4-h mean lactate clearance as a good predictor of adverse outcome in acute cardiogenic pulmonary edema: a pilot study

Introduction

Acute heart failure (AHF) is defined as a marked escalation in clinical symptoms or a sudden decrease in cardiac reserves requiring urgent therapy [1]. Acute cardiogenic pulmonary edema (ACPE), one of the most dramatic clinical forms of AHF, is an acute respiratory distress associated with rapid transudative fluid development in the pulmonary interstitium and alveoli, which often occurs in elderly patients. Pathophysiologically, ACPE is characterized by elevated pulmonary capillary pressure, which results from dysfunction of the left ventricle, either systolic or diastolic [2]. The leading etiology of ACPE is acute myocardial infarction (AMI)-induced left ventricular failure [3]. This condition causes myocardial damage, leading to a decrease in cardiac functional reserve. In addition, hypertensive emergencies lead to increased venous and pulmonary capillary pressures [4].

When cardiac arrhythmia, conduction abnormalities, or vasoconstriction cause more afterload and hypervolemia causes more preload, the heart’s ability to work is weakened.

These pathophysiological changes cause fluid to shift from the pulmonary capillaries to the interstitial spaces of the lung and alveolar spaces [5]. In clinical practice, patients with this condition usually present to the emergency department (ED) with severe episodes of dyspnea and hypoxia (SO₂<90 %) [6]. Rapid assessment and treatment of patients are vital for reducing the risk of respiratory failure and cardiovascular collapse.

Lactate is produced in the absence of oxygen as a byproduct of cellular metabolism and is indicative of tissue hypoxia and impaired perfusion. Although lactate levels are usually used in critically ill patients, mostly in shock, they can be elevated for several reasons [7]. Elevated blood lactate levels are strong prognostic markers for tissue hypoperfusion and are often used in therapeutic decision-making. This is an important part of evaluating the success of the treatment. Kawase et al. [8] reported that higher lactate levels on admission were associated with worse in-hospital mortality in patients with acute decompensated heart failure (HF) with and without acute coronary syndrome (ACS). Bakker et al. [9] indicated that measuring serial blood lactate levels could effectively serve as a predictor of multiple organ failure and mortality in patients with septic shock. In contrast to a single lactate level reflecting an instantaneous static state, Scott et al. [10] highlighted in a study on cardiac failure that serial measurements of lactate concentrations and dynamic changes in lactate levels may be more useful in indicating patient prognosis, arguing that 0–2 h lactate clearance (LACclr), is a good marker for prognosis prediction. However, no study has evaluated in-hospital clinical outcomes and treatment effectiveness in patients treated for ACPE in the ED after the first critical 4 h, the most intensive period of treatment. In this context, a single lactate or LACclr measurement in the early period in the ED may be insufficient to serve as a predictor of clinical outcome or treatment efficacy over time. In a systemic review published by Vincent et al. on lactate kinetics [11], decreasing lactate levels over time were consistently associated with lower mortality rates in critically ill patients. However, the authors suggested the use of a dynamic evaluation of serial lactate concentrations, regardless of the initial value, as lactate levels change, and these changes are relatively slow. Therefore, their recommendation was to measure lactate levels every 1–2 h for follow-up in acute conditions.

However, the question of which lactate clearance time interval is a good predictor of patient prognosis remains unclear. This study argues that the initial lactate levels before treatment in acute cardiogenic pulmonary edema are insufficient to assess resuscitation in these patients.

In light of the literature mentioned above, there is a need for better indicators that can be used to predict the clinical outcomes and treatment efficacy in critically ill emergency patients with ACPE. Studies of HF have often focused on unclassified HF. The present study is the first prospective study of 4-h LACclr in patients followed up at an AHF clinic in the ED.

The main objective of this study was to assess the predictive value of the 4-h mean LACclr level for in-hospital mechanical ventilation, in-hospital mortality, and 30-day mortality in patients with ACPE.

Materials and methods

Results

The study included 44 patients who met the inclusion criteria. The mean age of the patients was 70.8 ± 8.5 years, and the female and male percentages were 56.8 and 43.2 %, respectively. All patients (n=44) had hypertension (HT), 75 % (n=33) had coronary artery disease (CAD), 68.2 % (n=30) had diabetes mellitus (DM), 38.6 % (n=17) had HF, 20.5 % (n=9) had atrial fibrillation (AF) and chronic obstructive pulmonary disease (COPD), and 25 % (n=11) had chronic renal failure (CRF). Half of the patients (n=22) followed up in the critical unit of the ED received NIMV support. At the end of the 4-h follow-up period, three patients (6.8 %) were intubated. The median (IQR 25–75) lengths of ICU and ward hospitalizations were 2 (1–3) and 4 (3–6), respectively. There was no significant difference in ICU length of stay according to whether the patients had a negative outcome (intubation and/or in-hospital mortality), whereas there was a statistically significant difference in the total length of hospitalization (p=0.033) (Table 1). Eight patients (18.2 %) required invasive endotracheal intubation, and mortality occurred in three patients (6.8 %). A positive outcome (discharge from the hospital without needing IMV) was observed in 81.8 % (n=36) of patients. The 30-day mortality rate was 11.4 % (n=5). The demographic and clinical data of the patients are presented in Table 1. There were no statistically significant differences (p>0.05) in the in-hospital negative outcome, need for NIMV, or 30-day mortality in patients divided into two groups according to 4-h mean LAC levels (<2 mmol/L and ≥2 mmol/L) and initial lactate levels (<4 mmol/L and ≥4 mmol/L). Table 2 presents this relationship. When in-hospital mortality rates of patients were compared according to lactate levels and/or mean LACclr, no statistically significant difference was found between those who died and those who were discharged for initial lactate, 4-h mean lactate, LACclr 0–1 h, LACclr 0–2 h, and 4-h mean LACclr (p=0.913, p=0.397, 0.494, p=0.408, p=0.055, respectively). On the other hand, in 30-day mortality, LACclr 0–2 h and 4-h mean LACclr values showed a significant difference between the groups (p=0.013 and p=0.005, respectively). This relationship is presented in Table 3. Furthermore, the LACclr values at 0–2 h and 4-h mean LACclr were significantly different from the other indicators between the groups categorized according to the need for endotracheal intubation (p=0.024 and p<0.001, respectively). Table 4 presents this relationship. The mean values of aspartate aminotransferase (AST, IU/L) and alanine aminotransferase (ALT, IU/L) were 26.61 ± 16.04 and 20.53 ± 13.10, and the median values (IQR, 25–75) were 22.50 (16.25–32.75) and 16.50 (13–24), respectively. The mean creatinine (mg/dL) value of the patients was 1.40 ± 0.60) and the median value (IQR, 25–75) was 1.21 (0.98–1.67). AST, ALT, and creatinine levels were not significantly associated with the need for IMV, in-hospital mortality, or 30-day mortality in patients with and without negative outcomes (p>0.05). In the correlation analysis of continuous variables, a negative correlation was found between pH and total length of hospitalization, as well as between hemoglobin and Pro BNP (r=−0.0312, p=0.039; r=−0.453, p=0.002, respectively) and a positive correlation between troponin and Pro BNP (r=0.434, p=0.003). The optimum cut-off value of 4-h mean LACclr for predicting 30-day mortality was 5.57 %, with 80 % sensitivity and 66.7 % specificity (p=0.033). In addition, the optimum cut-off value of 2-h (0–2 h) LACclr in predicting 30-day mortality was 17.26 % with the same sensitivity and specificity. However, this result was not statistically significant (p=0.064). The ROC analysis results are shown in Figure 1. Furthermore, the threshold to rule out 30-day mortality for all patients was 18.85 with 100 % sensitivity and 30.2 % specificity (AUC, 0.795 95 % CI [0.546–1.000], p=0.033).

Figure 1:

ROC curves of 4-h mean LACclr and 2-h LACclr (0–2 h) to predict 30-day mortality. ROC, receiver operating characteristic; LACclr, lactate clearance. According to the figure, the optimal cutoff value for 4-h LACclr in predicting 30-day mortality was 5.57 %, with a sensitivity of 80 % and specificity of 66.7 % (p=0.033). For 2-h LACclr, the optimum cutoff value was 17.26 %, but it was not statistically significant (p=0.064).

Discussion

In this study, we investigated the relationship between the mean serum LACclr levels in the first 4 h after the initiation of treatment and prognosis in patients with ACPE. To our knowledge, this is the first prospective study to show that the 4-h mean LACclr level strongly predicts 30-day mortality in patients with ACPE. The main finding of our study was that patients with a negative in-hospital outcome or mortality within the first 30 days had a statistically lower 4-h mean LACclr rate than those without a negative outcome or mortality. There are many studies on lactate levels in critically ill patients in the literature [14], [15], [16], [17], [18], [19], [20]. However, limited studies attempt to explain the role of changes in initial lactate levels or LACclr concentrations on prognosis in AHF [21, 22].

Scott et al. [10] emphasized that serial measurements of lactate concentrations may be more useful in indicating prognosis; they argued that 0–2 h LACclr is a good marker for predicting prognosis. Bosso et al. also investigated how the 24-h time-weighted mean lactate level in AHF affects prognosis. They found that patients with values above 1.6 and 2.79 mmol/L had poor and worse outcomes compared to patients with values below 1.6 and 2.79 mmol/L [21]. However, we believe that lactate levels or clearance rates measured in a very short or relatively long period of time may be insufficient as predictors of treatment success or negative outcomes in serious clinics such as ACPE requiring intensive resuscitative treatment and follow-up. In the model we applied in our study, we proposed that the most appropriate time period for LACclr evaluation in this patient group is the first 4-h period when the treatment is applied most intensively and that the mean rate of change in lactate levels obtained hourly or bi-hourly during this period will be more useful in predicting in-hospital negative outcomes and 30-day mortality. In the current study, the 4-h mean LACclr was statistically more significant than the 2-h mean LACclr in predicting 30-day mortality and intubation requirement. As known, metabolic alterations such as increased glycolysis and catecholamine-stimulated Na-K pump activity can contribute to elevated blood lactate concentrations [23]. Reduced lactate clearance, primarily as a result of liver hypoperfusion, can also lead to elevated blood lactate levels [24].

However, this study revealed that the median values of liver and renal function tests were within normal limits, suggesting no significant abnormalities that could impact lactate clearance. Furthermore, no notable discrepancy was observed between the groups in terms of AST, ALT, and creatinine levels with respect to poor clinical outcomes.

The in-hospital mortality rates of patients with AHF are difficult to determine because the causes, severity, and treatment protocols are highly variable. According to the results of large-scale clinical trials, in-hospital and 1-year mortality rates vary widely, ranging from 4 % to nearly 40 % [25], [26], [27], [28]. High in-hospital mortality rates of 10–20 % have been reported for ACPE, particularly when associated with ACS. However, studies on in-hospital mortality rates in patients with ACPE but without AMI are limited. In our study, the in-hospital mortality rate was 6.8 %, and the 30-day mortality rate was 11.4 %, which is consistent with the literature. In a study by Zymliski et al., patients with AHF were analyzed in two groups according to lactate levels ≤2 mmol/L and >2 mmol/L, and 1-year mortality was found to be statistically significant in patients with high lactate levels [29]. However, in the present study, no significant difference was found between the patients in terms of early prognosis in terms of in-hospital and 1-month mortality at both the 2 mmol/L and 4 mmol/L thresholds.

In a study by Regnier et al., the AUC values for initial LACclr and 2-h LACclr in terms of mortality were found to be 0.78 and 0.70 for all patients [30]. In the study by Odom et al., these two values were found to be 0.88 and 0.78, respectively, in patients with >4 mmol/L, and these values were reported to be independent predictors for early (48 h) and late mortality [18].

In our study, the AUC value was 0.526 for the initial lactate level and 0.682 for the mean 4-h lactate in all the patients. However, these two values were not statistically significant in terms of 30-day mortality. In contrast, the AUC value for 4-h mean lactate level was 0.795 in all patients, which was statistically significant in predicting 30-day mortality. Even so, the AUC value for 2-h LACclr (0–2 h) was 0.756 in all patients, which was not statistically significant in predicting the 30-day mortality.

The ESC 2021 guidelines highlight the ongoing need for evidence-based biomarkers to enhance the diagnosis, treatment, and prognosis of patients with AHF, thereby improving organ perfusion and influencing clinical outcomes. ESC has suggested various biomarkers to identify the etiologic causes of AHF or to exclude the diagnosis when suspected, such as pro-BnP, NT-proBnP, troponin, procalcitonin, D-dimer, thyroid stimulating hormone (TSH), and lactate. For instance, troponin is used to diagnose ACS, D-dimer to detect or exclude pulmonary embolism, BNP to exclude non-cardiogenic causes, procalcitonin to identify infections, ferritin to assess anemia, and TSH to assess thyroid function [12]. Among these, lactate has emerged as a promising marker for assessing tissue perfusion, particularly in acute decompensated HF. Lactate monitoring, when combined with clinical assessment and vital signs, is expected to play a significant role in the management of acute severe pulmonary edema. Its potential utility lies in helping clinicians comprehensively assess the tissue perfusion status, contributing to more effective patient care and improved outcomes.

Palazzuoli et al. showed in their clinical studies that erythropoietin supplementation in patients with cardiorenal anemia increased BNP levels [31]. In our study, a statistically significant negative correlation was found between hemoglobin and BNP levels. Although high BNP and low hemoglobin levels are expected in patients with severe HF, correcting anemia can alleviate the symptoms of shortness of breath in patients with HF. In addition, Troponin and BNP provide potential utility as screening methods for identifying individuals at an elevated risk of chronic HF [32]. In the current study, a positive correlation was found between troponin and BNP levels. In this respect, these two markers may be important in screening for new HF as well as existing HF. This study focused on in-hospital mortality, intubation rates, and 30-day mortality as the primary endpoints. Future studies should consider additional outcome measures, such as long-term survival, quality of life, and healthcare utilization. Further validation of the findings through prospective, randomized controlled trials or meta-analyses is essential to confirm the predictive value of 4-h mean lactate clearance in patients with ACPE.

Conclusions

This study sheds light on the predictive potential of mean serum LACclr levels during the initial 4 h of treatment in patients diagnosed with ACPE. This pilot study marks the first endeavor of its kind, demonstrating a robust correlation between the 4-h mean LACclr level and 30-day mortality among patients with ACPE. Notably, patients who experience unfavorable outcomes during their hospital stay, particularly those who pass away within the first 30 days, often demonstrate a substantially lower 4-h mean LACclr than those who do not experience such adverse outcomes. The optimum cut-off value of 4-h mean LACclr for predicting 30-day mortality was 5.57 % with 80 % sensitivity and 66.7 % specificity. The threshold for ruling out 30-day mortality for all patients was 18.85, with 100 % sensitivity and 30.2 % specificity. Although prior studies have investigated lactate dynamics in critically ill populations, research exploring the impact of fluctuations in initial lactate levels or LACclr concentrations on the prognosis of AHF, particularly ACPE, remains scarce. Our study underscores the significance of serial lactate measurements and the dynamic changes in lactate levels over time, suggesting their potential superiority over isolated static measurements in prognostic assessments. In conclusion, in patients presenting with ACPE, 4-h LACclr calculated within the first 4 h after ED presentation may be used as a predictive indicator for in-hospital and 30-day mortality. In addition, in this patient group, 4-h LACclr may be a useful index for identifying patients with a higher risk of adverse outcomes that may require aggressive treatment in patients admitted from the ED to the ICU or ward. While the 4-h mean LACclr showed promising predictive value, further validation through clinical randomized studies is essential to confirm its efficacy as an index of adverse outcomes in patients with ACPE.