Best Evidence Topics
  • Send this BET as an Email
  • Make a Comment on this BET

Is skeletonized internal mammary harvest better than pedicle internal mammary harvest in coronary artery bypass grafting?

Three Part Question

In patients undergoing elective [coronary artery bypass grafting] does [skeletonised IMA] decrease [morbidity]?

Clinical Scenario

You are performing coronary artery bypass grafting on a 49-year-old diabetic with triple vessel disease and normal left ventricular function. He is overweight with a body mass index of 35. You would like to give him the best possible long-term results without causing increase in morbidity. Your colleague suggests the use of bilateral skeletonised internal mammary artery, thus giving long term results due to use of internal mammary artery. You decide to use a skeletonised LIMA and two vein grafts in this high-risk case but resolve to look up the evidence after the case.

Search Strategy

Medline 1966–Aug 2005 using the OVID interface.
[skeletoni$.mp] AND [exp Thoracic Arteries/OR exp Internal Mammary-Coronary Artery Anastomosis/OR exp Mammary Arteries/OR internal OR internal].

Search Outcome

One hundred and six papers were found of which 12 were selected

Relevant Paper(s)

Author, date and country Patient group Study type (level of evidence) Outcomes Key results Study Weaknesses
Athanasiou et al,
Systematic review of paper comparing Pedicle and skeletonised grafts searching Medline from 1966 to 2003Systematic review and meta-analysis (level 1a)Free flow Pedicle vs skeletonized from 4 studies with 494ptsWeighted mean difference is 32ml/min (CI 29-36) in favour of skeletonized IMAOnly Medline searched. Beneficial results were lacking due to the low number of high quality studies identified
Blood supply to sternumGrade A evidence that skeletonized IMA improves sternal blood supply
Mortaliy morbidity or angiographic patencyThere is no evidence to support improved Mortality/Morbidity ,angiographic patency or respiratory complications with skeletonized IMAs
Skeletonization in diabetic patientsSome evidence that BIMA skeletonization improves clinical outcome in diabetics but studies are inconclusive
Calafiore et al,
1146 patients receiving BIMA grafts 304 received pedicle BIMA from 1991-1994 842pts received skeletonized BIMA from 1994 to 1998Retrospective cohort study (level 2b)Sternal wound infectionPedicle 28/304 (2.5%). Skeletonized 14/842 (1.7%) p<0.005

DIABETICS-Pedicle 4/40 (10%). Skeletonized 5/223 (2.2%) p<0.05
Groups were not similar with the skeletonised group being older, having a lower ejection fraction and more redo procedures.
Event free survival (Mean 46months)Pedicle 91.4±0.8%. Skeletonized 95.4±0.7% p<0.001
Angiographic patencyPedicle 94%. Skeletonized 96.8% P<0.001
Number of BIMA anastomosesPedicle 2.1±0.4. Skeletonized 2.4±0.3 p<0.001
LIMA length after papaverine (28 patients assessed)Pedicle 16.4±1.7cms. Skeletonized 20.1±1.6cms p<0.001
Cartier et al,
640 patients undergoing Off pump LIMA harvest by one surgeon 440 had pedicle LIMA then next 200 had skeletonized LIMARetrospective cohort study (level 2b)Deep sternal infectionPedicle 1% Skeletonized 1.2% P=0.3846% had BIMA skeletonized, versus only 27% Pedicle BIMA Non-randomized study
30 day mortalityPedicle 1.6% Skeletonized 1.7% P=NS
Surgical timePedicle 172±46mins. Skeletonized 191±52mins P=0.001
Deja et al,
357 consecutive patients undergoing CABG 287pts had pedicle IMA, 70pts had skeletonized IMAProspective Cohort study (level 2b)Sternal DehiscencePedicle 5/287. Skeletonized 0/70 p=NSNon-randomized study No clinical benefits in outcome demonstrated
LIMA LengthPedicle 17.8±1.1cms. Skeletonized 20.3±0.5cms p=0.11
LIMA free blood flowPedicle 66.3±7.4ml/min. Skeletonized 100.3±14.8ml/min p=0.048
Acetylcholine induced relaxationPedicle 80.7±5.95%. Skeletonized 72.9±9.1% p=NS
Pevni et al,
1000 consecutive patients receiving skeletonized BIMA grafts 1996-1999 304 diabeticsRetrospective cohort study (level 2b)Deep sternal infection22/1000 (2.2%)No control groups
30-day mortality34/1000 (3.4%)
Subsets with increased risk of sternal infectionCOPD 6.2% EF<35% 4.5% Repeat CABG 15% Diabetes 8/306 (2.6%)
Matsa et al,
231 diabetic and 534 non diabetic patients had bilateral skeletonized IMA 1996-98.Retrospective cohortstudy (level 2b)Deep Sternal infectionDiabetics 2.6%. Non-diabetics 1.7%. p=0.40

obese diabetic women 3/20 (15%)
No comparison with pedicled IMAs
Peterson et al,
Retrospective review of 115 diabetics who underwent bilateral IMA CABG from 1990 to 2002 79 pts had skeletonized IMA 36 pedicled IMA 578 pts had bilateral IMAs in non diabeticsRetrospective cohort study (level 2b)Deep sternal wound infectionSkeletonized IMAs 1.3%. Pedicle IMAs 11.1% p=0.03Skeletonized IMA used only from 1999.
Any sternal wound infectionSkeletonized IMAs 5.1%. Pedicle IMAs 22.2% P =0.03
Operative timeSkeletonized IMAs 199.3±75.1 mins. Pedicle IMAs 184.7± 69.7 mins P=0.3
Lorberboym et al,
33 patients divided into 3 groups had pre and post operative 99m Tc-methlene diphosphate SPECT bone scan Gp I : Skeletonized LIMA, n=11 Gp II : Pedicle LIMA, n=12, Gp III : Semi-skeletonized LIMA, n=10RCT (level 2b)Ratio of unilateral sternal uptakeSignificant reduction in vascularity between pedicle group and skeletonized Gps.The length of time between operation and scan was different for each patient. Small study
Takami et al,
65 consecutive patients undergoing CABG 20 had pedicle LIMA 45 had skeletonized LIMA using a harmonic scalpelProspective cohort study (level 2b)Intraoperative flow , using a flwo probe after anastomosisPedicle 26.4±16ml/min. Skeletonized 42.6±29ml/min p=0.03No clinical correlation with outcomes
Diameter of LIMA proximal to anastomosis on angiographyPedicle 1.57±0.17mm. Skeletonized 1.77±0.28mm p=0.02
Wendle et al,
80 patients undergoing elective CABG by 2 surgeons. 40 skeletonized LIMA 40 pedicle LIMA . Papaverine injected into both LIMAsProspective Cohort study (level 2b)Free flow before PapaverinePedicle 68.7±54ml/min. Skeletonized 51.3±39ml/min p=NSNon-randomized study.
Free flow after papaverine injectionPedicle 147±70ml/min. Skeletonized 197±67ml/min p<0.05
Sternal dehiscencePedicle 1 Skeletonized 1
Sofer et al,
545 patients had bilateral skeletonized IMA in a 1 year period from 1996Retrospective cohort study (level 2b)Sternal wound InfectionDeep Sternal infection 1.7 % (9pts). Superficial infection 2.8% (15pts)No pedicle IMA control group or LIMA and SV group
Risk factors for sternal infectionCOPD odds ratio 13 Emergency operation OR 3.8
Bical et al,
712 consecutive patients under 70yrs with bilateral skeletonized IMA, 164 diabeticsRetrospective cohort study (level 2)Deep sternal wound infectionDiabetics 2/164 (1.1%). Non-Diabetics 6/548 (1.2%) p=NSLimitations of a retrospective analysis with no pedicle IMA control group
Superficial wound infectionDiabetics 9/164 (5.5%) Non-Diabetics 16/548 (2.9%) p=NS


Skeletonisation of the internal mammary artery involves mobilization of the arterial trunk from the satellite veins and surrounding tissue, usually using a non-diathermy technique. Advocates of skeletonisation of the IMA cite preservation of sternal blood flow thereby reducing the incidence of complications, longer graft length, larger graft caliber, and greater graft flow. Athanasiou et al. in 2004 published a systematic review and meta-analysis of studies comparing pedicled IMA harvest with skeletonised harvest. They identified a significantly increased level of free flow down skeletonised IMAs and good evidence for improved sternal vascularity, but failed to convincingly demonstrate a benefit in terms of mortality, morbidity, angiographic patency or respiratory complications for skeletonisation. Calafiore et al. in 1999 performed 304 pedicled BIMA anastomoses and then changed to skeletonised BIMA and performed a further 842 operations. They demonstrated that the sternal wound infection rate was lower in the skeletonised group (2.5% vs. 1.7%, P<0.005) and for diabetics pedicled harvest caused a sternal infection rate of 10%. Furthermore, they showed that the angiographic patency was similar at one year, the event-free survival was similar or superior with skeletonised grafts and an extra 4 cm was gained for each graft allowing more BIMA anastomoses. Cartier et al. reported their experience after a single surgeon changed from a pedicled to a skeletonised technique in 640 patients having off-pump CABG. There was no difference in sternal wound complications but more arterial anastomoses were possible with this technique. There was no difference in mortality and the skeletonising technique increased operative time by around 20 min. Deja et al. assessed their experience changing from pedicled to skeletonised LIMA in 357 patients. They demonstrated that an extra 3 cm of LIMA was available, and the LIMA flow was significantly increased. There was no difference in sternal wound complications or clinical short-term outcomes. Pevni et al. reported a sternal infection rate of only 2.2% in a consecutive cohort study of 1000 patients receiving skeletonised BIMA grafts. In particular there was no difference between diabetics and non-diabetics. Matsa et al. compared 231 diabetic with 534 non-diabetics who received bilateral skeletonised IMAs. They found that the sternal wound infection rate was 2.6% in diabetics compared to 1.7% in non-diabetics which was a non-significant difference. In a small retrospective study, Peterson et al. found that in 79 diabetics who received bilateral skeletonised IMAs, only one patient suffered a deep sternal wound infection (1.3%) compared to 11% of 36 diabetics who received pedicled BIMA Lorberboym et al. performed single photon emission computed tomography (SPECT) to determine sternal vascularity post LIMA harvest. They demonstrated a significant difference between skeletonised and pedicled groups in a study of 33 patients. Takami et al. prospectively evaluated patients receiving either a pedicled or skeletonised LIMA, using intraoperative LIMA flow after anastomosis and angiographic LIMA diameter 1 week post surgery. They found that the flow almost doubled in the LIMA graft and angiographically the LIMA diameter was slightly greater 1 week post-operatively. Wendler et al. studied the difference in flow between 40 pedicled LIMAs and 40 skeletonied LIMAs. While no difference was initially seen after harvest, skeletonised LIMAs demonstrated 25% more flow after papaverine injection. Sofer et al. published their findings after performing bilateral skeletonised IMAs in 545 patients. They found a sternal wound infection rate of only 1.7% with COPD and emergency operation but not diabetes increasing this risk. Bical et al. reported their findings on 712 patients under 70 years old receiving skeletonised BIMA grafts, comparing their diabetic and non-diabetic patients. The incidence of sternal wound infection was similar, with diabetics suffering a 1.1% deep infection rate compared to a 1.2% rate in non-diabetics. There is thus good evidence that flow and length is increased using a skeletonised technique. In addition, skeletonization is far superior to pedicled harvest for BIMA grafts in diabetics and reduces the sternal infection rate from 10% to around 2% in these patients. However, no significant differences in terms of sternal complications or vascular patency have been shown for LIMA harvest and consistently adds 15–20 min to the length of the operation.

Clinical Bottom Line

Skeletonisation increases the length of conduit by around 3 cm and may also increase flow and conduit diameter. Skeletonisation should be the technique of choice for diabetics in whom BIMA harvest is desired, but at the expense of an extra 15–20 min per operation, no convincing outcome benefits have been shown for single IMA harvest.


  1. Athanasiou T, Crossman MC, Asimakopoulos G et al. Should the internal thoracic artery be skeletonised? Ann Thorac Surg 2004;77:2238–2246.
  2. Calafiore AM, Vitolla G, Iaco AL et al. Bilateral internal mammary artery grafting: midterm results of pedicled vs. skeletonised conduits. Ann Thorac Surg 1999;67:1637–1642.
  3. Cartier R, Leacche M, Couture P. Changing pattern in beating heart operations: use of skeletonised internal thoracic artery. Ann Thorac Surg 2002;74:1548–1552.
  4. Deja MA, Wos S, Golba KS, Zurek P, Domaradzki W, Bachowski R, Spyt TJ. Intraoperative and laboratory evaluation of skeletonised vs. pedicled internal thoracic artery. Ann Thorac Surg 1999;68:2164–2168.
  5. Pevni D, Mohr R, Lev-Run O, Locer C, Paz Y, Kramer A, Shapira I. Influence of bilateral skeletonised harvesting on occurrence of deep sternal wound infection in 1,000 consecutive patients undergoing bilateral internal thoracic artery grafting. Ann Surg 2003; 237:277–280.
  6. Matsa M, Paz Y, Gurevitch J, Shapira I, Kramer A, Pevny D, Mohr R. Bilateral skeletonised internal thoracic artery grafts in patients with diabetes mellitus. J Thorac Cardiovasc Surg 2001;121:668–674.
  7. Peterson MD, Borger MA, Rao V, Peniston CM, Feindel CM. Skeletonisation of bilateral internal thoracic artery grafts lowers the risk of sternal infection in patients with diabetes. J Thorac Cardiovasc Surg 2003;126:1314–1319.
  8. Lorberboym M, Medalion B, Bder O, Lockman J, Cohen N, Schachner A, Cohen AJ. 99 mTc-MDP bone SPECT for the evaluation of sternal ischaemia following internal mammary artery dissection. Nuclear Med Com 2002;23:47–52.
  9. Takami Y, Ina H. Effects of skeletonization on intraoperative flow and anastomosis diameter of internal thoracic arteries in coronary artery bypass grafting. Ann Thorac Surg 2002;73:1441–1445.
  10. Wendler O, Tscholl D, Huang Q, Schafers HJ. Free flow capacity of skeletonised vs. pedicled internal thoracic artery grafts in coronary artery bypass grafts. Eur J Cardiothorac Surg 1999;15:247–250.
  11. Sofer D, Gurevitch J, Shapira I, Paz Y, Matsa M, Kramer A, Mohr R. Sternal wound infections in patients after coronary artery bypass grafting using bilateral skeletonised internal mammary arteries. Ann Surg 1999;229:585–590.
  12. Bical OM, Khoury W, Fromes Y, Fischer M, Sousa UM, Boccara G, Deleuze PH. Routine use of bilateral skeletonised internal thoracic artery grafts in middle-aged diabetic patients. Ann Thorac Surg 2004;78:2050–2053.