Damage incurred during ischemia and reperfusion is a major cause of acute tubular necrosis (ATN*) in transplanted kidneys (1). ATN may lead to acute renal failure, demanding a return to dialysis with a corresponding increase in costs or ultimately in retransplantation. In kidneys subjected to brief periods of ischemia, tubular necrosis is restricted mainly to the medullary thick ascending tubule and the pars recta (S3 segment) of the proximal tubule (2-4). Such damage seems to be a corollary of secondary ischemia during reflow, deriving from vascular erythrocyte congestion in the outer medulla (2-4). However, hemodilution has been shown to ameliorate the severity of medullary congestion without preventing necrosis of the medullary thick ascending tubule or the pars recta(5). Moreover, after prolonged ischemic insult, ATN becomes more widespread and involves other sections of the tubule, particularly the proximal S1 segment (6). Because superficial cortical blood flow is relatively unaffected even after prolonged ischemia (7,8), widespread cortical necrosis does not seem to be solely dependent on secondary ischemia.
Cultured proximal tubule cells subjected to hypoxia have been shown to be susceptible to free radical-mediated oxidative injury upon reoxygenation (9). Direct detection of free radicals by electron paramagnetic spin resonance studies have confirmed that free radicals are generated in vivo upon reperfusion of ischemic livers (10). That free radical-mediated injury may play a role in tubular necosis after renal transplantation is supported by much circumstantial evidence of improvements in tubular function or morphology as a result of antioxidant therapies (11-14). The segments of the proximal tubule seem to be differentially susceptible to free radical-mediated injury, the most susceptible being those most dependent on mitochondrial phosphorylation (6, 15). Primary function after ischemia seems to be dependent on mitochondrial capacity to resynthesize adenosine 5′-triphosphate (ATP) rather than postischemic ATP levels per se (16). Cardiac mitochondrial complex 1 activity has been shown to be unaffected by hypoxia alone but is significantly decreased on reoxygenation (17). Similarly, electron microscopical analysis of mitochondria in the S1 segment of the proximal tubule suggests that structural integrity is not compromised before reperfusion of rabbit kidneys (18). After 72 hours of cold ischemia, measurements of mitochondrial NADH fluorescence by surface fluorescence spectroscopy and of the oxidation state of cytochrome oxidase by near infrared spectroscopy suggested that severe mitochondrial dysfunction developed during the first 5 minutes of reperfusion (18, 19). However, it was not possible to determine whether the mitochondrial damage observed after reperfusion was a consequence of oxidative injury or of secondary ischemia during reflow.
The objective of this study was to determine the kinetics of hemoglobin oxygenation during in situ reperfusion of stored transplanted kidneys, to provide a framework against which the timing of renal metabolic injury could be considered. We have used the well established New Zealand White rabbit renal autograft model (20), in which nephrectomized kidneys were flushed with cold (0-2°C) hypertonic citrate solution (HCA), which is routinely used in clinical renal transplantation, and either autografted immediately (group 1) or stored at 0-2°C for 72 hours before autografting (group 2). From previous work in this model, at least 95% of group 1 recipients would be expected to survive with good renal function, but only 20-30% would be viable in group 2 (20). Direct noninvasive measurements of concentration changes in renal oxyhemoglobin and deoxyhemoglobin were made using near infrared spectroscopy (21-24) (see Methods). In addition, the distribution of infused barium sulfate (BaSO4), which gives a strikingly visual representation of regional vascular resistance within the kidney, was analyzed at discrete time points by angiography and phase contrast microscopy. Renal morphology was evaluated by light microscopy.
All animal procedures were performed according to Home Office regulations, Animals (Scientific Procedures) Act, 1986.
Anesthesia/physiological monitoring/surgical procedures. A total of 44 adult female New Zealand White rabbits (2.5-3 kg) in 2 groups of 22 animals were deeply sedated with an i.m. injection of ketamine (50 mg/kg) and xylazine (8 mg/kg), tracheotomized and artificially ventilated with a 50:50 oxygen:nitrous oxide mixture for periods of up to 4 hours. Surgical anesthesia was maintained by continuous i.v. infusion of ketamine and xylazine (50 mg:8 mg/kg/hour). The left femoral artery was carefully dissected and cannulated with an 18-gauge plastic cannula, which was inserted for a distance of 5 cm until the tip reached the base of the aorta. The cannula provided protective housing for a Continucath oxygen electrode (Bio-medical Sensors Ltd., High Wycombe, UK) coupled via a 3-way tap to a pressure transducer. Blood pressure and pO2 were thus continuously monitored throughout the study period. Blood pressure was maintained during reperfusion by continuous i.v. infusion of Haemaccel (30 ml/hour). Intermittent blood samples were taken for pCO2, pH, and hematocrit determination. Core temperature was monitored using an esophageal probe and was maintained between 37-39°C with a heated pad (37°C). In addition, the electrocardiogram, FiO2, and EtCO2 were continuously monitored. The transplant procedures have been described previously (20). Briefly, the right kidney was harvested on as long a renal pedicle as possible via a midline laparotomy, flushed with ice-chilled HCA and stored. When kidneys were to be stored for 72 hours, the donor rabbit was allowed to recover during the intervening 3 days before autografting. At the time of autografting, the left kidney was removed leaving the recipient renal pedicle as long as possible together with the ureter divided close to the kidney. The right kidney was then autografted onto the left pedicle using standard microsurgical techniques.
Experimental groups. Group 1: Freshly nephrectomized right kidneys were flushed with 40 ml cold (1-2°C) HCA (Baxter Health Care, Thetford, Norfolk, UK) via the renal artery and autografted immediately into the left renal bursa using standard microsurgical techniques. Group 1a: 6 autografted kidneys were monitored continuously in situ during reperfusion by near-infrared spectroscopy (NIRS); group 1b: 16 kidneys were perfused with micro-opaque BaSO4 either before reperfusion or after 5, 60, or 180 minutes of reperfusion (see below).
Group 2: Kidneys were flushed as in group 1 and then stored in HCA surrounded by ice to maintain a temperature of 1-2°C for 72 hours before autografting. Groups 2a and 2b were then subjected to the same protocol as groups 1a and 1b.
NIRS measurements of renal hemodynamics. NIRS measurements were made using a NIRO-500 monitor (Hamamatsu). This technique has been used for measurements of cerebral oxygenation for many years (25, 26); however, its use in renal physiology has been limited. NIRS transmission spectroscopy through tissues up to 10 cm is possible because of the relative transparency of biological tissues to NIR light (700-1000 nm). Absorption caused by oxyhemoglobin (HbO2) and deoxyhemoglobin (Hb) can be quantified using a modified form of the Beer-Lambert Law, which is applicable to a homogenous scattering medium in which absorption changes are linearly related to concentration, and which may be expressed as: Equation
OD is the optical density, Σ is the extinction coefficient of the chromophore (mM-1 cm-1), c is concentration (mM), L is the difference between the points where light enters and leaves the tissue (cm), B is a pathlength factor that takes account of the scattering of light in the tissue (which causes the optical pathlength to be greater than L) and G is a factor related to the geometry of the tissue. If L, B, and G remain constant, then changes in chromophore concentration can be obtained from the expression: Equation
An algorithm has been determined for the quantification of oxyhemoglobin and deoxyhemoglobin following measurements on infants and experimental animals (27). The extinction coefficients of HbO2 and Hb were obtained from studies of lysed human blood (28). However, although the optical pathlength (B) has been determined for the cerebellum by time of flight measurements of an ultra-short pulse of light across the head, there have been no estimates of B in kidneys. Because an estimate of B was not made in the present study, only relative changes in chromophore concentration (in μM) were measured across the kidney. However, because we have compared 2 groups of kidneys subjected to the same treatment and the optical pathlength (and differential pathlength factor) has been shown to be constant within the NIR region despite gross changes in oxygenation and perfusion before and after death (27), the relative changes in chromophore concentration determined in this study are directly comparable between groups.
The NIR probes were placed on either side of the kidney at a distance (L) of 1 cm and immobilized using a retort stand, to give an indication of hemoglobin oxygenation across both the cortex and medulla. Concentration changes in HbO2 and Hb across the kidney were calculated by linear summation of absorption changes (measured in absorbance units) at 772 nm, 830 nm, 842 nm and 909 nm, multiplied by factors obtained by spectroscopic and biochemical measurements of a wide range of hemoglobin concentrations in a scattering medium (27). Baseline measurements of HbO2 and Hb were measured every 5 seconds in situ for 5 minutes before reperfusion (i.e., before removal of the clamps but after completion of the anastomoses) and changes in these parameters were then measured during 3 hours of reperfusion. These values were used to calculate the following parameters: total hemoglobin (HbT = HbO2 + Hb); hemoglobin oxygenation index (HbD = HbO2 - Hb) and percentage mixed hemoglobin saturation (renal SaO2 = HbO2/HbT × 100). The use of this final parameter is dependent on changes from time 0, at which point the organ was blood free. Therefore, the maximum changes on reperfusion can be considered to be equivalent to total hemoglobin. Changes in concentration of oxidized cytochrome oxidase have been reported elsewhere (19).
BaSO4angiography. Kidneys were either transplanted immediately (group 1b) or stored for 72 hours before transplantation (group 2b). Kidneys were perfused via the renal artery with 1 ml of micro-opaque BaSO4 (50:50 w/v BaSO4:isotonic saline) from a bag suspended at a height of 1.5 m. The maximum perfusion pressure was therefore 90 mmHg. BaSO4 was infused immediately after flushing in group 1b or after 72 hours storage in group 2b (t=0); following transplantation after 5, 60, and 180 minutes of reperfusion in each group (n=4 for each time point). Kidneys were then sectioned and angiograms were taken at 22 kV with 30 seconds exposure. The angiograms were scored blind on a scale of 0 (no perfusion); 1 (minimal perfusion); 2 (patchy perfusion); 3 (wide-spread perfusion); and 4 (normal perfusion) in both the cortex and medulla.
Histology. Slices were taken from all kidneys soon after termination of reperfusion and fixed in 10% formol saline for processing to paraffin wax. Sections were stained with hematoxylin and eosin for light microscopical analysis. IR damage was assessed by several histological features including edema, congestion, hemorrhage, inflammation and necrosis in the cortex, corticomedullary junction and medulla. In addition, the microcirculatory distribution of barium sulfate particles was analyzed by phase contrast microscopy.
Statistics. In Figures 2 through 5, data are presented as mean ± SEM. Statistical significance within and between groups was tested by an unpaired Student's t test. In Figure 7, data are presented as mean ± range. Statistical differences in assessments of BaSO4 concentration between groups have been tested by Kruskal-Wallis 1-way analysis of variance to test the medians of all groups for equality followed by the Mann-Whitney U test to compare each group against group 1 controls (unstored untransplanted).
NIRS measurements of renal hemodynamics. In Figure 1, a representative time course plot of changes in concentration of oxyhemoglobin, deoxyhemoglobin, total hemoglobin, and the oxygenation index in a group 1 unstored transplanted kidney during 3 hours of reperfusion is shown. Baseline measurements of HbO2 and Hb were made for 5 minutes before reperfusion. Changes in HbO2 and Hb were then monitored for 3 hours of reperfusion in 6 group 1 (unstored) and 6 group 2 (72 hours stored) kidneys.
In Figure 2, the mean and SEM values for HbO2 at discrete time points in group 1 and group 2 kidneys are displayed. Reperfusion resulted in rapid increases in HbO2 within 1 minute in both groups; the rate of increase in group 1 kidneys was significantly greater (P<0.05). After 1 minute, there was a sharp decrease in HbO2 in group 1 kidneys from 1494±138 μM at 1 minute to 1077±224 μM at 3 minutes (P=0.053). Thereafter, HbO2 increased to 1482±285 μM at 20 minutes and did not change significantly during the remaining 140 minutes of reperfusion. In group 2 kidneys, HbO2 peaked at 1016±287 μM after 2 minutes of reperfusion and then decreased sharply to 590±86 μM at 5 minutes (P=0.059). The rate and magnitude of changes in HbO2 in the first 5 minutes of reperfusion of group 2 kidneys were smaller than group 1 kidneys, but the only significant difference occurred at 1 minute and the hemodynamic changes were similar to those in group 1 kidneys. However, there was no significant recovery in HbO2 after 5 minutes of reperfusion in group 2 kidneys and HbO2 was significantly lower compared with group 1 kidneys after 10 minutes of reperfusion (P<0.05). Thereafter, there was no significant change in group 2 HbO2, and the levels remained significantly lower than group 1 kidneys for the remainder of the reperfusion period tested.
In Figure 3, the mean and SEM changes in Hb at discrete time points in group 1 and group 2 kidneys are shown. There were no significant differences in the rates of increase of Hb in the first 2 minutes of reperfusion between the 2 groups. However, in group 1 kidneys, the Hb had plateaued by 3 minutes at 763±101 μM and dropped slowly thereafter to 473±138 μM at 3 hours. In group 2 kidneys, Hb continued to increase after 2 minutes to reach a maximum of 1756±322 μM at 10 minutes, a significantly greater concentration (P<0.05) than group 1 kidneys. Furthermore, unlike group 1 kidneys, the Hb in group 2 increased to levels significantly in excess of the HbO2 in the same kidneys by 10 minutes (P<0.05). Group 2 Hb remained higher than the HbO2 and the Hb in group 1 for the remainder of reperfusion. However, after 1 hour, group 2 Hb had fallen to 1123±243 μM, and the differences were no longer significant. After 1 hour of reperfusion, there were no further significant changes in Hb in either group.
Changes in total hemoglobin concentration (HbT) at discrete time points in group 1 and group 2 kidneys are shown in Figure 4. The rate of increase in HbT was significantly greater in group 1 compared to group 2 kidneys in the first minute of reperfusion; however, after 2 minutes and throughout the duration of reperfusion, there were no significant differences in HbT between the 2 groups. Nevertheless, there was a fall in HbT in group 2 kidneys from a peak at 10 minutes of 2465±317 μM to 1647±293 μM after 1 hour, which paralleled the decrease in Hb over the same period. Figure 5 shows the percentage of renal hemoglobin saturation in group 1 and group 2 kidneys (HbO2/HbT×100; see Methods). In both groups, saturation reached 69% (±2% and 1%, respectively) after 1 minute of reperfusion. In group 1 kidneys, there was a transient decrease in saturation to 57%±9% after 3 minutes of reperfusion before recovery to 78%±6% at 1 hour. In group 2 kidneys, there was a profound hemoglobin desaturation to 25%±9% at 10 minutes (P<0.005 compared to group 1) and no significant recovery in the remaining 170 minutes. Changes in percent hemoglobin saturation in group 2 kidneys did not correlate with changes in HbT.
Intrarenal distribution of blood flow. BaSO4 distribution after infusion of a 1 ml bolus (50/50 w/v micro-opaque BaSO4/isotonic saline) via the renal artery in (1) an unstored untransplanted kidney; (2) a transplanted 72 hours stored, 1 hour reperfused kidney; and (3) a transplanted 72 hours stored, 3 hours reperfused kidney is shown in Figure 6. In Figure 6A, the cortical and medullary vasculature are discernible by BaSO4 angiography. The mottled appearance of the cortex results from BaSO4 particles trapped in the glomeruli. The striations in the medulla are a result of BaSO4 in the vasa recta, which delineate the urinary tract. Phase contrast microscopy revealed that the BaSO4 particles remained within the vasculature rather than the urinary tract even in severely damaged kidneys. BaSO4 perfusion was scored as 4 (well perfused) in both the cortex and medulla. In Fig. 6B), cortical BaSO4 perfusion was scored at 3-4 and was not apparently different from unstored unreperfused controls. However, BaSO4 particles did not penetrate the medullary vasculature at a constant perfusion pressure of 90 mmHg (scored at 0-1). After 3 hours of reperfusion (Fig. 6C), cortical perfusion was good (scored at 3-4) and medullary perfusion was returning to normal (scored at 3).
The mean and range of BaSO4 scores for kidneys subjected to varying lengths of reperfusion are shown in Figure 7 (n=4 at each time point). Unstored transplanted (group 1) kidney scores have been omitted because there were no significant changes during the 3 hours of reperfusion, nor were there any significant disturbances in cortical perfusion (as judged by BaSO4 angiography) at any time point in 72 hours stored transplanted kidneys (group 2). By contrast, medullary BaSO4 perfusion was very poor, even before reperfusion: scores were significantly lower (P<0.05) than group 1 kidneys at all time points up to and including 1 hour of reperfusion. However, after 3 hours of reperfusion, medullary perfusion had improved and was not significantly different from group 1 controls.
Histology. In group 1 kidneys, there was very little edema; in isolated instances, there was mild to moderate focal intracellular edema, especially of the proximal tubule. There was no congestion, no hemorrhage, and no inflammatory infiltrate (e.g., neutrophils). In group 2, there was widespread severe cortical intracellular edema in almost every instance. In the medulla, there was moderate interstitial edema and severe congestion, but the cortex was not usually congested. There was no inflammatory infiltrate in any specimen.
The results demonstrate that near infrared spectroscopy can be used to monitor changes in renal hemoglobin oxygenation kinetics after hypothermic storage and transplantation. When used in transmission mode, NIRS measures changes in hemoglobin concentration across both the cortex and medulla. Although medullary blood volume is small compared with the cortex during normal conditions, medullary hyperemia after ischemia is likely to make a larger contribution to the total hemoglobin signal. Because the optical path-length factor caused by scatter in renal tissue is unknown, as far as we know, changes in hemoglobin concentration measured in this study are relative. However, in this study, we have compared between 2 groups of kidneys subjected to the same treatments with the exception of storage time. Because it has previously been shown that the optical pathlength is not significantly affected by gross differences in hemoglobin oxygenation (27), comparison between the 2 groups of kidneys is justified. This premise concurs with the fact that the mean total hemoglobin concentrations and the range of SEM values for HbT were similar in both groups (Fig. 4) despite large differences in hemoglobin oxygenation.
Reperfusion of group 1 kidneys resulted in rapid increases in HbO2 within 1 minute (Fig. 2), followed by a sharp decrease from 1494±138 μM at 1 minute to 1077±224 μM at 3 minutes. However, HbO2 had recovered by 20 minutes and did not change significantly thereafter. Group 1 Hb increased at a slower rate and had plateaued by 3 minutes (Fig. 3). The abrupt fall in HbO2 during the second and third minutes of reperfusion was accompanied by a less dramatic reduction in HbT (Fig. 4) and a small increase in Hb (Fig. 3), implying that an initial reactive hyperemia (signified by the fall in HbT) was associated with low tissue pO2 in the renal parenchyma (signified by the larger fall in HbO2). These changes are expected as a normal consequence of warm ischemia during the transplant procedure. We have previously shown that this brief period of ischemia also results in mild reversible mitochondrial dysfunction in the first 5 minutes of reperfusion of unstored transplanted kidneys (18).
The rate and the magnitude of changes in HbO2 in the first 5 minutes of reperfusion of group 2 kidneys seemed less than group 1, but the only significant difference (P<0.05) occurred at 1 minute, and the changes in NIRS parameters were essentially similar in the 2 groups (Fig. 2). The slower rate of increase of HbO2 and HbT (Fig. 4) in the first 1 minute of reperfusion is suggestive of greater vasoconstriction or vasospasm in the stored kidneys. Afferent and efferent arteriolar vasoconstriction has been reported in several models of renal ischemia and reperfusion (29, 30) and may account for the modest decrease previously observed in cortical blood flow after ischemia: cortical, unlike medullary, blood flow is not affected by hemodilution (8).
In this study, vasoconstriction seemed to be a transient response to reperfusion (or a delayed recovery from extended hypothermia), because group 2 HbO2 levels were not significantly different from group 1 between 2 and 10 minutes of reperfusion, and HbT remained similar in both groups throughout the 3 hours of reperfusion. This would question the role of vascular eicosanoids as promoters of renal vasoconstriction during reperfusion. Several studies have suggested that thromboxane A2, a potent vasoconstrictor, is a significant modulator of renal blood flow after ischemia (31-33). However, we have previously reported that thromboxane levels do not rise significantly after transplantation of 72 hours stored kidneys (34), and this finding is consistent with the hemodynamic changes reported here. These discrepancies may reflect the different models used; most studies of eicosanoid production in renal ischemia reperfusion injury have dealt with warm ischemia, in which the vasculature was congested with erythrocytes and other blood cells throughout the period of ischemia, and this scenario is very different from the cold ischemic/HCA flush model described here.
The sharp drop in group 2 HbO2 after 2 minutes of reperfusion is therefore likely to have been a response to parenchymal ischemia, as seen in group 1 kidneys, rather than a result of vasoconstriction. Unlike group 1 kidneys, however, there was no decrease in HbT corresponding to the drop in HbO2(Fig. 4). It is possible that early vasoconstriction precluded a reactive hyperemia in group 2 or that the sustained rise in Hb after 3 minutes of reperfusion (Fig. 2) masked any cortical hyperemia. This sustained increase in Hb may have been partly a result of prolonged ischemia in group 2 (with greater oxygen extraction in the more ischemic group) but is also likely to be a function of developing vascular congestion in the medulla. This phenomenon is sometimes referred to as medullary hyperemia (2-4); however, trapping of erythrocytes in the outer zone of the medulla is a pathological development rather than genuine reactive hyperemia (28).
The fact that the sustained rise in Hb during the first 10 minutes of reperfusion does indeed correlate with the development of vascular congestion in the medulla is confirmed by BaSO4 angiographical (Fig. 7) and histological analysis of intrarenal perfusion. Ex vivo BaSO4 angiography does not necessarily reflect distribution of blood flow within the kidney. Instead, it gives an indication of vascular resistance (35), because the micro-opaque suspension was infused at a maximum perfusion pressure of 90 mmHg and in a nonpulsatile manner; moreover, BaSO4 particles, although of similar size to erythrocytes, lack their plasticity and are more likely to become lodged in the glomeruli or the peritubular capillaries. Nevertheless, areas of tissue well perfused with BaSO4 are likely to have been well perfused with blood during reperfusion. Thus, cortical perfusion seems to have been normal in group 2 kidneys throughout the period of reperfusion (Fig. 7). On the other hand, it is likely that medullary perfusion was underestimated by angiography. After 72 hours of storage and before reperfusion, medullary vascular resistance was sufficiently high to prevent perfusion with BaSO4 (Fig. 7, t=0). During reperfusion, this situation persisted for at least the first 60 minutes. Nevertheless, both histological analysis and NIRS demonstrated that blood was able to penetrate (if not drain from) the medulla during the first 10 minutes of reperfusion.
After 30 minutes of reperfusion of group 2 kidneys, there were decreases in both Hb (P<0.005) and HbT but no significant change in HbO2. A similar reduction in Hb after 45 minutes of reperfusion has been reported using NIRS in a renal model of warm ischemia and reperfusion (36). Because there was no fall in HbO2(Fig. 2), it is likely that the decrease in Hb was a result of clearance of desaturated blood from the renal medulla. Furthermore, because the fall in total hemoglobin measured after 30 minutes of reperfusion coincided with an increase in serum levels of prostacyclin (a vasodilator that affects mainly the cortex) (34), it is likely that that the fall in total hemoglobin was a result of recovery from medullary hyperemia rather than cortical vasoconstriction. Although there was no significant increase in medullary BaSO4 perfusion even after 1 hour of reperfusion, it is likely that the angiographical studies underestimated medullary perfusion at this time point for the reasons discussed above. After 3 hours of reperfusion, vascular resistance in the medulla had dropped sufficiently to allow nearly normal BaSO4 perfusion (Figs. 6C and 7). Thus, it seems likely that tubular swelling in the outer medulla diminished during the course of reperfusion, resulting in a fall in vascular resistance and recovery from medullary congestion. The fact that clearance of congestion was not accompanied by an increase in trans-renal SaO2 presumably reflects the severity of parenchymal hypoxia in the medulla and suggests that there was an increase in medullary rather than cortical oxygen extraction. The stability of renal SaO2 and of cortical BaSO4 perfusion throughout this period also suggests that cortical reperfusion (and oxygen extraction) was not greatly affected by shifts in medullary perfusion. This inference is in agreement with Hellberg and co-workers, who demonstrated the independence of blood flow through the renal cortex and medulla after warm ischemia (7).
Considered together, these data point to several conclusions. First, vascular resistance was high in the medulla even before reperfusion, probably as a result of ischemic swelling of the medullary thick ascending tubules and the pars recta of the proximal tubules (as demonstrated in warm ischemic models by Mason and co-workers (8) and as seen in this model (37). Second, medullary vascular congestion developed during the first 10 minutes of reperfusion. The fact that hemoglobin in the trapped erythrocytes became severely desaturated during this time was confirmed by NIRS (Figs. 2 and 4). Medullary congestion began to clear after 30 minutes of reperfusion, but this was not accompanied by an increase in renal hemoglobin saturation, suggesting that tissue pO2 in the medulla remained very low. The low pO2 is likely to have stimulated increased oxygen extraction in the medulla, thereby suppressing rises in renal SaO2; increases in oxygen extraction do not necessarily indicate an increase in oxygen consumption. Finally, cortical vascular resistance was not significantly different from controls, nor were the hemoglobin oxygenation kinetics in group 2 significantly different from group 1 during the first 5 minutes of reperfusion, even across the entire kidney and despite the development of medullary congestion. Thus, it seems likely that cortical blood flow and hemoglobin oxygenation were very similar in the first 5 to 10 minutes of reperfusion and probably thereafter.
These conclusions are important in the context of mitochondrial dysfunction during reperfusion of 72 hours stored transplanted kidneys. We have previously shown, using surface fluorescence spectroscopy of superficial cortical mitochondrial NADH fluorescence (17) and NIRS measurements of cytochrome oxidase across the entire kidney (19), that severe mitochondrial energetic disruption occurs within the first 5 minutes of reperfusion. Moreover, there was no recovery in mitochondrial function after 3 hours of reperfusion (19). Because the changes measured were predominantly in the superficial cortex and this study has shown that cortical perfusion and hemoglobin oxygenation kinetics across the entire kidney were not significantly different from unstored transplanted kidneys, it may be concluded reasonably that mitochondrial dysfunction after renal storage was not only a result of secondary ischemia. Furthermore, this study demonstrates that medullary congestion is reversible over the first 3 hours of reperfusion, and this suggests that ischemic damage to the thick ascending tubules and the pars recta of the proximal tubules is reversible in this model. We conclude that attempts to improve renal function after hypothermic storage and transplantation should focus on means of improving mitochondrial function, particularly within the cortical proximal tubules.
Acknowledgments. We thank Caroline Doré for her help with the statistical analysis.
This work was supported in part by the Dunhill Medical Trust.
Abbreviations: ATN, acute tubular necrosis; BaSO4, barium sulfate; HbO2, oxygenated hemoglobin; Hb, deoxygenated hemoglobin; HbT, total hemoglobin; SaO2 percentage mixed renal hemoglobin saturation; HCA, hypertonic citrate solution; NIRS, near-infrared spectroscopy.
1. Rosenthal JT, Danovitch GM, Wilkinson A, Ettinger RB. The high cost of delayed graft function in cadaveric renal transplantation. Transplantation 1991; 51: 1115.
2. Brezis M, Seymour R, Silva P, Epstein FH. Renal ischemia: a new perspective. Kidney Int 1984; 26: 375.
3. Yamamoto K, Wilson DR, Baumal R. Outer medullary circulatory defect in ischemic acute renal failure. Am J Pathol 1984; 261: 253.
4. Mason J. The cause and consequences of the perfusion defect that follows renal ischaemia. In: Davidson AM, ed. Nephrology, Vol. II. Cambridge, UK: Cambridge University Press, 1988: 955.
5. Andersson G, Jennische E. Lack of causal relationship between medullary blood congestion and tubular necrosis in ostischaemic kidney damage. Acta Physiol Scand 1987; 130: 429.
6. Shanley F, Rosen MD, Brezis M, Silva P, Epstein FH, Rosen S. Topography of focal proximal tubular necrosis after ischemia with reflow in the rat kidney. Am J Pathol 1986; 122: 462.
7. Hellberg PAO, Kallskog O, Wolgast M. Red cell trapping and postischemic renal bloodflow. Differences between the cortex, outer and inner medulla. Kidney Int 1991; 40: 625.
8. Mason J, Welsch J, Takabatake T. Disparity between surface and deep nephron function early after renal ischemia. Kidney Int 1983; 24: 27.
9. Paller MS, Neumann TV. Reactive oxygen species and rat renal epithelial cells during hypoxia and reoxygenation. Kidney Int 1991; 40: 1041.
10. Togashi H, Shinzawa H, Yong H, Takahashi T, Noda H, Oikawa K, Kamada H. Ascorbic acid radical, superoxide and hydroxyl radical are detected in reperfusion injury of rat liver using electron spin resonance spectroscopy. Arch Biochem Biophys 1994; 48: 1.
11. Ratych RE, Bulkley GB. Free radical mediated postischemic reperfusion injury in the kidney. Free Radic Biol Med 1986; 2: 311.
12. Gower JD, Lane NJ, Goddard JG, Manek S, Ambrose IJ, Green CJ. Ebselen: antioxidant capacity in renal preservation. Biochem Pharmacol 1992; 43: 2341.
13. Schilling M, Saunder A, Southard JH, Belzer FO. Long term renal preservation and prevention of acute tubular necrosis by inhibition of arachidonate metabolism. Transplant Proc 1993; 25(4): 2534.
14. Rice-Evans CA, Diplock AT. Current status of antioxidant therapy. Free Radic Biol Med 1993; 15: 77.
15. Guder WG, Ross BD. Enzyme distribution along the nephron. Kidney Int 1984; 26: 101.
16. Taegtmeyer H, Roberts AFC, Raine AEG. Energy metabolism in reperfused heart muscle: metabolic correlates to return of function. J Am Coll Cardiol 1985; 6: 864.
17. Hardy L, Clark JB, Darley-Usmar VM, Smith DR, Stone D. Reoxygenation-dependent decrease in mitochondrial NADH: CoQ reductase (complex 1) activity in the hypoxic/reoxygenated rat heart. Biochem J 1991; 274: 133.
18. Lane NJ, Thorniley MS, Manek S, Fuller BJ, Green CJ. Iron chelation, respiratory chain function and tubular necrosis in renal transplantation. Biochem Soc Trans 1995; 23: 524S.
19. Thorniley MS, Lane NJ, Manek S, Green CJ. Non-invasive measurement of respiratory chain dysfunction following hypothermic renal storage and transplantation. Kidney Int 1994; 45: 1489.
20. Gower JD, Healing G, Fuller BJ, Simpkin S, Green CJ. Protection against oxidative damage in cold stored rabbit kidneys by desferrioxamine and indomethacin. Cryobiology 1989; 26: 309.
21. Jobsis FF. Non-invasive infrared monitoring of cerebral and myocardial oxygen sufficiency and circulatory parameters. Science 1977; 198; 1264.
22. Balaban RS, Sylvia AL. Spectrophotometric monitoring of O2
delivery to the exposed rat kidney. Am J Physiol 1981; 241: F257.
23. Thorniley MS, Wickramasinghe Y, Rolfe P. Near infra-red spectroscopy: a new technique for the non-invasive monitoring of tissue and blood oxygenation in vivo. Biochem Soc Trans 1988; 16: 978.
24. Brazy JE. Near infra-red spectroscopy. Clin Perinatol 1991; 18: 519.
25. Wray S, Cope M, Delpy DT, Wyatt JS, Reynolds EOR. characterization of the near infrared absorption spectra of cytochrome aa3
and haemoglobin for the non-invasive monitoring of cerebral oxygenation. Biochim Biophys Acta 1988; 933: 184.
26. Harris DNF, Bailey SM. Near infrared spectroscopy in adults. Anaesthesia 1993; 48: 694.
27. Cope M, Delpy DT, Reynolds EOR, Wray S, Wyatt JS, Van Der Zee P. Methods of quantitating cerebral near infrared spectroscopy data. Adv Exp Med Biol 1988; 222: 183.
28. Wyatt JS, Edwards AD, Azzopardi D, Reynolds EOR. Magnetic resonance and near infrared spectroscopy for investigation of perinatal hypoxic-ischemic brain injury. Arch Dis Child 1989; 64: 953.
29. Chan A, Chittinandana A, Shapiro JI, Shanley PF, Schrier RW. Effect of an endothelin-receptor antagonist on ischemic acute renal failure. Am J Physiol 1994; 266: F135.
30. Klausner JM, Paterson IS, Goldman G, et al. Postischemic renal injury is mediated by neutrophils and leukotrienes. Am J Physiol 1989; 256: F794.
31. Lelcuk S, Alexander F, Kobzik L, Valeri CR, Shepro D, Hechtman HB. Prostacyclin and thromboxane A2
moderate postischemic renal failure. Surgery 1985; 98: 207.
32. Halushka PB, Lefer AM. Thromboxane in health and disease. Fed Proc 1987; 46: 131.
33. Baud L, Ardaillou R. Involvement of reactive oxygen species in kidney damage. Br Med Bull 1993; 49: 621.
34. Lane NJ, Thorniley MS, Manek S, Fuller BJ, Green CJ. Inhibition of leukotriene B4
synthesis does not prevent development of acute renal failure following storage and transplantation. Transplantation 1994; 58: 1303.
35. Hollenberg NK, Epstein M, Rosen SM, Basch RI, Oken DE, Merrill JP. Acute oliguric renal failure in man: evidence for preferential renal cortical ischemia. Medicine 1968; 47: 455.
36. Vaughan DL, Wickramsinghe YABD, Russell GI, et al. Is allopurinol beneficial in the prevention of renal ischaemia-reperfusion injury in the rat? Evaluation by near-infrared spectroscopy. Clin Sci 1995; 88: 359.
37. Lane NJ, Thorniley MS, Manek S, Green CJ. Vascular congestion does not correlate with mitochondrial dysfunction following renal transplantation. J Pathol 1994; 172 (suppl): 138A.