Introduction

Hyperlactatemia after cardiac surgery with cardiopulmonary bypass (CPB) has long been recognized [1, 2, 3, 4, 5, 6] as a consequence of probable inadequate tissue perfusion [2, 7, 8, 9, 10, 11]. Indeed, despite some technical advances [2, 12, 13], insufficient tissue perfusion during CPB remains a concern [1, 2, 10, 11]. Recent reports suggest that elective coronary artery bypass grafting (CABG) with beating heart improves the tolerance [14, 15].

Blood lactate correlates with illness severity [16, 17], but it does not allow differentiation between excessive production and decreased consumption [9, 18]. The lactate turnover rate is increased in many pathological conditions, including cardiogenic shock [19], but decreased lactate clearance may also increase lactate concentration [20, 21]. Therefore, assessment of lactate metabolism from blood lactate concentration alone is ambiguous.

The infusion of an exogenous load of lactate makes it possible to evaluate lactate clearance and endogenous production [20]. This tool was used in the present study to investigate lactate metabolism after elective CABG.

Material and methods

Patients

Three groups of patients suffering from coronary heart disease and requiring CABG were investigated for lactate metabolism in a non-randomized prospective observational study. Exclusion criteria were: age under 19 or over 75 years, ejection fraction below 35%, renal failure (creatinine above 15 mg/l), liver failure (see Table 1), non-insulin and insulin-dependent diabetes, or patients requiring epinephrine or more than one inotropic drug or more than 10 µg/kg per min dobutamine or dopamine (see Table 2). One group was studied before surgery (preoperative, n=20) and two others 14–16 h after surgery, one with cardiopulmonary bypass (CPB, n=20) and one with beating heart (NO-CPB, n=20). All patients presented uneventful surgical procedures and the postoperative period was uncomplicated. The ethic committee of the National Cardiovascular Center, Harapan-Kita-Hospital, Jakarta Indonesia, approved the study protocol and informed written consent was obtained from every patient. The CABG procedures were performed with CPB or with the beating heart technique, depending on the surgeon's preference.

Table 1. Patient and surgery characteristics. (Preoperative before surgery (n=20), CPB postoperative after bypass (n=20), NO-CPB postoperative after surgery with beating heart (n=20)). Values are means ± SEM. Age, weight, height and body mass index (BMI) were not different within the three groups, ejection fraction [%, (end-diastolic volume end-systolic volume)×100/end-diastolic volume] was significantly lower in CPB patients as compared to both the preoperative group and NO-CPB (ANOVA and post-hoc Fischer's test 1 p<0.01), the number of grafts between CPB and NO-CPB was compared by a chi-square test, 2 p<0.05
Full size table
Table 2. Postoperative use of inotropes and/or vasoactive agents. (CPB postoperative after bypass (n=20), NO-CPB postoperative after surgery with beating heart (n=20))
Full size table

Surgical procedure and peri/postoperative care

Anesthesia was standardized: after premedication with 2.5–5 mg of midazolam i.v., induction was performed with propofol (3–5 mg/kg) and 1–2 µg/kg sufentanil before intubation with 0.1–0.2 mg/kg pancuronium bromide. Anesthesia was maintained with a continuous infusion of propofol, and i.v. bolus of sufentanil was administered when there was evidence of poor analgesia and/or when blood pressure or heart rate increased more than 30%. Nitroglycerin was continuously infused and the rate was modulated according to blood pressure: it was increased when blood pressure increased more than 30%.

For CPB, the extracorporeal circuit consisted of a roller pump (Sarns 8000) and a membrane oxygenator (Capiox Sx). The pump flow was 2.4 l/min per m2. During CPB, patients were cooled to 31–32°C and they received intermittent antegrade normothermic blood cardioplegia, which was a mixture of 400–600 ml of oxygenated blood with graduated doses of potassium-magnesium solution. Standard systemic heparinization (3 mg/kg) was performed and an activated clotting time of greater than 480 s was maintained during CPB; heparin was neutralized with protamine sulfate on discontinuation.

For the beating heart technique, traction sutures were applied to the pericardial edges, displacing the heart anteriorly. For exposure of the left anterior descending coronary artery or its diagonal branches, additional pericardial traction sutures were inserted anterior to the left phrenic nerve to rotate the heart and a moist sponge was placed behind its laterodorsal aspect, bringing the coronary artery into the operative field. For exposure of obtuse marginal or right coronary branches, two wet cotton tapes were passed through the transverse sinus with their right ends secured to surgical drapes. The two loose lengths of tape were then used to lift and rotate the heart toward the surgeon, as well as to stabilize the coronary artery. Patients were heparinized (1.5 mg/kg) and two 5-0 Prolene (Ethicon, Sommerville, NJ) sutures were used to occlude the coronary artery on either side of the anastomosis site temporarily. Heparin was neutralized with protamine sulfate after all anastomoses were completed.

Postoperative care was also standardized: (1) when the cardiac index was below 2.5 l/min per m2 despite sufficient volume expansion (pulmonary capillary wedge pressure 12–16 mmHg), dobutamine was infused; (2) the mean arterial pressure was maintained between 60 and 90 mmHg with norepinephrine and nitroglycerine; (3) hemoglobin concentration was maintained around 100 g/l. Patients were weaned from mechanical ventilation as soon as they had been rewarmed and were hemodynamically stable. All patients were extubated before the study, meaning that all were hemodynamically stable.

Exogenous lactate challenge test (ELCT)

Lactate metabolism was investigated at 8 a.m. after an overnight fast in preoperative patients or 14–16 h after surgery in the two postoperative groups, also under fasting conditions (Fig. 1). The total duration of the study was 180 min. In preoperative patients, one venous cannula was inserted into a wrist vein of an arm for the collection of blood samples. This hand was placed in a thermostabilized box heated at 50°C to achieve partial arterialization of venous blood [20]. Another venous cannula was inserted into an antecubital vein of the contralateral arm for lactate infusion. In postoperative patients lactate was infused through a central venous line, while blood samples were collected from a radial artery catheter. After a first blood sample was taken for basal lactate determination (T−15), 2.5 mmol/kg body weight sodium lactate was infused over a 15-min period. At the end of the infusion (T0) and after 5, 10, 15, 30, 60, 90 and 120 min, blood samples were collected for immediate bedside whole blood lactate determination (Analogs-LM-5 lactate analyzer, Analox Instruments, Hammersmith, London, UK), which was calibrated before each investigation. The response was linear between 0 and 20 mm, the coefficient of variation was1.6% and the identity curve comparing this method and the standard enzymatic procedure was y=1.01x0.09, r=0.998, n=56.

Fig. 1.
figure 1

Exogenous Lactate Challenge Test in three groups of patients before or after CABG surgery with or without cardiopulmonary bypass. Three groups of patients: preoperative patients (open circles, n=20) and postoperative patients (14–16 h after surgery) with cardiopulmonary bypass (CPB, closed circles, n=20) or without bypass (NO-CPB, closed squares, n=20) were challenged with an exogenous load of sodium lactate continuously infused from −15 min to 0. Curves represent the best fits of bi-exponential modeling (see Materials and Methods) of the decay of the plasma lactate from T0 up to 120 min

Full size image

Calculations

The clearance of exogenous lactate (ml/kg per min) was the ratio of the lactate load (mmol/kg) to the area under the lactate concentration curve (μmol/min per ml). Areas were calculated graphically from the unadjusted experimental points by using Kaleidagraph (Abelbeck Software, Reading, PA, USA) as described in [20]. The endogenous lactate production (μmol/kg per min) was calculated as the product of basal lactate concentration and lactate clearance rate. There is no fitting of the data and the area under the curve (representing the part of exogenous lactate not cleared at any given time) is calculated graphically using the method of sum of trapezes. Therefore, estimation of endogenous lactate production by this method is relatively robust, depending only on two assumptions: (1) the dilution space of lactate was estimated to be 60% of body mass (a value widely used) and (2) exogenous lactate infusion does not significantly affect endogenous lactate production. Data from the literature supports the latter fact [22].

The kinetics of exogenous plasma lactate disappearance could be assessed with either a mono-compartmental model [22, 23, 24, 25, 26] or a two-compartmental model [20, 26, 27, 28]. From a theoretical point of view, a two-compartmental modeling seems more appropriate since lactate utilization is related to two major processes, i.e. oxidation and transformation into glucose. We have compared the mathematical fitting resulting from either mono- or bi-compartmental pharmacokinetic modeling in healthy controls and in postoperative patients [20, 26, 27, 28]. Bi-exponential fitting resulted in a better description of the lactate decay, therefore this modeling was applied to describe the kinetics of plasma lactate decay over time from the end of exogenous lactate infusion (T0 = peak of lactate) to 120 min (T120). Experimental values were fitted by means of the least-square method, using commercial software (Kaleidagraph, Abelbeck Software, Reading, PA, USA). In each experiment, mathematical fitting of the experimental data of lactate decay permits the calculation of two different half-lives corresponding to the two exponential curves (HL-I and HL-II).

Statistical analysis

The results are presented as means ± SEM. When variables were normally distributed, statistical comparisons were made as indicated either by a paired Student t-test or by an unpaired Student t-test or, when three groups were compared, by an ANOVA followed by post-hoc analysis (Fisher's PLSD) if a significant difference (p<0.05) was found by ANOVA. In the case of non-normally distributed variables (HL-1 and HL-2), comparisons were first analyzed by a Kruskal-Wallis test. When Kruskal-Wallis analysis showed significant differences among the three groups a post-hoc comparison by a Mann-Whitney test was then performed. Non-parametric variables (number of grafts, use of inotropes) were compared by a chi-square analysis.

Results

The baseline characteristics of the participants and of the surgical procedure are listed in Table 1. The three groups were matched for age, height, weight and body mass index (BMI). Ejection fraction was similar between preoperative and NO-CPB, while it was lower in the CPB group (p<0.05). In the CPB group the bypass time was approximately 100 min, while the mean aortic cross-clamping time was 68 min. During this procedure, patients were placed in hypothermia (31°C), but body temperature at the time of admission to the ICU was the same for CPB and NO-CPB. The number of grafts was significantly higher (p<0.05) in CPB, while the total length of anesthesia was shorter (p<0.05) as compared with NO-CPB. Table 2 shows the use of inotropes or vasoactive agents, which was not different for CPB and NO-CPB.

Hemodynamic parameters (Table 3) at the time of the lactate test were identical in the two postoperative groups. As shown in Table 4, basal lactate was not different in the three groups. At the end of lactate infusion, the peak value of blood lactate was steadily and significantly higher as compared to the basal value as a result of the lactate load. The peak value was significantly higher in the CPB group as compared to both preoperative and NO-CPB groups, while there was no significant difference between the latter two groups. After 120 min blood lactate returned to the basal value in both preoperative and NO-CPB groups, while it remained significantly higher in the CPB group. Hence, as compared to preoperative patients, equivalent lactate load resulted in higher peak and T-120 blood lactate values in the CPB group.

Table 3. Hemodynamics parameters. (CPB postoperative after bypass (n=20), NO-CPB postoperative after surgery with beating heart (n=20), VO 2 oxygen consumption, DO 2 oxygen delivery, MAP mean arterial pressure, lowest MAP lowest mean arterial pressure observed during the 12 h following surgery) Values are means ± SEM, no significant difference (unpaired Student's t-test) was observed between the two groups except for cardiac index, *p<0.05
Full size table
Table 4. Parameters of lactate metabolism. (Lactate (T15) blood lactate before lactate infusion, lactate (T0) peak of blood lactate at the end of lactate infusion, lactate (T120) blood lactate 120 min after the end of lactate infusion).). Groups were the same as in Table 1. Values are means ± SEM, for non-normal distributed parameters (HL-1 and HL-2) minimal and maximal values were added in parentheses (min-max).
Full size table

The evolution of blood lactate was different between CPB and both preoperative and NO-CPB groups (Table 4). The lactate clearance was significantly lower in the CPB group as compared to both preoperative and NO-CPB groups, but there was no difference between the latter two groups. Endogenous lactate production among the three groups was not significantly different. Fitting of mathematical modeling was assessed by correlation coefficients (r), which were similar in the three groups: 0.98±0.003, 0.99±0.002 and 0.98±0.002 for preoperative, CPB and NO-CPB, respectively. Bi-exponential fitting allowed the calculation of two distinct half-lives (HL-1 and HL-2), which were similar in preoperative and NO-CPB groups, while both parameters were significantly different in the CPB group. As compared with preoperative and NO-CPB, HL-I was lower in the CPB group, whereas HL-II was higher. This indicated that total lactate clearance and the balance between the two different states were affected in the CPB group.

Discussion

The data indicate that 14–16 h after surgery, patients undergoing a CPB for CABG exhibit altered lactate metabolism as compared with patients undergoing CABG without bypass or with preoperative patients. Although the selection of patients between CPB and NO-CPB groups was not based on randomization, strict identical criteria of inclusion have been used regarding patient selection as well as perioperative and postoperative management. In fact, the surgeon's choice between NO-CPB or CPB was based on the anticipation of the surgical difficulty of the procedure, as judged from preoperative angiography, and not from the severity of the disease. Due to lack of randomization and the limited number of patients in each group (20 patients), a difference in ejection fraction was found between NO-CPB and both preoperative and CPB groups. However, though lower by 10% in the CPB group (45% as compared with the other groups' 55%), this value of ejection fraction was clearly higher than 35%, a value recognized as a threshold for postoperative complications. Cardiac index before the test was lower in the NO-CPB group as compared to the CPB group, indicating that if the preoperative ejection fraction was higher in NO-CPB as compared to CPB, the cardiac function before the test was not better. In addition, there was no significant difference between the two postoperative groups concerning the use of inotropes (Table 2) and hemodynamic data (Table 3).

The alteration in lactate metabolism is minor since neither basal lactate nor endogenous lactate production differed much among the three groups and the alteration was only revealed by the exogenous lactate infusion. Interestingly, lactate turnover is lower in all three groups of the present study as compared to data from the current literature, including our own previous study [20]. We have recently investigated Indonesian healthy controls and we found much lower values than those currently found in a healthy European population (unpublished). Hence basal lactate turnover appears to be lower in an Asian population living in Indonesia as compared to a Caucasian one living in Europe.

Lactate clearance and metabolism (HL-1 and HL-2) were affected in the CPB group, and the difference persisted when adjusted for the number of independent parameters of lactate metabolism. Considered all together, these results are very coherent since all five parameters investigating lactate metabolism: (1) maximal lactate concentration, (2) lactate recovery after 120 min, (3) lactate clearance, (4) HL-1 and (5) HL-2, significantly differed between the CPB and both preoperative and NO-CPB groups, while none of them differed between the preoperative and NO-CPB groups (Table 4). Hence, the CPB group was, indeed, different from the other two regarding lactate metabolism. The decrease in lactate clearance was most probably responsible for the increase in both peak and residual value after 120 min in the CPB group. A decrease in the volume distribution of lactate might contribute to increase the lactate peak. However, the peak was approximately 50% higher in the CPB group as compared to the other two and it is unlikely that such a difference between CPB and NO-CPB patients is related to hypovolemia in view of the similar hemodynamic data (Table 3).

The minor alteration in clearance after bypass was insignificant in these uncomplicated postoperative patients, unless lactate production increased, as is mimicked here by the exogenous lactate infusion. Hence, any rise in lactate production, e.g. in shock, hypoxia, epinephrine administration, etc., would increase blood lactate, an effect amplified by the moderate decrease in the clearance. Therefore, in these conditions of cardiac surgery, increased blood lactate probably suggests an increased lactate production. The lack of effect of beating heart surgery (NO-CPB) is another important finding. In this group, lactate metabolism was very similar to preoperative patients.

The recycling level of lactate depends on liver function, and the significant increase in the HL-II observed in CBP patients could be interpreted as decreased lactate metabolism through gluconeogenesis as was proposed after hepatectomy [20]. Accordingly, the significant decrease in HL-I in these CPB patients may indicate a higher probability of lactate being oxidized. Hence, the balance between oxidation and recycling is probably slightly modified in CPB patients: lactate oxidation is favored while lactate recycling is impaired. Again, these changes have been observed only when lactate metabolism was driven to an extremely high value by the exogenous lactate infusion. The understanding of the consequences of increased lactate oxidation by various tissues needs to be considered in the frame of the complex metabolic network of the whole body, i.e. inter-organ substrate cycles.

On the basis of several considerations [9, 29], it can be proposed that increased lactate turnover (i.e. production by some tissues and simultaneous consumption by others) represents a metabolic way of sharing substrate oxidation (respiration) among different tissues or cells. Indeed, when a net flux of lactate is released by some cells, this implies a net efflux of reducing equivalents (i.e. a non-complete oxidation), whereas a net uptake of lactate by others means a net import of exogenous-reducing equivalents, which are then oxidized. This could be viewed as an adaptive mechanism: some cells supplying others with regards to oxidative metabolism. Conversely, when the rise in lactate is mainly the consequence of a decreased lactate clearance, this is probably a negative event since it indicates a reduction of lactate clearance at the level of the body as a whole, i.e. a decrease in the oxidative capacity. This was shown recently in critically ill patients, where a decreased lactate clearance was associated with poor outcome [30].

Liver hypoperfusion and hypothermia during surgery may participate in the liver alteration. It was recently shown that hypothermia during CPB was responsible for metabolic disturbances [32]. In this report, increased blood lactate was found during and after surgery in hypothermic conditions associated with a reduction of acetoacetate/3-hydroxybutyrate ratio indicating a potential liver dysfunction [32]. Interestingly, hepatic blood flow was not affected by temperature while liver oxygen extraction was decreased in the hypothermic group, suggesting a metabolic effect. Although hypothermia decreases the liver respiratory rate and metabolic activity, it is not yet clear how these effects can persist 1 or 2 days after the end of the hypothermic period.

An inflammatory response could also explain the deleterious consequence of CPB on liver metabolism. Decreased splanchnic flow could cause endotoxin or bacterial translocations and endotoxins are present in the inferior vena cava during CPB [33, 34]. Endotoxin exerts some metabolic effects on liver cells [35], but the use of steroid pretreatment reduces their release [36]. Extracorporeal circulation exposes blood to artificial membranes and to mechanical damage of the circulating cells responsible for activation of complement [37, 38] and platelet decrease. By increasing the regional generation of free oxygen radicals, CPB has been associated with some postoperative dysfunctions [39, 40]. Activated granulocytes and free radicals were shown to be drained from the inferior vena cava [40]. Nevertheless, the surgical procedure per se could also be responsible for the inflammatory response and, indeed, some authors did not find any difference according to the surgical procedure with or without CPB [41, 42].

In conclusion, the CPB procedure for CABG alters lactate metabolism by decreasing its clearance, this effect being possibly located in the liver, but the exact mechanism of such a deleterious effect is not clear. In uneventful elective surgery this alteration is minor since exogenous lactate infusion is necessary to substantiate lactate concentration abnormality. This could indicate that a lactate production increase is probably responsible for a high lactate concentration after cardiac surgery. The lack of consequence on lactate metabolism of the beating heart procedure could also be emphasized.