- Email: [email protected]
Comparison of saphenous vein graft relaxation between Plasma-Lyte solution and normal saline solution
Comparison of saphenous vein graft relaxation between Plasma-Lyte solution and normal saline solution Venospasm of saphenous vein grafts may damage endothelial ceUs and compromise early and late graft performance. Hence it is desirable to identify and use storage solutions that minimize vascular spasm during vein preparation. In view of this, we initiated isometric tension-recording studies in isolated canine and human saphenous vein to evaluate the acute, vasoactive effects of two storage solutions, Plasma-Lyte solution and normal saline solution. In initial experiments, canine saphenous veins were mounted in tissue baths containing physiologic salt solution and tonicaUy constricted by 2 x 10-6 mol/L norepinephrine. The physiologic salt solution in the bath was then replaced by Plasma-Lyte solution or normal saline solution containing the same norepinephrine concentration, and changes in contraction amplitude were recorded for 90 minutes. Storage in Plasma-Lyte solution at 37° C completely relaxed norepinephrine-activated canine saphenous vein within 20 minutes, whereas veins remained partiaUy constricted in normal saline solution. Both Plasma-Lyte solution and normal saline solution relaxed canine saphenous vein less at room temperature (25° C) than at 37° C, implying that warming of storage solutions in the operating room may promote graft dilation. To identify the mechanism by which Plasma-Lyte solution induced relaxation, we replaced its putative vasodilator components of gluconate and acetate- with NaO, but this alteration did not reduce relaxation induced by Plasma-Lyte solution. However, adding 1.6 mmol/L Ca02 to Plasma-Lyte solution completely reversed the venodilation, suggesting that the low Ca H content of Plasma-Lyte solution confers its relaxant action. FinaUy, we tested the vasoactive effect of Plasma-Lyte solution on human saphenous vein obtained by discard from coronary bypass operations. Plasma-Lyte solution at 37° C effectively dilated norepinephrine-activated human saphenous vein, inducing complete relaxation within 20 minutes. On this basis, we recommend the use of Plasma-Lyte solution as a venodilating storage solution during coronary bypass operations to optimize vein graft relaxation before implantation. (J THoRAe CARDIOVASC SURG 1994;107:1445-53)
Alexandra M. Sanchez,* Theresa A. Wooldridge, BS,a, b Lawrence E. Boerboom, PhD,a, b Gordon N. Olinger, MD,a G. Hossein Almassi, MD,a and Nancy J. Rusch, PhD,a, b Milwaukee, Wis.
From the Departments of Cardiothoracic Surgery" and Physiology.> Medical College of Wisconsin, Milwaukee, Wis. Supported by National Institute of Health grant HL-41840 (L.E.B.), National Institute of Health grant HL-40474 (NJ.R.), and an award from the Max Baer Heart Fund of the Fraternal Order of Eagles. Received for publication Aug. 3, 1993. Accepted for publication Oct. 25, 1993. Address for reprints: Nancy J. Rusch, PhD, Associate Professor, Department of Physiology, Medical College of Wisconsin, 8701 Watertown Plank Rd., Milwaukee, WI 53226.
* ApprenticeshipIn Research Scholar, Minority Student Affairs, Medical Collegeof Wisconsin, Milwaukee, Wis. Copyright © 1994by Mosby-Year Book, Inc. 0022-5223/94 $3.00 + 0
12/1/52620
Minimizing vascular smooth muscle and endothelial cell damage during coronary bypass operations improves the longevity of in situ vein grafts. Multiple factors may influence graft integrity during the intraoperative period, including the inherent condition of the vein, the level of surgical manipulation, and the type of solution used for venous storage.!:" With regard to storage solutions, much attention has focused on defining which solutes and temperatures provide optimal vein storage. On the basis of graft morphology, autologous whole blood, Plasma-Lyte solution, (Baxter Healthcare Corp., Deerfield, Ill.), normal saline solution (NSS), and cell culture media have been promoted as optimal solutions.>"? Storage solution temperatures from 4 0 C to 37 0 C also have been advanced 1445
I 446
Sanchez et al.
as appropriate for preserving vascular smooth muscle and endothelial cell integrity. 15-20 The design of a venous storage solution should consider both technical and physiologic factors. From the technical perspective, the solution must induce rapid and sufficient venous relaxation to permit vein implantation. From the physiologic viewpoint, the solution composition should preserve or enhance long-term graft viability. However, despite the need to evaluate the acute venodilating action of storage solutions, there is little practical information about this issue. Most studies have focused on the anatomic impact, rather than the acute vasodilating effect, of storage media. Thus whether or not common storage solutions, such as Plasma-Lyte solution and NSS, help dilate vein grafts during intraoperative storage is unknown. The goal of this study was to evaluate the vasodilating effect and mechanisms ofPlasma-Lyte solution and NSS on isolated canine saphenous veins. On the basis of this initial screening, we further tested for optimal vasodilating efficacy in human saphenous veins obtained from coronary bypass operations.
Methods Preparation of veins. Canine saphenous veins were removed from anesthetized dogs (sodium pentobarbital, 30 mg/kg intravenously) using protocols approved by National Institutes of Health guidelines and the Institutional Animal Care and Use Committee. In later experiments, segments of human saphenous vein discarded during coronary bypass operations were obtained and handled according to standards suggested by the Institutional Pathogen Committee. All veins were handled and prepared carefully to preserve vascular endothelium and smooth muscle cell function. Immediately after removal, all veins were divided into 3 to 4 mm segments and placed at the optimal point of their length-tension relationship in an isometric tensionrecording system." Standard bath solution was physiologicsalt solution (PSS) composed of (in millimoles per liter) NaCl119, KCI4.7,MgS041.l7,CaCh 1.6,NaH 2P041.l8,NaHC0324, ethylenediaminetetraacetic acid 0.026, and glucose ~.5. Experiments in canine saphenous vein. After equilibration for 30 to 60 minutes at 37° C, the temperature of the PSS in some tissue baths was reduced to 25° C, whereas other baths were maintained at 37° C. This permitted us to compare the vasoactive effect of storage solutions between canine saphenous vein at physiologic temperature (37° C) and the ambient temperature of the operating room (25° C). After temperature equilibration,2 X 10- 6 mol/L norepinephrine was added to the PSS in the tissue baths to obtain a sustained, half-maximal contraction of canine saphenous veins at 37° Cor 25° C. These norepinephrine-induced contractions were produced to mimic the venospasm observed during surgical vein removaJ.21 Subsequently, the norepinephrine-containing PSS in the tissue baths was equilibrated with room air and then replaced with new norepinephrine-containing solutions, which were (I) the same norepinephrine-containing PSS, which served as a time control, (2) Plasma-Lyte solution, or (3) NSS. Solution exchange was accomplished by emptying the tissue baths of initial PSS and
The Journal of Thoracic and Cardiovascular Surgery June 1994
then perfusing the new solutions into the baths from upper reservoirs. It should be emphasized that constant o-adrencceptor activation of canine saphenous veins was maintained throughout the recording period, inasmuch as all storage solutions had the same norepinephrine concentration (2 X 10- 6 moll L) as the initial PSS. After replacement of norepinephrine-containing PSS with norepinephrine-containing storage solutions, tension was recorded at 5-minute intervals for 30 minutes and then at 15-minute intervals for 60 to 90 minutes to assess the effect of Plasma-Lyte solution and NSS on norepinephrine-induced contractions. In one series of experiments, 10- 4 mol/L papaverine was added to the Plasma-Lyte solution or NSS to evaluate whether this vasodilator drug was able to induce further venorelaxation. In other experiments, the PSS was replaced by a modified Plasma-Lyte solution identical to the commercially available solution, but containing equimolar replacement of acetate and gluconate by NaCI or having the addition of 1.6 mmol/L CaCl 2 to its standard composition. In these instances, the pH of the modified solutions was kept constant at 7.4 with the use of NaOH as a buffer. Experiments in human saphenous vein. After the venodilating effect ofPlasma-Lyte solution and NSS was compared at 37° C and 25° C in canine saphenous veins, the dilating effect of Plasma-Lyte solution was further compared in human saphenous veins at the same temperatures. Experimental protocols were identical to methods used to evaluate relaxations in canine saphenous veins induced by Plasma-Lyte solution. Data acquisition and analysis. Changes in tension were displayed continuously on an enhanced graphics adapter color monitor and simultaneously digitized and stored on an IBMcompatible 386SX computer (Citus Corp., Menomonee Falls, Wis.). Data were acquired and analyzed with CODAS software (Dataq Instruments, Akron, Ohio). Contraction amplitude was expressed as mean ± standard error of the mean and plotted as percent of the initial contractile response to norepinephrine. Significant differences in contraction amplitude between storage solution types at matched time intervals were determined by two-way analyses of variance with repeated measures on one factor with the use of a CSS:General Manova Stats Plus Software program (Statsoft, Tulsa, Okla.). Significant differences between individual means of specificgroups were determined by Duncan's new multiple range test. A p value less than 0.05 was considered statistically significant. Solutions and drugs. The composition of solutions used in this study were (I) Plasma-Lyte solution (Plasma-Lyte A injection solution, Baxter),22 composed of (in millimoles per liter) Na+ 140, K+ 5, Mg 2+ 3, Cl- 98, acetate 27, andgluconate 23 (pH = 7.4), and (2) NSS (Baxter), composed of NaCI 154 mfiq/L (pH = 5.0). The composition of Plasma-Lyte solution was modified in some experiments as explained previously. Norepinephrine ([-]-arterenol) and papaverine were obtained from Sigma Chemical Co., St. Louis, Missouri.
Results Experiments in canine saphenous veins Time-dependent relaxations to Plasma-Lyte solution and NSS at 37° Cand 25° C. Segments of canine saphenous vein were constricted with a median effective dose of 2 X 10-6 moljL norepinephrine in PSS at 37° C, after which the bath solution was exchanged for Plasma-Lyte
The Journal of Thoracic and Cardiovascular Surgery Volume 107, Number 6
Sanchez et al.
1 447
,
PL.PAP
A
~~~ NE
B
0
:?::
z 0
..
;:::: u
z 0
40
U
w
...z 0
~
15
30 (min)
20 0
45
60
NE
0..
\_.370C : ,0-0 37°C (+PAP)
1'. ~,
10
-,
0
--
60
.. ..
.'
30 45607590 20 TIME (minutes)
~
0
15
30
30 (min)
45
I
45
60
•
60
NSS
100
.-.370C 0-0 37°C (+PAP)
z
80
;:::: u
60
T
40
~T
.
15
'"
NE
D If) If)
60
0::
l-
0
PL
100 80
45
I"""
...
C Q.
30
~dNE
If) If)
15
0
0::
I-
z
0
U
w
...z 0
il'1
20 0
·-e "
...... T
i i-i ° + -·-·~li-1'9.....+ + + + + + +
10
-
tl;l l;l Ii' iii 20 30 45607590 TIME (minutes)
Fig. 1. Recordings showing effectof Plasma-Lytesolution (PL) and NSS on caninesaphenous veins precontracted by 2 X 10- 6 mol/L norepinephrine (NE) at 37° C. A, Plasma-Lytesolution inducedrapid and completerelaxation (left trace), which was not affected by 10- 4 mol/L papaverine (PAP) (right trace). B, NSS induced only partial relaxation (left trace), whichwasenhancedby 10- 4 mol/L papaverine (right trace). Plotsin C and D showrelaxation as function of time after PSS was replacedwith Plasma-Lytesolution or NSS, respectively. Data are expressed as percent of initial norepinephrine-induced constriction and represent mean ± standard error of mean for six veins. +p < 0.05 for control versus papaverine-containing solution at same time interval. solution or NSS containing the same norepinephrine concentration. Remaining tension was measured every 5 minutes for the first 30 minutes after solution exchange and then at 15-minute intervals for 1 hour. Exposure to Plasma-Lyte solution at 37° C induced immediate relaxation, which was complete within 20 minutes (Fig. 1, A, left trace). In parallel canine saphenous vein segments, addition of 10-4 moljL papaverine to Plasma-Lyte solutiondid not enhance relaxations induced by Plasma-Lyte solution at 37° C (Fig. 1, A, right trace). The plot of contraction amplitude as a function of storage time in Plasma-Lyte solution shows that Plasma-Lyte solution is an effective venorelaxant at 37° C and indicates that the venodilating efficacy of Plasma-Lyte solution was not enhanced by papaverine (Fig. 1, C; n = 6 veins). In contrast to Plasma-Lyte solution, storage in NSS at 37° C did not induce complete relaxation of canine saphenous vein (Fig. 1, B, left trace). The summarized data show that 25% ± 5% of the initial norepinephrine-induced contraction remained after 90 minutes in NSS (Fig. 1, D;
n = 6 veins). Under these conditions of incomplete relaxation, 10-4 mol/L papaverine enhanced the venorelaxant effect ofNSS at 37° C (Fig. 1, B, right trace, and Fig. 1, D). The solution exchange procedure per se did not affect contraction amplitude, as determined by monitoring contraction during replacement of norepinephrine-containing PSS with identical norepinephrine-containing solution. A sample recording from such a control experiment is presented later in the results. Similar experiments were done at 25 ° C, a temperature level comparable to that of the operating room. At 25 ° C, Plasma-Lyte solution only partially relaxed canine saphenous vein during a 90-minute recording (Fig. 2, A, left trace) and 56% ± 1%, 29% ± 9%, and 20% ± 7% of the initial norepinephrine-induced contraction remained after 30, 60, and 90 minutes of Plasma-Lyte solution storage, respectively (Fig. 2, C; n = 6 veins). This response contrasted with the complete and rapid relaxation induced by Plasma-Lyte solution at 37° C (Fig. 1, C). This residual contraction to norepinephrine at 25° C
I 44 8
The Journal of Thoracic and Cardiovascular Surgery June 1994
Sanchez et al.
A
~OC~ ..
NE
,__,__,n,__,
0
15
30
45
60
j ..
-1.-------,=-=c=:~ NE
0
15
30
45
60
NSS.PAP
B
C==~ ...
I
I
NE
0
15
I
I
30 45 (min)
,'''
GO
NE
I
I
I
0
15
30 (min)
o o
~
u
60
~ o
40
~
20
o
o
u
u, ~
,
60
NSS
lI'I lI'I
a. 100 :i':: z 80
I
45
\..
.-.250C 0'_ 0 - 0 25°C (+PAP) +\.+..................
°'\+ -i ?,+ +
+
10
3045607590
-.~.TT
?'O
-.-w-. TIT
+ + + + ' ....9."0_ _
20
TIME (minutes)
Fig. 2. Recordings showingeffectof Plasma-Lyte solution (PL) and NSS on canine saphenousveinsprecontracted by 2 X 10- 6 mol/L norepinephrine (NE) at 25° C. A, Compared with tracings at 37° C, reiaxatiOlfinduced by Plasma-Lyte solution was slowerand less effective at 25° C (left tracing) and its effect was potentiated by 10- 4 mol/L papaverine (PAP) (right tracing). B, NSS also inducedslowerand lesserrelaxation at 25° C (left tracing), whichwas potentiated by 10- 4 mol/L papaverine (right tracing). Plots in C and D showrelaxation as function of time after PSS was replaced by Plasma-Lyte solution or NSS, respectively. Expression of data is same as in Fig. I.
was eliminated by 10-4 mOI/L papaverine (Fig. 2, A, right trace, and Fig. 2, C). Similarly, storage in NSS at 25° C relaxed norepinephrine-contracted canine saphenous vein less than storage at 37° C (Fig. 2, B, left trace), and relaxation did not progress after the initial 45 minutes ofNSS exposure (Fig. 2, D; n = 6 veins). The addition of 10-4 mol/L papaverine also reduced venoconstriction in NSS at this temperature, confirming that NSS at 25° C did not eliminate a large component of active venous con traction.
Removal ofacetate and gluconate from Plasma-Lyte solution. Because our initial experiments (Figs. I and 2) showed that optimal relaxation of canine saphenous vein occurred in Plasma-Lyte solution at 37° C, we designed experiments to identify the vasodilatory component of Plasma-Lyte solution. It has been suggested that two anions in Plasma-Lyte solution, acetate and gluconate, account for its relaxant action.V To test this hypothesis, we determined the effect of anion-modified Plasma-Lyte solutions on norepinephrine-induced constrictions at 37° C. Canine saphenous vein relaxed rapidly after exposure to standard Plasma-Lyte solution (Fig. 3, A), as shown earlier in this study at 37° C (Fig. I, A). Parallel record-
ings showed that removal of 27 mmol/L acetate (Fig. 3, B), removal of 23 mmol/L gluconate (Fig. 3, C), or removal of both anions (Fig. 3, D) from standard Plasma-Lyte solution did not inhibit the rapid relaxation induced by this solution. Surprisingly, in view of the literature, the most rapid relaxation was in canine saphenous vein exposed to Plasma-Lyte solution without acetate or gluconate (Fig. 3, E; n = 6 veins). Addition of 1.6 mmolfI. CaCl2 to Plasma-Lyte solution. In view of the absence of evidence for an anion-induced venorelaxant mechanism for Plasma-Lyte solution, we postulated that relaxation induced by Plasma-Lyte solution was due to the low ionized calcium concentration in this solution. This was a reasonable hypothesis, inasmuch as Plasma-Lyte solution contains no added calcium and canine saphenous vein relies on calcium influx for sustained contraction.P'?" To test this hypothesis, we added a physiologic level of calcium (1.6 mrnol/L CaCI 2) to standard Plasma-Lyte solution at 37° C and evaluated its effect on norepinephrine-contracted veins. As in other experiments, we documented that the solution exchange procedure per se did not affect contraction amplitude. Thus canine saphenous vein constricted by norepineph-
The Journal of Thoracic and Cardiovascular Surgery Volume 107, Number 6
Sanchez et al.
A
,
PL - Acetate
~L
~= . NE
0
15
.
.s",.....90
30 (min)
NE
o
fl_,.,..... . •
0
45
30 (min)
15
E
15
30
(min)
_ (Gluconate/Acetate)
PL - Gluconate
NE
1 449
90 NE
o
30
15
(min)
100 0-
~
80
z
0
i=
o 60 -c
It:
I-
Z 0
u
40
w
Z u,
0
. - . PL
i~.
Vl Vl
20
,\~ '\~
,,\"
T
PL - Acetate v - v PL - Gluconate .- - . PL - (Gluconate/Acetate)
0--0
,, "t"" '" '-,\~\" ,
1 ....
llt
5
-c T
~,,, -,
"
v~""i
---- ----.=_.. ___
.. ! ·0----.- - - - • ~====, 10
15
20
~
25
. 30
TIME (minutes) Fig. 3. Recordings showing effect of Plasma-Lyte solution (PL) and modified Plasma-Lyte solution on canine saphenous veins precontracted with norepinephrine (NE). In all cases, rapid and complete relaxation was induced by Plasma-Lyte solution (A),Plasma-Lyte solution without acetate (8), Plasma-Lyte solution without gluconate (C), and Plasma-Lyte solution without both gluconate and acetate (D). E, Normalized plot of relaxation of norepinephrine-induced contraction as function of time in standard and modified Plasma-Lyte solutions. Data are expressed as mean ± standard error of mean for six veins. For clarity, statistical comparison between solutions is not shown, although analysis showed that Plasma-Lyte solution without acetate and gluconate was more effective at 5 and 10 minutes than some other solution types (p < 0.05).
rine in PSS did not change contraction amplitude after solution exchange with identical norepinephrine-containing PSS (Fig. 4, A). Other canine saphenous vein exposed to standard Plasma-Lyte solution showed the typical rapid relaxation induced by this solution at 37° C (Fig. 4, B). In contrast, canine saphenous vein exposed to modified Plasma-Lyte solution containing 1.6 mmol/L CaCh showed only minimal relaxation (Fig. 4, C), and the summated data show that contraction amplitude after
storage in CaCh-containing Plasma-Lyte solution for 60 minutes was sti1184% ± 8% of the initial norepinephrineinduced constriction (Fig. 4, D). Together, these findings imply that the low calcium concentration in Plasma-Lyte solution likely accounts for its relaxing action on canine saphenous vein. To further validate this finding, we also performed a series of "calcium-return" experiments. Canine saphenous vein segments contracted by norepinephrine in PSS
1 45 0
The Journal of Thoracic and Cardiovascular Surgery June 1994
Sanchez et al.
o --/
ISg
•
O-OPl e-ePl+CoIcium
lit
lit ....
~
z
0
I... 0
u
B~
/Sg
t_-
----/Sg
;-----------o
15
...z •
0
PL.Calclum
c
. - . PSS
30
45 (min)
60
75
90
20 10
50
E
e - e Pl 0-0 Pl+Cok:ium
120 lit
:
60
100
I: \,;.~~I ~
./rr1Tfi!1i1
...
z l!;
•
20
......
10
20
30
40
50
60
70
lID
lKl
llME (minut••)
Fig. 4. Recordingsin canine saphenousveins showing effectof replacingPSS containingnorepinephrine (NE) with other solutions containingsame norepinephrine concentration(2 X 10- 6 moljL). PSS wasreplacedwithsimilar PSS, which acted as a time control (A), standard Plasma-Lyte solution (B), and Plasma-Lyte solution containing 1.6 mmoljL CaCh (C). D, Line graph showing vasoactive effectsof PSS, Plasma-Lyte solution,and CaCh-containing Plasma-Lyte solutionas function of time (*p < 0.05 for PSS versusPlasma-Lyte solutionwith calcium at same time interval; +p < 0.05 for Plasma-Lyte solution versusPlasma-Lyte solution with calcium). E, Inset is original recording that showsthat adding 1.6mmoljL CaCl2completely reversed relaxationinducedby Plasma-Lytesolution. Main graph showsaveraged data plotted at 5-minute intervals (+p < 0.05 for Plasma-Lytesolution or Plasma-Lyte solution with calcium versus original norepinephrine-induced contraction).
were exposed to standard Plasma-Lyte solution for 30 minutes to obtain maximal relaxation, and then standard Plasma-Lyte solution was replaced with modified Plasma-Lyte solution containing 1.6 mmol/L CaCI 2• Under these conditions, as shown by an original trace (Fig. 4, E, inset), relaxation induced by Plasma-Lyte solution at 30 minutes was reversed completely by exposure to CaChcontaining Plasma-Lyte solution. The summarized data show that veins exposed to standard Plasma-Lyte solution relaxed to 5% ± 1% of initial contraction amplitude by 30 minutes, but reconstricted to 100% ± 12% of the original contraction in PSS after exposure to Plasma-Lyte solution containing 1.6 mmol/L CaCh (Fig. 4, E; n = 6 veins). Contractions stabilized at 92% ± 12% and 98% ± 24% of initial control amplitude at 30 and 60 minutes after solution exchange, respectively, and these values were not different from initial values in PSS.
Experiments in human saphenous veins Time-dependent relaxation to Plasma-Lyte solution at 37° C and 25° C in human saphenous vein. Our results in canine saphenous veins suggested that storage at 37° C in Plasma-Lyte solution was the optimal condition for venodilation. To validate the efficacy of this method in human saphenous veins, we obtained veins that had been discarded from coronary bypass operations and prepared them for isometric tension recording. Segments of human saphenous vein were preconstricted with 2 X 10-6 mol/L norepinephrine in PSS at 37° C or 25° C, after which the bath solution was exchanged for norepinephrine-containing Plasma-Lyte solution. Changes in tension were recorded at the same intervals as for the canine saphenous veins. Similar to results with canine saphenous veins, Plasma-Lyte solution at 37° C induced immediate relaxation in human saphenous veins, which was complete
The Journal of Thoracic and Cardiovascular Surgery Volume 107, Number 6
Sanchez et al.
I45I
PL
'"
I
37·C
A
,
0
I
15
30
(min)
,
45
SO
cs PL
B
c
5,
.,.
NE
I
o
I
15
,
30
,
45
(min)
I
60
120
~
100 .-0...... _
~
",
g so ~ 60
8
w
40
~
20
M
0
T\*
. - . PL(J7"C) 0 - 0 PL (~OC)
9.......
1
9 <, <,c}.. . .
*
\.
?---? ), * * * ---r--?-?-?-? * * * * * *
'.......
o
20
---.
40
TIIAE (minutes)
10
Fig. 5. Recordings showing effect of Plasma-Lyte solution on human saphenous veins precontracted by 2 X 10- 6 moljL norepinephrine (NE) in PSS at 37° C (A)and 25° C (B).C, Graph showing relaxation as function of time after PSS was replaced by Plasma-Lyte solution at 37° Cor 25° C. Data are expressed as mean ± standard error of mean for six veins. *p < 0.05 at same time interval for Plasma-Lyte solution at 37° C and 25° C.
within 20 minutes (Fig. 5, A). In Plasma-Lyte solution at 25° C, the vessels significantlyrelaxed, but were still constricted by 40% ± II % after 60 minutes (Fig. 5, Band C). The graph of contraction as a function of time emphasizes that Plasma-Lyte solution rapidly and effectivelyrelaxed human saphenousveinsat 37° C (Fig. 5, C; n = 6 veins). Discussion
Surgical excision of human saphenous vein segments induces venoconstriction, which may be severe and protracted. The cause of this venospasm is unknown, but it likely triggers endothelial and smooth muscle cell injury.4, II Thus it is important to identify an effectivevenodilating solution for the intraoperative storage of human saphenous vein grafts. Techniques designed to reduce venospasm include intraluminal pressurization,14, 16, 18 application of venodilator drugs.'- 11-18,20,21 and vein storage in different solution types.t '" However, the acute vasoactiveeffectsof commonly used storage solutionsare apparently unknown. Thus we screened for the optimal storage condition to induce relaxation of canine saphenous veins and then tested for relevance of these findings to human saphenous veins. The initial finding of this study was that Plasma-Lyte solutioninduced rapid and complete relaxation of canine saphenous veinat 37° C. Papaverine did not enhance this acute relaxation induced by Plasma-Lyte solution, emphasizing the high efficacy of Plasma-Lyte solution
and suggesting that the practice of adding papaverine to Plasma-Lyte solution at lower temperatures I 1-15 likely is unwarranted at 37° C. NSS relaxed canine saphenous vein less than Plasma-Lyte solution at 37° C, and papaverine was required to completely relax NSS-exposed veins. The vasodilating effect of both Plasma-Lyte solution and NSS was greatly reduced at 25° C compared with that at 37° C, suggesting that warming storage solutions from 25° C to 37° C may be a simple method for enhancing their venodilating efficacy. The complete relaxation of isolated canine saphenous vein by Plasma-Lyte solution prompted us to identify the mechanism of this relaxation. Generally the in vivo vasodilatory action of Plasma-Lyte solution has been attributed to its anionic composition.F: 25-30 but our data showed that the acetate and gluconate in Plasma-Lyte solution do not confer its venorelaxant effect. Consequently, we postulated that the lack of calcium in Plasma-Lyte solution accounted for its venodilating action, inasmuch as the canine saphenous vein relies primarily on Ca2+ influx for vascular contraction.P 24, 31 Our data support this hypothesis, because addition of 1.6 mmoljL CaCl 2 to Plasma-Lyte solution prevented 80% to 90% of the relaxation incurred by standard PlasmaLyte solution and completely reversed relaxation induced by this solution. However, low calcium per se may not be sufficientto eliminate vein activation, because NSS also contains low calcium but dilated the canine saphenous vein less than Plasma-Lyte solution. The reason NSS
I 452
The Journal of Thoracic and Cardiovascular Surgery June 1994
Sanchez et al.
relaxed canine saphenous vein less is not readily apparent. NSS has a low pH compared with Plasma-Lyte solution, but acidosis reportedly promotes, rather than inhibits, canine saphenous vein relaxation.F Alternatively, physiologic levels of critical ions such as K+ and Mg2+ may be required for full venodilation.V: 34 With regard to temperature, our study suggests that the venorelaxant effect of Plasma-Lyte solution is enhanced at 37° C compared with 25 ° C, implying that physical factors, as well as the ionic profile of storage solutions, likely promote optimal relaxation. Finally, we evaluated the vasoactive effect of PlasmaLyte solution on human saphenous vein. Our data showed that Plasma-L yte solution rapidly and completely relaxed norepinephrine-contracted human saphenous vein at 37° C, but only induced slow and incomplete relaxation at 25 ° C. Plasma-Lyte solution reportedly causes less vascular smooth muscle and endothelial cell damage than NSS. 14, 19 Furthermore, the option of storing vein grafts in a drug-free venodilating solution is appealing, because the addition of venodilators is both costly and potentially injurious to graft integrity. 13, 20 On the basis of our results showing that Plasma-Lyte solution effectively relaxes human saphenous vein at 37° C, we recommend and currently use warm Plasma-Lyte solution as a venodilating storage solution in the operating room.
12.
We thank Ms. Anne Runnells for technical and graphics support.
16.
REFERENCES 1. Panetta TF, Marin ML, Veith FJ, et al. Unsuspected preexisting saphenous vein disease: an unrecognized cause of vein bypass failure. J Vase Surg 1992;15:102-12. 2. Donaldson MC, Mannick JA, Whittemore AD. Causes of primary graft failure after in situ saphenous vein bypass grafting. J Vase Surg 1992;15:113-20. 3. Angelini GD, Bryan AJ, Hunter S, et al. A surgical technique that preserves human saphenous vein functional integrity. Ann Thorac Surg 1992;53:871-4. 4. Olinger GN. Consequences of trauma to the saphenous vein incurred during preparation for use as grafts. In: Bernhard VM, Towne JB, eds. Complications in vascular surgery. Orlando, Fla.: Grune & Stratton, 1985;143-56. 5. Gundry SR, Jones M, Ishihara T, et al. Optimal preparation techniques for human saphenous vein grafts. Surgery 1980;88:785-94. 6. Sottiurai VS, Stanley JC, Fry W J. Ultrastructure of human and transplanted canine veins: effects of different preparation media. Surgery 1983;93:28-38. 7. Bush HL, Jakubowski JA, Curl GR, et al. The natural history of endothelial structure and function in arterialized vein grafts. J Vase Surg 1986;3:204-15. 8. Adcock GD, Adcock OT, Wheeler JR, et al. Arterialization
17.
9.
10.
II.
13.
14.
15.
18.
19.
20.
21.
22.
23.
of reversed autogenous vein grafts: quantitative light and electron microscopy of canine jugular vein grafts harvested and implanted by standard or improved techniques. J Vase Surg 1987;6:283-95. Hoover EL, Ross M, Fani K, et al. Biochemical and histopathologic comparison between blood and saline storage of canine veins. J Vase Surg 1988;7:543-8. Barner HB, Fisher VW. Endothelial preservation in human saphenous veins harvested for coronary grafting. J THORAC CARDIOVASC SURG 1990;100:148-9. Baumann FG, Catinella FP, Cunningham IN, et al. Vein contraction and smooth muscle cell extensions as causes of endothelial damage during graft preparation. Ann Surg 1981;194:199-211. Catinella FP, Cunningham IN, Srungaram RK, et al. The factors influencing early patency of coronary artery bypass vein grafts. J THORAC CARDIOVASC SURG 1982;83:686700. Roberts AJ, Hay DA, Mehta JL, et al. Biochemical and ultrastructural integrity of the saphenous vein conduit during coronary artery bypass grafting: preliminary results of the effects of papaverine. J THORAC CARDIOVASC SURG 1984;88:39-48. Lawrie GM, Weilbacher DE, Henry PD. Endothelium-dependent relaxation in human saphenous vein grafts. J THORAC CARDIOVASC SURG 1990;100:612-20. Quist WC, Haudenschild CC, LoGerfo FW. Qualitative microscopy of implanted vein grafts: effects of graft integrity on morphologic fate. J THORAC CARDIOVASC SURG 1992;103:671-7. LoGerfo FW, Quist WC, Crawshaw HM, et al. An improved technique for preservation of endothelial morphology in vein grafts. Surgery 1981;90:1015-24. Haudenschild CC, Gould KE, Quist WC, et al. Protection of endothelium in vesselsegments excised for grafting. Circulation 1981;64(Suppl):II101-7. LoGerfo FW, Quist WC, Cantelmo NL, et al. Integrity of vein grafts as a function of initial intimal and medial preservation. Circulation 1983;68(Suppl):II 117-24. Schwartz LB, Radic ZS, O'Donohoe MK, et al. Functional and morphologic endothelial damage in rabbit external jugular veins stored in heparinized normal saline. Blood Vessels 1991;28:511-9. Dries D, Mohammed SF, Woodward SC, et al. The influence of harvesting technique on endothelial preservation in saphenous veins. J Surg Res 1992;52:219-25. Rusch NJ, Wooldridge TA, Olinger GN, et al. The effect of ambient temperature on papaverine-induced relaxation in canine saphenous veins. J THORAC CARDIOVASC SURG 1992;104:1289-93. Olinger GN, Werner PH, Bonchek LI, et al. Vasocilator effectsof the sodium acetate in pooled protein fraction. Ann Surg 1979;198:305-11. Cooke JP, Rimele TJ, Flavahan NA, Vanhoutte PM. Nimodipineand inhibition of alpha adrenergic activation of the isolated canine saphenous vein. J Pharmacol Exp Ther 1985;234:598-602.
The Journal of Thoracic and Cardiovascular Surgery Volume 107, Number 6
24. HeGW, AngusJA, Rosenfeldt FL. Reactivityofthecanine isolated internal mammary artery, saphenous vein, and coronary artery to constrictor and dilator substances: relevance to coronary bypass graft surgery. J Cardiovasc PharmacoI1988;12:12-22. 25. Kirkendol PL, Pearson JE, Boer JD, et al. Myocardial depressant effects of sodium acetate. Cardiovasc Res 1978; 12:127-36. 26. Aizawa Y, Ohmori T, Imai K, et al. Depressant action of acetate upon the human cardiovascular system. Clin N ephrol 1977;8:477-80. 27. Molnar JI, Scott JB, Frohlich ED, et al. Local effects of various anions and H+ on dog limb and coronary vascular resistance. Am J Physiol 1962;203:125-32. 28. Frohlich ED. Vascular effects of the Krebs intermediate metabolites. Am J PhysioI1965;208:149-53. 29. Steffen RP, O'Neill JT, Haddy FJ. Effect of theophylline on renal vasoactivityof acetate and adenosine. J Cardiovasc Pharmacol 1988;II :682-6.
Sanchez et al.
I453
30. Daugirdas JT, Swanson V, Islam S, et al. Acetate causes endothelium-independent increases in cyclic AMP in rat caudal artery. Am J Physiol 1988;255:H 1378-83. 31. Kock P, Wilffert B, Wilhelm D, et al. An approach to differentiate between noradrenaline-elicited contractile processes in the rat isolated aorta. Naunyn-Schmiedebergs Arch Pharmacol 1990;342:454-61. 32. Vanhoutte P, Clement D. Effect of pH and Pco, changes on the reactivity of isolated venous smooth muscle. Arch Int Physiol Biochim 1968;76:144-6. 33. Chiavarelli M, Fabi F, Stati T, et al. Effects of cardioplegic solutions and their components on human saphenous vein contractility. Ann Thorac Surg 1992;53:455-9. 34. Hermsmeyer K. Electrogenesis of increased norepinephrine sensitivity of arterial vascular muscle in hypertension. Circ Res 1976;38:362-7.
Alexandra M. Sanchez,* Theresa A. Wooldridge, BS,a, b Lawrence E. Boerboom, PhD,a, b Gordon N. Olinger, MD,a G. Hossein Almassi, MD,a and Nancy J. Rusch, PhD,a, b Milwaukee, Wis.
From the Departments of Cardiothoracic Surgery" and Physiology.> Medical College of Wisconsin, Milwaukee, Wis. Supported by National Institute of Health grant HL-41840 (L.E.B.), National Institute of Health grant HL-40474 (NJ.R.), and an award from the Max Baer Heart Fund of the Fraternal Order of Eagles. Received for publication Aug. 3, 1993. Accepted for publication Oct. 25, 1993. Address for reprints: Nancy J. Rusch, PhD, Associate Professor, Department of Physiology, Medical College of Wisconsin, 8701 Watertown Plank Rd., Milwaukee, WI 53226.
* ApprenticeshipIn Research Scholar, Minority Student Affairs, Medical Collegeof Wisconsin, Milwaukee, Wis. Copyright © 1994by Mosby-Year Book, Inc. 0022-5223/94 $3.00 + 0
12/1/52620
Minimizing vascular smooth muscle and endothelial cell damage during coronary bypass operations improves the longevity of in situ vein grafts. Multiple factors may influence graft integrity during the intraoperative period, including the inherent condition of the vein, the level of surgical manipulation, and the type of solution used for venous storage.!:" With regard to storage solutions, much attention has focused on defining which solutes and temperatures provide optimal vein storage. On the basis of graft morphology, autologous whole blood, Plasma-Lyte solution, (Baxter Healthcare Corp., Deerfield, Ill.), normal saline solution (NSS), and cell culture media have been promoted as optimal solutions.>"? Storage solution temperatures from 4 0 C to 37 0 C also have been advanced 1445
I 446
Sanchez et al.
as appropriate for preserving vascular smooth muscle and endothelial cell integrity. 15-20 The design of a venous storage solution should consider both technical and physiologic factors. From the technical perspective, the solution must induce rapid and sufficient venous relaxation to permit vein implantation. From the physiologic viewpoint, the solution composition should preserve or enhance long-term graft viability. However, despite the need to evaluate the acute venodilating action of storage solutions, there is little practical information about this issue. Most studies have focused on the anatomic impact, rather than the acute vasodilating effect, of storage media. Thus whether or not common storage solutions, such as Plasma-Lyte solution and NSS, help dilate vein grafts during intraoperative storage is unknown. The goal of this study was to evaluate the vasodilating effect and mechanisms ofPlasma-Lyte solution and NSS on isolated canine saphenous veins. On the basis of this initial screening, we further tested for optimal vasodilating efficacy in human saphenous veins obtained from coronary bypass operations.
Methods Preparation of veins. Canine saphenous veins were removed from anesthetized dogs (sodium pentobarbital, 30 mg/kg intravenously) using protocols approved by National Institutes of Health guidelines and the Institutional Animal Care and Use Committee. In later experiments, segments of human saphenous vein discarded during coronary bypass operations were obtained and handled according to standards suggested by the Institutional Pathogen Committee. All veins were handled and prepared carefully to preserve vascular endothelium and smooth muscle cell function. Immediately after removal, all veins were divided into 3 to 4 mm segments and placed at the optimal point of their length-tension relationship in an isometric tensionrecording system." Standard bath solution was physiologicsalt solution (PSS) composed of (in millimoles per liter) NaCl119, KCI4.7,MgS041.l7,CaCh 1.6,NaH 2P041.l8,NaHC0324, ethylenediaminetetraacetic acid 0.026, and glucose ~.5. Experiments in canine saphenous vein. After equilibration for 30 to 60 minutes at 37° C, the temperature of the PSS in some tissue baths was reduced to 25° C, whereas other baths were maintained at 37° C. This permitted us to compare the vasoactive effect of storage solutions between canine saphenous vein at physiologic temperature (37° C) and the ambient temperature of the operating room (25° C). After temperature equilibration,2 X 10- 6 mol/L norepinephrine was added to the PSS in the tissue baths to obtain a sustained, half-maximal contraction of canine saphenous veins at 37° Cor 25° C. These norepinephrine-induced contractions were produced to mimic the venospasm observed during surgical vein removaJ.21 Subsequently, the norepinephrine-containing PSS in the tissue baths was equilibrated with room air and then replaced with new norepinephrine-containing solutions, which were (I) the same norepinephrine-containing PSS, which served as a time control, (2) Plasma-Lyte solution, or (3) NSS. Solution exchange was accomplished by emptying the tissue baths of initial PSS and
The Journal of Thoracic and Cardiovascular Surgery June 1994
then perfusing the new solutions into the baths from upper reservoirs. It should be emphasized that constant o-adrencceptor activation of canine saphenous veins was maintained throughout the recording period, inasmuch as all storage solutions had the same norepinephrine concentration (2 X 10- 6 moll L) as the initial PSS. After replacement of norepinephrine-containing PSS with norepinephrine-containing storage solutions, tension was recorded at 5-minute intervals for 30 minutes and then at 15-minute intervals for 60 to 90 minutes to assess the effect of Plasma-Lyte solution and NSS on norepinephrine-induced contractions. In one series of experiments, 10- 4 mol/L papaverine was added to the Plasma-Lyte solution or NSS to evaluate whether this vasodilator drug was able to induce further venorelaxation. In other experiments, the PSS was replaced by a modified Plasma-Lyte solution identical to the commercially available solution, but containing equimolar replacement of acetate and gluconate by NaCI or having the addition of 1.6 mmol/L CaCl 2 to its standard composition. In these instances, the pH of the modified solutions was kept constant at 7.4 with the use of NaOH as a buffer. Experiments in human saphenous vein. After the venodilating effect ofPlasma-Lyte solution and NSS was compared at 37° C and 25° C in canine saphenous veins, the dilating effect of Plasma-Lyte solution was further compared in human saphenous veins at the same temperatures. Experimental protocols were identical to methods used to evaluate relaxations in canine saphenous veins induced by Plasma-Lyte solution. Data acquisition and analysis. Changes in tension were displayed continuously on an enhanced graphics adapter color monitor and simultaneously digitized and stored on an IBMcompatible 386SX computer (Citus Corp., Menomonee Falls, Wis.). Data were acquired and analyzed with CODAS software (Dataq Instruments, Akron, Ohio). Contraction amplitude was expressed as mean ± standard error of the mean and plotted as percent of the initial contractile response to norepinephrine. Significant differences in contraction amplitude between storage solution types at matched time intervals were determined by two-way analyses of variance with repeated measures on one factor with the use of a CSS:General Manova Stats Plus Software program (Statsoft, Tulsa, Okla.). Significant differences between individual means of specificgroups were determined by Duncan's new multiple range test. A p value less than 0.05 was considered statistically significant. Solutions and drugs. The composition of solutions used in this study were (I) Plasma-Lyte solution (Plasma-Lyte A injection solution, Baxter),22 composed of (in millimoles per liter) Na+ 140, K+ 5, Mg 2+ 3, Cl- 98, acetate 27, andgluconate 23 (pH = 7.4), and (2) NSS (Baxter), composed of NaCI 154 mfiq/L (pH = 5.0). The composition of Plasma-Lyte solution was modified in some experiments as explained previously. Norepinephrine ([-]-arterenol) and papaverine were obtained from Sigma Chemical Co., St. Louis, Missouri.
Results Experiments in canine saphenous veins Time-dependent relaxations to Plasma-Lyte solution and NSS at 37° Cand 25° C. Segments of canine saphenous vein were constricted with a median effective dose of 2 X 10-6 moljL norepinephrine in PSS at 37° C, after which the bath solution was exchanged for Plasma-Lyte
The Journal of Thoracic and Cardiovascular Surgery Volume 107, Number 6
Sanchez et al.
1 447
,
PL.PAP
A
~~~ NE
B
0
:?::
z 0
..
;:::: u
z 0
40
U
w
...z 0
~
15
30 (min)
20 0
45
60
NE
0..
\_.370C : ,0-0 37°C (+PAP)
1'. ~,
10
-,
0
--
60
.. ..
.'
30 45607590 20 TIME (minutes)
~
0
15
30
30 (min)
45
I
45
60
•
60
NSS
100
.-.370C 0-0 37°C (+PAP)
z
80
;:::: u
60
T
40
~T
.
15
'"
NE
D If) If)
60
0::
l-
0
PL
100 80
45
I"""
...
C Q.
30
~dNE
If) If)
15
0
0::
I-
z
0
U
w
...z 0
il'1
20 0
·-e "
...... T
i i-i ° + -·-·~li-1'9.....+ + + + + + +
10
-
tl;l l;l Ii' iii 20 30 45607590 TIME (minutes)
Fig. 1. Recordings showing effectof Plasma-Lytesolution (PL) and NSS on caninesaphenous veins precontracted by 2 X 10- 6 mol/L norepinephrine (NE) at 37° C. A, Plasma-Lytesolution inducedrapid and completerelaxation (left trace), which was not affected by 10- 4 mol/L papaverine (PAP) (right trace). B, NSS induced only partial relaxation (left trace), whichwasenhancedby 10- 4 mol/L papaverine (right trace). Plotsin C and D showrelaxation as function of time after PSS was replacedwith Plasma-Lytesolution or NSS, respectively. Data are expressed as percent of initial norepinephrine-induced constriction and represent mean ± standard error of mean for six veins. +p < 0.05 for control versus papaverine-containing solution at same time interval. solution or NSS containing the same norepinephrine concentration. Remaining tension was measured every 5 minutes for the first 30 minutes after solution exchange and then at 15-minute intervals for 1 hour. Exposure to Plasma-Lyte solution at 37° C induced immediate relaxation, which was complete within 20 minutes (Fig. 1, A, left trace). In parallel canine saphenous vein segments, addition of 10-4 moljL papaverine to Plasma-Lyte solutiondid not enhance relaxations induced by Plasma-Lyte solution at 37° C (Fig. 1, A, right trace). The plot of contraction amplitude as a function of storage time in Plasma-Lyte solution shows that Plasma-Lyte solution is an effective venorelaxant at 37° C and indicates that the venodilating efficacy of Plasma-Lyte solution was not enhanced by papaverine (Fig. 1, C; n = 6 veins). In contrast to Plasma-Lyte solution, storage in NSS at 37° C did not induce complete relaxation of canine saphenous vein (Fig. 1, B, left trace). The summarized data show that 25% ± 5% of the initial norepinephrine-induced contraction remained after 90 minutes in NSS (Fig. 1, D;
n = 6 veins). Under these conditions of incomplete relaxation, 10-4 mol/L papaverine enhanced the venorelaxant effect ofNSS at 37° C (Fig. 1, B, right trace, and Fig. 1, D). The solution exchange procedure per se did not affect contraction amplitude, as determined by monitoring contraction during replacement of norepinephrine-containing PSS with identical norepinephrine-containing solution. A sample recording from such a control experiment is presented later in the results. Similar experiments were done at 25 ° C, a temperature level comparable to that of the operating room. At 25 ° C, Plasma-Lyte solution only partially relaxed canine saphenous vein during a 90-minute recording (Fig. 2, A, left trace) and 56% ± 1%, 29% ± 9%, and 20% ± 7% of the initial norepinephrine-induced contraction remained after 30, 60, and 90 minutes of Plasma-Lyte solution storage, respectively (Fig. 2, C; n = 6 veins). This response contrasted with the complete and rapid relaxation induced by Plasma-Lyte solution at 37° C (Fig. 1, C). This residual contraction to norepinephrine at 25° C
I 44 8
The Journal of Thoracic and Cardiovascular Surgery June 1994
Sanchez et al.
A
~OC~ ..
NE
,__,__,n,__,
0
15
30
45
60
j ..
-1.-------,=-=c=:~ NE
0
15
30
45
60
NSS.PAP
B
C==~ ...
I
I
NE
0
15
I
I
30 45 (min)
,'''
GO
NE
I
I
I
0
15
30 (min)
o o
~
u
60
~ o
40
~
20
o
o
u
u, ~
,
60
NSS
lI'I lI'I
a. 100 :i':: z 80
I
45
\..
.-.250C 0'_ 0 - 0 25°C (+PAP) +\.+..................
°'\+ -i ?,+ +
+
10
3045607590
-.~.TT
?'O
-.-w-. TIT
+ + + + ' ....9."0_ _
20
TIME (minutes)
Fig. 2. Recordings showingeffectof Plasma-Lyte solution (PL) and NSS on canine saphenousveinsprecontracted by 2 X 10- 6 mol/L norepinephrine (NE) at 25° C. A, Compared with tracings at 37° C, reiaxatiOlfinduced by Plasma-Lyte solution was slowerand less effective at 25° C (left tracing) and its effect was potentiated by 10- 4 mol/L papaverine (PAP) (right tracing). B, NSS also inducedslowerand lesserrelaxation at 25° C (left tracing), whichwas potentiated by 10- 4 mol/L papaverine (right tracing). Plots in C and D showrelaxation as function of time after PSS was replaced by Plasma-Lyte solution or NSS, respectively. Expression of data is same as in Fig. I.
was eliminated by 10-4 mOI/L papaverine (Fig. 2, A, right trace, and Fig. 2, C). Similarly, storage in NSS at 25° C relaxed norepinephrine-contracted canine saphenous vein less than storage at 37° C (Fig. 2, B, left trace), and relaxation did not progress after the initial 45 minutes ofNSS exposure (Fig. 2, D; n = 6 veins). The addition of 10-4 mol/L papaverine also reduced venoconstriction in NSS at this temperature, confirming that NSS at 25° C did not eliminate a large component of active venous con traction.
Removal ofacetate and gluconate from Plasma-Lyte solution. Because our initial experiments (Figs. I and 2) showed that optimal relaxation of canine saphenous vein occurred in Plasma-Lyte solution at 37° C, we designed experiments to identify the vasodilatory component of Plasma-Lyte solution. It has been suggested that two anions in Plasma-Lyte solution, acetate and gluconate, account for its relaxant action.V To test this hypothesis, we determined the effect of anion-modified Plasma-Lyte solutions on norepinephrine-induced constrictions at 37° C. Canine saphenous vein relaxed rapidly after exposure to standard Plasma-Lyte solution (Fig. 3, A), as shown earlier in this study at 37° C (Fig. I, A). Parallel record-
ings showed that removal of 27 mmol/L acetate (Fig. 3, B), removal of 23 mmol/L gluconate (Fig. 3, C), or removal of both anions (Fig. 3, D) from standard Plasma-Lyte solution did not inhibit the rapid relaxation induced by this solution. Surprisingly, in view of the literature, the most rapid relaxation was in canine saphenous vein exposed to Plasma-Lyte solution without acetate or gluconate (Fig. 3, E; n = 6 veins). Addition of 1.6 mmolfI. CaCl2 to Plasma-Lyte solution. In view of the absence of evidence for an anion-induced venorelaxant mechanism for Plasma-Lyte solution, we postulated that relaxation induced by Plasma-Lyte solution was due to the low ionized calcium concentration in this solution. This was a reasonable hypothesis, inasmuch as Plasma-Lyte solution contains no added calcium and canine saphenous vein relies on calcium influx for sustained contraction.P'?" To test this hypothesis, we added a physiologic level of calcium (1.6 mrnol/L CaCI 2) to standard Plasma-Lyte solution at 37° C and evaluated its effect on norepinephrine-contracted veins. As in other experiments, we documented that the solution exchange procedure per se did not affect contraction amplitude. Thus canine saphenous vein constricted by norepineph-
The Journal of Thoracic and Cardiovascular Surgery Volume 107, Number 6
Sanchez et al.
A
,
PL - Acetate
~L
~= . NE
0
15
.
.s",.....90
30 (min)
NE
o
fl_,.,..... . •
0
45
30 (min)
15
E
15
30
(min)
_ (Gluconate/Acetate)
PL - Gluconate
NE
1 449
90 NE
o
30
15
(min)
100 0-
~
80
z
0
i=
o 60 -c
It:
I-
Z 0
u
40
w
Z u,
0
. - . PL
i~.
Vl Vl
20
,\~ '\~
,,\"
T
PL - Acetate v - v PL - Gluconate .- - . PL - (Gluconate/Acetate)
0--0
,, "t"" '" '-,\~\" ,
1 ....
llt
5
-c T
~,,, -,
"
v~""i
---- ----.=_.. ___
.. ! ·0----.- - - - • ~====, 10
15
20
~
25
. 30
TIME (minutes) Fig. 3. Recordings showing effect of Plasma-Lyte solution (PL) and modified Plasma-Lyte solution on canine saphenous veins precontracted with norepinephrine (NE). In all cases, rapid and complete relaxation was induced by Plasma-Lyte solution (A),Plasma-Lyte solution without acetate (8), Plasma-Lyte solution without gluconate (C), and Plasma-Lyte solution without both gluconate and acetate (D). E, Normalized plot of relaxation of norepinephrine-induced contraction as function of time in standard and modified Plasma-Lyte solutions. Data are expressed as mean ± standard error of mean for six veins. For clarity, statistical comparison between solutions is not shown, although analysis showed that Plasma-Lyte solution without acetate and gluconate was more effective at 5 and 10 minutes than some other solution types (p < 0.05).
rine in PSS did not change contraction amplitude after solution exchange with identical norepinephrine-containing PSS (Fig. 4, A). Other canine saphenous vein exposed to standard Plasma-Lyte solution showed the typical rapid relaxation induced by this solution at 37° C (Fig. 4, B). In contrast, canine saphenous vein exposed to modified Plasma-Lyte solution containing 1.6 mmol/L CaCh showed only minimal relaxation (Fig. 4, C), and the summated data show that contraction amplitude after
storage in CaCh-containing Plasma-Lyte solution for 60 minutes was sti1184% ± 8% of the initial norepinephrineinduced constriction (Fig. 4, D). Together, these findings imply that the low calcium concentration in Plasma-Lyte solution likely accounts for its relaxing action on canine saphenous vein. To further validate this finding, we also performed a series of "calcium-return" experiments. Canine saphenous vein segments contracted by norepinephrine in PSS
1 45 0
The Journal of Thoracic and Cardiovascular Surgery June 1994
Sanchez et al.
o --/
ISg
•
O-OPl e-ePl+CoIcium
lit
lit ....
~
z
0
I... 0
u
B~
/Sg
t_-
----/Sg
;-----------o
15
...z •
0
PL.Calclum
c
. - . PSS
30
45 (min)
60
75
90
20 10
50
E
e - e Pl 0-0 Pl+Cok:ium
120 lit
:
60
100
I: \,;.~~I ~
./rr1Tfi!1i1
...
z l!;
•
20
......
10
20
30
40
50
60
70
lID
lKl
llME (minut••)
Fig. 4. Recordingsin canine saphenousveins showing effectof replacingPSS containingnorepinephrine (NE) with other solutions containingsame norepinephrine concentration(2 X 10- 6 moljL). PSS wasreplacedwithsimilar PSS, which acted as a time control (A), standard Plasma-Lyte solution (B), and Plasma-Lyte solution containing 1.6 mmoljL CaCh (C). D, Line graph showing vasoactive effectsof PSS, Plasma-Lyte solution,and CaCh-containing Plasma-Lyte solutionas function of time (*p < 0.05 for PSS versusPlasma-Lyte solutionwith calcium at same time interval; +p < 0.05 for Plasma-Lyte solution versusPlasma-Lyte solution with calcium). E, Inset is original recording that showsthat adding 1.6mmoljL CaCl2completely reversed relaxationinducedby Plasma-Lytesolution. Main graph showsaveraged data plotted at 5-minute intervals (+p < 0.05 for Plasma-Lytesolution or Plasma-Lyte solution with calcium versus original norepinephrine-induced contraction).
were exposed to standard Plasma-Lyte solution for 30 minutes to obtain maximal relaxation, and then standard Plasma-Lyte solution was replaced with modified Plasma-Lyte solution containing 1.6 mmol/L CaCI 2• Under these conditions, as shown by an original trace (Fig. 4, E, inset), relaxation induced by Plasma-Lyte solution at 30 minutes was reversed completely by exposure to CaChcontaining Plasma-Lyte solution. The summarized data show that veins exposed to standard Plasma-Lyte solution relaxed to 5% ± 1% of initial contraction amplitude by 30 minutes, but reconstricted to 100% ± 12% of the original contraction in PSS after exposure to Plasma-Lyte solution containing 1.6 mmol/L CaCh (Fig. 4, E; n = 6 veins). Contractions stabilized at 92% ± 12% and 98% ± 24% of initial control amplitude at 30 and 60 minutes after solution exchange, respectively, and these values were not different from initial values in PSS.
Experiments in human saphenous veins Time-dependent relaxation to Plasma-Lyte solution at 37° C and 25° C in human saphenous vein. Our results in canine saphenous veins suggested that storage at 37° C in Plasma-Lyte solution was the optimal condition for venodilation. To validate the efficacy of this method in human saphenous veins, we obtained veins that had been discarded from coronary bypass operations and prepared them for isometric tension recording. Segments of human saphenous vein were preconstricted with 2 X 10-6 mol/L norepinephrine in PSS at 37° C or 25° C, after which the bath solution was exchanged for norepinephrine-containing Plasma-Lyte solution. Changes in tension were recorded at the same intervals as for the canine saphenous veins. Similar to results with canine saphenous veins, Plasma-Lyte solution at 37° C induced immediate relaxation in human saphenous veins, which was complete
The Journal of Thoracic and Cardiovascular Surgery Volume 107, Number 6
Sanchez et al.
I45I
PL
'"
I
37·C
A
,
0
I
15
30
(min)
,
45
SO
cs PL
B
c
5,
.,.
NE
I
o
I
15
,
30
,
45
(min)
I
60
120
~
100 .-0...... _
~
",
g so ~ 60
8
w
40
~
20
M
0
T\*
. - . PL(J7"C) 0 - 0 PL (~OC)
9.......
1
9 <, <,c}.. . .
*
\.
?---? ), * * * ---r--?-?-?-? * * * * * *
'.......
o
20
---.
40
TIIAE (minutes)
10
Fig. 5. Recordings showing effect of Plasma-Lyte solution on human saphenous veins precontracted by 2 X 10- 6 moljL norepinephrine (NE) in PSS at 37° C (A)and 25° C (B).C, Graph showing relaxation as function of time after PSS was replaced by Plasma-Lyte solution at 37° Cor 25° C. Data are expressed as mean ± standard error of mean for six veins. *p < 0.05 at same time interval for Plasma-Lyte solution at 37° C and 25° C.
within 20 minutes (Fig. 5, A). In Plasma-Lyte solution at 25° C, the vessels significantlyrelaxed, but were still constricted by 40% ± II % after 60 minutes (Fig. 5, Band C). The graph of contraction as a function of time emphasizes that Plasma-Lyte solution rapidly and effectivelyrelaxed human saphenousveinsat 37° C (Fig. 5, C; n = 6 veins). Discussion
Surgical excision of human saphenous vein segments induces venoconstriction, which may be severe and protracted. The cause of this venospasm is unknown, but it likely triggers endothelial and smooth muscle cell injury.4, II Thus it is important to identify an effectivevenodilating solution for the intraoperative storage of human saphenous vein grafts. Techniques designed to reduce venospasm include intraluminal pressurization,14, 16, 18 application of venodilator drugs.'- 11-18,20,21 and vein storage in different solution types.t '" However, the acute vasoactiveeffectsof commonly used storage solutionsare apparently unknown. Thus we screened for the optimal storage condition to induce relaxation of canine saphenous veins and then tested for relevance of these findings to human saphenous veins. The initial finding of this study was that Plasma-Lyte solutioninduced rapid and complete relaxation of canine saphenous veinat 37° C. Papaverine did not enhance this acute relaxation induced by Plasma-Lyte solution, emphasizing the high efficacy of Plasma-Lyte solution
and suggesting that the practice of adding papaverine to Plasma-Lyte solution at lower temperatures I 1-15 likely is unwarranted at 37° C. NSS relaxed canine saphenous vein less than Plasma-Lyte solution at 37° C, and papaverine was required to completely relax NSS-exposed veins. The vasodilating effect of both Plasma-Lyte solution and NSS was greatly reduced at 25° C compared with that at 37° C, suggesting that warming storage solutions from 25° C to 37° C may be a simple method for enhancing their venodilating efficacy. The complete relaxation of isolated canine saphenous vein by Plasma-Lyte solution prompted us to identify the mechanism of this relaxation. Generally the in vivo vasodilatory action of Plasma-Lyte solution has been attributed to its anionic composition.F: 25-30 but our data showed that the acetate and gluconate in Plasma-Lyte solution do not confer its venorelaxant effect. Consequently, we postulated that the lack of calcium in Plasma-Lyte solution accounted for its venodilating action, inasmuch as the canine saphenous vein relies primarily on Ca2+ influx for vascular contraction.P 24, 31 Our data support this hypothesis, because addition of 1.6 mmoljL CaCl 2 to Plasma-Lyte solution prevented 80% to 90% of the relaxation incurred by standard PlasmaLyte solution and completely reversed relaxation induced by this solution. However, low calcium per se may not be sufficientto eliminate vein activation, because NSS also contains low calcium but dilated the canine saphenous vein less than Plasma-Lyte solution. The reason NSS
I 452
The Journal of Thoracic and Cardiovascular Surgery June 1994
Sanchez et al.
relaxed canine saphenous vein less is not readily apparent. NSS has a low pH compared with Plasma-Lyte solution, but acidosis reportedly promotes, rather than inhibits, canine saphenous vein relaxation.F Alternatively, physiologic levels of critical ions such as K+ and Mg2+ may be required for full venodilation.V: 34 With regard to temperature, our study suggests that the venorelaxant effect of Plasma-Lyte solution is enhanced at 37° C compared with 25 ° C, implying that physical factors, as well as the ionic profile of storage solutions, likely promote optimal relaxation. Finally, we evaluated the vasoactive effect of PlasmaLyte solution on human saphenous vein. Our data showed that Plasma-L yte solution rapidly and completely relaxed norepinephrine-contracted human saphenous vein at 37° C, but only induced slow and incomplete relaxation at 25 ° C. Plasma-Lyte solution reportedly causes less vascular smooth muscle and endothelial cell damage than NSS. 14, 19 Furthermore, the option of storing vein grafts in a drug-free venodilating solution is appealing, because the addition of venodilators is both costly and potentially injurious to graft integrity. 13, 20 On the basis of our results showing that Plasma-Lyte solution effectively relaxes human saphenous vein at 37° C, we recommend and currently use warm Plasma-Lyte solution as a venodilating storage solution in the operating room.
12.
We thank Ms. Anne Runnells for technical and graphics support.
16.
REFERENCES 1. Panetta TF, Marin ML, Veith FJ, et al. Unsuspected preexisting saphenous vein disease: an unrecognized cause of vein bypass failure. J Vase Surg 1992;15:102-12. 2. Donaldson MC, Mannick JA, Whittemore AD. Causes of primary graft failure after in situ saphenous vein bypass grafting. J Vase Surg 1992;15:113-20. 3. Angelini GD, Bryan AJ, Hunter S, et al. A surgical technique that preserves human saphenous vein functional integrity. Ann Thorac Surg 1992;53:871-4. 4. Olinger GN. Consequences of trauma to the saphenous vein incurred during preparation for use as grafts. In: Bernhard VM, Towne JB, eds. Complications in vascular surgery. Orlando, Fla.: Grune & Stratton, 1985;143-56. 5. Gundry SR, Jones M, Ishihara T, et al. Optimal preparation techniques for human saphenous vein grafts. Surgery 1980;88:785-94. 6. Sottiurai VS, Stanley JC, Fry W J. Ultrastructure of human and transplanted canine veins: effects of different preparation media. Surgery 1983;93:28-38. 7. Bush HL, Jakubowski JA, Curl GR, et al. The natural history of endothelial structure and function in arterialized vein grafts. J Vase Surg 1986;3:204-15. 8. Adcock GD, Adcock OT, Wheeler JR, et al. Arterialization
17.
9.
10.
II.
13.
14.
15.
18.
19.
20.
21.
22.
23.
of reversed autogenous vein grafts: quantitative light and electron microscopy of canine jugular vein grafts harvested and implanted by standard or improved techniques. J Vase Surg 1987;6:283-95. Hoover EL, Ross M, Fani K, et al. Biochemical and histopathologic comparison between blood and saline storage of canine veins. J Vase Surg 1988;7:543-8. Barner HB, Fisher VW. Endothelial preservation in human saphenous veins harvested for coronary grafting. J THORAC CARDIOVASC SURG 1990;100:148-9. Baumann FG, Catinella FP, Cunningham IN, et al. Vein contraction and smooth muscle cell extensions as causes of endothelial damage during graft preparation. Ann Surg 1981;194:199-211. Catinella FP, Cunningham IN, Srungaram RK, et al. The factors influencing early patency of coronary artery bypass vein grafts. J THORAC CARDIOVASC SURG 1982;83:686700. Roberts AJ, Hay DA, Mehta JL, et al. Biochemical and ultrastructural integrity of the saphenous vein conduit during coronary artery bypass grafting: preliminary results of the effects of papaverine. J THORAC CARDIOVASC SURG 1984;88:39-48. Lawrie GM, Weilbacher DE, Henry PD. Endothelium-dependent relaxation in human saphenous vein grafts. J THORAC CARDIOVASC SURG 1990;100:612-20. Quist WC, Haudenschild CC, LoGerfo FW. Qualitative microscopy of implanted vein grafts: effects of graft integrity on morphologic fate. J THORAC CARDIOVASC SURG 1992;103:671-7. LoGerfo FW, Quist WC, Crawshaw HM, et al. An improved technique for preservation of endothelial morphology in vein grafts. Surgery 1981;90:1015-24. Haudenschild CC, Gould KE, Quist WC, et al. Protection of endothelium in vesselsegments excised for grafting. Circulation 1981;64(Suppl):II101-7. LoGerfo FW, Quist WC, Cantelmo NL, et al. Integrity of vein grafts as a function of initial intimal and medial preservation. Circulation 1983;68(Suppl):II 117-24. Schwartz LB, Radic ZS, O'Donohoe MK, et al. Functional and morphologic endothelial damage in rabbit external jugular veins stored in heparinized normal saline. Blood Vessels 1991;28:511-9. Dries D, Mohammed SF, Woodward SC, et al. The influence of harvesting technique on endothelial preservation in saphenous veins. J Surg Res 1992;52:219-25. Rusch NJ, Wooldridge TA, Olinger GN, et al. The effect of ambient temperature on papaverine-induced relaxation in canine saphenous veins. J THORAC CARDIOVASC SURG 1992;104:1289-93. Olinger GN, Werner PH, Bonchek LI, et al. Vasocilator effectsof the sodium acetate in pooled protein fraction. Ann Surg 1979;198:305-11. Cooke JP, Rimele TJ, Flavahan NA, Vanhoutte PM. Nimodipineand inhibition of alpha adrenergic activation of the isolated canine saphenous vein. J Pharmacol Exp Ther 1985;234:598-602.
The Journal of Thoracic and Cardiovascular Surgery Volume 107, Number 6
24. HeGW, AngusJA, Rosenfeldt FL. Reactivityofthecanine isolated internal mammary artery, saphenous vein, and coronary artery to constrictor and dilator substances: relevance to coronary bypass graft surgery. J Cardiovasc PharmacoI1988;12:12-22. 25. Kirkendol PL, Pearson JE, Boer JD, et al. Myocardial depressant effects of sodium acetate. Cardiovasc Res 1978; 12:127-36. 26. Aizawa Y, Ohmori T, Imai K, et al. Depressant action of acetate upon the human cardiovascular system. Clin N ephrol 1977;8:477-80. 27. Molnar JI, Scott JB, Frohlich ED, et al. Local effects of various anions and H+ on dog limb and coronary vascular resistance. Am J Physiol 1962;203:125-32. 28. Frohlich ED. Vascular effects of the Krebs intermediate metabolites. Am J PhysioI1965;208:149-53. 29. Steffen RP, O'Neill JT, Haddy FJ. Effect of theophylline on renal vasoactivityof acetate and adenosine. J Cardiovasc Pharmacol 1988;II :682-6.
Sanchez et al.
I453
30. Daugirdas JT, Swanson V, Islam S, et al. Acetate causes endothelium-independent increases in cyclic AMP in rat caudal artery. Am J Physiol 1988;255:H 1378-83. 31. Kock P, Wilffert B, Wilhelm D, et al. An approach to differentiate between noradrenaline-elicited contractile processes in the rat isolated aorta. Naunyn-Schmiedebergs Arch Pharmacol 1990;342:454-61. 32. Vanhoutte P, Clement D. Effect of pH and Pco, changes on the reactivity of isolated venous smooth muscle. Arch Int Physiol Biochim 1968;76:144-6. 33. Chiavarelli M, Fabi F, Stati T, et al. Effects of cardioplegic solutions and their components on human saphenous vein contractility. Ann Thorac Surg 1992;53:455-9. 34. Hermsmeyer K. Electrogenesis of increased norepinephrine sensitivity of arterial vascular muscle in hypertension. Circ Res 1976;38:362-7.