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

What is the optimal vasodilator for preventing spasm in the left internal mammary artery during coronary arterial bypass grafting?

Three Part Question

In [patients undergoing CABG using the left internal mammary artery] do [vasodilators] improve [graft flow]

Clinical Scenario

You have just started working with a consultant in a new firm. The consultant you had previously worked with uses topical papaverine to prevent vasospasm of the left internal mammary artery. Your new consultant never does this. On the first theatre day in your new firm you have completed harvesting the internal mammary and noted the flow to be poor. You are contemplating using topical papavarine to improve the vasospasm of the mammary artery. However, your consultant stops you and asks you to show him the evidence that topical vasodilators significantly improve mammary arterial flow before using any vasodilators in his cases.

Search Strategy

Medline 1966–March 2005 using the Ovid interface
[exp Mammary Arteries/OR OR Mammary art$.mp OR thoracic art$.mp.] AND [ or OR OR OR] AND [ or sodium OR or OR exp Nitroglycerin/OR OR exp Papaverine/OR OR exp Phosphoric Diester Hydrolases/OR OR exp Vasodilator Agents/].

Search Outcome

A total of 200 papers were found from the above search. Case reports and in vitro studies were excluded. Studies investigating systemic vasodilators were also excluded. Thirteen studies represented the best evidence to answer our question. These papers are listed in the table

Relevant Paper(s)

Author, date and country Patient group Study type (level of evidence) Outcomes Key results Study Weaknesses
Mills and Bringaze,
31 patients had internal mammary harvested for elective CABG Group I (n=14) Topical papaverine in soaked sponge (60 mg in 40 ml n/sal) for 20 mins and then intraluminal papaverine (10 ml) instillation and hydrostatic dilatation with sequential finger occlusion up to its origin Group II (n=17) Pedicle injection of papaverine and 20 mins delay then intraluminal instillation of papaverine and hydrostatic instillation (as above)Cohort study with no control group (level 3b)Group I Blood flow in ml per min. (Mean and range)No treatment: 18 ml/min (5-44 ml/min). Topical papaverine: 51 ml/min (10-108 ml/min). Intraluminal papaverine in addition to topical papaverine:229 ml/min (150-333 ml/min)Non-randomised study Small sample size No control group with no treatment to compare with the papaverine groups
Group II Blood flow in ml per min (mean and range)After pedicle injection 69 ml/min (28-132ml/min). Intraluminal papaverine in addition to pedicle injection 198 ml/min (144-280 ml/min). With pedicle injection to 188 ml/min with intraluminal injection
Sasson et al,
50 patients had internal mammary harvested for elective CABG Group I (n=10) Topical saline Group II (n=10) Topical papaverine (5 mg/10 ml Nsal) Group III (n=10) Topical nitroglycerine (5 mg/10 ml Nsal) Group IV (n=10) Periarterial injection of papaverine (5 mg/5 ml Nsal)Non-randomised cohort study (level 3b)Pre and post treatment flow (all pre CPB)Group I (Saline) pre 59 ± 20.7 ml/min. Post 72.8 ± 17 ml/min. Group II (papaverine) pre 86 ± 25 ml/min post 118 ± 53 ml/min. Group III (GTN) pre 71 ± 24 ml/min post 90 ± 30 ml/min. Group IV (SNP) pre 43 ± 27 ml/min post 67 ± 27 ml/min. Group V (papaverine) pre 57 ± 33 ml/min post 83 ± 49 ml/min. No significant flow increase seen compared to controlNon-randomised study Small sample size in each group Discrepancy in age, mean arterial pressures and time of measurement between the groups
Complication8 patients were withdrawn in the sodium nitroprusside group due to clinical hypotension
Takeuchi et al,
80 patients had mammary artery harvested for elective CABG. 2ml of test drug were directly injected to the proximal segment of the LIMA at 1 ml/s, distal end was clipped Group I (n=20) Intra arterial injection of 2 ml saline Group II (n=20) Intra arterial injection of 2 ml papaverine Group III (n=20) Intra arterial injection of 2 ml Isosorbide Dinitrate Group IV (n=20) Intra arterial injection of 2 ml Phosphodiesterase III inhibitorSingle blind PRCT (level 1b)Graft free flow pre and 1 min post injectionGroup I (saline) pre 37 ± 14 ml/min post 36 ± 15 ml/min. Group II (papaverine) pre 37 ± 17 ml/min post 40 ± 19 ml/min. Group III (ISDN) pre 37 ± 16 ml/min post 48 ± 20 ml/min. Group IV (PDE-III-I) pre 36 ± 18 ml/min post 57 ± 18 ml/min. Graft free flow increased significantly post injection in Group III (P<0.05) and Group IV (P<0.0001)Short time duration of measurement of graft flow before and after drug injection (papaverine may take 10 min for maximal action)
Girard et al,
58 patients undergoing primary, redo CABG and CABG + valves Group I (n=20) Control group Group II (n=18) Periarterial papaverine injection with a blunt needle before dissection (10 mg in 10 ml Nsal) Group III (n=20) Periarterial papaverine injection after dissectionPRCT (level 2b)Graft free flow just before performing the anastomosis with bypass pressures at 70 mmHgGroup I (control) 86.2 ml/min. Group II (papaverine) 86.2 ml/min I vs II p=0.0874. Group III (post papaverine) 139.7 ml/min I vs III p=0.0457Small sample size No uniformity in the type of surgery in different groups Harvesting of artery by 4 different surgeons Confidence intervals for results not given
Nili et al,
80 patients having first time CABG LIMA immersed in a tube containing 20 ml solution Group I (n=16) Normal saline Group II (n=16) Papaverine 2mg/ml Group III (n=16) Verapamil 0.5mg/ml Group IV (n=16) Nitroglycerine 1 mg/ml Group V (n=16) Nitroprusside 0.5 mg/mlDouble blind PRCT (level 1b)Graft free flow using electromagnetometer immediately after division and just prior to anastomosis (on CPB)Group I (saline) pre 39 ± 10 ml/min, post 85 ± 16 ml/min. Group II (papaverine) pre 42 ± 8 ml/min, post 82 ± 6 ml/min. Group III (verapamil) pre 41 ± 8 ml/min, post 78 ± 9 ml/min. Group IV (nitroglycerine) pre 39 ± 11 ml/min, post 84 ± 24 ml/min. Group V (SNP) pre 41 ± 5 ml/min, post 99 ± 12 ml/min. No significant improvement in flow compared to placebo.Well conducted study Mammary artery harvested by one surgeon Mammary artery soaked in applicators for uniform distribution
Yavuz et al,
150 patients undergoing CABG ± valves ± Redo surgery. 3 methods of giving 60 mg of papverine in 40 ml Nsal Group I (n=50) Intraluminal papaverine applied retrogradely to unclamped LIMA Group II (n=50) Topical papaverine Group III (n=50) Periarterial papaverinePRCT (level 2b)Graft free flow, pre injection and prior to CPBGroup I (intraluminal) pre 63 ± 6 ml/min, post 129 ± 10 ml/min. Group II (topical) pre 60 ± 6 ml/min, post 88 ± 4 ml/min. Group III (peri-arterial) pre 60 ± 6 ml/min, post 131 ± 9 ml/min. Groups I and III significantly better than group II.No control to compare with No uniformity of surgery in the different groups 6 patients (4%) did not have mammary artery used due to poor flow in spite of papaverine flow
ComplicationsIntra-luminal injection caused intimal dissection in 3 patients
Cooper et al,
50 patients mammary artery harvested for CABG Group I (n=10) Topical 0.9% saline Group II (n=10) Topical papaverine 6mg in 4 ml Nsal Group III (n=10) Topical Nifedipine 400 µg in 4 ml Group IV (n=10) Topical Nitroglycerine 2 mg in 4 ml Nsal Group V (n=10) Topical sodium Nitroprusside 2 mg in 4 ml 5% DextUnblinded PRCT (level 2b)Graft free flow on dissection and 19 min after topical application of study drug (mean and range)Group I (saline) pre 23(17-88) ml/min, post 38(20-84) ml/min. Group II (papaverine) pre 25(16-78) ml/min, post 43(43-112) ml/min, I vs II p<0.01. Group III (Nifedipine) pre 23(14-66) ml/min, post 71(45-118) ml/min, I vs III p<0.001. Group IV (Nitroglycerine) pre 23(14-28) ml/min, post 62(46-126) ml/min I vs IV p<0.001. Group V (SNP) pre 26(10-58) ml/min, post 108(46-196) ml/min, I vs V p<0.001.Small sample size Superiority of SNP over the other vasodilators claimed by this paper is not supported by significant inter-group findings between the vasodilator drugs
Yorgancioglu et al,
86 patients undergoing CABG Group I (n=42) Topical nitroprusside 3mg in 10 ml 5% Dext Group II (n=44) Topical application of 5 ml SNP then 5 ml SNP peri-arterial injectionPRCT (level 2b)Graft free flow after transection, and before anastomosisGroup I (topical SNP) pre 22 ± 18 ml/min, post 70 ± 36 ml/min. Group II (periarterial SNP) pre 20 ± 17 ml/min, post 107 ± 60 ml/min, p<0.01 between two groups at end of studyNo control group to compare findings with
Vilandt et al,
75 patients undergoing elective CABG Group I (n=26) Intraluminal papaverine (60mg in 2 ml Nsal) at 18 degrees injected into LIMA lumen Group II (n=26) Intraluminal saline Group III (n=25) No injectionPRCT (level 2b)Graft free flow after dissection and before anastomosisGroup I (papaverine) pre 40 ± 12 ml/min, post 154 ± 364 ml/min. Group II (saline injection) pre 42 ± 18 ml/min, post 84 ± 26 ml/min. Group III (control) pre 36 ± 10 ml/min, post 66 ± 17 ml/min. Post papaverine flow was twice as high compared to controls.Some patients in control group had high flows and papaverine group had low flows All patients were on nitroglycerine infusion during procedure Papaverine solution was found to have a PH of 3 in 2 samples
ComplicationsNo incidence of hypotension
Dregelid et al,
51 patients had mammary artery harvested for elective CABG Group I (n=17) Covered with sponge soaked in papaverine (0.8 mg/ml) with intact distal ends until anastomosis Group II (n=17) Covered with sponge soaked in papaverine and distal end disconnected Group III (n=17) Pedicle soaked in glove with papaverinePRCT (level 2b)Graft free flow prior to anastomosisGroup I (no disconnection) 44ml/min. Group II (disconnected) 30ml/min. Group III (glove) 60ml/minPatients on systemic vasodilators during procedure No control to compare papaverine injection to
Morphometric measurement such luminal area and fold index of the internal elastic laminaMedian luminal area and fold index larger in group III than I and II
Dregelid et al,
78 patients undergoing elective CABG Group I (n=26) Pedicle placed in bag containing papaverine (1.5 mg/ml) Group II (n=26) Intraluminal injection of 2 ml heparinised blood with 0.2 ml of 40 mg/ml papaverine then placed in bag containing papaverine Group III (n=26) Intraluminal injection of heparinised blood + papaverine after harvesting and again just before going on bypass and placed in bag containing papaverinePRCT (level 2b)Graft free flow. Mean.Group I (bag) 58 ml/min. Group II (injection and bag) 82 ml/min. Group III (2 injections and bag) 68 ml/minMechanical injury to lumen by intraluminal injection Patients were on systemic nitroglycerine during the procedure Small sample size No comparison with control group who did not have papaverine
Morphometric measurementLess folding of internal elastic lamina and larger luminal area in Group II and III (p<0.02)
ComplicationsMicroscopic mechanical injury in 8(15%) patients with intraluminal injection. Five cases of dissection, 1 disruptions of media, and 2 invaginated medias into lumen.
Hausmann et al,
106 patients for primary, redo and CABG + valve Group I (n=38) Intraluminal injection of papaverine (50mg) Group II (n=46) Topical application of papaverine (50mg) Group III (n=22) Periarterial injection of papaverine (50mg)PRCT (level 2b)Graft free flow before harvesting and 12-17 min after papaverine application. Mean.Group I (Intraluminal) pre 66.1 ml/min, post 105 ml/min. Group II (topical ) pre 53 ml/min, post 85 ml/min. Group III (periarterial) pre 64 ml/min, post 136 ml/min. Groups I and III significantly better flow than group IIInappropriate sample selection Analysis of data among the 3 groups were done by student's t-test and not one way analysis of variance Times to second measurement varied between groups from 12 min in Gp III to 17 min in Gp II
Bilgen et al,
60 patients undergoing CABG Group I (n=20) Topical normal saline at 20 °C (6 mg in 4ml Nsal) Group III (n=20) Topical papaverine at 37 °C (6 mg in 4 ml Nsal)PRCT (level 2b)Graft free flow immediately after harvesting and median of 16 min after application of vasodilatorsGroup I (Nsal) pre 38 ± 8 ml/min, post 79 ± 21 ml/min. Group III (37 °C papaverine) pre 37 ± 13 ml/min, post 103 ± 45 ml/min. Grp III superior to Grp II with p=0.0174Small sample size Median values calculated by student t-test and Mann-Whitney U test


Eleven Randomized Controlled Trials and 2 multi-arm prospective cohort studies were found investigating the effects of topical, intraluminal and periarterial vasodilators. These studies compared the mode of administration the concentration of the drug administered and the temperature in which they have been administered for papaverine, sodium nitroprusside, nitroglycerine and phosphodiesterase inhibitors. Among the studies papaverine has been used in all studies except one [Yorgancioglu]. Papaverine has been shown to increase blood flow compared to control in some studies [Mills, Girard, Vilandt, Dregelid 1995] but not others [Sasson, Takeuchi, Nili, Cooper]. Flow prior to bypass in control groups varies from 36 ml per min to 85 ml per min among all studies reporting a control group. Positive papaverine studies demonstrate a mean flow from 100 to 229 ml/min with the highest pre-anastomosis flow rate being from intraluminal application followed by hydrostatic pressure dilatation [Mills]. Perivascular and intraluminal instillation of papaverine significantly increased blood flow compared to topical papaverine [Yavuz, Vilandt, Hausmann]. However, the microscopic analysis by Dregelid [1995] showed that intraluminal instillation caused mechanical injury to the lumen of the mammary artery, with 5 dissections, 1 medial disruption and 2 medial invaginations into the lumen in their study. In addition Yavuz identified 6 patients (4%) who were noted to have poor flow in the mammary artery after intraluminal injection and 3 were found to have suffered a dissection. Sodium nitroprusside has also been widely investigated [Sasson, Nili, Cooper, Yorgancioglu]. Two studies failed to show a significant improvement compared to control [Sasson, Nili]. Cooper et al. was the only study to show a significant improvement to controls and also the only study to demonstrate a benefit in comparison to other topical vasodilators, although the finding did not reach significance. Sasson et al found that topical application brought about systemic hypotension in 8 of the 10 patients receiving topical SNP requiring withdrawal of these patients from the study. Yorgancioglu et al found that periarterial injection of sodium nitroprusside brought about a greater increase in mammary flow in comparison to topical spraying although they had no control group. Topical nitroglycerine was not shown to significantly increase blood flow in 3 studies [Sasson, Takeuchi, Nili], but was shown to increase blood flow compared to controls in one study [Cooper]. No studies have shown nitroglycerine to be superior to any other vasodilators. Takeuchi et al is the only study to show that topical phosphodiesterase III inhibitors also increase mammary artery flow although the improvements compared to control were small. In summary there is surprisingly little strong evidence that vasodilators significantly improve LIMA graft flow compared to no treatment. All studies that use a control show that the flow can often initially be low but the flow invariably doubles after 15–20 min. Only one study has demonstrated a significant benefit using SNP or GTN. The strongest evidence for benefit is for Papaverine with 4 studies showing a significant benefit. This benefit is greatest if periarterial or intraluminal injection is performed although there have been several reports of damage to the mammary artery with intraluminal injection.

Clinical Bottom Line

Mammary arteries often have low flow initially, but invariably will double their flow after 15–20 min even with no treatment. The strongest evidence for safe prevention of spasm is for papaverine given topically and periarterially, however, many studies have also shown no benefit and thus no treatment at all is an entirely acceptable strategy.


  1. Mills NL, Bringaze WL 3rd. Preparation of the internal mammary artery graft. Which is the best method? J Thorac Cardiovasc Surg 1989;98:73–79.
  2. Sasson L, Cohen A, Hauptmann E, Schachner A. Effect of topical vasodilators on Internal mammary arteries. Ann Thorac Surg 1995;59:494–496.
  3. Takeuchi K, Sakamoto S, Nagayoshi Y, Nishizawa H, Matsubara J. Reactivity of the human internal thoracic artery to vasodilators in coronary artery bypass grafting. Eur J Cardiothorac Surg 2004;26:956–959.
  4. Girard DS, Sutton JP III, Williams TH, Crumbley AJ III, Zellner JL, Kratz JM, Crawford FA. Papaverine delivery to the internal mammary artery pedicle effectively treats spasm. Ann Thorac Surg 2004;78:1295–1298.
  5. Nili M, Stamler A, Sulkes J, Vidne B. Preparation of the internal thoracic artery by vasodilator drugs: is it really necessary? A randomized double-blind placebo-controlled clinical study. Eur J Cardiothoracic Surg 1999;16:560–563.
  6. Yavuz S, Celkan A, Goncu T, Turk T, Ozdemir A. Effect of papaverine applications on blood flow of the internal mammary artery. Ann Thorac Cardiovasc Surg 2001;7:84–88.
  7. Cooper GJ, Wilkinson GA, Angelini GD. Overcoming perioperative spasm of the internal mammary artery: Which is the best vasodilator? J Thorac Cardiovasc Surg 1992;104:465–468.
  8. Yorgancioglu C, Tokmakoglu H, Gunaydin S, Catav Z, Suzer K. An alternative application of sodium nitroprusside to overcome perioperative spasm of the internal thoracic artery. Cardiovasc Surg 2001;9:64–67.
  9. Vilandt J, Kjaergard H, Aggestrup S, Andreasen J, Olesen A. Intraluminal papaverine with pH 3 doubles blood flow in the internal mammary artery. Scand Cardiovasc J 1999;33:330–332.
  10. Dregelid E, Heldal K, Andersen K, Stangeland L, Svendsen E. Dilation of the internal mammary artery by external papaverine application to the pedicle: an improved method. Eur J Cardiothorac Surg 1993;7:158–163.
  11. Dregelid E, Heldal K, Resch F, Stangeland L, Breivik K, Svendsen E. Dilation of the internal mammary artery by external and intraluminal papaverine application. Journal of Thoracic & Cardiovascular Surgery. 1995;110:697–703.
  12. Hausmann H, Photiadis J, Hetzer R. Blood flow in the internal mammary artery after the administraion of papaverine during coronary artery bypass grafting. Texas Heart Institute J 1996;23:279–283.
  13. Bilgen F, Yapici F, Serbetcioglu A, Tarhan A, Coruh T, Ozler A. Effect of Normothermic papaverine to relieve intraoperative spasm of the internal thoracic artery. Ann Thorac Surg 1996;62:769–771.