Lack of change in the respiratory quotient during oxygen supply dependence in endotoxemic shock: a subanalysis of an experimental controlled study

Objective To evaluate if the reductions in systemic and renal oxygen consumption are associated with the development of evidence of anaerobic metabolism. Methods This is a subanalysis of a previously published study. In anesthetized and mechanically ventilated sheep, we measured the respiratory quotient by indirect calorimetry and its systemic, renal, and intestinal surrogates (the ratios of the venous-arterial carbon dioxide pressure and content difference to the arterial-venous oxygen content difference. The Endotoxemic Shock Group (n = 12) was measured at baseline, after 60 minutes of endotoxemic shock, and after 60 and 120 minutes of fluid and norepinephrine resuscitation, and the values were compared with those of a Control Group (n = 12) without interventions. Results Endotoxemic shock decreased systemic and renal oxygen consumption (6.3 [5.6 - 6.6] versus 7.4 [6.3 - 8.5] mL/minute/kg and 3.7 [3.3 - 4.5] versus 5.4 [4.6 - 9.4] mL/minute/100g; p < 0.05 for both). After 120 minutes of resuscitation, systemic oxygen consumption was normalized, but renal oxygen consumption remained decreased (6.3 [5.9 - 8.2] versus 7.1 [6.1 - 8.6] mL/minute/100g; p = not significance and 3.8 [1.9 - 4.8] versus 5.7 [4.5 - 7.1]; p < 0.05). The respiratory quotient and the systemic, renal and intestinal ratios of the venous-arterial carbon dioxide pressure and content difference to the arterial-venous oxygen content difference did not change throughout the experiments. Conclusion In this experimental model of septic shock, oxygen supply dependence was not associated with increases in the respiratory quotient or its surrogates. Putative explanations for these findings are the absence of anaerobic metabolism or the poor sensitivity of these variables in detecting this condition.


INTRODUCTION
Shock states are characterized by the failure of the cardiovascular system to meet metabolic oxygen demands.Regardless of the different hemodynamic patterns, the distinctive and common feature of shock is the presence of tissue hypoperfusion, which results in tissue hypoxia and anaerobic metabolism.Hence, the dependence of oxygen consumption (VO2) on oxygen delivery (DO2) is considered characteristic of all types of shock. (1)n patients with septic shock, VO2/DO2 dependence has been repeatedly described. (2)In experimental models, oxygen supply dependence has also been found at both the systemic and organ levels-such as in the gut and kidney. (3,4)owever, the meaning of this phenomenon is controversial.Although the fall in VO2 is usually considered an expression of anaerobic metabolism leading to organ dysfunction, other explanations are possible.In septic shock, while Arnaldo Dubin 1 , Juan Francisco Caminos Eguillor 1 , Gonzalo Ferrara 1 , María Guillermina Buscetti 1 , Héctor Saúl Canales 1 , Bernardo Lattanzio 1 , Luis Gatti 1 , Facundo Javier Gutierrez 1 , Vanina Siham Kanoore Edul 1 1.Cátedras de Terapia Intensiva y Farmacología Aplicada, Facultad de Ciencias Médicas, Universidad Nacional de La Plata -La Plata, Argentina.
the mitochondrial ability to generate cellular adenosinetriphosphate (ATP) is decreased, this is not associated with significant organ necrosis. (5)Therefore, the reduction in VO2 might be an adaptive response that allows survival in the face of an overwhelming insult.The suppression of nonessential functions-such as glomerular filtration rate and consequent tubular energy demand-might thus be a mechanism to avoid death by dysoxia.From this standpoint, organ failure could be a reactive and potentially protective mechanism. (6)n approach that might help to elucidate the meaning of the VO2/DO2 dependence is the analysis of the respiratory quotient (RQ).The RQ is the ratio between carbon dioxide (CO2) production (VCO2) and VO2.In animal models of tissue hypoxia, the beginning of anaerobic metabolism is signaled by the abrupt increase in the RQ. (7,8)Although VCO2 and VO2 decrease secondary to the compromise of aerobic metabolism, there is anaerobic VCO2 due to the buffering of protons derived from anaerobically generated acids by bicarbonate.Therefore, the relative increase in VCO2 in relation to VO2 increases the RQ.
In an experimental model of endotoxemic shock and severe kidney injury, we previously found the presence of systemic and renal oxygen supply dependence. (4)The decreases in renal VO2 were still present after resuscitation.Notwithstanding this, the renal oxygen ratio extraction (O2ER) remained stable and eventually decreased, suggesting a primary reduction in metabolic oxygen needs.
The goal of this subanalysis was to evaluate whether the reductions in systemic and renal VO2 are associated with the development of evidence of anaerobic metabolism.For this purpose, we examined the changes in RQ and its systemic and regional surrogates, the ratios of venousarterial CO2 pressure and the content difference to the arterial-venous O2 content difference (Pv-aCO2/Ca-vO2 and Cv-aCO2/Ca-vO2, respectively). (9)Our hypothesis was that VO2/DO2 dependence is not associated with anaerobic metabolism, as reflected by the RQ and its surrogates.

METHODS
We used original data from a previously published study. (4)The local research committee approved this study [protocol P01-05-2016].Care of animals was in accordance with the National Research Council's Guide for the Care and Use of Laboratory Animals.

Anesthesia and ventilation
Twenty-four sheep (24 [22 -27] kg, median [25 th -75 th percentiles]) were anesthetized with 30mg.kg - sodium pentobarbital, intubated and mechanically ventilated with a Servo Ventilator 900C (Siemens -Elema AB, Solna, Sweden) with a tidal volume of 10mL/kg, an FiO2 of 0.21 and a positive end-expiratory pressure of 6cmH2O.The initial respiratory rate was set to keep the arterial PCO2 between 35 -40mmHg.This respiratory setting was maintained during the rest of the experiment.Neuromuscular blockade was performed with pancuronium bromide (0.06mg.kg -1 ).Additional pentobarbital boluses (1mg/kg) were administered hourly and when clinical signs of inadequate depth of anesthesia were evident.Analgesia was provided by fentanyl as a bolus of 2µg/kg, followed by 1µg/kg/h.These drugs were administered intravenously.

Surgical preparation
A 7.5 French Swan -Ganz Standard Thermodilution Pulmonary Artery Catheter (Edwards Life Sciences, Irvine, CA, USA) was inserted in the right external jugular vein to obtain mixed venous samples.Catheters were placed in the descending aorta via the left femoral artery to measure blood pressure and obtain blood samples and in the inferior vena cava to administer fluids and drugs.
A midline laparotomy was performed, followed by a gastrostomy to drain the gastric contents and a splenectomy to avoid spleen contraction during shock.Perivascular ultrasonic flow probes were placed around the superior mesenteric artery and the left renal artery to measure intestinal blood flow (IBF) and renal blood flow (RBF).Catheters were introduced in the left renal and mesenteric veins to draw blood samples and to measure venous pressure.Catheters were also positioned in the abdomen for intraabdominal pressure measurement and into the bladder to monitor urinary output.To allow renal cortical videomicroscopy, the left kidney was gently decapsulated, and 5cm incision was made in the left flank of the abdominal wall.A 10-to 15cm segment of the ileum was mobilized, placed outside the abdomen, and opened 2cm on the antimesenteric border to allow examination of mucosal microcirculation.The exteriorized intestinal segment was covered and moistened, and the temperature was preserved by an external heating device.Finally, after complete hemostasis, the midline abdominal wall incision was closed, except for a short segment for externalization of the ileal loop.

Measurements and derived calculations
Systemic VO2 and the RQ were measured by analysis of expired gases (MedGraphics CPX Ultima, Medical Graphics Corporation, St. Paul, MN) and adjusted to body weight.
Arterial, mixed venous, renal venous, and mesenteric venous PO2, PCO2, pH, Hb, and O2 saturation were measured with a blood gas analyzer and a co-oximeter in sheep mode (ABL 5 and OSM 3, Radiometer, Copenhagen, Denmark).Oxygen-derived variables were calculated by standard formulae.
Since the thermodilution method overestimates low cardiac output, the cardiac index (CI) was calculated as VO2 divided by the arterial-mixed venous O2 content difference (Ca-mvO2).Oxygen delivery was calculated as the CI multiplied by the arterial O2 content (CaO2).Systemic O2ER was calculated as Ca-mvO2 divided by CaO2.
Intestinal blood flow and RBF were measured by an ultrasonic flowmeter (One Channel Perivascular Flowmeter, Transonics Systems Inc., Ithaca, NY, USA) and normalized to the organ weight.
Intestinal and renal DO2 and VO2 were calculated as the product of the respective flow index multiplied by either the CaO2 or arterial-venous oxygen content difference.Intestinal and renal O2ER were calculated as the respective arteriovenous oxygen content difference divided by CaO2 (Ca-ivO2 and Ca-rvO2, respectively).
As surrogates of the systemic, renal, and intestinal RQ, we calculated the systemic, renal, and intestinal Pv-aCO2/Ca-vO2.In addition, the corresponding Cv-aCO2/Ca-vO2 values were calculated by means of the Douglas algorithm (10) to assess the changes in the CO2 dissociation curve.
Arterial lactate was measured with a point-of-care analyzer (Stat Profile Critical Care Xpress, Nova Biomedical, Waltham, MA, USA).
Creatinine clearance was calculated as the urinary creatinine level multiplied by the urine output in 60 minutes divided by the plasma creatinine level.

Experimental procedure
Basal measurements were taken after a period of no less than 30 minutes after blood pressure, heart rate, systemic VO2, and renal and intestinal flow became stable.Animals were then randomly assigned to the endotoxemic shock (n = 12) or control (n = 12) groups.In the endotoxemic shock group, shock was induced by intravenous injection of Escherichia coli lipopolysaccharide (5µg/kg followed by 2.5µg/kg/hour for 180 minutes).After 60 minutes of shock, 30mL/kg of 0.9% sodium chloride (NaCl) solution was infused, and norepinephrine was titrated to reach a mean arterial pressure (MAP) of 70mmHg.In the sham group, the same experimental preparation was carried out, and 0.9% NaCl was infused to maintain hemodynamic variables at basal values without further interventions.Measurements were performed at baseline (0 minutes), after 60 minutes of endotoxemic shock without resuscitation, and after 60 and 120 minutes of resuscitation.Blood temperature was kept constant throughout the study with a heating lamp.
At the end of the experiment, animals were killed with an additional dose of pentobarbital and a potassium chloride (KCl) bolus.A catheter was inserted in the superior mesenteric artery, and Indian ink was instilled through the catheter.Dyed intestinal segments were dissected, washed, and weighed.We also weighed the left kidney.Consequently, renal and intestinal VO2 and DO2 are expressed as indices based on organ weight.

Data analysis
Because of the small numbers of animals, nonparametric tests were used.Data expressed as medians [25 th -75 th percentiles] were analyzed using generalized estimating equations (GEE), followed by Mann-Whitney and Wilcoxon tests with Bonferroni correction for between-and withingroup pairwise comparisons.The association of the RQ with the systemic Pv-aCO2/Ca-vO2 and Cv-aCO2/Ca-vO2 was assessed by means of the Spearman correlation.Agreement between the RQ and Cv-aCO2/Ca-vO2 was evaluated by the Bland and Altman method.A p value < 0.05 was considered statistically significant.

RESULTS
The effect of endotoxemic shock and subsequent resuscitation on systemic, regional, and microvascular hemodynamics and oxygen transport has been reported elsewhere. (4)Briefly, endotoxin administration decreased blood pressure, CI, RBF, and IBF (Table 1).Systemic VO2 and DO2 fell, and O2ER increased.Renal VO2 and DO2 decreased, but renal O2ER did not change.At the intestinal level, DO2 was reduced but due to the increase in O2ER, VO2 remained stable (Table 1 and Figures 1S,  2S, and 3S -Supplementary Material).Microcirculatory alterations arose in the sublingual mucosa, intestinal villi and, especially, renal peritubular capillaries. (4)Oliguria and severe acute kidney injury were also present (Table 1).
Resuscitation normalized the CI and systemic VO2 and DO2.Renal blood flow and renal DO2 and VO2 remained low, whereas renal O2ER never increased and dropped at 60 min of resuscitation.Intestinal DO2 improved, but O2ER remained high (Table 1 and Figures 1S, 2S, and 3S -Supplementary Material).Most of the renal microvascular abnormalities appearing during shock were still present in the resuscitation period.In the intestinal and sublingual mucosa, only minor alterations persisted. (4)n the endotoxemic group, hyperlactatemia and increased anion gap metabolic acidosis developed during resuscitation (Table 2).
In both groups, RQ did not change throughout the experiments.The systemic, renal, and intestinal Pv-aCO2/Ca-vO2 and Cv-aCO2/Ca-vO2 also remained unchanged (Figures 1 and 2).

DISCUSSION
The main finding of this study was that the RQ and its systemic and regional surrogates did not change, despite the severe hemodynamic compromise with oxygen supply dependence, tissue hypoperfusion, and acute kidney injury produced by the administration of endotoxin.The lack of increase in the RQ and its surrogates might suggest the absence of anaerobic metabolism but also the inability of these variables to reflect tissue hypoperfusion.
Our experimental model is relevant and resembles many components of human septic shock, including derangements in systemic and microvascular hemodynamics.In addition, it produced severe renal failure, which was unresponsive to resuscitation.Remarkably, the renal O2ER never increased and eventually dropped at 60 minutes after the start of the resuscitation period.
The concurrent alterations in RBF and peritubular microcirculation could be considered the cause of renal failure or a reflex compensation for metabolic shutdown.
Since previous studies could not detect overt necrotic lesions, (11) septic acute kidney injury has been related to bioenergetic failure.This hypothesis states that mitochondrial dysfunction and insufficient adenosine triphosphate lead to reduced cellular metabolism.Organ failures might thus be primarily functional rather than structural.Indeed, this could act as a potentially protective, reactive mechanism against inflammatory stress. (6)In an experimental study, proximal tubular cells exposed to endotoxin developed an irreversible reduction in VO2 as a sign of pathologic metabolic downregulation. (12)Even though this process is usually described at several hours or days after septic challenge, the intravenous administration of endotoxin is associated with almost immediate reductions in the intestinal redox state of mitochondrial cytochrome aa3. (13)Likewise, within 1 hour of endotoxin exposure, renal cells show decreased expression of genes involved in mitochondrial processes. (14)The lack of changes in the RQ and systemic and regional Pv-aCO2/Ca-vO2 and Cv-aCO2/Ca-vO2 might be linked to bioenergetic failure, but this is merely speculative because mitochondrial function was not assessed in our study.
Normal RQ values range from 0.67 to 1.10, which depends on the type of substrate utilized. (15)For this reason, a sharp increase-rather than isolated high values-of the RQ signals the beginning of anaerobic metabolism during progressive exercise load and during reductions in oxygen transport in critically ill patients. (16)Acute increases in the RQ have been described in ischemic, hypoxic, and anemic hypoxia. (7,8,17,18)nother explanation for the lack of changes in the RQ during endotoxemic shock might be a switch in the source of energy, i.e., from carbohydrates to lipids.In this case, however, the RQ should be lower during resuscitation.
In septic shock, there are conflicting results about the behavior of the RQ and its surrogates.In our study, these variables remained constant.In a similar model of endotoxemic shock with systemic and intestinal oxygen supply dependence, we found that the corresponding Cv-aCO2/Ca-vO2 did not increase. (19)Apart from ours, only two studies, which were carried out in rodent models of sepsis, have assessed the RQ calculated from the measurement of expired gases during VO2/DO2 dependence. (20,21)In contrast to our results, endotoxin injection resulted in an increase in the RQ.This discrepancy could be related to the species studied (rats, guinea pigs, and sheep).Severely hypodynamic murine models of sepsis have been considered poorly representative of human sepsis (5) Another explanation might reside in the fact that our animals were on mechanical ventilation, whereas the rodents breathed spontaneously.Spontaneous breathing is a major contributor to the development of muscle anaerobic metabolism and lactic acidosis in shock states, regardless of hemodynamic changes. (22,23)n another study performed in septic pigs, the systemic and regional Pv-aCO2/Ca-vO2 and Cv-aCO2/Ca-vO2 did not change, but VO2/DO2 dependence was absent. (24)In patients with septic shock, the RQ had a similar time course in survivors and nonsurvivors. (25,26)Although mortality was associated with temporal decreases in VCO2 and VO2, the RQ was stable over time.29) Pv-aCO2/Ca-vO2 and Cv-aCO2/Ca-vO2 have been used as surrogates for the RQ.Experimental studies have shown that both variables increase during states of ischemic, hypoxic, and anemic hypoxia. (30,31)In patients with septic shock, a Pv-aCO2/Ca-vO2 higher than 1.4 was a predictor of mortality, hyperlactatemia, and oxygen supply dependence. (9)(32) Factors that enhance the dissociation of CO2 from hemoglobin, such as anemia, metabolic acidosis, and the Haldane effect, can account for the increase in Pv-aCO2/Ca-vO2.In our study, Pv-aCO2/Ca-vO2 only showed a weak correlation with the RQ.The calculation of Cv-aCO2/Ca-vO2 should overcome these difficulties, but this approach could also be misleading.In the validation of the Douglas algorithm, an excellent correlation between the tonometric and calculated CO2 content was found. (10)However, using data from the abovementioned study, the 95% limits of agreement between the measured and calculated CO2 contents are as large as 4.7mL/100mL.In addition, there is a propagation error linked to the calculation of Cv-aCO2.For these reasons, this calculation can occasionally result in spurious negative values of Cv-aCO2.Accordingly, we found no correlation and wide 95% limits of agreement between the RQ and Cv-aCO2/Ca-vO2.Previous experimental and clinical studies showed that Pv-aCO2/Ca-vO2 is a misleading surrogate for the RQ. (28,31,32)It might exhibit high sensitivity but low specificity to detect increases in the RQ.Despite the high sensitivity, this variable remained markedly stable in our experiments, even in the presence of systemic and renal oxygen supply dependence.
The increase in the RQ in anaerobic states results from the anaerobic VCO2 produced by bicarbonate buffering of anaerobically generated acids, such as lactate. (16)In our experiments, lactate only showed marginal increases in the initial phase of shock, which is congruent with the lack of changes in the RQ and its surrogates and might imply the preservation of aerobic metabolism.In contrast, after the restoration of systemic VO2 and DO2 by fluid and norepinephrine resuscitation, severe hyperlactatemia and metabolic acidosis secondary to an increased anion gap arose.A large body of evidence shows that hyperlactatemia in septic shock, especially after the normalization of blood pressure and cardiac output, depends mainly on the increased aerobic glycolysis secondary to catecholamine stimulation of Na + /K + -ATPase activity. (33)Furthermore, energetic reprogramming from fatty acid oxidation and oxidative phosphorylation toward aerobic glycolysis might be contributing factors. (34)his study has weaknesses.Secondary analyses pose inherent limitations that have been subject to critiques. (35)dditionally, endotoxemic shock might not completely resemble human septic shock.In addition, our research lacks measurements of tissue oxygenation, bioenergetics, and mitochondrial function.Therefore, we could not completely rule out the occurrence of anaerobic metabolism.Another drawback is the lack of histologic examinations.

Figure 1 -
Figure 1 -Respiratory quotient (panel A), ratio of the mixed venous-arterial carbon dioxide pressure difference to arterial-mixed venous oxygen content difference (panel B), ratio of the renal venous-arterial carbon dioxide pressure difference to arterial-renal venous oxygen content difference (panel C), and ratio of the intestinal venousarterial carbon dioxide pressure difference to arterial-intestinal venous oxygen content difference (panel D) in the control and endotoxemic shock groups.CO2 -carbon dioxide; O2 -oxygen.

Figure 2 -
Figure 2 -Ratio of the mixed venous-arterial carbon dioxide content difference to arterial-mixed venous oxygen content difference (panel A), ratio of the renal venousarterial carbon dioxide content difference to arterial-renal venous oxygen content difference (panel B), and ratio of the intestinal venous-arterial carbon dioxide content difference to arterial-intestinal venous oxygen content difference (panel C) in the control and endotoxemic shock groups.

Table 1 -
Values of systemic, intestinal, and renal hemodynamic and oxygen transport variables in the control and endotoxemic shock groups O2 -oxygen.Data are shown as the median [25-75 interquartile range].* p < 0.05 versus basal.† p < 0.05 versus control.