Cardiopulmonary bypass (CPB) is a technique in which an external system temporarily takes over the function of the heart and lungs during cardiac surgery, maintaining the circulation of blood and the oxygen content of the body. Cardiac and major vascular surgeries require CPB. The procedure results in the deterioration of normal vascular physiology, as well as the release of a variety of debris into the bloodstream, including fragments of blood cells, tubing, and plaques. For example, when surgeons clamp and connect the aorta to tubing, the resulting emboli may block blood flow and cause mini strokes [1]. Despite its importance, the monitoring of tissue perfusion during CPB is challenging because mixed venous oxygen saturation (SvO₂) cannot be measured during CPB, and pulse oximetry does not function correctly in a hypothermic state. Therefore, conventional methods, such as mean blood pressure measurement, blood gas analysis, and intake/output, are commonly used during CPB. However, these conventional methods often fail to detect occult hypoperfusion. Abou-Khalil et al. [2] reported that conventional methods could not detect occult hypoperfusion in traumatic patients and recommended early invasive monitoring methods, such as serum lactate sampling and pulmonary catheterization, in critically ill patients.

Measurement of serum lactate, which is the most commonly used biomarker for detecting hypoperfusion, provides valuable clinical information during CPB. However, the procedure for monitoring serum lactate levels is limited by the delay in obtaining results, which renders continuous monitoring challenging. SvO₂, which is mostly used for monitoring tissue perfusion in critical patients, cannot be measured during CPB because the pulmonary artery catheter is retracted before cardiotomy. Therefore, a continuous and noninvasive method for monitoring perfusion during CPB is required.

Near-infrared spectroscopy (NIRS) is a spectroscopic method that uses the near-infrared region of the electromagnetic spectrum (from about 700 to 2500 nm). This technique was initially applied for detecting the state of regional tissue perfusion by attaching electrodes at specific parts of the body, such as the frontal head, abdomen, thenar muscle, or thigh. However, the concept that small areas of tissue perfusion can be representative of whole-body perfusion has been suggested [3].

Therefore, in this study, we attempted to validate the clinical utility of regional perfusion saturation for detecting whole body hypoperfusion during CPB. We analyzed the correlation of skeletal muscle (StO₂) and cerebral (SctO₂) tissue oxygen saturation with levels of serum lactate and delivery oxygen (DO₂).

The mean durations of operation and CPB were 453.5 (107.8) min and 162.7 (47.5) min, respectively (Table 2). The analysis of the monitoring data revealed a periodic increase in serum lactate levels (P < 0.05) at all stages of CPB (Fig. 1A). In addition, both StO₂ and SctO₂ were decreased during CPB compared to pre-CPB values (Figs. 1B and 1C). Although the pre-CPB values of DO₂ could not be obtained, the DO₂ levels at T3 and T4 were found to be lower compared to those at T2 (Fig. 2). The results of correlation analysis revealed that the serum lactate levels were negatively correlated with the StO₂ levels at T4 (R = −0.384) and T5 (R = −0.370) and positively correlated with the SctO₂ levels at T3 (R = 0.445), with the correlations exhibiting statistical significance (P < 0.05; Fig. 3). Additionally, StO₂ was positively correlated with DO₂ at T4 (R = 0.466, P < 0.05; Fig. 4).

C-reactive protein, interleukin-8, CD64, and lactate are currently used as biomarkers for detecting poor organ perfusion [5], while lactate is the most commonly used biomarker for evaluating hypoxic tissue state. Many studies have reported that blood lactate levels are highly correlated with prognosis in critically ill patients [6–9]. Vitek and Cowley [6] reported that blood lactate levels were helpful for predicting the prognosis of various forms of shock. Bakker et al. [7] confirmed that blood lactate levels are superior to oxygen-derived variables, such as DO₂ and oxygen consumption, for predicting outcomes in human septic shock. In addition, Marecaux et al. [9] showed that blood lactate levels are better prognostic indicators than levels of tumor necrosis factor and interleukin-6 in patients with septic shock. Moreover, according to the recently published Surviving Sepsis Campaign guidelines (2012) for sepsis treatment, the monitoring of serum lactate is recommended over that of other biological markers in the early sepsis state [10]. Initial serum lactate levels had been found to be correlated with mortality rates, with up to 75% mortality in critically ill patients who had lactate levels > 4.4 mM [11]. Abramson et al. [12] also reported that trauma patients with 48-hour serum lactate levels > 2 mM exhibited greater mortality and morbidity than those with serum lactate levels < 2 mM. In the present study, serum lactate levels were over 2 mM from cooling state (2.5 [1.0] mM) to post-CPB (3.9 [1.5] mM). Considering the duration of CPB (162.7 [47.5] min), serum lactate levels were high enough to suggest the occurrence of occult hypoperfusion during CPB. Park et al. [13] reported that pediatric cardiac patients who had major adverse events showed mean serum lactate levels of 4.2 mM at weaning from CPB, whereas pediatric cardiac patients who did not have major adverse events showed mean levels of 2.1 mM. In our study, the mean serum lactate level during the weaning period was 3.1 (1.1) mM.

Although blood lactate level is a good biomarker for tissue hypoperfusion, it has some limitations. Lactate level monitoring entails intermittent blood sampling and a time delay until results are obtained; however, continuous perfusion monitoring has been advocated for clinically ill patients. Sakr et al. [14] reported the development of multi-organ failure, as well as increases in lethality and mortality, consequent to increased tissue hypoxia, which had been caused by a continuous in vivo perfusion deficit. Therefore, we consider that lactate level monitoring does not provide sufficient information to prevent occult hypoperfusion during CPB.

NIRS was introduced into clinical practice for the continuous and noninvasive monitoring of tissue perfusion. This technique was first applied in the cerebral regions [15,16] and the value of cerebral perfusion was expressed as the oxygen saturation value, SctO₂. SctO₂ is commonly used to detect cerebral hypoperfusion during CPB and has become one of the essential values in cardiac anesthetic monitoring, but its purpose was limited to the protection of the cerebral region. Recently, the purpose of cerebral-NIRS was extended to detect systemic perfusion. Murkin et al. [3] showed that management of SctO₂ could reduce the duration of stay in the ICU, as well as the occurrence of renal failure and myocardial ischemia, and suggested that SctO₂ is correlated with whole-body perfusion. Moreover, Mittnacht [17] reported that that cerebral-NIRS reflects the systemic hypoperfusion in high-risk pediatric patients. Recently, NIRS was developed to include the skeletal muscles. This device measures oxygen saturation in peripheral tissues as a means of estimating systemic perfusion [18]. McKinley et al. [19] evaluated the oxygen saturation levels of peripheral tissues, including the thenar and biceps muscles, in 24 patients under ventilator care and found that oxygen saturation data collected from the thenar muscle exhibited greater accuracy compared to those collected from the biceps. Skarda et al. [20] also reported that the value of thenar muscle saturation provides appropriate information about the systemic perfusion state in patients with severe sepsis. Following this line of reasoning, we applied NIRS sensors to the thenar muscles in the present study. Crookes et al. [21] validated the StO₂ of the thenar muscle in 707 healthy, ambulatory volunteers and 150 traumatic patients. The authors recommended 87% (6%) as the normal range of StO₂ and concluded that a decreased thenar StO₂ reflects the presence of severe systemic hypoperfusion. The mean pre-CPB value of StO₂ in our patients was 81% (6%). Cohn et al. [22], in a study that included 151 patients with trauma, also evaluated whether the oxygen saturation of the thenar muscle could predict multiple organ failure. The authors reported that StO₂ predicted multiple organ failure with a sensitivity of 78% and a specificity of 39%, and it also exhibited an efficacy similar to that of other parameters, such as base deficit and blood pressure, in the prediction of multiple organ failure or death. In this study, we were able to confirm that serum lactate levels increased, while StO₂, SctO₂, and DO₂ simultaneously decreased during CPB. We also observed that StO₂ was negatively correlated with lactate concentration during the rewarming and reperfusion stages of CPB and was positively correlated with DO₂ in the rewarming stage of CPB; these findings, we believe, suggest that StO₂ is reliable in the detection of systemic hypoperfusion during CPB. Although SctO₂ decreased significantly during CPB, we could not observe a negative correlation between SctO₂ and lactate levels. Instead, we observed a positive correlation between lactate and SctO₂ in the hypothermic stage of CPB. At our cardiac surgical center, we control the acid-base balance during CPB in a pH-stat manner. We add carbon dioxide to the oxygenator gas inflow to maintain pH-stat content during the hypothermic stage of CPB. Under these conditions, cerebral blood flow becomes more dependent on PACO₂. Consequently, additional carbon dioxide could increase the total blood flow in the brain [23]. These vulnerabilities of cerebral flow due to PACO₂ might have masked the systemic value of SctO₂ in the present study. However, the value of StO₂ reflected the serum lactate level during the hypothermic period of CPB. Thus, we believe that StO₂ could reflect systemic hypoperfusion in the hypothermic stage of CPB because the hypothermia caused microcirculation disturbances.

This retrospective study has several limitations. First, the study was conducted at a single center, which renders its reproducibility at other hospitals uncertain. Second, we did not limit the inclusion criteria to a single type of surgery; instead, we analyzed various types of cardiac surgery. In addition, we analyzed only data obtained during surgery because we did not routinely monitor tissue oxygen saturation in the ICU. Third, the relatively small sample size limited this study in terms of the performance of a more categorized analysis and significance testing of some variables, although we proved the significance of the study by increasing our sample size to above 80%, the generally accepted standard. Fourth, patients’ predisposing diseases, such as chronic kidney disease, hypertension, and diabetes, might affect their blood flow reserve. Therefore, these conditions make patients more vulnerable to occult hypoperfusion during CPB. In this study, the demographic data showed 55% of patients had hypertension and 31% had diabetes. However, the purpose of this study was to evaluate the correlation between serum lactate levels and StO₂ during CPB, not to estimate the absolute values of lactate level. We assumed that hypertension and diabetes would only have a minimal effect on our results.

In conclusion, in this study, we found a significant correlation between StO₂ and serum lactate levels during CPB, which may indicate the reliability of StO₂ for the detection of systemic hypoperfusion during CPB. Therefore, we suggest that StO₂ is a valuable parameter for detecting a systemic hypoperfusion state during CPB.