- Email: [email protected]
Functional assessment of human femoral arteries after cryopreservation
Functional assessment of human femoral arteries after cryopreservation Françoise Stanke, PharmD, Danièle Riebel, Sessa Carmine, MD, Jean-Luc Cracowski, MD, F. Caron, Jean-Luc Magne, MD, Harald Egelhoffer, PharmD, Germain Bessard, MD, and Philippe Devillier, MD, PhD, La Tronche Cedex, France Purpose: An established method of cryostorage that might preserve the vascular and endothelial responses of human femoral arteries (HFAs) to be transplanted as allografts was studied. Methods: HFAs were harvested from multiorgan donors and stored at 4°C in Belzer solution before cryostorage. One hundred eleven HFA rings were isolated and randomly assigned to 1 control group of unfrozen HFAs and 2 groups of HFAs cryopreserved for 7 and 30 days, respectively. Cryopreservation was performed in Elohes solution containing dimethyl sulfoxide (1.8 mmol/L), and the rate of cooling was 1.6°C/min, until –141°C was reached. The contractile and relaxant responses of unfrozen and frozen/thawed arteries were assessed in organ bath by measurement of isometric force generated by the HFAs. Results: After thawing, the maximal contractile responses to all the contracting agonists tested (KCl, U46619 [a thromboxane A2-mimetic], norepinephrine, serotonin, and endothelin-1) were in the range of 7% to 34% of the responses in unfrozen HFAs. The endothelium-independent relaxant responses to forskolin and verapamil were weakly altered, whereas the endothelium-independent relaxant responses to sodium nitroprusside were markedly reduced. Cryostorage of HFAs also resulted in a loss of the endothelium-dependent relaxant response to acetylcholine. The vascular and endothelial responses were similarly altered in the HFAs cryopreserved for 7 and 30 days. Conclusion: The cryopreservation method used provided a limited preservation of HFAs contractility, a good preservation of the endothelium-independent relaxant responses, but no apparent preservation of the endothelium-dependent relaxation. It is possible that further refinements of the cryopreservation protocol, such as a slower rate of cooling and a more controlled stepwise addition of dimethyl sulfoxide, might allow better post-thaw functional recovery of HFAs. J Vasc Surg 1998;28:273-283.)
INTRODUCTION Cryopreservation of tissue at –70°C to –196°C offers the prospect of virtually indefinite storage. Improvements to the preservation techniques, such From the Laboratory of Pharmacology, PCEBM, Faculté de Médecine de Grenoble, and the Blood Transfusion Center (Riebel and Dr Egelhoffer), and the Department of Vascular Surgery (Dr Magne), Centre Hospitalier Universitaire Grenoble. Supported by a grant from the Commission de la Recherche Clinique du Centre Hospitalier de Grenoble, France. Reprint requests: Françoise Stanke, Laboratoire de Pharmacologie, PCEBM, Faculté de Médecine, 38706 La Tronche Cedex, France. Copyright © 1998 by The Society for Vascular Surgery and Internation Society for Cardiovascular Surgery, North American Chapter. 0741-5214/98/$5.00 + 0 24/1/91875
as the introduction of cryoprotectant agents and a controlled rate of freezing, have enabled the development of human blood vessel banks. Cryopreserved allogenic saphenous veins1 and human femoral arteries (HFAs)2,3 are used for the bypass of diseased peripheral arteries and the treatment of limb-threatening ischemia or infected grafts. The reactivity of human saphenous veins has been extensively studied on both unfrozen and cryopreserved tissues.4–8 However, no data exists concerning the in vitro reactivity of HFAs, whereas the arteries appear considerably more susceptible to cryoinjury than the veins.6 The responses to various contracting and relaxing agents of both unfrozen HFAs and HFAs cryopreserved for 7 days and 30 days was assessed. The harvesting, freezing, and thawing procedures were those routinely applied to 273
274 Stanke et al
HFAs that are to be transplanted as allografts. The HFAs were stored at –141°C in Elohes solution containing dimethyl sulfoxide (DMSO). MATERIALS AND METHODS Human femoral arteries preparation and storage method. Samples of HFAs were carefully harvested from 9 multiorgan donors after permission was obtained from the local ethical committee. The tissues were immediately placed in Belzer solution at +4°C for 4 to 36 hours. The Belzer solution (ViaSpan) had the following composition (mmol/L): hydroxyethylstarch (5.0), lactobionic acid (100.0), KHPO3 (28.5), MgSO4, 7 H2O (5.0), raffinose pentahydrate (30.0), adenosine (5.0), allopurinol (1.0), total glutathione (3.0), KOH q.s., NaOH adjusted to pH 7.4. The femoral arteries were then cut into 3 equal segments approximately 3 cm long and were randomly divided into 3 groups. The control group (D0) consisted of fresh, “unfrozen” HFAs that were transferred to the laboratory for organ bath studies. The 2 cryopreserved groups were stored at –141°C in gas nitrogen for 7 days (D7) or 30 days (D30). In both cryostored groups, individual arterial segments were first transferred in Elohes solution (hydroxyethylstarch 200,000 [0.3 mmol/L], sodium chloride [15.4 mmol/L]) and then progressively loaded with DMSO in 3 concentration steps (4%, 9%, and 14% [weight/weight]), each lasting 2 minutes. The segments were stored in separated cryotubes containing 2 ml of the Elohes solution containing DMSO 14% (1.8 mmol/L) and were allowed to equilibrate at 4°C for approximately 30 minutes before their transfer to an automatic freezing machine (minidigitcool IMV, Bicef, France). The mean cooling rate was about 1.6°C/min, until –141°C was reached. After the predetermined period of cryostorage of 7 days or 30 days, individual blood vessel segments were quickly thawed by means of total immersion of the cryotubes in a water bath (41°C) for 2 minutes. Organ-bath technique. Immediately after thawing, the arterial segments were precisely cut into rings 5 mm long. The number of rings obtained from the femoral artery segments varied from 4 to 6. The rings were placed in a 10 mL organ bath filled with KrebsHenseleit solution (mmol/L: NaCl 118, KCl 4.7, CaCl2 2.5, MgSO4 1, KH2PO4 1, glucose 11, NaHCO3 25), which was continuously bubbled with a gas mixture (95% O2/5% CO2) and maintained at 37°C. The arterial rings were suspended on wires; the lower wire was fixed to a micrometer (Mitutoyo, Japan), and the upper was attached to a force trans-
JOURNAL OF VASCULAR SURGERY August 1998
ducer (UF-1, Pioden, U.K.) through which changes in isometric force were displayed on recorders (Linseis 200, Bioblock, France). After being cautiously mounted on the wires to preserve the endothelium, the rings were allowed to equilibrate unstretched for 15 to 30 minutes. A normalization procedure was then performed to set the vascular rings under in vitro stress comparable to the in vivo pressure, as previously reported.9,10 Briefly, the rings were progressively stretched with the micrometer at intervals of 5 g to determine the internal circumference (calculated from the distance between the wires)-tension exponential curve for each ring. When the transmural pressure on the rings reached 100 mm Hg, this stretch procedure was stopped, and the rings were then released to approximately 90 of their internal circumference at 100 mm Hg (0.9 D100). The isometric force at this setting is the “passive” or “resting” force in the absence of constrictor tone. This normalized degree of stress was then maintained throughout the experiment. Internal diameters at a pressure of 100 mm Hg and the resting forces at 0.9 D100 were recorded. PROTOCOL After the normalization procedure, the arterial rings were allowed to equilibrate for 60 minutes. The rings were then contracted with KCl 90 mmol/L, rinsed with Krebs solution, and allowed to equilibrate for 1 hour. This initial maximal contraction was performed to stabilize the preparations and to ensure reproducible contractile responsiveness. During all the equilibration periods, the Krebs solution was changed every 15 minutes. Thereafter, concentration-response curves were established by cumulative addition of increasing concentrations (0.5-log increments) of the drugs, the following concentration being added when the maximal effect had been produced by the previous concentration. Concentration-contraction curves were established for the following ranges of concentrations of contracting substances: KCl (7.5-120 mmol/L), thromboxane A2 (TXA2)-mimetic U46619 (1 nmol/L-3 µmol/L), norepinephrine (1 nmol/L-100 µmol/L), 5-hydroxytryptamine (5-HT) (1 nmol/L100 µmol/L), and endothelin-1 (1 nmol/L-0.3 µmol/L). When relaxing agents were investigated, the rings were precontracted with 30 mmol/L KCl. Endothelium-independent relaxations were assessed by establishing concentration-relaxation curves for the following ranges of concentrations of relaxing substances: sodium nitroprusside (SNP) (1 nmol/L-100 µmol/L), forskolin (1 nmol/L-30 µmol/L), and ver-
JOURNAL OF VASCULAR SURGERY Volume 28, Number 2
Stanke et al
275
apamil (1 nmol/L-30 µmol/L). Endotheliumdependent relaxation was assessed by measuring relaxation to acetylcholine (1 µmol/L)11. After the maximal effect of each drug was obtained, papaverine (300 µmol/L) was added to determine the maximum relaxation achievable. On each arterial ring, only one contracting agonist was tested, followed by investigation of one relaxant compound. DATA ANALYSIS Contractions are expressed as grams of isometric force developed by the arterial rings in response to the contracting agonists. Relaxation elicited by papaverine (300 µmol/L) was considered 100% relaxation. Endothelium-dependent relaxations and smooth muscle-dependent relaxations are expressed as percentages of the maximal relaxations achieved with papaverine. The smooth muscle-dependent relaxations were compared by determining the concentration (IC40) of agonist producing 40% of this maximal papaverine-response. Maximum effect (Emax) was the greatest response obtained with an agonist. For the vasoconstrictor agonists studied, the effective concentration of substance that caused 50% of the maximal response (EC50) was determined from each curve by a logistic, curve-fitting equation. The pD2 value represents the negative logarithm of EC50. Drug concentrations are expressed as final molar concentrations in the organ bath. Data are given as mean ± SEM. Statistical analysis was performed using one-way analysis of variance (ANOVA) for multiple comparisons, followed by Bonferroni corrected t-test and individual comparisons were made by Student’s t-test for paired or impaired data. P<.05 was considered to be significant. Drugs. TXA2 stable mimetic U46619 (9,11dihydroxy-11, 9-epoxy-methanoprostaglandin F2), sodium nitroprusside (sodium nitroferricyanide), 5hydroxytryptamine (5-hydroxytryptamine hydrochloride), papaverine, and forskolin were obtained from Sigma (L’isle d’Abeau, France); norepinephrine bitartrate was obtained from Assistance Publique des Hôpitaux de Paris (France); endothelin-1 was obtained from Euromedex (Souffelweyersheim, France); KCl Normapur was obtained from Prolabo (Paris, France); verapamil was obtained from Knoll (Levallois Perret, France); Elohes solution was obtained from Fresenius (Louviers, France); and Belzer solution (ViaSpan) was obtained from Dupont (Herts, Netherlands). All drugs were dissolved in distilled water, except forskolin, for which a stock solution at 100 µmol/L
Fig. 1. Resting force-internal diameter relationships for rings of fresh human femoral arteries (solid circle, n = 12) and of human femoral arteries cryopreserved for 7 days (open circle, n = 12) and for 30 days (open box, n = 12).
was prepared in DMSO. Drugs were kept at –20°C and were freshly dissolved in distilled water to the appropriate final concentrations. RESULTS Resting vessel parameters of the femoral arteries. One hundred eleven femoral arterial rings taken from 9 multiorgan donors were used. The mean internal diameters at an equivalent transmural pressure of 100 mm Hg were 5.3 ± 0.5 mm (D0), 5.1 ± 0.3 mm (D7), and 4.5 ± 0.4 mm (D30), and the resting forces at 0.9 D100 were 17.0 ± 1.0 g (D0), 16.8 ± 0.6 g (D7), and 15.8 ± 0.7 g (D30). No statistically significant difference was found between the resting vessel parameters of unfrozen and cryopreserved arteries. The relations between resting force and vessel internal diameter are shown in Fig. 1. Contractile responses. The contractile force as assessed by the maximal response in grams of the HFAs rings was markedly reduced in preparations from frozen/thawed arteries. In fresh HFAs, the maximal contractile response was diminished up to 34% for KCl (Fig. 2, A), up to 23% for U46619 (Fig. 2, B), up to 7% for norepinephrine (Fig. 2, C), up to 13% for 5-HT (Fig. 2, D), and up to 23% for endothelin-1 (Fig. 2, E). Nevertheless, KCl, endothelin-1, and U46619 (at day 30) stimulated the fresh and frozen/thawed arteries with similar
276 Stanke et al
JOURNAL OF VASCULAR SURGERY August 1998
Fig. 2. Cumulative concentration-contractile response curves on rings from unfrozen (solid circle) and frozen/thawed for 7 days (open circle) or 30 days (solid triangle) human femoral arteries in response to KCl (A), U46619 (B), norepinephrine (C), 5-HT (D), and endothelin1 (E). Results are expressed as gram force developed by ring vessels. The vertical bars represent mean ± SEM of 4 to 7 experiments.
JOURNAL OF VASCULAR SURGERY Volume 28, Number 2
Stanke et al
277
Table I. Responses to contractile agents of unfrozen controls and human femoral arteries cryopreserved for 7 (D7) or 30 days (D30) Unfrozen D0 U46619 KCl Norepinephrine 5-HT Endothelin-1
pD2 Emax pD2 Emax pD2 Emax pD2 Emax pD2 Emax
(g) (g) (g) (g) (g)
8.7 24.2 1.9 14.1 6.1 22.5 7.5 17.8 7.8 21.3
± ± ± ± ± ± ± ± ± ±
0.4 5.0 0.2 3.0 0.1 4.2 0.3 4.5 0.2 2.2
Frozen/Thawed D7 7.7 12.7 1.7 3.3 5.7 3.6 6.4 3.2 7.5 4.9
± ± ± ± ± ± ± ± ± ±
0.4* 1.4* 0.0 0.8* 0.1* 1.6* 0.8* 1.1* 0.4 2.0*
D30 8.2 8.3 1.6 4.6 5.3 1.6 5.4 2.3 7.6 5.9
± ± ± ± ± ± ± ± ± ±
0.4 1.3* 0.1 2.4* 0.4* 0.5* 0.2* 0.7* 0.4 0.4*
Emax = maximal contraction expressed as gram isometric force developed by the rings. pD2 = –logEC50, negative logarithm of the molar concentration producing 50% of maximal response. Data are given as means ± SEM of 4 to 7 experiments. *Significant differences (P<.05) against values determined in unfrozen controls. No significant difference was observed between HFAs cryostored for 7 or 30 days.
potencies. However, the apparent pD2 values were significantly reduced at day 7 for U46619 and at both day 7 and day 30 for norepinephrine and 5-HT (Table I). In terms of maximal contraction and potency for all these contracting agents, there was no significant difference between HFAs cryopreserved for 7 days or those cryopreserved for 30 days. Endothelium-independent relaxant responses. Relaxant responses to various agents, known to act at the vascular smooth muscle through different mechanisms, were investigated on fresh and frozen/thawed HFAs rings that had been precontracted with KCl 30 mmol/L. In preliminary experiments, it was observed that the maximal relaxation induced by papaverine (300 µmol/L) was unchanged in fresh and frozen/thawed arteries. The magnitude of the relaxations induced by papaverine, expressed as a percentage of the active tone induced by KCl, was 148 ± 12% (n = 16) in D0, 182 ± 16% (n = 14) in D7, and 184 ± 28% (n = 12) in D30. Therefore, the responses to forskolin, verapamil, and SNP in fresh and frozen/thawed arteries were expressed as a percentage of the maximal relaxation elicited by papaverine. Verapamil (Fig. 3, A) and forskolin (Fig. 3, B) elicited similar relaxant responses from both fresh and cryopreserved arteries. The –log IC40 values and maximal responses were not statistically different in fresh and frozen/thawed femoral arteries (Table II). In contrast, marked differences of both potency (expressed as –log IC40) and efficacy between fresh and frozen/thawed arteries were observed when SNP-induced relaxation was investigated (Fig.3, C; Table II). In terms of maximal relaxation and potency for all these relaxant agents, there was no signifi-
cant difference between cryopreserved HFAs for 7 or 30 days. Endothelium-dependent relaxant responses. Endothelium-dependent relaxations were investigated on fresh and frozen/thawed HFAs precontracted with 30 mmol/L of KCl. Acetylcholine (1 µmol/L) induced a relaxant response in fresh HFAs that reached 40 ± 11% (n = 8) of the maximal relaxation elicited by papaverine, whereas arteries cryopreserved for 7 or 30 days failed to relax in response to acetylcholine. Duration of storage at +4°C and recovery after cryopreservation. The duration of storage at +4°C in Belzer solution of the HFAs before cryopreservation varied from 4 to 36 hours. The relationship between the time of cold storage and the postthaw functional recovery has been evaluated in terms of both recovery of contractile response to the different vasoconstrictors tested (Fig. 4, A) and recovery of endothelium-independent relaxant response to SNP (Fig. 4, B and C). No relationship was found between the time of cold storage in Belzer and the reduction in these 2 types of response in the frozen/thawed femoral arteries. DISCUSSION Cryopreserved human arteries are used for the bypass of diseased peripheral arteries, and cryopreserved femoral arteries have been used for the treatment of limb-threatening ischemia2 or infection of arterial bypass grafts.3 It was therefore important to evaluate the functional alteration resulting from cryostorage. Our study is the first to compare the contractile responses as well as the endotheliumdependent and endothelium-independent relaxant
JOURNAL OF VASCULAR SURGERY August 1998
278 Stanke et al
Table II. Responses to relaxant agents of unfrozen controls and human femoral arteries cryopreserved for 7 (D7) or 30 days (D30) Unfrozen D0 SNP Forskolin Verapamil
–log IC40 Emax –log IC40 Emax –log IC40 Emax
8.1 93.5 6.1 75.3 6.8 75.4
± ± ± ± ± ±
0.1 1.9 0.4 10.5 0.2 7.3
Frozen/Thawed D7 6.3 79.3 5.3 73.1 6.9 91.9
± ± ± ± ± ±
D30 0.4* 4.6* 0.2 5.7 0.2 2.8
5.4 65.0 5.3 65.6 6.5 83.8
± ± ± ± ± ±
0.4* 6.5* 0.2 9.7 0.3 4.8
Emax = maximal relaxation expressed as percentage of maximal response to papaverine 300 µM. IC40 are expressed as the molar concentration producing 40% of relaxation. Data are given as means ± SEM of 4 to 7 experiments. *Significant differences (P<.05) against values determined in unfrozen controls. Statistical analysis of the IC40 was performed on the –log IC40. No significant difference was observed between HFAs cryostored for 7 or 30 days.
responses of both unfrozen and cryopreserved HFAs. The resting vessel parameters (internal diameter at an equivalent transmural pressure of 100 mm Hg and resting force) were similar in unfrozen and frozen/thawed HFAs, allowing a comparison of the responses with vasoactive agents at a comparable basis. In addition, the resting force-internal diameter relationship was the same for both frozen and cryopreserved HFAs, suggesting that the passive mechanical properties of the HFAs were not altered by cryopreservation. This assumption is in agreement with the results of previous studies in which mechanical testing with a traction-compression apparatus12,13 and viscoelastic properties evaluation with a mock circulation loop14 showed that cryopreservation does not alter the mechanical and viscoelastic characteristics of human descending thoracic aorta12,13 and superficial femoral artery14. Cryopreservation was associated with a marked reduction of the contractility. The maximal contractile responses were in the range of 7% to 34% of the responses in unfrozen HFAs with the different receptor-mediated stimuli and also with KCl, which provokes a contraction strictly dependent on Ca2+influx through voltage-operated channels. A reduction of the post-thaw contractility to less than 40% of fresh tissue has been observed in human coronary,15 internal mammary,16 and postmortem basilar arteries,17 whereas the maximal response to norepinephrine was only reduced to 75% in human pulmonary arteries.18 Freezing of vascular preparations generally results in a loss of contractile responsiveness.17,19 A post-thaw contractile force of about 60% to 80% of the maximum achieved in fresh tissue has been obtained following cryopreservation of animal arteries,19–21 although post-thaw recoveries of less than 50% have also been reported.22,23 In addition, Ku et al.15,20 have shown that an identical cryo-
preservation procedure applied to canine and human coronary arteries resulted in a greater loss of contractility in human arteries, indicating vessel to vessel variability in the susceptibility to cryoinjury. In contrast with the considerable reductions of the maximal contractile responses, potencies in terms of apparent pD2 values for KCl and endothelin-1 were not altered, and values for U46619, norepinephrine, and 5-HT were only moderately diminished after the cryopreservation procedure. These results are in line with previous studies showing that the sensitivities of human arteries and veins and animal blood vessels for various contracting agonists were well-maintained after cryopreservation.4,16,18,20,21,23–25 In addition, the endothelium-independent relaxant responses appear weakly altered by the cryopreservation procedure. The relaxations induced by forskolin, which acts through direct activation of adenylate cyclase, and by verapamil, which acts through calcium channel blockade, were similar in fresh and cryopreserved HFAs. However, the cryopreserved HFAs were less sensitive than fresh HFAs to the relaxant effect of SNP, which acts through direct activation of guanylate cyclase. A decrease in responsiveness to nitric oxide was previously reported in canine saphenous graft after cryopreservation.25 Moreover, cryopreserved human coronary arteries were recently reported to be significantly less sensitive to SNP than unfrozen tissues, whereas a slight, although nonsignificant, supersensitivity to SNP was observed in cryopreserved human mesenteric arteries.26 These results suggest that the cyclic guanosine monophosphate (cGMP)-dependent pathway of relaxation is susceptible to alteration by the cryopreservation process and that the development of supersensitivity or hyposensitivity to nitrovasodilators is tissue-specific. After cryoconservation, a weaker sensitivity has also been reported for potassium chan-
JOURNAL OF VASCULAR SURGERY Volume 28, Number 2
Stanke et al
279
Fig. 3. Cumulative concentration-relaxant response curves on rings from unfrozen human femoral arteries (solid circle) and human femoral arteries frozen/thawed for 7 days (open circle) or 30 days (solid triangle) for verapamil (A), forskolin (B), sodium nitroprusside (C). Results are expressed as percentages of the maximal relaxation induced by papaverine 3.10-4 M. The vertical bars represent mean ± SEM of 4 to 6 experiments.
nel activators in human pulmonary arteries18 and for isoproterenol in human coronary arteries.15 Again, an identical cryopreservation procedure applied to canine and human coronary arteries resulted in a reduction of the relaxant response to isoproterenol in
human arteries, whereas the response was well-preserved in canine arteries.15,20 In contrast to the marked reduction of the maximal contractile responses, the maximal relaxant responses were not altered or were altered much less after cryoconservation.15,20–22
280 Stanke et al
JOURNAL OF VASCULAR SURGERY August 1998
Fig. 4. Effect of the duration of storage at +4°C in Belzer solution on the functional recovery of human femoral arteries cryopreserved for 7 or 30 days. A, Effect of the duration of storage on the reduction of the maximal contractile response to U46619 (open circle), norepinephrine (open box), 5-HT (crossed box), KCl (solid triangle), and endothelin-1 (open triangle). B, Effect of the duration of storage on the reduction of the relaxant response to SNP as percentage of the reduction in the response to 10-5 M SNP (open circle). C, Effect of the duration of storage on the reduction of the relaxant response to SNP as log-shift to the right of the concentration-response (C/R) curve to SNP (open circle).
Cryostorage of HFAs resulted also in a loss of the endothelium-dependent relaxant response to acetylcholine. The alteration of the cGMP-dependent pathway of relaxation may, at least in part, account for this result because acetylcholine induced an endotheliumdependent relaxation by stimulating the production and release of nitric oxide from the endothelial cells, which, in turn, induces the relaxation of the vascular smooth muscle through the cGMP-dependent pathway.27 However, the loss of the endothelium-depen-
dent relaxation may also be caused by alterations of the cryostored HFAs endothelial cells. An incomplete preservation of the vascular smooth muscle cell response and of the endothelial cell function have been previously evoked to explain the reduction in the endothelium-dependent relaxant responses to histamine, thrombin, and substance P after cryopreservation of human coronary arteries.15 In addition, a reduction in the endothelium-dependent relaxant response to substance P and bradykinin has been
JOURNAL OF VASCULAR SURGERY Volume 28, Number 2
reported after cryopreservation of human mesenteric and coronary arteries.26 However, retention of endothelium-dependent relaxant responses and endothelial release of prostacyclin have been demonstrated following cryopreservation in human28 and canine23 internal mammary arteries, in canine coronary arteries,20 and in human saphenous veins.5,29 Further experiments using other endothelium-dependent relaxants (substance P, bradykinin) and a histologic examination are required to better evaluate the cryoinjury of the endothelium of HFAs. Therefore, the cryopreservation method used in this report provided a limited preservation of HFAs contractility, a good preservation of the endothelium-independent relaxation in response to verapamil and forskolin, a reduced preservation of the SNPinduced relaxation, and no apparent preservation of the endothelium-dependent relaxation to acetylcholine. The vehicle solution for cryoconservation contained DMSO (1.8 mmol/L) and a high-molecular-weight polymer, hydroxyethylstarch, as cryoprotecting agents. Hydroxyethylstarch is the colloidalosmotic support of the University of Wisconsin solution for organ preservation and exerts a potent cryoprotective effect.30 In most studies on the cryopreservation of blood vessels, DMSO was used at concentrations in the narrow range of 11% to 16%, ie, 1.4 to 2.0 mmol/L, and the equilibration times of the tissues in DMSO varied from 10 to 60 minutes.5,7,15,20,29 With respect to the equilibration times in DMSO, optimal post-thaw functional recovery in terms of contractions and endothelium-independent relaxations was obtained with canine femoral arteries that had been frozen within 10 to 20 minutes of being placed in the cryomedium.21 Retention of endothelium function in human saphenous vein5,29 and canine coronary arteries20 following cryoconservation has been demonstrated after equilibration times in DMSO of 10 or 30 minutes. Moreover, the measure of the permeation kinetics of DMSO in the rabbit carotid artery has shown that the times required for 95% and 99% equilibration at 2°C were 18 and 32 minutes, respectively.31 Therefore, the equilibration time of 30 minutes in DMSO 1.8 mmol/L in this study does not appear to be the cause of the weak post-thaw functional recovery. However, we have not assessed the effect of 30 minute exposure to Elohes solution on the endothelial and smooth muscle functions, and, therefore, we cannot exclude the possibility that this short-term exposure before cryopreservation may have influenced the post-thaw functional recovery. In addition to using the appropriate concentration of DMSO
Stanke et al
281
and time to equilibration, an optimal rate of cooling is also important. In a recent study, a cooling rate of 0.69°C/min was associated with a maximal postthaw recovery of the contractile response and of both the endothelium-dependent and endothelium-independent relaxant responses of the rabbit carotid arteries, whereas a significant reduction of the postthaw recovery was observed for a cooling rate of 2.2°C/min.32 Thus, a slower cooling rate for HFAs (0.69°C/min instead of 1.6°C/min) may provide better preservation of endothelial and smooth cell functions. In other respects, the duration of storage did not appear to influence the functional recovery of cryopreserved femoral arteries, because no statistical difference was found between HFAs cryopreserved for 7 days and those cryopreserved for 30 days. The thawing step is generally performed by placing the storage container in a 37°C to 42°C water bath for 2 minutes, and then removing DMSO by rinsing the blood vessels with appropriate solutions. The removal of DMSO by a stepwise dilution protocol to avoid damage by osmotic shock has not been shown to improve the post-thaw functional recovery of blood vessels.19 The thawing procedure applied to the HFAs was in line with the standard procedure applied to blood vessels and was therefore probably not responsible for the weak post-thaw functional recovery of the cryopreserved HFAs. However, it was reported recently that fractures in cryopreserved arteries, which are a major barrier to surgical use, can be prevented by warming slowly (< 50°C/min) until –100°C is reached,33,34 suggesting the interest of a more controlled thawing process. Although the occurrence of fractures was studied, inspection by naked eye did not identify the presence of fracture in the 3 cm long segments of frozen/thawed HFAs. In addition to the freezing/thawing process by itself, tissue handling during surgery and time of cold storage before cryopreservation may influence tissue viability and post-thaw recovery. Short-term storage (1 hour) of human saphenous veins in University of Wisconsin solution (Belzer solution) has been reported to markedly attenuate endothelium-dependent relaxation to acetylcholine and to increase response to vasoconstrictors.35 The attenuation of the relaxation to acetylcholine was attributed to high potassium content of the Belzer solution, which may cause functional and morphologic damages to vascular endothelium.35 Sharp relaxations to acetylcholine were observed in all the HFAs rings tested after storage in Belzer solution, but alteration of the endothelium cannot be ruled out in absence of control relaxations to acetylcholine before storage in Belzer solu-
282 Stanke et al
tion. In addition, storage of rat aortas or canine coronary arteries at 4°C in University of Wisconsin solution for up to 24 hours has been shown to preserve endothelial and vascular-smooth muscle functions.36,37 However, cold storage for 48 hours in a modified Ringer solution was associated with a decrease of the contractile force of human mammary arteries to KCl and norepinephrine.16 In the present study, the delay between removal from the body and cryostorage varied from 4 to 36 hours (median, 17.6 hours); however, only 1 HFAs was stored at +4°C for more than 24 hours (ie, 36 hours). Furthermore, we did not observe a relationship between the time of cold anoxia in Belzer solution and functional recovery of the cryopreserved arteries. In addition, a preliminary study conducted on human internal mammary arteries has shown that the contractile responses to norepinephrine and U46619 were similar or even increased in segments stored in Belzer solution at +4°C for 24 hours instead of 2 hours (Emax Norepinephrine: 4.3 ± 0.8 g [24 hours in Belzer, n = 4] vs. 3.6 g [2 hours in Belzer, n = 4]; Emax U46619: 7.0 ± 1.1 g [24 hours in Belzer, n = 4] vs. 4.8 ± 0.8 g [2 hours in Belzer, n = 4]). These data are in agreement with the results reported in human saphenous veins by Anastasiou et al.35 Therefore, the reduction of the post-thaw functional recovery of cryopreserved HFAs appears to be more related to the freezing/thawing process than to the time of cold storage in Belzer solution. In conclusion, allogenic human femoral arteries banking using cryopreservation techniques has enabled the availability of allografts. Despite certain problems, such as reduced contractile forces and loss of overall endothelial cell function, cryopreserved human femoral arteries remain useful grafts for limb salvage when autologous blood vessels are not suitable.2 It is possible that further refinements of the present cryostorage methodology, such as a slower rate of cooling and a more controlled stepwise addition of DMSO32 or a slower rate of thawing, might allow better post-thaw functional recovery of HFAs. However, it is not known whether optimum conditions of cryopreservation of endothelium and smooth muscle functions would affect the long-term patency of cryopreserved HFAs allografts. We thank Mark Hunt for helping to translate the manuscript. REFERENCES 1. Leseche G, Penna C, Bouttier S, Joubert S, Andreassian B. Femorodistal bypass using cryopreserved venous allografts for limb salvage. Ann Vasc Surg 1997;11:230-6.
JOURNAL OF VASCULAR SURGERY August 1998
2. Gournier JP, Favre JP, Gay JL, Barral X. Cryopreserved arterial allografts for limb salvage in the absence of suitable saphenous vein: Two-year results in 20 cases. Ann Vasc Surg 1995;9 Suppl:S7-S14. 3. Wagstaff SA, Grigg MJ. Arterial homografts—a possible solution to an infective dilemma. Cardiovasc Surg 1996;4:796-800. 4. Müller-Schweinitzer E, Tapparelli C, Victorzon M. Functional studies on human veins after storage at –190°C. Br J Pharmacol 1986;88:685-7. 5. Louagie YA, Legrand-Monsieur A, Lavenne-Pardonge E, Remacle C, Delvaux P, Maldague P, et al. Viability of long term cryopreserved human saphenous veins. J Cardiovasc Surg 1990;31:92-100. 6. Müller-Schweinitzer E, Ellis P. Sucrose promotes the functional activity of blood vessels after cryopreservation in DMSO containing fetal calf serum. Naunyn Schmiedebergs Arch Pharmacol 1992;345:594-7. 7. Schilling A, Glusa E, Müller-Schweinitzer E. Nature of the vehicle solution for cryopreservation of human peripheral veins. Preservation of reactivity to pharmacological stimuli. Cryobiology 1995;32:109-13. 8. Grohs JG, Kadletz M, Wodratzka M, Wolner E, Raberger G. Contractile function of human veins after long-term storage in different media. J Cardiovasc Pharmacol 1996;28:89-93. 9. Mulvany MJ, Halpern W. Contractile properties of small arterial resistance vessels in spontaneously hypertensive and normotensive rats. Circ Res 1977;41:19-25. 10. He GW, Rosenfeld FL, Buxton BF, Angus JA. Reactivity of human isolated internal mammary artery to constrictor and dilator agents. Circulation 1989;80 Suppl I:I141-I150. 11. Furchgott RF, Zawadzki JV. The obligatory role of the endothelial cells in the relaxation of arterial smooth muscle by acetylcholine. Nature 1980;288:373-6. 12. Gournier JP, Adham M, Favre JP, Raba M, Bancel B, Lepetit JC, et al. Cryopreserved arterial homografts: Preliminary study. Ann Vasc Surg 1993;7:503-11. 13. Adham M, Gournier JP, Favre JP, De La Roche E, Ducerf C, Baulieux J, et al. Mechanical characteristics of fresh and frozen human descending thoracic aorta. J Surg Res 1996;64:32-4. 14. Rosset E, Friggi A, Novakovitch G, Rolland PH, Rieu R, Pellissier JF, et al. Effects of cryopreservation on the viscoelastic properties of human arteries. Ann Vasc Surg 1996;10:262-72. 15. Ku DD, Winn MJ, Grigsby T, Caulfield JB. Human coronary vascular smooth muscle and endothelium-dependent responses after storage at –75°C. Cryobiology 1992;29:199-209. 16. Nataf P, Hadjiisky P, Lechat P, Mougenot N, Peuchmaurd M, Gouezo R, et al. Effect of cold anoxia on cryopreservation on metabolic and contractile functions of human internal mammary artery. Cryobiology 1995;32:327-33. 17. Müller-Schweinitzer E. Cryopreservation: A useful technique for storing tissues for pharmacological investigations. Trends Pharmacol Sci 1988;9:221-3. 18. Ellis P, Müller-Schweinitzer E. Maintenance of functional activity of human pulmonary arteries after cryopreservation. Br J Pharmacol 1991;103:1377-80. 19. Müller-Schweinitzer E. Applications for cryopreserved blood vessels in pharmacological research. Cryobiology 1994; 31:57-62. 20. Ku DD, Willis WL, Caulfield JB. Retention of endotheliumdependent vasodilatory responses in canine coronary arteries following cryopreservation. Cryobiology 1990;27:511-20. 21. Müller-Schweinitzer E. Arterial smooth muscle function after prolonged exposure to a medium containing dimethyl sul-
JOURNAL OF VASCULAR SURGERY Volume 28, Number 2
22.
23.
24.
25.
26.
27.
28.
29.
foxide (Me2SO) and storage at –196°C. Cryobiology 1994; 31:330-5. Schoeffter P, Müller-Schweinitzer E. The preservation of functional activity of smooth muscle and endothelium in pig coronary arteries after storage at –190°C. J Pharm Pharmacol 1990;42:646-51. Almassi GH, Farahbakhsh B, Wooldridge T, Rusch NJ, Olinger GN. Endothelium and vascular smooth muscle function in internal mammary artery after cryopreservation. J Surg Res 1996;60:355-60. Ebeigbe A, Müller-Schweinitzer E, Vogel A. Effect of calcium channel blockade in canine saphenous veins after storage at –190°C. Br J Pharmacol 1988;94:381-8. Elmore JR, Gloviczki P, Brockbank KGM, Miller VM. Cryopreservation affects endothelial and smooth muscle function of canine autogenous saphenous vein grafts. J Vasc Surg 1991;13:584-92. Müller-Schweinitzer E, Mihatsch MJ, Schilling M, Haefeli WE. Functional recovery of human mesenteric and coronary arteries after cryopreservation at –196°C in a serum-free medium. J Vasc Surg 1997;25:743-50. Schoeffter P, Dion R, Godfraind T. Modulatory role of the vascular endothelium in the contractility of human isolated internal mammary artery. Br J Pharmacol 1988;95:531-43. Pompilio G, Polvani GL, Antona C, Rossoni G, Guarino A, Porqueddu M, et al. Retention of endothelium-dependent properties in human mammary arteries after cryopreservation. Ann Thorac Surg 1996;61:667-73. Passani SL, Angelini GD, Breckenridge IM, Newby AC. Endothelial function can be preserved during cryo-storage of human saphenous vein. Eur J Cardio-Thorac Surg 1988; 2:233-6.
Stanke et al
283
30. Neveux N, De Bandt JP, Charrueau C, Savier E, Chaumeil JC, Hannoun L, et al. Deletion of hydroxyethylstarch from University of Wisconsin solution induces cell shrinkage and proteolysis during and after cold storage of rat liver. Hepatology 1997;25:678-82. 31. Bateson EAJ, Busza AL, Pegg DE, Taylor MJ. Permeation of rabbit common carotid arteries with dimethyl sulfoxide. Cryobiology 1994;31:393-7. 32. Song YC, Pegg DE, Hunt CJ. Cryopreservation of the common carotid artery of the rabbit: Optimization of dimethyl sulfoxide concentration and cooling rate. Cryobiology 1995;32:405-21. 33. Pegg DE, Wustman MC, Boylan S. Fractures in cryopreserved elastic arteries. Cryobiology 1997;34:183-92. 34. Hunt CJ, Song YC, Bateson EAJ, Pegg DE. Fractures in cryopreserved arteries. Cryobiology 1994;3:506-15. 35. Anastasiou N, Allen S, Paniagua R, Chester A, Yacoub M. Altered endothelial and smooth muscle cell reactivity caused by University of Wisconsin preservation solution in human saphenous vein. J Vasc Surg 1997;25:713-21. 36. Abebe W, Cavallari N, Agrawal DK, Rowley J, Thorpe PE, Hunter WJ, et al. Functional and morphological assessment of rat aorta stored in University of Wisconsin and Eurocollins solutions. Transplantation 1993;56:808-16. 37. Sorajja P, Cable DG, Schaff HV. Hypothermic storage with University of Wisconsin solution preserves endothelial and vascular smooth-muscle function. Circulation 1997;96 Suppl II:II297-II303.
Submitted Feb 13, 1998; accepted May 20, 1998.
INTRODUCTION Cryopreservation of tissue at –70°C to –196°C offers the prospect of virtually indefinite storage. Improvements to the preservation techniques, such From the Laboratory of Pharmacology, PCEBM, Faculté de Médecine de Grenoble, and the Blood Transfusion Center (Riebel and Dr Egelhoffer), and the Department of Vascular Surgery (Dr Magne), Centre Hospitalier Universitaire Grenoble. Supported by a grant from the Commission de la Recherche Clinique du Centre Hospitalier de Grenoble, France. Reprint requests: Françoise Stanke, Laboratoire de Pharmacologie, PCEBM, Faculté de Médecine, 38706 La Tronche Cedex, France. Copyright © 1998 by The Society for Vascular Surgery and Internation Society for Cardiovascular Surgery, North American Chapter. 0741-5214/98/$5.00 + 0 24/1/91875
as the introduction of cryoprotectant agents and a controlled rate of freezing, have enabled the development of human blood vessel banks. Cryopreserved allogenic saphenous veins1 and human femoral arteries (HFAs)2,3 are used for the bypass of diseased peripheral arteries and the treatment of limb-threatening ischemia or infected grafts. The reactivity of human saphenous veins has been extensively studied on both unfrozen and cryopreserved tissues.4–8 However, no data exists concerning the in vitro reactivity of HFAs, whereas the arteries appear considerably more susceptible to cryoinjury than the veins.6 The responses to various contracting and relaxing agents of both unfrozen HFAs and HFAs cryopreserved for 7 days and 30 days was assessed. The harvesting, freezing, and thawing procedures were those routinely applied to 273
274 Stanke et al
HFAs that are to be transplanted as allografts. The HFAs were stored at –141°C in Elohes solution containing dimethyl sulfoxide (DMSO). MATERIALS AND METHODS Human femoral arteries preparation and storage method. Samples of HFAs were carefully harvested from 9 multiorgan donors after permission was obtained from the local ethical committee. The tissues were immediately placed in Belzer solution at +4°C for 4 to 36 hours. The Belzer solution (ViaSpan) had the following composition (mmol/L): hydroxyethylstarch (5.0), lactobionic acid (100.0), KHPO3 (28.5), MgSO4, 7 H2O (5.0), raffinose pentahydrate (30.0), adenosine (5.0), allopurinol (1.0), total glutathione (3.0), KOH q.s., NaOH adjusted to pH 7.4. The femoral arteries were then cut into 3 equal segments approximately 3 cm long and were randomly divided into 3 groups. The control group (D0) consisted of fresh, “unfrozen” HFAs that were transferred to the laboratory for organ bath studies. The 2 cryopreserved groups were stored at –141°C in gas nitrogen for 7 days (D7) or 30 days (D30). In both cryostored groups, individual arterial segments were first transferred in Elohes solution (hydroxyethylstarch 200,000 [0.3 mmol/L], sodium chloride [15.4 mmol/L]) and then progressively loaded with DMSO in 3 concentration steps (4%, 9%, and 14% [weight/weight]), each lasting 2 minutes. The segments were stored in separated cryotubes containing 2 ml of the Elohes solution containing DMSO 14% (1.8 mmol/L) and were allowed to equilibrate at 4°C for approximately 30 minutes before their transfer to an automatic freezing machine (minidigitcool IMV, Bicef, France). The mean cooling rate was about 1.6°C/min, until –141°C was reached. After the predetermined period of cryostorage of 7 days or 30 days, individual blood vessel segments were quickly thawed by means of total immersion of the cryotubes in a water bath (41°C) for 2 minutes. Organ-bath technique. Immediately after thawing, the arterial segments were precisely cut into rings 5 mm long. The number of rings obtained from the femoral artery segments varied from 4 to 6. The rings were placed in a 10 mL organ bath filled with KrebsHenseleit solution (mmol/L: NaCl 118, KCl 4.7, CaCl2 2.5, MgSO4 1, KH2PO4 1, glucose 11, NaHCO3 25), which was continuously bubbled with a gas mixture (95% O2/5% CO2) and maintained at 37°C. The arterial rings were suspended on wires; the lower wire was fixed to a micrometer (Mitutoyo, Japan), and the upper was attached to a force trans-
JOURNAL OF VASCULAR SURGERY August 1998
ducer (UF-1, Pioden, U.K.) through which changes in isometric force were displayed on recorders (Linseis 200, Bioblock, France). After being cautiously mounted on the wires to preserve the endothelium, the rings were allowed to equilibrate unstretched for 15 to 30 minutes. A normalization procedure was then performed to set the vascular rings under in vitro stress comparable to the in vivo pressure, as previously reported.9,10 Briefly, the rings were progressively stretched with the micrometer at intervals of 5 g to determine the internal circumference (calculated from the distance between the wires)-tension exponential curve for each ring. When the transmural pressure on the rings reached 100 mm Hg, this stretch procedure was stopped, and the rings were then released to approximately 90 of their internal circumference at 100 mm Hg (0.9 D100). The isometric force at this setting is the “passive” or “resting” force in the absence of constrictor tone. This normalized degree of stress was then maintained throughout the experiment. Internal diameters at a pressure of 100 mm Hg and the resting forces at 0.9 D100 were recorded. PROTOCOL After the normalization procedure, the arterial rings were allowed to equilibrate for 60 minutes. The rings were then contracted with KCl 90 mmol/L, rinsed with Krebs solution, and allowed to equilibrate for 1 hour. This initial maximal contraction was performed to stabilize the preparations and to ensure reproducible contractile responsiveness. During all the equilibration periods, the Krebs solution was changed every 15 minutes. Thereafter, concentration-response curves were established by cumulative addition of increasing concentrations (0.5-log increments) of the drugs, the following concentration being added when the maximal effect had been produced by the previous concentration. Concentration-contraction curves were established for the following ranges of concentrations of contracting substances: KCl (7.5-120 mmol/L), thromboxane A2 (TXA2)-mimetic U46619 (1 nmol/L-3 µmol/L), norepinephrine (1 nmol/L-100 µmol/L), 5-hydroxytryptamine (5-HT) (1 nmol/L100 µmol/L), and endothelin-1 (1 nmol/L-0.3 µmol/L). When relaxing agents were investigated, the rings were precontracted with 30 mmol/L KCl. Endothelium-independent relaxations were assessed by establishing concentration-relaxation curves for the following ranges of concentrations of relaxing substances: sodium nitroprusside (SNP) (1 nmol/L-100 µmol/L), forskolin (1 nmol/L-30 µmol/L), and ver-
JOURNAL OF VASCULAR SURGERY Volume 28, Number 2
Stanke et al
275
apamil (1 nmol/L-30 µmol/L). Endotheliumdependent relaxation was assessed by measuring relaxation to acetylcholine (1 µmol/L)11. After the maximal effect of each drug was obtained, papaverine (300 µmol/L) was added to determine the maximum relaxation achievable. On each arterial ring, only one contracting agonist was tested, followed by investigation of one relaxant compound. DATA ANALYSIS Contractions are expressed as grams of isometric force developed by the arterial rings in response to the contracting agonists. Relaxation elicited by papaverine (300 µmol/L) was considered 100% relaxation. Endothelium-dependent relaxations and smooth muscle-dependent relaxations are expressed as percentages of the maximal relaxations achieved with papaverine. The smooth muscle-dependent relaxations were compared by determining the concentration (IC40) of agonist producing 40% of this maximal papaverine-response. Maximum effect (Emax) was the greatest response obtained with an agonist. For the vasoconstrictor agonists studied, the effective concentration of substance that caused 50% of the maximal response (EC50) was determined from each curve by a logistic, curve-fitting equation. The pD2 value represents the negative logarithm of EC50. Drug concentrations are expressed as final molar concentrations in the organ bath. Data are given as mean ± SEM. Statistical analysis was performed using one-way analysis of variance (ANOVA) for multiple comparisons, followed by Bonferroni corrected t-test and individual comparisons were made by Student’s t-test for paired or impaired data. P<.05 was considered to be significant. Drugs. TXA2 stable mimetic U46619 (9,11dihydroxy-11, 9-epoxy-methanoprostaglandin F2), sodium nitroprusside (sodium nitroferricyanide), 5hydroxytryptamine (5-hydroxytryptamine hydrochloride), papaverine, and forskolin were obtained from Sigma (L’isle d’Abeau, France); norepinephrine bitartrate was obtained from Assistance Publique des Hôpitaux de Paris (France); endothelin-1 was obtained from Euromedex (Souffelweyersheim, France); KCl Normapur was obtained from Prolabo (Paris, France); verapamil was obtained from Knoll (Levallois Perret, France); Elohes solution was obtained from Fresenius (Louviers, France); and Belzer solution (ViaSpan) was obtained from Dupont (Herts, Netherlands). All drugs were dissolved in distilled water, except forskolin, for which a stock solution at 100 µmol/L
Fig. 1. Resting force-internal diameter relationships for rings of fresh human femoral arteries (solid circle, n = 12) and of human femoral arteries cryopreserved for 7 days (open circle, n = 12) and for 30 days (open box, n = 12).
was prepared in DMSO. Drugs were kept at –20°C and were freshly dissolved in distilled water to the appropriate final concentrations. RESULTS Resting vessel parameters of the femoral arteries. One hundred eleven femoral arterial rings taken from 9 multiorgan donors were used. The mean internal diameters at an equivalent transmural pressure of 100 mm Hg were 5.3 ± 0.5 mm (D0), 5.1 ± 0.3 mm (D7), and 4.5 ± 0.4 mm (D30), and the resting forces at 0.9 D100 were 17.0 ± 1.0 g (D0), 16.8 ± 0.6 g (D7), and 15.8 ± 0.7 g (D30). No statistically significant difference was found between the resting vessel parameters of unfrozen and cryopreserved arteries. The relations between resting force and vessel internal diameter are shown in Fig. 1. Contractile responses. The contractile force as assessed by the maximal response in grams of the HFAs rings was markedly reduced in preparations from frozen/thawed arteries. In fresh HFAs, the maximal contractile response was diminished up to 34% for KCl (Fig. 2, A), up to 23% for U46619 (Fig. 2, B), up to 7% for norepinephrine (Fig. 2, C), up to 13% for 5-HT (Fig. 2, D), and up to 23% for endothelin-1 (Fig. 2, E). Nevertheless, KCl, endothelin-1, and U46619 (at day 30) stimulated the fresh and frozen/thawed arteries with similar
276 Stanke et al
JOURNAL OF VASCULAR SURGERY August 1998
Fig. 2. Cumulative concentration-contractile response curves on rings from unfrozen (solid circle) and frozen/thawed for 7 days (open circle) or 30 days (solid triangle) human femoral arteries in response to KCl (A), U46619 (B), norepinephrine (C), 5-HT (D), and endothelin1 (E). Results are expressed as gram force developed by ring vessels. The vertical bars represent mean ± SEM of 4 to 7 experiments.
JOURNAL OF VASCULAR SURGERY Volume 28, Number 2
Stanke et al
277
Table I. Responses to contractile agents of unfrozen controls and human femoral arteries cryopreserved for 7 (D7) or 30 days (D30) Unfrozen D0 U46619 KCl Norepinephrine 5-HT Endothelin-1
pD2 Emax pD2 Emax pD2 Emax pD2 Emax pD2 Emax
(g) (g) (g) (g) (g)
8.7 24.2 1.9 14.1 6.1 22.5 7.5 17.8 7.8 21.3
± ± ± ± ± ± ± ± ± ±
0.4 5.0 0.2 3.0 0.1 4.2 0.3 4.5 0.2 2.2
Frozen/Thawed D7 7.7 12.7 1.7 3.3 5.7 3.6 6.4 3.2 7.5 4.9
± ± ± ± ± ± ± ± ± ±
0.4* 1.4* 0.0 0.8* 0.1* 1.6* 0.8* 1.1* 0.4 2.0*
D30 8.2 8.3 1.6 4.6 5.3 1.6 5.4 2.3 7.6 5.9
± ± ± ± ± ± ± ± ± ±
0.4 1.3* 0.1 2.4* 0.4* 0.5* 0.2* 0.7* 0.4 0.4*
Emax = maximal contraction expressed as gram isometric force developed by the rings. pD2 = –logEC50, negative logarithm of the molar concentration producing 50% of maximal response. Data are given as means ± SEM of 4 to 7 experiments. *Significant differences (P<.05) against values determined in unfrozen controls. No significant difference was observed between HFAs cryostored for 7 or 30 days.
potencies. However, the apparent pD2 values were significantly reduced at day 7 for U46619 and at both day 7 and day 30 for norepinephrine and 5-HT (Table I). In terms of maximal contraction and potency for all these contracting agents, there was no significant difference between HFAs cryopreserved for 7 days or those cryopreserved for 30 days. Endothelium-independent relaxant responses. Relaxant responses to various agents, known to act at the vascular smooth muscle through different mechanisms, were investigated on fresh and frozen/thawed HFAs rings that had been precontracted with KCl 30 mmol/L. In preliminary experiments, it was observed that the maximal relaxation induced by papaverine (300 µmol/L) was unchanged in fresh and frozen/thawed arteries. The magnitude of the relaxations induced by papaverine, expressed as a percentage of the active tone induced by KCl, was 148 ± 12% (n = 16) in D0, 182 ± 16% (n = 14) in D7, and 184 ± 28% (n = 12) in D30. Therefore, the responses to forskolin, verapamil, and SNP in fresh and frozen/thawed arteries were expressed as a percentage of the maximal relaxation elicited by papaverine. Verapamil (Fig. 3, A) and forskolin (Fig. 3, B) elicited similar relaxant responses from both fresh and cryopreserved arteries. The –log IC40 values and maximal responses were not statistically different in fresh and frozen/thawed femoral arteries (Table II). In contrast, marked differences of both potency (expressed as –log IC40) and efficacy between fresh and frozen/thawed arteries were observed when SNP-induced relaxation was investigated (Fig.3, C; Table II). In terms of maximal relaxation and potency for all these relaxant agents, there was no signifi-
cant difference between cryopreserved HFAs for 7 or 30 days. Endothelium-dependent relaxant responses. Endothelium-dependent relaxations were investigated on fresh and frozen/thawed HFAs precontracted with 30 mmol/L of KCl. Acetylcholine (1 µmol/L) induced a relaxant response in fresh HFAs that reached 40 ± 11% (n = 8) of the maximal relaxation elicited by papaverine, whereas arteries cryopreserved for 7 or 30 days failed to relax in response to acetylcholine. Duration of storage at +4°C and recovery after cryopreservation. The duration of storage at +4°C in Belzer solution of the HFAs before cryopreservation varied from 4 to 36 hours. The relationship between the time of cold storage and the postthaw functional recovery has been evaluated in terms of both recovery of contractile response to the different vasoconstrictors tested (Fig. 4, A) and recovery of endothelium-independent relaxant response to SNP (Fig. 4, B and C). No relationship was found between the time of cold storage in Belzer and the reduction in these 2 types of response in the frozen/thawed femoral arteries. DISCUSSION Cryopreserved human arteries are used for the bypass of diseased peripheral arteries, and cryopreserved femoral arteries have been used for the treatment of limb-threatening ischemia2 or infection of arterial bypass grafts.3 It was therefore important to evaluate the functional alteration resulting from cryostorage. Our study is the first to compare the contractile responses as well as the endotheliumdependent and endothelium-independent relaxant
JOURNAL OF VASCULAR SURGERY August 1998
278 Stanke et al
Table II. Responses to relaxant agents of unfrozen controls and human femoral arteries cryopreserved for 7 (D7) or 30 days (D30) Unfrozen D0 SNP Forskolin Verapamil
–log IC40 Emax –log IC40 Emax –log IC40 Emax
8.1 93.5 6.1 75.3 6.8 75.4
± ± ± ± ± ±
0.1 1.9 0.4 10.5 0.2 7.3
Frozen/Thawed D7 6.3 79.3 5.3 73.1 6.9 91.9
± ± ± ± ± ±
D30 0.4* 4.6* 0.2 5.7 0.2 2.8
5.4 65.0 5.3 65.6 6.5 83.8
± ± ± ± ± ±
0.4* 6.5* 0.2 9.7 0.3 4.8
Emax = maximal relaxation expressed as percentage of maximal response to papaverine 300 µM. IC40 are expressed as the molar concentration producing 40% of relaxation. Data are given as means ± SEM of 4 to 7 experiments. *Significant differences (P<.05) against values determined in unfrozen controls. Statistical analysis of the IC40 was performed on the –log IC40. No significant difference was observed between HFAs cryostored for 7 or 30 days.
responses of both unfrozen and cryopreserved HFAs. The resting vessel parameters (internal diameter at an equivalent transmural pressure of 100 mm Hg and resting force) were similar in unfrozen and frozen/thawed HFAs, allowing a comparison of the responses with vasoactive agents at a comparable basis. In addition, the resting force-internal diameter relationship was the same for both frozen and cryopreserved HFAs, suggesting that the passive mechanical properties of the HFAs were not altered by cryopreservation. This assumption is in agreement with the results of previous studies in which mechanical testing with a traction-compression apparatus12,13 and viscoelastic properties evaluation with a mock circulation loop14 showed that cryopreservation does not alter the mechanical and viscoelastic characteristics of human descending thoracic aorta12,13 and superficial femoral artery14. Cryopreservation was associated with a marked reduction of the contractility. The maximal contractile responses were in the range of 7% to 34% of the responses in unfrozen HFAs with the different receptor-mediated stimuli and also with KCl, which provokes a contraction strictly dependent on Ca2+influx through voltage-operated channels. A reduction of the post-thaw contractility to less than 40% of fresh tissue has been observed in human coronary,15 internal mammary,16 and postmortem basilar arteries,17 whereas the maximal response to norepinephrine was only reduced to 75% in human pulmonary arteries.18 Freezing of vascular preparations generally results in a loss of contractile responsiveness.17,19 A post-thaw contractile force of about 60% to 80% of the maximum achieved in fresh tissue has been obtained following cryopreservation of animal arteries,19–21 although post-thaw recoveries of less than 50% have also been reported.22,23 In addition, Ku et al.15,20 have shown that an identical cryo-
preservation procedure applied to canine and human coronary arteries resulted in a greater loss of contractility in human arteries, indicating vessel to vessel variability in the susceptibility to cryoinjury. In contrast with the considerable reductions of the maximal contractile responses, potencies in terms of apparent pD2 values for KCl and endothelin-1 were not altered, and values for U46619, norepinephrine, and 5-HT were only moderately diminished after the cryopreservation procedure. These results are in line with previous studies showing that the sensitivities of human arteries and veins and animal blood vessels for various contracting agonists were well-maintained after cryopreservation.4,16,18,20,21,23–25 In addition, the endothelium-independent relaxant responses appear weakly altered by the cryopreservation procedure. The relaxations induced by forskolin, which acts through direct activation of adenylate cyclase, and by verapamil, which acts through calcium channel blockade, were similar in fresh and cryopreserved HFAs. However, the cryopreserved HFAs were less sensitive than fresh HFAs to the relaxant effect of SNP, which acts through direct activation of guanylate cyclase. A decrease in responsiveness to nitric oxide was previously reported in canine saphenous graft after cryopreservation.25 Moreover, cryopreserved human coronary arteries were recently reported to be significantly less sensitive to SNP than unfrozen tissues, whereas a slight, although nonsignificant, supersensitivity to SNP was observed in cryopreserved human mesenteric arteries.26 These results suggest that the cyclic guanosine monophosphate (cGMP)-dependent pathway of relaxation is susceptible to alteration by the cryopreservation process and that the development of supersensitivity or hyposensitivity to nitrovasodilators is tissue-specific. After cryoconservation, a weaker sensitivity has also been reported for potassium chan-
JOURNAL OF VASCULAR SURGERY Volume 28, Number 2
Stanke et al
279
Fig. 3. Cumulative concentration-relaxant response curves on rings from unfrozen human femoral arteries (solid circle) and human femoral arteries frozen/thawed for 7 days (open circle) or 30 days (solid triangle) for verapamil (A), forskolin (B), sodium nitroprusside (C). Results are expressed as percentages of the maximal relaxation induced by papaverine 3.10-4 M. The vertical bars represent mean ± SEM of 4 to 6 experiments.
nel activators in human pulmonary arteries18 and for isoproterenol in human coronary arteries.15 Again, an identical cryopreservation procedure applied to canine and human coronary arteries resulted in a reduction of the relaxant response to isoproterenol in
human arteries, whereas the response was well-preserved in canine arteries.15,20 In contrast to the marked reduction of the maximal contractile responses, the maximal relaxant responses were not altered or were altered much less after cryoconservation.15,20–22
280 Stanke et al
JOURNAL OF VASCULAR SURGERY August 1998
Fig. 4. Effect of the duration of storage at +4°C in Belzer solution on the functional recovery of human femoral arteries cryopreserved for 7 or 30 days. A, Effect of the duration of storage on the reduction of the maximal contractile response to U46619 (open circle), norepinephrine (open box), 5-HT (crossed box), KCl (solid triangle), and endothelin-1 (open triangle). B, Effect of the duration of storage on the reduction of the relaxant response to SNP as percentage of the reduction in the response to 10-5 M SNP (open circle). C, Effect of the duration of storage on the reduction of the relaxant response to SNP as log-shift to the right of the concentration-response (C/R) curve to SNP (open circle).
Cryostorage of HFAs resulted also in a loss of the endothelium-dependent relaxant response to acetylcholine. The alteration of the cGMP-dependent pathway of relaxation may, at least in part, account for this result because acetylcholine induced an endotheliumdependent relaxation by stimulating the production and release of nitric oxide from the endothelial cells, which, in turn, induces the relaxation of the vascular smooth muscle through the cGMP-dependent pathway.27 However, the loss of the endothelium-depen-
dent relaxation may also be caused by alterations of the cryostored HFAs endothelial cells. An incomplete preservation of the vascular smooth muscle cell response and of the endothelial cell function have been previously evoked to explain the reduction in the endothelium-dependent relaxant responses to histamine, thrombin, and substance P after cryopreservation of human coronary arteries.15 In addition, a reduction in the endothelium-dependent relaxant response to substance P and bradykinin has been
JOURNAL OF VASCULAR SURGERY Volume 28, Number 2
reported after cryopreservation of human mesenteric and coronary arteries.26 However, retention of endothelium-dependent relaxant responses and endothelial release of prostacyclin have been demonstrated following cryopreservation in human28 and canine23 internal mammary arteries, in canine coronary arteries,20 and in human saphenous veins.5,29 Further experiments using other endothelium-dependent relaxants (substance P, bradykinin) and a histologic examination are required to better evaluate the cryoinjury of the endothelium of HFAs. Therefore, the cryopreservation method used in this report provided a limited preservation of HFAs contractility, a good preservation of the endothelium-independent relaxation in response to verapamil and forskolin, a reduced preservation of the SNPinduced relaxation, and no apparent preservation of the endothelium-dependent relaxation to acetylcholine. The vehicle solution for cryoconservation contained DMSO (1.8 mmol/L) and a high-molecular-weight polymer, hydroxyethylstarch, as cryoprotecting agents. Hydroxyethylstarch is the colloidalosmotic support of the University of Wisconsin solution for organ preservation and exerts a potent cryoprotective effect.30 In most studies on the cryopreservation of blood vessels, DMSO was used at concentrations in the narrow range of 11% to 16%, ie, 1.4 to 2.0 mmol/L, and the equilibration times of the tissues in DMSO varied from 10 to 60 minutes.5,7,15,20,29 With respect to the equilibration times in DMSO, optimal post-thaw functional recovery in terms of contractions and endothelium-independent relaxations was obtained with canine femoral arteries that had been frozen within 10 to 20 minutes of being placed in the cryomedium.21 Retention of endothelium function in human saphenous vein5,29 and canine coronary arteries20 following cryoconservation has been demonstrated after equilibration times in DMSO of 10 or 30 minutes. Moreover, the measure of the permeation kinetics of DMSO in the rabbit carotid artery has shown that the times required for 95% and 99% equilibration at 2°C were 18 and 32 minutes, respectively.31 Therefore, the equilibration time of 30 minutes in DMSO 1.8 mmol/L in this study does not appear to be the cause of the weak post-thaw functional recovery. However, we have not assessed the effect of 30 minute exposure to Elohes solution on the endothelial and smooth muscle functions, and, therefore, we cannot exclude the possibility that this short-term exposure before cryopreservation may have influenced the post-thaw functional recovery. In addition to using the appropriate concentration of DMSO
Stanke et al
281
and time to equilibration, an optimal rate of cooling is also important. In a recent study, a cooling rate of 0.69°C/min was associated with a maximal postthaw recovery of the contractile response and of both the endothelium-dependent and endothelium-independent relaxant responses of the rabbit carotid arteries, whereas a significant reduction of the postthaw recovery was observed for a cooling rate of 2.2°C/min.32 Thus, a slower cooling rate for HFAs (0.69°C/min instead of 1.6°C/min) may provide better preservation of endothelial and smooth cell functions. In other respects, the duration of storage did not appear to influence the functional recovery of cryopreserved femoral arteries, because no statistical difference was found between HFAs cryopreserved for 7 days and those cryopreserved for 30 days. The thawing step is generally performed by placing the storage container in a 37°C to 42°C water bath for 2 minutes, and then removing DMSO by rinsing the blood vessels with appropriate solutions. The removal of DMSO by a stepwise dilution protocol to avoid damage by osmotic shock has not been shown to improve the post-thaw functional recovery of blood vessels.19 The thawing procedure applied to the HFAs was in line with the standard procedure applied to blood vessels and was therefore probably not responsible for the weak post-thaw functional recovery of the cryopreserved HFAs. However, it was reported recently that fractures in cryopreserved arteries, which are a major barrier to surgical use, can be prevented by warming slowly (< 50°C/min) until –100°C is reached,33,34 suggesting the interest of a more controlled thawing process. Although the occurrence of fractures was studied, inspection by naked eye did not identify the presence of fracture in the 3 cm long segments of frozen/thawed HFAs. In addition to the freezing/thawing process by itself, tissue handling during surgery and time of cold storage before cryopreservation may influence tissue viability and post-thaw recovery. Short-term storage (1 hour) of human saphenous veins in University of Wisconsin solution (Belzer solution) has been reported to markedly attenuate endothelium-dependent relaxation to acetylcholine and to increase response to vasoconstrictors.35 The attenuation of the relaxation to acetylcholine was attributed to high potassium content of the Belzer solution, which may cause functional and morphologic damages to vascular endothelium.35 Sharp relaxations to acetylcholine were observed in all the HFAs rings tested after storage in Belzer solution, but alteration of the endothelium cannot be ruled out in absence of control relaxations to acetylcholine before storage in Belzer solu-
282 Stanke et al
tion. In addition, storage of rat aortas or canine coronary arteries at 4°C in University of Wisconsin solution for up to 24 hours has been shown to preserve endothelial and vascular-smooth muscle functions.36,37 However, cold storage for 48 hours in a modified Ringer solution was associated with a decrease of the contractile force of human mammary arteries to KCl and norepinephrine.16 In the present study, the delay between removal from the body and cryostorage varied from 4 to 36 hours (median, 17.6 hours); however, only 1 HFAs was stored at +4°C for more than 24 hours (ie, 36 hours). Furthermore, we did not observe a relationship between the time of cold anoxia in Belzer solution and functional recovery of the cryopreserved arteries. In addition, a preliminary study conducted on human internal mammary arteries has shown that the contractile responses to norepinephrine and U46619 were similar or even increased in segments stored in Belzer solution at +4°C for 24 hours instead of 2 hours (Emax Norepinephrine: 4.3 ± 0.8 g [24 hours in Belzer, n = 4] vs. 3.6 g [2 hours in Belzer, n = 4]; Emax U46619: 7.0 ± 1.1 g [24 hours in Belzer, n = 4] vs. 4.8 ± 0.8 g [2 hours in Belzer, n = 4]). These data are in agreement with the results reported in human saphenous veins by Anastasiou et al.35 Therefore, the reduction of the post-thaw functional recovery of cryopreserved HFAs appears to be more related to the freezing/thawing process than to the time of cold storage in Belzer solution. In conclusion, allogenic human femoral arteries banking using cryopreservation techniques has enabled the availability of allografts. Despite certain problems, such as reduced contractile forces and loss of overall endothelial cell function, cryopreserved human femoral arteries remain useful grafts for limb salvage when autologous blood vessels are not suitable.2 It is possible that further refinements of the present cryostorage methodology, such as a slower rate of cooling and a more controlled stepwise addition of DMSO32 or a slower rate of thawing, might allow better post-thaw functional recovery of HFAs. However, it is not known whether optimum conditions of cryopreservation of endothelium and smooth muscle functions would affect the long-term patency of cryopreserved HFAs allografts. We thank Mark Hunt for helping to translate the manuscript. REFERENCES 1. Leseche G, Penna C, Bouttier S, Joubert S, Andreassian B. Femorodistal bypass using cryopreserved venous allografts for limb salvage. Ann Vasc Surg 1997;11:230-6.
JOURNAL OF VASCULAR SURGERY August 1998
2. Gournier JP, Favre JP, Gay JL, Barral X. Cryopreserved arterial allografts for limb salvage in the absence of suitable saphenous vein: Two-year results in 20 cases. Ann Vasc Surg 1995;9 Suppl:S7-S14. 3. Wagstaff SA, Grigg MJ. Arterial homografts—a possible solution to an infective dilemma. Cardiovasc Surg 1996;4:796-800. 4. Müller-Schweinitzer E, Tapparelli C, Victorzon M. Functional studies on human veins after storage at –190°C. Br J Pharmacol 1986;88:685-7. 5. Louagie YA, Legrand-Monsieur A, Lavenne-Pardonge E, Remacle C, Delvaux P, Maldague P, et al. Viability of long term cryopreserved human saphenous veins. J Cardiovasc Surg 1990;31:92-100. 6. Müller-Schweinitzer E, Ellis P. Sucrose promotes the functional activity of blood vessels after cryopreservation in DMSO containing fetal calf serum. Naunyn Schmiedebergs Arch Pharmacol 1992;345:594-7. 7. Schilling A, Glusa E, Müller-Schweinitzer E. Nature of the vehicle solution for cryopreservation of human peripheral veins. Preservation of reactivity to pharmacological stimuli. Cryobiology 1995;32:109-13. 8. Grohs JG, Kadletz M, Wodratzka M, Wolner E, Raberger G. Contractile function of human veins after long-term storage in different media. J Cardiovasc Pharmacol 1996;28:89-93. 9. Mulvany MJ, Halpern W. Contractile properties of small arterial resistance vessels in spontaneously hypertensive and normotensive rats. Circ Res 1977;41:19-25. 10. He GW, Rosenfeld FL, Buxton BF, Angus JA. Reactivity of human isolated internal mammary artery to constrictor and dilator agents. Circulation 1989;80 Suppl I:I141-I150. 11. Furchgott RF, Zawadzki JV. The obligatory role of the endothelial cells in the relaxation of arterial smooth muscle by acetylcholine. Nature 1980;288:373-6. 12. Gournier JP, Adham M, Favre JP, Raba M, Bancel B, Lepetit JC, et al. Cryopreserved arterial homografts: Preliminary study. Ann Vasc Surg 1993;7:503-11. 13. Adham M, Gournier JP, Favre JP, De La Roche E, Ducerf C, Baulieux J, et al. Mechanical characteristics of fresh and frozen human descending thoracic aorta. J Surg Res 1996;64:32-4. 14. Rosset E, Friggi A, Novakovitch G, Rolland PH, Rieu R, Pellissier JF, et al. Effects of cryopreservation on the viscoelastic properties of human arteries. Ann Vasc Surg 1996;10:262-72. 15. Ku DD, Winn MJ, Grigsby T, Caulfield JB. Human coronary vascular smooth muscle and endothelium-dependent responses after storage at –75°C. Cryobiology 1992;29:199-209. 16. Nataf P, Hadjiisky P, Lechat P, Mougenot N, Peuchmaurd M, Gouezo R, et al. Effect of cold anoxia on cryopreservation on metabolic and contractile functions of human internal mammary artery. Cryobiology 1995;32:327-33. 17. Müller-Schweinitzer E. Cryopreservation: A useful technique for storing tissues for pharmacological investigations. Trends Pharmacol Sci 1988;9:221-3. 18. Ellis P, Müller-Schweinitzer E. Maintenance of functional activity of human pulmonary arteries after cryopreservation. Br J Pharmacol 1991;103:1377-80. 19. Müller-Schweinitzer E. Applications for cryopreserved blood vessels in pharmacological research. Cryobiology 1994; 31:57-62. 20. Ku DD, Willis WL, Caulfield JB. Retention of endotheliumdependent vasodilatory responses in canine coronary arteries following cryopreservation. Cryobiology 1990;27:511-20. 21. Müller-Schweinitzer E. Arterial smooth muscle function after prolonged exposure to a medium containing dimethyl sul-
JOURNAL OF VASCULAR SURGERY Volume 28, Number 2
22.
23.
24.
25.
26.
27.
28.
29.
foxide (Me2SO) and storage at –196°C. Cryobiology 1994; 31:330-5. Schoeffter P, Müller-Schweinitzer E. The preservation of functional activity of smooth muscle and endothelium in pig coronary arteries after storage at –190°C. J Pharm Pharmacol 1990;42:646-51. Almassi GH, Farahbakhsh B, Wooldridge T, Rusch NJ, Olinger GN. Endothelium and vascular smooth muscle function in internal mammary artery after cryopreservation. J Surg Res 1996;60:355-60. Ebeigbe A, Müller-Schweinitzer E, Vogel A. Effect of calcium channel blockade in canine saphenous veins after storage at –190°C. Br J Pharmacol 1988;94:381-8. Elmore JR, Gloviczki P, Brockbank KGM, Miller VM. Cryopreservation affects endothelial and smooth muscle function of canine autogenous saphenous vein grafts. J Vasc Surg 1991;13:584-92. Müller-Schweinitzer E, Mihatsch MJ, Schilling M, Haefeli WE. Functional recovery of human mesenteric and coronary arteries after cryopreservation at –196°C in a serum-free medium. J Vasc Surg 1997;25:743-50. Schoeffter P, Dion R, Godfraind T. Modulatory role of the vascular endothelium in the contractility of human isolated internal mammary artery. Br J Pharmacol 1988;95:531-43. Pompilio G, Polvani GL, Antona C, Rossoni G, Guarino A, Porqueddu M, et al. Retention of endothelium-dependent properties in human mammary arteries after cryopreservation. Ann Thorac Surg 1996;61:667-73. Passani SL, Angelini GD, Breckenridge IM, Newby AC. Endothelial function can be preserved during cryo-storage of human saphenous vein. Eur J Cardio-Thorac Surg 1988; 2:233-6.
Stanke et al
283
30. Neveux N, De Bandt JP, Charrueau C, Savier E, Chaumeil JC, Hannoun L, et al. Deletion of hydroxyethylstarch from University of Wisconsin solution induces cell shrinkage and proteolysis during and after cold storage of rat liver. Hepatology 1997;25:678-82. 31. Bateson EAJ, Busza AL, Pegg DE, Taylor MJ. Permeation of rabbit common carotid arteries with dimethyl sulfoxide. Cryobiology 1994;31:393-7. 32. Song YC, Pegg DE, Hunt CJ. Cryopreservation of the common carotid artery of the rabbit: Optimization of dimethyl sulfoxide concentration and cooling rate. Cryobiology 1995;32:405-21. 33. Pegg DE, Wustman MC, Boylan S. Fractures in cryopreserved elastic arteries. Cryobiology 1997;34:183-92. 34. Hunt CJ, Song YC, Bateson EAJ, Pegg DE. Fractures in cryopreserved arteries. Cryobiology 1994;3:506-15. 35. Anastasiou N, Allen S, Paniagua R, Chester A, Yacoub M. Altered endothelial and smooth muscle cell reactivity caused by University of Wisconsin preservation solution in human saphenous vein. J Vasc Surg 1997;25:713-21. 36. Abebe W, Cavallari N, Agrawal DK, Rowley J, Thorpe PE, Hunter WJ, et al. Functional and morphological assessment of rat aorta stored in University of Wisconsin and Eurocollins solutions. Transplantation 1993;56:808-16. 37. Sorajja P, Cable DG, Schaff HV. Hypothermic storage with University of Wisconsin solution preserves endothelial and vascular smooth-muscle function. Circulation 1997;96 Suppl II:II297-II303.
Submitted Feb 13, 1998; accepted May 20, 1998.