- Email: [email protected]
Comparison of the Safety of Right Internal Mammary Artery Versus Radial Artery for the Second Arterial Conduit During Coronary Artery Bypass Grafting
ORIGINAL ARTICLE
Heart, Lung and Circulation (2014) 23, 1179–1186 1443-9506/04/$36.00 http://dx.doi.org/10.1016/j.hlc.2014.05.017
Comparison of the Safety of Right Internal Mammary Artery Versus Radial Artery for the Second Arterial Conduit During Coronary Artery Bypass Grafting Chenyang Dai, MD, Wentian Zhang, MD, Zhexin Lu, MD, Hongsheng Zhu, MD, PhD, Song Xue, MD, PhD, Feng Lian, MD, PhD * Department of Cardiovascular Surgery, Ren Ji Hospital, School of Medicine, Shanghai Jiao Tong University, Shanghai, People’s Republic of China Received 20 February 2014; received in revised form 12 May 2014; accepted 20 May 2014; online published-ahead-of-print 23 June 2014
Objective
To compare the safety of procedure in patients receiving right internal mammary artery (RIMA) versus radial artery (RA) as the second arterial graft during left internal mammary artery bypass surgery.
Methods
A literature search was performed, and observational studies comparing RA and RIMA as a second arterial conduit were included. The end points included in-hospital mortality, sternal wound infection (SWI), myocardial infarction (MI), reoperation for bleeding, stroke and low cardiac output syndrome. Sensitivity analysis was performed, and publication bias was assessed.
Results
Analysis of nine studies comprising 1,738 RIMA patients and 1,906 RA patients showed a similar risk of inhospital mortality (relative risk [RR] 1.13, 95% confidence interval [CI] 0.64 to 1.98) between two groups. The risk of SWI in RIMA patients was significantly higher than that in RA patients (RR 3.19, 95% CI 1.64 to 6.20). The risk of the rest end points did not differ.
Conclusions
The postoperative risk seems to be similar between RIMA group and RA group, except the higher risk of SWI in RIMA patients. The surgeon relies on his own experience with these two arterial grafts and assessment of the patient profile to decide the choice of the second arterial conduit.
Keywords
Coronary artery bypass Meta-analysis Radial artery Mammary arteries Postoperative complications Surgical wound infection
Introduction The benefits of the use of left internal mammary artery (LIMA) are well established when compared with saphenous vein grafts. Therefore, logic seems to dictate that adding a second arterial graft, which might further improve the longterm effect is an issue of increased interest. RIMA and RA are most commonly used as a second graft of choice. In fact, several studies have demonstrated that adding either RA or RIMA is associated with higher patency rate and better
long-term survival than vein grafts during LIMA bypass surgery [1–4]. Nevertheless, the postoperative high risk restricts the routine adoption of RA or RIMA. Many surgeons are reluctant to adopt bilateral internal mammary artery grafting (BIMA) due to the increased risk of sternal wound infection, including life threatening mediastinitis, which not only increases the length and the expense of hospital stay but also the in-hospital mortality significantly [5]. There is a high propensity of vasospasm at the early stage after radial artery procedure. Once spasm
*Corresponding author at: Department of Cardiovascular Surgery, Ren Ji Hospital, School of Medicine, Shanghai Jiao Tong University, No. 160 Pujian Road, Shanghai, 200127, P.R. China; Fax: +86-21-68383764., Email: [email protected] © 2014 Australian and New Zealand Society of Cardiac and Thoracic Surgeons (ANZSCTS) and the Cardiac Society of Australia and New Zealand (CSANZ). Published by Elsevier Inc. All rights reserved.
1180
occurs, the decreased blood supply might lead to a higher postoperative risk since RA graft is often used as T connection, providing blood supply to more than one target vessel in sequential fashion. Hence, the safety of procedure is often an important issue for surgeons. Yet, it is still unclear which arterial graft is accompanied by lower postoperative risk. The aim of this meta-analysis is to assess the postoperative safety by comparing early outcomes in patients receiving RIMA versus RA as the second arterial graft during LIMA bypass surgery.
Methods Literature Search A literature search was performed in PubMed, Embase, and CENTRAL with ‘‘Coronary artery bypass’’ as the MESH search heading. Headings ‘‘right internal mammary artery’’, ‘‘right internal thoracic artery’’, ‘‘bilateral internal mammary artery’’, ‘‘bilateral internal thoracic artery’’ and ‘‘radial artery’’ were also applied. The articles were restricted to English and published until the end of January 10, 2014. To broaden the search, we also screened the references of articles and made use of the ‘‘Related articles’’ function in PubMed.
Data Extraction and Quality Assessment Two reviewers (C.D. and W.Z) extracted these essential data from each study including publication year, first author, population characteristics, sample size, and the mortality and morbidity after operation (Tables 2 and 3). The quality of each study was assessed by a modified Newcastle-Ottawa scale (Table 1) [6]. Total score was nine stars, and a study with
C. Dai et al.
more than four stars was graded as high level of quality. Data extraction and quality assessment were carried out independently by two reviewers.
Inclusion and Exclusion Criteria The included studies should satisfy the following criteria: (1) observational studies comparing LIMA+RIMA versus LIMA+RA with or without concomitant saphenous vein grafts; (2) when several studies were reported from the same institution, the most recent study or study with most information would be included. If there was a potential possibility that a third arterial conduit was used, the study would be excluded if a confirmation email was received from the author or correspondence with the author failed.
Outcome of Interest The primary end point was in-hospital mortality defined as a death occurring within 30 days after operation. Secondary end points included SWI and MI. SWI includes superficial SWI and deep SWI. The former was defined as infection and was limited in subcutaneous tissue, and the latter was defined as infection involving sternal osteomyelitis with or without mediastinitis. MI was defined as the appearance of new Q wave in at least two leads or left bundle branch block on the electrocardiogram or the creatine kinase myocardial band (CKMB) isoenzyme fraction increased more than 30 IU/L or more than 7% of the total creatine kinase level. Additional supplementary end points included reoperation for bleeding, stroke and low cardiac output syndrome.
Statistical Analysis Statistical analysis was performed by taking relative risk (RR) as the summary statistic. The RR represents the risk of
Table 1 Check List for Quality Assessment. Selection 1. Were any criteria for assignment of patients to treatment reported? (If yes, 1 star). 2. How representative was right internal mammary artery group? (1 star if basically representative of general population for surgery, no star if the patients were selected or selection of group was not reported). 3. How was radial artery group selected? (1 star if selected from the same population as exposed group, no star if chosen from a different population or selection of group was not represented). Comparability 1. Group comparable for age, sex and diabetes. (2 stars if no differences of the above characteristics between 2 groups; 1 star if one of the 3 characteristics was not represented or one characteristic was different between 2 groups; no star if 2 groups differed) 2. Group comparable for chronic obstructive pulmonary disease, body mass index, and ventricular function. (2 stars if no differences of the above characteristics between 2 groups; 1 star if one of the 3 characteristics was not represented or one characteristic was different between 2 groups; no star if 2 groups differed) Outcomes 1. Assessment of outcome (1 star if outcome was ascertained by medical record or interview, no star if not reported). 2. Adequacy of cohort follow-up (1 star if no patient or fewer than 10% of patients were lost to follow-up; no star if more than 10% of patients were lost).
Study
Year
Patients, n RIMA/RA
Age, mean years RIMA/RA
Sex (% men) RIMA/RA
Diabetes (%) RIMA/RA
COPD(%) RIMA/RA
BMI(mean) RIMA/RA
Ventricular function (mean or %) Classification
RIMA/RA
Score
Borger et al. [8]
1998
132/171
NS
92/88
11/27
NS
NS
Ejection fraction < 40%
14/18
Wendler et al. [9]
1999
73/178
61.4/60.5
89/78.7
NS
NS
NS
Ejection fraction
58.1/55.3
3
Lemma et al. [10]
2001
94/156
57/60
91.5/87.8
7.4/19.2
3.2/15.3
NS
Ejection fraction
54.2/56.5
4
Caputo et al. [11]
2003
336/325
55.5/57.7
91.4/84.3
6.3/18.2
NS
27.5/28.2
Poor ejection fraction
13.8/26.2
5
Vicol et al. [12]
2003
129/84
59.2/61.7
89.2/76.2
19.4/27.4
NS
NS
Ejection fraction
58.1/62
4
Miana et al. [13]
2007
20/38
57/56
80/76.3
15/21
NS
NS
Left ventricular dysfunction
15/18.4
7
Ruttmann et al. [14]
2011
277/277
56.6/57.8
89.9/89.9
21.3/22.4
33.2/33.2
27.4/27.7
Ejection fraction
54.9/52.9
8
Navia et al. [15] Tranbaugh et al. [16]
2014 2014
149/149 528/528
68/67 61/60
82.5/81.2 76.9/78.4
30.8/32.2 35.6/36.6
NS 10/10.4
NS 28.7/28.8
Severe ventricular function Ejection fraction
30.2/30.2 47.5/47.3
6 8
RIMA versus RA
Table 2 Characteristics of studies used in the meta-analysis.
5
RIMA, right internal mammary artery; RA, radial artery; NS, not stated; COPD, chronic obstructive pulmonary disease; BMI, body mass index.
Table 3
Incidence of end points: RIMA versus RA as the second arterial conduit.
Study
Patients, n
In-hospital
SWI, n(%)
MI, n(%)
Reoperation for
Stroke, n(%)
Low cardiac output
RIMA/RA
mortality, n(%)
RIMA/RA
RIMA/RA
bleeding, n(%)
RIMA/RA
syndrome, n(%)
RIMA/RA
RIMA/RA
RIMA/RA
Borger et al. [8]
132/171
0(0)/1(0.58)
7(5.30)/1(0.58)
2(1.52)/2(1.17)
1(0.76)/1(0.58)
1(0.76)/0(0)
Wendler et al. [9]
73/178
2(2.74)/4(2.25)
0(0)/1(0.56)
2(2.74)/4(2.25)
0(0)/1(0.56)
0(0)/4(2.25)
8(6.06)/10(5.85) 1(1.37)/3(1.69)
Lemma et al. [10]
94/156
0(0)/3(1.92)
8(8.51)/2(1.28)
2(2.13)/6(3.85)
1(1.06)/1(0.64)
NS
2(2.13)/3(1.92)
Caputo et al. [11]
336/325
3(0.89)/1(0.31)
NS
7(2.08)/1(0.31)
NS
2(0.6)/2(0.62)
NS
Vicol et al. [12]
129/84
1(0.78)/1(1.19)
1(0.78)/1(1.19)
4(3.1)/0(0)
4(3.1)/1(1.19)
NS
3(2.33)/1(1.19)
Miana et al. [13]
20/38
0(0)/1(2.63)
0(0)/0(0)
0(0)/1(2.63)
0(0)/1(2.63)
0(0)/1(2.63)
0(0)/1(2.63)
Ruttmann et al. [14] Navia et al. [15]
277/277 149/149
2(0.72)/5(1.81)a 6(4.03)/5(3.36)
NS NS
1(0.36)/10(3.61) NS
11(3.97)/18(6.5) NS
1(0.36)/10(3.61) NS
NS NS
Tranbaugh et al. [16]
528/528
9(1.7)/3(0.6)
14(2.7)/6(1.1)
3(0.6)/6(1.1)
10(1.9)/8(1.5)
11(2.1)/4(0.8)
NS
RIMA, right internal mammary artery; RA, radial artery; SWI, sternal wound infection; MI, myocardial infarction; NS, not stated; a cardiac related death.
1181
1182
postoperative event in RIMA group divided by that in RA group. RR less than 1 favours RIMA, whereas more than 1 favours RA. The point estimate of RR is considered statistically significant at the level of p value less than 0.05 if the 95% confidence interval (CI) does not include the value 1. We chose I2 statistic to measure the heterogeneity between studies. According to the results of I2 statistic, the level of heterogeneity was classified as low (less than 30%), moderate (31% to 50%) and high (more than 50%) [7]. When the heterogeneity between studies was low or moderate, a fixedeffect model would be chosen to combine the results; if not, a random-effect model would be adopted. Sensitivity analysis was undertaken to examine the stability of the results. Firstly, the results were recalculated by using another effect model. Secondly, two subgroup analyses were assessed: (1) excluded the study with small series (less than 100 patients in either group); (2) included the studies of high quality. Also, publication bias was evaluated by drawing funnel plot and performing Egger’s test and Begg’s test. All these statistical analyses were conducted by using Review Manager 5.1 (The Cochrane Collaboration, Oxford, UK).
C. Dai et al.
Primary End Point: Mortality All of nine included studies reported in-hospital mortality. The incidence of mortality was 1.32% (23 of 1738 patients) in RIMA group versus 1.26% (24 of 1906 patients) in RA group, and the RR was calculated as 1.13 (95% CI 0.64 to 1.98) (Fig. 1) and there was no significant heterogeneity among studies (I2 = 0%).
Second End Point: SWI and MI
Results
Incidence of SWI was available from six of nine included studies. The incidence of SWI was 3.07% in RIMA group and 0.95% in RA group. The RR showed that RIMA group was associated with a higher risk of SWI versus RA group (RR 3.19, 95% CI 1.64 to 6.20) (Fig. 2A) with low heterogeneity (I2 = 6%). A total of eight studies reported the data on MI, and the incidence in RIMA group was 1.32% compared with 1.71% in RA group. As shown in Figure 2B, the RR was calculated as 0.83 (95% CI 0.48 to 1.46) with moderate heterogeneity (I2 = 36%). It should be pointed out that the definition of MI from seven of the included studies was the same or similar to ours, and the remaining one from Ruttmann was defined as the appearance of new Q wave or the elevation of troponin which was more precise than CKMB. Even after the exclusion of this study, there was still no significant difference in the risk of MI among studies (p = 0.48, RR 1.26, 95% CI 0.66 to 2.37).
Selected Studies
Other End Points
A total of 3427 articles were searched from three electronic databases, and 11 studies met the inclusion criteria after review. Two studies with the potential possibility of using a third arterial conduit were excluded, because the authors failed to answer our query letter by email [17,18]. In all, nine observational studies with 3,644 patients were included in this meta-analysis, of whom 1,738 received RIMA graft and 1,906 received RA graft (Tables 2 and 3).
Seven studies demonstrated the similar risk of reoperation for bleeding between RIMA group and RA group (2.15% versus 2.16%, RR 0.91, 95% CI 0.55 to 1.50) with no heterogeneity (I2 = 0%) (Fig. 3A). Six studies reported the incidence of stroke between RIMA group and RA group (1.1% versus 1.39%), and the RR calculated from these studies was 0.84 (95% CI 0.44 to 1.61), also there was no high heterogeneity (I2 = 49%) (Fig. 3B). In the five studies, the risk of low cardiac
Figure 1 Meta-analysis of in-hospital mortality comparing right internal mammary artery versus radial artery as the second arterial conduit.
1183
RIMA versus RA
Figure 2 Meta-analysis of sternal wound infection (A) and myocardial infarction (B) comparing right internal mammary artery versus radial artery as the second arterial conduit.
output syndrome in RIMA group was comparable with that in RA group (3.13% versus 2.87%, RR 1.07, 95% CI 0.53 to 2.13), and no heterogeneity was found (I2 = 0%) (Fig. 3C).
Table 4 revealed that the outcomes of all end points were hardly changed and quite stable.
Sensitivity Analysis
Funnel plot was drawn for in-hospital mortality reported by all nine studies showing basic symmetry (Fig. 4). In addition, the results of both Egger’s test (p = 0.182) and Begg’s test
Different strategies were adopted to perform the sensitivity analysis for these six end points. Those results shown in
Assessment on Publication Bias
Table 4 The results of sensitivity analysis. Changing effect model
Excluding the small studies
High-quality studies
No. studies
RR(95% CI)
P value
No. studies
RR(95% CI)
P value
No. studies
RR(95% CI)
P value
In-hospital mortality
9
1.15(0.62-2.10)
0.66
5
1.34(0.69-2.58)
0.38
6
1.29(0.68-2.45)
0.43
SWI
6
3.01(1.41-6.42)
0.004
2
3.19(1.38-7.36)
0.007
3
3.19(1.38-7.36)
0.007
MI
8
0.89(0.37-2.10)
0.78
4
0.79(0.17-3.73)
0.76
5
0.76(0.20-2.85)
0.68
Reoperation
7
0.89(0.53-1.50)
0.67
3
0.82(0.47-1.43)
0.49
4
0.82(0.47-1.41)
0.46
for bleeding Stroke
6
0.81(0.23-2.83)
0.74
4
0.98(0.19-5.11)
0.98
5
0.93(0.23-3.77)
0.92
Low cardiac
5
1.06(0.53-2.13)
0.87
1
1.04(0.42-2.55)
0.94
2
0.99(0.42-2.36)
0.98
output syndrome SWI, sternal wound infection; MI, myocardial infarction; RR, relative risk; CI, confidence interval
1184
C. Dai et al.
Figure 3 Meta-analysis of reoperation for bleeding (A), stroke (B) and low cardiac output syndrome (C) comparing right internal mammary artery versus radial artery as the second arterial conduit.
(p = 0.251) demonstrated that no notable publication bias was identified.
Discussion
Figure 4 Funnel plot for in-hospital mortality showing a symmetrical plot. (SE = standard error).
Since the benefits of less trauma in nature and improving therapeutic outcome achieved by the interventional cardiologist, surgeons have been aiming to improve the long-term outcomes of bypass surgery. Adding a RIMA or RA as a second arterial conduit is believed an effective strategy to optimise the prognosis of bypass surgery. In spite of the encouraging report early this year by Tranbaugh showing the five, 10, 15-year survival rate reached 92%, 80% and 71% for RIMA patients and 95%, 85% and 76% for RA patients respectively [16], only 4% of bypass operations adopted BIMA grafting, and the use of RA graft was 9% in clinical
1185
RIMA versus RA
practice [19]. The concern about the perioperative safety of procedure is one of the reasons for this present clinical status. In fact our study shows that the in-hospital mortality, as a primary indicator of safety is acceptable in both two groups (1.32% in RIMA and 1.26% in RA). As to the protection for surgical safety, both arterial grafts showed a similar risk of postoperative adverse events, except the higher risk of SWI in RIMA group. A recent survey reported that SWI was the common concern regarding the extensive use of RIMA even among experienced cardiac surgeons [20]. Our meta-analysis shows that their concerns are valid and well founded. The blood deprivation of hemi-sternum reached 90% after the harvest of one internal mammary artery [21], so the chance of SWI increased significantly when both LIMA and RIMA were utilised simultaneously. Several studies also demonstrated that patients with advanced age, diabetes mellitus or chronic obstructive pulmonary disease endured much higher risk of SWI [22,23]. Hence, SWI seems to be a negative impact on BIMA grafting. Even so, it does not mean that BIMA should be fully abandoned in clinical practice. The increased risk of SWI had been observed after BIMA surgery, while morbidity was relatively low especially deep SWI (ranging from 1.2% to 2.5%) [24]. In fact, BIMA was just one of the initial factors for postoperative SWI. The risk of infection can be reduced to a minimum if effort is made to prevent other risk factors, such as the overall discovery of latent inflammatory foci and effective control of metabolic disturbance preoperatively, combined with the meticulous manipulation and thorough haemostasis intraoperatively, followed by the strict implementation of standard wound care postoperatively, and the appropriate use of antibiotic during the entire perioperative period [24,25]. More than anything else, a recent metaanalysis showed that the utilisation of skeletonisation technique preserved more sternal perfusion after bypass surgery [26]. Saso and colleagues found that skeletonised BIMA was associated with a notably lower risk of SWI than pedicled BIMA [27]. Moreover, the lower risk was also comparable with that of single internal mammary artery grafting [23], and this favourable effect even extended to diabetic patients [28]. As to the safety of RA graft, surgeons often worry about its susceptibility to vasospasm during the early period after operation. Our result shows that the incidence of MI is not significantly higher in RA group, which is attributed to the intraluminal irrigation of vasodilator drugs, for example, papavarine, verapamil and diltiazem, either singly or in combination [8–10,12–16]. Gentle harvesting of RA such as light tractions, ultrasonicscalpel dissection or endoscopic technique also contributes to prevent vasospasm [29]. In addition, some other potential complications for RA like superficial radial nerve injury even reported as less than 10% can be prevented by avoiding excessive lateral traction of brachioradialis muscle [30], and haematoma formation can also be prevented by careful and secure ligation of the branches and the stumps. Even the rare but devastating compartment syndrome can also be avoided by close observation and regular check of the hand
after radial artery harvesting for unrestricted movement and intact sensation at routine intervals. Since the negative aspect of RA graft can be managed, it should be considered as a reasonable and safe option. Furthermore if a patient is complicated with multiple risk factors of SWI, the risk of SWI increases exponentially when BIMA grafting is adopted [31]. However, SWI is not a major concern for RA graft, and its contraindications including carpal tunnel syndrome, Dupuytren disease, etc. are quite rare. Hence, RA also can be a competent substitution for RIMA to these patients at high risk for SWI.
Limitations The nine included studies are all observational studies, and selection bias and performance bias are inevitable among studies. Due to the inherent differences between study designs of observational study and randomised controlled trial, it is inappropriate to pool the only one randomised trial by Hayward [32] with other observational studies [33], thus it was excluded finally. With the successive publication of analogous studies, an updated meta-analysis with only randomised controlled trials in the future is anticipated to verify our results. Moreover, it is also a topic of interest to compare the survival rate between RIMA and RA. Though the vast majority of studies showed a similar survival between two grafts [8,10,11,16], the required data was not available from these studies for us to make a meta-analysis. At last, the morbidity of SWI from the study by Lemma included the incidence of sternal dehiscence, however the outcome was still stable (RR 2.61, 95% CI 1.23 to 5.55) after the exclusion of this study.
Conclusion The postoperative risk seems to be similar between RIMA group and RA group, except the higher risk of SWI in RIMA patients. The surgeon relies on his own experience with these two arterial grafts and assessment of the patient profile to decide the choice of the second arterial conduit.
Conflict of interest None.
Acknowledgements This work was supported by the project of Shanghai Municipal Bureau of Health (No.20114243), Shanghai Education Commission (11YZ48 and 13ZZ091), and Science and Technology Commission of Shanghai Municipality (12nm0502100).
References [1] Desai ND, Cohen EA, Naylor CD, Fremes SE. A randomized comparison of radial-artery and saphenous-vein coronary bypass grafts. N Engl J Med 2004;351:2302–9.
1186
[2] Gansera B, Schmidtler F, Angelis I, Kiask T, Kemkes BM, Botzenhardt F. Patency of internal thoracic artery compared to vein grafts - postoperative angiographic findings in 1189 symptomatic patients in 12 years. Thorac Cardiovasc Surg 2007;55:412–7. [3] Puskas JD, Sadiq A, Vassiliades TA, Kilgo PD, Lattouf OM. Bilateral internal thoracic artery grafting is associated with significantly improved long-term survival, even among diabetic patients. Ann Thorac Surg 2012;94:710–6. [4] Tranbaugh RF, Dimitrova KR, Friedmann P, Geller CM, Harris LJ, Stelzer P, et al. Radial artery conduits improve long-term survival after coronary artery bypass grafting. Ann Thorac Surg 2010;90:1165–72. [5] El Oakley RM, Wright JE. Postoperative mediastinitis: classification and management. Ann Thorac Surg 1996;61:1030–6. [6] Wells GA, Shea B, O’connell D, Peterson J, Welch V, Losos M, et al. The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analyses. Accessed January 10 2014. Available from: http://www.ohri.ca/programs/clinical_epidemiology. [7] Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med 2002;21:1539–58. [8] Borger MA, Cohen G, Buth KJ, Rao V, Bozinovski J, Liaghati-Nasseri N, et al. Multiple arterial grafts. Radial versus right internal thoracic arteries. Circulation 1998;98(supply.19):II7–II14. [9] Wendler O, Hennen B, Markwirth T, Konig J, Tscholl D, Huang Q, et al. T grafts with the right internal thoracic artery to left internal thoracic artery versus the left internal thoracic artery and radial artery: flow dynamics in the internal thoracic artery main stem. J Thorac Cardiovasc Surg 1999;118:841–8. [10] Lemma M, Gelpi G, Mangini A, Vanelli P, Carro C, Condemi A, et al. Myocardial revascularization with multiple arterial grafts: comparison between the radial artery and the right internal thoracic artery. Ann Thorac Surg 2001;71:1969–73. [11] Caputo M, Reeves B, Marchetto G, Mahesh B, Lim K, Angelini GD. Radial versus right internal thoracic artery as a second arterial conduit for coronary surgery: early and midterm outcomes. J Thorac Cardiovasc Surg 2003;126:39–47. [12] Vicol C, Raab S, Beyer M, Reichart B. Myocardial revascularization using the arterial T graft: which conduit should be chosen for the free graft? Heart Surg Forum 2003;6:353–7. [13] Miana LA, Lima DS, Whitaker JF, Passos PH, Loures JB, Miana AA. Immediate results of right internal thoracic artery and radial artery as the second arterial graft in myocardial revascularization. Rev Bras Cir Cardiovasc 2007;22:60–7. [14] Ruttmann E, Fischler N, Sakic A, Chevtchik O, Alber H, Schistek R, et al. Second internal thoracic artery versus radial artery in coronary artery bypass grafting: a long-term, propensity score-matched follow-up study. Circulation 2011;124:1321–9. [15] Navia D, Vrancic M, Piccinini F, Camporrotondo M, Thierer J, Gil C, et al. Is total the second internal thoracic artery better than the radial artery in arterial off-pump coronary artery bypass grafting? A propensity scorematched follow-up study. J Thorac Cardiovasc Surg 2014;147:632–8. [16] Tranbaugh RF, Dimitrova KR, Lucido DJ, Hoffman DM, Dincheva GR, Geller CM, et al. The second best arterial graft: A propensity analysis of the radial artery versus the free right internal thoracic artery to bypass the circumflex coronary artery. J Thorac Cardiovasc Surg 2014;147: 133–42. [17] Cho WC, Yoo DG, Kim JB, Lee SH, Jung SH, Chung CH, et al. Left internal thoracic artery composite grafting with the right internal
C. Dai et al.
[18]
[19]
[20]
[21]
[22]
[23]
[24] [25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
thoracic versus radial artery in coronary artery bypass grafting. J Card Surg 2011;26:579–85. Davierwala PM, Leontyev S, Misfeld M, Rastan A, Holzhey D, Lehmann S, et al. No-touch aorta off-pump coronary bypass operation: arteriovenous composite grafts may be used as a last resort. Ann Thorac Surg 2013;95:846–52. Tranbaugh RF, Dimitrova KR, Friedmann P, Geller CM, Harris LJ, Stelzer P, et al. Coronary artery bypass grafting using the radial artery: clinical outcomes, patency, and need for reintervention. Circulation 2012;126 (suppl.1):S170–5. Mastrobuoni S, Gawad N, Price J, Chan V, Ruel M, Mesana TG, et al. Use of bilateral internal thoracic artery during coronary artery bypass graft surgery in Canada: The bilateral internal thoracic artery survey. J Thorac Cardiovasc Surg 2012;144:874–9. Seyfer AE, Shriver CD, Miller TR, Graeber GM. Sternal blood flow after median sternotomy and mobilization of the internal mammary arteries. Surgery 1988;104:899–904. Buja A, Zampieron A, Cavalet S, Chiffi D, Sandona P, Vinelli A, et al. An update review on risk factors and scales for prediction of deep sternal wound infections. Int Wound J 2012;9:372–86. Dai C, Lu Z, Zhu H, Xue S, Lian F. Bilateral internal mammary artery grafting and risk of sternal wound infection: evidence from observational studies. Ann Thorac Surg 2013;95:1938–45. Tatoulis J, Buxton BF, Fuller JA. The right internal thoracic artery: is it underutilized? Curr Opin Cardiol 2011;26:528–35. Haycock C, Laser C, Keuth J, Montefour K, Wilson M, Austin K, et al. Implementing evidence-based practice findings to decrease postoperative sternal wound infections following open heart surgery. J Cardiovasc Nurs 2005;20:299–305. Hu X, Zhao Q. Skeletonized internal thoracic artery harvest improves prognosis in high-risk population after coronary artery bypass surgery for good quality grafts. Ann Thorac Surg 2011;92:48–58. Saso S, James D, Vecht JA, Kidher E, Kokotsakis J, Malinovski V, et al. Effect of skeletonization of the internal thoracic artery for coronary revascularization on the incidence of sternal wound infection. Ann Thorac Surg 2010;89:661–70. Deo SV, Shah IK, Dunlay SM, Erwin PJ, Locker C, Altarabsheh SE, et al. Bilateral internal thoracic artery harvest and deep sternal wound infection in diabetic patients. Ann Thorac Surg 2013;95:862–9. Nezic DG, Knezevic AM, Milojevic PS, Dukanovic BP, Jovic MD, Borzanovic MD, et al. The fate of the radial artery conduit in coronary artery bypass grafting surgery. Eur J Cardiothorac Surg 2006;30:341–6. Acar C, Jebara VA, Portoghese M, Beyssen B, Pagny JY, Grare P, et al. Revival of the radial artery for coronary artery bypass grafting. Ann Thorac Surg 1992;54:652–60. De Paulis R, de Notaris S, Scaffa R, Nardella S, Zeitani J, Del Giudice C, et al. The effect of bilateral internal thoracic artery harvesting on superficial and deep sternal infection: The role of skeletonization. J Thorac Cardiovasc Surg 2005;129:536–43. Hayward PA, Hare DL, Gordon I, Matalanis G, Buxton BF. Which arterial conduit? Radial artery versus free right internal thoracic artery: six-year clinical results of a randomized controlled trial. Ann Thorac Surg 2007; 84:493–7. Kirtane AJ, Gupta A, Iyengar S, Moses JW, Leon MB, Applegate R, et al. Safety and efficacy of drug-eluting and bare metal stents: comprehensive meta-analysis of randomized trials and observational studies. Circulation 2009;119:3198–206.
Heart, Lung and Circulation (2014) 23, 1179–1186 1443-9506/04/$36.00 http://dx.doi.org/10.1016/j.hlc.2014.05.017
Comparison of the Safety of Right Internal Mammary Artery Versus Radial Artery for the Second Arterial Conduit During Coronary Artery Bypass Grafting Chenyang Dai, MD, Wentian Zhang, MD, Zhexin Lu, MD, Hongsheng Zhu, MD, PhD, Song Xue, MD, PhD, Feng Lian, MD, PhD * Department of Cardiovascular Surgery, Ren Ji Hospital, School of Medicine, Shanghai Jiao Tong University, Shanghai, People’s Republic of China Received 20 February 2014; received in revised form 12 May 2014; accepted 20 May 2014; online published-ahead-of-print 23 June 2014
Objective
To compare the safety of procedure in patients receiving right internal mammary artery (RIMA) versus radial artery (RA) as the second arterial graft during left internal mammary artery bypass surgery.
Methods
A literature search was performed, and observational studies comparing RA and RIMA as a second arterial conduit were included. The end points included in-hospital mortality, sternal wound infection (SWI), myocardial infarction (MI), reoperation for bleeding, stroke and low cardiac output syndrome. Sensitivity analysis was performed, and publication bias was assessed.
Results
Analysis of nine studies comprising 1,738 RIMA patients and 1,906 RA patients showed a similar risk of inhospital mortality (relative risk [RR] 1.13, 95% confidence interval [CI] 0.64 to 1.98) between two groups. The risk of SWI in RIMA patients was significantly higher than that in RA patients (RR 3.19, 95% CI 1.64 to 6.20). The risk of the rest end points did not differ.
Conclusions
The postoperative risk seems to be similar between RIMA group and RA group, except the higher risk of SWI in RIMA patients. The surgeon relies on his own experience with these two arterial grafts and assessment of the patient profile to decide the choice of the second arterial conduit.
Keywords
Coronary artery bypass Meta-analysis Radial artery Mammary arteries Postoperative complications Surgical wound infection
Introduction The benefits of the use of left internal mammary artery (LIMA) are well established when compared with saphenous vein grafts. Therefore, logic seems to dictate that adding a second arterial graft, which might further improve the longterm effect is an issue of increased interest. RIMA and RA are most commonly used as a second graft of choice. In fact, several studies have demonstrated that adding either RA or RIMA is associated with higher patency rate and better
long-term survival than vein grafts during LIMA bypass surgery [1–4]. Nevertheless, the postoperative high risk restricts the routine adoption of RA or RIMA. Many surgeons are reluctant to adopt bilateral internal mammary artery grafting (BIMA) due to the increased risk of sternal wound infection, including life threatening mediastinitis, which not only increases the length and the expense of hospital stay but also the in-hospital mortality significantly [5]. There is a high propensity of vasospasm at the early stage after radial artery procedure. Once spasm
*Corresponding author at: Department of Cardiovascular Surgery, Ren Ji Hospital, School of Medicine, Shanghai Jiao Tong University, No. 160 Pujian Road, Shanghai, 200127, P.R. China; Fax: +86-21-68383764., Email: [email protected] © 2014 Australian and New Zealand Society of Cardiac and Thoracic Surgeons (ANZSCTS) and the Cardiac Society of Australia and New Zealand (CSANZ). Published by Elsevier Inc. All rights reserved.
1180
occurs, the decreased blood supply might lead to a higher postoperative risk since RA graft is often used as T connection, providing blood supply to more than one target vessel in sequential fashion. Hence, the safety of procedure is often an important issue for surgeons. Yet, it is still unclear which arterial graft is accompanied by lower postoperative risk. The aim of this meta-analysis is to assess the postoperative safety by comparing early outcomes in patients receiving RIMA versus RA as the second arterial graft during LIMA bypass surgery.
Methods Literature Search A literature search was performed in PubMed, Embase, and CENTRAL with ‘‘Coronary artery bypass’’ as the MESH search heading. Headings ‘‘right internal mammary artery’’, ‘‘right internal thoracic artery’’, ‘‘bilateral internal mammary artery’’, ‘‘bilateral internal thoracic artery’’ and ‘‘radial artery’’ were also applied. The articles were restricted to English and published until the end of January 10, 2014. To broaden the search, we also screened the references of articles and made use of the ‘‘Related articles’’ function in PubMed.
Data Extraction and Quality Assessment Two reviewers (C.D. and W.Z) extracted these essential data from each study including publication year, first author, population characteristics, sample size, and the mortality and morbidity after operation (Tables 2 and 3). The quality of each study was assessed by a modified Newcastle-Ottawa scale (Table 1) [6]. Total score was nine stars, and a study with
C. Dai et al.
more than four stars was graded as high level of quality. Data extraction and quality assessment were carried out independently by two reviewers.
Inclusion and Exclusion Criteria The included studies should satisfy the following criteria: (1) observational studies comparing LIMA+RIMA versus LIMA+RA with or without concomitant saphenous vein grafts; (2) when several studies were reported from the same institution, the most recent study or study with most information would be included. If there was a potential possibility that a third arterial conduit was used, the study would be excluded if a confirmation email was received from the author or correspondence with the author failed.
Outcome of Interest The primary end point was in-hospital mortality defined as a death occurring within 30 days after operation. Secondary end points included SWI and MI. SWI includes superficial SWI and deep SWI. The former was defined as infection and was limited in subcutaneous tissue, and the latter was defined as infection involving sternal osteomyelitis with or without mediastinitis. MI was defined as the appearance of new Q wave in at least two leads or left bundle branch block on the electrocardiogram or the creatine kinase myocardial band (CKMB) isoenzyme fraction increased more than 30 IU/L or more than 7% of the total creatine kinase level. Additional supplementary end points included reoperation for bleeding, stroke and low cardiac output syndrome.
Statistical Analysis Statistical analysis was performed by taking relative risk (RR) as the summary statistic. The RR represents the risk of
Table 1 Check List for Quality Assessment. Selection 1. Were any criteria for assignment of patients to treatment reported? (If yes, 1 star). 2. How representative was right internal mammary artery group? (1 star if basically representative of general population for surgery, no star if the patients were selected or selection of group was not reported). 3. How was radial artery group selected? (1 star if selected from the same population as exposed group, no star if chosen from a different population or selection of group was not represented). Comparability 1. Group comparable for age, sex and diabetes. (2 stars if no differences of the above characteristics between 2 groups; 1 star if one of the 3 characteristics was not represented or one characteristic was different between 2 groups; no star if 2 groups differed) 2. Group comparable for chronic obstructive pulmonary disease, body mass index, and ventricular function. (2 stars if no differences of the above characteristics between 2 groups; 1 star if one of the 3 characteristics was not represented or one characteristic was different between 2 groups; no star if 2 groups differed) Outcomes 1. Assessment of outcome (1 star if outcome was ascertained by medical record or interview, no star if not reported). 2. Adequacy of cohort follow-up (1 star if no patient or fewer than 10% of patients were lost to follow-up; no star if more than 10% of patients were lost).
Study
Year
Patients, n RIMA/RA
Age, mean years RIMA/RA
Sex (% men) RIMA/RA
Diabetes (%) RIMA/RA
COPD(%) RIMA/RA
BMI(mean) RIMA/RA
Ventricular function (mean or %) Classification
RIMA/RA
Score
Borger et al. [8]
1998
132/171
NS
92/88
11/27
NS
NS
Ejection fraction < 40%
14/18
Wendler et al. [9]
1999
73/178
61.4/60.5
89/78.7
NS
NS
NS
Ejection fraction
58.1/55.3
3
Lemma et al. [10]
2001
94/156
57/60
91.5/87.8
7.4/19.2
3.2/15.3
NS
Ejection fraction
54.2/56.5
4
Caputo et al. [11]
2003
336/325
55.5/57.7
91.4/84.3
6.3/18.2
NS
27.5/28.2
Poor ejection fraction
13.8/26.2
5
Vicol et al. [12]
2003
129/84
59.2/61.7
89.2/76.2
19.4/27.4
NS
NS
Ejection fraction
58.1/62
4
Miana et al. [13]
2007
20/38
57/56
80/76.3
15/21
NS
NS
Left ventricular dysfunction
15/18.4
7
Ruttmann et al. [14]
2011
277/277
56.6/57.8
89.9/89.9
21.3/22.4
33.2/33.2
27.4/27.7
Ejection fraction
54.9/52.9
8
Navia et al. [15] Tranbaugh et al. [16]
2014 2014
149/149 528/528
68/67 61/60
82.5/81.2 76.9/78.4
30.8/32.2 35.6/36.6
NS 10/10.4
NS 28.7/28.8
Severe ventricular function Ejection fraction
30.2/30.2 47.5/47.3
6 8
RIMA versus RA
Table 2 Characteristics of studies used in the meta-analysis.
5
RIMA, right internal mammary artery; RA, radial artery; NS, not stated; COPD, chronic obstructive pulmonary disease; BMI, body mass index.
Table 3
Incidence of end points: RIMA versus RA as the second arterial conduit.
Study
Patients, n
In-hospital
SWI, n(%)
MI, n(%)
Reoperation for
Stroke, n(%)
Low cardiac output
RIMA/RA
mortality, n(%)
RIMA/RA
RIMA/RA
bleeding, n(%)
RIMA/RA
syndrome, n(%)
RIMA/RA
RIMA/RA
RIMA/RA
Borger et al. [8]
132/171
0(0)/1(0.58)
7(5.30)/1(0.58)
2(1.52)/2(1.17)
1(0.76)/1(0.58)
1(0.76)/0(0)
Wendler et al. [9]
73/178
2(2.74)/4(2.25)
0(0)/1(0.56)
2(2.74)/4(2.25)
0(0)/1(0.56)
0(0)/4(2.25)
8(6.06)/10(5.85) 1(1.37)/3(1.69)
Lemma et al. [10]
94/156
0(0)/3(1.92)
8(8.51)/2(1.28)
2(2.13)/6(3.85)
1(1.06)/1(0.64)
NS
2(2.13)/3(1.92)
Caputo et al. [11]
336/325
3(0.89)/1(0.31)
NS
7(2.08)/1(0.31)
NS
2(0.6)/2(0.62)
NS
Vicol et al. [12]
129/84
1(0.78)/1(1.19)
1(0.78)/1(1.19)
4(3.1)/0(0)
4(3.1)/1(1.19)
NS
3(2.33)/1(1.19)
Miana et al. [13]
20/38
0(0)/1(2.63)
0(0)/0(0)
0(0)/1(2.63)
0(0)/1(2.63)
0(0)/1(2.63)
0(0)/1(2.63)
Ruttmann et al. [14] Navia et al. [15]
277/277 149/149
2(0.72)/5(1.81)a 6(4.03)/5(3.36)
NS NS
1(0.36)/10(3.61) NS
11(3.97)/18(6.5) NS
1(0.36)/10(3.61) NS
NS NS
Tranbaugh et al. [16]
528/528
9(1.7)/3(0.6)
14(2.7)/6(1.1)
3(0.6)/6(1.1)
10(1.9)/8(1.5)
11(2.1)/4(0.8)
NS
RIMA, right internal mammary artery; RA, radial artery; SWI, sternal wound infection; MI, myocardial infarction; NS, not stated; a cardiac related death.
1181
1182
postoperative event in RIMA group divided by that in RA group. RR less than 1 favours RIMA, whereas more than 1 favours RA. The point estimate of RR is considered statistically significant at the level of p value less than 0.05 if the 95% confidence interval (CI) does not include the value 1. We chose I2 statistic to measure the heterogeneity between studies. According to the results of I2 statistic, the level of heterogeneity was classified as low (less than 30%), moderate (31% to 50%) and high (more than 50%) [7]. When the heterogeneity between studies was low or moderate, a fixedeffect model would be chosen to combine the results; if not, a random-effect model would be adopted. Sensitivity analysis was undertaken to examine the stability of the results. Firstly, the results were recalculated by using another effect model. Secondly, two subgroup analyses were assessed: (1) excluded the study with small series (less than 100 patients in either group); (2) included the studies of high quality. Also, publication bias was evaluated by drawing funnel plot and performing Egger’s test and Begg’s test. All these statistical analyses were conducted by using Review Manager 5.1 (The Cochrane Collaboration, Oxford, UK).
C. Dai et al.
Primary End Point: Mortality All of nine included studies reported in-hospital mortality. The incidence of mortality was 1.32% (23 of 1738 patients) in RIMA group versus 1.26% (24 of 1906 patients) in RA group, and the RR was calculated as 1.13 (95% CI 0.64 to 1.98) (Fig. 1) and there was no significant heterogeneity among studies (I2 = 0%).
Second End Point: SWI and MI
Results
Incidence of SWI was available from six of nine included studies. The incidence of SWI was 3.07% in RIMA group and 0.95% in RA group. The RR showed that RIMA group was associated with a higher risk of SWI versus RA group (RR 3.19, 95% CI 1.64 to 6.20) (Fig. 2A) with low heterogeneity (I2 = 6%). A total of eight studies reported the data on MI, and the incidence in RIMA group was 1.32% compared with 1.71% in RA group. As shown in Figure 2B, the RR was calculated as 0.83 (95% CI 0.48 to 1.46) with moderate heterogeneity (I2 = 36%). It should be pointed out that the definition of MI from seven of the included studies was the same or similar to ours, and the remaining one from Ruttmann was defined as the appearance of new Q wave or the elevation of troponin which was more precise than CKMB. Even after the exclusion of this study, there was still no significant difference in the risk of MI among studies (p = 0.48, RR 1.26, 95% CI 0.66 to 2.37).
Selected Studies
Other End Points
A total of 3427 articles were searched from three electronic databases, and 11 studies met the inclusion criteria after review. Two studies with the potential possibility of using a third arterial conduit were excluded, because the authors failed to answer our query letter by email [17,18]. In all, nine observational studies with 3,644 patients were included in this meta-analysis, of whom 1,738 received RIMA graft and 1,906 received RA graft (Tables 2 and 3).
Seven studies demonstrated the similar risk of reoperation for bleeding between RIMA group and RA group (2.15% versus 2.16%, RR 0.91, 95% CI 0.55 to 1.50) with no heterogeneity (I2 = 0%) (Fig. 3A). Six studies reported the incidence of stroke between RIMA group and RA group (1.1% versus 1.39%), and the RR calculated from these studies was 0.84 (95% CI 0.44 to 1.61), also there was no high heterogeneity (I2 = 49%) (Fig. 3B). In the five studies, the risk of low cardiac
Figure 1 Meta-analysis of in-hospital mortality comparing right internal mammary artery versus radial artery as the second arterial conduit.
1183
RIMA versus RA
Figure 2 Meta-analysis of sternal wound infection (A) and myocardial infarction (B) comparing right internal mammary artery versus radial artery as the second arterial conduit.
output syndrome in RIMA group was comparable with that in RA group (3.13% versus 2.87%, RR 1.07, 95% CI 0.53 to 2.13), and no heterogeneity was found (I2 = 0%) (Fig. 3C).
Table 4 revealed that the outcomes of all end points were hardly changed and quite stable.
Sensitivity Analysis
Funnel plot was drawn for in-hospital mortality reported by all nine studies showing basic symmetry (Fig. 4). In addition, the results of both Egger’s test (p = 0.182) and Begg’s test
Different strategies were adopted to perform the sensitivity analysis for these six end points. Those results shown in
Assessment on Publication Bias
Table 4 The results of sensitivity analysis. Changing effect model
Excluding the small studies
High-quality studies
No. studies
RR(95% CI)
P value
No. studies
RR(95% CI)
P value
No. studies
RR(95% CI)
P value
In-hospital mortality
9
1.15(0.62-2.10)
0.66
5
1.34(0.69-2.58)
0.38
6
1.29(0.68-2.45)
0.43
SWI
6
3.01(1.41-6.42)
0.004
2
3.19(1.38-7.36)
0.007
3
3.19(1.38-7.36)
0.007
MI
8
0.89(0.37-2.10)
0.78
4
0.79(0.17-3.73)
0.76
5
0.76(0.20-2.85)
0.68
Reoperation
7
0.89(0.53-1.50)
0.67
3
0.82(0.47-1.43)
0.49
4
0.82(0.47-1.41)
0.46
for bleeding Stroke
6
0.81(0.23-2.83)
0.74
4
0.98(0.19-5.11)
0.98
5
0.93(0.23-3.77)
0.92
Low cardiac
5
1.06(0.53-2.13)
0.87
1
1.04(0.42-2.55)
0.94
2
0.99(0.42-2.36)
0.98
output syndrome SWI, sternal wound infection; MI, myocardial infarction; RR, relative risk; CI, confidence interval
1184
C. Dai et al.
Figure 3 Meta-analysis of reoperation for bleeding (A), stroke (B) and low cardiac output syndrome (C) comparing right internal mammary artery versus radial artery as the second arterial conduit.
(p = 0.251) demonstrated that no notable publication bias was identified.
Discussion
Figure 4 Funnel plot for in-hospital mortality showing a symmetrical plot. (SE = standard error).
Since the benefits of less trauma in nature and improving therapeutic outcome achieved by the interventional cardiologist, surgeons have been aiming to improve the long-term outcomes of bypass surgery. Adding a RIMA or RA as a second arterial conduit is believed an effective strategy to optimise the prognosis of bypass surgery. In spite of the encouraging report early this year by Tranbaugh showing the five, 10, 15-year survival rate reached 92%, 80% and 71% for RIMA patients and 95%, 85% and 76% for RA patients respectively [16], only 4% of bypass operations adopted BIMA grafting, and the use of RA graft was 9% in clinical
1185
RIMA versus RA
practice [19]. The concern about the perioperative safety of procedure is one of the reasons for this present clinical status. In fact our study shows that the in-hospital mortality, as a primary indicator of safety is acceptable in both two groups (1.32% in RIMA and 1.26% in RA). As to the protection for surgical safety, both arterial grafts showed a similar risk of postoperative adverse events, except the higher risk of SWI in RIMA group. A recent survey reported that SWI was the common concern regarding the extensive use of RIMA even among experienced cardiac surgeons [20]. Our meta-analysis shows that their concerns are valid and well founded. The blood deprivation of hemi-sternum reached 90% after the harvest of one internal mammary artery [21], so the chance of SWI increased significantly when both LIMA and RIMA were utilised simultaneously. Several studies also demonstrated that patients with advanced age, diabetes mellitus or chronic obstructive pulmonary disease endured much higher risk of SWI [22,23]. Hence, SWI seems to be a negative impact on BIMA grafting. Even so, it does not mean that BIMA should be fully abandoned in clinical practice. The increased risk of SWI had been observed after BIMA surgery, while morbidity was relatively low especially deep SWI (ranging from 1.2% to 2.5%) [24]. In fact, BIMA was just one of the initial factors for postoperative SWI. The risk of infection can be reduced to a minimum if effort is made to prevent other risk factors, such as the overall discovery of latent inflammatory foci and effective control of metabolic disturbance preoperatively, combined with the meticulous manipulation and thorough haemostasis intraoperatively, followed by the strict implementation of standard wound care postoperatively, and the appropriate use of antibiotic during the entire perioperative period [24,25]. More than anything else, a recent metaanalysis showed that the utilisation of skeletonisation technique preserved more sternal perfusion after bypass surgery [26]. Saso and colleagues found that skeletonised BIMA was associated with a notably lower risk of SWI than pedicled BIMA [27]. Moreover, the lower risk was also comparable with that of single internal mammary artery grafting [23], and this favourable effect even extended to diabetic patients [28]. As to the safety of RA graft, surgeons often worry about its susceptibility to vasospasm during the early period after operation. Our result shows that the incidence of MI is not significantly higher in RA group, which is attributed to the intraluminal irrigation of vasodilator drugs, for example, papavarine, verapamil and diltiazem, either singly or in combination [8–10,12–16]. Gentle harvesting of RA such as light tractions, ultrasonicscalpel dissection or endoscopic technique also contributes to prevent vasospasm [29]. In addition, some other potential complications for RA like superficial radial nerve injury even reported as less than 10% can be prevented by avoiding excessive lateral traction of brachioradialis muscle [30], and haematoma formation can also be prevented by careful and secure ligation of the branches and the stumps. Even the rare but devastating compartment syndrome can also be avoided by close observation and regular check of the hand
after radial artery harvesting for unrestricted movement and intact sensation at routine intervals. Since the negative aspect of RA graft can be managed, it should be considered as a reasonable and safe option. Furthermore if a patient is complicated with multiple risk factors of SWI, the risk of SWI increases exponentially when BIMA grafting is adopted [31]. However, SWI is not a major concern for RA graft, and its contraindications including carpal tunnel syndrome, Dupuytren disease, etc. are quite rare. Hence, RA also can be a competent substitution for RIMA to these patients at high risk for SWI.
Limitations The nine included studies are all observational studies, and selection bias and performance bias are inevitable among studies. Due to the inherent differences between study designs of observational study and randomised controlled trial, it is inappropriate to pool the only one randomised trial by Hayward [32] with other observational studies [33], thus it was excluded finally. With the successive publication of analogous studies, an updated meta-analysis with only randomised controlled trials in the future is anticipated to verify our results. Moreover, it is also a topic of interest to compare the survival rate between RIMA and RA. Though the vast majority of studies showed a similar survival between two grafts [8,10,11,16], the required data was not available from these studies for us to make a meta-analysis. At last, the morbidity of SWI from the study by Lemma included the incidence of sternal dehiscence, however the outcome was still stable (RR 2.61, 95% CI 1.23 to 5.55) after the exclusion of this study.
Conclusion The postoperative risk seems to be similar between RIMA group and RA group, except the higher risk of SWI in RIMA patients. The surgeon relies on his own experience with these two arterial grafts and assessment of the patient profile to decide the choice of the second arterial conduit.
Conflict of interest None.
Acknowledgements This work was supported by the project of Shanghai Municipal Bureau of Health (No.20114243), Shanghai Education Commission (11YZ48 and 13ZZ091), and Science and Technology Commission of Shanghai Municipality (12nm0502100).
References [1] Desai ND, Cohen EA, Naylor CD, Fremes SE. A randomized comparison of radial-artery and saphenous-vein coronary bypass grafts. N Engl J Med 2004;351:2302–9.
1186
[2] Gansera B, Schmidtler F, Angelis I, Kiask T, Kemkes BM, Botzenhardt F. Patency of internal thoracic artery compared to vein grafts - postoperative angiographic findings in 1189 symptomatic patients in 12 years. Thorac Cardiovasc Surg 2007;55:412–7. [3] Puskas JD, Sadiq A, Vassiliades TA, Kilgo PD, Lattouf OM. Bilateral internal thoracic artery grafting is associated with significantly improved long-term survival, even among diabetic patients. Ann Thorac Surg 2012;94:710–6. [4] Tranbaugh RF, Dimitrova KR, Friedmann P, Geller CM, Harris LJ, Stelzer P, et al. Radial artery conduits improve long-term survival after coronary artery bypass grafting. Ann Thorac Surg 2010;90:1165–72. [5] El Oakley RM, Wright JE. Postoperative mediastinitis: classification and management. Ann Thorac Surg 1996;61:1030–6. [6] Wells GA, Shea B, O’connell D, Peterson J, Welch V, Losos M, et al. The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analyses. Accessed January 10 2014. Available from: http://www.ohri.ca/programs/clinical_epidemiology. [7] Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med 2002;21:1539–58. [8] Borger MA, Cohen G, Buth KJ, Rao V, Bozinovski J, Liaghati-Nasseri N, et al. Multiple arterial grafts. Radial versus right internal thoracic arteries. Circulation 1998;98(supply.19):II7–II14. [9] Wendler O, Hennen B, Markwirth T, Konig J, Tscholl D, Huang Q, et al. T grafts with the right internal thoracic artery to left internal thoracic artery versus the left internal thoracic artery and radial artery: flow dynamics in the internal thoracic artery main stem. J Thorac Cardiovasc Surg 1999;118:841–8. [10] Lemma M, Gelpi G, Mangini A, Vanelli P, Carro C, Condemi A, et al. Myocardial revascularization with multiple arterial grafts: comparison between the radial artery and the right internal thoracic artery. Ann Thorac Surg 2001;71:1969–73. [11] Caputo M, Reeves B, Marchetto G, Mahesh B, Lim K, Angelini GD. Radial versus right internal thoracic artery as a second arterial conduit for coronary surgery: early and midterm outcomes. J Thorac Cardiovasc Surg 2003;126:39–47. [12] Vicol C, Raab S, Beyer M, Reichart B. Myocardial revascularization using the arterial T graft: which conduit should be chosen for the free graft? Heart Surg Forum 2003;6:353–7. [13] Miana LA, Lima DS, Whitaker JF, Passos PH, Loures JB, Miana AA. Immediate results of right internal thoracic artery and radial artery as the second arterial graft in myocardial revascularization. Rev Bras Cir Cardiovasc 2007;22:60–7. [14] Ruttmann E, Fischler N, Sakic A, Chevtchik O, Alber H, Schistek R, et al. Second internal thoracic artery versus radial artery in coronary artery bypass grafting: a long-term, propensity score-matched follow-up study. Circulation 2011;124:1321–9. [15] Navia D, Vrancic M, Piccinini F, Camporrotondo M, Thierer J, Gil C, et al. Is total the second internal thoracic artery better than the radial artery in arterial off-pump coronary artery bypass grafting? A propensity scorematched follow-up study. J Thorac Cardiovasc Surg 2014;147:632–8. [16] Tranbaugh RF, Dimitrova KR, Lucido DJ, Hoffman DM, Dincheva GR, Geller CM, et al. The second best arterial graft: A propensity analysis of the radial artery versus the free right internal thoracic artery to bypass the circumflex coronary artery. J Thorac Cardiovasc Surg 2014;147: 133–42. [17] Cho WC, Yoo DG, Kim JB, Lee SH, Jung SH, Chung CH, et al. Left internal thoracic artery composite grafting with the right internal
C. Dai et al.
[18]
[19]
[20]
[21]
[22]
[23]
[24] [25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
thoracic versus radial artery in coronary artery bypass grafting. J Card Surg 2011;26:579–85. Davierwala PM, Leontyev S, Misfeld M, Rastan A, Holzhey D, Lehmann S, et al. No-touch aorta off-pump coronary bypass operation: arteriovenous composite grafts may be used as a last resort. Ann Thorac Surg 2013;95:846–52. Tranbaugh RF, Dimitrova KR, Friedmann P, Geller CM, Harris LJ, Stelzer P, et al. Coronary artery bypass grafting using the radial artery: clinical outcomes, patency, and need for reintervention. Circulation 2012;126 (suppl.1):S170–5. Mastrobuoni S, Gawad N, Price J, Chan V, Ruel M, Mesana TG, et al. Use of bilateral internal thoracic artery during coronary artery bypass graft surgery in Canada: The bilateral internal thoracic artery survey. J Thorac Cardiovasc Surg 2012;144:874–9. Seyfer AE, Shriver CD, Miller TR, Graeber GM. Sternal blood flow after median sternotomy and mobilization of the internal mammary arteries. Surgery 1988;104:899–904. Buja A, Zampieron A, Cavalet S, Chiffi D, Sandona P, Vinelli A, et al. An update review on risk factors and scales for prediction of deep sternal wound infections. Int Wound J 2012;9:372–86. Dai C, Lu Z, Zhu H, Xue S, Lian F. Bilateral internal mammary artery grafting and risk of sternal wound infection: evidence from observational studies. Ann Thorac Surg 2013;95:1938–45. Tatoulis J, Buxton BF, Fuller JA. The right internal thoracic artery: is it underutilized? Curr Opin Cardiol 2011;26:528–35. Haycock C, Laser C, Keuth J, Montefour K, Wilson M, Austin K, et al. Implementing evidence-based practice findings to decrease postoperative sternal wound infections following open heart surgery. J Cardiovasc Nurs 2005;20:299–305. Hu X, Zhao Q. Skeletonized internal thoracic artery harvest improves prognosis in high-risk population after coronary artery bypass surgery for good quality grafts. Ann Thorac Surg 2011;92:48–58. Saso S, James D, Vecht JA, Kidher E, Kokotsakis J, Malinovski V, et al. Effect of skeletonization of the internal thoracic artery for coronary revascularization on the incidence of sternal wound infection. Ann Thorac Surg 2010;89:661–70. Deo SV, Shah IK, Dunlay SM, Erwin PJ, Locker C, Altarabsheh SE, et al. Bilateral internal thoracic artery harvest and deep sternal wound infection in diabetic patients. Ann Thorac Surg 2013;95:862–9. Nezic DG, Knezevic AM, Milojevic PS, Dukanovic BP, Jovic MD, Borzanovic MD, et al. The fate of the radial artery conduit in coronary artery bypass grafting surgery. Eur J Cardiothorac Surg 2006;30:341–6. Acar C, Jebara VA, Portoghese M, Beyssen B, Pagny JY, Grare P, et al. Revival of the radial artery for coronary artery bypass grafting. Ann Thorac Surg 1992;54:652–60. De Paulis R, de Notaris S, Scaffa R, Nardella S, Zeitani J, Del Giudice C, et al. The effect of bilateral internal thoracic artery harvesting on superficial and deep sternal infection: The role of skeletonization. J Thorac Cardiovasc Surg 2005;129:536–43. Hayward PA, Hare DL, Gordon I, Matalanis G, Buxton BF. Which arterial conduit? Radial artery versus free right internal thoracic artery: six-year clinical results of a randomized controlled trial. Ann Thorac Surg 2007; 84:493–7. Kirtane AJ, Gupta A, Iyengar S, Moses JW, Leon MB, Applegate R, et al. Safety and efficacy of drug-eluting and bare metal stents: comprehensive meta-analysis of randomized trials and observational studies. Circulation 2009;119:3198–206.