Surgical and early-term outcomes in skeletonized vs. pedicled IMA conduits for CABG Authors: Rui Fan, Yifei Zhou, Yifan Zhu, Wuwei Wang, Fei Xiang, Wei Qin, Xin Chen Published: 2025 ABSTRACT AbstractObjectiveThe aim of this study was to compare the outcomes of skeletonized vs. pedicled internal mammary artery graft in coronary artery bypass grafting.MethodsWe retrospectively enrolled patients undergoing coronary artery bypass grafting in Nanjing First Hospital from 2020 to 2023. Patients were divided into two groups, skeletonized and pedicled internal mammary artery group. The outcomes were in-hospital mortality, deep sternal wound infection, blood flow and surgery duration. Kaplan-Meier estimation was used to depict the survival situation. Propensity score matching was used to minimize the effect of cofounders. Subgroup analysis was used to explore the outcomes in different population.ResultsA total of 2353 patients undergoing coronary artery bypass grafting were enrolled in this study. 1360 patients (57.8%) received skeletonized internal mammary artery and 993 patients (42.2%) received pedicled internal mammary artery. In primary and matched cohorts, the outcomes were not significantly different, except for blood flow. We found that blood flow of conduit and pulse index in skeletonized internal mammary artery were better that those in pedicle internal mammary artery in primary cohort(blood flow: 33vs31, P = 0.002 and pulse index: 1.9vs2.0, P < 0.001) and matched cohort(blood flow: 32vs30, P = 0.002 and pulse index: 1.9vs2.1, P < 0.001). The Kaplan-Merier curves showed that the survival rates were not significant difference between the two groups (P = 0.75). The results of subgroup analysis showed the similar results in different population.ConclusionWhether skeletonized or pedicled internal mammary artery conduits did not affect the in-hospital mortality, deep sternal wound infection, surgical duration and survival rate. Skeletonized internal mammary artery harvesting may offer physiological advantages in terms of graft flow in experienced hands, but further studies are needed to confirm its safety and clinical benefit.Supplementary InformationThe online version contains supplementary material available at 10.1186/s13019-025-03737-8.Keywords: Coronary artery bypass grafting, Skeletonized internal mammary artery, Pedicled internal mammary artery Introduction Due to the long-term patency, internal mammary artery (IMA) to left anterior descending coronary (LAD) has become a gold standard for CABG, bringing less heart adverse events, such as recurrent angina, late myocardial infarction and repeat coronary intervention, and improvement in long-term survival [1–3]. A lower rates of lung complications and sternal pain after CABG with IMA conduits were observed in several researches [4]. There are two main IMA harvesting techniques, skeletonized and pedicled IMA. Compared with pedicled IMA harvesting, Skeletonized IMA harvesting can preserve the blood supply to the sternum and avoid the thermal injury from electronic cauterization, which decrease the risk of sternal wound infection [5]. As known, sternal wound infection can be life-threatening and prolong the hospital staying duration, even can cause second surgical intervention [6, 7]. What’s more, surgeons can get longer IMA through skeletonization, which provides more possibility for revascularization without using vein and provides long-term patency as soon as possible [5]. Diameter limitation decreases the flow rate in IMA. Skeletonization can help to acquire bigger diameter and get higher flow rate in conduits [8]. However, this procedure is more complex and require longer learning curve. When harvesting, it is easy to stress the vessel and result in vaso-spasm, which might affect postoperative results [9]. Harvesting pedicled IMA is a relatively simple and easy technique to master for surgeons. The conduits were harvested with the surrounding veins, fascia, adipose tissue and lymphatics, which avoids IMA being injuried and keep the artery in the normal physiological environment [10]. However, harvesting pedicled IMA reduce the blood supply to sternum and also cause sternal vessel dissection, which would cause sternal infection [11]. Shorter length and smaller diameter of conduits would also have an impact on the blood flow. There were no many researches to give the strong evidence to show that the surgical outcomes or follow-up outcomes between the two different harvesting technique. What’s more, no consensus or guideline was constructed on how to choose the harvesting technique in patients. The purpose of this study was to compare the in-hospital mortality and early-term survival between the two groups. Method Population This study collected a consecutive group of CABG patients from June 2020 to December 2023 in Nanjing First Hospital. Finally, 2353 patients were enrolled from database in our department. All of these patients received IMA to LAD graft. Patients with age < 18 or > 80 and missing data were excluded. According to the harvest technique, patients were divided into two groups, skeletonized IMA group(n = 1360) and pedicled IMA group(n = 993). The two groups were matched in a 1:1 ratio using propensity score. The patients’ allocation was shown in Fig. 1. This study was performed and approved by the Institutional Review Board (IRB) of Nanjing First Hospital (IRB No: KY-20240603-KS-03). The flow chart of this study Outcomes The outcomes were in-hospital mortality, deep sternal wound infection (DSWI), blood flow and operative duration. Grafting strategy In our center, surgeons always try their best to use LIMA to LAD, due to its perfect long-term patency. If following situation comes, SVG will be used to graft LAD, such as LIMA injury when harvesting, or blood flow of LIMA not good, or small diameter of LIMA, or emergency surgery. If patients were young and Allen test was negative, surgeons would try to use radial artery as conduit. Gastroepiploic artery is not a common choice in our center. SVG is always a choice for grafting to other branches. Surgical procedure All IMA harvesting were performed by skilled doctors. After standard median sternotomy, the left hemisternum was elevated by IMA retractor (Favaloro retractor or Rultract system retractor). The voltage levels for electrocautery was changed to 15 J. The pleura was usually opened longitudinally. The pedicled IMA was accompanied by veins, fat, fascia and lymphatics using high voltage electrocautery. When IMA was harvested as a skeletonized technique, the incision was performed between the artery and accompanying vein, which was performed with low voltage electrocautery to avoid the thermal damage to the artery wall. The branches would be occluded by clips or transected by scissors. The full length of IMA was harvested from proximal part to the distal bifurcation, usually at the sixth intercostal space. When CABG was done, transit time flow probes were used and that they were used in all patients. Surgeons performed TTFM after all anastomosis finishing and before closing the pericardium. When the mean arterial pressure (MAP) exceeding 80 mmHg and the heart rhythm and cardiac output were stable, the probe (Medistim) was used to test blood flow. Statistical analysis Continuous variables are presented as mean ± standard deviation (SD) for normally distributed parameters and median with interquartile range for non-normally distributed parameters. Categorical variables are showed as percentages. Patients were matched in a ratio 1:1 by using propensity score matching. The caliper was 0.02, which is higher demanding when matching. Matching covariates were sex, age, body mass index, smoking, hypertension, diabetes, hypelipidemia, dialysis, peripheral artery disease, chronic obstructive pulmonary disease, cerebral vessel disease, myocardial infarction, acute coronary syndrome, left main branch disease. Kaplan-Meier method was used to present the survival situation. Subgroup analysis was used to explore the outcomes in different population. P value less than 0.05 was considered statistically significant. SPSS version 27 and Rstudio version 4.2.1 were used to analyze. Results The primary cohort 2353 patients were enrolled in this study. 1360 patients (57.8%) received skeletonized IMA and 993 patients (42.2%) received pedicled IMA. In skeletonized IMA group, more patients underwent AVR procedure (8.8% vs. 6.3%, P = 0.031) in primary cohort. However, to ventricular aneurysm resection, more patients underwent this procedure in pedicled IMA group (2.4% vs. 1.2%, P = 0.022). The blood flow of conduit in skeletonized IMA was better that that in pedicled IMA group (33 vs 31, P = 0.002). PI of conduit in two groups showed the similar results (1.9 vs 2.0, P < 0.001). Between two groups, the outcomes were not significantly different (Tables 1, 2 and 3). The median follow-up time was 358 days in primary cohort. The Kaplan-Merier curves showed that the survival rates were not significant difference between the two groups (P = 0.75) (Fig. 2). Table 1Baseline characteristics in primary cohortVariablesSkeletonized IMA(n = 1360)Pedicle IMA(n = 993)P valueSex(male)1017(74.8%)730(73.5%)0.488Age63.75 ± 9.2564.25 ± 9.30.195BMI24.72 ± 3.1124.87 ± 3.450.261Smoking637(58.6%)450(45.3%)0.465Hypertension922(57.1%)693(69.8%)0.303Diabetes572(42.1%)426(42.9%)0.683Hyperlipidemia216(15.9%)229(23.1%)< 0.001Dialysis7(0.5%)19(1.9%)0.001COPD33(2.4%)28(2.8%)0.553PAD45(3.3%)51(5.1%)0.027CVD150(11%)131(13.2%)0.110Stroke9(0.7%)9(0.9%)0.501MI276(20.3%)228(23%)0.119ACS381(28%)349(35.1%)< 0.001Cardiogenic shock8(0.6%)9(0.9%)0.368Pre-op intubation6(0.4%)4(0.4%)0.888PCI history152(11.2%)111(11.2%)0.999Numbers of lesion vessel0.804 182(6%)59(5.9%) 2168(12.4%)114(11.5%) 31110(81.6%)820(82.6%)Left main lesion vessel363(26.7%)342(34.4%)< 0.001Pre-op Scr75(64,90.1)75.2(65,90.1)0.169Pre-op LVEF62(49,65)62(50,65)0.205Euroscore1.4(0.96,2.4)1.8(1.12,2.9)< 0.001BMI: Body Mass Index, PAD: Peripheral Artery Disease, COPD: Chronic Obstructive Pulmonary Disease, CVD: Cerebral Vessel Disease, MI: Myocardial Infarction, ACS: Acute Coronary Syndrome, PCI: Percutaneous Coronary Intervention, Scr: Serum Creatinine, LVEF: Left Ventricular Eject Fraction Baseline characteristics in primary cohort BMI: Body Mass Index, PAD: Peripheral Artery Disease, COPD: Chronic Obstructive Pulmonary Disease, CVD: Cerebral Vessel Disease, MI: Myocardial Infarction, ACS: Acute Coronary Syndrome, PCI: Percutaneous Coronary Intervention, Scr: Serum Creatinine, LVEF: Left Ventricular Eject Fraction Table 2Intraoperative parameters in primary cohortVariablesSkeletonized IMA(n = 1360)Pedicle IMA(n = 993)P valueEmergency surgery42(3.1%)29(2.9%)0.814Radial artery80(5.9%)69(6.9%)0.294AVR119(8.8%)63(6.3%)0.031MVR55(4%)29(2.9%)0.147Aortic procedure29(2.1%)15(1.5%)0.271Ventricular aneurysm resection16(1.2%)24(2.4%)0.022Operation duration(min)275(240,315)287.5(250,330)0.076CPB duration(min)102(85,127)107(83,132)0.08Cross-clamp time(min)67(53,87)71(54,91.8)0.118Blood flow(ml/min)33(22,48)31(19,43)0.002PI1.9(1.5,2.4)2.0(1.6,2.7)< 0.001AVR: Aortic Valve Replacement, MVR: Mitral Valve Replacement, CPB: Cardiopulmonary Bypass, PI: Pulse Index Intraoperative parameters in primary cohort AVR: Aortic Valve Replacement, MVR: Mitral Valve Replacement, CPB: Cardiopulmonary Bypass, PI: Pulse Index Table 3Postoperative outcomes in primary cohortVariablesSkeletonized IMA(n = 1360)Pedicle IMA(n = 993)P valueRe-operation for bleeding18(1.3%)12(1.2%)0.806ICU stay time(d)1(1,2)1(1,2)0.556Stroke18(1.3%)19(1.9%)0.256Perioperative MI4(0.3%)3(0.3%)1.000Cardiac arrest10(0.7%)11(1.1%)0.343New on-set dialysis12(0.9%)13(1.3%)0.319DSWI14(1%)14(1.4%)0.401In-hospital death26(1.9%)12(1.2%)0.181Follow up time358(179,722)361(178,728)0.002All-cause death65(4.8%)57(5.7%)0.299ICU: Intensive Care Unit, DSWI: Deep Sternal Wound Infection Postoperative outcomes in primary cohort ICU: Intensive Care Unit, DSWI: Deep Sternal Wound Infection The survival curve of two matched groups. Group 1 is skeletonized IMA group and Group 2 is pedicled IMA group The matched cohort To minimize the effects of confounders, we used prospensity score matching to keep comparable between the two groups. Finally, 539 pairs of patients in the two groups were matched. The baseline characteristics were balanced well, even though the two variables, numbers of lesion vessel and Euroscore II, were still significant differences(P = 0.028 and P = 0.002, respectively). To operative values, ventricular aneurysm resection, blood flow and PI of conduits still showed the significant differences between two groups(P = 0.03, P = 0.002 and P < 0.001, respectively). The outcomes were similar with those in primary cohort (Tables 4, 5 and 6). Table 4Baseline characteristics in matched cohortVariablesSkeletonized IMA(n = 539)Pedicle IMA(n = 539)P valueSex(male)409(75.9%)402(74.6%)0.621Age64.11 ± 9.3263.95 ± 9.270.776BMI25.0 ± 3.1424.9 ± 3.260.648Smoking252(46.8%)248(46%)0.807Hypertension358(66.4%)357(66.2%)0.949Diabetes211(39.1%)209(38.8%)0.901Hyperlipidemia98(18.2%)104(19.3%)0.640Dialysis---COPD13(2.4%)14(2.6%)0.845PAD20(3.7%)18(3.3%)0.741CVD68(12.6%)67(12.4%)0.927Stroke4(0.7%)5(0.9%)1.000MI113(21%)115(21.3%)0.881ACS171(31.7%)169(31.4%)0.896Cardiogenic shock3(0.6%)5(0.9%)0.726Pre-op intubation4(0.7%)2(0.4%)0.687PCI history57(10.6%)56(10.4%)0.921Numbers of lesion vessel0.028 120(3.7%)40(7.4%) 265(12.1%)65(12.1%) 3454(84.2%)434(80.5%)Left main lesion vessel173(32.1%)169(31.4%)0.793Pre-op Scr74(65,83.5)73.75(65,83.78)0.967Pre-op LVEF62(50,65)63(52.75,65)0.107Euroscore1.35(0.95,2.32)1.59(1.05,2.52)0.002BMI: Body Mass Index, PAD: Peripheral Artery Disease, COPD: Chronic Obstructive Pulmonary Disease, CVD: Cerebral Vessel Disease, MI: Myocardial Infarction, ACS: Acute Coronary Syndrome, PCI: Percutaneous Coronary Intervention, Scr: Serum Creatinine, LVEF: Left Ventricular Eject Fraction Baseline characteristics in matched cohort BMI: Body Mass Index, PAD: Peripheral Artery Disease, COPD: Chronic Obstructive Pulmonary Disease, CVD: Cerebral Vessel Disease, MI: Myocardial Infarction, ACS: Acute Coronary Syndrome, PCI: Percutaneous Coronary Intervention, Scr: Serum Creatinine, LVEF: Left Ventricular Eject Fraction Table 5Intraoperative parameters in matched cohortVariablesSkeletonized IMA(n = 539)Pedicle IMA(n = 539)P valueEmergency surgery16(3%)10(1.9%)0.234Radial artery36(6.7%)43(8%)0.413AVR52(9.6%)42(7.8%)0.280MVR23(4.3%)14(2.6%)0.132Aortic procedure13(2.4%)9(1.7%)0.389Ventricular aneurysm resection4(0.7%)14(2.6%)0.03Operation duration(min)275(240,315)285(240,330)0.486CPB duration(min)101(84,125)108(82,132)0.177Cross-clamp time(min)66(53,85)72(54,92)0.077Blood flow(ml/min)32(22,46)30(18,40)0.002PI1.9(1.5,2.4)2.1(1.7,2.7)< 0.001AVR: Aortic Valve Replacement, MVR: Mitral Valve Replacement, CPB: Cardiopulmonary Bypass, PI: Pulse Index Intraoperative parameters in matched cohort AVR: Aortic Valve Replacement, MVR: Mitral Valve Replacement, CPB: Cardiopulmonary Bypass, PI: Pulse Index Table 6Postoperative outcomes in matched cohortVariablesSkeletonized IMA(n = 539)Pedicle IMA(n = 539)P valueRe-operation for bleeding8(1.5%)9(1.7%)0.807ICU stay time(d)1(1,2)1(1,2)0.528Stroke7(1.3%)8(1.5%)0.795Perioperative MI1(0.2%)1(0.2%)1.000Cardiac arrest5(0.9%)7(1.3%)0.562New on-set dialysis4(0.7%)4(0.7%)1.000DSWI5(0.9%)6(1.1%)0.762In-hospital death8(1.5%)6(1.1%)0.591Follow up time359(181,724)377(180,823)0.020All-cause death22(4.1%)29(5.4%)0.315ICU: Intensive Care Unit, DSWI: Deep Sternal Wound Infection Postoperative outcomes in matched cohort ICU: Intensive Care Unit, DSWI: Deep Sternal Wound Infection Subgroup analysis To explore the outcomes in different population, we also used subgroup analysis. We selected several specific population, such as male and female, age > 65 and < = 65 years old, BMI > 25 and < = 25, smoking and non-smoking, patients with diabetes and without diabetes, patients with hypertension and without hypertension. Finally, the in-hospital and all-cause mortality was not significant difference in skeletonized and pedicled IMA groups among overall group and subgroups (Figures 3 and 4). The subgroup analysis to in-hospital mortality The subgroup analysis to all-cause mortality Discussion This paper included 2353 patients in primary cohort and 539 pairs of patients in PSM cohort, comparing the surgical outcomes between skeletonized and pedicled IMA harvesting technique. No significant difference was shown in in-hospital mortality, early-term mortality, deep sternal wound infection, operation duration. However, the results showed that the blood flow of conduits was significantly different between two groups. The conduits in skeletonized group got higher blood flow rate. Our data showed no significant difference was observed in terms of in-hospital mortality and early-term mortality, which is consistent with the result of a latest meta-analysis. Aimen et al. published a meta-analysis and found that there was no significant difference between the two harvesting techniques [12]. Carter et al. also reported that no difference in mortality between skeletonization and pedicled IMA [13]. These results break the consensus that skeletonization can decrease mortality. Skeletonization is commonly recommended to be used in population with diabetes or obesity, who have higher risk in experiencing sternal wound infection [14]. The latest meta-analysis demonstrated that skeletonized IMA significantly reduced the sternal wound infection compared to the pedicled IMA, which contrasted with our result. The recent researches give the view that skeletonized IMA can preserve the sternal perfusion and reduce the risk of sternal wound infection especially when the bilateral IMA are harvested [15, 16]. A RCT reported that skeletonization preserved the retrosternal microcirculation in the postoperative period to a greater extent [17, 18]. Another meta-analysis focusing on SWI also showed that the rate of sternal wound infection was lower in SIMA than in PIMA [19]. The different result was observed in our study may be that in our center, most of patients received left IMA as conduit to LAD. Therefore, no matter skeletonized or pedicled harvesting technique, the sternum can still acquire good blood supply. This may be the reason why there was no significant difference in terms of deep sternal wound infection between the two groups in our center. Pedicled IMA harvesting technique is a simple and requires a shorter learning curve [20]. Skeletonized harvesting is more meticulous and time-consuming. Some research found that skeletonization technique would increase around 20 min compared with pedicled technique [13]. However, a meta-analysis showed that there was no obvious difference in the duration of operation, which is same with our result [16]. In our center, the IMA harvesting was done by senior surgeon, who are more familiar with the skill and spend less time in harvesting. This may be the reason why the operation duration was no significant difference between two groups. Our result showed that skeletonized IMA conduit can present higher blood flow and lower PI compared with pedicled IMA conduit. The meta-anaylsis also demonstrated higher flow rates could be presented in skeletonized IMA [12]. Deja et al. found that when they changing from pedicled to skeletonized LIMA, the LIMA flow was significant increased. Another paper also discovered that free flow was greater in the skeletonized mammary [21]. The reason may be that skeletonization can increase the conduit diameter and then the blood flow will also increase [22]. A paper showed that the risk of SWI decreased while using longer conduits with higher flow rates [15, 16, 22]. However, our results showed that skeletonized IMA had higher blood flow, but no difference in SWI incidence. Limitations This study is a retrospective and single-center observational research. The survival analysis was limited and longer follow-up data need to be collected. In our database, we did not collect the information about length of conduits, angina again and revascularization again. The most important aspect was the lack of graft patency images. In China, patients are not always willing to receive the CTA check, considering the fee and the harm to body. This makes us no idea about the the impact of blood flow on conduits’ patency. Conclusion In our study, we found that no matter skeletonized or pedicled IMA conduits, there was no significant differences in the aspects of in-hospital mortality, DSWI, surgical duration and early-term survival rate. Skeletonized IMA harvesting may offer physiological advantages in terms of graft flow in experienced hands, but further studies are needed to confirm its safety and clinical benefit. Supplementary Information Supplementary Material 1. Supplementary Material 2. Supplementary Material 3. Supplementary Material 4. Acknowledgements None. Author contributions Rui Fan analysed data and wrote the manuscript; Yifei Zhou and Yifan Zhu collected data; Wuwei Wang gave support in data analysis; Xin Chen, Wei Qin and Fei Xiang reviewed manuscript. Funding There was no funding in this paper. Data availability The datasets generated and analysed during the current study are available from the corresponding author on reasonable request. Declarations Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Rui Fan, Yifei Zhou and Yifan Zhu contributed equally in this paper. Contributor Information Fei Xiang, Email: jackie-xf@163.com. Wei Qin, Email: edwardqinwei@163.com. Xin Chen, Email: stevecx@njmu.edu.cn. References Associated Data Supplementary Materials Supplementary Material 1. Supplementary Material 2. Supplementary Material 3. Supplementary Material 4. Data Availability Statement The datasets generated and analysed during the current study are available from the corresponding author on reasonable request.