patient and helps to justify the use of a strategy that is not the standard of care for triple-vessel CAD.
More rapid recovery for the hybrid cohort suggests that factors other than pain play a role in their quicker recovery. Less blood transfusions and reduced systemic inflammation have been linked to improved postoperative morbidity in a number of studies comparing minimally invasive and conventional surgery and likely play a role in outcomes after the hybrid procedure. In addition, the degree of cardiac manipulation varies between the hybrid procedure, where the heart is left in its native position, versus OPCAB, which frequently requires the heart to be rotated into positions that compromise hemodynamics, as noted in our patients. In addition, coronary occlusion for the hybrid procedure is limited to that required for placement of a single LAD graft and less than 20-second intervals for stenting. OPCAB requires 8- to 12-minute periods of coronary occlusion during each of 3 to 4 distal anastomoses (total ischemic time, 25–40 minutes). Better myocardial protection, reflected by a reduction in regional myoglobin and systemic troponin I release, might be an additional mechanism for quicker recovery after the hybrid procedure.
Compared with OPCAB, patients undergoing the hybrid procedure showed a reduced transcardiac gradient of F1.2, a marker of thrombin production. This methodology has been used to assess coagulation activity within the upstream coronary circulation in a wide range of clinical studies.10–12 F1.2 is a proved risk factor for thrombosis of stents13 or bypass grafts14 and predicts morbidity after cardiac surgery. Potential triggers for thrombin in patients undergoing OPCAB are the obligatory periods of warm ischemia15 or the grafting of the more thrombogenic SVG versus using all arterial conduits.16 Additionally, reversal of heparin with protamine, avoided in the hybrid group, is known to provoke a transient “rebound” increase in thrombin formation. As evidenced by a significant difference in postoperative ACT, the heparin effect resolved in the patients undergoing the hybrid procedure after the time point at which transcardiac F1.2 levels were analyzed. This assay might reflect changes in the kinetics of thrombin formation and not just a determination of the relative thrombogenicity of each procedure. On the other hand, the safe use of more aggressive antithrombotic therapies, such as heparin administration without reversal, is the result of less risk of bleeding after the hybrid procedure. Undoubtedly, there are a variety of factors that influence the calculus of regional thrombin production after surgical revascularization.
A final advantage of the hybrid procedure is that it provides surgical revascularization without requiring an SVG. Although our study was not sufficiently powered to document a difference in target-vessel patency between groups, the use of the SVG is well known to limit the longevity of surgical treatment for multivessel CAD. All arterial grafting provides a theoretic advantage over the hybrid procedure by avoiding patient exposure to the potential risks of intracoronary restenosis, thrombosis, or both from either stents or SVGs. However, all-arterial grafting has not been enthusiastically adopted by most cardiac surgeons, in part because the best choice for arterial conduits after the LITA remains controversial. The right internal thoracic artery has shown excellent patency rates and increased survival advantage17–19 but is not often used because of the fear of sternal infection and a variety of other technical considerations.20,21 The radial artery is a versatile arterial conduit accustomed to pulsatile pressure but has demonstrated highly variable early patency results, preventing a consensus about its role in CABG.22,23 As a result, a recent review of the Society of Thoracic Surgeons database indicates that 95% of CABG cases nationwide involve an SVG. Therefore the hybrid procedure might be a more practical solution for avoiding the well-known limitations of the SVG than all-arterial grafting.
This pilot study has limitations that need to be considered when interpreting the data. First, patient assignment to groups was nonrandomized. Although this could serve to bias the selection of lower-risk patients for the minimally invasive approach, our hybrid cohort had a higher incidence of chronic obstructive pulmonary disease, peripheral vascular disease, and prior myocardial infarction and stroke compared with the OPCAB control group.
Second, subjective measures, such as the patient’s satisfaction and time to return to work, could have been influenced by preconceived notions about the superiority of the hybrid approach. However, randomization is not able to blind the patient to whether a minithoracotomy for the hybrid procedure or sternotomy for OPCAB was performed. The risk of this bias can only be controlled by comparing the hybrid group with one that is revascularized with multiple arterial grafts through a minithoracotomy.
Third, comparison of target-vessel patency between groups over the first postoperative year might not be sufficient to demonstrate the longevity advantages of bypass grafting over PCI that typically become evident only after 3 to 5 years of follow-up. These data provide reassurance, however, that the technical success of LAD grafting does not appear to be compromised by using a minimally invasive approach.
Finally, hybrid operating rooms with permanent fluoroscopic equipment are currently available at only a few centers, limiting the generalizability of our protocol. Our results cannot be extrapolated to the hybrid procedure performed as a staged procedure. Proliferation of percutaneous procedures for vascular stenting and valvular disease might make rooms for the simultaneous hybrid procedure more common in the future. Although our data support the feasibility of the hybrid approach, it is best interpreted as hypothesis generating to support the design of more appropriately powered, randomized studies in the future.
In conclusion, the treatment of multivessel CAD with the simultaneous hybrid approach provides a rapid recovery and a level of patient satisfaction that compares favorably with that of traditional OPCAB. These results were seen without significantly increasing costs or compromising graft patency. In addition, local activation of inflammation and coagulation was minimized by the hybrid procedure, perhaps because of reduced myocardial manipulation and warm ischemia. These promising early findings warrant further investigation of the simultaneous hybrid procedure.
RP is funded by a grant from the National Institutes of Health (R01HL084080), a Scientist Development Grant from the American Heart Association, an Intramural Grant from the University of Maryland, and a faculty pilot grant from the Tobacco Restitution Fund at the University of Maryland.
The treatment of multivessel coronary artery disease with the simultaneous hybrid procedure, a minimally invasive technique for revascularization, provides for clinical outcomes and patient satisfaction that are very competitive with those of traditional off-pump coronary artery bypass. In addition, local activation of inflammation and coagulation is minimized, perhaps because of less myocardial manipulation and regional bouts of warm ischemia.
Read at the Eighty-Seventh Annual Meeting of the American Association for Thoracic Surgery, Washington, DC, May 5–9, 2007.