Konstantinos N. Malizos, Anthony V. Seaber, and James R. Urbaniak
A COMPARATIVE STUDY OF THE EFFECT OF ARTERIAL AND Downloaded by: Georgetown University Medical Center. Copyrighted material.
VENOUS OCCLUSION AFTER VARIOUS PERIODS OF ISCHEMIA ABSTRACT The author attempted to quantify any additional tissue damage resulting from one hour of venous occlusion after controlled periods of ischemia in skeletal muscle, compared to equal time periods of ischemia. The rat quadriceps muscle was used. pH and temperature were monitored during the experiment and the damage was evaluated on histochemical sections. Of the other two parameters monitored in this study, the pH recovery rate during reperfusion best reflected tissue perfusion and served as a predictor for the extent of tissue damage, when global ischemia or the combination of ischemia and postischemic venous stasis did not exceed five hours. Apart from the nonstained lethally damaged areas, interstitial tissue edema and cellular infiltration were constant findings in all muscles that sustained ischemia or ischemia and subsequent venous occlusion. However, both findings were more pronounced in the muscles that underwent venous obstruction. Muscle fibers were more edematous after venous occlusion, compared to ischemic muscles. Thrombosed veins and collapsed arteries, surrounded by excessive edema and inflammatory reaction, were common in muscles after ischemia and venous occlusion. Comparing the difference in percent of lethal damage from one hour of postischemic venous stasis versus the damage from additional ischemia of equal duration, findings demonstrated comparable damage when reperfusion was established in less than five hours.
Venous stasis is a common complication in emergency or elective microsurgery and may occur either as a thrombotic complication soon after revascularization 1 or secondary to clamping for anastomosis of the vein after the arterial flow has been reestablished and the artery is still supplying blood. Vascular congestion of any cause compromises tissue perfusion in a normal muscle. 2 - 4 In an unconstricted skeletal muscle, 5 hr of independent venous occlusion has been shown to produce damage similar to 3 hr of ischemic injury.5 Prolonged antecedent ischemia leads to endothelial damage, vessel wall impairment, clotting disturbances, and parenchymal tissue damage. During reperfusion, the ischemic injury is further exacerbated by
the production of short-lived but highly reactive oxygen free radicals that can affect membrane integrity and permeability.^ 1 0 2 9 3 0 3 2 In clinical situations, such as after major limb revascularization or elective muscle transfer, venous drainage obstruction occurring in a tissue that has previously sustained an ischemic insult may often lead to a poor functional outcome. The extent of the damage from a temporary postischemic venous stasis has been related to the extent of the antecedent ischemia and the duration of vascular congestion, but the duration of additional ischemia necessary to produce the same results as various periods of venous stasis has not been determined.
Orthopaedic Research Laboratories, Department of Surgery, Duke University Medical Center, Durham, North Carolina Reprint requests-. Anthony V. Seaber, Dept. of Surgery, P.O. Box 3093, Duke University Medical Center, Durham, NC 27710 Accepted for publication lanuary 29, 1990 Copyright © 1990 by Thieme Medical Publishers, Inc., 381 Park Avenue South, New York, NY 10016. All rights reserved.
271
The purpose of this study was to quantify any additional damage resulting from 1 hr of venous occlusion in skeletal muscles after controlled periods of ischemia, and to compare this to the damage caused by equal time periods of additional ischemia. The rat quadriceps muscle was used. pH and temperature were monitored during the experiment and the damage was evaluated on histochemical sections.
MATERIALS AND METHODS An isolated quadriceps muscle model was used in 80 male Sprague Dawley rats (280 to 400 g). Each animal was anesthetized with an intraperitoneal injection of pentobarbital (50 mg/kg b.w). The quadriceps was approached through an anteromedial incision from the knee to the ipsilateral groin. The neurovascular pedicle of the muscle was isolated under a high-power operating microscope, and the femoral nerve was divided. The quadriceps was separated from all surrounding tissues, including both the origin and insertion sites, thus isolating the muscle on its vascular pedicle and excluding any other vascular supply. After the dissection was completed, the muscle was placed into its original bed, with each end sutured to its origin or insertion (Fig. 1). Temperature was monitored using a needle probe connected to an OMEGA thermistor thermometer. The tip of the probe was inserted through the distal end of the muscle, up to its middle third (see Fig. 1). Intramuscular pH was measured with a glass needle microelectrode (MI 408B, Microelectrodes, Inc.) inserted in the lateral side of the muscle to a depth of 1 cm. A skin electrode was used as a reference and both were connected to an ORION 407A Ionalyzer (see Fig. 1). A microclamp was placed on the nutrient artery of the quadriceps and a patency test was performed to
JULY 1990
verify total ischemia. The skin was closed with subcuticular sutures, thereby maintaining the muscle in the normothermic environment of the body. The approximate time required for this preparation was 30 min. During venous occlusion, the arterial clamp was released to supply blood to the quadriceps. Three initial ischemic periods of different duration were studied, as outlined in Figure 2. In Group IA, ten animals underwent 2 hr of arterial ischemia. In Group IB, an equal number of animals had 1 hr of venous occlusion, in addition to the initial 2-hr ischemia period. Eight more animals in Group IC underwent 3 hr of ischemia. Group IIA animals were subjected to an ischemic period of 4 hr (n = 10). In Group IIB, ten animals received 4 hr of ischemia and one hour of venous occlusion. The eight animals in Group IIC underwent an additional hour of ischemia for a total of 5 hr. Initial ischemia lasted 6 hr in Group IIIA animals (n = 8), while the venous occlusion Group IIIB animals had 6 hr of ischemia and 1 hr of venous occlusion (n = 8). Eight more animals in Group IIIC were subjected to an ischemia period of 7 hr (see Fig. 2). Temperature and pH were monitored every 15 min with needle probes during ischemia, venous occlusion, and the first hour of reperfusion. All animals were sacrificed 48 hr after reperfusion. At sacrifice, the quadriceps muscles were removed from the animal and sectioned into three pieces (proximal, middle, and distal third). A frozen section was obtained from each third (see Fig. 2). The sections were incubated in nitroblue tetrazolium (NBT) dye with substrate NADH for 30 min at 37° C. Normal areas of the muscle showed dark
I
Groups
II
III
A
B
c
A
B
C
A
B
2
2
2
4
4
4
6
6
Animals (No.) Ischemia (Hrs.)
8
Venous Occlusion (Hrs.)
1
Arterial Occlusion (Hrs.) Reperfusion established after (Hrs.)
C
1 2
t
i
1 4
5
6
1
1 5
1 6
7
7
• Temperature and pH monitored every 15 minutes through the first hour of reperfusion. • Reperfusion lasted 48 hours. • Muscle damage was quantified on histochemically stained sections with computerized planimetry.
Prox.
272
Figure 1. After complete dissection and evaluation to eliminate any other source of blood supply except from its pedicle, the quadriceps was replaced and resutured at its resting length and the nerve was sectioned. Temperature and pH probes measured the parameters in the core of the Figure 2. Diagram presenting the material methodolmiddle third of the muscle. ogy and muscle sectioning.
Downloaded by: Georgetown University Medical Center. Copyrighted material.
JOURNAL OF RECONSTRUCTIVE MICROSURGERY/VOLUME 6, NUMBER 3
VESSEL OCCLUSION AFTER ISCHEMIA/MALIZOS, SEABER, URBANIAK
RESULTS Analysis of intramuscular temperature measurements obtained during the different flow conditions in the muscle showed an increase of 0.69° C/hr (± 0.12 Standard Error [S.E.j) in the first 2 hr of ischemia, 0.65° C/hr (±0.12 S.E.) between the second and fourth hour, 0.43° C/hr (± 0.12 S.E.) from the fourth to sixth hour, and 0.32° C/hr (± 0.18 S.E.) during the seventh hour of ischemia (Fig. 3). During the hour of venous occlusion, temperature increased 0.59° C (± 0.15 S.E.) in Group IC, 0.49° C (± 0.27 S.E.) in Group IIB, and 0.36° C (± 0.18 S.E.) in Group IIIB animals. In the first hour of reperfusion, the temperature increased at a rate of 1.21° C (± 0.18 S.E.) in Group IA, 1.05° C (± 0.13 S.E.) in Group IB venous occlusion muscles, and 1.11° C (± 0.16 S.E.) in Group IC 3-hr ischemia muscles. In Group II, the rates of temperature increase were: 1.08° C (± 0.13 S.E.) in Group IIA, 0.84° C (± 0.13 S.E.) in Group IIB, and 0.89° C (± 0.16 S.E.) in Group IIC. In Group III, the rates were: 0.96° C (± 0.14 S.E.) in Group IIIA, 0.54° C (± 0.14 S.E.) in Group IIIB
and 0.49° C (± 0.19 S.E.) in Group IIIC (see Fig. 3). Statistical analysis of variance, testing temperature by time during the postischemic arterial and venous occlusion and during reperfusion, demonstrated no significant difference among any of the B and C groups. The mean pH value before clamping of the artery was 7.37 ± 0.03 U in all muscles studied. In the first 30 min of ischemia, the pH rapidly declined at a mean rate of 0.95 U/hr (± 0.05 S.E.) in all muscles. A further, slower, decline occurred in the following 90 min, during which the rate was 0.29 U/hr (± 0.03 S.E.) (Fig. 4). After 2 hr of ischemia, the mean pH was 6.45 ± 0.06 U, after 4 hr the mean was 6.33 ± 0.06 U, and after 7 hr, the mean was 6.19 ± 0.05 U. At the end of the venous stasis period, the pH began to increase in all groups. During the first hour of reperfusion, in Group IA the pH recovered at a rate of 0.59 U/hr (± 0.017 S.E.), and muscle pH in Group IB animals recovered at a rate of 0.37 U/hr (± 0.017 S.E.) after venous occlusion. In Group IC, after 3 hr of ischemia, pH increased at 0.39 U/hr (± 0.017 S.E.). In Group II, the pH recovered at a rate of 0.38 U/hr (± 0.017 S.E.) after 4 hr of ischemia (Group IIA), 0.28 U/hr (± 0.017 S.E.) after venous occlusion (Group IIB), and 0.31 U/hr (± 0.017 S.E.) after 5 hr of ischemia (Group IIC). In Groups IB, IC, IIB, and IIC, postischemic venous occlusion or one more hour of ischemia significantly delayed the pH recovery rate (p < 0.0001) during reperfusion, compared to Groups IA and IIA, respectively. In Group IIIA which underwent 6 hr of ischemia, pH increased 0.15 U/hr (± 0.019 S.E.) in the first hour of reperfusion. In Group IIIB, pH increased at 0.12 U/hr (± 0.018 S.E.) after postischemic venous occlusion, and in Group IIIC, the increase was 0.10 U/hr (± 0.019 S.E.) after 7 hr of ischemia (see Fig. 4).
Downloaded by: Georgetown University Medical Center. Copyrighted material.
blue staining of type I fibers and lighter blue staining of type II fibers. Damaged areas were unstained. Using a light microscope under low power (7x), the images were projected through a video camera (C.C.D. Color) to a video screen attached to a computer (N.E.C. video monitor to a Computer Innovation Corporation personal computer device). The total area and nonstained areas were quantified on each section using appropriately designed software for computerized planimetry. The percentage of necrotic area on each slide and the mean value for each muscle were calculated. All data from pH, histology, and temperature evaluations were analyzed using the ANOVA multifactorial test for the analysis of variance.
HISTOCHEMICAL EVALUATION. Areas of necrosis were quantified with computerized planimetry in a total of 240 sections. All data are expressed as the least square mean of the percentage of nonstained areas in all three sections from each group of muscles (Fig. 5). In Group IA, 5.05 percent (± 1.62 S.E.) of the total
1.6 1.4 1.2
1.0
C/hr o.8
I 1
A Ai
Reperfusion
7.2
Venous Occlusion
7.0 6.8
0.6
;:
1 -S^^:::'l
0.4 0.2 _ . . . . . . . . . Reperfusion mmm mmm ~ ~ Venous Occlusion i i i i 0.0 0 1 2 3 4
Ischemia
7.4
6.6
JF
6.4 i
i
5
i
6
i
7
i
8
Time (Hrs) Figure 3. The rate of temperature changes in ° C/hour. Nonsignificant difference was found among all B and C groups during reperfusion (p = .2462).
6.2
0
1
2
3
4
5
6
7
Time (Hrs) Figure 4. pH rate changes measured in U/hr.
273
50 Ischemia 40
Venous Occlusion
e 30 u & 20 10
1
2
3
4 5 6 7 8 Time (Hrs) Figure 5. Mean percent of necrotic areas (± S.E measured in all three muscle sections of each group.
cross-sectional area was unstained. The unstained area in Group IB animals with venous occlusion averaged 19.5 percent (± 1.58 S.E.), and in Group IC after 3 hr of ischemia, 18.9 percent (± 1.68 S.E.). The unstained area in Groups IB and IC differed significantly from Group IA (p < 0.0001). In Group II, muscles with 4 hr of ischemia (Group IIA) had 24.13 percent (± 1.58 S.E.) unstained area, while postischemic venous occlusion muscles (Group IIB) showed a damaged area of 35.28 percent (± 1.50 S.E.). After 5 hr of ischemia (Group IIC), 30.19 percent (± 1.68 S.E.) of the total area was unstained. In Group IIIA, 6 hr of ischemia resulted in a 43.08 percent (± 1.77 S.E.) unstained area. After postischemic venous occlusion (Group IIIB), 40.98 percent (± 1.77 S.E.) of the muscle failed to stain, and after 7 hr of ischemia (Group IIIC), 47.91 percent (± 1.69 S.E.) of the muscle area failed to stain. Histochemical evaluation data from the venous occlusion Groups IB, IIB, and IIIB, and the extended ischemia Groups IC, IIC, and IIIC, were also analyzed pairwise. The percentage of nonstained areas after 2 hr of ischemia and 1 hr of venous occlusion was comparable to that resulting from 3 hr of ischemia (p = 0.8057). The difference in percentage of damaged area between muscles subjected to 4 hr of ischemia and 1 hr of venous occlusion and those subjected to 5 hr of ischemia was also not significant (p = 0.0284). In contrast, 7 hr of ischemia created significantly higher damage to the muscles, compared to that caused by 6 hr of ischemia and 1 hr of venous occlusion [p < 0.0064), (see Fig. 5).* A qualitative analysis of the histologic character-
*In the last three tests, significance was evaluated with an adjusted p rate according to the Bonferoni adjustment rule: a
274
=
0.05
= 0.0167
number of tests 3 3 = adjusted probability rate
JULY 1990
istics was performed from the histochemically stained sections and additional sections stained with hematoxylin and eosin. Apart from the nonstained lethally damaged areas (Fig. 6A), interstitial tissue edema and cellular infiltration were constant findings in all muscles that sustained ischemia or ischemia and subsequent venous occlusion. However, both findings were more pronounced in the muscles that underwent venous obstruction. Muscle fibers were more edematous after venous occlusion, compared to ischemic muscles. Thrombosed veins and collapsed arteries (Fig. 7A-F), surrounded by excessive edema and inflammatory reaction, were common in muscles after ischemia and venous occlusion. The type II muscle fibers showed different staining characteristics after ischemia. They appeared more swollen and were more weakly stained, compared to the type I fibers, as the ischemia was prolonged (see Fig. 6B and 7A-F).
DISCUSSION The isolated in vivo quadriceps muscle of the rat provides a suitable experimental model 5 in which to measure both the ischemic and reperfusion components of skeletal muscle injury with subsequent venous occlusion. Unlike the tourniquet ischemia model, there is no soft-tissue injury, no fascial compartmentalization, and no sources of blood flow except that coming from the pedicle. In addition, systemic toxicity is significantly less than that caused by the tourniquet or the amputated hind limb ischemic models. 511 To create conditions similar to those existing in amputated limbs or elective transfers, we denervated the muscles in ourstudy. Erikssonetd. and Fleming1213 agree that denervation eliminates sympathetic control from the precapillary arterioles and increase resting blood flow. Reed14 demonstrated that postvenous stasis edema occurred in denervated and immobilized skeletal muscles at lower venous pressures, compared to muscles with intact innervation. The nitroblue tetrazolium dye employed for the histochemical detection of damaged areas in our study is an oxidation-reduction indicator that, on reduction, produces a dark blue formazan. This is bound firmly to the normal tissue in frozen sections and contrasts with the nonstained necrotic areas. 1516 The dye is reduced by the substrate through the action of the tissue diaphorares and dehydrogenases, that are inactive in dead cells (see Fig. 6A). Electron microscopy has demonstrated that NBT accurately discriminates between viable and necrotic skeletal muscle and spatially resolves to 1 mm, even at the border of the necrotic zone. This allowed us to use computerized planimetry to quantify the extent of tissue necrosis. 1516 Other investigators have found that cellular en-
Downloaded by: Georgetown University Medical Center. Copyrighted material.
JOURNAL OF RECONSTRUCTIVE MICROSURGERY/VOLUME 6, NUMBER 3
VESSEL OCCLUSION AFTER ISCHEMIA/MAUZOS, SEABER, URBANIAK
•••y
Qb>
;• I
B
zyme depletion persists for two days after an ischemic episode. To avoid false staining from inadequate reperfusion, in this study we allowed reperfusion of the muscle up to 48 hr.15-17 Temperature measurements in this model revealed poor correlation between the alterations of blood flow during ischemia and venous occlusion, possibly due to the containment of the muscle in a warmer tissue bed. During the first hour of reperfusion, the temperature increased faster, but no significant differences were found. These findings indicate that temperature monitoring is an unreliable method for the assessment of blood flow in elective muscle transfers. Raskin18 demonstrated similar findings that were attributed to the wide surface of the grafts and the warmer underlying tissues. Ischemic acidosis in our study, as reflected by pH,19 was profound and directly related to the length of ischemia. The onset of ischemic acidosis in the first 30 min was three times faster, compared to the following 90 min of ischemia. During the hour of venous stasis after ischemia, pH stopped declining (see Fig. 4). A probable explanation of this change may be that the hydrogen ion concentration in the muscle decreased from dilution, due to the temporary arterial inflow in the muscle after arterial clamp release, with venous drainage obstructed. The pH curve was reproducible and reflected the alterations of the blood circulation in the muscles from ischemia to venous stasis and the early reperfusion period (see Fig. 4). Many investigators18-22 describe pH as an indicator of perfusion disturbances in the skeletal muscles and suggest pH monitoring for evaluation of the adequacy of perfusion in elective tissue transfer. The significance and reliability of the findings in the early reperfusion period, and their correlation with the final outcome of revascularized tissues, have been extensively investigated.23-2527 In our study, the duration of ischemia affected the recovery rate of pH in the muscles. Longer ischemia was followed by slower recovery of the pH during the first hour of reperfusion
and, in turn, by more extended lethal damage (see Fig. 5). Histochemical evaluation of the muscle revealed no harmful effect from 2 hr of arterial ischemia, as has been shown in many other studies. However, either 1 hr of venous occlusion or extension of ischemia for one more hour (Groups IB and IC) caused more severe damage, approaching one-fifth of the examined tissue (see Fig. 5). The percentage of necrotic tissue following ischemia combined with venous stasis was comparable to that of equally extended ischemic insult. After 4 hr of ischemia alone, the muscle sustained lethal damage in up to one-fourth of its mass. Damage was further increased from temporary venous stasis in an amount comparable to that caused by an equal period of additional ischemia (see Fig. 5). The six-hour ischemia group demonstrated a high percentage of lethal tissue damage. Although 1 hr of venous stasis subsequent to ischemia failed to increase the damage, an equal period of arterial ischemia extended the damage significantly (see Fig. 5). Qualitative findings in the stained areas revealed excessive intersitial edema, cellular infiltration and a heterogenous reaction of the type II muscle fibers that were more weakly stained or unstained. The degree of abnormality was proportional to the duration of the ischemic episode. 23 These findings are in agreement with the reports of Jennishe et al.28 and Sahlin et al2A In muscles that underwent 1 hr of venous occlusion rather than an equivalent additional ischemic period, these findings were more pronounced, accompanied by numerous thrombosed veins and collapsed arteries (see Figs. 6B, 7A, 7F). The necrotic areas tended to be located in the core of the muscle and extended more proximally, with no significant difference in damage among the levels at which the sections were obtained. 2326 Walker et al.& and Labbe et al26 found similar patterns of damage in isolated muscle models. Wright and colleagues demonstrated a milder insult in the distal part of the gracilis muscle in dogs, attributed to the lower temperature of that part during ischemia.7
Downloaded by: Georgetown University Medical Center. Copyrighted material.
Figure 6. A, Nonstained necrotic muscle (NBT, x 10). B, Different staining characteristics for type II fibers (swollen, weakly or nonstained), excessive edema, and cellular infiltration (NBT, x 40). Type I fibers are stained darker (4 hr ischemia).
275
JOURNAL OF RECONSTRUCTIVE MICROSURGERY/VOLUME 6, NUMBER 3
JULY 1990
•
•
A COMPARATIVE STUDY OF THE EFFECT OF ARTERIAL AND Downloaded by: Georgetown University Medical Center. Copyrighted material.
VENOUS OCCLUSION AFTER VARIOUS PERIODS OF ISCHEMIA ABSTRACT The author attempted to quantify any additional tissue damage resulting from one hour of venous occlusion after controlled periods of ischemia in skeletal muscle, compared to equal time periods of ischemia. The rat quadriceps muscle was used. pH and temperature were monitored during the experiment and the damage was evaluated on histochemical sections. Of the other two parameters monitored in this study, the pH recovery rate during reperfusion best reflected tissue perfusion and served as a predictor for the extent of tissue damage, when global ischemia or the combination of ischemia and postischemic venous stasis did not exceed five hours. Apart from the nonstained lethally damaged areas, interstitial tissue edema and cellular infiltration were constant findings in all muscles that sustained ischemia or ischemia and subsequent venous occlusion. However, both findings were more pronounced in the muscles that underwent venous obstruction. Muscle fibers were more edematous after venous occlusion, compared to ischemic muscles. Thrombosed veins and collapsed arteries, surrounded by excessive edema and inflammatory reaction, were common in muscles after ischemia and venous occlusion. Comparing the difference in percent of lethal damage from one hour of postischemic venous stasis versus the damage from additional ischemia of equal duration, findings demonstrated comparable damage when reperfusion was established in less than five hours.
Venous stasis is a common complication in emergency or elective microsurgery and may occur either as a thrombotic complication soon after revascularization 1 or secondary to clamping for anastomosis of the vein after the arterial flow has been reestablished and the artery is still supplying blood. Vascular congestion of any cause compromises tissue perfusion in a normal muscle. 2 - 4 In an unconstricted skeletal muscle, 5 hr of independent venous occlusion has been shown to produce damage similar to 3 hr of ischemic injury.5 Prolonged antecedent ischemia leads to endothelial damage, vessel wall impairment, clotting disturbances, and parenchymal tissue damage. During reperfusion, the ischemic injury is further exacerbated by
the production of short-lived but highly reactive oxygen free radicals that can affect membrane integrity and permeability.^ 1 0 2 9 3 0 3 2 In clinical situations, such as after major limb revascularization or elective muscle transfer, venous drainage obstruction occurring in a tissue that has previously sustained an ischemic insult may often lead to a poor functional outcome. The extent of the damage from a temporary postischemic venous stasis has been related to the extent of the antecedent ischemia and the duration of vascular congestion, but the duration of additional ischemia necessary to produce the same results as various periods of venous stasis has not been determined.
Orthopaedic Research Laboratories, Department of Surgery, Duke University Medical Center, Durham, North Carolina Reprint requests-. Anthony V. Seaber, Dept. of Surgery, P.O. Box 3093, Duke University Medical Center, Durham, NC 27710 Accepted for publication lanuary 29, 1990 Copyright © 1990 by Thieme Medical Publishers, Inc., 381 Park Avenue South, New York, NY 10016. All rights reserved.
271
The purpose of this study was to quantify any additional damage resulting from 1 hr of venous occlusion in skeletal muscles after controlled periods of ischemia, and to compare this to the damage caused by equal time periods of additional ischemia. The rat quadriceps muscle was used. pH and temperature were monitored during the experiment and the damage was evaluated on histochemical sections.
MATERIALS AND METHODS An isolated quadriceps muscle model was used in 80 male Sprague Dawley rats (280 to 400 g). Each animal was anesthetized with an intraperitoneal injection of pentobarbital (50 mg/kg b.w). The quadriceps was approached through an anteromedial incision from the knee to the ipsilateral groin. The neurovascular pedicle of the muscle was isolated under a high-power operating microscope, and the femoral nerve was divided. The quadriceps was separated from all surrounding tissues, including both the origin and insertion sites, thus isolating the muscle on its vascular pedicle and excluding any other vascular supply. After the dissection was completed, the muscle was placed into its original bed, with each end sutured to its origin or insertion (Fig. 1). Temperature was monitored using a needle probe connected to an OMEGA thermistor thermometer. The tip of the probe was inserted through the distal end of the muscle, up to its middle third (see Fig. 1). Intramuscular pH was measured with a glass needle microelectrode (MI 408B, Microelectrodes, Inc.) inserted in the lateral side of the muscle to a depth of 1 cm. A skin electrode was used as a reference and both were connected to an ORION 407A Ionalyzer (see Fig. 1). A microclamp was placed on the nutrient artery of the quadriceps and a patency test was performed to
JULY 1990
verify total ischemia. The skin was closed with subcuticular sutures, thereby maintaining the muscle in the normothermic environment of the body. The approximate time required for this preparation was 30 min. During venous occlusion, the arterial clamp was released to supply blood to the quadriceps. Three initial ischemic periods of different duration were studied, as outlined in Figure 2. In Group IA, ten animals underwent 2 hr of arterial ischemia. In Group IB, an equal number of animals had 1 hr of venous occlusion, in addition to the initial 2-hr ischemia period. Eight more animals in Group IC underwent 3 hr of ischemia. Group IIA animals were subjected to an ischemic period of 4 hr (n = 10). In Group IIB, ten animals received 4 hr of ischemia and one hour of venous occlusion. The eight animals in Group IIC underwent an additional hour of ischemia for a total of 5 hr. Initial ischemia lasted 6 hr in Group IIIA animals (n = 8), while the venous occlusion Group IIIB animals had 6 hr of ischemia and 1 hr of venous occlusion (n = 8). Eight more animals in Group IIIC were subjected to an ischemia period of 7 hr (see Fig. 2). Temperature and pH were monitored every 15 min with needle probes during ischemia, venous occlusion, and the first hour of reperfusion. All animals were sacrificed 48 hr after reperfusion. At sacrifice, the quadriceps muscles were removed from the animal and sectioned into three pieces (proximal, middle, and distal third). A frozen section was obtained from each third (see Fig. 2). The sections were incubated in nitroblue tetrazolium (NBT) dye with substrate NADH for 30 min at 37° C. Normal areas of the muscle showed dark
I
Groups
II
III
A
B
c
A
B
C
A
B
2
2
2
4
4
4
6
6
Animals (No.) Ischemia (Hrs.)
8
Venous Occlusion (Hrs.)
1
Arterial Occlusion (Hrs.) Reperfusion established after (Hrs.)
C
1 2
t
i
1 4
5
6
1
1 5
1 6
7
7
• Temperature and pH monitored every 15 minutes through the first hour of reperfusion. • Reperfusion lasted 48 hours. • Muscle damage was quantified on histochemically stained sections with computerized planimetry.
Prox.
272
Figure 1. After complete dissection and evaluation to eliminate any other source of blood supply except from its pedicle, the quadriceps was replaced and resutured at its resting length and the nerve was sectioned. Temperature and pH probes measured the parameters in the core of the Figure 2. Diagram presenting the material methodolmiddle third of the muscle. ogy and muscle sectioning.
Downloaded by: Georgetown University Medical Center. Copyrighted material.
JOURNAL OF RECONSTRUCTIVE MICROSURGERY/VOLUME 6, NUMBER 3
VESSEL OCCLUSION AFTER ISCHEMIA/MALIZOS, SEABER, URBANIAK
RESULTS Analysis of intramuscular temperature measurements obtained during the different flow conditions in the muscle showed an increase of 0.69° C/hr (± 0.12 Standard Error [S.E.j) in the first 2 hr of ischemia, 0.65° C/hr (±0.12 S.E.) between the second and fourth hour, 0.43° C/hr (± 0.12 S.E.) from the fourth to sixth hour, and 0.32° C/hr (± 0.18 S.E.) during the seventh hour of ischemia (Fig. 3). During the hour of venous occlusion, temperature increased 0.59° C (± 0.15 S.E.) in Group IC, 0.49° C (± 0.27 S.E.) in Group IIB, and 0.36° C (± 0.18 S.E.) in Group IIIB animals. In the first hour of reperfusion, the temperature increased at a rate of 1.21° C (± 0.18 S.E.) in Group IA, 1.05° C (± 0.13 S.E.) in Group IB venous occlusion muscles, and 1.11° C (± 0.16 S.E.) in Group IC 3-hr ischemia muscles. In Group II, the rates of temperature increase were: 1.08° C (± 0.13 S.E.) in Group IIA, 0.84° C (± 0.13 S.E.) in Group IIB, and 0.89° C (± 0.16 S.E.) in Group IIC. In Group III, the rates were: 0.96° C (± 0.14 S.E.) in Group IIIA, 0.54° C (± 0.14 S.E.) in Group IIIB
and 0.49° C (± 0.19 S.E.) in Group IIIC (see Fig. 3). Statistical analysis of variance, testing temperature by time during the postischemic arterial and venous occlusion and during reperfusion, demonstrated no significant difference among any of the B and C groups. The mean pH value before clamping of the artery was 7.37 ± 0.03 U in all muscles studied. In the first 30 min of ischemia, the pH rapidly declined at a mean rate of 0.95 U/hr (± 0.05 S.E.) in all muscles. A further, slower, decline occurred in the following 90 min, during which the rate was 0.29 U/hr (± 0.03 S.E.) (Fig. 4). After 2 hr of ischemia, the mean pH was 6.45 ± 0.06 U, after 4 hr the mean was 6.33 ± 0.06 U, and after 7 hr, the mean was 6.19 ± 0.05 U. At the end of the venous stasis period, the pH began to increase in all groups. During the first hour of reperfusion, in Group IA the pH recovered at a rate of 0.59 U/hr (± 0.017 S.E.), and muscle pH in Group IB animals recovered at a rate of 0.37 U/hr (± 0.017 S.E.) after venous occlusion. In Group IC, after 3 hr of ischemia, pH increased at 0.39 U/hr (± 0.017 S.E.). In Group II, the pH recovered at a rate of 0.38 U/hr (± 0.017 S.E.) after 4 hr of ischemia (Group IIA), 0.28 U/hr (± 0.017 S.E.) after venous occlusion (Group IIB), and 0.31 U/hr (± 0.017 S.E.) after 5 hr of ischemia (Group IIC). In Groups IB, IC, IIB, and IIC, postischemic venous occlusion or one more hour of ischemia significantly delayed the pH recovery rate (p < 0.0001) during reperfusion, compared to Groups IA and IIA, respectively. In Group IIIA which underwent 6 hr of ischemia, pH increased 0.15 U/hr (± 0.019 S.E.) in the first hour of reperfusion. In Group IIIB, pH increased at 0.12 U/hr (± 0.018 S.E.) after postischemic venous occlusion, and in Group IIIC, the increase was 0.10 U/hr (± 0.019 S.E.) after 7 hr of ischemia (see Fig. 4).
Downloaded by: Georgetown University Medical Center. Copyrighted material.
blue staining of type I fibers and lighter blue staining of type II fibers. Damaged areas were unstained. Using a light microscope under low power (7x), the images were projected through a video camera (C.C.D. Color) to a video screen attached to a computer (N.E.C. video monitor to a Computer Innovation Corporation personal computer device). The total area and nonstained areas were quantified on each section using appropriately designed software for computerized planimetry. The percentage of necrotic area on each slide and the mean value for each muscle were calculated. All data from pH, histology, and temperature evaluations were analyzed using the ANOVA multifactorial test for the analysis of variance.
HISTOCHEMICAL EVALUATION. Areas of necrosis were quantified with computerized planimetry in a total of 240 sections. All data are expressed as the least square mean of the percentage of nonstained areas in all three sections from each group of muscles (Fig. 5). In Group IA, 5.05 percent (± 1.62 S.E.) of the total
1.6 1.4 1.2
1.0
C/hr o.8
I 1
A Ai
Reperfusion
7.2
Venous Occlusion
7.0 6.8
0.6
;:
1 -S^^:::'l
0.4 0.2 _ . . . . . . . . . Reperfusion mmm mmm ~ ~ Venous Occlusion i i i i 0.0 0 1 2 3 4
Ischemia
7.4
6.6
JF
6.4 i
i
5
i
6
i
7
i
8
Time (Hrs) Figure 3. The rate of temperature changes in ° C/hour. Nonsignificant difference was found among all B and C groups during reperfusion (p = .2462).
6.2
0
1
2
3
4
5
6
7
Time (Hrs) Figure 4. pH rate changes measured in U/hr.
273
50 Ischemia 40
Venous Occlusion
e 30 u & 20 10
1
2
3
4 5 6 7 8 Time (Hrs) Figure 5. Mean percent of necrotic areas (± S.E measured in all three muscle sections of each group.
cross-sectional area was unstained. The unstained area in Group IB animals with venous occlusion averaged 19.5 percent (± 1.58 S.E.), and in Group IC after 3 hr of ischemia, 18.9 percent (± 1.68 S.E.). The unstained area in Groups IB and IC differed significantly from Group IA (p < 0.0001). In Group II, muscles with 4 hr of ischemia (Group IIA) had 24.13 percent (± 1.58 S.E.) unstained area, while postischemic venous occlusion muscles (Group IIB) showed a damaged area of 35.28 percent (± 1.50 S.E.). After 5 hr of ischemia (Group IIC), 30.19 percent (± 1.68 S.E.) of the total area was unstained. In Group IIIA, 6 hr of ischemia resulted in a 43.08 percent (± 1.77 S.E.) unstained area. After postischemic venous occlusion (Group IIIB), 40.98 percent (± 1.77 S.E.) of the muscle failed to stain, and after 7 hr of ischemia (Group IIIC), 47.91 percent (± 1.69 S.E.) of the muscle area failed to stain. Histochemical evaluation data from the venous occlusion Groups IB, IIB, and IIIB, and the extended ischemia Groups IC, IIC, and IIIC, were also analyzed pairwise. The percentage of nonstained areas after 2 hr of ischemia and 1 hr of venous occlusion was comparable to that resulting from 3 hr of ischemia (p = 0.8057). The difference in percentage of damaged area between muscles subjected to 4 hr of ischemia and 1 hr of venous occlusion and those subjected to 5 hr of ischemia was also not significant (p = 0.0284). In contrast, 7 hr of ischemia created significantly higher damage to the muscles, compared to that caused by 6 hr of ischemia and 1 hr of venous occlusion [p < 0.0064), (see Fig. 5).* A qualitative analysis of the histologic character-
*In the last three tests, significance was evaluated with an adjusted p rate according to the Bonferoni adjustment rule: a
274
=
0.05
= 0.0167
number of tests 3 3 = adjusted probability rate
JULY 1990
istics was performed from the histochemically stained sections and additional sections stained with hematoxylin and eosin. Apart from the nonstained lethally damaged areas (Fig. 6A), interstitial tissue edema and cellular infiltration were constant findings in all muscles that sustained ischemia or ischemia and subsequent venous occlusion. However, both findings were more pronounced in the muscles that underwent venous obstruction. Muscle fibers were more edematous after venous occlusion, compared to ischemic muscles. Thrombosed veins and collapsed arteries (Fig. 7A-F), surrounded by excessive edema and inflammatory reaction, were common in muscles after ischemia and venous occlusion. The type II muscle fibers showed different staining characteristics after ischemia. They appeared more swollen and were more weakly stained, compared to the type I fibers, as the ischemia was prolonged (see Fig. 6B and 7A-F).
DISCUSSION The isolated in vivo quadriceps muscle of the rat provides a suitable experimental model 5 in which to measure both the ischemic and reperfusion components of skeletal muscle injury with subsequent venous occlusion. Unlike the tourniquet ischemia model, there is no soft-tissue injury, no fascial compartmentalization, and no sources of blood flow except that coming from the pedicle. In addition, systemic toxicity is significantly less than that caused by the tourniquet or the amputated hind limb ischemic models. 511 To create conditions similar to those existing in amputated limbs or elective transfers, we denervated the muscles in ourstudy. Erikssonetd. and Fleming1213 agree that denervation eliminates sympathetic control from the precapillary arterioles and increase resting blood flow. Reed14 demonstrated that postvenous stasis edema occurred in denervated and immobilized skeletal muscles at lower venous pressures, compared to muscles with intact innervation. The nitroblue tetrazolium dye employed for the histochemical detection of damaged areas in our study is an oxidation-reduction indicator that, on reduction, produces a dark blue formazan. This is bound firmly to the normal tissue in frozen sections and contrasts with the nonstained necrotic areas. 1516 The dye is reduced by the substrate through the action of the tissue diaphorares and dehydrogenases, that are inactive in dead cells (see Fig. 6A). Electron microscopy has demonstrated that NBT accurately discriminates between viable and necrotic skeletal muscle and spatially resolves to 1 mm, even at the border of the necrotic zone. This allowed us to use computerized planimetry to quantify the extent of tissue necrosis. 1516 Other investigators have found that cellular en-
Downloaded by: Georgetown University Medical Center. Copyrighted material.
JOURNAL OF RECONSTRUCTIVE MICROSURGERY/VOLUME 6, NUMBER 3
VESSEL OCCLUSION AFTER ISCHEMIA/MAUZOS, SEABER, URBANIAK
•••y
Qb>
;• I
B
zyme depletion persists for two days after an ischemic episode. To avoid false staining from inadequate reperfusion, in this study we allowed reperfusion of the muscle up to 48 hr.15-17 Temperature measurements in this model revealed poor correlation between the alterations of blood flow during ischemia and venous occlusion, possibly due to the containment of the muscle in a warmer tissue bed. During the first hour of reperfusion, the temperature increased faster, but no significant differences were found. These findings indicate that temperature monitoring is an unreliable method for the assessment of blood flow in elective muscle transfers. Raskin18 demonstrated similar findings that were attributed to the wide surface of the grafts and the warmer underlying tissues. Ischemic acidosis in our study, as reflected by pH,19 was profound and directly related to the length of ischemia. The onset of ischemic acidosis in the first 30 min was three times faster, compared to the following 90 min of ischemia. During the hour of venous stasis after ischemia, pH stopped declining (see Fig. 4). A probable explanation of this change may be that the hydrogen ion concentration in the muscle decreased from dilution, due to the temporary arterial inflow in the muscle after arterial clamp release, with venous drainage obstructed. The pH curve was reproducible and reflected the alterations of the blood circulation in the muscles from ischemia to venous stasis and the early reperfusion period (see Fig. 4). Many investigators18-22 describe pH as an indicator of perfusion disturbances in the skeletal muscles and suggest pH monitoring for evaluation of the adequacy of perfusion in elective tissue transfer. The significance and reliability of the findings in the early reperfusion period, and their correlation with the final outcome of revascularized tissues, have been extensively investigated.23-2527 In our study, the duration of ischemia affected the recovery rate of pH in the muscles. Longer ischemia was followed by slower recovery of the pH during the first hour of reperfusion
and, in turn, by more extended lethal damage (see Fig. 5). Histochemical evaluation of the muscle revealed no harmful effect from 2 hr of arterial ischemia, as has been shown in many other studies. However, either 1 hr of venous occlusion or extension of ischemia for one more hour (Groups IB and IC) caused more severe damage, approaching one-fifth of the examined tissue (see Fig. 5). The percentage of necrotic tissue following ischemia combined with venous stasis was comparable to that of equally extended ischemic insult. After 4 hr of ischemia alone, the muscle sustained lethal damage in up to one-fourth of its mass. Damage was further increased from temporary venous stasis in an amount comparable to that caused by an equal period of additional ischemia (see Fig. 5). The six-hour ischemia group demonstrated a high percentage of lethal tissue damage. Although 1 hr of venous stasis subsequent to ischemia failed to increase the damage, an equal period of arterial ischemia extended the damage significantly (see Fig. 5). Qualitative findings in the stained areas revealed excessive intersitial edema, cellular infiltration and a heterogenous reaction of the type II muscle fibers that were more weakly stained or unstained. The degree of abnormality was proportional to the duration of the ischemic episode. 23 These findings are in agreement with the reports of Jennishe et al.28 and Sahlin et al2A In muscles that underwent 1 hr of venous occlusion rather than an equivalent additional ischemic period, these findings were more pronounced, accompanied by numerous thrombosed veins and collapsed arteries (see Figs. 6B, 7A, 7F). The necrotic areas tended to be located in the core of the muscle and extended more proximally, with no significant difference in damage among the levels at which the sections were obtained. 2326 Walker et al.& and Labbe et al26 found similar patterns of damage in isolated muscle models. Wright and colleagues demonstrated a milder insult in the distal part of the gracilis muscle in dogs, attributed to the lower temperature of that part during ischemia.7
Downloaded by: Georgetown University Medical Center. Copyrighted material.
Figure 6. A, Nonstained necrotic muscle (NBT, x 10). B, Different staining characteristics for type II fibers (swollen, weakly or nonstained), excessive edema, and cellular infiltration (NBT, x 40). Type I fibers are stained darker (4 hr ischemia).
275
JOURNAL OF RECONSTRUCTIVE MICROSURGERY/VOLUME 6, NUMBER 3
JULY 1990
•
•
