- Email: [email protected]
Blood pH level modulates organ metabolism of lactate in septic shock in dogs
Blood pH Level Modulates Organ Metabolism of Lactate in Septic Shock in Dogs Carla Chrusch, Edgar Bautista, Hans K. Jacobs, R. Bruce Light, Deepak Bose, Krika Duke, and Steven N. Mink Objective: Lactic acidosis is an important complication of septic shock. Alkali treatment such as sodium bicarbonate is often used to treat the low pH level that develops in sepsis. The effect of this treatment on lactate (Lac) clearance is not clear. In the present study, the objective was to examine whether blood pH level alters Lac metabolism in sepsis. Measurements were determined in a canine model of Escherichia coli sepsis after bolus infusion (5 mmol/kg) of either lactic acid (LA) or sodium lactate (NaL). In one preparation, Lac uptake by the splanchnic organs (SP), liver, lung, kidneys (Kid), and soft tissues of the lower extremity (SOL) was primarily determined, whereas in another preparation, Lac uptake by the head and neck region and lung was obtained. Methods: The dogs were studied while anesthetized and ventilated. After 4 hours of sepsis, either LA or NaL was given through a catheter positioned in the abdominal aorta in respective sepsis (SepLA, SepNaL) and nonsepsis groups (ConLA, ConNaL) (n 6 in each preparation). Catheters and flow probes were used to measure organ Lac uptake. Measurements were obtained at end infusion and at 15-minute intervals after infusion until 75 minutes after infusion.
Results: Arterial clearance of Lac in the sepsis groups was slower as compared with the nonsepsis groups. In the liver, sepsis inhibited the uptake of LA as compared with the nonseptic group. In SP, both sepsis and pH affected Lac uptake in which an increase in uptake was found only after NaL infusion in the nonseptic group. In the head and neck region, Lac uptake was pH-level dependent and was found after LA infusion in the sepsis and nonsepsis groups. In the lung, Lac was produced after either LA or NaL infusion in all groups. Neither Kid nor SOL contributed to Lac uptake in any of the groups. Conclusion: Lactate clearance was reduced in sepsis. Both effects of pH level and sepsis modulated the organ uptake of Lac in septic shock. Only a small amount of the total Lac infused could be accounted for by the organs measured in the present study. This suggests that additional organs may account for lactate removal in sepsis. Copyright 2002, Elsevier Science (USA). All rights reserved.
N SEPTIC SHOCK, hyperlactemia and acidosis are often observed in severely ill patients.1-3 The factors that regulate the development of lactic acidosis in septic shock are complex and our understanding of the mechanisms responsible for its development are not clear. On the one hand, the mechanism of lactic acidosis in septic shock may reflect splanchnic overproduction caused by an impairment of aerobic metabolism4,5 occurring as a consequence of a decrease in bulk oxygen flow, a malredistribution of microvascular flow, or to a defect in the usage of oxygen at the mitochondrial level.6,7
On the other hand, rather than an increase in production, abnormalities in lactate (Lac) extraction by the liver and other organs may contribute to the hyperlactemia that accompanies septic shock.8-10 Uptake of Lac by the liver is thought to be regulated by an active transport process as well as by diffusion. At blood concentrations between 1 to 5 mmol/L, hepatic Lac uptake appears to be mediated primarily by an active transport process.11-15 Lac is actively cotransported with a hydrogen ion, so that when Lac anion is uptake up by the liver, the concomitant uptake of hydrogen ion results in the formation of bicarbonate ion in the plasma. At high Lac concentrations, diffusion of the nondissociated anion as well as other nonspecific active transport processes appear important in its metabolism. Although the factors that regulate hepatic uptake of Lac by the Lachydrogen cotransporter are still being worked out, blood pH level may be one important factor in the regulation of Lac uptake. When the pH level external to the hepatocyte is less than the pH level inside the cell, Lac uptake is increased as compared with the opposite situation.11-16 However, as pH level in the hepatocyte falls to very low values (ie, 6.9), the hepatocyte becomes a Lac producer
I
From the Departments of Medicine, Biochemistry and Molecular Biology; and Pharmacology and Therapeutics and Anesthesiology, University of Manitoba, Winnipeg, Manitoba, Canada. Supported by the Manitoba Heart and Stroke Foundation and by a fellowship from the Manitoba Lung Association (E.B.). Address reprint requests to Steven Mink, MD, Health Sciences Centre, GF-221, 700 William Ave, Winnipeg, Manitoba, Canada R3E-0Z3. E-mail: [email protected]. Copyright 2002, Elsevier Science (USA). All rights reserved. 0883-9441/02/1703-0007$35.00/0 doi:10.1053/jcrc.2002.35816 188
Journal of Critical Care, Vol 17, No 3 (September), 2002: pp 188-202
LACTATE METABOLISM IN SEPSIS
rather than a Lac consumer, and this may contribute to the increase in Lac and acidemia observed in sepsis.17,18 Although blood pH level has theoretical implications on modulating the metabolism of Lac in septic shock, there is little information available as to whether this effect would be clinically important. Nevertheless, alkali treatment such as sodium bicarbonate is often used to treat the low pH level that develops in sepsis.19 The lack of beneficial response observed with this treatment may occur as a consequence of the resulting acid-base changes that may negatively impact on Lac clearance among the different organs.19,20 Under nonseptic conditions, the liver and kidney are thought to be the major sources of Lac uptake. pH level has been found to alter Lac uptake by these organs,2,17,18,21-23 as well as skeletal muscle,24,25 and brain.25-27 Moreover, the lung has been shown to produce Lac under many pathologic conditions28-30 and the effect of pH level on this production needs to be considered in the understanding of Lac clearance in sepsis. Accordingly, because blood pH level may be important in determining the extent to which different organs may metabolize Lac in septic shock, this question was examined in a canine model of Escherichia coli sepsis.10,31 Either hydrogen Lac or its salt, sodium lactate (NaL), was administered as a bolus infusion after 4 hours of bacteremia in respective groups of dogs. Major organs that are associated with Lac metabolism including the liver, gut, lung, kidneys, as well as the brain and skeletal muscle were compared in respective septic and nonseptic groups. The objectives were to examine which organs were responsible for Lac metabolism in sepsis and to delineate the effect of pH level on its metabolism. METHODOLOGY
Animal Preparation This study was performed in accordance with protocols approved by the University Animal Care Committee and conforms with the Guide for the Care and Use of Laboratory Animals published by the National Institutes of Health (NIH Publication No. 85-23, revised 1996). Mongrel dogs (25–30 kg) were anesthetized with thiopental sodium (20 mg/kg, intravenously) and were given a loading dose of sufentanil (10 ug/kg), followed by maintenance doses of sufentanil (1 ug/min) and midazolam (5 ug/kg/min).31,32 The trachea was intubated with an endotracheal tube and the lungs were mechanically ventilated (Harvard Apparatus, Hollison, MA).
189
The ventilator settings (approximate rate of 10 breaths/min; tidal volume, 15 mL/kg) were adjusted to maintain initial arterial PCO2 at approximately 35 mm Hg and pH level at approximately 7.35. These settings were not changed throughout the entire experiment. The inspired oxygen concentration was titrated to maintain arterial PO2 greater than 150 mm Hg, so that arterial hypoxemia did not contribute to any of the findings observed over the course of the experiment. In this study, 2 preparations were used because it was often not feasible to obtain measurements of lactate metabolism from the many organs examined in a single preparation in most cases. In setting up both preparations, the animal was placed in the supine position, and vascular catheters were inserted percutaneously under sterile conditions. An arterial catheter was placed into the left femoral artery and advanced into the midabdominal aorta. This catheter was used to obtain arterial lactate concentration (LacArt), oxygen content (CaO2) (see later), and arterial blood gases. It was also used to measure mean arterial pressure (MAP), and for infusion of one of the lactate solutions (see later). Into the left jugular vein, a thermistor-tipped catheter was advanced into the pulmonary artery to measure mean pulmonary arterial pressure (Ppa), mean pulmonary wedge pressure (Pwp), mean right arterial pressure (Pra), and cardiac output (CO) by the thermodilution technique (Columbus Instruments, Columbus, OH). The average of 3 determinations was used in the reporting of CO for each condition. Samples for mixed venous lactate (LacMV) and oxygen content (CMVO2 ) were obtained from the distal port of this catheter. In one preparation, splanchnic lactate metabolism (SLM), hepatic lactate metabolism (HLM), kidney lactate metabolism (KLM), and lower extremity (LEM) lactate metabolism were primarily determined. This preparation was similar to that described by Kellum et al33 and has previously been described.10 The abdomen was opened by a midline incision, after which the spleen was removed so that the splenic vein could be cannulated. A catheter was advanced into the splenic vein and positioned in the portal vein at the point at which it enters the liver. The portal vein catheter measured 0.2 cm in external diameter and contained one endhole and 2 sideholes, the latter of which were within 1.0 cm of the tip. The portal vein catheter was used to obtain blood samples for portal vein lactate (Lacpov) and portal vein oxygen content CpovO2 ) (see later). In addition to the catheter, a flow probe (Transonic, Ithaca, NY) was placed around the proximal portion of the portal vein to measure portal blood flow (Qpov). Another catheter with characteristics similar to the portal vein catheter was advanced from the right jugular vein into the hepatic vein. The hepatic vein catheter was usually placed into the right hepatic vein because of easier placement, although hepatic Lac concentrations and oxygen contents were similar between right and left hepatic veins when determinations were compared in a previous study.10 The hepatic vein catheter was positioned into the hepatic vein by palpation, and proper positioning of the catheter was verified at autopsy. The hepatic vein catheter was used to obtain hepatic vein lactate (Lachv) and hepatic oxygen content (ChvO2 ). A flow probe (Transonic) was placed around the hepatic artery to measure hepatic artery flow (Qha). Two catheters identical in size to the hepatic and portal catheters (0.2 cm) were inserted percutaneously through the right and left femoral veins, respectively, and advanced into the inferior vena cava. One catheter was placed above the renal veins
190
(IVCupper), whereas the other catheter was placed below the renal veins (IVClower). A flow probe was placed around the IVC at a position below the renal veins to measure IVC flow (QIVC). From Lac determinations made above and below the renal veins, KLM was assessed. LEM was calculated from the difference between arterial Lac and the blood Lac concentration in the lower vena cava (ie, LacArt LacIVClower). In a second preparation, Lac metabolism by the thorax and cephalic region (to assess primarily brain uptake) was determined. The spleen was removed for consistency with the previous preparation. A right jugular venous catheter (0.2 cm) was inserted and was advanced in a retrograde manner to the base of the brain to measure jugular venous lactate (LacJug) and jugular venous oxygen concentration. A flow probe was placed around one of the carotid arteries to measure carotid artery flow (Qcar). In addition, because the proximal port of the Swanz-Ganz catheter (Edwards Life Sciences Corp., Irvine, CA) was located in the left brachiocephalic vein, not too far from the superior vena cava (SVC), this port was used to obtain LacSVC, so that the lactate difference measured between the jugular vein and SVC (LacJug) LacSVC) could be determined. The fluid-filled catheters to measure MAP and pulmonary artery pressures were connected to transducers (Cobe, Lakewood, CO) that were referenced relative to the left atrium. Measurements were obtained at end expiration during which the ventilator was turned off for approximately 3 to 5 seconds. All signals were displayed on an 8-channel recorder (Astro Med, West Warwick, RI). Furthermore, the ultrasonic flow probes used in this study have been found to be stable with little 0 drift over the course of the experiment; before and after each experiment, 0 flow that was obtained electrically was checked with that measured when the vessel was transiently occluded.
Protocols Baseline measurements (presepsis) were obtained 1 hour after completion of the preparation when hemodynamics (ie, MAP and CO) were stable. The animals were randomized into sepsis and nonsepsis groups. In the sepsis groups, a bolus of 1010 colony forming units of live E. coli (serotype 0111:B4) was given over a 30-minute interval.10,31 A constant infusion of approximately 5 109 colony forming units of E. coli per hour was then maintained for the remainder of the study. After 4 hours of infusion, measurements were again made and this interval was termed the 4-hour condition. The reason for choosing the 4-hour period was that systemic hypotension has been shown to occur at the 4-hour period in this model.31 In the nonseptic group (control group), the protocol was the same except that a similar amount of normal saline was administered over this period. After the 4-hour period, further randomization was performed. In the septic and nonseptic groups, one half of the animals in each group received 5 mmol/kg of lactic acid, whereas the other half received a similar dose of NaL (Sigma, St. Louis, MO). NaL or LA was mixed in .500 L of 5% dextrose in water (D5W). The 4 groups were designated as follows: (1) septic group with lactic acid (SepLA), (2) septic group with sodium lactate (SepNaL), (3) control group with lactic acid (ConLA), and (4) control group with sodium lactate (ConNaL). There were approximately 12 animals in each group, and, for the most part, approximately 6 animals were studied in preparation 1, while the other 6 were examined in preparation 2. Sometimes, there was overlap in the 2 preparations, so that additional measure-
CHRUSCH ET AL
ments could be obtained in a given experiment. In SepLA, HLM and LEM were also studied in 2 experiments as part of preparation 2, whereas SLM was studied in one experiment as part of preparation 2. In ConNaL, SLM and HLM were studied in one experiment as part of preparation 2. In ConLA, KLM was studied in 3 experiments as part of preparation 2 and in 3 experiments as part of preparation 1, whereas in SepLA, KLM was studied in 4 experiments as part of preparation 1 and in 2 experiments as part of preparation 2. Differences were not observed when the 2 preparations were compared. In all groups, either the LA or NaL treatment was given over 30 minutes through the catheter positioned in the lower abdominal aorta. Lac was infused into the aorta rather than into the venous circulation based on the work of others that showed that the infusion of lactate into the venous circulation of normal dogs resulted in intravascular hemolysis and the formation of microemboli in the lungs.34 This effect leads to pulmonary hypertension, which does not occur when Lac is infused into the aorta. In the protocol, hemodynamics were measured at end infusion (during which the infusion was completely off) and at 15, 30, 45, 60, and 75 minutes after infusion. Hemodynamic measurements included CO, Ppa, Pwp, Pra, MAP, arterial hematocrit (Hct), as well as organ hematocrits, and flows. Arterial samples for Lac were also obtained at the latter intervals, whereas samples for mixed venous and organ Lacs were measured at all of these intervals except for 60 and 75 minutes after infusion. Similarly, arterial blood samples for PO2, PCO2, pH levels, and oxygen contents were determined at all intervals, whereas samples for mixed venous and organ samples were determined at all of the earlier-mentioned intervals except for 60 and 75 minutes after infusion. Moreover, in 5 other nonseptic dogs (ConNSS), and in 5 other septic dogs (SepNSS), time control experiments were performed. In these 2 groups, at the 4-hour period, a similar amount of sodium chloride (5 mmol/kg) mixed in .5 L D5W was infused over 30 minutes. Then measurements were determined at intervals identical to the other groups. The rationale for ConNSS and SepNSS was to contrast the effect of lactate versus saline solution on the parameters measured. In these experiments, both preparations 1 and 2 were combined because KLM and LEM were found to have little effect on Lac metabolism in the treatment groups (see Discussion section). In all groups, oxygen contents were directly measured by the carbon monoxide scrubbing technique.35 Lac determination was made by an automated Lac dehydrogenase–based assay. Previous experiments have indicated that the coefficient of intraassay and interassay variation of Lac measurements were 0.57% 0.21% and 1.2% 0.33%, respectively,10 and therefore only one sample was obtained during each measurement. When measurements were obtained, moreover, the catheter was agitated in a to-and-fro motion to reduce any effect of streaming.
Data Analysis In the determination of organ Lac metabolism, uptake of Lac from plasma rather than blood was used based on the work of Naylor et al.36 These investigators showed that erythrocyte Lac equilibrates only slowly with plasma, and therefore little net erythrocyte Lac would be removed from the red cell during flow through the various organs examined. SLM was calculated from (LacArt Lacpov) Qpov (1 Hctpov). HLM was calculated
LACTATE METABOLISM IN SEPSIS
from [Qha LacArt (1 HctArt) Qpov Lacpov (1 Hctpov) (Qpov Qha) Lachv (1 Hcthv]. Lung Lac metabolism (LLM) was calculated by (LacMV LacArt) CO (1 HctMV). Cephalic lactate metabolism (CLM) was calculated by (LacArt LacJug) 2 Qcar (1 HctJug). It is important to note that when Lac concentration differences across an organ were determined, total blood Lac was used because blood flow was not considered in the equation. KLM was calculated by (LacIVCupper LacIVClower); LEM was calculated by (LacArt LacIVClower). Lac difference between the brain and SVC was calculated from (LacJug LacSVC). After either NaL or LA was infused, the decay in arterial blood Lac was analyzed in terms of 1- and 2-compartment models (NW StatPak, Portland, OR).9 Of the mathematic functions tested (linear, log, exponential, and power), the results indicated that a 1-compartment exponential model gave the best fit. The equation used to describe the decay was Aebx where “b” is the parameter of the curve fit, x time, and A is the Y intercept. Whole-animal and organ oxygen consumption were calculated from the respective arterial and venous content differences and flows.10 Whole-body consumption was calculated from (CaO2 CmvO2) CO; splanchnic oxygen consumption was calculated from (CaO2 CpovO2) Qpov; hepatic oxygen consumption was calculated from [(Qha CaO2 Qpov CpovO2) (Qha Qpov) ChvO2)]; and cephalic oxygen consumption was calculated from 2Qcar (CaO2 CjugO2). Furthermore, in conjunction with arterial pH and PCO2 levels, arterial bicarbonate was calculated at each measurement interval from the Henderson-Hasselbalch equation in which pH level 6.1 log([HCO3]/0.03 PCO2).
Statistics Between-group analyses were determined by 2-way repeated measures analysis of variance (ANOVA) split plot design (factor A, different treatment groups; factor B, different time periods), in which the interaction between the 2 factors was assessed. If a significant interaction was present, a StudentNewman-Keuls’ (SNK) multiple range test was used to determine where differences occurred. Furthermore, within a given group, statistical analyses between all of the conditions were not reported because the comparison between baseline and sepsis was the primary measurement examined. Finally, when a single parameter between groups was examined, a one-way nonrepeated ANOVA and SNK were used. Results are expressed as mean 1SE.
RESULTS
Lactate Metabolism The decay in LacArt measured in the 4 treatment groups is shown in Fig 1 (left upper panel). In the septic and nonseptic groups, LacArt in SepLA were higher than those in ConLA over the entire decay. The curves were plotted as an exponential function in which the parameter “b” was an expression of the curve fit. In SepLA, “b” was less negative than that found in ConLA (.0116 .0015 vs. .0204 .0023; P .05), which indicates that the decay in concentrations was slower in SepLA. Further-
191
more, at end infusion, peak lactate concentration in SepLA was higher than that measured in ConLA and, consistent with this finding, the parameter “A” in the curve fit was also higher in SepLA (9.5 56 vs. 7.045 .48 mmol/L; P .05). In SepNaL, LacArt was higher at 15 and 30 minutes after infusion as compared with all of the other groups, and LacArt remained higher than those in ConNaL over the remainder of the experiment. The parameter “b” in SepNaL was again less negative than that found in ConNaL (.0117 .0014 vs. .0180 .0017; P .05), but was not different than that calculated in SepLA. In SepNaL, “A” was higher than that found in ConNaL, but did not reach statistical significance between groups (9.12 .44 vs 7.65 .57 mmol/L). In terms of the exponential model used, the correlation coefficient (R) was similar in the 4 groups and measured .93 .009 in SepLa, .95 .008 in ConLA, .92 .03 in SepNaL, and .95 .012 in ConNaL. Furthermore, after infusion of either NaL or LA, it is important to note that steadystate plasma concentrations of Lac were reached within 15 minutes from the start of the infusion. LacArt was measured at 15 minutes and then at 30 minutes after infusion. The results (n 9) were similar between the 2 time periods and measured 10.3 1.4 and 11 2 mmol/L, respectively. SLM was different among the 4 groups (see Fig. 2, left upper panel). After infusion of NaL, Lac consumption increased in ConNaL, but not in SepNaL. The increase in splanchnic uptake measured in ConNaL was higher than that found in all of the other groups. In SepLA and ConLA, SLM remained at net 0 flux over the course of the experiment. HLM is shown in Fig. 2 (left lower panel). After LA was infused, uptake of Lac increased in ConLA, but not in SepLA. After infusion of NaL, HLM was not significantly altered as compared with baseline in either the septic or nonseptic group. Lung Lac metabolism was determined from the difference between mixed venous lactate (LacMV) and arterial lactate (LacArt) times cardiac output. Because small differences in (LacMV LacArt), when multiplied by CO, may account for large differences in lung metabolism (LacMV LacArt) and the product of (LacMV LacArt) CO were analyzed separately (see Fig 2 right upper and lower panels). There were significant increases in (LacMV LacArt) and lung lactate production at
192
CHRUSCH ET AL
Fig 1. Arterial lactate, arterial pH, arterial PCO2, and HCO3 levels are shown for the 4 groups in which sodium Lac or lactic acid was infused. Statistics were performed by 2-way ANOVA and SNK. ——, Septic group lactic acid; ——, control group lactic acid; ——, septic group sodium lactate; ——, control group sodium lactate. (A) *P .05 vs baseline; !P .05 vs ConLA and ConNaL; $$$P .05 vs ConLA; #P .05 vs all other groups. (B) *P .05 vs baseline; #P .05 vs all other groups; ?P .05 vs. SepNaL, ConNaL. (C) *P .05 vs baseline; ?P .05 vs SepNaL, ConNaL; #P .05 vs all other groups; ??P .05 vs SepLA; !P .05 vs ConLA, ConNaL; %P .05 vs ConNaL. (D) %P .05 vs ConNaL; ?P .05 vs SepNaL, ConNaL; *P .05 vs baseline; $$$P .05 vs ConLA.
end infusion and at 15 minutes after infusion in all groups. Cephalic Lac extraction is shown in Fig 3 (left upper panel). There was a marked pH level dependence of Lac metabolism by the cephalic region. After LA was infused, Lac extraction in both the sepsis and nonsepsis groups was substantially increased at end infusion as compared with baseline. However, no Lac extraction was observed after NaL infusion in the septic and nonseptic groups. The Lac difference between the jugular vein (LacJug) and SVC (LacSVC) was also determined at each measurement interval (Fig 3; left lower panel).
The proximal port of the Swanz-Ganz catheter was located in the left brachiocephalic vein near the SVC, so that a venous sample could be obtained from this site to measure the difference in Lac concentrations between Jug and SVC. In all the treatment groups, a significant decrease in Lac concentration was observed at end infusion in the SVC sample. Kidney Lac difference is shown in Fig 3 (right upper panel). The results showed that no uptake of Lac was observed at any of the intervals in the 4 groups. Similar findings were observed in the uptake of Lac by the lower extremity (see Fig 3, right
LACTATE METABOLISM IN SEPSIS
193
Fig 2. Splanchnic Lac metabolism, hepatic Lac metabolism, mixed venous minus arterial Lac, and lung Lac metabolism are shown for the 4 groups in which sodium Lac or lactic acid was infused. Positive values indicate uptake, whereas negative values indicate production. Statistics by 2-way ANOVA and SNK. ——, Septic group lactic acid; ——, control group lactic acid; ——, septic group sodium lactate; ——, control group sodium lactate. (A) *P .05 vs baseline; #P .05 vs all other groups. (B) *P .05 vs baseline. (C) *P .05 vs baseline; @P .05 vs SepLA, SepNSS. (D) *P .05 vs baseline.
lower panel) in which no evidence of extraction was observed among the treatment groups. In the septic and nonseptic saline groups (ie, SepNSS and ConNSS), normal saline solution was infused to contrast the difference in organ metabolism between the 2 solutions. The arterial lactates are shown in Table 1. The increases in LacArt observed in this model are modest. In SepNSS, although LacArt significantly increased at 4 hours as compared with baseline (see Table 1), and although slight further increases in arterial Lac occurred over the course of the experiment, LacArt were not different between SepNSS and ConNSS. During saline infusion, SLM and HLM showed only small changes over the course of the experiment. At end infusion, lung Lac production did not increase as
compared with the Lac treatment groups. Similarly, cephalic Lac uptake and (LacJug LacSVC) did not change over the course of the experiment in either SepNSS or ConNSS. Acid Base Changes The changes in pH level found with either LA or NaL infusion were different among the septic versus nonseptic groups (see Fig 1; left lower panel). It is important to note that in the design of the experimental protocol, the ventilator settings were not changed over the entire measurement period. pH level in SepLA decreased to approximately 6.9 at end infusion, which was significantly lower than that found in ConLA. Values of pH in SepLA remained lower than those found in ConLA
194
CHRUSCH ET AL
Fig 3. Cephalic Lac extraction, jugular Lac minus superior vena cava Lac, kidney Lac metabolism, and lower extremity Lac are shown for the 4 groups in which sodium Lac or lactic acid was infused. Positive values indicate uptake, whereas negative values indicate production. Statistics by 2-way ANOVA and SNK. ——, Septic group lactic acid; ——, control group lactic acid; ——, septic group sodium lactate; ——, control group sodium lactate. (A) *P .05 vs baseline; ?P .05 vs SepNaL, ConNaL. (C) *P .05 vs baseline.
for most of the postinfusion interval. In SepNaL, pH level did not change as compared with baseline over the entire measurement period. In contrast, in ConNaL, pH level gradually rose as compared with the baseline value. These changes were significantly different between the septic and nonseptic groups. The changes in PCO2 found over the course of the experiment are shown in Fig 1 (right upper panel). After infusion of LA, there were large initial increases in PCO2 that occurred to similar extents in both the septic and nonseptic groups. After end infusion, PCO2 in SepLA declined more rapidly as compared with ConLA, such that PCO2 values were significantly lower for most of the postinfusion period in SepLA. In contrast, after infusion of NaL, PCO2 increased to similar degrees
as compared with baseline in both the septic and nonseptic groups. The calculated arterial bicarbonates are shown in Fig 1 (right lower panel). In ConLA, HCO3 remained unchanged as compared with baseline over the entire experimental protocol, whereas in SepLA, HCO3 decreased from 10 to 6 mmol/L between 4 hours and 75 minutes after infusion. In contrast, when NaL was infused in the septic and nonseptic groups, HCO3 increased steadily over the course of experiment, and this occurred to a slightly greater extent in the nonseptic group. In the saline groups, pH level decreased in both SepNSS and ConNSS after saline infusion and remained relatively constant over the subsequent intervals (see Table 2). Corresponding changes in PCO2 and HCO3 are also shown in Table 2.
LACTATE METABOLISM IN SEPSIS
195
Table 1. Lactate Metabollism in ConNSS and SepNSS
ConNSS (n 5) Arterial Lac (mmol/L) Splanchnic Lac metabolism (mmol/min) Hepatic Lac metabolism (mmol/min) Lung Lac metabolism (mmol/min) Cephalic Lac metabolism (mmol/min) (Jug SVC) Lac (mmol/L) SepNSS (n 5) Arterial Lac (mmol/L) Splanchnic Lac metabolism (mmol/min) Hepatic Lac metabolism (mmol/min) Lung Lac metabolism (mmol/min) Cephalic Lac metabolism (mmol/min) (Jug SVC) Lac (mmol/L)
Baseline
4 Hours
End Infusion
15 Minutes After Infusion
30 Minutes After Infusion
1.3 .2 .009 .005
1.5 .3 .011 .01
2.1 3* .14 .1||
2.3 3*,† .025 .02
2.1 3†,‡ .004 .02
1.9 1†,§ .002 .01
.02 .02
.02 .01
.07 .1¶
.05 .02
.05 .02
.03 .02
.27 .08
.22 .06
.44 .12
.38 .21
.014 .002 .05 .1
.014 .004 0 .15
.41 .22*
.84 .35†
.034 .013#,¶
.034 .008
.26 .1*
1.8 .2 .02 .01
1.9 .3 .007 .01
3.1 3*,† .009 .02**
.01 .02
.01 .02
.06 .08¶
.08 .07
.11 .06
.07 .22*
.016 .07
.012 .00
.04 .04
.03 .06
.1 .2 3.1 3*,† .012 .01 .04 .04
.01 .008 .05 .1 2.6 3†,§ .022 .01
45 Minutes After Infusion
.01 .006 .05 .1 2.2 4§ .0146 .005
.01 .04
.05 .02
.42 .07†
.30 .07
.25 .09
.014 .01#,¶
.04 .01
.016 .01
.018 .01
.03 .04*
.09 .06
.07 .1
.07 .1
NOTE: Mean SE. ConNSS and SepNSS are control and septic normal saline groups, respectively. Positive values are uptake, whereas minus values are production. By split plot ANOVA and SNK. *P .05 vs SepLA, SepNaL, ConLA, and ConNaL. † P .05 vs baseline. ‡ P .05 vs SepNaL, SepLA, ConNaL. § P .05 vs SepNaL and SepLA. || P .05 vs SepNaL. ¶ P .05 vs ConLA. # P .05 vs SepLA. **P .05 vs ConNaL.
Organ Blood Flow and Oxygen Consumption and Systemic Hemodynamics In Fig 4, the organ blood flows measured during infusion for the 4 treatment groups are shown. The response of hepatic artery blood flow (Qha) to either LA or NaL infusion was different between the septic and nonseptic groups. In the septic treatment groups, the infusion of LA or NaL caused an increase in Qha, that occurred to the greatest extent in SepNaL, whereas there was no change in Qha in the nonseptic groups (see Fig 4, left upper panel). In contrast, infusion caused the opposite changes in portal blood flow (Qpov) between the septic and nonseptic groups (Fig 4, left lower panel). There were no differences in IVC flow after LA or NaL infusion among the 4 treatment groups studied (see Fig 4, right upper panel). In all the treatment groups, carotid blood flow increased during Lac in-
fusion, and this occurred to the greatest extent in SepNaL (right upper panel). In SepLA, whole-animal oxygen consumption decreased at end infusion as compared with baseline and as compared with end infusion in ConNaL (Table 3). Splanchnic oxygen consumption, hepatic oxygen consumption, and cephalic oxygen consumption were not different between conditions and among groups over the course of the study. Systemic hemodynamics are shown in Fig 5. In the septic treatment groups, MAP (upper panel) measured at 4-hours decreased to a similar extent from baseline in all groups. At end infusion, MAP increased as compared to the 4-hour measurement, and then gradually declined over the remainder of the study. There were no changes in MAP over the course of the experiment in the nonseptic groups. In Fig 5 (middle panel), Pwp are shown for the NaL and LA groups. Pwp were slightly lower for the
196
CHRUSCH ET AL
Table 2. Acid Base and Hemodynamics in ConNSS and SepNSS
ConNSS (n 5) Arterial pH level Arterial PCO2 (mm Hg) Calculated arterial bicarbonate (mmol/L) Qha (L/M) Qpov (L/M) MAP (mm Hg) SepNSS (n 5) Arterial pH level Arterial PCO2 (mm Hg) Calculated arterial bicarbonate (mmol/L) Qha (L/M) Qpov (L/M) MAP (mm Hg)
End Infusion
15 Minutes After Infusion
30 Minutes After Infusion
45 Minutes After Infusion
Baseline
4 Hours
7.36 .01 26 2 9.4 1.7
7.36 .02 23 .05 9 .3
7.21 .01*,† 30 2|| 9.6 1.4#
7.21 .01*,‡ 27 3 8 1.4§
7.22 00*,§ 27 2¶ 7.6 1.2§
7.23 .01*,§ 26 ¶ 7.6 1.3§
.14 .02 .14 .01 110 10
.12 .04 .13 .01 113 12#
.17 .04** .33 .03#,†† 125 12#
.15 .04 .24 .02# 117 12#
.16 .04 .24 .02 118 12#
.15 00 .21 .02 120 12#
7.37 .01 25 1 8.4 .5
7.27*,†† 27 1 7.4 .6
.16 .08 .33 .03‡‡ 112 3
.21 .08 .20 .05 61 8*,‡‡,|| ||
7.17 .01*,† 36 2* 12.4 1.4* .37 .06*#,††,‡‡ .21 .05#,†† 95 11*#,‡‡
7.17 .01*,‡ 33 2*,|| || 10.4 1.0§ .35 .05*#,††,‡‡ .17 .04 91 11*,#
7.14 01*,§,‡‡ 33 2¶¶ 10.4 1.2§
7.15 .01*,‡‡,§§ 33 2¶¶ 9.8 7§
.36 .07*,‡‡ .17 .04 86 8*,‡‡,|| ||
.37 08*,‡‡ .15 .04 87 10*,‡‡,|| ||
NOTE: Mean SE. By 2-way ANOVA and SNK. *P .05 vs baseline. † P .05 vs SepLA, SepNaL, ConLA, ConNaL. ‡ P .05 vs SepNaL, SepLA, ConNaL. § P .05 vs SepNaL and ConNaL. || P .05 vs SepLA and ConLA. ¶ P .05 vs SepNaL. # P .05 vs ConNaL. **P .05 vs SepNaL and SepLA. †† P .05 vs ConLA. ‡‡ P .05 vs ConNSS. §§ P .05 vs ConLA, SepNaL, ConNaL. || || P .05 vs ConLA and ConNaL. ¶¶ P .05 vs SepLA.
septic groups at the 4-hour period as compared with the nonseptic groups, so that for similar amounts of fluid given during volume infusion, Pwp increased to higher values at end infusion in the nonseptic groups. For the most part, the changes in CO (see Fig 5, bottom) followed those in Pwp, and CO measured at the end-infusion period in SepLa was the lowest as compared with the other treatment groups. DISCUSSION
In the present study, the results showed that both blood pH level and sepsis had differential effects on Lac metabolism in this canine model of E. coli–induced sepsis. After LA or NaL was infused in the nonseptic groups, the half-life of the infusate as assessed from the arterial Lac decay was approximately 15 minutes, as compared with the nonseptic groups in which the half-life of arterial Lac decay approximated 75 minutes. The modeling constant “b” was more negative in the nonseptic
versus septic groups, suggesting that blood Lac given by either LA or NaL is cleared less effectively in sepsis. In addition, the peak Lac levels measured at end infusion in the sepsis groups were higher than those found in the nonsepsis groups. This did not appear because of a change in the volume of distribution of lactate (VolLac) among the groups because a similar dose of lactate (5 mmol/kg) was given in all experiments. VolLac (ie, 5 mmol/kg divided by lactate found between end infusion and 4 h) was not different and measured 0.69 .04 L/kg in ConLA, 0.65 .08 L/kg in SepLA, 0.689 .05 L/kg in ConLA, and 0.629 .037 L/kg in SepNaL. Because steady-state infusion had been reached, during infusion the high VolLac may indicate that Lac was being metabolized at this time. Based on the average weight of the animals (ie, 25 kg), approximately 125 mmol of Lac was given in each experiment. In ConLA, the half-life of Lac was approximately 15 minutes, so that approxi-
LACTATE METABOLISM IN SEPSIS
197
Fig 4. Hepatic artery flow, portal blood flow, IVC flow, and carotid artery flow are shown for the 4 groups in which sodium Lac or lactic acid was infused. The carotid artery flow includes flow from both arteries. Statistics by 2-way ANOVA and SNK. ——, Septic group lactic acid; ——, control group lactic acid; ——, septic group sodium lactate; ——, control group sodium lactate. (A) *P .05 vs baseline; !P .05 vs ConLA and ConNaL; %%P .05 vs SepNaL; #P .05 vs all other groups; $$$P .05 vs ConLA. (B) *P .05 vs baseline. (C) *P .05 vs baseline; %P .05 vs ConLA; ??P .05 vs SepLA; !P .05 vs ConLA, ConNaL; #P .05 vs all other groups. (D) *P .05 vs baseline; #P .05 vs all other groups.
mately 62 mmol was metabolized during this interval. In Fig 2 (left lower panel), between end infusion and 15 minutes after infusion, the absolute hepatic uptake of Lac could be approximated by the area under the curve for Lac concentrations measured. Then the amount extracted would be 0.4 mmol/min 15 min/2 3 mmol. Based on the work of Lupo et al11 and Naylor et al,36 the normal canine liver would be able to metabolize approximately 1 to 2 mmol/min of lactate under nonseptic conditions. In ConLA, the maximal rate of lactate uptake was 0.4 mmol/min, which was slightly less than that predicted based on the maximal uptake as reported in other studies. A major factor determining the hepatic uptake of Lac appears to be blood pH level. When pH level external to the hepatocyte is less than pH level in-
side the cell, Lac uptake has been found to increase as compared with the opposite situation.16,18 Thus, the initial large increase in hepatic uptake found at end infusion in ConLA would be consistent with the effect of blood acidemia that resulted in an accelerated uptake of lactate. However, it has also been shown that when hepatic intracellular pH level decreases to less than 6.9, then hepatic lactate uptake is subsequently curtained. Blood pH level approached 6.9 at end infusion in ConLA. The rapid fall in Lac uptake observed at the subsequent intervals could be explained by the development of a low intracellular pH level. In contrast to what was found in the nonseptic groups, hepatic Lac uptake was not observed in the septic groups, despite the fact that blood pH level was changed to a similar extent among the comparable groups. As previously
198
CHRUSCH ET AL
Table 3. Oxygen Consumption in the Treatment Groups Baseline
Hepatic oxygen consumption SepLA (n 7) ConLA (n 5) SepNaL (n 8) ConNaL (n 6) Splanchnic oxygen consumption SepLA (n 7) ConLA (n 5) SepNaL (n 8) ConNaL (n 6) Cephalic oxygen consumption SepLA (n 5) ConLA (n 5) SepNaL (n 4) ConNaL (n 6) Whole-body oxygen consumption SepLA (n 12) ConLA (n 12) SepNaL (n 12) ConNaL (n 12)
4 Hours
End Infusion
30 Minutes After Infusion
16 6 15 4 19 6 14 4
5.4 11 13 15
5 3 3 2
7 7.9 20 12
4 2 5 4
15 11 25 22
4 3 3 2
18 10 16 22
16 2 17 7 14 3 16 3
17 10 10 13
4 3 4 2
62 63 93 12 4
9 8 5 10
4 2 1 3
10 4 61 71 16 7
8 9 5 9
2 3 1 3
4 8 7 12
1 3 1 3
5 5 3 8
2 1 2 3
214 190 210 223
18 26 25 26
140 203 198 276
16* 23 30 29
172 181 182 277
15 21 21 39
8 8 8 12
2 1 3 3
266 221 207 244
18 19 14 22
15 Minutes After Infusion
4 1 4 10
4 5 5 9
1 1 3 3
207 186 206 312
23 23 27 85
45 Minutes After Infusion
16 11 15 19
4 2 4 2
8 7 8 11
2 1 3 2
93 63 32 83 180 162 183 255
21 23 27 26
NOTE: Values presented as mL/min; mean SE. *P .05 vs baseline and ConNaL at end infusion.
indicated, uptake of Lac by the liver may be regulated by an active transport process.11-15 In sepsis, either the Lac membrane transport system became impaired, or the internal environment of the hepatocyte prevented uptake. These results emphasize, however, that the liver was not able to augment Lac extraction in sepsis. Although the renal cortex is stated to be an important consumer of Lac during acidosis,22 renal uptake of Lac also was not apparent in any of the groups. Under nonseptic conditions, previous studies that used lactic acid loading in rats and sheep have estimated that the kidney extracts approximately 20% to 30% of the load and that the excess Lac enters both oxidative and biosynthetic pathways.2,22,36 Under conditions of hyperlactemia, renal excretion contributes to Lac removal, but accounts for less than 1% to 2% of the infused load, and most of the extraction occurs by cellular metabolism.22,23,36 Yet, despite this severe degree of acidosis produced in the present study, little uptake of Lac by the kidney was observed (see Fig 3). Conceivably, the discrepancy between what is found in the literature and the results reported here may be related to the magnitude of the acidosis produced. Yudkin and Cohen21 showed that Lac removal by the kidneys of rats was reduced when the
perfusate pH level was reduced below 7.1. The extremely low pH level produced during LA infusion in the present study may have resulted in an inhibition of renal uptake of Lac, although this would not explain why clearance was not observed in the NaL groups. However, most likely, the explanation for the low Lac clearance observed in the present study is that any contribution of clearance by the kidney would be diluted from other sources that enter the vena cava. Based on previous studies in this model,10 renal blood flow approximates 10% to 20% IVC flow. In the present study, Lac concentrations were measured in the IVC in which catheters were positioned above (IVCupper) and below (IVClower) the kidney. Any contribution by the kidney would be diluted by other sources. These results point out, therefore, that the kidney may not contribute in large part to Lac clearance in sepsis because its blood flow is only a small amount of the total flow. Lac use by the soft tissues of the lower extremity was not apparent in this model. Although normally a lactate-producing organ, resting skeletal muscle can also extract Lac from the blood and use it primarily as fuel for oxidation during hyperlactatemia produced by exogenous infusion.2,24,25 Nevertheless, the results showed no increase in Lac uptake after
LACTATE METABOLISM IN SEPSIS
199
Fig 5. MAP, mean pulmonary capillary wedge pressure, and cardiac output are shown for the 4 groups in which sodium Lac or lactic acid was infused. Statistics by 2-way ANOVA and SNK. ——, Septic group lactic acid; ——, control group lactic acid; ——, septic group sodium lactate; ——, control group sodium lactate. (A) *P .05 vs baseline; %P .05 vs ConNaL; !P .05 vs ConLA, ConNaL; (B) *P .05 vs baseline; ??P .05 vs SepLA; %P .05 vs. ConNaL; !P .05 vs ConLA, ConNaL; #P .05 vs all other groups. (C) *P .05 vs baseline; #P .05 vs all other groups; %P .05 vs ConNaL; !P .05 vs ConLA, ConNaL.
NaL or LA infusion in the present study to show that this mechanism may be important in sepsis. The splanchnic organs are considered to be an important source of Lac production in sepsis. This production may result as a consequence of a reduction in oxygen delivery either owing to a decrease in bulk splanchnic flow, or because microvascular injury leads to a redistribution of perfusion to some areas while other areas remain perfused.4,5 Alternatively, a defect in the use of oxygen at a mitochondrial level may be primarily involved in sepsis,6,7 a conclusion that would agree with the investigations of Curtis and Cain,6 Hurtado et al,7 and our previous study.10 In terms of this model, it is further important to note that the increase in splanchnic Lac production observed between baseline and 4 hours is relatively small (.06 mmol/min), and the results shown in Fig 2 are similar to those previously reported,10 but did not reach statistical significance in the present study.
In the current study, SLM was different among the 4 groups studied (see Fig 2). A low pH level appeared to have a negative effect on Lac uptake, such that no uptake of Lac was observed in either the sepsis or nonsepsis groups after LA infusion. On the other hand, when NaL was infused, a large increase in splanchnic Lac uptake was observed in ConNaL, but not in SepNaL. Over the first 15minute interval after infusion, approximately 3 mmol/min Lac was consumed in ConNaL. The explanation for the different effect of NaL on splanchnic Lac uptake between the septic and nonseptic groups is not clear, but an impairment in the uptake of Lac by the monocarboxylate transport system or a defect in the mitochondrial metabolism may be responsible for this lack of an increase.6,37 In all groups, when either LA or NaL was infused, lung Lac production increased at end infusion (see Fig 2). The amount of Lac produced by the lung between end infusion and 15 minutes after infusion was
200
quite large and could account for 20 mmol over this interval. In another study, Bellomo et al28 reported that during endotoxemia in dogs, the lung produced Lac. Similar increases in lung Lac production have been reported in humans with acute lung injury.29,30 Although the mechanism responsible for this increase has not been identified,38 the present results show that lung production may contribute to delayed clearance of Lac in sepsis. As part of the protocol, an estimate of cephalic Lac consumption was determined. In both the septic and nonseptic groups, the results showed that after LA infusion, but not after NaL infusion, there was a significant increase in cephalic Lac consumption (see Fig 3). Although other organs may have contributed, most of the Lac metabolism was probably via the brain, although it is recognized that total brain blood flow was not measured. Because neuronal cells metabolize glucose and release Lac during conditions of ischemia, it was surprising that Lac uptake was detected from the brain. However, active transport of Lac occurred in cultured astroglial cells, particularly when the pH medium level was low.26,27 Thus, after LA infusion, when pH level was low, cephalic extraction measured between end infusion and 15 minutes after infusion accounted for approximately 3 mmol of uptake in the septic and nonseptic groups. Another interesting finding is that when LA or NaL was infused, a significant step down in lactate was observed between Lacjug and LacSVC in all groups (see Fig 3). Although this difference implies that there was mixing from several sources, the particular source of extraction is not clear. The venous effluent from the unidentified source could come from any organ in the chest, head, or neck, as well as the upper extremities. One possibility is that the astroglial cells of the spinal cord that drained by means of the vertebral system accounted for this step down.26,27 Whereas the site importantly awaits further evaluation, this source accounted for a significant amount of Lac extraction after infusion of either LA and NaL in both septic and nonseptic groups. From the results obtained in this study, it is clear that blood pH level may modulate Lac metabolism in sepsis. Nevertheless, although multiple organ sites were examined, it was not possible to account for all of the Lac metabolized over the experimental interval. For instance, in ConLA, approximately
CHRUSCH ET AL
62 mmol of lactate was consumed between end infusion and 15 minutes after infusion. Yet, over this interval, of the organs examined, 3 mmol was consumed by the liver, 3 mmol was consumed by the brain, and there was a further step down in Lac concentration between the brain and the SVC (ie, JugSVC). Excluding the increase in production by the lung, these measurements can only account for 20% of the Lac extracted over this interval. In ConNaL, the splanchnic organs accounted for 3 mmol or approximately 6% of the consumption. However, besides the step down found between Jug-SVC, no other organ system could be identified as a consumer of Lac in ConNaL. In SepLA, over the 15minute interval, the amount of Lac consumed was approximately one third of the total dose given (ie, 1/3 125) 41 mmol Lac. Of the multiple organs examined, only the brain could be identified as a metabolizer of Lac and accounted for approximately 8% of the total metabolized. In SepNaL, 25%, or approximately 31 mmol of 125 mmol, was metabolized over the initial 15-minute interval. Yet, besides the step down observed between Jug and SVC, no specific organ system of Lac metabolism could be identified. In terms of our analysis, moreover, it is important to note that the approach of Naylor et al36 was used in which plasma Lac rather than whole-blood Lac concentrations was substituted into the calculations. During Lac loading, it takes 1 or 2 hours for plasma Lac to equilibrate within the red cell. Conceivably, some of the decline in arterial decay may be related to uptake of Lac by the red cell. However, it would be expected that this contribution would occur to a similar degree in all 4 groups. Importantly, the present study points out that in sepsis, our knowledge of the extent to which different organs metabolize Lac is not clear. The different changes found in arterial pH, PCO2, and HCO3 levels in the treatment groups are shown in Fig 1. After LA was infused in ConLA and SepLA, PCO2 levels rose dramatically in both groups and then subsequently declined over the remaining course of the experiment. Interestingly, the rate of decline in PCO2 levels in the septic group was greater than that found in the nonseptic group. The explanation for this finding is not clear because the ventilator rate was not changed in either group. The decrease in whole-animal oxygen consumption shown in Table 3 may indicate that lactic acid infusion caused a decrease in cellular oxygen metabolism that resulted
LACTATE METABOLISM IN SEPSIS
201
in a reduction in CO2 production and, hence, arterial PCO2, although the organ responsible for this reduction in metabolism is not clear. On the other hand, after NaL infusion, PCO2 levels gradually rose over the course of the experiment to a similar extent in both SepNaL and ConNaL, and these increases in PCO2 levels were related to the generation of HCO3 that occurred in the 2 groups. In terms of the extensive preparation used in this study, it is recognized that this factor may have impacted on the results observed. Moreover, in the design of the study, there could be an effect other than pH level alone that contributed to the differences observed in Lac metabolism when sodium Lac versus lactic acid was administered in this model. When sodium Lac was given, the sodium concen-
tration would obviously be higher than when lactic acid was given. It is possible that this effect contributed to the differences between NaL and LA metabolism observed in this model. In summary, both effects of pH level and sepsis modulate the organ uptake of Lac in septic shock. Only a small amount of the Lac infused could be accounted for by the organs monitored in this model. The liver and kidney were not important metabolizers of Lac in sepsis. The lung contributed to Lac production, whereas Lac extraction by the brain and the organs that drain into the upper thorax contributed to Lac uptake in sepsis. These findings suggest that additional organs may account for Lac metabolism in sepsis. The net result of these changes is that the clearance of Lac is reduced in septic shock.
REFERENCES 1. Mizock BA, Falk JL: Lactic acidosis in critical illness. Crit Care Med 20:80-93, 1992 2. Madias NE: Lactic acidosis. Kidney Int 29:752-774, 1986 3. Nelson DP, King CE, Dodd SL, et al: Systemic and intestinal limits of O2 extraction in the dog. J Appl Physiol 63:387394, 1987 4. Drazenovic R, Samsel RW, Wylam ME, et al: Regulation of perfused capillary density in canine intestinal mucosa during endotoxemia. J Appl Physiol 72:259-265, 1992 5. Gutierrez G: Derangements of oxygen transport in shock states and sepsis, in Rippe JM, Irwin RS, Fink MP, et al (eds): Intensive Care Medicine, 3rd ed. Boston, MA, Little Brown and Company, 1996, pp 1972-1979 6. Curtis SE, Cain SM: Regional and systemic oxygen delivery/uptake relations and lactate flux in hyperdynamic, endotoxin-treated dogs. Am Rev Respir Dis 45:348-354, 1992 7. Hurtado FJ, Gutierrez AM, Silva N, et al: Role of tissue hypoxia as the mechanism of lactic acidosis during E. coli endotoxemia. J Appl Physiol 72:1895-1901, 1992 8. Falk JL, Rackow EC, Leavy J, et al: Delayed lactate clearance in patients surviving circulatory shock. Acute Care 11:212215, 1985 9. Levraut J, Ciebiera J, Chave S, et al: Mild hyperlactemia in stable septic patients is due to impaired clearance rather than overproduction. Am J Respir Crit Care Med 157:1021-1026, 1998 10. Chrusch C, Bands C, Bose D, et al: Impaired hepatic extraction and increased splanchnic production contribute to lactic acidosis in canine sepsis. Am J Respir Crit Care Med 161:517-526, 2000 11. Lupo MA, Cefalu WT, Pardridge WM: Kinetics of lactate transport into rat liver in vivo. Metab Clin Exp 39:374-377, 1990 12. Poole RC, Halestrap AP: Transport of lactate and other monocarboxylates across mammalian plasma membranes. Am J Physiol 264:C761-C782, 1993 13. Jackson VN, Halestrap AP: The kinetics, substrate, and inhibitor specificity of the monocarboxylate (lactate) transporter of rat liver cells determined using the fluorescent intracellular pH indicator 2 ,7 Bis(carboxyethyl-5(6)-carboxyfluorescein. J Biol Chem 270:1843-1849, 1995
14. Edlund GL, Halestrap AP: The kinetics of transport of lactate and pyruvate into rat hepatocytes. Biochem J 249:117126, 1988 15. Fafournoux P, Demingné C, Rémésy C: Carrier-mediated uptake of lactate in rat hepatocytes. J Biol Chem 260:292-299, 1985 16. Sestoft L, Marshall MO: Hepatic lactate uptake is enhanced by low pH at low lactate concentrations in perfused rat liver. Clin Sci 70:19-22, 1986 17. Berry MN: The liver and lactic acidosis. Proc R Soc Med 60:1260-1262, 1967 18. Baron PG, Iles RA, Cohen RD: Effect of varying PCO2 on intracellular pH and lactate consumption in the isolated perfused rat liver. Clin Sci 55:175-181, 1978 19. Forsythe SM, Schmidt GA: Sodium bicarbonate for the treatment of lactic acidosis. Chest 117:260-267, 2000 20. Cooper DJ, Walley KR, Wiggs BR, et al: Bicarbonate does not improve hemodynamics in critically ill patients who have lactic acidosis: A prospective controlled clinical study. Ann Intern Med 112:492-498, 1990 21. Yudkin J, Cohen RD: The effect of acidosis on lactate removal by the perfused rat kidney. Clin Sci 50:185-194, 1983 22. Yudkin J, Cohen RD: The contribution of the kidney to the removal of a lactic acid load under normal and acidotic conditions in the conscious rat. Clin Sci 48:121-131, 1975 23. Connor H, Woods HF, Ledingham JGG, et al: A model of L () lactate metabolism in normal man. Ann Nutr Metab 26:254-263, 1982 24. Gladden LB, Yates JW: Lactic acid infusion in dogs: Effects of varying infusate pH. J Appl Physiol 54:1254-1260, 1983 25. Brooks GA, Gaesser GA: End points of lactate and glucose metabolism after exhausting exercise. J Appl Physiol 49:1057-1069, 1980 26. Hamprech TB, Dringer R: Energy metabolism, in Kettenmann H, Ransom BR (eds): Neuroglia. New York, Oxford University Press, 1995, pp 473-487 27. Waniewski RA, Martin DL: Preferential utilization of acetate by astrocytes is attributable to transport. J Neurosci 18: 5225-5233, 1998
202
28. Bellomo R, Kellum JA, Pinsky MR: Transvisceral lactate fluxes during early endotoxemia. Chest 110:198-204, 1986 29. Brown SD, Clark C, Gutierrez G: Pulmonary lactate release in patients with sepsis and ARDS. J Crit Care 11:2-8, 1996 30. Kellum JA, Kramer DJ, Lee K, et al: Release of lactate by the lung in acute lung injury. Chest 111:1301-1305, 1997 31. Gomez A, Wang R, Unruh H, et al: Hemofiltration reverses left ventricular dysfunction during sepsis in dogs. Anesthesiology 73:671-685, 1990 32. Flecknell P: Laboratory Animal Anaesthesia, 2nd ed. San Diego, CA, Academic Press Ltd. 1999, p 112 33. Kellum JA, Bellomo R, Kramer DJ, et al: Hepatic anion flux during acute endotoxemia. J Appl Physiol 78:2212-2217, 1995 34. Presberg KW, Sznajder JL, Melendres J, et al: Distribu-
CHRUSCH ET AL
tion of pulmonary vascular resistance during lactic acid infusion in dogs. J Appl Physiol 68:1328-1336, 1990 35. Kirk BW, Raber MB: A practical apparatus for rapid determination of blood oxygen content. J Appl Physiol 34:724725, 1973 36. Naylor JM, Kronfeld DS, Freeman DE, et al: Hepatic and extrahepatic lactate metabolism in sheep: Effects of lactate loading and pH. Am J Physiol 247:E747-E755, 1984 37. Faeilli A, Orsenigo MD, Verri A, et al: Cl/HCO3 antiport affects H-lactate symport activity at the basolateral pole of jejunum enterocyte. Cell Physiol Biochem 8:151-157, 1988 38. Kilpatrick-Smith L, Erecinska DJ: Cellular effects of endotoxin in-vitro: 1. Effect of endotoxin on mitochondrial substrate metabolism and intracellular calcium. Circ Shock 11:8599, 1983
Results: Arterial clearance of Lac in the sepsis groups was slower as compared with the nonsepsis groups. In the liver, sepsis inhibited the uptake of LA as compared with the nonseptic group. In SP, both sepsis and pH affected Lac uptake in which an increase in uptake was found only after NaL infusion in the nonseptic group. In the head and neck region, Lac uptake was pH-level dependent and was found after LA infusion in the sepsis and nonsepsis groups. In the lung, Lac was produced after either LA or NaL infusion in all groups. Neither Kid nor SOL contributed to Lac uptake in any of the groups. Conclusion: Lactate clearance was reduced in sepsis. Both effects of pH level and sepsis modulated the organ uptake of Lac in septic shock. Only a small amount of the total Lac infused could be accounted for by the organs measured in the present study. This suggests that additional organs may account for lactate removal in sepsis. Copyright 2002, Elsevier Science (USA). All rights reserved.
N SEPTIC SHOCK, hyperlactemia and acidosis are often observed in severely ill patients.1-3 The factors that regulate the development of lactic acidosis in septic shock are complex and our understanding of the mechanisms responsible for its development are not clear. On the one hand, the mechanism of lactic acidosis in septic shock may reflect splanchnic overproduction caused by an impairment of aerobic metabolism4,5 occurring as a consequence of a decrease in bulk oxygen flow, a malredistribution of microvascular flow, or to a defect in the usage of oxygen at the mitochondrial level.6,7
On the other hand, rather than an increase in production, abnormalities in lactate (Lac) extraction by the liver and other organs may contribute to the hyperlactemia that accompanies septic shock.8-10 Uptake of Lac by the liver is thought to be regulated by an active transport process as well as by diffusion. At blood concentrations between 1 to 5 mmol/L, hepatic Lac uptake appears to be mediated primarily by an active transport process.11-15 Lac is actively cotransported with a hydrogen ion, so that when Lac anion is uptake up by the liver, the concomitant uptake of hydrogen ion results in the formation of bicarbonate ion in the plasma. At high Lac concentrations, diffusion of the nondissociated anion as well as other nonspecific active transport processes appear important in its metabolism. Although the factors that regulate hepatic uptake of Lac by the Lachydrogen cotransporter are still being worked out, blood pH level may be one important factor in the regulation of Lac uptake. When the pH level external to the hepatocyte is less than the pH level inside the cell, Lac uptake is increased as compared with the opposite situation.11-16 However, as pH level in the hepatocyte falls to very low values (ie, 6.9), the hepatocyte becomes a Lac producer
I
From the Departments of Medicine, Biochemistry and Molecular Biology; and Pharmacology and Therapeutics and Anesthesiology, University of Manitoba, Winnipeg, Manitoba, Canada. Supported by the Manitoba Heart and Stroke Foundation and by a fellowship from the Manitoba Lung Association (E.B.). Address reprint requests to Steven Mink, MD, Health Sciences Centre, GF-221, 700 William Ave, Winnipeg, Manitoba, Canada R3E-0Z3. E-mail: [email protected]. Copyright 2002, Elsevier Science (USA). All rights reserved. 0883-9441/02/1703-0007$35.00/0 doi:10.1053/jcrc.2002.35816 188
Journal of Critical Care, Vol 17, No 3 (September), 2002: pp 188-202
LACTATE METABOLISM IN SEPSIS
rather than a Lac consumer, and this may contribute to the increase in Lac and acidemia observed in sepsis.17,18 Although blood pH level has theoretical implications on modulating the metabolism of Lac in septic shock, there is little information available as to whether this effect would be clinically important. Nevertheless, alkali treatment such as sodium bicarbonate is often used to treat the low pH level that develops in sepsis.19 The lack of beneficial response observed with this treatment may occur as a consequence of the resulting acid-base changes that may negatively impact on Lac clearance among the different organs.19,20 Under nonseptic conditions, the liver and kidney are thought to be the major sources of Lac uptake. pH level has been found to alter Lac uptake by these organs,2,17,18,21-23 as well as skeletal muscle,24,25 and brain.25-27 Moreover, the lung has been shown to produce Lac under many pathologic conditions28-30 and the effect of pH level on this production needs to be considered in the understanding of Lac clearance in sepsis. Accordingly, because blood pH level may be important in determining the extent to which different organs may metabolize Lac in septic shock, this question was examined in a canine model of Escherichia coli sepsis.10,31 Either hydrogen Lac or its salt, sodium lactate (NaL), was administered as a bolus infusion after 4 hours of bacteremia in respective groups of dogs. Major organs that are associated with Lac metabolism including the liver, gut, lung, kidneys, as well as the brain and skeletal muscle were compared in respective septic and nonseptic groups. The objectives were to examine which organs were responsible for Lac metabolism in sepsis and to delineate the effect of pH level on its metabolism. METHODOLOGY
Animal Preparation This study was performed in accordance with protocols approved by the University Animal Care Committee and conforms with the Guide for the Care and Use of Laboratory Animals published by the National Institutes of Health (NIH Publication No. 85-23, revised 1996). Mongrel dogs (25–30 kg) were anesthetized with thiopental sodium (20 mg/kg, intravenously) and were given a loading dose of sufentanil (10 ug/kg), followed by maintenance doses of sufentanil (1 ug/min) and midazolam (5 ug/kg/min).31,32 The trachea was intubated with an endotracheal tube and the lungs were mechanically ventilated (Harvard Apparatus, Hollison, MA).
189
The ventilator settings (approximate rate of 10 breaths/min; tidal volume, 15 mL/kg) were adjusted to maintain initial arterial PCO2 at approximately 35 mm Hg and pH level at approximately 7.35. These settings were not changed throughout the entire experiment. The inspired oxygen concentration was titrated to maintain arterial PO2 greater than 150 mm Hg, so that arterial hypoxemia did not contribute to any of the findings observed over the course of the experiment. In this study, 2 preparations were used because it was often not feasible to obtain measurements of lactate metabolism from the many organs examined in a single preparation in most cases. In setting up both preparations, the animal was placed in the supine position, and vascular catheters were inserted percutaneously under sterile conditions. An arterial catheter was placed into the left femoral artery and advanced into the midabdominal aorta. This catheter was used to obtain arterial lactate concentration (LacArt), oxygen content (CaO2) (see later), and arterial blood gases. It was also used to measure mean arterial pressure (MAP), and for infusion of one of the lactate solutions (see later). Into the left jugular vein, a thermistor-tipped catheter was advanced into the pulmonary artery to measure mean pulmonary arterial pressure (Ppa), mean pulmonary wedge pressure (Pwp), mean right arterial pressure (Pra), and cardiac output (CO) by the thermodilution technique (Columbus Instruments, Columbus, OH). The average of 3 determinations was used in the reporting of CO for each condition. Samples for mixed venous lactate (LacMV) and oxygen content (CMVO2 ) were obtained from the distal port of this catheter. In one preparation, splanchnic lactate metabolism (SLM), hepatic lactate metabolism (HLM), kidney lactate metabolism (KLM), and lower extremity (LEM) lactate metabolism were primarily determined. This preparation was similar to that described by Kellum et al33 and has previously been described.10 The abdomen was opened by a midline incision, after which the spleen was removed so that the splenic vein could be cannulated. A catheter was advanced into the splenic vein and positioned in the portal vein at the point at which it enters the liver. The portal vein catheter measured 0.2 cm in external diameter and contained one endhole and 2 sideholes, the latter of which were within 1.0 cm of the tip. The portal vein catheter was used to obtain blood samples for portal vein lactate (Lacpov) and portal vein oxygen content CpovO2 ) (see later). In addition to the catheter, a flow probe (Transonic, Ithaca, NY) was placed around the proximal portion of the portal vein to measure portal blood flow (Qpov). Another catheter with characteristics similar to the portal vein catheter was advanced from the right jugular vein into the hepatic vein. The hepatic vein catheter was usually placed into the right hepatic vein because of easier placement, although hepatic Lac concentrations and oxygen contents were similar between right and left hepatic veins when determinations were compared in a previous study.10 The hepatic vein catheter was positioned into the hepatic vein by palpation, and proper positioning of the catheter was verified at autopsy. The hepatic vein catheter was used to obtain hepatic vein lactate (Lachv) and hepatic oxygen content (ChvO2 ). A flow probe (Transonic) was placed around the hepatic artery to measure hepatic artery flow (Qha). Two catheters identical in size to the hepatic and portal catheters (0.2 cm) were inserted percutaneously through the right and left femoral veins, respectively, and advanced into the inferior vena cava. One catheter was placed above the renal veins
190
(IVCupper), whereas the other catheter was placed below the renal veins (IVClower). A flow probe was placed around the IVC at a position below the renal veins to measure IVC flow (QIVC). From Lac determinations made above and below the renal veins, KLM was assessed. LEM was calculated from the difference between arterial Lac and the blood Lac concentration in the lower vena cava (ie, LacArt LacIVClower). In a second preparation, Lac metabolism by the thorax and cephalic region (to assess primarily brain uptake) was determined. The spleen was removed for consistency with the previous preparation. A right jugular venous catheter (0.2 cm) was inserted and was advanced in a retrograde manner to the base of the brain to measure jugular venous lactate (LacJug) and jugular venous oxygen concentration. A flow probe was placed around one of the carotid arteries to measure carotid artery flow (Qcar). In addition, because the proximal port of the Swanz-Ganz catheter (Edwards Life Sciences Corp., Irvine, CA) was located in the left brachiocephalic vein, not too far from the superior vena cava (SVC), this port was used to obtain LacSVC, so that the lactate difference measured between the jugular vein and SVC (LacJug) LacSVC) could be determined. The fluid-filled catheters to measure MAP and pulmonary artery pressures were connected to transducers (Cobe, Lakewood, CO) that were referenced relative to the left atrium. Measurements were obtained at end expiration during which the ventilator was turned off for approximately 3 to 5 seconds. All signals were displayed on an 8-channel recorder (Astro Med, West Warwick, RI). Furthermore, the ultrasonic flow probes used in this study have been found to be stable with little 0 drift over the course of the experiment; before and after each experiment, 0 flow that was obtained electrically was checked with that measured when the vessel was transiently occluded.
Protocols Baseline measurements (presepsis) were obtained 1 hour after completion of the preparation when hemodynamics (ie, MAP and CO) were stable. The animals were randomized into sepsis and nonsepsis groups. In the sepsis groups, a bolus of 1010 colony forming units of live E. coli (serotype 0111:B4) was given over a 30-minute interval.10,31 A constant infusion of approximately 5 109 colony forming units of E. coli per hour was then maintained for the remainder of the study. After 4 hours of infusion, measurements were again made and this interval was termed the 4-hour condition. The reason for choosing the 4-hour period was that systemic hypotension has been shown to occur at the 4-hour period in this model.31 In the nonseptic group (control group), the protocol was the same except that a similar amount of normal saline was administered over this period. After the 4-hour period, further randomization was performed. In the septic and nonseptic groups, one half of the animals in each group received 5 mmol/kg of lactic acid, whereas the other half received a similar dose of NaL (Sigma, St. Louis, MO). NaL or LA was mixed in .500 L of 5% dextrose in water (D5W). The 4 groups were designated as follows: (1) septic group with lactic acid (SepLA), (2) septic group with sodium lactate (SepNaL), (3) control group with lactic acid (ConLA), and (4) control group with sodium lactate (ConNaL). There were approximately 12 animals in each group, and, for the most part, approximately 6 animals were studied in preparation 1, while the other 6 were examined in preparation 2. Sometimes, there was overlap in the 2 preparations, so that additional measure-
CHRUSCH ET AL
ments could be obtained in a given experiment. In SepLA, HLM and LEM were also studied in 2 experiments as part of preparation 2, whereas SLM was studied in one experiment as part of preparation 2. In ConNaL, SLM and HLM were studied in one experiment as part of preparation 2. In ConLA, KLM was studied in 3 experiments as part of preparation 2 and in 3 experiments as part of preparation 1, whereas in SepLA, KLM was studied in 4 experiments as part of preparation 1 and in 2 experiments as part of preparation 2. Differences were not observed when the 2 preparations were compared. In all groups, either the LA or NaL treatment was given over 30 minutes through the catheter positioned in the lower abdominal aorta. Lac was infused into the aorta rather than into the venous circulation based on the work of others that showed that the infusion of lactate into the venous circulation of normal dogs resulted in intravascular hemolysis and the formation of microemboli in the lungs.34 This effect leads to pulmonary hypertension, which does not occur when Lac is infused into the aorta. In the protocol, hemodynamics were measured at end infusion (during which the infusion was completely off) and at 15, 30, 45, 60, and 75 minutes after infusion. Hemodynamic measurements included CO, Ppa, Pwp, Pra, MAP, arterial hematocrit (Hct), as well as organ hematocrits, and flows. Arterial samples for Lac were also obtained at the latter intervals, whereas samples for mixed venous and organ Lacs were measured at all of these intervals except for 60 and 75 minutes after infusion. Similarly, arterial blood samples for PO2, PCO2, pH levels, and oxygen contents were determined at all intervals, whereas samples for mixed venous and organ samples were determined at all of the earlier-mentioned intervals except for 60 and 75 minutes after infusion. Moreover, in 5 other nonseptic dogs (ConNSS), and in 5 other septic dogs (SepNSS), time control experiments were performed. In these 2 groups, at the 4-hour period, a similar amount of sodium chloride (5 mmol/kg) mixed in .5 L D5W was infused over 30 minutes. Then measurements were determined at intervals identical to the other groups. The rationale for ConNSS and SepNSS was to contrast the effect of lactate versus saline solution on the parameters measured. In these experiments, both preparations 1 and 2 were combined because KLM and LEM were found to have little effect on Lac metabolism in the treatment groups (see Discussion section). In all groups, oxygen contents were directly measured by the carbon monoxide scrubbing technique.35 Lac determination was made by an automated Lac dehydrogenase–based assay. Previous experiments have indicated that the coefficient of intraassay and interassay variation of Lac measurements were 0.57% 0.21% and 1.2% 0.33%, respectively,10 and therefore only one sample was obtained during each measurement. When measurements were obtained, moreover, the catheter was agitated in a to-and-fro motion to reduce any effect of streaming.
Data Analysis In the determination of organ Lac metabolism, uptake of Lac from plasma rather than blood was used based on the work of Naylor et al.36 These investigators showed that erythrocyte Lac equilibrates only slowly with plasma, and therefore little net erythrocyte Lac would be removed from the red cell during flow through the various organs examined. SLM was calculated from (LacArt Lacpov) Qpov (1 Hctpov). HLM was calculated
LACTATE METABOLISM IN SEPSIS
from [Qha LacArt (1 HctArt) Qpov Lacpov (1 Hctpov) (Qpov Qha) Lachv (1 Hcthv]. Lung Lac metabolism (LLM) was calculated by (LacMV LacArt) CO (1 HctMV). Cephalic lactate metabolism (CLM) was calculated by (LacArt LacJug) 2 Qcar (1 HctJug). It is important to note that when Lac concentration differences across an organ were determined, total blood Lac was used because blood flow was not considered in the equation. KLM was calculated by (LacIVCupper LacIVClower); LEM was calculated by (LacArt LacIVClower). Lac difference between the brain and SVC was calculated from (LacJug LacSVC). After either NaL or LA was infused, the decay in arterial blood Lac was analyzed in terms of 1- and 2-compartment models (NW StatPak, Portland, OR).9 Of the mathematic functions tested (linear, log, exponential, and power), the results indicated that a 1-compartment exponential model gave the best fit. The equation used to describe the decay was Aebx where “b” is the parameter of the curve fit, x time, and A is the Y intercept. Whole-animal and organ oxygen consumption were calculated from the respective arterial and venous content differences and flows.10 Whole-body consumption was calculated from (CaO2 CmvO2) CO; splanchnic oxygen consumption was calculated from (CaO2 CpovO2) Qpov; hepatic oxygen consumption was calculated from [(Qha CaO2 Qpov CpovO2) (Qha Qpov) ChvO2)]; and cephalic oxygen consumption was calculated from 2Qcar (CaO2 CjugO2). Furthermore, in conjunction with arterial pH and PCO2 levels, arterial bicarbonate was calculated at each measurement interval from the Henderson-Hasselbalch equation in which pH level 6.1 log([HCO3]/0.03 PCO2).
Statistics Between-group analyses were determined by 2-way repeated measures analysis of variance (ANOVA) split plot design (factor A, different treatment groups; factor B, different time periods), in which the interaction between the 2 factors was assessed. If a significant interaction was present, a StudentNewman-Keuls’ (SNK) multiple range test was used to determine where differences occurred. Furthermore, within a given group, statistical analyses between all of the conditions were not reported because the comparison between baseline and sepsis was the primary measurement examined. Finally, when a single parameter between groups was examined, a one-way nonrepeated ANOVA and SNK were used. Results are expressed as mean 1SE.
RESULTS
Lactate Metabolism The decay in LacArt measured in the 4 treatment groups is shown in Fig 1 (left upper panel). In the septic and nonseptic groups, LacArt in SepLA were higher than those in ConLA over the entire decay. The curves were plotted as an exponential function in which the parameter “b” was an expression of the curve fit. In SepLA, “b” was less negative than that found in ConLA (.0116 .0015 vs. .0204 .0023; P .05), which indicates that the decay in concentrations was slower in SepLA. Further-
191
more, at end infusion, peak lactate concentration in SepLA was higher than that measured in ConLA and, consistent with this finding, the parameter “A” in the curve fit was also higher in SepLA (9.5 56 vs. 7.045 .48 mmol/L; P .05). In SepNaL, LacArt was higher at 15 and 30 minutes after infusion as compared with all of the other groups, and LacArt remained higher than those in ConNaL over the remainder of the experiment. The parameter “b” in SepNaL was again less negative than that found in ConNaL (.0117 .0014 vs. .0180 .0017; P .05), but was not different than that calculated in SepLA. In SepNaL, “A” was higher than that found in ConNaL, but did not reach statistical significance between groups (9.12 .44 vs 7.65 .57 mmol/L). In terms of the exponential model used, the correlation coefficient (R) was similar in the 4 groups and measured .93 .009 in SepLa, .95 .008 in ConLA, .92 .03 in SepNaL, and .95 .012 in ConNaL. Furthermore, after infusion of either NaL or LA, it is important to note that steadystate plasma concentrations of Lac were reached within 15 minutes from the start of the infusion. LacArt was measured at 15 minutes and then at 30 minutes after infusion. The results (n 9) were similar between the 2 time periods and measured 10.3 1.4 and 11 2 mmol/L, respectively. SLM was different among the 4 groups (see Fig. 2, left upper panel). After infusion of NaL, Lac consumption increased in ConNaL, but not in SepNaL. The increase in splanchnic uptake measured in ConNaL was higher than that found in all of the other groups. In SepLA and ConLA, SLM remained at net 0 flux over the course of the experiment. HLM is shown in Fig. 2 (left lower panel). After LA was infused, uptake of Lac increased in ConLA, but not in SepLA. After infusion of NaL, HLM was not significantly altered as compared with baseline in either the septic or nonseptic group. Lung Lac metabolism was determined from the difference between mixed venous lactate (LacMV) and arterial lactate (LacArt) times cardiac output. Because small differences in (LacMV LacArt), when multiplied by CO, may account for large differences in lung metabolism (LacMV LacArt) and the product of (LacMV LacArt) CO were analyzed separately (see Fig 2 right upper and lower panels). There were significant increases in (LacMV LacArt) and lung lactate production at
192
CHRUSCH ET AL
Fig 1. Arterial lactate, arterial pH, arterial PCO2, and HCO3 levels are shown for the 4 groups in which sodium Lac or lactic acid was infused. Statistics were performed by 2-way ANOVA and SNK. ——, Septic group lactic acid; ——, control group lactic acid; ——, septic group sodium lactate; ——, control group sodium lactate. (A) *P .05 vs baseline; !P .05 vs ConLA and ConNaL; $$$P .05 vs ConLA; #P .05 vs all other groups. (B) *P .05 vs baseline; #P .05 vs all other groups; ?P .05 vs. SepNaL, ConNaL. (C) *P .05 vs baseline; ?P .05 vs SepNaL, ConNaL; #P .05 vs all other groups; ??P .05 vs SepLA; !P .05 vs ConLA, ConNaL; %P .05 vs ConNaL. (D) %P .05 vs ConNaL; ?P .05 vs SepNaL, ConNaL; *P .05 vs baseline; $$$P .05 vs ConLA.
end infusion and at 15 minutes after infusion in all groups. Cephalic Lac extraction is shown in Fig 3 (left upper panel). There was a marked pH level dependence of Lac metabolism by the cephalic region. After LA was infused, Lac extraction in both the sepsis and nonsepsis groups was substantially increased at end infusion as compared with baseline. However, no Lac extraction was observed after NaL infusion in the septic and nonseptic groups. The Lac difference between the jugular vein (LacJug) and SVC (LacSVC) was also determined at each measurement interval (Fig 3; left lower panel).
The proximal port of the Swanz-Ganz catheter was located in the left brachiocephalic vein near the SVC, so that a venous sample could be obtained from this site to measure the difference in Lac concentrations between Jug and SVC. In all the treatment groups, a significant decrease in Lac concentration was observed at end infusion in the SVC sample. Kidney Lac difference is shown in Fig 3 (right upper panel). The results showed that no uptake of Lac was observed at any of the intervals in the 4 groups. Similar findings were observed in the uptake of Lac by the lower extremity (see Fig 3, right
LACTATE METABOLISM IN SEPSIS
193
Fig 2. Splanchnic Lac metabolism, hepatic Lac metabolism, mixed venous minus arterial Lac, and lung Lac metabolism are shown for the 4 groups in which sodium Lac or lactic acid was infused. Positive values indicate uptake, whereas negative values indicate production. Statistics by 2-way ANOVA and SNK. ——, Septic group lactic acid; ——, control group lactic acid; ——, septic group sodium lactate; ——, control group sodium lactate. (A) *P .05 vs baseline; #P .05 vs all other groups. (B) *P .05 vs baseline. (C) *P .05 vs baseline; @P .05 vs SepLA, SepNSS. (D) *P .05 vs baseline.
lower panel) in which no evidence of extraction was observed among the treatment groups. In the septic and nonseptic saline groups (ie, SepNSS and ConNSS), normal saline solution was infused to contrast the difference in organ metabolism between the 2 solutions. The arterial lactates are shown in Table 1. The increases in LacArt observed in this model are modest. In SepNSS, although LacArt significantly increased at 4 hours as compared with baseline (see Table 1), and although slight further increases in arterial Lac occurred over the course of the experiment, LacArt were not different between SepNSS and ConNSS. During saline infusion, SLM and HLM showed only small changes over the course of the experiment. At end infusion, lung Lac production did not increase as
compared with the Lac treatment groups. Similarly, cephalic Lac uptake and (LacJug LacSVC) did not change over the course of the experiment in either SepNSS or ConNSS. Acid Base Changes The changes in pH level found with either LA or NaL infusion were different among the septic versus nonseptic groups (see Fig 1; left lower panel). It is important to note that in the design of the experimental protocol, the ventilator settings were not changed over the entire measurement period. pH level in SepLA decreased to approximately 6.9 at end infusion, which was significantly lower than that found in ConLA. Values of pH in SepLA remained lower than those found in ConLA
194
CHRUSCH ET AL
Fig 3. Cephalic Lac extraction, jugular Lac minus superior vena cava Lac, kidney Lac metabolism, and lower extremity Lac are shown for the 4 groups in which sodium Lac or lactic acid was infused. Positive values indicate uptake, whereas negative values indicate production. Statistics by 2-way ANOVA and SNK. ——, Septic group lactic acid; ——, control group lactic acid; ——, septic group sodium lactate; ——, control group sodium lactate. (A) *P .05 vs baseline; ?P .05 vs SepNaL, ConNaL. (C) *P .05 vs baseline.
for most of the postinfusion interval. In SepNaL, pH level did not change as compared with baseline over the entire measurement period. In contrast, in ConNaL, pH level gradually rose as compared with the baseline value. These changes were significantly different between the septic and nonseptic groups. The changes in PCO2 found over the course of the experiment are shown in Fig 1 (right upper panel). After infusion of LA, there were large initial increases in PCO2 that occurred to similar extents in both the septic and nonseptic groups. After end infusion, PCO2 in SepLA declined more rapidly as compared with ConLA, such that PCO2 values were significantly lower for most of the postinfusion period in SepLA. In contrast, after infusion of NaL, PCO2 increased to similar degrees
as compared with baseline in both the septic and nonseptic groups. The calculated arterial bicarbonates are shown in Fig 1 (right lower panel). In ConLA, HCO3 remained unchanged as compared with baseline over the entire experimental protocol, whereas in SepLA, HCO3 decreased from 10 to 6 mmol/L between 4 hours and 75 minutes after infusion. In contrast, when NaL was infused in the septic and nonseptic groups, HCO3 increased steadily over the course of experiment, and this occurred to a slightly greater extent in the nonseptic group. In the saline groups, pH level decreased in both SepNSS and ConNSS after saline infusion and remained relatively constant over the subsequent intervals (see Table 2). Corresponding changes in PCO2 and HCO3 are also shown in Table 2.
LACTATE METABOLISM IN SEPSIS
195
Table 1. Lactate Metabollism in ConNSS and SepNSS
ConNSS (n 5) Arterial Lac (mmol/L) Splanchnic Lac metabolism (mmol/min) Hepatic Lac metabolism (mmol/min) Lung Lac metabolism (mmol/min) Cephalic Lac metabolism (mmol/min) (Jug SVC) Lac (mmol/L) SepNSS (n 5) Arterial Lac (mmol/L) Splanchnic Lac metabolism (mmol/min) Hepatic Lac metabolism (mmol/min) Lung Lac metabolism (mmol/min) Cephalic Lac metabolism (mmol/min) (Jug SVC) Lac (mmol/L)
Baseline
4 Hours
End Infusion
15 Minutes After Infusion
30 Minutes After Infusion
1.3 .2 .009 .005
1.5 .3 .011 .01
2.1 3* .14 .1||
2.3 3*,† .025 .02
2.1 3†,‡ .004 .02
1.9 1†,§ .002 .01
.02 .02
.02 .01
.07 .1¶
.05 .02
.05 .02
.03 .02
.27 .08
.22 .06
.44 .12
.38 .21
.014 .002 .05 .1
.014 .004 0 .15
.41 .22*
.84 .35†
.034 .013#,¶
.034 .008
.26 .1*
1.8 .2 .02 .01
1.9 .3 .007 .01
3.1 3*,† .009 .02**
.01 .02
.01 .02
.06 .08¶
.08 .07
.11 .06
.07 .22*
.016 .07
.012 .00
.04 .04
.03 .06
.1 .2 3.1 3*,† .012 .01 .04 .04
.01 .008 .05 .1 2.6 3†,§ .022 .01
45 Minutes After Infusion
.01 .006 .05 .1 2.2 4§ .0146 .005
.01 .04
.05 .02
.42 .07†
.30 .07
.25 .09
.014 .01#,¶
.04 .01
.016 .01
.018 .01
.03 .04*
.09 .06
.07 .1
.07 .1
NOTE: Mean SE. ConNSS and SepNSS are control and septic normal saline groups, respectively. Positive values are uptake, whereas minus values are production. By split plot ANOVA and SNK. *P .05 vs SepLA, SepNaL, ConLA, and ConNaL. † P .05 vs baseline. ‡ P .05 vs SepNaL, SepLA, ConNaL. § P .05 vs SepNaL and SepLA. || P .05 vs SepNaL. ¶ P .05 vs ConLA. # P .05 vs SepLA. **P .05 vs ConNaL.
Organ Blood Flow and Oxygen Consumption and Systemic Hemodynamics In Fig 4, the organ blood flows measured during infusion for the 4 treatment groups are shown. The response of hepatic artery blood flow (Qha) to either LA or NaL infusion was different between the septic and nonseptic groups. In the septic treatment groups, the infusion of LA or NaL caused an increase in Qha, that occurred to the greatest extent in SepNaL, whereas there was no change in Qha in the nonseptic groups (see Fig 4, left upper panel). In contrast, infusion caused the opposite changes in portal blood flow (Qpov) between the septic and nonseptic groups (Fig 4, left lower panel). There were no differences in IVC flow after LA or NaL infusion among the 4 treatment groups studied (see Fig 4, right upper panel). In all the treatment groups, carotid blood flow increased during Lac in-
fusion, and this occurred to the greatest extent in SepNaL (right upper panel). In SepLA, whole-animal oxygen consumption decreased at end infusion as compared with baseline and as compared with end infusion in ConNaL (Table 3). Splanchnic oxygen consumption, hepatic oxygen consumption, and cephalic oxygen consumption were not different between conditions and among groups over the course of the study. Systemic hemodynamics are shown in Fig 5. In the septic treatment groups, MAP (upper panel) measured at 4-hours decreased to a similar extent from baseline in all groups. At end infusion, MAP increased as compared to the 4-hour measurement, and then gradually declined over the remainder of the study. There were no changes in MAP over the course of the experiment in the nonseptic groups. In Fig 5 (middle panel), Pwp are shown for the NaL and LA groups. Pwp were slightly lower for the
196
CHRUSCH ET AL
Table 2. Acid Base and Hemodynamics in ConNSS and SepNSS
ConNSS (n 5) Arterial pH level Arterial PCO2 (mm Hg) Calculated arterial bicarbonate (mmol/L) Qha (L/M) Qpov (L/M) MAP (mm Hg) SepNSS (n 5) Arterial pH level Arterial PCO2 (mm Hg) Calculated arterial bicarbonate (mmol/L) Qha (L/M) Qpov (L/M) MAP (mm Hg)
End Infusion
15 Minutes After Infusion
30 Minutes After Infusion
45 Minutes After Infusion
Baseline
4 Hours
7.36 .01 26 2 9.4 1.7
7.36 .02 23 .05 9 .3
7.21 .01*,† 30 2|| 9.6 1.4#
7.21 .01*,‡ 27 3 8 1.4§
7.22 00*,§ 27 2¶ 7.6 1.2§
7.23 .01*,§ 26 ¶ 7.6 1.3§
.14 .02 .14 .01 110 10
.12 .04 .13 .01 113 12#
.17 .04** .33 .03#,†† 125 12#
.15 .04 .24 .02# 117 12#
.16 .04 .24 .02 118 12#
.15 00 .21 .02 120 12#
7.37 .01 25 1 8.4 .5
7.27*,†† 27 1 7.4 .6
.16 .08 .33 .03‡‡ 112 3
.21 .08 .20 .05 61 8*,‡‡,|| ||
7.17 .01*,† 36 2* 12.4 1.4* .37 .06*#,††,‡‡ .21 .05#,†† 95 11*#,‡‡
7.17 .01*,‡ 33 2*,|| || 10.4 1.0§ .35 .05*#,††,‡‡ .17 .04 91 11*,#
7.14 01*,§,‡‡ 33 2¶¶ 10.4 1.2§
7.15 .01*,‡‡,§§ 33 2¶¶ 9.8 7§
.36 .07*,‡‡ .17 .04 86 8*,‡‡,|| ||
.37 08*,‡‡ .15 .04 87 10*,‡‡,|| ||
NOTE: Mean SE. By 2-way ANOVA and SNK. *P .05 vs baseline. † P .05 vs SepLA, SepNaL, ConLA, ConNaL. ‡ P .05 vs SepNaL, SepLA, ConNaL. § P .05 vs SepNaL and ConNaL. || P .05 vs SepLA and ConLA. ¶ P .05 vs SepNaL. # P .05 vs ConNaL. **P .05 vs SepNaL and SepLA. †† P .05 vs ConLA. ‡‡ P .05 vs ConNSS. §§ P .05 vs ConLA, SepNaL, ConNaL. || || P .05 vs ConLA and ConNaL. ¶¶ P .05 vs SepLA.
septic groups at the 4-hour period as compared with the nonseptic groups, so that for similar amounts of fluid given during volume infusion, Pwp increased to higher values at end infusion in the nonseptic groups. For the most part, the changes in CO (see Fig 5, bottom) followed those in Pwp, and CO measured at the end-infusion period in SepLa was the lowest as compared with the other treatment groups. DISCUSSION
In the present study, the results showed that both blood pH level and sepsis had differential effects on Lac metabolism in this canine model of E. coli–induced sepsis. After LA or NaL was infused in the nonseptic groups, the half-life of the infusate as assessed from the arterial Lac decay was approximately 15 minutes, as compared with the nonseptic groups in which the half-life of arterial Lac decay approximated 75 minutes. The modeling constant “b” was more negative in the nonseptic
versus septic groups, suggesting that blood Lac given by either LA or NaL is cleared less effectively in sepsis. In addition, the peak Lac levels measured at end infusion in the sepsis groups were higher than those found in the nonsepsis groups. This did not appear because of a change in the volume of distribution of lactate (VolLac) among the groups because a similar dose of lactate (5 mmol/kg) was given in all experiments. VolLac (ie, 5 mmol/kg divided by lactate found between end infusion and 4 h) was not different and measured 0.69 .04 L/kg in ConLA, 0.65 .08 L/kg in SepLA, 0.689 .05 L/kg in ConLA, and 0.629 .037 L/kg in SepNaL. Because steady-state infusion had been reached, during infusion the high VolLac may indicate that Lac was being metabolized at this time. Based on the average weight of the animals (ie, 25 kg), approximately 125 mmol of Lac was given in each experiment. In ConLA, the half-life of Lac was approximately 15 minutes, so that approxi-
LACTATE METABOLISM IN SEPSIS
197
Fig 4. Hepatic artery flow, portal blood flow, IVC flow, and carotid artery flow are shown for the 4 groups in which sodium Lac or lactic acid was infused. The carotid artery flow includes flow from both arteries. Statistics by 2-way ANOVA and SNK. ——, Septic group lactic acid; ——, control group lactic acid; ——, septic group sodium lactate; ——, control group sodium lactate. (A) *P .05 vs baseline; !P .05 vs ConLA and ConNaL; %%P .05 vs SepNaL; #P .05 vs all other groups; $$$P .05 vs ConLA. (B) *P .05 vs baseline. (C) *P .05 vs baseline; %P .05 vs ConLA; ??P .05 vs SepLA; !P .05 vs ConLA, ConNaL; #P .05 vs all other groups. (D) *P .05 vs baseline; #P .05 vs all other groups.
mately 62 mmol was metabolized during this interval. In Fig 2 (left lower panel), between end infusion and 15 minutes after infusion, the absolute hepatic uptake of Lac could be approximated by the area under the curve for Lac concentrations measured. Then the amount extracted would be 0.4 mmol/min 15 min/2 3 mmol. Based on the work of Lupo et al11 and Naylor et al,36 the normal canine liver would be able to metabolize approximately 1 to 2 mmol/min of lactate under nonseptic conditions. In ConLA, the maximal rate of lactate uptake was 0.4 mmol/min, which was slightly less than that predicted based on the maximal uptake as reported in other studies. A major factor determining the hepatic uptake of Lac appears to be blood pH level. When pH level external to the hepatocyte is less than pH level in-
side the cell, Lac uptake has been found to increase as compared with the opposite situation.16,18 Thus, the initial large increase in hepatic uptake found at end infusion in ConLA would be consistent with the effect of blood acidemia that resulted in an accelerated uptake of lactate. However, it has also been shown that when hepatic intracellular pH level decreases to less than 6.9, then hepatic lactate uptake is subsequently curtained. Blood pH level approached 6.9 at end infusion in ConLA. The rapid fall in Lac uptake observed at the subsequent intervals could be explained by the development of a low intracellular pH level. In contrast to what was found in the nonseptic groups, hepatic Lac uptake was not observed in the septic groups, despite the fact that blood pH level was changed to a similar extent among the comparable groups. As previously
198
CHRUSCH ET AL
Table 3. Oxygen Consumption in the Treatment Groups Baseline
Hepatic oxygen consumption SepLA (n 7) ConLA (n 5) SepNaL (n 8) ConNaL (n 6) Splanchnic oxygen consumption SepLA (n 7) ConLA (n 5) SepNaL (n 8) ConNaL (n 6) Cephalic oxygen consumption SepLA (n 5) ConLA (n 5) SepNaL (n 4) ConNaL (n 6) Whole-body oxygen consumption SepLA (n 12) ConLA (n 12) SepNaL (n 12) ConNaL (n 12)
4 Hours
End Infusion
30 Minutes After Infusion
16 6 15 4 19 6 14 4
5.4 11 13 15
5 3 3 2
7 7.9 20 12
4 2 5 4
15 11 25 22
4 3 3 2
18 10 16 22
16 2 17 7 14 3 16 3
17 10 10 13
4 3 4 2
62 63 93 12 4
9 8 5 10
4 2 1 3
10 4 61 71 16 7
8 9 5 9
2 3 1 3
4 8 7 12
1 3 1 3
5 5 3 8
2 1 2 3
214 190 210 223
18 26 25 26
140 203 198 276
16* 23 30 29
172 181 182 277
15 21 21 39
8 8 8 12
2 1 3 3
266 221 207 244
18 19 14 22
15 Minutes After Infusion
4 1 4 10
4 5 5 9
1 1 3 3
207 186 206 312
23 23 27 85
45 Minutes After Infusion
16 11 15 19
4 2 4 2
8 7 8 11
2 1 3 2
93 63 32 83 180 162 183 255
21 23 27 26
NOTE: Values presented as mL/min; mean SE. *P .05 vs baseline and ConNaL at end infusion.
indicated, uptake of Lac by the liver may be regulated by an active transport process.11-15 In sepsis, either the Lac membrane transport system became impaired, or the internal environment of the hepatocyte prevented uptake. These results emphasize, however, that the liver was not able to augment Lac extraction in sepsis. Although the renal cortex is stated to be an important consumer of Lac during acidosis,22 renal uptake of Lac also was not apparent in any of the groups. Under nonseptic conditions, previous studies that used lactic acid loading in rats and sheep have estimated that the kidney extracts approximately 20% to 30% of the load and that the excess Lac enters both oxidative and biosynthetic pathways.2,22,36 Under conditions of hyperlactemia, renal excretion contributes to Lac removal, but accounts for less than 1% to 2% of the infused load, and most of the extraction occurs by cellular metabolism.22,23,36 Yet, despite this severe degree of acidosis produced in the present study, little uptake of Lac by the kidney was observed (see Fig 3). Conceivably, the discrepancy between what is found in the literature and the results reported here may be related to the magnitude of the acidosis produced. Yudkin and Cohen21 showed that Lac removal by the kidneys of rats was reduced when the
perfusate pH level was reduced below 7.1. The extremely low pH level produced during LA infusion in the present study may have resulted in an inhibition of renal uptake of Lac, although this would not explain why clearance was not observed in the NaL groups. However, most likely, the explanation for the low Lac clearance observed in the present study is that any contribution of clearance by the kidney would be diluted from other sources that enter the vena cava. Based on previous studies in this model,10 renal blood flow approximates 10% to 20% IVC flow. In the present study, Lac concentrations were measured in the IVC in which catheters were positioned above (IVCupper) and below (IVClower) the kidney. Any contribution by the kidney would be diluted by other sources. These results point out, therefore, that the kidney may not contribute in large part to Lac clearance in sepsis because its blood flow is only a small amount of the total flow. Lac use by the soft tissues of the lower extremity was not apparent in this model. Although normally a lactate-producing organ, resting skeletal muscle can also extract Lac from the blood and use it primarily as fuel for oxidation during hyperlactatemia produced by exogenous infusion.2,24,25 Nevertheless, the results showed no increase in Lac uptake after
LACTATE METABOLISM IN SEPSIS
199
Fig 5. MAP, mean pulmonary capillary wedge pressure, and cardiac output are shown for the 4 groups in which sodium Lac or lactic acid was infused. Statistics by 2-way ANOVA and SNK. ——, Septic group lactic acid; ——, control group lactic acid; ——, septic group sodium lactate; ——, control group sodium lactate. (A) *P .05 vs baseline; %P .05 vs ConNaL; !P .05 vs ConLA, ConNaL; (B) *P .05 vs baseline; ??P .05 vs SepLA; %P .05 vs. ConNaL; !P .05 vs ConLA, ConNaL; #P .05 vs all other groups. (C) *P .05 vs baseline; #P .05 vs all other groups; %P .05 vs ConNaL; !P .05 vs ConLA, ConNaL.
NaL or LA infusion in the present study to show that this mechanism may be important in sepsis. The splanchnic organs are considered to be an important source of Lac production in sepsis. This production may result as a consequence of a reduction in oxygen delivery either owing to a decrease in bulk splanchnic flow, or because microvascular injury leads to a redistribution of perfusion to some areas while other areas remain perfused.4,5 Alternatively, a defect in the use of oxygen at a mitochondrial level may be primarily involved in sepsis,6,7 a conclusion that would agree with the investigations of Curtis and Cain,6 Hurtado et al,7 and our previous study.10 In terms of this model, it is further important to note that the increase in splanchnic Lac production observed between baseline and 4 hours is relatively small (.06 mmol/min), and the results shown in Fig 2 are similar to those previously reported,10 but did not reach statistical significance in the present study.
In the current study, SLM was different among the 4 groups studied (see Fig 2). A low pH level appeared to have a negative effect on Lac uptake, such that no uptake of Lac was observed in either the sepsis or nonsepsis groups after LA infusion. On the other hand, when NaL was infused, a large increase in splanchnic Lac uptake was observed in ConNaL, but not in SepNaL. Over the first 15minute interval after infusion, approximately 3 mmol/min Lac was consumed in ConNaL. The explanation for the different effect of NaL on splanchnic Lac uptake between the septic and nonseptic groups is not clear, but an impairment in the uptake of Lac by the monocarboxylate transport system or a defect in the mitochondrial metabolism may be responsible for this lack of an increase.6,37 In all groups, when either LA or NaL was infused, lung Lac production increased at end infusion (see Fig 2). The amount of Lac produced by the lung between end infusion and 15 minutes after infusion was
200
quite large and could account for 20 mmol over this interval. In another study, Bellomo et al28 reported that during endotoxemia in dogs, the lung produced Lac. Similar increases in lung Lac production have been reported in humans with acute lung injury.29,30 Although the mechanism responsible for this increase has not been identified,38 the present results show that lung production may contribute to delayed clearance of Lac in sepsis. As part of the protocol, an estimate of cephalic Lac consumption was determined. In both the septic and nonseptic groups, the results showed that after LA infusion, but not after NaL infusion, there was a significant increase in cephalic Lac consumption (see Fig 3). Although other organs may have contributed, most of the Lac metabolism was probably via the brain, although it is recognized that total brain blood flow was not measured. Because neuronal cells metabolize glucose and release Lac during conditions of ischemia, it was surprising that Lac uptake was detected from the brain. However, active transport of Lac occurred in cultured astroglial cells, particularly when the pH medium level was low.26,27 Thus, after LA infusion, when pH level was low, cephalic extraction measured between end infusion and 15 minutes after infusion accounted for approximately 3 mmol of uptake in the septic and nonseptic groups. Another interesting finding is that when LA or NaL was infused, a significant step down in lactate was observed between Lacjug and LacSVC in all groups (see Fig 3). Although this difference implies that there was mixing from several sources, the particular source of extraction is not clear. The venous effluent from the unidentified source could come from any organ in the chest, head, or neck, as well as the upper extremities. One possibility is that the astroglial cells of the spinal cord that drained by means of the vertebral system accounted for this step down.26,27 Whereas the site importantly awaits further evaluation, this source accounted for a significant amount of Lac extraction after infusion of either LA and NaL in both septic and nonseptic groups. From the results obtained in this study, it is clear that blood pH level may modulate Lac metabolism in sepsis. Nevertheless, although multiple organ sites were examined, it was not possible to account for all of the Lac metabolized over the experimental interval. For instance, in ConLA, approximately
CHRUSCH ET AL
62 mmol of lactate was consumed between end infusion and 15 minutes after infusion. Yet, over this interval, of the organs examined, 3 mmol was consumed by the liver, 3 mmol was consumed by the brain, and there was a further step down in Lac concentration between the brain and the SVC (ie, JugSVC). Excluding the increase in production by the lung, these measurements can only account for 20% of the Lac extracted over this interval. In ConNaL, the splanchnic organs accounted for 3 mmol or approximately 6% of the consumption. However, besides the step down found between Jug-SVC, no other organ system could be identified as a consumer of Lac in ConNaL. In SepLA, over the 15minute interval, the amount of Lac consumed was approximately one third of the total dose given (ie, 1/3 125) 41 mmol Lac. Of the multiple organs examined, only the brain could be identified as a metabolizer of Lac and accounted for approximately 8% of the total metabolized. In SepNaL, 25%, or approximately 31 mmol of 125 mmol, was metabolized over the initial 15-minute interval. Yet, besides the step down observed between Jug and SVC, no specific organ system of Lac metabolism could be identified. In terms of our analysis, moreover, it is important to note that the approach of Naylor et al36 was used in which plasma Lac rather than whole-blood Lac concentrations was substituted into the calculations. During Lac loading, it takes 1 or 2 hours for plasma Lac to equilibrate within the red cell. Conceivably, some of the decline in arterial decay may be related to uptake of Lac by the red cell. However, it would be expected that this contribution would occur to a similar degree in all 4 groups. Importantly, the present study points out that in sepsis, our knowledge of the extent to which different organs metabolize Lac is not clear. The different changes found in arterial pH, PCO2, and HCO3 levels in the treatment groups are shown in Fig 1. After LA was infused in ConLA and SepLA, PCO2 levels rose dramatically in both groups and then subsequently declined over the remaining course of the experiment. Interestingly, the rate of decline in PCO2 levels in the septic group was greater than that found in the nonseptic group. The explanation for this finding is not clear because the ventilator rate was not changed in either group. The decrease in whole-animal oxygen consumption shown in Table 3 may indicate that lactic acid infusion caused a decrease in cellular oxygen metabolism that resulted
LACTATE METABOLISM IN SEPSIS
201
in a reduction in CO2 production and, hence, arterial PCO2, although the organ responsible for this reduction in metabolism is not clear. On the other hand, after NaL infusion, PCO2 levels gradually rose over the course of the experiment to a similar extent in both SepNaL and ConNaL, and these increases in PCO2 levels were related to the generation of HCO3 that occurred in the 2 groups. In terms of the extensive preparation used in this study, it is recognized that this factor may have impacted on the results observed. Moreover, in the design of the study, there could be an effect other than pH level alone that contributed to the differences observed in Lac metabolism when sodium Lac versus lactic acid was administered in this model. When sodium Lac was given, the sodium concen-
tration would obviously be higher than when lactic acid was given. It is possible that this effect contributed to the differences between NaL and LA metabolism observed in this model. In summary, both effects of pH level and sepsis modulate the organ uptake of Lac in septic shock. Only a small amount of the Lac infused could be accounted for by the organs monitored in this model. The liver and kidney were not important metabolizers of Lac in sepsis. The lung contributed to Lac production, whereas Lac extraction by the brain and the organs that drain into the upper thorax contributed to Lac uptake in sepsis. These findings suggest that additional organs may account for Lac metabolism in sepsis. The net result of these changes is that the clearance of Lac is reduced in septic shock.
REFERENCES 1. Mizock BA, Falk JL: Lactic acidosis in critical illness. Crit Care Med 20:80-93, 1992 2. Madias NE: Lactic acidosis. Kidney Int 29:752-774, 1986 3. Nelson DP, King CE, Dodd SL, et al: Systemic and intestinal limits of O2 extraction in the dog. J Appl Physiol 63:387394, 1987 4. Drazenovic R, Samsel RW, Wylam ME, et al: Regulation of perfused capillary density in canine intestinal mucosa during endotoxemia. J Appl Physiol 72:259-265, 1992 5. Gutierrez G: Derangements of oxygen transport in shock states and sepsis, in Rippe JM, Irwin RS, Fink MP, et al (eds): Intensive Care Medicine, 3rd ed. Boston, MA, Little Brown and Company, 1996, pp 1972-1979 6. Curtis SE, Cain SM: Regional and systemic oxygen delivery/uptake relations and lactate flux in hyperdynamic, endotoxin-treated dogs. Am Rev Respir Dis 45:348-354, 1992 7. Hurtado FJ, Gutierrez AM, Silva N, et al: Role of tissue hypoxia as the mechanism of lactic acidosis during E. coli endotoxemia. J Appl Physiol 72:1895-1901, 1992 8. Falk JL, Rackow EC, Leavy J, et al: Delayed lactate clearance in patients surviving circulatory shock. Acute Care 11:212215, 1985 9. Levraut J, Ciebiera J, Chave S, et al: Mild hyperlactemia in stable septic patients is due to impaired clearance rather than overproduction. Am J Respir Crit Care Med 157:1021-1026, 1998 10. Chrusch C, Bands C, Bose D, et al: Impaired hepatic extraction and increased splanchnic production contribute to lactic acidosis in canine sepsis. Am J Respir Crit Care Med 161:517-526, 2000 11. Lupo MA, Cefalu WT, Pardridge WM: Kinetics of lactate transport into rat liver in vivo. Metab Clin Exp 39:374-377, 1990 12. Poole RC, Halestrap AP: Transport of lactate and other monocarboxylates across mammalian plasma membranes. Am J Physiol 264:C761-C782, 1993 13. Jackson VN, Halestrap AP: The kinetics, substrate, and inhibitor specificity of the monocarboxylate (lactate) transporter of rat liver cells determined using the fluorescent intracellular pH indicator 2 ,7 Bis(carboxyethyl-5(6)-carboxyfluorescein. J Biol Chem 270:1843-1849, 1995
14. Edlund GL, Halestrap AP: The kinetics of transport of lactate and pyruvate into rat hepatocytes. Biochem J 249:117126, 1988 15. Fafournoux P, Demingné C, Rémésy C: Carrier-mediated uptake of lactate in rat hepatocytes. J Biol Chem 260:292-299, 1985 16. Sestoft L, Marshall MO: Hepatic lactate uptake is enhanced by low pH at low lactate concentrations in perfused rat liver. Clin Sci 70:19-22, 1986 17. Berry MN: The liver and lactic acidosis. Proc R Soc Med 60:1260-1262, 1967 18. Baron PG, Iles RA, Cohen RD: Effect of varying PCO2 on intracellular pH and lactate consumption in the isolated perfused rat liver. Clin Sci 55:175-181, 1978 19. Forsythe SM, Schmidt GA: Sodium bicarbonate for the treatment of lactic acidosis. Chest 117:260-267, 2000 20. Cooper DJ, Walley KR, Wiggs BR, et al: Bicarbonate does not improve hemodynamics in critically ill patients who have lactic acidosis: A prospective controlled clinical study. Ann Intern Med 112:492-498, 1990 21. Yudkin J, Cohen RD: The effect of acidosis on lactate removal by the perfused rat kidney. Clin Sci 50:185-194, 1983 22. Yudkin J, Cohen RD: The contribution of the kidney to the removal of a lactic acid load under normal and acidotic conditions in the conscious rat. Clin Sci 48:121-131, 1975 23. Connor H, Woods HF, Ledingham JGG, et al: A model of L () lactate metabolism in normal man. Ann Nutr Metab 26:254-263, 1982 24. Gladden LB, Yates JW: Lactic acid infusion in dogs: Effects of varying infusate pH. J Appl Physiol 54:1254-1260, 1983 25. Brooks GA, Gaesser GA: End points of lactate and glucose metabolism after exhausting exercise. J Appl Physiol 49:1057-1069, 1980 26. Hamprech TB, Dringer R: Energy metabolism, in Kettenmann H, Ransom BR (eds): Neuroglia. New York, Oxford University Press, 1995, pp 473-487 27. Waniewski RA, Martin DL: Preferential utilization of acetate by astrocytes is attributable to transport. J Neurosci 18: 5225-5233, 1998
202
28. Bellomo R, Kellum JA, Pinsky MR: Transvisceral lactate fluxes during early endotoxemia. Chest 110:198-204, 1986 29. Brown SD, Clark C, Gutierrez G: Pulmonary lactate release in patients with sepsis and ARDS. J Crit Care 11:2-8, 1996 30. Kellum JA, Kramer DJ, Lee K, et al: Release of lactate by the lung in acute lung injury. Chest 111:1301-1305, 1997 31. Gomez A, Wang R, Unruh H, et al: Hemofiltration reverses left ventricular dysfunction during sepsis in dogs. Anesthesiology 73:671-685, 1990 32. Flecknell P: Laboratory Animal Anaesthesia, 2nd ed. San Diego, CA, Academic Press Ltd. 1999, p 112 33. Kellum JA, Bellomo R, Kramer DJ, et al: Hepatic anion flux during acute endotoxemia. J Appl Physiol 78:2212-2217, 1995 34. Presberg KW, Sznajder JL, Melendres J, et al: Distribu-
CHRUSCH ET AL
tion of pulmonary vascular resistance during lactic acid infusion in dogs. J Appl Physiol 68:1328-1336, 1990 35. Kirk BW, Raber MB: A practical apparatus for rapid determination of blood oxygen content. J Appl Physiol 34:724725, 1973 36. Naylor JM, Kronfeld DS, Freeman DE, et al: Hepatic and extrahepatic lactate metabolism in sheep: Effects of lactate loading and pH. Am J Physiol 247:E747-E755, 1984 37. Faeilli A, Orsenigo MD, Verri A, et al: Cl/HCO3 antiport affects H-lactate symport activity at the basolateral pole of jejunum enterocyte. Cell Physiol Biochem 8:151-157, 1988 38. Kilpatrick-Smith L, Erecinska DJ: Cellular effects of endotoxin in-vitro: 1. Effect of endotoxin on mitochondrial substrate metabolism and intracellular calcium. Circ Shock 11:8599, 1983