Ann Thorac Surg 1997;64:1075-1081
© 1997 The Society of Thoracic Surgeons
Original Article: Cardiovascular
Effect of Flush-Perfusion on Vascular Endothelial and Smooth Muscle Function
Richard Ingemansson, MD, PhD,
Algimantas Budrikis, MD,
Ramunas Bolys, MD,
Trygve Sjöberg, PhD,
Stig Steen, MD, PhD
Department of Cardiothoracic Surgery, University Hospital, Lund, Sweden
Accepted for publication April 16, 1997.
 |
Abstract
|
|---|
Background. The aim of this study was to investigate how much perfusion pressure an artery can tolerate without significant loss of endothelium-dependent relaxation (EDR) and vascular contractility.
Methods. The abdominal aortas of 396 Sprague-Dawley rats were used. One hundred twenty aortas were flush-perfused for 1 or 5 minutes with cold St. Thomas' Hospital cardioplegic (STHC) solution or with the same solution but modified by the addition of 3.5% dextran 40. Three perfusion pressures were tested: 50, 100, and 150 mm Hg. Two hundred eighty vessels were subjected to pressures of 50, 150, or 300 mm Hg using saline or STHC solution at 22°C or STHC solution at 4°C, for 10 or 60 seconds. The vessels were investigated in organ baths. Contractility was tested with the thromboxane analogue U-46619, acetylcholine was used to investigate EDR, and papaverine to elicit endothelium-independent relaxation.
Results. Flush-perfusion with cold STHC solution for 5 minutes at a perfusion pressure of 50 or 100 mm Hg affected neither contractility nor EDR. Vessels exposed to a flush-perfusion pressure of 150 mm Hg for 1 or 5 minutes lost 39% (p < 0.001) and 53% (p < 0.001) of their contractility, respectively. Flush-perfusion at 150 mm Hg for 1 minute did not affect EDR, whereas 5 minutes' perfusion caused a reduction of 7% (p < 0.05). A repetition of these experiments using STHC solution with 3.5% dextran 40 added gave no significantly different results. The impairment in contractility and EDR seen after perfusion at 150 mm Hg for 5 minutes disappeared after transplantation and reperfusion for 7 days. The vessels could be distended with saline or STHC solution at a pressure of 150 mm Hg without affecting contractility at 22°C. At 4°C, however, this pressure was harmful to contractility. Distention at a pressure of 300 mm Hg almost abolished contractility and 7 days after transplantation there had not yet been any recovery of contractility, but 30 days after transplantation the grafts had regained their normal contractility.
Conclusions. Cold STHC solution, with or without dextran 40, can be used with a perfusion pressure of 100 but not 150 mm Hg without impairing EDR or vascular smooth muscle function.
|
Introduction
|
|---|
In heart operations, cardioplegia can be obtained by perfusing the coronary arteries with a cold cardioplegic solution (eg, St. Thomas' Hospital Cardioplegic solution). The flush-perfusion pressure used ranges from 50 to 300 mm Hg. If there is an occlusion in a coronary vessel there will be no flow in that vessel and a situation with pressure without flow will occur.
The aim of this study was to evaluate how different flush-perfusion pressures or pressures without flow affect vascular smooth muscle and endothelial function. It is of the utmost importance to know how much flush-perfusion pressure can be used without causing vascular damage, because, for example, endothelial dysfunction could lead to postoperative coronary spasm, thrombosis, and, later on, intimal hyperplasia and accelerated atherosclerosis.
The infrarenal rat aorta was used as a model because its diameter of 1 to 1.5 mm is the same as that of a medium-sized coronary artery. We have previously shown that the rat aorta can be dissected out and investigated without disturbing its endothelium-dependent relaxation and smooth muscle function [1]. Furthermore, it enabled us to reimplant flush-perfused segments into syngeneic rats for investigation after in vivo reperfusion [2, 3].
|
Material and Methods
|
|---|
Three hundred ninety-six Sprague-Dawley rats (275 to 325 g) were used in this study. Sixty animals served as donors and 60 as recipients. Twenty-four were used to provide samples for fresh controls. Twelve animals were used to provide samples for light microscopy. Ninety-six animals served in the flush-perfusion part and 144 in the distention part of the study. The animals were treated in compliance with the "Guide for Care and Use of Laboratory Animals" published by the National Institutes of Health (NIH publication 85–23, revised 1985).
|
Harvesting Procedure
|
|---|
After exposing the abdominal aorta, the segment between the renal arteries and the iliac bifurcation was dissected free from the inferior vena cava, using a dissecting microscope (Leika WILD M 691, Wild Leitz Ltd, Heerbrugg, Switzerland) for visualization. Two microvascular clamps were placed on the freed aorta in such a way as to isolate a segment 12 to 15 mm in length. A 10- to 12-mm-long graft segment was excised with intact iliac bifurcation and blood was removed by dripping Krebs solution (composition in mmol/L: K+, 4.6; Na+, 135; Mg2+, 1.2; Ca2+, 1.5; Cl-, 129; HCO3-, 15; H2PO4- + HPO42-, 1.2; glucose, 11) at room temperature through the lumen. In a previous study [4] we have shown that Krebs solution at room temperature does not impair endothelium-dependent relaxation or smooth muscle function during 2 hours of storage.
|
Flush-Perfusion Procedure
|
|---|
In 96 rats the abdominal aorta was dissected free and removed with an intact iliac bifurcation. In the proximal part of the aorta, a catheter, 1 mm in diameter, was inserted and fixed with a ligature. In the right iliac artery another catheter, 1 mm in diameter, was introduced and fixed with a ligature, and in the left iliac artery a third catheter, attached to a pressure transducer connected to a pressure recorder for continuous recording of pressure, was inserted (Fig 1
). This procedure was performed with the vessels lying in 37°C Krebs solution. St. Thomas' Hospital cardioplegic solution (composition in mmol/L: Na+, 120; K+, 16; Ca2+, 1.2; Mg2+, 16; Cl-, 170; Tris buffer to pH = 7.40) with and without 3.5% dextran 40 at 4°C was used to apply a flush-perfusion pressure of 50, 100, and 150 mm Hg for 1 or 5 minutes (Table 1). The pressure was regulated and standardized by changing the height of the tip of the output catheter. Two ring segments, each 1.0 to 1.2 mm in length, were taken from the midportion of the excised aorta immediately after perfusion and transferred to organ baths, or else the iliac bifurcation was cut away from the excised aorta and was then transplanted to a syngeneic rat and investigated after 7 or 30 days of reperfusion in vivo.
|
Distention Procedure
|
|---|
In 144 rats the abdominal aorta was dissected free and a length of 15 mm removed. A thin catheter, approximately 1 mm in diameter, was fixed to the proximal part of the aorta with a ligature (Fig 2). The catheter was connected to a pressure transducer and the signal was continuously monitored on a Hewlett-Packard fluoroscope (HP 78342 B, Andover, MA). A clamp was placed on the distal part of the aorta. The vessels were then exposed to St. Thomas' Hospital cardioplegic solution or saline solution at a pressure of 50, 150, or 300 mm Hg and at a temperature of 22°C for 10 or 60 seconds. The experiment was repeated with St. Thomas' Hospital cardioplegic solution at 4°C. Immediately after the distention two ring segments, each 1.0 to 1.2 mm in length, were taken from the midportion of the graft and transferred to the organ baths, or else the whole graft was immediately transplanted into a syngeneic rat and investigated after 7 or 30 days of in vivo reperfusion.
|
Recipient Procedure
|
|---|
All animals were given Streptocillin vet (dihydrostreptomycin and bensylpenicillinprokain; Boehringer-Ingelheim, Ingelheim/Rhein, Germany) 0.1 mL subcutaneously before the operation was started. The aorta was then prepared and clamped in the same way as in the donor and divided midway between the clamps. The graft segment was then interposed in its original orientation and sutured end-to-end with resorbable 9-0 Vicryl sutures (Ethicon, Somerville, NJ). Each anastomosis consisted of eight to ten interrupted sutures. Sixty animals served as donors and 60 as recipients.
Seven or 30 days later, the graft was removed and immediately placed in warm (37°C) and oxygenated (95% oxygen and 5% carbon dioxide) Krebs buffer solution and freed from any surrounding connective tissue. Two ring segments, each 1.0 to 1.2 mm in length, were then taken from the midportion of the graft and transferred to the organ baths.
|
Controls
|
|---|
The infrarenal aorta was extirpated and immediately investigated in organ baths as controls.
|
Recording Contractility and Endothelium-Dependent or -Independent Relaxation
|
|---|
Isometric tension was measured in organ baths that were water-mantled to keep the temperature of the bath solution at 37°C. The bath solution (Krebs) was bubbled with 95% oxygen and 5% carbon dioxide, giving it a pH of approximately 7.4. Each ring segment was suspended between two metal holders (0.2 mm in diameter). One holder was attached to a Grass FT 03 (Grass Instrument Co, Quincy, MA) transducer connected to a Grass polygraph for continuous recording of isometric tension. The other metal holder was fixed to an adjustable unit by means of which the vessel segments were repeatedly stretched until a basal tension of about 8 millinewton (mN) was reached. In separate experiments it has been found that the maximum response is obtained at this tension. Contraction was then induced with the thromboxane A2 analogue U-46619 (Upjohn Co, Kalamazoo, MI) added at a concentration of 10-6.5 mmol/L. In separate experiments, concentration–response curves have shown 10-8 mmol/L U-46619 to induce half-maximum contractions, and 10-6.5 mmol/L U-46619 to induce contractions ranging in strength from 90% to 100% of the maximum. After repeated washes, resulting in a restoration of the basal tension, a new contraction was induced with the same concentration of U-46619. When the degree of contraction had reached a stable plateau, increasing concentrations of acetylcholine (acetylcholine chloride; Sigma Chemical Co, St. Louis, MO) were cumulatively added to the baths. Acetylcholine causes the release of endothelium-derived relaxing factor by stimulating receptors in the endothelium. In previous experiments the endothelium was removed by gently rubbing the intimal surface over a pair of microtweezers and in these cases acetylcholine elicited no relaxation [3]. In each segment, the response to the different concentrations of acetylcholine was expressed as a percentage of the U-46619-induced contraction. If the relaxation obtained with acetylcholine in the preserved vessels was impaired, compared with the fresh controls, the endothelium-independent vasodilator papaverine (10-4 mmol/L) was added to the bath to ascertain whether complete relaxation could be obtained.
|
Histology
|
|---|
Six segments, each 1 mm in length, were taken for light microscopy from six fresh aortas and served as fresh controls. Twelve segments from 12 different rats were taken for microscopy after 5 minutes cold (4°C) flush perfusion, with a pressure of 100 mm Hg or 150 mm Hg, using St. Thomas' Hospital cardioplegic solution. The segments were fixed in 4% phosphate-buffered formalin–saline solution (pH, 7.0). They were dehydrated and embedded in paraffin, and sectioned and stained with hematoxylin and eosin.
|
Data Analysis
|
|---|
Results were expressed as the mean ± the standard error of the mean, n being the number of animals used in each group (n = 8 in all groups except in the transplanted groups in which n = 6). Statistical evaluation was carried out with one-way analysis of variance using Dunnet's test for the multiple comparisons. A p value of less than 0.05 was considered statistically significant. All experiments were randomized.
|
Results
|
|---|
Graft Patency
All grafts were patent 7 or 30 days after transplantation.
|
Contractile Response After Flush-Perfusion
|
|---|
As seen in Figures 3 and 4, perfusion with cold (4°C) St. Thomas' Hospital cardioplegic solution, with or without 3.5% dextran 40, for 1 or 5 minutes with a flush-perfusion pressure of 50 or 100 mm Hg did not significantly affect contractility. Vessels exposed to 1 minute of flush-perfusion, with or without 3.5% dextran 40, with a perfusion pressure of 150 mm Hg lost 29% (p < 0.05) and 39% (p < 0.001) of their contractile response, respectively, compared with fresh controls. Five minutes of cold flush-perfusion (4°C), with or without dextran 40, with a perfusion pressure of 150 mm Hg gave a significant reduction of 52% (p < 0.001) and 53% (p < 0.001), respectively, compared with fresh controls.
View larger version (42K):
[in this window]
[in a new window]
|
Fig 3. . Contractile response to the thromboxane analogue U-46619 (upper panel) and the maximum endothelium-dependent relaxation elicited by acetylcholine (lower panel). The bars show the results from fresh controls and from vessels exposed to 1 and 5 minutes of flush-perfusion at 4°C with modified St. Thomas' Hospital solution containing 3.5% dextran 40. Each bar represents the mean ± the standard error of the mean (n = 8 animals in each group). (*p < 0.05; **p < 0.01; ***p < 0.001.)
|
|
View larger version (43K):
[in this window]
[in a new window]
|
Fig 4. . Contractile response to the thromboxane analogue U-46619 (upper panel) and the maximum endothelium-dependent relaxation elicited by acetylcholine (lower panel). The bars show the results from fresh controls and from vessels exposed to 1 and 5 minutes of flush-perfusion at 4°C with St. Thomas' Hospital solution. Each bar represents the mean ± the standard error of the mean (n = 8 animals in each group). (*p < 0.05; ***p < 0.001.)
|
|
|
Contractile Response After Pressure Without Flow
|
|---|
As seen in Figure 5, with both saline solution and St. Thomas' Hospital cardioplegic solution at 22°C, the vessels could be subjected to a pressure of 150 mm Hg, but not 300 mm Hg, without significantly affecting contractility. Only St. Thomas' Hospital cardioplegic solution was tested at 4°C, and at that temperature contractility was reduced by 29% (p < 0.05) after distention at 150 mm Hg for 60 seconds.
View larger version (51K):
[in this window]
[in a new window]
|
Fig 5. . Contractile response to the thromboxane analogue U-46619 (upper panel) and the maximum endothelium-dependent relaxation elicited by acetylcholine (lower panel). Each bar represents the mean ± standard error of the mean (n = 8 animals in each group). Endothelium-dependent relaxation could not be investigated in vessels exposed to 300 mm Hg for 60 seconds, as they had lost almost all contractility (A). (*p < 0.05; **p < 0.01; ***p < 0.001 compared with fresh controls.)
|
|
|
Endothelium-Dependent Relaxation After Flush-Perfusion
|
|---|
As seen in Figures 3 and 4
, perfusion with cold St. Thomas' Hospital cardioplegic solution, with or without 3.5% dextran 40, for 1 or 5 minutes with a flush-perfusion pressure of 50 or 100 mm Hg gave no significant decrease in endothelium-dependent relaxation, compared with fresh controls. One minute of cold flush-perfusion, with or without 3.5% dextran 40, and a flush-perfusion pressure of 150 mm Hg gave no significant decrease in endothelium-dependent relaxation. Five minutes of flush-perfusion with St. Thomas' Hospital cardioplegic solution, with or without added dextran 40, with a flush-perfusion pressure of 150 mm Hg gave a reduction of 17% (p < 0.01) and 7% (p < 0.05) in endothelium-dependent relaxation capacity, respectively, compared with fresh controls. However, the differences between these groups were not statistically significant.
|
Endothelium-Dependent Relaxation After Pressure Without Flow
|
|---|
As seen in Figure 5, distention with saline solution was injurious to endothelium-dependent relaxation. Endothelium-dependent relaxation was preserved after subjection to St. Thomas' Hospital cardioplegic solution at 22°C and 150 mm Hg, but not at 22°C and 300 mm Hg (p < 0.01). At 4°C, endothelium-dependent relaxation was impaired at both pressures.
|
Endothelium-Dependent Relaxation and Contractile Function 7 and 30 Days After Transplantation
|
|---|
As seen in Figure 6, vessels flush-perfused for 5 minutes with cold St. Thomas' Hospital cardioplegic solution containing 3.5% dextran 40 with a flush-perfusion pressure of 100 and 150 mm Hg, and then transplanted and investigated after 7 or 30 days, showed no significant decrease in contractility or endothelium-dependent relaxation, compared with fresh controls. As seen in Figure 7, vessels distended with a pressure of 300 mm Hg had not regained their contractility 7 days after transplantation, but 30 days after transplantation full recovery of contractility had occurred.
View larger version (45K):
[in this window]
[in a new window]
|
Fig 6. . Contractile response to the thromboxane analogue U-46619 (upper panel) and the maximum endothelium-dependent relaxation elicited by acetylcholine (lower panel). The bars show the results from fresh controls and from vessels exposed to 5 minutes of flush-perfusion at 4°C with modified St. Thomas' Hospital solution containing 3.5% dextran 40, transplanted and reperfused for 7 and 30 days before being investigated. Each bar represents the mean ± standard error of the mean (n = 6 animals in each group). (**p < 0.01; ***p < 0.001.)
|
|
View larger version (43K):
[in this window]
[in a new window]
|
Fig 7. . Contractile response to the thromboxane analogue U-46619 (upper panel) and the maximum endothelium-dependent relaxation elicited by acetylcholine (lower panel). The bars show the results from vessels exposed to St. Thomas' Hospital cardioplegic solution at 50, 150, and 300 mm Hg for 60 seconds at 22°C followed by transplantation and investigation after 7 or 30 days. Each bar represents the mean ± standard error of the mean (n = 6 animals in each group). Endothelium-dependent relaxation could not be investigated in vessels exposed to 300 mm Hg for 60 seconds, as they had lost almost all contractility (A). Exposure to 50 mm Hg and reimplantation for 30 days was not performed. (Reimpl. = reimplantation; *p < 0.05; **p < 0.01; ***p < 0.001 compared with fresh controls.)
|
|
|
Preservation of Endothelium-Independent Relaxation
|
|---|
In cases in which full relaxation was not obtained with acetylcholine, the endothelium-independent vasodilator papaverine (10-4 mol/L) elicited complete relaxation in all cases.
|
Structural Changes in the Grafts
|
|---|
The infrarenal aortas of 6 rats, investigated after flush-perfusion with St. Thomas' Hospital cardioplegic solution (4°C) with 3.5% dextran 40 for 5 minutes at 100 mm Hg and reimplanted for 30 days, showed no signs of intimal hyperplasia. However, after 5 minutes flush-perfusion at 150 mm Hg followed by reimplantation for 30 days, one vessel of six evidenced intimal hyperplasia (Fig 8).
View larger version (174K):
[in this window]
[in a new window]
|
Fig 8. . Light microscopy on a vessel investigated immediately after harvesting (A, B); light microscopy on a vessel investigated after flush-perfusion for 5 minutes with St. Thomas' Hospital cardioplegic solution (4°C), with 3.5% added dextran 40, at a pressure of 100 mm Hg (C, D); and light microscopy on a vessel investigated after flush-perfusion for 5 minutes with St. Thomas' Hospital cardioplegic solution (4°C), with 3.5% added dextran 40, at a pressure of 150 mm Hg (E, F). A–D demonstrate a normal endothelium but E and F, which come from the same segment, show signs of intimal thickening (arrows). The same magnification (x190) was used in A, C, and E, but in higher magnification (x740) was used B, D, and F.
|
|
|
Comment
|
|---|
St. Thomas' Hospital cardioplegic solution is widely used in routine heart operations. This solution does not contain colloids. The reasons for using colloids such as dextran 40 in a cardioplegic solution are several. One is to create oncotic pressure and thus counteract interstitial edema formation during the flush-perfusion procedure. Protein molecules are the only dissolved substances in plasma that do not readily penetrate the pores of the capillary membrane and they are therefore responsible for the osmotic pressure at the capillary membrane [5]. Albumin (molecular weight around 69,000 daltons) is, because of its amount and low molecular weight compared with the other plasma proteins, responsible for about 70% of the oncotic pressure of plasma. The normal concentration of albumin in plasma is about 4 g/100 mL, ie, 4%. Because the modified St. Thomas' Hospital cardioplegic solution in this study contains 3.5% dextran 40 with an average molecular weight of 40,000 daltons, there will be more dextran 40 molecules in the modified St. Thomas' Hospital cardioplegic solution than albumin molecules in an equivalent volume of normal plasma. Dextran 40, like albumin, does not readily penetrate capillary membranes of the heart. Therefore, St. Thomas' Hospital cardioplegic solution modified by the addition of 3.5% dextran 40 has an oncotic pressure higher than that of normal plasma.
Flush-perfusion of the vessels with cold (4°C) St. Thomas' Hospital cardioplegic solution, with or without 3.5% dextran 40, using a perfusion pressure 100 mm Hg for 5 minutes resulted in normal vascular smooth muscle and endothelium-dependent relaxing factor function. However, exposure of vessels to a perfusion pressure of 150 mm Hg for 1 minute gave a significant decrease in smooth muscle function, although the endothelium-dependent relaxation remained intact both with and without dextran 40. With a flush-perfusion pressure of 150 mm Hg for 5 minutes, an impairment of endothelium-dependent relaxation was seen both with and without dextran 40. However, after 7 or 30 days of in vivo reperfusion the vessels showed normal endothelium-dependent relaxation and contractile properties compared with fresh controls. Vessels investigated after flush-perfusion with St. Thomas' Hospital cardioplegic solution (4°C) with 3.5% dextran 40 for 5 minutes at 100 mm Hg and reimplanted for 30 days demonstrated no sign of intimal hyperplasia. However, when the pressure was raised to 150 mm Hg one vessel of six demonstrated intimal hyperplasia (Figs 7, 8). Even if the dextran 40 does not improve endothelium-dependent relaxation after flush-perfusion in our model it might still be interesting to use in cardioplegic solutions because of other well-documented positive effects. There are several studies demonstrating the potentially positive effects of an increased oncotic pressure in counteracting interstitial edema caused by leakage of the preservation solution through the capillary pores. Dextran 40 also causes an improved microcirculatory flow by coating the surface of blood cells and endothelial cells and may have an oxygen free radical scavenging effect [5–10].
A loss in contractility in, for example, coronary arteries after a heart operation is not necessarily bad because it might prevent spasm in the vessels in the postoperative period. But excellent preservation of the endothelium is of the utmost importance because dysfunctional, swollen endothelial cells at the capillary level might lead to impaired reperfusion [11]. Intimal hyperplasia may also develop after trauma to the endothelium [12]. A decrease in endothelial function might contribute to increased vascular resistance [13].
Vessels exposed to St. Thomas' Hospital cardioplegic solution with a distention pressure of 150 mm Hg for 1 minute at 22°C did not lose any vascular smooth muscle or endothelial function. However, when the temperature was lowered to 4°C, there was a significant loss in both contractility and endothelium-dependent relaxation. Thus, it seems that cold flush-perfusion is more harmful than perfusion at higher temperatures. We have previously shown that cold storage (4°C) causes endothelial dysfunction that is completely reversible after 2 hours of reperfusion in vivo [3]. Exposure to low temperatures can impair the basal and stimulated release of endothelium-dependent relaxing factor [14]. Studies on cultured human endothelial cells have shown that structural changes are induced by hypothermia, but rewarming elicits a rapid and nearly complete reversal of these changes [15]. The fact that low temperatures, high pressures, and ischemia, which are the result of using crystalloid cardioplegia, damage the endothelium might speak for the advantage of blood cardioplegia, in which extreme cooling is not necessary and the heart is not subjected to ischemia.
We wanted to evaluate if a toxic solution, such as saline solution [4], is detrimental to vascular function even after short exposure periods (1 minute) with different distention pressures. Saline solution caused a significant decrease in endothelium-dependent relaxation after 1 minute with a distention pressure of 50 mm Hg. St. Thomas' Hospital cardioplegic solution caused no significant decrease in endothelium-dependent relaxation after 1 minute of distention at 150 mm Hg. St. Thomas' Hospital cardioplegic solution was clearly superior to saline solution in the present study. We have previously shown that storage of vessels in normal saline solution, which is still widely used for irrigation and short-term storage of vascular grafts, causes a significant reduction in endothelium-dependent relaxation after 2 hours, both at 4°C and 22°C [4]. In a study by O'Connell and coworkers [12], it was shown that infusion of arteries with normal saline solution produces immediate topographic changes, demonstrable by scanning electron microscopy; when examined 1 month after infusion of saline solution, the arteries manifested moderate to severe intimal thickening.
Vessels exposed to St. Thomas' Hospital cardioplegic solution at 300 mm Hg for 10 seconds demonstrated a significant decrease in vascular smooth muscle and endothelial function. If the exposure time was extended to 1 minute almost all contractile capacity was lost and endothelium-dependent relaxation could not be tested. This is of importance in a situation in which a coronary vessel is occluded. No movement of fluid occurs in the vessel and theoretically the pressure in the vessel can rise to 300 mm Hg, which is the pressure applied to the cardioplegia bag. This demonstrates that perfusion pressure, perfusion time, and perfusion temperature are important variables to consider if good vascular smooth muscle and endothelial function are to be obtained when using flush-perfusion.
In conclusion, cold St. Thomas' Hospital cardioplegic solution can be used with flush-perfusion pressures of 100 mm Hg but not 150 mm Hg without impairing endothelium-dependent relaxation or smooth muscle function. Furthermore, we have demonstrated that the perfusion pressure, perfusion time, perfusion temperature, and type of solution used seem to be important variables for preserving endothelium-dependent relaxation and smooth muscle func
Top
Abstract
Introduction
