- Email: [email protected]
Myocardial Revascularization Exclusively With Bilateral Internal Thoracic Arteries in T-Graft Configuration: Effects on Late Survival
Daniel O. Navia, MD, Mariano Vrancic, MD, Fernando Piccinini, MD, Mariano Camporrotondo, MD, Alberto Dorsa, MD, Juan Espinoza, MD, Mariano Benzadon, MD, and Juan Camou, MD Cardiac Surgery Department, Instituto Cardiovascular de Buenos Aires, Buenos Aires, Argentina
Background. We studied long-term survival using bilateral internal thoracic artery (BITA) grafting in a T-configuration exclusively versus using single internal thoracic artery (SITA) grafting in patients with multivessel disease. Methods. Consecutive coronary operations performed at a single center between 1996 and 2014 were reviewed. Long-term survival among patients receiving coronary revascularization exclusively with BITA grafting in a T-configuration (n [ 2,098) versus SITA grafts plus other types of conduits (saphenous vein graft [SVG] and radial artery [RA]) grafts (n [ 1,659). In patients who underwent BITA grafting, the left internal thoracic artery (LITA) was grafted mainly to the left anterior descending artery, whereas the right internal thoracic artery (RITA) was used more commonly to graft the circumflex (Cx) artery and the right coronary system as T-grafts. A total of 485 pairs of patients were matched using propensity scores. Cox proportional hazard models were generated to examine the association of arterial BITA grafting with mortality. Results. Patients in the BITA group were more likely to be younger (BITA, 63.7 ± 9.1 versus SITA, 65.0 ± 9.9;
p < 0.0001). At 30 days, patients who underwent BITA grafting experienced reduced unadjusted mortality (BITA, 1.2% versus SITA, 4.4%; p < 0.0001). At 10 years, patients who underwent BITA grafting experienced superior unadjusted survival (BITA, 82.6% ± 1.8% versus SITA, 76.1% ± 1.3%; p [ 0.001). Cox regression analysis in the entire study cohort showed that BITA grafting was associated with improved survival (hazard ratio [HR], 0.71; 95% confidence interval [CI], 0.58–0.87; p < 0.001). In the propensity-score–adjusted analysis, patients who underwent BITA grafting had similar in-hospital mortality (BITA, 1.6% versus SITA, 2.9%; p [ 0.196). Patients who underwent BITA grafting still showed improved survival at 10 years (BITA, 81.0% ± 4.1% versus SITA, 71.8% ± 2.5%; p [ 0.039). Conclusions. This study suggests that coronary artery operations exclusively with BITA grafting in a T-configuration may be associated with better long-term survival than grafting with SITA plus other types of conduits.
T
only randomized controlled trial to date evaluating longterm survival (the Arterial Revascularization Trial) is still ongoing [5]. Tector and colleagues [6, 7] popularized myocardial revascularization exclusively with BITA grafts in a composite T-configuration. In an 8.5-year follow-up report on 897 patients operated on with this technique, they demonstrated the safety and midterm benefit of the procedure. The purpose of this study was to evaluate our experience with myocardial revascularization exclusively with BITA grafts in a T-configuration and compare it with a group of patients receiving SITA plus other types of conduits, ie, SVG or radial artery (RA) grafts, or both.
he use of a single internal thoracic artery (SITA) to revascularize the left anterior descending (LAD) artery as part of coronary artery bypass grafting (CABG) is associated with improved survival, improved quality of life, and a lower incidence of late cardiac events compared with the use of saphenous vein grafts (SVGs) [1]. To further improve the outcomes after surgical myocardial revascularization, other arterial conduits, especially the right internal thoracic artery (RITA), have been used with favorable results [2, 3]. Although increasing evidence from observational cohort studies supports a survival advantage for bilateral internal thoracic artery (BITA) relative to SITA revascularization [4], there remains doubt about whether BITA grafting is the better choice for patients in the long term, because the
Accepted for publication Oct 26, 2015. Address correspondence to Dr Navia, Blanco Encalada 1543, C1428DCO, Buenos Aires, Argentina; email: [email protected].
Ó 2016 by The Society of Thoracic Surgeons Published by Elsevier
(Ann Thorac Surg 2016;101:1775–81) Ó 2016 by The Society of Thoracic Surgeons
Patients and Methods This is a retrospective analysis of prospectively gathered data over a period of 18 years of all patients undergoing exclusively CABG at the Institute Cardiovascular of Buenos Aires. This entire group of patient represents all the experience in myocardial revascularization of our 0003-4975/$36.00 http://dx.doi.org/10.1016/j.athoracsur.2015.10.074
ADULT CARDIAC
Myocardial Revascularization Exclusively With Bilateral Internal Thoracic Arteries in T-Graft Configuration: Effects on Late Survival
ADULT CARDIAC
1776
NAVIA ET AL MYOCARDIAL REVASCULARIZATION WITH BITA T-GRAFS
Abbreviations and Acronyms BITA CABG Cx Dg DSWI HR IQR LAD LITA OR PCI RA RCA RITA SITA SVG
= = = = = = = = = = = = = = = =
bilateral internal thoracic artery coronary artery bypass grafting circumflex coronary artery diagonal artery deep sternal wound infection hazard ratio interquartile range left anterior descending left internal thoracic artery odds ratio percutaneous coronary intervention radial artery right coronary artery right internal thoracic artery single internal thoracic artery saphenous vein graft
Ann Thorac Surg 2016;101:1775–81
identify independent predictors for the incidence of in-hospital mortality. Long-term survival and events were assessed by direct communication with the patient, the family, and attending physicians; medical records were also reviewed. The interview investigated survival, symptoms, long-term medical management, and the incidence of new acute myocardial infarction or the need for percutaneous coronary intervention (PCI) (or both). The institutional ethical committee approved the trial, and surgical consent was obtained from each patient with respect to surgical method and postoperative evaluations.
Surgical Techniques
cardiac surgery department and includes a historical overview of the different surgical techniques used over time. From November 1996 through May 2014, a total of 3,757 consecutive patients who were scheduled for urgent or elective CABG fulfilled the inclusion criteria. We included patients undergoing surgical procedures since 1996, because this coincided with the introduction of our institutional database. Of these cases, 2,098 (55.8%) patients underwent myocardial revascularization exclusively with BITA grafting in a T-configuration and 1,659 (44.2%) received a SITA graft plus other types of conduits (SVG or RA graft, or both). These 2 groups of patients represent the core of the study. For further analysis, the SITA group was divided into 2 subgroups: (1) patients who underwent operation with LITA and RA with or without an additional SVG (n ¼ 1,242) and (2) patients receiving a LITA supplemented by SVG only (n ¼ 388). All the patient data were prospectively collected using our custom-made database (Microsoft Access; Microsoft Corp, Redmond, WA), which is used daily for clinical data management. Preoperative, operative, and postoperative data were obtained by retrospective review of clinical and pathologic reports from the database and were cross-checked with all medical charts. Patients were included for analysis if they had 2- or 3vessel coronary artery disease and received at least 1 ITA graft in situ. Complete arterial revascularization was attempted in all cases and was defined as all diseased coronary systems (stenosis > 70%) receiving at least 1 (artery or vein) graft insertion. This relationship (graft insertion/lesion >70%) was evaluated for the LAD, Cx, and RCA coronary artery systems in all patients.
This study represents an evolving process of how we performed myocardial revascularization over 18 years in our department. In the early part of the experience, we used SITA grafts plus SVGs on pump until 1998 when we introduced the RA graft for the non-LAD system. In early 2002 and with the introduction of the available commercial stabilizer, we started using the same surgical strategy with an off-pump technique. Finally, from 2003 until the present, we have performed off-pump BITA grafting in a T-configuration. ITAs have been dissected as skeletonized conduits since 2003. In the BITA group, the LITA was anastomosed to the LAD artery and the RITA was connected as a sequential T-graft to the Cx artery and the distal right coronary artery (RCA). RITA anastomosis was localized in the anterior aspect of the attached LITA at the level of the left atrial appendage [8]. The LITA was used to graft the LAD territory in 2,090 patients (98.8%) and the diagonal (Dg) artery in 198 patients (9.5%) in sequential grafts. In 8 patients, the LITA was used in a non-LAD territory. The RITA was used from the LITA (T-graft) in 2,023 patients (96.4%) and in situ and as a free graft in the remaining patients. Altogether, 2,098 patients received the RITA as part of a BITA strategy. Considering 2,960 sequential anastomoses, a total of 4,271 distal anastomoses were performed with the RITA—to the Dg artery in 416 patients (19.8%), to the Cx artery in 1,982 patients (94.4%), and to the RCA in 1,270 patients (60.4%). In the SITA group, the LITA was grafted to the LAD artery in all patients; the RA was connected to the Cx artery in 1,141 patients (68.8%) and to the RCA in 107 patients (6.5%). The RA was used mainly as a composite T-configuration from the side of the LITA. Finally, an SVG was used from the aorta in 1,184 patients (71.3%)—to the Cx artery in 396 patients (23.8%) and to the RCA in 610 patients (36.7%). The quality of the anastomosis was assessed by transit-time flow measurement and Doppler velocity probes (Medistim VeriQ system; Medistem, Oslo, Norway).
Study End Points
Propensity-Score Matching and Statistical Analysis
We examined early outcomes: in-hospital mortality, deep sternal wound infection (DSWI), postoperative myocardial infarction, postoperative stroke, reoperation for bleeding, and acute renal impairment (dialysis required). Multivariate analysis was performed to
Preoperative characteristics of patients in these study groups were summarized as mean standard deviation, median and interquartile range (IQR), or prevalence (percentage), as appropriate. Student t tests for independent samples or Mann-Whitney U tests for continuous
1777
NAVIA ET AL MYOCARDIAL REVASCULARIZATION WITH BITA T-GRAFS
variables and c2 tests for categorical variables were applied to examine differences between groups. A propensity-score–matched sample was constructed. A propensity score for having BITA grafting was calculated for each patient using a logistic regression model that included all the preoperative variables listed in Table 1. Patients were matched 1:1 by the propensity scores using the greedy matching technique without replacement. A nearest-neighbor–matching algorithm was used with a caliper distance of 0.0001. Outcomes of interest between the matched groups were compared using the paired t test for continuous variables and the McNemar test for categorical variables. After matching we examined the balance of all observed covariates, interactions among all covariates, and quadratic terms of all covariates. Nearly no imbalances remained as assessed through univariate and multivariate tests. The overall c2 balance test (Hansen and Bowers) was also not significant (c2 [degrees of freedom: 21] ¼ 7.071; p ¼ 0.998) [9]. The multivariate imbalance measure (Iacus, King, and Porro, L1) was larger in the unmatched sample (0.861) than
in the matched sample (0.641), also indicating that matching improved overall balance [10]. Event-free survival curves were estimated using the Kaplan-Meier method. The log-rank test was used to assess the differences in survival between the 2 groups. Univariate and multivariate Cox proportional hazard analyses were performed to investigate the significant predictors of late mortality. The variables used for univariate analysis were the clinical variables listed in Table 1. Variables with p less than 0.2 on univariate analysis were included in the multivariate model.
Results The preoperative clinical profile of the study population is presented in Table 1. Patients in the BITA group were more likely to be younger (BITA, 63.7 9.1 years versus SITA, 65.0 9.9 years; p < 0.0001) and to be men. Intraoperatively, the off-pump technique was less frequently used in patients undergoing SITA grafting (p < 0.0001). Early postoperative outcomes are presented in
Table 1. Preoperative Clinical Profile and Early Postoperative Outcomes in Unmatched Group of Patients BITA (n ¼ 2,098)
SITA (n ¼ 1,659)
p Value
Age (y, SD) Female sex Diabetes Hypertension Dyslipidemia Smoking habit Family history Elective operation On-pump operation Left ventricular dysfunction (moderate/severe) Left main coronary artery disease Three-vessel disease (%) Redo operation Previous MI Previous percutaneous coronary intervention Peripheral vascular disease Carotid artery disease Abdominal aortic aneurysm Chronic obstructive pulmonary disease Cerebrovascular disease Previous renal dysfunction
63.7 9.1 205 (9.8) 544 (25.9) 1,605 (76.5) 1,668 (79.5%) 1,363 (65) 592 (28.2%) 1,385 (66) 8 (0.4) 313 (15) 472 (22) 1,721 (82) 13 (0.6) 605 (28) 444 (21) 62 (3) 102 (4.9) 20 (1) 88 (4.2) 80 (3.8) 88 (4.2)
65 9.9 261 (15) 434 (26) 1,157 (69) 1,085 (65) 868 (52) 153 (9) 1,064 (64) 1,014 (61) 126 (8) 325 (19) 1,293 (77.9) 103 (6.2) 490 (29) 306 (18) 83 (5) 48 (2.9) 21 (1.3) 72 (4.3) 50 (3) 49 (3)
0.000 0.000 0.873 0.000 0.000 0.000 0.000 0.230 0.000 0.000 0.030 0.002 0.000 0.640 0.038 0.001 0.002 0.360 0.826 0.183 0.130
Unadjusted Data Early Outcomes
BITA (n ¼ 2,098) (%)
SITA (n ¼ 1,659) (%)
p Value
1.2 1.9 1.0 0.5 2.1 0.7
4.4 1.6 3.2 0.7 2.4 1.3
0.000 0.521 0.000 0.325 0.586 0.06
Unadjusted-Risk Data
30-day mortality Deep sternal wound infection Postoperative MI Postoperative stroke Reoperation for bleeding Renal impairment requiring dialysis BITA ¼ bilateral internal thoracic artery;
MI ¼ myocardial infarction;
SD ¼ standard deviation;
SITA ¼ single internal thoracic artery.
ADULT CARDIAC
Ann Thorac Surg 2016;101:1775–81
ADULT CARDIAC
1778
NAVIA ET AL MYOCARDIAL REVASCULARIZATION WITH BITA T-GRAFS
Ann Thorac Surg 2016;101:1775–81
Table 1. Unadjusted in-hospital mortality was lower in patients who underwent BITA grafting (1.2% versus 4.4%; p < 0.0001). In multiple logistic regression analyses, significant predictors for in-hospital mortality were age (odds ratio [OR], 1.08; 95% confidence interval [CI], 1.05– 1.11; p < 0.0001) and on-pump operations (OR, 2.43; 95% CI, 1.47–4.01; p < 0.001). Elective operations were identified as a protector for hospital mortality (OR, 0.42; 95% CI, 0.26–0.68; p < 0.001). Follow-up of hospital survivors (n ¼ 3,658) was 95.0% complete: BITA grafting, 1,963 (94.7%) patients and SITA grafting, 1,513 (95.4%) patients. Median follow-up of all patients was 2,016 days (IQR, 935–3,223 days). Median follow-up in the BITA group was 1,464 days (IQR, 691– 2,437 days). For the SITA group, the median follow-up was 3,096 days (IQR, 1,455–4,337 days; p < 0.001). There was no difference in the proportion of patients who completed follow up: BITA: 94.7% versus SITA: 95.4% (p ¼ 0.365). During follow-up, there were 405 (11.1%) late deaths. Figure 1A displays the postoperative outcomes of patients at 5 and 10 years for the full unmatched patient population. Postoperative survival was 92.4% 0.7% and 82.6% 1.8%, respectively, in the BITA group and 89.1% 0.8% and 76.1% 1.3%, respectively, in the SITA group (log-rank p ¼ 0.001). The hazard ratio [HR] was 0.71; the 95% CI was 0.581–0.869, and p was less than 0.001 for the BITA group (Table 3). The HR was 1.457; the 95% CI was 1.182-1.796, and p was less than 0.001 for the SITA group. Overall, the BITA group had a significantly better longterm survival than did the SITA group. In the subanalysis of the SITA group at 10 years, patients who
underwent BITA grafting had a superior survival compared with the group of patients with SITA supplemented by SVG only. Also the first subgroup (LITA þ RA þ SVG) had significantly better long-term survival than the second subgroup (LITA-SVG) (log-rank p < 0.0001). Finally, patients in the first SITA subgroup had almost identical long-term survival as patients who underwent BITA grafting (log-rank p ¼ 0.055); however, there was a superior effect gained by using BITA over LITA-RA-SVG at 10 years, expressed by Cox regression analysis (BITA: HR, 0.48, 95% CI, 0.37–0.63 versus first SITA subgroup: HR, 0.60; 95% CI, 0.47–0.78) (Fig 1B). Patients who underwent BITA grafting had a superior postoperative survival period free of the need for PCI compared with the SITA group: 94 0.5% versus 90 0.7% (p < 0.001; HR, 0.574; 95% CI, 0.442–0.744). Table 2 displays the patient characteristics of the propensity-score–matched population. In-hospital mortality occurred in 1.6% of the BITA group and in 2.9% of the SITA group (p ¼ 0.19). During follow-up, there were 139 (14.3%) late deaths (BITA, 38 [7.8%] patients versus SITA, 101 (20.8%) patients; p < 0.001). Figure 2 displays the postoperative patient outcomes at 5 and 10 years for the propensity-score–matched group. Postoperative survival was 92.1% 1.5% and 81.0% 4.1%, respectively, in the BITA group and 87.9% 1.6% and 71.8% 2.5% in the SITA group (log-rank p ¼ 0.039). Overall, in this propensity-score–matched population, the use of BITAs had significantly better long-term outcomes than the use of SITAs. Completeness of revascularization was higher in the BITA group than in the SITA group: the mean
Fig 1. (A) Full-unadjusted 10-year survival. (B) Full unadjusted 10-year survival comparing patients who underwent BITA grafting with the 2 SITA subgroups: first subgroup, ITA þ RA SVG and second subgroup, SITA þ SVG (BITA ¼ bilateral internal thoracic arteries; ITA ¼ internal thoracic artery; RA ¼ radial artery; SITA ¼ single internal thoracic artery; SVG ¼ saphenous vein graft.)
NAVIA ET AL MYOCARDIAL REVASCULARIZATION WITH BITA T-GRAFS
1779
Table 2. Preoperative Clinical Profile and Early Postoperative Outcomes of Propensity-Score–Matched Cases Adjusted-Risk Data
BITA (n ¼ 485)
SITA (n ¼ 485)
Age (y, SD) Female sex Diabetes Hypertension Dyslipidemia Smoking habits (current or former smoker) Family history Elective operation On-pump operation Left ventricular dysfunction (moderate/severe) Left main coronary artery disease Three-vessel disease Redo operation Previous MI Previous percutaneous coronary intervention Peripheral vascular disease Carotid artery disease Abdominal aortic aneurysm Chronic obstructive pulmonary disease Cerebrovascular disease Previous renal dysfunction
65.4 66 131 369 355 291 122 301 8 74 102 389 2 126 82 22 13 6 23 11 17
65.5 67 141 370 347 280 115 303 8 76 101 395 2 125 85 11 15 5 22 15 23
Adjusted Data Early Outcomes
BITA (n ¼ 485) (%)
SITA (n ¼ 485) (%)
p Value
1.6 2.3 1.4 0.6 2.9 0.8
2.9 1.4 1.9 1.0 1.6 0.8
0.196 0.341 0.614 0.478 0.196 1.000
30-day mortality Deep sternal wound infection Postoperative MI Postoperative stroke Reoperation for bleeding Renal impairment requiring dialysis BITA ¼ bilateral internal thoracic artery;
MI ¼ myocardial infarction;
number of arterial bypass grafts was 3.2 0.6 versus 2.9 0.7, respectively (p < 0.001). Our surgical strategy has been always to perform a complete myocardial revascularization, and considering that in both groups, 20% of the patients had 2-vessel disease, this strategy was achieved in the total group of patients.
Comment This single-center analysis directly compares a strategy of myocardial revascularization exclusively with BITAs as composite T-grafts (Tector’s technique) with a more “traditional” approach of SITA grafts supplemented by RA grafts or SVGs and confirms better long-term survival in patients receiving BITA grafts [6]. BITA grafting in CABG has been associated with improved long-term survival and freedom from late cardiac events. The advantages and disadvantages of the BITA technique have been rigorously studied through multiple cohort studies. A meta-analysis of 7 observational studies that included 15,962 patients (11,269 LITA
(8.9) (13) (27) (76) (73) (60) (25) (62) (1.6) (15) (21) (80.2) (0.4) (26) (16.5) (4.5) (2.7) (1.2) (4.7) (2.3) (3.5)
SD ¼ standard deviation;
(9.5) (14) (29) (76) (71) (57) (23) (62) (1.6) (15.7) (20.8) (81.4) (0.4) (25.8) (17.5) (2.3) (3.1) (1.0) (4.5) (3.1) (4.7)
p Value 0.948 0.926 0.475 0.940 0.566 0.473 0.601 0.895 10.000 0.859 0.937 0.625 10.000 0.942 0.799 0.051 0.701 0.762 0.879 0.427 0.382
SITA ¼ single internal thoracic artery.
and 4,693 BITA procedures) performed in 2001 by Taggart and coworkers [11] demonstrated significantly better survival in patients who underwent BITA grafting compared with patients who underwent a LITA procedure (HR, 0.81; 95% CI, 0.70–0.94). This sparked an even greater number of nonrandomized studies. In an updated meta-analysis, Weiss and colleagues [12] demonstrated an increase in long-term survival in patients receiving BITAs as a primary grafting strategy over those receiving LITAs. In our study, an off-pump technique was performed in more than 95% of the patients who underwent BITA grafting, with lower hospital mortality compared with patients who underwent SITA grafting. Adding off-pump techniques to myocardial revascularization exclusively with a BITA technique creates a more technically demanding procedure. However, we consider that this surgical approach is more patient friendly: 88% of patients were extubated in the operating room [13], and the use of a no-touch aortic technique in the off-pump group significantly decreased the incidence of
ADULT CARDIAC
Ann Thorac Surg 2016;101:1775–81
ADULT CARDIAC
1780
NAVIA ET AL MYOCARDIAL REVASCULARIZATION WITH BITA T-GRAFS
Table 3. Cox Proportional Hazard Regression Analysis for Long-Term (10 Years) All-Cause Mortality Covariate
HR
95% CI
p Value
Age (y) Diabetes mellitus Cerebrovascular disease Previous renal dysfunction Smoking habit Elective operation Left ventricular dysfunction (moderate/severe) BITA
1.07 1.69 2.16 2.12 1.47 0.78 2.47
1.06–1.08 1.39–2.06 1.49–3.11 1.58–2.85 1.21–1.78 0.64–0.94 1.92–3.19
0.000 0.000 0.000 0.000 0.000 0.009 0.000
0.71
0.58–0.87
0.000
BITA ¼ bilateral internal thoracic arteries; HR ¼ hazard ratio.
CI ¼ confidence interval;
postoperative stroke [14]. Finally, the possibility to perform BITA CABG with a single skin incision (no other conduit harvesting site) and with a short hospital stay (mean 4.8 days in elective cases) is better for the patient. Concerns about DSWI have discouraged the use of BITA grafting in the general population, particularly in patients with diabetes; however, in this study, there was no difference between the groups. It has been suggested that the method of ITA harvesting (skeletonized) and maintaining strict glycemic control during the entire procedure and hospitalization influences the incidence of postoperative DSWI [15]. There were no differences in the early hospital outcomes in the propensity-matched group. At 10 years, patients who underwent BITA grafting experienced greater unadjusted and adjusted survival than patients who underwent SITA
Fig 2. Ten-year survival among 485 propensity-score–matched patient pairs (BITA ¼ bilateral internal thoracic arteries; SITA ¼ single internal thoracic artery.)
Ann Thorac Surg 2016;101:1775–81
grafting. There have been other reports of excellent perioperative outcomes with good late survival using on-pump full myocardial revascularization purely with BITA composite grafts [16]. Long-term survival after CABG may be related to the late patency of conduits, and this may partially explain the improved survival seen in the patients who underwent BITA grafting. It is well known that the ITA is more resistant to atherosclerosis because of the endothelium’s capability of producing nitric oxide, which, together with its small diameter that creates better size matching with target vessels, may contribute to its durability [17, 18]. In a large multicenter study, Buxton and associates [19] suggested that a strategy of total arterial revascularization with multiple RA grafts as a second arterial conduit is associated with improved long-term survival compared with the use of SITA grafts and SVGs. The results of our previous study provide evidence for the superiority of the RITA graft compared with RA grafts as a second conduit in multiple arterial grafting [20]. However, in this study, patients in the SITA group with RA grafts plus SVGs had better long-term survival than patients with LITA-SVGs only, raising the speculation that RA grafts may indeed provide a prognostic benefit. Also, patients in the first SITA subgroup had long-term survival at 10 years almost identical to that of the BITA group; however, BITA grafting still shows greater beneficial effects on survival, as expressed in Cox proportional hazards regression analysis. The proposed superiority of arterial grafts may lie in their effect on the native circulation compared with vein grafts. The supraphysiological flow provided by vein grafts may create turbulence that promotes progression of atherosclerotic disease [21]. The use of off-pump ITA T-grafts with multiple sequential side-to-side ITA to coronary artery anastomoses introduces the concept of coronary arterial tree reconstruction by the most appropriate conduit; yet the procedure is more technically demanding [22]. Using this surgical approach, the total coronary bypass flow depends on the flow of the proximal LITA; this scenario has led to concern about whether coronary flow reserve in the LITA is sufficient to supply more than 1 coronary anastomosis. Several reports have already concluded that ITA T-grafts allow complete myocardial revascularization with good perioperative results and that the coronary flow reserve of the proximal LITA is adequate for multiple coronary anastomoses [23]. Regarding T-graft configuration and competitive flow, Glineur and coworkers [24, 25] stated that the resistance is similar in both branches of the graft, which excludes the possibility of steal from 1 branch to the other during periods of high myocardial blood flow demand. However, the longer the arterial assembly, the lower the pressure at the distal anastomosis, considering that this phenomenon is more frequent at the right coronary anastomosis of composite T-grafts because of the lower pressure in the distal portion of the T-branch. In our series of postoperative angiograms, the patency rate of the distal graft to the RCA/posterior descending artery was 91%, and only 8
1781
NAVIA ET AL MYOCARDIAL REVASCULARIZATION WITH BITA T-GRAFS
patients presented with competitive flow in the RITA to the right coronary artery or its branches. In these patients, the opacification of the native coronary arteries, visualized by retrograde flow, showed perfect patency and good diameter of the RITA graft. This scenario was more frequent in the case of a large dominant RCA. However, these angiographic abnormalities had no postoperative consequences in our group of patients [8]. There are several study limitations to mention. First, this is a retrospective study, and although the propensityscore–matching algorithm produced rather comparable groups, the study was not randomized, and we cannot rule out additional effects of missing covariates. Second, the study period was long, and most of the SITA procedures were done in the earlier part of the study because our experience in BITA grafting came later; this issue may have played a role in the final results. As was mentioned earlier, we started in 1996 using SITA plus SVGs on pump until 1998 when we introduced the use of RA grafts for the non–LAD artery system, and from 2003 until the present we have performed off-pump BITA grafting in a T-graft configuration. All the staff surgeons in the department have adopted these changes in the surgical technique over time. If we had included the variable “time,” we would have obtained a matched sample that would have included all the late SITA and early BITA procedures, so we would not have really obtained 2 comparable groups, which was necessary to answer the objective of the study. In conclusion, this single-center study suggests that myocardial revascularization exclusively with BITA in a T-graft configuration and an off-pump technique can be safely performed with low in-hospital mortality and may be associated with improved long-term survival compared with the use of SITA grafts plus other types of conduits.
8. Navia D, Vrancic M, Vaccarino G, et al. Total arterial offpump coronary revascularization using bilateral internal thoracic arteries in triple-vessel disease: surgical technique and clinical outcomes. Ann Thorac Surg 2008;86: 524–30. 9. Hansen BB, Bowers J. Covariate balance in simple, stratified and clustered comparative studies. Statistical Science 2008;23:219–36. 10. Iacus SM, King G, Porro G. CEM: Coarsened exact matching software. Journal of Statistical Software 2009;30:1–27. 11. Taggart DP, D’Amico R, Altman DG. Effect of arterial revascularization on survival: a systematic review of studies comparing bilateral and single internal mammary arteries. Lancet 2001;358:870–5. 12. Weiss A, Zhao S, Tian D, Taggart D, Yan T. A meta-analysis comparing bilateral internal mammary artery with left internal mammary artery for coronary artery bypass grafting. Ann Cardiothorac Surg 2013;2:390–400. 13. Dorsa A, Rossi A, Thierer J, et al. Extubation after off-pump coronary artery bypass graft surgery in 1,196 consecutive patients: feasibility, safety and predictors of when not to attempt it. J Cardiothorac Vasc Anesth 2011;25:431–6. 14. Moss E, Puskas J, Thourani V, et al. Avoiding aortic clamping during coronary artery bypass grafting reduces postoperative stroke. J Thorac Cardiovasc Surg 2015;149:175–80. 15. Barros de Oliveira Saa M, Ferraza P, Escobara R, et al. Skeletonized versus pedicled internal thoracic artery and risk of sternal wound infection after coronary bypass surgery: meta-analysis and meta-regression of 4817 patients. Interact Cardiovasc Thorac Surg 2013;16:849–57. 16. Paterson H, Naidoo R, Byth K, Chen C, Denniss R. Full myocardial revascularization with bilateral internal mammary artery Y grafts. Ann Cardiothorac Surg 2013;2:444–52. 17. Otsuka F, Yahagi K, Sakakura K, Virmani R. Why is the mammary artery so special and what protects it from atherosclerosis? Ann Cardiothorac Surg 2013;2:519–26. 18. Locker C, Schaff HV, Dearani JA, et al. Multiple arterial grafts improve late survival of patients undergoing coronary artery bypass graft surgery: analysis of 8622 patients with multivessel disease. Circulation 2012;126:1023–30. 19. Buxton BF, Shi WY, Tatoulis J, Fuller JA, Rosalion A, Hayward PA. Total arterial revascularization with internal thoracic and radial artery grafts in triple-vessel coronary artery disease is associated with improved survival. J Thorac Cardiovasc Surg 2014;148:1238–43. 20. Navia D, Vrancic M, Piccinini F, et al. Is the second internal thoracic artery better than the radial artery in total arterial off-pump coronary artery bypass grafting? A propensity score-matched follow-up study. J Thorac Cardiovasc Surg 2014;147:632–8. 21. Dimitrova KR, Hoffman DM, Geller CM, Dincheva G, Ko W, Tranbaugh RF. Arterial grafts protect the native coronary vessels from atherosclerotic disease progression. Ann Thorac Surg 2012;94:475–81. 22. Tarrío RF, Cuenca JJ, Gomes V, et al. Off-pump total arterial revascularization: our experience. J Card Surg 2004;19: 389–95. 23. Wendler O, Hennen B, Markwirth T, et al. T grafts with the right internal thoracic artery to the left internal thoracic artery versus the left internal thoracic and radial artery: flow dynamics in the internal thoracic artery main stem. J Thorac Cardiovasc Surg 1999;118:841–8. 24. Glineur D, Hanet C. Competitive flow in coronary bypass surgery: is it a problem? Curr Opin Cardiol 2012;27:620–8. 25. Glineur D, Hanet C, D’hoore W, et al. Causes of nonfunctioning right internal mammary used in a Y-graft configuration: insight from a 6-month systematic angiographic trial. Eur J Cardiothorac Surg 2009;36:129–35; discussion 135–6.
References 1. Loop FD, Lytle BW, Cosgrove DM, et al. Influence of the internal-mammary-artery graft on 10-year survival and other cardiac events. N Engl J Med 1986;314:1–6. 2. Barner HB. Double internal mammary-coronary artery bypass. Arch Surg 1974;109:627–30. 3. Fiore A, Naunheim K, Dean P, et al. Results of internal thoracic artery grafting over 15 years: single versus double grafts. Ann Thorac Surg 1990;49; 202–2. 4. Lytle BW, Blackstone EH, Loop FD, et al. Two internal thoracic arteries are better than one. J Thorac Cardiovasc Surg 1999;117:855–72. 5. Taggart DP, Altman DG, Gray AM, et al, ART Investigators. Randomized trial to compare bilateral versus single internal mammary coronary artery bypass grafting: 1-year results of the Arterial Revascularisation Trial (ART). Eur Heart J 2010;31:2470–81. 6. Tector AJ, Amundsen S, Schmahl TM, et al. Total revascularization with T grafts. Ann Thorac Surg 1994;57:33–8; discussion 39. 7. Tector AJ, McDonald ML, Kress DC, et al. Purely internal thoracic artery grafts: outcomes. Ann Thorac Surg 2001;72: 450–5.
ADULT CARDIAC
Ann Thorac Surg 2016;101:1775–81
Background. We studied long-term survival using bilateral internal thoracic artery (BITA) grafting in a T-configuration exclusively versus using single internal thoracic artery (SITA) grafting in patients with multivessel disease. Methods. Consecutive coronary operations performed at a single center between 1996 and 2014 were reviewed. Long-term survival among patients receiving coronary revascularization exclusively with BITA grafting in a T-configuration (n [ 2,098) versus SITA grafts plus other types of conduits (saphenous vein graft [SVG] and radial artery [RA]) grafts (n [ 1,659). In patients who underwent BITA grafting, the left internal thoracic artery (LITA) was grafted mainly to the left anterior descending artery, whereas the right internal thoracic artery (RITA) was used more commonly to graft the circumflex (Cx) artery and the right coronary system as T-grafts. A total of 485 pairs of patients were matched using propensity scores. Cox proportional hazard models were generated to examine the association of arterial BITA grafting with mortality. Results. Patients in the BITA group were more likely to be younger (BITA, 63.7 ± 9.1 versus SITA, 65.0 ± 9.9;
p < 0.0001). At 30 days, patients who underwent BITA grafting experienced reduced unadjusted mortality (BITA, 1.2% versus SITA, 4.4%; p < 0.0001). At 10 years, patients who underwent BITA grafting experienced superior unadjusted survival (BITA, 82.6% ± 1.8% versus SITA, 76.1% ± 1.3%; p [ 0.001). Cox regression analysis in the entire study cohort showed that BITA grafting was associated with improved survival (hazard ratio [HR], 0.71; 95% confidence interval [CI], 0.58–0.87; p < 0.001). In the propensity-score–adjusted analysis, patients who underwent BITA grafting had similar in-hospital mortality (BITA, 1.6% versus SITA, 2.9%; p [ 0.196). Patients who underwent BITA grafting still showed improved survival at 10 years (BITA, 81.0% ± 4.1% versus SITA, 71.8% ± 2.5%; p [ 0.039). Conclusions. This study suggests that coronary artery operations exclusively with BITA grafting in a T-configuration may be associated with better long-term survival than grafting with SITA plus other types of conduits.
T
only randomized controlled trial to date evaluating longterm survival (the Arterial Revascularization Trial) is still ongoing [5]. Tector and colleagues [6, 7] popularized myocardial revascularization exclusively with BITA grafts in a composite T-configuration. In an 8.5-year follow-up report on 897 patients operated on with this technique, they demonstrated the safety and midterm benefit of the procedure. The purpose of this study was to evaluate our experience with myocardial revascularization exclusively with BITA grafts in a T-configuration and compare it with a group of patients receiving SITA plus other types of conduits, ie, SVG or radial artery (RA) grafts, or both.
he use of a single internal thoracic artery (SITA) to revascularize the left anterior descending (LAD) artery as part of coronary artery bypass grafting (CABG) is associated with improved survival, improved quality of life, and a lower incidence of late cardiac events compared with the use of saphenous vein grafts (SVGs) [1]. To further improve the outcomes after surgical myocardial revascularization, other arterial conduits, especially the right internal thoracic artery (RITA), have been used with favorable results [2, 3]. Although increasing evidence from observational cohort studies supports a survival advantage for bilateral internal thoracic artery (BITA) relative to SITA revascularization [4], there remains doubt about whether BITA grafting is the better choice for patients in the long term, because the
Accepted for publication Oct 26, 2015. Address correspondence to Dr Navia, Blanco Encalada 1543, C1428DCO, Buenos Aires, Argentina; email: [email protected].
Ó 2016 by The Society of Thoracic Surgeons Published by Elsevier
(Ann Thorac Surg 2016;101:1775–81) Ó 2016 by The Society of Thoracic Surgeons
Patients and Methods This is a retrospective analysis of prospectively gathered data over a period of 18 years of all patients undergoing exclusively CABG at the Institute Cardiovascular of Buenos Aires. This entire group of patient represents all the experience in myocardial revascularization of our 0003-4975/$36.00 http://dx.doi.org/10.1016/j.athoracsur.2015.10.074
ADULT CARDIAC
Myocardial Revascularization Exclusively With Bilateral Internal Thoracic Arteries in T-Graft Configuration: Effects on Late Survival
ADULT CARDIAC
1776
NAVIA ET AL MYOCARDIAL REVASCULARIZATION WITH BITA T-GRAFS
Abbreviations and Acronyms BITA CABG Cx Dg DSWI HR IQR LAD LITA OR PCI RA RCA RITA SITA SVG
= = = = = = = = = = = = = = = =
bilateral internal thoracic artery coronary artery bypass grafting circumflex coronary artery diagonal artery deep sternal wound infection hazard ratio interquartile range left anterior descending left internal thoracic artery odds ratio percutaneous coronary intervention radial artery right coronary artery right internal thoracic artery single internal thoracic artery saphenous vein graft
Ann Thorac Surg 2016;101:1775–81
identify independent predictors for the incidence of in-hospital mortality. Long-term survival and events were assessed by direct communication with the patient, the family, and attending physicians; medical records were also reviewed. The interview investigated survival, symptoms, long-term medical management, and the incidence of new acute myocardial infarction or the need for percutaneous coronary intervention (PCI) (or both). The institutional ethical committee approved the trial, and surgical consent was obtained from each patient with respect to surgical method and postoperative evaluations.
Surgical Techniques
cardiac surgery department and includes a historical overview of the different surgical techniques used over time. From November 1996 through May 2014, a total of 3,757 consecutive patients who were scheduled for urgent or elective CABG fulfilled the inclusion criteria. We included patients undergoing surgical procedures since 1996, because this coincided with the introduction of our institutional database. Of these cases, 2,098 (55.8%) patients underwent myocardial revascularization exclusively with BITA grafting in a T-configuration and 1,659 (44.2%) received a SITA graft plus other types of conduits (SVG or RA graft, or both). These 2 groups of patients represent the core of the study. For further analysis, the SITA group was divided into 2 subgroups: (1) patients who underwent operation with LITA and RA with or without an additional SVG (n ¼ 1,242) and (2) patients receiving a LITA supplemented by SVG only (n ¼ 388). All the patient data were prospectively collected using our custom-made database (Microsoft Access; Microsoft Corp, Redmond, WA), which is used daily for clinical data management. Preoperative, operative, and postoperative data were obtained by retrospective review of clinical and pathologic reports from the database and were cross-checked with all medical charts. Patients were included for analysis if they had 2- or 3vessel coronary artery disease and received at least 1 ITA graft in situ. Complete arterial revascularization was attempted in all cases and was defined as all diseased coronary systems (stenosis > 70%) receiving at least 1 (artery or vein) graft insertion. This relationship (graft insertion/lesion >70%) was evaluated for the LAD, Cx, and RCA coronary artery systems in all patients.
This study represents an evolving process of how we performed myocardial revascularization over 18 years in our department. In the early part of the experience, we used SITA grafts plus SVGs on pump until 1998 when we introduced the RA graft for the non-LAD system. In early 2002 and with the introduction of the available commercial stabilizer, we started using the same surgical strategy with an off-pump technique. Finally, from 2003 until the present, we have performed off-pump BITA grafting in a T-configuration. ITAs have been dissected as skeletonized conduits since 2003. In the BITA group, the LITA was anastomosed to the LAD artery and the RITA was connected as a sequential T-graft to the Cx artery and the distal right coronary artery (RCA). RITA anastomosis was localized in the anterior aspect of the attached LITA at the level of the left atrial appendage [8]. The LITA was used to graft the LAD territory in 2,090 patients (98.8%) and the diagonal (Dg) artery in 198 patients (9.5%) in sequential grafts. In 8 patients, the LITA was used in a non-LAD territory. The RITA was used from the LITA (T-graft) in 2,023 patients (96.4%) and in situ and as a free graft in the remaining patients. Altogether, 2,098 patients received the RITA as part of a BITA strategy. Considering 2,960 sequential anastomoses, a total of 4,271 distal anastomoses were performed with the RITA—to the Dg artery in 416 patients (19.8%), to the Cx artery in 1,982 patients (94.4%), and to the RCA in 1,270 patients (60.4%). In the SITA group, the LITA was grafted to the LAD artery in all patients; the RA was connected to the Cx artery in 1,141 patients (68.8%) and to the RCA in 107 patients (6.5%). The RA was used mainly as a composite T-configuration from the side of the LITA. Finally, an SVG was used from the aorta in 1,184 patients (71.3%)—to the Cx artery in 396 patients (23.8%) and to the RCA in 610 patients (36.7%). The quality of the anastomosis was assessed by transit-time flow measurement and Doppler velocity probes (Medistim VeriQ system; Medistem, Oslo, Norway).
Study End Points
Propensity-Score Matching and Statistical Analysis
We examined early outcomes: in-hospital mortality, deep sternal wound infection (DSWI), postoperative myocardial infarction, postoperative stroke, reoperation for bleeding, and acute renal impairment (dialysis required). Multivariate analysis was performed to
Preoperative characteristics of patients in these study groups were summarized as mean standard deviation, median and interquartile range (IQR), or prevalence (percentage), as appropriate. Student t tests for independent samples or Mann-Whitney U tests for continuous
1777
NAVIA ET AL MYOCARDIAL REVASCULARIZATION WITH BITA T-GRAFS
variables and c2 tests for categorical variables were applied to examine differences between groups. A propensity-score–matched sample was constructed. A propensity score for having BITA grafting was calculated for each patient using a logistic regression model that included all the preoperative variables listed in Table 1. Patients were matched 1:1 by the propensity scores using the greedy matching technique without replacement. A nearest-neighbor–matching algorithm was used with a caliper distance of 0.0001. Outcomes of interest between the matched groups were compared using the paired t test for continuous variables and the McNemar test for categorical variables. After matching we examined the balance of all observed covariates, interactions among all covariates, and quadratic terms of all covariates. Nearly no imbalances remained as assessed through univariate and multivariate tests. The overall c2 balance test (Hansen and Bowers) was also not significant (c2 [degrees of freedom: 21] ¼ 7.071; p ¼ 0.998) [9]. The multivariate imbalance measure (Iacus, King, and Porro, L1) was larger in the unmatched sample (0.861) than
in the matched sample (0.641), also indicating that matching improved overall balance [10]. Event-free survival curves were estimated using the Kaplan-Meier method. The log-rank test was used to assess the differences in survival between the 2 groups. Univariate and multivariate Cox proportional hazard analyses were performed to investigate the significant predictors of late mortality. The variables used for univariate analysis were the clinical variables listed in Table 1. Variables with p less than 0.2 on univariate analysis were included in the multivariate model.
Results The preoperative clinical profile of the study population is presented in Table 1. Patients in the BITA group were more likely to be younger (BITA, 63.7 9.1 years versus SITA, 65.0 9.9 years; p < 0.0001) and to be men. Intraoperatively, the off-pump technique was less frequently used in patients undergoing SITA grafting (p < 0.0001). Early postoperative outcomes are presented in
Table 1. Preoperative Clinical Profile and Early Postoperative Outcomes in Unmatched Group of Patients BITA (n ¼ 2,098)
SITA (n ¼ 1,659)
p Value
Age (y, SD) Female sex Diabetes Hypertension Dyslipidemia Smoking habit Family history Elective operation On-pump operation Left ventricular dysfunction (moderate/severe) Left main coronary artery disease Three-vessel disease (%) Redo operation Previous MI Previous percutaneous coronary intervention Peripheral vascular disease Carotid artery disease Abdominal aortic aneurysm Chronic obstructive pulmonary disease Cerebrovascular disease Previous renal dysfunction
63.7 9.1 205 (9.8) 544 (25.9) 1,605 (76.5) 1,668 (79.5%) 1,363 (65) 592 (28.2%) 1,385 (66) 8 (0.4) 313 (15) 472 (22) 1,721 (82) 13 (0.6) 605 (28) 444 (21) 62 (3) 102 (4.9) 20 (1) 88 (4.2) 80 (3.8) 88 (4.2)
65 9.9 261 (15) 434 (26) 1,157 (69) 1,085 (65) 868 (52) 153 (9) 1,064 (64) 1,014 (61) 126 (8) 325 (19) 1,293 (77.9) 103 (6.2) 490 (29) 306 (18) 83 (5) 48 (2.9) 21 (1.3) 72 (4.3) 50 (3) 49 (3)
0.000 0.000 0.873 0.000 0.000 0.000 0.000 0.230 0.000 0.000 0.030 0.002 0.000 0.640 0.038 0.001 0.002 0.360 0.826 0.183 0.130
Unadjusted Data Early Outcomes
BITA (n ¼ 2,098) (%)
SITA (n ¼ 1,659) (%)
p Value
1.2 1.9 1.0 0.5 2.1 0.7
4.4 1.6 3.2 0.7 2.4 1.3
0.000 0.521 0.000 0.325 0.586 0.06
Unadjusted-Risk Data
30-day mortality Deep sternal wound infection Postoperative MI Postoperative stroke Reoperation for bleeding Renal impairment requiring dialysis BITA ¼ bilateral internal thoracic artery;
MI ¼ myocardial infarction;
SD ¼ standard deviation;
SITA ¼ single internal thoracic artery.
ADULT CARDIAC
Ann Thorac Surg 2016;101:1775–81
ADULT CARDIAC
1778
NAVIA ET AL MYOCARDIAL REVASCULARIZATION WITH BITA T-GRAFS
Ann Thorac Surg 2016;101:1775–81
Table 1. Unadjusted in-hospital mortality was lower in patients who underwent BITA grafting (1.2% versus 4.4%; p < 0.0001). In multiple logistic regression analyses, significant predictors for in-hospital mortality were age (odds ratio [OR], 1.08; 95% confidence interval [CI], 1.05– 1.11; p < 0.0001) and on-pump operations (OR, 2.43; 95% CI, 1.47–4.01; p < 0.001). Elective operations were identified as a protector for hospital mortality (OR, 0.42; 95% CI, 0.26–0.68; p < 0.001). Follow-up of hospital survivors (n ¼ 3,658) was 95.0% complete: BITA grafting, 1,963 (94.7%) patients and SITA grafting, 1,513 (95.4%) patients. Median follow-up of all patients was 2,016 days (IQR, 935–3,223 days). Median follow-up in the BITA group was 1,464 days (IQR, 691– 2,437 days). For the SITA group, the median follow-up was 3,096 days (IQR, 1,455–4,337 days; p < 0.001). There was no difference in the proportion of patients who completed follow up: BITA: 94.7% versus SITA: 95.4% (p ¼ 0.365). During follow-up, there were 405 (11.1%) late deaths. Figure 1A displays the postoperative outcomes of patients at 5 and 10 years for the full unmatched patient population. Postoperative survival was 92.4% 0.7% and 82.6% 1.8%, respectively, in the BITA group and 89.1% 0.8% and 76.1% 1.3%, respectively, in the SITA group (log-rank p ¼ 0.001). The hazard ratio [HR] was 0.71; the 95% CI was 0.581–0.869, and p was less than 0.001 for the BITA group (Table 3). The HR was 1.457; the 95% CI was 1.182-1.796, and p was less than 0.001 for the SITA group. Overall, the BITA group had a significantly better longterm survival than did the SITA group. In the subanalysis of the SITA group at 10 years, patients who
underwent BITA grafting had a superior survival compared with the group of patients with SITA supplemented by SVG only. Also the first subgroup (LITA þ RA þ SVG) had significantly better long-term survival than the second subgroup (LITA-SVG) (log-rank p < 0.0001). Finally, patients in the first SITA subgroup had almost identical long-term survival as patients who underwent BITA grafting (log-rank p ¼ 0.055); however, there was a superior effect gained by using BITA over LITA-RA-SVG at 10 years, expressed by Cox regression analysis (BITA: HR, 0.48, 95% CI, 0.37–0.63 versus first SITA subgroup: HR, 0.60; 95% CI, 0.47–0.78) (Fig 1B). Patients who underwent BITA grafting had a superior postoperative survival period free of the need for PCI compared with the SITA group: 94 0.5% versus 90 0.7% (p < 0.001; HR, 0.574; 95% CI, 0.442–0.744). Table 2 displays the patient characteristics of the propensity-score–matched population. In-hospital mortality occurred in 1.6% of the BITA group and in 2.9% of the SITA group (p ¼ 0.19). During follow-up, there were 139 (14.3%) late deaths (BITA, 38 [7.8%] patients versus SITA, 101 (20.8%) patients; p < 0.001). Figure 2 displays the postoperative patient outcomes at 5 and 10 years for the propensity-score–matched group. Postoperative survival was 92.1% 1.5% and 81.0% 4.1%, respectively, in the BITA group and 87.9% 1.6% and 71.8% 2.5% in the SITA group (log-rank p ¼ 0.039). Overall, in this propensity-score–matched population, the use of BITAs had significantly better long-term outcomes than the use of SITAs. Completeness of revascularization was higher in the BITA group than in the SITA group: the mean
Fig 1. (A) Full-unadjusted 10-year survival. (B) Full unadjusted 10-year survival comparing patients who underwent BITA grafting with the 2 SITA subgroups: first subgroup, ITA þ RA SVG and second subgroup, SITA þ SVG (BITA ¼ bilateral internal thoracic arteries; ITA ¼ internal thoracic artery; RA ¼ radial artery; SITA ¼ single internal thoracic artery; SVG ¼ saphenous vein graft.)
NAVIA ET AL MYOCARDIAL REVASCULARIZATION WITH BITA T-GRAFS
1779
Table 2. Preoperative Clinical Profile and Early Postoperative Outcomes of Propensity-Score–Matched Cases Adjusted-Risk Data
BITA (n ¼ 485)
SITA (n ¼ 485)
Age (y, SD) Female sex Diabetes Hypertension Dyslipidemia Smoking habits (current or former smoker) Family history Elective operation On-pump operation Left ventricular dysfunction (moderate/severe) Left main coronary artery disease Three-vessel disease Redo operation Previous MI Previous percutaneous coronary intervention Peripheral vascular disease Carotid artery disease Abdominal aortic aneurysm Chronic obstructive pulmonary disease Cerebrovascular disease Previous renal dysfunction
65.4 66 131 369 355 291 122 301 8 74 102 389 2 126 82 22 13 6 23 11 17
65.5 67 141 370 347 280 115 303 8 76 101 395 2 125 85 11 15 5 22 15 23
Adjusted Data Early Outcomes
BITA (n ¼ 485) (%)
SITA (n ¼ 485) (%)
p Value
1.6 2.3 1.4 0.6 2.9 0.8
2.9 1.4 1.9 1.0 1.6 0.8
0.196 0.341 0.614 0.478 0.196 1.000
30-day mortality Deep sternal wound infection Postoperative MI Postoperative stroke Reoperation for bleeding Renal impairment requiring dialysis BITA ¼ bilateral internal thoracic artery;
MI ¼ myocardial infarction;
number of arterial bypass grafts was 3.2 0.6 versus 2.9 0.7, respectively (p < 0.001). Our surgical strategy has been always to perform a complete myocardial revascularization, and considering that in both groups, 20% of the patients had 2-vessel disease, this strategy was achieved in the total group of patients.
Comment This single-center analysis directly compares a strategy of myocardial revascularization exclusively with BITAs as composite T-grafts (Tector’s technique) with a more “traditional” approach of SITA grafts supplemented by RA grafts or SVGs and confirms better long-term survival in patients receiving BITA grafts [6]. BITA grafting in CABG has been associated with improved long-term survival and freedom from late cardiac events. The advantages and disadvantages of the BITA technique have been rigorously studied through multiple cohort studies. A meta-analysis of 7 observational studies that included 15,962 patients (11,269 LITA
(8.9) (13) (27) (76) (73) (60) (25) (62) (1.6) (15) (21) (80.2) (0.4) (26) (16.5) (4.5) (2.7) (1.2) (4.7) (2.3) (3.5)
SD ¼ standard deviation;
(9.5) (14) (29) (76) (71) (57) (23) (62) (1.6) (15.7) (20.8) (81.4) (0.4) (25.8) (17.5) (2.3) (3.1) (1.0) (4.5) (3.1) (4.7)
p Value 0.948 0.926 0.475 0.940 0.566 0.473 0.601 0.895 10.000 0.859 0.937 0.625 10.000 0.942 0.799 0.051 0.701 0.762 0.879 0.427 0.382
SITA ¼ single internal thoracic artery.
and 4,693 BITA procedures) performed in 2001 by Taggart and coworkers [11] demonstrated significantly better survival in patients who underwent BITA grafting compared with patients who underwent a LITA procedure (HR, 0.81; 95% CI, 0.70–0.94). This sparked an even greater number of nonrandomized studies. In an updated meta-analysis, Weiss and colleagues [12] demonstrated an increase in long-term survival in patients receiving BITAs as a primary grafting strategy over those receiving LITAs. In our study, an off-pump technique was performed in more than 95% of the patients who underwent BITA grafting, with lower hospital mortality compared with patients who underwent SITA grafting. Adding off-pump techniques to myocardial revascularization exclusively with a BITA technique creates a more technically demanding procedure. However, we consider that this surgical approach is more patient friendly: 88% of patients were extubated in the operating room [13], and the use of a no-touch aortic technique in the off-pump group significantly decreased the incidence of
ADULT CARDIAC
Ann Thorac Surg 2016;101:1775–81
ADULT CARDIAC
1780
NAVIA ET AL MYOCARDIAL REVASCULARIZATION WITH BITA T-GRAFS
Table 3. Cox Proportional Hazard Regression Analysis for Long-Term (10 Years) All-Cause Mortality Covariate
HR
95% CI
p Value
Age (y) Diabetes mellitus Cerebrovascular disease Previous renal dysfunction Smoking habit Elective operation Left ventricular dysfunction (moderate/severe) BITA
1.07 1.69 2.16 2.12 1.47 0.78 2.47
1.06–1.08 1.39–2.06 1.49–3.11 1.58–2.85 1.21–1.78 0.64–0.94 1.92–3.19
0.000 0.000 0.000 0.000 0.000 0.009 0.000
0.71
0.58–0.87
0.000
BITA ¼ bilateral internal thoracic arteries; HR ¼ hazard ratio.
CI ¼ confidence interval;
postoperative stroke [14]. Finally, the possibility to perform BITA CABG with a single skin incision (no other conduit harvesting site) and with a short hospital stay (mean 4.8 days in elective cases) is better for the patient. Concerns about DSWI have discouraged the use of BITA grafting in the general population, particularly in patients with diabetes; however, in this study, there was no difference between the groups. It has been suggested that the method of ITA harvesting (skeletonized) and maintaining strict glycemic control during the entire procedure and hospitalization influences the incidence of postoperative DSWI [15]. There were no differences in the early hospital outcomes in the propensity-matched group. At 10 years, patients who underwent BITA grafting experienced greater unadjusted and adjusted survival than patients who underwent SITA
Fig 2. Ten-year survival among 485 propensity-score–matched patient pairs (BITA ¼ bilateral internal thoracic arteries; SITA ¼ single internal thoracic artery.)
Ann Thorac Surg 2016;101:1775–81
grafting. There have been other reports of excellent perioperative outcomes with good late survival using on-pump full myocardial revascularization purely with BITA composite grafts [16]. Long-term survival after CABG may be related to the late patency of conduits, and this may partially explain the improved survival seen in the patients who underwent BITA grafting. It is well known that the ITA is more resistant to atherosclerosis because of the endothelium’s capability of producing nitric oxide, which, together with its small diameter that creates better size matching with target vessels, may contribute to its durability [17, 18]. In a large multicenter study, Buxton and associates [19] suggested that a strategy of total arterial revascularization with multiple RA grafts as a second arterial conduit is associated with improved long-term survival compared with the use of SITA grafts and SVGs. The results of our previous study provide evidence for the superiority of the RITA graft compared with RA grafts as a second conduit in multiple arterial grafting [20]. However, in this study, patients in the SITA group with RA grafts plus SVGs had better long-term survival than patients with LITA-SVGs only, raising the speculation that RA grafts may indeed provide a prognostic benefit. Also, patients in the first SITA subgroup had long-term survival at 10 years almost identical to that of the BITA group; however, BITA grafting still shows greater beneficial effects on survival, as expressed in Cox proportional hazards regression analysis. The proposed superiority of arterial grafts may lie in their effect on the native circulation compared with vein grafts. The supraphysiological flow provided by vein grafts may create turbulence that promotes progression of atherosclerotic disease [21]. The use of off-pump ITA T-grafts with multiple sequential side-to-side ITA to coronary artery anastomoses introduces the concept of coronary arterial tree reconstruction by the most appropriate conduit; yet the procedure is more technically demanding [22]. Using this surgical approach, the total coronary bypass flow depends on the flow of the proximal LITA; this scenario has led to concern about whether coronary flow reserve in the LITA is sufficient to supply more than 1 coronary anastomosis. Several reports have already concluded that ITA T-grafts allow complete myocardial revascularization with good perioperative results and that the coronary flow reserve of the proximal LITA is adequate for multiple coronary anastomoses [23]. Regarding T-graft configuration and competitive flow, Glineur and coworkers [24, 25] stated that the resistance is similar in both branches of the graft, which excludes the possibility of steal from 1 branch to the other during periods of high myocardial blood flow demand. However, the longer the arterial assembly, the lower the pressure at the distal anastomosis, considering that this phenomenon is more frequent at the right coronary anastomosis of composite T-grafts because of the lower pressure in the distal portion of the T-branch. In our series of postoperative angiograms, the patency rate of the distal graft to the RCA/posterior descending artery was 91%, and only 8
1781
NAVIA ET AL MYOCARDIAL REVASCULARIZATION WITH BITA T-GRAFS
patients presented with competitive flow in the RITA to the right coronary artery or its branches. In these patients, the opacification of the native coronary arteries, visualized by retrograde flow, showed perfect patency and good diameter of the RITA graft. This scenario was more frequent in the case of a large dominant RCA. However, these angiographic abnormalities had no postoperative consequences in our group of patients [8]. There are several study limitations to mention. First, this is a retrospective study, and although the propensityscore–matching algorithm produced rather comparable groups, the study was not randomized, and we cannot rule out additional effects of missing covariates. Second, the study period was long, and most of the SITA procedures were done in the earlier part of the study because our experience in BITA grafting came later; this issue may have played a role in the final results. As was mentioned earlier, we started in 1996 using SITA plus SVGs on pump until 1998 when we introduced the use of RA grafts for the non–LAD artery system, and from 2003 until the present we have performed off-pump BITA grafting in a T-graft configuration. All the staff surgeons in the department have adopted these changes in the surgical technique over time. If we had included the variable “time,” we would have obtained a matched sample that would have included all the late SITA and early BITA procedures, so we would not have really obtained 2 comparable groups, which was necessary to answer the objective of the study. In conclusion, this single-center study suggests that myocardial revascularization exclusively with BITA in a T-graft configuration and an off-pump technique can be safely performed with low in-hospital mortality and may be associated with improved long-term survival compared with the use of SITA grafts plus other types of conduits.
8. Navia D, Vrancic M, Vaccarino G, et al. Total arterial offpump coronary revascularization using bilateral internal thoracic arteries in triple-vessel disease: surgical technique and clinical outcomes. Ann Thorac Surg 2008;86: 524–30. 9. Hansen BB, Bowers J. Covariate balance in simple, stratified and clustered comparative studies. Statistical Science 2008;23:219–36. 10. Iacus SM, King G, Porro G. CEM: Coarsened exact matching software. Journal of Statistical Software 2009;30:1–27. 11. Taggart DP, D’Amico R, Altman DG. Effect of arterial revascularization on survival: a systematic review of studies comparing bilateral and single internal mammary arteries. Lancet 2001;358:870–5. 12. Weiss A, Zhao S, Tian D, Taggart D, Yan T. A meta-analysis comparing bilateral internal mammary artery with left internal mammary artery for coronary artery bypass grafting. Ann Cardiothorac Surg 2013;2:390–400. 13. Dorsa A, Rossi A, Thierer J, et al. Extubation after off-pump coronary artery bypass graft surgery in 1,196 consecutive patients: feasibility, safety and predictors of when not to attempt it. J Cardiothorac Vasc Anesth 2011;25:431–6. 14. Moss E, Puskas J, Thourani V, et al. Avoiding aortic clamping during coronary artery bypass grafting reduces postoperative stroke. J Thorac Cardiovasc Surg 2015;149:175–80. 15. Barros de Oliveira Saa M, Ferraza P, Escobara R, et al. Skeletonized versus pedicled internal thoracic artery and risk of sternal wound infection after coronary bypass surgery: meta-analysis and meta-regression of 4817 patients. Interact Cardiovasc Thorac Surg 2013;16:849–57. 16. Paterson H, Naidoo R, Byth K, Chen C, Denniss R. Full myocardial revascularization with bilateral internal mammary artery Y grafts. Ann Cardiothorac Surg 2013;2:444–52. 17. Otsuka F, Yahagi K, Sakakura K, Virmani R. Why is the mammary artery so special and what protects it from atherosclerosis? Ann Cardiothorac Surg 2013;2:519–26. 18. Locker C, Schaff HV, Dearani JA, et al. Multiple arterial grafts improve late survival of patients undergoing coronary artery bypass graft surgery: analysis of 8622 patients with multivessel disease. Circulation 2012;126:1023–30. 19. Buxton BF, Shi WY, Tatoulis J, Fuller JA, Rosalion A, Hayward PA. Total arterial revascularization with internal thoracic and radial artery grafts in triple-vessel coronary artery disease is associated with improved survival. J Thorac Cardiovasc Surg 2014;148:1238–43. 20. Navia D, Vrancic M, Piccinini F, et al. Is the second internal thoracic artery better than the radial artery in total arterial off-pump coronary artery bypass grafting? A propensity score-matched follow-up study. J Thorac Cardiovasc Surg 2014;147:632–8. 21. Dimitrova KR, Hoffman DM, Geller CM, Dincheva G, Ko W, Tranbaugh RF. Arterial grafts protect the native coronary vessels from atherosclerotic disease progression. Ann Thorac Surg 2012;94:475–81. 22. Tarrío RF, Cuenca JJ, Gomes V, et al. Off-pump total arterial revascularization: our experience. J Card Surg 2004;19: 389–95. 23. Wendler O, Hennen B, Markwirth T, et al. T grafts with the right internal thoracic artery to the left internal thoracic artery versus the left internal thoracic and radial artery: flow dynamics in the internal thoracic artery main stem. J Thorac Cardiovasc Surg 1999;118:841–8. 24. Glineur D, Hanet C. Competitive flow in coronary bypass surgery: is it a problem? Curr Opin Cardiol 2012;27:620–8. 25. Glineur D, Hanet C, D’hoore W, et al. Causes of nonfunctioning right internal mammary used in a Y-graft configuration: insight from a 6-month systematic angiographic trial. Eur J Cardiothorac Surg 2009;36:129–35; discussion 135–6.
References 1. Loop FD, Lytle BW, Cosgrove DM, et al. Influence of the internal-mammary-artery graft on 10-year survival and other cardiac events. N Engl J Med 1986;314:1–6. 2. Barner HB. Double internal mammary-coronary artery bypass. Arch Surg 1974;109:627–30. 3. Fiore A, Naunheim K, Dean P, et al. Results of internal thoracic artery grafting over 15 years: single versus double grafts. Ann Thorac Surg 1990;49; 202–2. 4. Lytle BW, Blackstone EH, Loop FD, et al. Two internal thoracic arteries are better than one. J Thorac Cardiovasc Surg 1999;117:855–72. 5. Taggart DP, Altman DG, Gray AM, et al, ART Investigators. Randomized trial to compare bilateral versus single internal mammary coronary artery bypass grafting: 1-year results of the Arterial Revascularisation Trial (ART). Eur Heart J 2010;31:2470–81. 6. Tector AJ, Amundsen S, Schmahl TM, et al. Total revascularization with T grafts. Ann Thorac Surg 1994;57:33–8; discussion 39. 7. Tector AJ, McDonald ML, Kress DC, et al. Purely internal thoracic artery grafts: outcomes. Ann Thorac Surg 2001;72: 450–5.
ADULT CARDIAC
Ann Thorac Surg 2016;101:1775–81