Effect of topical vasodilators on gastroepiploic artery graft

Background . Mobilization of the gastroepiploic artery (GEA) often results in a vasospasm with reduction of early graft flow. In order to prevent or suppress this highly reactive artery's spasm, we have compared the effect of 4 vasodilators, used in external application to prepare the GEA graft, prior to myocardial revascularization.

Methods . We performed a double-blind clinical study to compare the effects of external application of vasodilators on gastroepiploic artery grafts. Fifty patients, whose gastroepiploic artery was used for coronary artery bypass grafting, were randomized into 5 groups of 10 patients. Gastroepiploic artery free flow and hemodynamic measurements were evaluated immediately after harvesting, before any pharmacological manipulation, and 10 minutes after the topical application of vasodilators, respectively: papaverine, linsidomine, nicardipine, glyceryl trinitrate, and normal saline solution.

Results . A significant increase in free flow occurred in all groups except for the normal saline solution group with measurements from 26.1 ± 3.6 mL/min to 26.4 ± 6.5 mL/min; p = 0.9. The most important increase in flow before and after local application occurred with glyceryl trinitrate and papaverine: from 25.5 ± 2 mL/min to 50 ± 6.1 mL/min ( p 0.01) and from 36.8 ± 3.2 mL/min to 62 ± 7.8 mL/min ( p < 0.01) respectively. Nicardipine and linsidomine produced a less significant increase in flow: from 33.1 ± 3.6 mL/min to 47.7 ± 8.9 mL/min ( p < 0.05) and from 28 ± 3.8 mL/min to 39.8 ± 7.5 mL/min ( p < 0.05) respectively. When comparing percentage of flow increase, glyceryl trinitrate appeared to be significantly more efficient than nicardipine and linsidomine ( p < 0.01 versus both groups). Although papaverine was more efficient than nicardipine and linsidomine, it did not reach statistical significance.

Conclusions . During intraoperative preparation of the GEA graft, glyceryl trinitrate and papaverine to a lesser extent, used as topical vasodilators, appear to be more efficient in external application to increase the free flow of the GEA.

 The superiority of arterial grafts for myocardial revascularization, based on their long-term patency rates and patient survival, compared to venous conduits, has led to a dramatically expanded use of the internal thoracic artery (ITA) [1] [2] . After the first reports by Pym and Suma in 1987 [3] [4] , gastroepiploic artery (GEA) grafting became popular. It has been used as a suitable arterial conduit for coronary artery bypass grafting because of its similarity to the IMA in terms of diameter, flow, and atherosclerosis, although intimal thickening is slightly greater in the GEA [5] [6] . Furthermore, its ability to reach the coronary arteries on the inferior ventricular wall allows total arterial revascularization for three-vessel coronary artery disease [7] [8] . Mobilization and operative preparation of arterial conduits often results in vasospasms in spite of careful dissection and minimal handling, given that this handling is more difficult to minimize for the GEA pedicle than during ITA harvesting. Vasospasm results in inadequate flow which may be detrimental for both the perioperative period and the long term patency rate. Furthermore, an adequately dilated artery graft allows to evaluate the function of the graft and facilitates placement of anastomotic sutures. If the local preparation of the ITA has been well described [9] [10] [11] , no clinical comparison has been made on vasodilators used for the GEA, which is known to be very prone to vasospasm because of its muscular status [12] [13] . In this study, we compared the effect of the local external applications of papaverine, linsidomine, nicardipine, glyceryl trinitrate, and normal saline solution. The aim was to find which vasodilator best promoted dilation of the GEA.

We studied 50 consecutive patients undergoing elective coronary artery bypass operation from October 1996 to December 1997. The protocol was approved in January 1996 by the local institutional human research committee, at the Michallon Hospital in Grenoble. Criteria for admission into the study group were elective coronary artery bypass operation using the GEA. All patients received nitrates, calcium antagonist, and beta-blockers until the date of operation. All patients were anesthetized with 50% oxygen, sufentanil, propofol, and atracurium. They all had a Swan-Ganz catheter (Baxter, Maurepas, France) inserted into the right internal jugular vein. Each patient was randomly allocated to 1 of the 5 groups. Patients who were in unstable hemodynamic condition or showed evidence of ischemia during the operation were excluded from the study. A pilot study was performed in order to test the reproducibility of GEA flow measurements. GEA flow was measured using the same methodology as in the final study, at a 10 minute interval, in 10 patients. The mean free flow was 30.1 ± 3.2 mL/min and 30.5 ± 3.1 mL/min for the 1st and 2nd measurements, respectively. The intraindividual coefficient of variation was 13% or 4.2 mL/min, demonstrating the good reproducibility of the method. The GEA was dissected on a pedicle from the pylorus to the middle of the greater curvature of the stomach with the aid of metal ligature clips. After systemic heparinization and aortic and right atrial cannulation, the artery was divided distally after a 5 minute stabilization period, to restore adequate hemodynamic status. The first free flow was determined by measuring the volume of blood expelled from the end of the freely bleeding artery in a 30 second period, using a protocol previously used in a ITA flow study [14] . Heart rate, systolic, mean and diastolic systemic arterial pressures, and central venous pressure were also measured. The GEA pedicle base was then occluded with a metallic bulldog clamp, the artery was sprinkled with the vasodilator with a syringe and needle, wrapped in a moist swab with topical solutions, and finally laid in the peritoneal cavity.

Topical solutions: All topical solutions were at room temperature (18°C) when applied:

1.Group 1: papaverine 80 mg in 10 mL of 0.9% sodium chloride solution (Aguettant, Lyon, France); 2.Group 2: nicardipine 10 mg in 10 mL of 0.9% sodium chloride solution (Novartis, Rueil-Malmaison, France); 3.Group 3: glyceryl trinitrate 15 mg in 10 mL of 0.9% sodium chloride solution (Besins-Iscovesco, Paris, France); 4.Group 4: linsidomine, 10 mg in 10 mL of 0.9% sodium chloride solution (Hoechst, Paris, France); 5.Group 5: normal saline solution, 10 mL of 0.9% sodium chloride solution.

Ten minutes after the first measurement, the GEA was unwrapped and the bulldog clamp was taken off. The second GEA flow measurement was obtained and hemodynamic data were recorded again. To ensure that the experiment was double-blind, the surgeon measuring the blood flow was never aware of the drug administered. After the second measurement, the blind was removed for all patients. For the normal saline solution group, a complementary application of papaverine was made.

Each patient served as his or her own control. All values were expressed as mean ± standard error. The significance of the changes in GEA flow and hemodynamic data was tested by two-tailed Wilcoxon's rank sum test. The difference between each group was tested by variance analysis, with the Bonferroni test as post hoc test. Probability values inferior to 0.05 were considered as statistically significant.

There was no statistically significant difference among each of the 5 groups concerning sex ratio and age ( [Table 1] ).

Hemodynamic data (heart rate, central venous pressure, mean arterial pressure), at the time of the first and second flow measurements, was not significantly different either ( [Table 2] ).

Mean flow for the first and second measurement, for each group, is shown in . [Table 3] There was no significant difference among the first flow measurements in the 5 groups. Normal saline solution produced no increase in flow before and after local application, whereas the most effective increase in flow occurred with papaverine and glyceryl trinitrate ( p p p

In the present study, glyceryl trinitrate and papaverine to a lesser extent, appear to be more efficient than nicardipine and linsidomine as topical vasodilators on GEA flow. Indeed, GEA is a highly vasoactive artery, which is prone to spasm because of its muscular histological status [12] [13] and the method of preparation of the GEA graft appears to be the major determinant of GEA total flow capacity [15] . The decrease in diameter results in lesser arterial flow, with possible clinical consequences such as functional insufficiency. In vitro comparative study between ITA and GEA showed stronger contraction for the GEA to vasoconstrictors [13] [16] . If clinical evaluation of vasodilators used for the preparation of the ITA is now well documented, with controversial results as to which topical vasodilator better relieves intraoperative spasm of the ITA graft [10] [11] , to our knowledge, there are no formal studies testing other topical vasodilators such as papaverine, available for GEA graft preparation.

Free flow of the GEA depends both on the perfusion pressure and the vasomotor properties of the graft. This is the reason why measuring it is a good method to compare the effect of topical vasodilators. In this study, the time (10 minutes) between the 2 measurements may have underestimated their maximal effect. But, a quick response is essential with regard to operative time and for ethical reasons. We chose to perform an external administration of these treatments since intraluminal administration of papaverine has been shown to induce detrimental effects on the ITA wall [17] , or direct mechanical trauma [18] . Furthermore, chemical damage due to the acidity of some solutions (papaverine, nicardipine) could have been detrimental to the endothelium. Proximal placement of the clamp on the pedicle was performed because we only wanted to assess the drug's chemical effect. Indeed, distal placement of the bulldog could have been more efficient in promoting maximal dilation of the conduct. This would be due to the vasodilator and to a mechanical effect induced by exposure to arterial pressure.

The results of the present study confirm that the GEA is in spasm, with reduced flow, immediately after mobilization as shown in previous studies [15] . The fact that normal saline solution did not increase the flow shows that the vessel does not undergo substantial spontaneous relaxation in the time between end of mobilization and start of cardiopulmonary bypass, as demonstrated by our pilot study. The 5 groups are homogeneous, and, according to clinical data, the hemodynamic status is comparable between the 2 measurements. This suggests that for the vasodilator concentration used, there was no significant absorption through the surrounding soft tissues resulting in a systemic blood pressure drop. We performed aortic and right atrial cannulation before the first measurement because these maneuvers may be associated with important blood volume loss, leading to variation in GEA flow as previously noted in ITA flow measurement [19] .

Nicardipine and linsidomine produced a similar increase in GEA flow, but were both less potent than papaverine and glyceryl trinitrate. Papaverine, a lipophilic vasodilator used to prepare the ITA, is the classical agent to protect against GEA spasm, although no clinical study has compared other topical vasodilators. All human studies concerning the effect of vasodilators in the GEA where carried out in vitro, mainly in GEA rings precontracted with various vasoconstrictors in organ-bath chambers. Uydes-Dogan showed that nitrovasodilators (sodium nitroprusside and glyceryl trinitrate), as well as papaverine and nifedipine, reversed the spasm induced by norepinephrine or endothelin-1 [20] . Ali demonstrated on porcine GEA segments that, when given externally on the perivascular fat of the GEA, papaverine prevented graft spasm for up to 2 hours, but in contrast when given intraluminally, papaverine did not show graft protection against norepinephrine-induced spasm [21] . This study also evaluated three calcium channel blockers (diltiazem, nifedipine, verapamil) with verapamil as the most potent vasodilator whether applied externally or internally. Another agent was tested, levcromakalim a potassium channel opener. It proved to be a highly relaxant agent on GEA rings precontracted with K+ or endothelin-1 [22] . In our study, glyceryl trinitrate, a NO-donor, was found to be at least as effective as papaverine, a nonselective phosphodiesterase inhibitor, in increasing GEA graft flow. This interesting result was demonstrated by Cooper for the ITA. Conversely, both drugs appeared to be more effective than nicardipine, a calcium channel blocker, and linsidomine, a direct NO-donor. Recently, Bilgen showed that papaverine at 37°C proved to be more effective than papaverine at room temperature, probably because enzyme activity is progressively reduced at lower temperature [23] . In our study, all topical vasodilators were used in routine condition, ie, with room temperature around 18°C. Vasodilators at 37°C would probably be more effective.

In conclusion, our study shows that in order to prevent or reverse the intraoperative spasm of the GEA graft in myocardial revascularization, our preference goes to glyceryl trinitrate or papaverine topical vasodilators. Furthermore, the possibility of administering a low dose of glyceryl trinitrate intravenously should be taken in consideration, and further clinical studies are required to determine the effect of systemic glyceryl trinitrate infusion on GEA flow, as previously demonstrated for ITA graft [24] .

 References

1.Dion R., Etienne P.Y., Verhelst R.. Bilateral mammary grafting. Clinical functional and angiographic assessment in 400 consecutive patients. Eur J Cardiothorac Surg 1993;7:287−294.

2.Cameron A., Davis K.B., Green G., Schaff H.V.. Coronary bypass surgery with internal-thoracic-artery-grafts. Effects on survival over a 15−year period. N Engl J Med 1996;334:216−219.

3.Pym J., Brown P.M., Charrette E.J.P., Parker J.O., West R.O.. Gastroepiploic-coronary anastomosis. A viable alternative bypass graft. J Thorac Cardiovasc Surg 1987;94:256−259.

4.Suma H., Fukumoto H., Takeuchi A.. Coronary artery bypass grafting by utilizing in situ right gastroepiploic artery. Ann Thorac Surg 1987;44:394−397.

5.Suma H., Wanibuchi Y., Futura S., Isshiki T., Yamaguchi T., Takanashi R.. Comparative study between the gastroepiploic and the internal thoracic artery as a coronary bypass graft. Size, flow, patency, histology. Eur J Cardiothorac Surg 1991;5:244−247.

6.Suma H., Takanashi R.. Arteriosclerosis of the gastroepiploic and internal thoracic arteries. Ann Thorac Surg 1990;50:413−416.

7.Grandjean J.G., Boonstra P.W., Heyer P.D., Ebels T.. Arterial revascularization with the right gastroepiploic artery and internal mammary arteries in 300 patients. J Thorac Cardiovasc Surg 1994;107:1309−1316.

8.Jegaden O., Eker A., Montagna P.. Risk and results of bypass grafting using bilateral internal mammary and right gastroepiploic arteries. Ann Thorac Surg 1995;59:955−960.

9.Mills N.L., Bringaze W.L.. Preparation of the internal mammary artery graft. Which is the best method?. J Thorac Cardiovasc Surg 1989;98:73−79.

10.Cooper G.J., Wilkinson G.A.L., Angelini G.D.. Overcoming perioperative spasm of the internal mammary artery. J Thorac Cardiovasc Surg 1992;104:465−468.

11.Sasson L., Cohen A.J., Hauptman E., Schachner A.. Effect of topical vasodilators on internal mammary arteries. Ann Thorac Surg 1995;59:494−496.

12.Suma H.. Spasm of the right gastroepiploic artery. Ann Thorac Surg 1990;49:168−169.

13.He G.W., Yang C.Q.. Comparison among arterial grafts and coronary artery. An attempt at functional classification. J Thorac Cardiovasc Surg 1995;109:707−715.

14.Cracowski J.L., Chavanon O., Durand M.. Effect of low-dose positive inotropic drugs on human internal mammary artery flow. Ann Thorac Surg 1997;64:1742−1746.

15.Mills N.L., Hockmuth D.R., Everson C.T., Robart C.C.. Right gastroepiploic artery used for coronary artery bypass grafting. Evaluation of flow characteristics and size. J Thorac Cardiovasc Surg 1993;106:579−585.

16.Dignan R.J., Yeh T., Dyke C.M.. Reactivity of gastroepiploic and internal mammary arteries. Relevance to coronary artery bypass grafting. J Thorac Cardiovasc Surg 1992;103:116−122.

17.Van Son J.A.M., Tavilla G., Noyez L.. Detrimental sequelae on the wall of internal mammary artery caused by hydrostatic dilation with diluted papaverine solution. J Thorac Cardiovasc Surg 1992;104:972−976.

18.Dregelid E., Heldal K., Resch F., Stangeland L., Breivik K., Svendsen E.. Dilation of the internal mammary artery by external and intraluminal papaverine application. J Thorac Cardiovasc Surg 1995;110:697−703.

19.Von Segesser L.K., Lehmann K., Turina M.. Deleterious effects of shock in internal mammary artery anatomoses. Ann Thorac Surg 1989;47:575−579.

20.Uydes-Dogan B.S., Negibil M., Aslamaci S., Onuk E., Kanzik I., Akar F.. The comparison of vascular reactivities of arterial and venous grafts to vasodilators. Int J Cardiol 1996;53:137−145.

21.Ali A.T., Montgomery W.D., Santamore W.P., Spence P.A.. Preventing gastroepiploic artery spasm. J Surg Res 1997;71:41−48.

22.Akar F., Uydes-Dogan B.S., Tufan H., Aslamaci S., Koksoy C., Kanzik I.. The comparison of the responsiveness of human isolated internal mammary and gastroepiploic arteries to levcromakalim. Br J Clin Pharmacol 1997;44:49−56.

23.Bilgen F., Yapici M.F., Serbetcioglu A., Tarhan I.A., Coruh T., Цzler A.. Effect of normothermic papaverine to relieve intraoperative spasm of the internal thoracic artery. Ann Thorac Surg 1996;62:769−771.

24.Arnaudov D., Cohen A.J., Zabeeda D.. Effect of systemic vasodilators on internal mammary flow during coronary bypass grafting. Ann Thorac Surg 1996;62:1816−1819.