If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
A meta-analysis of adjusted hazard ratios from 20 observational studies of bilateral versus single internal thoracic artery coronary artery bypass grafting
Address for reprints: Hisato Takagi, MD, PhD, Shizuoka Medical Center, Department of Cardiovascular Surgery, 762-1 Nagasawa, Shimizu-cho, Sunto-gun, Shizuoka 411-8611, Japan.
In 2001, a landmark meta-analysis of bilateral internal thoracic artery (BITA) versus single internal thoracic artery (SITA) coronary artery bypass grafting for long-term survival included 7 observational studies (only 3 of which reported adjusted hazard ratios [HRs]) enrolling approximately 16,000 patients. Updating the previous meta-analysis to determine whether BITA grafting reduces long-term mortality relative to SITA grafting, we exclusively abstracte, then combined in a meta-analysis, adjusted (not unadjusted) HRs from observational studies.
Methods
MEDLINE and EMBASE were searched until September 2013. Eligible studies were observational studies of BITA versus SITA grafting and reporting an adjusted HR for long-term (≥4 years) mortality as an outcome. Meta-regression analyses were performed to determine whether the effects of BITA grafting were modulated by the prespecified factors.
Results
Twenty observational studies enrolling 70,897 patients were identified and included. A pooled analysis suggested a significant reduction in long-term mortality with BITA relative to SITA grafting (HR, 0.80; 95% confidence interval, 0.77 to 0.84). When data from 6 pedicled and 6 skeletonized internal thoracic artery studies were separately pooled, BITA grafting was associated with a statistically significant 26% and 16% reduction in mortality relative to SITA grafting, respectively (P for subgroup differences = .04). A meta-regression coefficient was significantly negative for the proportion of men (−0.00960; −0.01806 to −0.00114).
Conclusions
Based on an updated meta-analysis of exclusive adjusted HRs from 20 observational studies enrolling more than 70,000 patients, BITA grafting seems to significantly reduce long-term mortality. As the proportion of men increases, BITA grafting is more beneficial in reducing mortality.
of bilateral internal thoracic artery (BITA) versus single internal thoracic artery (SITA) grafting in coronary artery bypass grafting (CABG) was published in the Lancet. The BITA group had significantly better long-term survival than the SITA group (hazard ratio [HR] for death, 0.81; 95% confidence interval [CI], 0.70 to 0.94). The meta-analysis,
that had enrolled 15,962 patients and were published in the 1990s. On the one hand, the Cox model was used in an attempt to adjust for the characteristics of the patients (eg, mean age, sex distribution, ventricular function, diabetic status, and so forth) that differed between the 2 groups in 3 reports
from the analysis substantively altered the overall result (statistically significant benefit in the BITA group than the SITA group for long-term mortality) to no statistically significant difference (HR, 0.85; 95% CI, 0.68 to 1.05; calculated by us) between the 2 groups. Since the previous meta-analysis
was performed, several studies on BITA versus SITA grafting in CABG have been published. In a recent large study published in 2013 by Parsa and colleagues,
), adjusted mortality was similar (HR, 0.95; 95% CI, 0.83 to 1.08) between the BITA and SITA groups. To the best of our knowledge, only 2 randomized controlled trials
ART Investigators Randomized trial to compare bilateral vs. single internal mammary coronary artery bypass grafting: 1-year results of the Arterial Revascularisation Trial (ART).
randomizing only 162 patients with a median 7.5-year follow-up, 5-year survival was not different (96.3% [95% CI, 92.0% to 98.7%] vs 93.8% [95% CI, 88.4% to 97.3%]; P = .39) between the BITA and SITA groups. Another relatively large trial randomizing 3102 patients, the Arterial Revascularization Trial,
ART Investigators Randomized trial to compare bilateral vs. single internal mammary coronary artery bypass grafting: 1-year results of the Arterial Revascularisation Trial (ART).
also demonstrated similar mortality at only 1 year (2.5% for BITA vs 2.3% for SITA; relative risk [RR], 1.06; 95% CI, 0.68 to 1.67). Accordingly, evidence for long-term survival in BITA versus SITA grafting in CABG from randomized controlled trials remains insufficient.
Updating the meta-analysis by Taggart and colleagues
to determine whether BITA grafting reduces long-term mortality relative to SITA grafting in CABG, we exclusively abstracted, then combined in a meta-analysis, adjusted (not unadjusted [crude]) risk estimates for long-term mortality from 20 observational studies (more than 70,000 patients were enrolled, which was more than quadruple the number of patients included in the previous meta-analysis
). Moreover, meta-regression analyses were performed to determine whether the effects of BITA grafting were modulated by prespecified factors.
Methods
Search Strategy and Study Selection
All observational studies of BITA versus SITA grafting that enrolled patients undergoing isolated CABG and reported adjusted risk estimates for long-term mortality were identified using the same search strategy as in the previous meta-analysis by Taggart and colleagues.
The MEDLINE and EMBASE databases were searched for publications containing the words “internal,” “mammary,” “thoracic,” “single,” “unilateral,” “bilateral,” “multiple,” “artery,” “arteries,” “singly” and in combination, between 1972 (MEDLINE) or 1980 (EMBASE) and September 2013. Two of us (H.T. and S.G.) independently inspected the electronic reports identified by the searches. We included published studies that had at least 100 patients in each group, and had a median (or mean) follow-up of at least 4 years. Only data from the last publication of centers that had produced sequential reports were included. We inspected the references of all studies to identify further studies.
Data Abstraction and Statistical Analysis
Data regarding detailed inclusion criteria, duration of follow-up, and all-cause long-term mortality (adjusted HRs for BITA vs SITA grafting and 95% CIs) were abstracted from each individual study. All study-specific estimates were combined using inverse variance-weighted averages of logarithmic HRs in both fixed- and random-effects models (primary meta-analysis). Between-study heterogeneity was analyzed by means of standard χ2 tests. Where nonsignificant statistical heterogeneity was identified, the fixed-effects estimate was used preferentially as the summary measure. Sensitivity analyses were performed to assess the contribution of each study to the pooled estimate by excluding individual studies 1 at a time and recalculating the pooled HR estimates for the remaining studies (1-study-removed meta-analysis). To assess the impact of differential internal thoracic artery (ITA) harvesting techniques among the studies on the pooled estimate, the effects of BITA grafting on long-term mortality were explored separately in studies using the pedicled and skeletonized ITA. Publication bias was assessed graphically using a funnel plot and mathematically using adjusted rank correlation and linear regression tests. Mixed-effects (unrestricted maximum likelihood) meta-regression analyses were performed to determine whether the effects of BITA grafting were modulated by prespecified factors: that is, the mean length of follow-up or age (years), and proportion of men or diabetes (%). Meta-regression graphs depict the effect of BITA grafting on the outcome (plotted as logHR on the y-axis) as a function of a given factor (plotted as a mean or proportion of that factor on the x-axis). Meta-regression coefficients (slopes of the meta-regression line) show the estimated increase in logHR per unit increase in the covariate. Because logHR >0 corresponds to HR >1 and log HR <0 corresponds to HR <1, a negative coefficient indicates that as a given factor increases, the HR decreases; that is, BITA grafting is more beneficial in reducing the outcome of interest. All analyses were conducted using Review Manager version 5.2 (Nordic Cochrane Centre, Copenhagen, Denmark) and Comprehensive Meta-Analysis version 2 (Biostat, Englewood, NJ).
Propensity matched analysis of bilateral internal mammary artery versus single left internal mammary artery grafting at 17-year follow-up: validation of a contemporary surgical experience.
Bilateral internal thoracic artery grafting is superior to other forms of multiple arterial grafting in providing survival benefit after coronary bypass surgery.
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.
Single versus bilateral internal thoracic artery grafts with concomitant saphenous vein grafts for multivessel coronary artery bypass grafting: effects on mortality and event-free survival.
of BITA versus SITA grafting that enrolled patients undergoing CABG and reported adjusted risk estimates for long-term mortality. In total, our meta-analysis included data on 70,897 patients undergoing CABG assigned to BITA or SITA grafting. The baseline characteristics for the patients enrolled in each study are summarized in Table 1. The most notable difference in some studies
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.
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.
respectively. Despite the noted heterogeneity in design among studies, there was sufficient similarity between the populations and the hypotheses to merit inclusion of all 20 studies in the quantitative meta-analysis. The most dissimilar studies (Bonacchi and colleagues, 2006,
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.
Propensity matched analysis of bilateral internal mammary artery versus single left internal mammary artery grafting at 17-year follow-up: validation of a contemporary surgical experience.
Bilateral internal thoracic artery grafting is superior to other forms of multiple arterial grafting in providing survival benefit after coronary bypass surgery.
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.
Single versus bilateral internal thoracic artery grafts with concomitant saphenous vein grafts for multivessel coronary artery bypass grafting: effects on mortality and event-free survival.
A pooled analysis of all 20 studies demonstrated a statistically significant 20% reduction in long-term mortality with BITA relative to SITA grafting in the fixed-effects model (HR, 0.80; 95% CI, 0.77 to 0.84; P < .00001; Figure 1). There was minimal study heterogeneity of the results (P = .30) and, accordingly, little difference in the pooled result from random-effects modeling (HR, 0.80; 95% CI, 0.76 to 0.84; P < .00001).
Figure 1Forest plot of adjusted hazard ratios for long-term mortality among patients assigned to bilateral versus single internal thoracic artery grafting. BITA, Bilateral internal thoracic artery; SITA, single internal thoracic artery grafting; SE, standard error; CI, confidence interval; IV, inverse variance; RA, radial artery; SV, saphenous vein.
To assess the impact of qualitative heterogeneity in the study design and patient selection on the pooled effect estimate, we performed several sensitivity analyses. First, we sequentially excluded Bonacchi and colleagues, 2006,
respectively; combining the remaining 19 studies still generated a statistically significant result favoring BITA grafting (fixed-effects HR for exclusion of Bonacchi and colleagues, 2006,
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.
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.
Without them, there was still a statistically significant benefit for BITA grafting in pooled analysis of the remaining 18 studies (fixed-effects HR, 0.80; 95% CI, 0.77 to 0.84; P < .00001). In general, exclusion of any single study from the analysis did not substantively alter the overall result of our analysis (1-study-removed meta-analysis; Figure 2). When data from 6 pedicled
Propensity matched analysis of bilateral internal mammary artery versus single left internal mammary artery grafting at 17-year follow-up: validation of a contemporary surgical experience.
Bilateral internal thoracic artery grafting is superior to other forms of multiple arterial grafting in providing survival benefit after coronary bypass surgery.
Single versus bilateral internal thoracic artery grafts with concomitant saphenous vein grafts for multivessel coronary artery bypass grafting: effects on mortality and event-free survival.
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.
(enrolling 7,831 patients) were separately pooled using the fixed-effects model, BITA grafting was associated with a statistically significant 26% (HR, 0.74; 95% CI, 0.68 to 0.81; P for effect < .00001; P for heterogeneity = .51) and 16% (HR, 0.84; 95% CI, 0.78 to 0.91; P for effect < .0001; P for heterogeneity = .65) reduction in long-term mortality relative to SITA grafting (P for subgroup differences = .04; Figure 3).
Figure 2One-study-removed meta-analysis of adjusted hazard ratios for long-term mortality among patients assigned to bilateral versus single internal thoracic artery grafting. CI, Confidence interval; BITA, bilateral internal thoracic artery; SITA, single internal thoracic artery grafting.
Figure 3Forest plot of adjusted hazard ratios for long-term mortality among patients assigned to bilateral versus single internal thoracic artery grafting pooled separately in studies using the pedicled and skeletonized internal thoracic artery. BITA, Bilateral internal thoracic artery; SITA, single internal thoracic artery grafting; SE, standard error; IV, inverse variance; CI, confidence interval; ITA, internal thoracic artery.
To assess publication bias we generated a funnel plot of the logarithm of effect size versus the precision (reciprocal of the standard error) for each study (data not shown). There was no evidence of significant publication bias (2-tailed P with continuity correction = .95787 by the adjusted rank correlation test and 2-tailed P = .21818 by the linear regression test).
Meta-Regression Analyses
Meta-regression coefficients were not statistically significant for the mean length of follow-up (coefficient, −0.00194; 95% CI, −0.01526 to 0.01137; P = .77472), mean age (coefficient, 0.01087; 95% CI, −0.00448 to 0.02622; P = .16528), and proportion of diabetes (coefficient, −0.00167; 95% CI, −0.00137 to 0.00471; P = .28178). However, a meta-regression coefficient was significantly negative for the proportion of men (coefficient, −0.00960; 95% CI, −0.01806 to −0.00114; P = .02621; Figure 4), which would indicate that as the proportion of men increases, the HR decreases; that is, BITA grafting is more beneficial in reducing long-term mortality.
Figure 4Meta-regression of proportion of men (%) on logarithmic hazard ratio for long-term mortality among patients assigned to bilateral versus single internal thoracic artery grafting.
The results of our analysis suggest that BITA grafting may reduce long-term mortality by 20% relative to SITA grafting in patients undergoing CABG, which is most compelling with adjusted data from 70,897 patients assigned in 20 different observational studies and robust in sensitivity analyses. Because the present meta-analysis overcomes the limitations of the analysis Taggart and colleagues
mentioned earlier, the results could provide better (not the best) evidence until a large randomized controlled trial reports long-term outcomes in the future.
In 1 of our sensitivity analyses, BITA grafting was associated with a 26% reduction in long-term mortality relative to SITA grafting in pedicled ITA studies, which was better (P for subgroup differences = .04) than a 16% reduction in skeletonized ITA studies. The 6 skeletonized ITA studies
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.
Propensity matched analysis of bilateral internal mammary artery versus single left internal mammary artery grafting at 17-year follow-up: validation of a contemporary surgical experience.
Bilateral internal thoracic artery grafting is superior to other forms of multiple arterial grafting in providing survival benefit after coronary bypass surgery.
Single versus bilateral internal thoracic artery grafts with concomitant saphenous vein grafts for multivessel coronary artery bypass grafting: effects on mortality and event-free survival.
Although theoretically skeletonization of the ITA might adversely affect its long-term resistance to atherosclerosis, the skeletonized ITA has not been used long enough to establish whether a decline in patency will occur after several years.
Also, only 4 were comparative studies of skeletonized versus pedicled conduits, in which skeletonization patency was at least comparable with pedicled conduits and in 2 studies even higher.
Sternal wound infection (SWI), the most serious manifestation of which is mediastinitis, is recognized as an important complication of CABG, and mediastinitis can be severe and potentially lethal. Skeletonized ITA grafting seems to reduce the incidence of postoperative SWI in comparison with pedicled ITA grafting after CABG. A recent meta-analysis
Skeletonized versus pedicled internal thoracic artery and risk of sternal wound infection after coronary bypass surgery: meta-analysis and meta-regression of 4817 patients.
of 22 studies (involving 4817 patients) showed a statistically significant difference in favor of skeletonized ITA grafting (odds ratio [OR], 0.443; 95% CI, 0.323 to 0.608, P < .001), which was also observed in subgroups of BITA grafting (OR, 0.381; 95% CI, 0.257 to 0.565; P < .001) and diabetic patients with BITA grafting (OR, 0.188; 95% CI, 0.098 to 0.360; P < .001). To confirm the long-term survival benefit of BITA grafting in pedicled versus skeletonized ITA studies suggested in our sensitivity analysis, further investigations are required.
One of our meta-regression analyses indicated that as the proportion of men increases, BITA grafting is more beneficial in reducing long-term mortality. A post hoc analysis
(included in the present meta-analysis) may support our result. In multivariable Cox proportional hazard regression analyses for unmatched patients, the choice of the BITA conduit was identified as a predictor of late survival among men (HR, 0.8; 95% CI, 0.7 to 0.9; P < .001) but not associated with late survival among women.
In the propensity score matched male groups, the median survival for SITA patients was 14.2 years (95% CI, 13.2 to 15.2) compared with 15.8 years (95% CI, 15.0 to 16.8) for BITA patients. The equality of the survival distribution for these 2 groups of male patients demonstrated a significant difference (P < .001).
These results provide further evidence of the survival benefits achieved in male patients with BITA grafting. In the propensity score matched female groups, however, the median survival for SITA patients was 14.0 years (95% CI, 12.8 to 15.4) and that for BITA patients was 13.7 years (95% CI, 12.8 to 14.7), with no significant difference in survival distribution (P = .571).
Therefore, BITA grafting seems to confer an incremental survival benefit relative to SITA grafting in men but not in women. Although these results represent the largest experience in the current literature that specifically addresses long-term survival of BITA grafting in women, it may be that 329 patients, compared with the male cohorts of more than 1000 patients, is just not sufficient to demonstrate a survival advantage (type II error).
of 32 studies (consisting of 172,880 patients) showed that BITA grafting increases the risk of SWI compared with SITA grafting. The risk of SWI in the BITA group was higher (RR for SITA vs BITA, 0.62; 95% CI, 0.55 to 0.71) than that in the SITA group. This adverse effect further extends to diabetic and elderly patients
; BITA grafting was also associated with a higher risk of SWI in diabetic patients (RR for SITA vs BITA, 0.65; 95% CI, 0.52 to 0.81; 12 studies consisting of 128,109 patients) as well as elderly patients (RR for SITA vs BITA, 0.45; 95% CI, 0.33 to 0.62; 5 studies consisting of 8206 patients). Also in another recent meta-analysis
of 1 randomized controlled trial and 10 observational studies (enrolling 126,235 diabetic patients), deep SWI occurred in 3.1% and 1.6% for the BITA and LITA cohorts, respectively (RR, 1.71; 95% CI, 1.37 to 2.14). Despite the higher risk of SWI in BITA grafting, the present meta-analysis demonstrated the obvious benefit for BITA grafting for long-term survival.
Our analysis must be viewed in the context of its limitations. We used only data from observational studies, not randomized controlled trials. Although patients enrolled in randomized trials may not be representative of patients typically seen in clinical practice, this is the study design least vulnerable to bias because randomized trials balance both known and unknown confounders across treatment groups. Potential biases are likely to be greater for observational studies compared with randomized trials, however, so results should always be interpreted with caution when they are included in reviews and meta-analyses.
Reeves BC, Deeks JJ, Higgins JP, Wells GA. Including non-randomized studies. In: Higgins JP, Green S, eds. Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 (updated March 2011). The Cochrane Collaboration; 2011, Chapter 13. Available at: www.cochrane-handbook.org.
Particular concerns arise with respect to differences between patients in different intervention groups (selection bias). Unlike for randomized trials, it would usually be appropriate to analyze adjusted (rather than unadjusted) effect estimates, that is, analyses that attempt to control for confounding.
Reeves BC, Deeks JJ, Higgins JP, Wells GA. Including non-randomized studies. In: Higgins JP, Green S, eds. Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 (updated March 2011). The Cochrane Collaboration; 2011, Chapter 13. Available at: www.cochrane-handbook.org.
To reduce the effect of treatment selection bias and potential confounding in observational studies, rigorous adjustment for significant differences in the baseline characteristics of patients should be conducted. Furthermore, adjusted (not unadjusted) estimates ought to be pooled in a meta-analysis that includes observational studies. In the present meta-analysis, we strictly abstracted, then combined in a meta-analysis, exclusive adjusted (not unadjusted) risk estimates from observational studies. Meanwhile, our results may be influenced by a publication bias favoring BITA grafting. This risk was minimized through an exhaustive search of the available literature. Although the statistical tests did not indicate publication bias, there is clearly limited power to detect such bias, given the small number of studies examined.
Despite these acknowledged limitations, we found that, based on an updated meta-analysis of exclusive adjusted risk estimates from 20 observational studies enrolling more than 70,000 patients, BITA grafting is likely effective in the prevention of long-term mortality in patients undergoing CABG. Because mortality reduction must imply the greatest clinical benefit among patients undergoing CABG, BITA grafting should be considered for patients who meet the criteria for enrollment in the observational studies discussed earlier.
References
Taggart D.P.
D'Amico R.
Altman D.G.
Effect of arterial revascularisation on survival: a systematic review of studies comparing bilateral and single internal mammary arteries.
Randomized trial to compare bilateral vs. single internal mammary coronary artery bypass grafting: 1-year results of the Arterial Revascularisation Trial (ART).
Propensity matched analysis of bilateral internal mammary artery versus single left internal mammary artery grafting at 17-year follow-up: validation of a contemporary surgical experience.
Bilateral internal thoracic artery grafting is superior to other forms of multiple arterial grafting in providing survival benefit after coronary bypass surgery.
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.
Single versus bilateral internal thoracic artery grafts with concomitant saphenous vein grafts for multivessel coronary artery bypass grafting: effects on mortality and event-free survival.
Skeletonized versus pedicled internal thoracic artery and risk of sternal wound infection after coronary bypass surgery: meta-analysis and meta-regression of 4817 patients.
Reeves BC, Deeks JJ, Higgins JP, Wells GA. Including non-randomized studies. In: Higgins JP, Green S, eds. Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 (updated March 2011). The Cochrane Collaboration; 2011, Chapter 13. Available at: www.cochrane-handbook.org.
00052-X&id=gr1.jpg){kind=link}
00052-X&id=gr2.jpg){kind=link}
00052-X&id=gr3.jpg){kind=link}
00052-X&id=gr4.jpg){kind=link}