This study retrospectively analyzed changes in the lactate kinetics of patients with hyperlactatemia and showed that lactate kinetics at 6 h, 12 h, 24 h, 48 h, and 72 h were 21%, 40%, 57%, 66%, and 72%, respectively. Using these values as standards, their predictive value for patient death gradually increased (AUC 0.684–0.848). Examination of the factors that influenced achieving 6 h, 12 h, and 24 h lactate kinetics targets showed that the APACHE II score, SOFA score, HR, and blood glucose were independent risk factors atthe time points that we measured. These results suggest that disease severity and the level of organ dysfunction affect the ability to achieve lactate kinetics targets. After stratification using the APACHE II score, the results showed that in critically ill patients (APACHE II ≥ 15), appropriate fluid balance and norepinephrine doses were beneficial for achieving lactate kinetics targets, whereas excessive positive fluid balance and large norepinephrine doses were harmful. Additionally, the effects of continuously achieving lactate kinetics targets on the prognosis were further analyzed and classified into eight clinical phenotypes. The results showed clinical phenotype VIII (T6, T12, and T24 targets were unachieved) had the highest odds ratio of patient death (OR = 4.39;95%CI 2.4–8.03). Even when the lactate kinetics target was achieved at 6 h but not at the following timepoints(clinical phenotype IV and II), the odds ratio still increased(OR = 4.2;95%CI 1.69–10.48 and OR = 5.27;95%CI 2.33–11.88, respectively).
Although some studies have explored the relationship between lactate kinetics and the prognosis of critically ill patients, some key issues, such as (1) the optimal cutoff value of lactate kinetics at different times and(2) the appropriate duration of monitoring lactate kinetics, remain unclear. To solve these problems, we first reviewed and analyzed the optimal cutoff value for prognosis at different timepoints.The results showed that the optimal cutoff values corresponding to these time points increased gradually. The EMShockNet investigators reported noninferiority in terms of reduction in hospital mortality among the group with lactate kinetics greater than 10% at 6 h and in the group with ScvO2 ≥ 70% at 6 h (17% vs 23%) 11. Craig et al. reported in a retrospective study that resuscitation within 6 h and lactate kinetics of 36% could predict the prognoses of patients with sepsis12.Marynaet al. reported that lactate kineticsT24h≤19% was associated with increased ICU mortality (15% vs 43%; OR 4.11)13. In addition to the specific cutoff value differences, our results are consistent with previous studies because, on the one hand, the sample size of these studies is different;on the other hand, the prognosis and lactate kinetics of critically ill patients are closely related to disease heterogeneityand treatment differences in different centers14,15. From the lessons learned from the failure of studies on supernormal goal-oriented therapy in the last century, we realized the importance of setting reasonable resuscitation goals 16,17. In fact, recovery of organ function, tissue perfusion, and even cell function during resuscitation requires time. Reasonable lactate kinetic goals can both produce the internal driving force to promote resuscitation and meet the physiological needs of the body to avoid excessive therapy caused by inappropriate and excessively high goals.
For patients with hyperlactatemia, how long should we monitor the lactate kinetics? Hernandez et al.18 confirmed that only 52% of septic shock patients had normal blood lactate levels within 24 h. In a study by Maryna et al., for patients with lactate kinetics less than 19%, the average lactate level for the first 24 h was 5.25 mmol/L, and the average for the second 24 h was 6.43 mmol/L. Even in patients with lactate kinetics greater than 19%, the average lactate level for the first 24 h was 5.10 mmol/L, and the average for the second 24 h was 2.47 mmol/L. Thus, even after 24 h of resuscitation, a large number of patients still have hyperlactatemia and hypoperfusion. Therefore, monitoring 6 h, 12 h, or 24 h lactate kinetics alone is not sufficient to guide the entire process of resuscitation therapy. Based on the above reasons, our retrospective analysis of previous patients determined lactate kinetics cutoff values at five time points, from 6 h to 72 h. In addition, with the passage of time, the lactate kinetics gradually increased, and the ability to predict the survival rate of patients was also more evident.
Our study further examined the risk factors that influence whether the lactate kinetics at each time point reach these cutoff values. Various factors affect the achievement of lactate kinetics targets in clinical practice. Disease severity and the level of organ dysfunction are important components from our dataset. As representatives of these two aspects, the APACHE II score and SOFA score both show direct influences on achieving lactate kinetics goals, indicating that they are still reliable and indispensable evaluation tools for critically ill patients. Furthermore, additional attention should be paid to reductions in stress responses in critically ill patients, and the HR is a sensitive indicator of stress in the body. Many studies in recent years have confirmed that sepsis patients obtained excellent effects after applying β-receptor blockers to control the ventricular rate19–21 to reduce stress responses. For cardiogenic shock patients, reduction in the ventricular rate can improve ventricular diastolic function to further improve the ventricular stroke volume and overall cardiac efficiency, which is beneficial for improving tissue perfusion and the prognosis22,23. In our study, HR was an independent risk factor for achieving lactate kinetics targets from 6 h to 24 h, which again confirms that the influence of HR on the treatment of critically ill patients requires attention. High blood glucose is also a presentation of stress responses in critically ill patients. One recent study showed that a high blood glucose level was an independent risk factor for in-hospital death of cardiogenic shock patients 24. Another study showed that, for both cardiogenic shock and septic shock, hyperlactatemia was mainly caused by an increase in lactate production and that the increase in lactate production was usually accompanied by high blood glucose and an increase in glucose turnover, indicating that the latter had great impacts on lactate metabolism 25. Our study also showed that blood glucose affected achieving lactate kinetics targets, suggesting that blood glucose control should be a focus during shock resuscitation.
After the lactate kinetics targets were confirmed, we evaluated the effects of clinical resuscitation measures on dynamic attainment. The results showed that for patients with critical illness (APACHE II score ≥ 15 points), conservative fluid resuscitation strategies were more likely to achieve lactate kinetics targets. A meta-analysis published in 2017 confirmed that for patients with acute respiratory distress syndrome (ARDS), sepsis, or systemic inflammatory response syndrome (SIRS), a conservative fluid resuscitation strategy can reduce the mechanical ventilation time and length of ICU stay 26. Our study results further validated that conservative fluid resuscitation also had positive effects on lactate metabolism. Previous studies showed that the application of norepinephrine to increase the blood pressure of septic shock patients to 85 mmHg did not benefit tissue oxygen metabolism, skin capillary blood flow, and urine output 27. In addition, excessive norepinephrine can cause increased peripheral blood vessel contraction, resulting in microcirculation disturbances and terminal necrosis. Our study also suggested that the norepinephrine doses were lower in the group that achieved lactate kinetics targets, which is consistent with the “less is more” concept promoted in recent years 28, with an emphasis on a balance between treatment and reinjury.
By observing the change in the trajectory of lactate kinetics while reaching the cutoff values,we confirmed that the group that continuously reached the cutoff valueshad an obviously better outcome than that of the group that reached the cutoff values at any time point. According to these dynamic changesin lactate kinetics, we proposed clinical phenotypes of lactate kinetics to identify the most critical points and the phenotype for the best prognosis. The effect of the previously reported 6 h lactate kinetics attainment rate on the prognosis was not as good as that of the 12 h and 24 h attainment rates, which was partially consistent with results from a previous study 29. Clinically, the phenotype of lactate changes can be used to determine the patient prognosis. These results suggested that, under limited resource conditions, greater focus should be placed on achieving 12 and 24 h lactate kinetics goals rather than 6 h lactate kinetics goals. Additionally, attaining high lactate kinetics goals might require more fluid infusion during treatment, the application of more inotropic drugs to increase cardiac output, and setting a higher arterial pressure, which might cause harm 16,17. The phenotype theory based on the lactate kinetics cutoff values at different timepoints represents a milestone for the entire resuscitation process; thus, the goals during resuscitation are clearer, and insufficient or excessive resuscitation can be avoided.
Our study had some limitations. First, this study was a retrospective, single-center study. The confirmed lactate kinetics cutoff values at different timepoints lack broad representation. Factors such as differences in therapeutic strategies at different medical institutions (such as blood transfusion and cardiotonic therapy) and timeliness of treatment might influence the determination of cutoff values. Therefore, a multicenter study with a clear and unified treatment plan is needed for further verification. Second, this study targeted all critically ill patients. In the future, the lactate kinetics of patients with certain diseases, such as septic shock and cardiogenic shock, can be explored according to disease classification in order to more accurately guide clinical treatment. Third, this study included a 72 h period,but some patients did not remain in the ICU for this length of time due to death and transfer out. Over time, fewer patients were included in the analysis. Therefore, we can only include all of the parameters that we can obtain, and we cannot rule out actual factors that affect the data.