For many years, hyperlactatemia in critically ill patients, and particularly those in shock, has been interpreted as a marker of oxygen debt, global tissue hypoxia, and thereafter, as a gauge of anaerobic metabolism (1). This view has recently been challenged with the demonstration that lactate production during shock states is, at least in part, linked to an accelerated aerobic glycolysis through β2 stimulation (2, 3). Indeed, several human and animal studies have demonstrated that epinephrine, via β2 stimulation, increases cyclic adenosine monophosphate production, thereby inducing stimulation of glycogenolysis and glycolysis with a concomitant production of adenosine triphosphate (ATP) and activation of the Na+K+-ATPase pump (4, 5). This activation consumes ATP, thereby leading to the generation of adenosine diphosphate (ADP). Generated ADP, via phosphofructokinase stimulation, reactivates glycolysis and hence generates more pyruvate and, consequently, more lactate. Muscle tissue, which represents approximately 40% of total body cell mass, is particularly involved in this mechanism (6). In a recent report, we demonstrated that in human septic shock, muscle is a net producer of lactate, and that this production can be totally inhibited by ouabain, thus confirming a Na+K+-ATPase-dependent mechanism yet remaining clearly independent of tissue hypoxia (5). We also recently demonstrated in low-flow (hemorrhagic) and normal- to high-flow models of shock (endotoxin and peritonitis) that lactate production is related, at least partly, to an increase in Na+K+-ATPase activity upon β2 stimulation (7).
We recently suggested that testing the ability of the body to produce a physiological response under catecholamine stimulation may be useful in identifying potential survivors from nonsurvivors (8). We hypothesized that the intensity of the metabolic response to epinephrine, which actually translates into hyperglycemia and arterial lactate elevation, may be of good prognosis as to whether aerobic glycolysis can be further activated, thereby indicating the existence of a metabolic reserve.
We thus conducted a retrospective analysis of 100 consecutive patients admitted in our intensive care unit (ICU) in shock state and receiving epinephrine infusion to decipher whether early (within 4 h) arterial lactate elevation could be an indicator of a good prognosis.
MATERIALS AND METHODS
Setting and study design
Collected registries of 100 consecutive patients admitted in a 16-bed medical ICU of a French university hospital between 2004 and 2008 were retrospectively analyzed. The retrospective and noninterventional nature of this study waived the need for an ethics committee's approval.
Patients were included if they met the following criteria: (a) shock state, irrespective of etiology, defined by a MAP less than 65 mmHg with no improvement after fluid therapy and associated with tissue hypoperfusion (persistent oliguria, peripheral vasoconstriction, etc); (b) no epinephrine infusion at the time of enrollment; and (c) arterial lactate measurement performed immediately before epinephrine administration. Patients were not enrolled if (a) they received epinephrine administration before ICU admission, (b) were admitted after a cardiac arrest, (c) were younger than 18 years, and (d) were moribund or when an initial decision for limitation of care was made.
The treatment of shock was conducted according to international recommendations. Of noted importance, the administration of epinephrine as a first-line vasopressor agent was not, and still is not, the usual practice in our ICU, especially in septic shock patients. Indeed, this drug was preferred over norepinephrine (a) in patients presenting a low cardiac output state despite adequate volume resuscitation or (b) in patients at high risk of cardiac failure in whom hemodynamic evaluation had not already been performed (history of cardiac insufficiency, myocardial infarct, etc). None of the patients received insulin within the first 4 h.
Adequate volume expansion was assessed by pulse pressure variation when feasible or when at a given level additional fluid challenge was no longer associated with an increase in cardiac index. Cardiac index was measured using a transesophageal Doppler, a pulmonary artery catheter, or echocardiography. Insulin was used to treat hyperglycemia according to Sepsis Campaign recommendations.
Clinical and laboratory characteristics were studied at admission and during ICU stay: severity score (Sequential Organ Failure Assessment [SOFA] score), McCabe score (used to assess the heaviness of comorbidities) (9), age, sex, reason for admission in the ICU, blood pressure, as well as requirement in vasopressor, respiratory, extrarenal, and steroid (hydrocortisone, 300 mg/day) support. Recorded biological data included arterial pH, lactate (automatic analyzer ABL 700; Radiometer, Copenhagen), potassium, glucose, Pao2, as well as markers of renal, hepatic, and hematological dysfunction. Specifically, arterial lactate was obtained at the time of epinephrine infusion (herein defined as H0) and thereafter at H4 (i.e., 4 h after the start of epinephrine infusion), 10, 24, 32, and 48 h, according to our routine sampling protocol. We defined Δlactate as (100 × [arterial lactateH4 − arterial lactateH0]/arterial lactateH0) and expressed as a percentage. There were no missing data.
Normally distributed (Kolmogorov-Smirnov test) quantitative variables were analyzed by the Student t test, and results were expressed as mean ± SD. Non-normally distributed variables were studied with the nonparametric Mann Whitney U test, and results were expressed as median (interquartile range). The evolution of arterial lactate concentration was analyzed using ANOVA. A receiver operating characteristic (ROC) curve was constructed to determine the sensitivity and specificity of Δlactate in predicting mortality. Upon establishment of the optimal Δlactate cutoff value, a survival analysis was performed using the Kaplan-Meier survival curve and log-rank test. Finally, a stepwise forward and backward multiple regression analysis was performed, including all data associated with a P < 0.1 in univariate analysis. A Cox model was then constructed to validate this latter analysis. A two-tailed P < 0.05 was deemed to be significant. All analyses were performed using GraphPad software, version 5 (La Jolla, Calif).
Characteristics of the studied population
One hundred consecutive patients were included in this study for which there were no missing values. The baseline characteristics of the studied population are depicted in Table 1. Most patients presented with associated comorbidities, mostly cardiac insufficiency of ischemic origin. Septic shock was the most frequent reason for admission (82%), and 58% of patients had already received antibiotics before ICU referral (of course, all septic shock patients received antibiotics once in the ICU). All patients required mechanical ventilation, and approximately 20% received norepinephrine and/or dobutamine in association with epinephrine. Thus, 28-day mortality rate was very high at 72% (deaths occurring after the 28th day were not recorded). Among septic shock patients, lung and abdomen were the principal sources of infection (Table 2).
Evolution of arterial lactate concentration
At H0, arterial lactate concentration was elevated (4.96 ± 3.8 mmol/L) and was further increased upon epinephrine administration, reaching a peak at H4, and followed by a progressive decline thereafter (Fig. 1). When patients were stratified according to their outcome, nonsurvivors displayed the same pattern as survivors, although with a significant upward shift in values (ANOVA, P = 0.0003), except at 4 h after epinephrine institution, where lactate levels did not differ between survivors and nonsurvivors.
Survivors versus nonsurvivors
We next sought to elucidate the clinicobiological differences between survivors and nonsurvivors. Univariate analysis revealed that age, SOFA score, epinephrine dose, arterial lactate at H0, Δlactate, arterial pH, and H4 arterial glucose and potassium differed significantly between groups (Table 3). Among these parameters, SOFA score, H4 arterial glucose, and Δlactate were those most significantly different. Indeed, Δlactate was 137 (range, 85-287) in survivors and 45 (range, 12-115) in nonsurvivors (Fig. 2). Moreover, all survivors showed a positive Δlactate, whereas eight nonsurvivors (11.1%) showed a decrease in lactate concentration between H0 and H4. At 24 h, lactate concentration was higher in nonsurvivors (4.55 [2.27-7.90] mmol/L) than in survivors (2.3 [1.25-3.70] mmol/L) (P = 0.0011).
ΔLactate as a prognosis factor
The significance of these parameters was further explored by performing a multiple logistic regression analysis (Table 4). Only two factors remained significantly associated with the risk of death: SOFA score at admission, with an odds ratio (OR) at 1.32 (95% confidence interval [CI], 1.06-1.65, P = 0.01) and Δlactate, with an OR at 0.99 (95% CI, 0.99-0.99, P = 0.03). In other words, for each 1-point increment of SOFA, the risk of death increased by 32%; and for each 1% increase in Δlactate, the risk of death decreased by 1%. This analysis was further validated by a Cox model, with 82% of patients correctly classified (survival/death). Of note, no differences were observed between patients with Δlactate greater than 100 and those with Δlactate less than 100 in terms of MAP increase between H0 and H4 (17% vs. 17% increase, respectively; P = 0.95) or in terms of epinephrine requirement between H4 and H0 (241% vs. 228% increase, respectively; P = 0.94). Arterial lactate, pH, and epinephrine dose considered after H4 were not independently associated with outcome (not shown).
A ROC curve was constructed to determine the optimal cutoff value of Δlactate in predicting outcome (Fig. 3). At a value of 100%, Δlactate predicted death with a sensitivity of 71% (95% CI, 51%-87%) and a specificity of 67% (95% CI, 43%-85%). The positive likelihood ratio was thus established at 2.14. Hence, when arterial lactate concentration doubled between H4 and H0, the risk of death was reduced by a factor of 2.
Finally, Kaplan-Meier survival analysis further confirmed this finding, with a 52.4% death rate among patients with Δlactate greater than 100 as compared with 84.3% in patients presenting a Δlactate less than 100 (log-rank test, P = 0.0002; Fig. 4).
Indirect evidence for epinephrine-induced Na+ K+-ATPase activation and increased aerobic glycolysis
Two indirect findings favor the hypothesis that an increase in Δlactate upon epinephrine infusion in survivors was indeed a reflection of Na+K+-ATPase pump activation. First, potassium concentration decreased between H0 and H4 in survivors while remaining unchanged in nonsurvivors (P = 0.001). Second, arterial glucose increased in survivors as compared with nonsurvivors between H0 and H4 (P = 0.0001; see Table 3).
The key result of the present study is that the ability to increase aerobic glycolysis and, therefore, to produce lactate upon epinephrine stimulation during shock state is associated with a better prognosis, evoking a preserved physiological response to catecholaminergic stress. Results also confirm the already demonstrated link between epinephrine, hyperlactatemia, and increased Na+K+-ATPase activity.
Monitoring lactate concentration during the treatment of shock is highly recommended because the persistence of high lactate levels is often associated with a bad prognosis (10-12). Indeed, ancillary reports derived from the study by Rivers et al. (13) clearly demonstrated that the decrease in lactate concentration during the first 6 h of septic shock treatment was associated with better prognosis. It is important to note, however, that epinephrine was not used in the Rivers study.
Herein, we found that after 4 h of epinephrine administration, the higher the epinephrine-associated increase in lactate, the better the prognosis. Such results have also been observed in stable normolactatemic septic shock patients by Levraut et al. (14). In their study, the authors used an infusion of exogenous lactate to estimate lactate production and clearance and found that survivors produced more lactate than nonsurvivors. Similarly, survivors were also better able to clear this lactate "overproduction" (14).
This apparent discrepancy between these latter results and previous published studies such as the Rivers study may be explained by the unique properties of epinephrine. Epinephrine, which is one of the main stress mediators, acts on α, β1, as well as β2 adrenoreceptors. The latter is considered to be responsible for the metabolic actions of epinephrine including hyperglycemia, hyperlactatemia, and hypokalemia (15). Because we previously demonstrated that epinephrine-induced lactate production is a transient phenomenon peaking at about 4 to 6 h and persisting 12 to 18 h (16), we therefore chose to measure the increase in lactate concentration after 4 h of epinephrine infusion and to use this increase as a test to estimate the ability of cells to adequately respond to a pharmacological trigger such as epinephrine. Finally, and as previously demonstrated, we found that after 24 h of treatment, lactate levels were higher in nonsurvivors than in survivors. Thus, measuring Δlactate after 4 h likely tests the so-called physiological reserve.
The use of catecholamines to test cellular response is not new. For example, the prognostic value of a positive response to dobutamine challenge estimated by the DO2/VO2 relationship (17) or the increase in cardiac index (18) has already been demonstrated. Moreover, we previously demonstrated that dopamine-sensitive patients had a better prognosis (78% survival) as compared with dopamine-resistant patients (16% survival) (19). Hence, the response to catecholamines can be studied at various levels, including the heart using β1 stimulation, at the vessel level using α stimulation and at the metabolic level using β2 stimulation. In the present study, we also observed a large difference in the amount of epinephrine used when comparing survivors with that of nonsurvivors. In survivors, epinephrine was increased by 80%, whereas in nonsurvivors, the dose of epinephrine was increased by 190% from baseline to H4. Hence, there seems to be a major difference in α-receptor sensitivity between survivors and nonsurvivors. However, it remains unknown whether the altered response to catecholamine observed in nonsurvivors is caused by a signaling problem at the receptor or postreceptor level, or by the tissue's inability to mount a physiological response, or both. An alternative view to explain the relatively smaller rise in lactate in nonsurvivors than in survivors may be that the metabolic reserves (i.e., glucose, glycogen) in moribund patients may be so depleted that they are not able to produce more lactate near the very end. This latter explanation appears unlikely, however, because moribund patients were excluded from this study, and patients were studied in the first 4 h after their admission in the ICU. Nevertheless, the ability to increase energy expenditure during shock is part of the general response to stress, with the delivery of substrates to vital organs to allow fight-or-flight responses (20). From a general perspective, these metabolic changes may be viewed as an adaptive attempt to survive during adverse situations. Whether these results disagree with the concept of hibernation strategy proposed by Singer et al. (21) is not clear. We may argue that patients studied herein were at an early phase of shock, during the nadir of sympathoadrenal response and cell activation. Hibernation, which is largely linked to mitochondrial cytopathy, is likely a later phenomenon (22).
Some limitations of this study should be pointed out. At the outset, it is clear that epinephrine is not the recommended first-choice vasopressor during shock state. Clearly, the administration of norepinephrine to treat vascular failure (in combination with dobutamine to treat associated cardiac failure) is preferentially recommended and is of daily use in our ICU (23). Nevertheless, in a large population of shock subjects, neither Myburgh et al. (24) nor Annane et al. (25) were able to find differences between epinephrine and norepinephrine in terms of mortality. Second, it is important to note that our selected population had a very high mortality rate in relation with a high number of comorbidities and a high degree of organ failure at admission. Most patients died early (within 4 days) from refractory shock and not from delayed multiple organ failure. Whether these characteristics may have influenced the present results and their generalizability to a less severe population is not known. The fact that nonsurvivors presented with higher baseline lactate concentrations than survivors could explain why they failed to further rise upon epinephrine and one could argue that this study is just another way of showing that a high lactate concentration is associated with a poor outcome. Nevertheless, multivariate analysis showed that baseline lactate concentration was not an independent predictor of death in this population. Finally, other factors that may have influenced the outcome (protective mechanical ventilation, activated protein C infusion, etc) were not specifically sought for.
Obviously, we are not trying to demonstrate that epinephrine should be administered as a prognostic test in shock patients; we just tried to address a physiological hypothesis.
In conclusion, an early and adapted response (lactate production) to a physiological trigger (epinephrine) is a protective factor during shock of various etiologies. Nevertheless, the persistence of hyperlactatemia in shock patients remains of poor prognosis.
1. Cohen RD, Woods HF: Lactic acidosis revisited. Diabetes
2. James JH, Fang CH, Schrantz SJ, Hasselgren PO, Paul RJ, Fischer JE: Linkage of aerobic glycolysis to sodium-potassium transport in rat skeletal muscle. Implications for increased muscle lactate
production in sepsis. J Clin Invest
3. James JH, Luchette FA, McCarter FD, Fischer JE: Lactate
is an unreliable indicator of tissue hypoxia in injury or sepsis. Lancet
4. James JH, Wagner KR, King JK, Leffler RE, Upputuri RK, Balasubramaniam A, Friend LA, Shelly DA, Paul RJ, Fischer JE: Stimulation of both aerobic glycolysis and Na(+)-K(+)-ATPase activity in skeletal muscle by epinephrine
or amylin. Am J Physiol
5. Levy B, Gibot S, Franck P, Cravoisy A, Bollaert PE: Relation between muscle Na+K+ ATPase activity and raised lactate
concentrations in septic shock
: a prospective study. Lancet
6. Levy B: Lactate
state: the metabolic view. Curr Opin Crit Care
7. Levy B, Desebbe O, Montemont C, Gibot S: Increased aerobic glycolysis through beta2 stimulation is a common mechanism involved in lactate
formation during shock
8. Collin S, Sennoun N, Levy B: Cardiovascular and metabolic responses to catecholamine and sepsis prognosis: a ubiquitous phenomenon? Crit Care
9. McCabe WR, Jackson GG: Gram negative bacteremia: I. Etiology and ecology. Arch Intern Med
10. Levy B, Sadoune LO, Gelot AM, Bollaert PE, Nabet P, Larcan A: Evolution of lactate
/pyruvate and arterial ketone body ratios in the early course of catecholamine-treated septic shock
. Crit Care Med
11. Vincent JL: End-points of resuscitation: arterial blood pressure, oxygen delivery, blood lactate
, or...? Intensive Care Med
12. Kompanje EJ, Jansen TC, van der Hoven B, Bakker J: The first demonstration of lactic acid in human blood in shock
by Johann Joseph Scherer (1814 - 1869) in January 1843. Intensive Care Med
13. Rivers E, Nguyen B, Havstad S, Ressler J, Muzzin A, Knoblich B, Peterson E, Tomlanovich M, Early Goal-Directed Therapy Collaborative Group: Early goal-directed therapy in the treatment of severe sepsis and septic shock
. N Engl J Med
14. Levraut J, Ichai C, Petit I, Ciebiera JP, Perus O, Grimaud D: Low exogenous lactate
clearance as an early predictor of mortality in normolactatemic critically ill septic patients. Crit Care Med
15. Liggett SB, Shah SD, Cryer PE: Characterization of beta-adrenergic receptors of human skeletal muscle obtained by needle biopsy. Am J Physiol
16. Levy B, Bollaert PE, Charpentier C, Nace L, Audibert G, Bauer P, Nabet P, Larcan A: Comparison of norepinephrine and dobutamine to epinephrine
for hemodynamics, lactate
metabolism, and gastric tonometric variables in septic shock
: a prospective, randomized study. Intensive Care Med
17. Vallet B, Chopin C, Curtis SE, Dupuis BA, Fourrier F, Mehdaoui H, LeRoy B, Rime A, Santre C, Herbecq P, et al: Prognostic value of the dobutamine test in patients with sepsis syndrome and normal lactate
values: a prospective, multicenter study. Crit Care Med
18. Rhodes A, Lamb FJ, Malagon I, Newman PJ, Grounds RM, Bennett ED: A prospective study of the use of a dobutamine stress test to identify outcome in patients with sepsis, severe sepsis, or septic shock
. Crit Care Med
19. Levy B, Dusang B, Annane D, Gibot S, College Interregional des Réanimateurs du Nord-Est: Cardiovascular response to dopamine and early prediction of outcome in septic shock
: a prospective multiple-center study. Crit Care Med
20. Tappy L, Chiolero R: Substrate utilization in sepsis and multiple organ failure. Crit Care Med
21. Singer M, De Santis V, Vitale D, Jeffcoate W: Multiorgan failure is an adaptive, endocrine-mediated, metabolic response to overwhelming systemic inflammation. Lancet
22. Leverve XM: Mitochondrial function and substrate availability. Crit Care Med
23. Dellinger RP, Levy MM, Carlet JM, Bion J, Parker MM, Jaeschke R, Reinhart K, Angus DC, Brun-Buisson C, Beale R, et al: Surviving Sepsis Campaign: international guidelines for management of severe sepsis and septic shock
: 2008. Intensive Care Med
24. Myburgh JA, Higgins A, Jovanovska A, Lipman J, Ramakrishnan N, Santamaria J, CAT Study investigators: A comparison of epinephrine
and norepinephrine in critically ill patients. Intensive Care Med
25. Annane D, Vignon P, Regnault A, Bollaert PE, Charpentier C, Martin C, Troché G, Ricard JD, Nitenberg G, Papazian L, et al: Norepinephrine plus dobutamine versus epinephrine
alone for management of septic shock
: a randomized trial. Lancet