In our study, no association was found between serum lactate levels and AMI diagnosis. To our knowledge, our study is the first to compare the prognostic value of serum lactate on ICU mortality between AMI and control patients.
In a previous study, serum lactate was associated with the diagnosis of AMI . As the increase in serum lactate was initially thought to be due to intestinal cell hypoxia, serum lactate was considered to be the best diagnostic biomarker of AMI . However, several experimental studies suggested that AMI was not systematically associated with increased serum lactate [22, 23]. Furthermore, lactate metabolism is complex. In critical situations, such as in severe AMI, the increase in serum lactate may include 1) an intestinal hypoxic component, 2) a component related to the systemic inflammatory response and its associated hypermetabolism and 3) a component related to decreased hepatic metabolism [24, 25]. Finally, severe AMI may itself be associated with septic complications (digestive perforation, peritonitis) and hemodynamic events, which may themselves be associated with elevated serum lactate.
In 2012, Demir et al. reviewed seven clinical studies testing serum lactate as a diagnostic tool for AMI . T he sensitivity of lactate as a diagnostic marker of AMI in these studies ranged from 33–78% for a specificity ranging from 36–72%. Other studies confirmed that lactate is not a diagnostic marker of AMI [6, 8, 26, 27]. The ESVS and the ESTES therefore recommended that serum lactate is not used to diagnose or rule out AMI [2, 3]. In the largest published cohort of 780 AMI patients, 23% had normal serum lactate levels at the time of AMI diagnosis , confirming the danger of not making this diagnosis when there is no increase in serum lactate. Unfortunately, serum lactate is still widely used  and the diagnosis of AMI may still be delayed when serum lactate is normal, leading to delayed management and a loss of opportunity for patients . To date, neither serum lactate nor any other biomarker has been shown to be useful in the diagnosis of AMI [2, 3].
In our study, the cohort included only ICU patients with a median SAPS II score of 56 (predicted mortality 59.8%) and an ICU mortality of about 40%. There was no significant difference in serum lactate between the two groups. This lack of significant difference between the two groups translates into poor performance of serum lactate to diagnose AMI. Our study therefore confirms the studies showing that lactatemia is a poor diagnostic tool in AMI . As a result, a normal lactatemia should not rule out the diagnosis of AMI . In other words, we observed no specific association between serum lactate and AMI diagnosis.
Increased serum lactate predicted ICU mortality in both groups suggesting that there is no specific association between serum lactate and AMI prognosis. In a recent single-center retrospective study including 221 AMI patients, Sindall et al. observed that an increase of each mmol/L of lactate above the mean (3.04 mmol/L) was associated with an odds ratio for mortality of 1.36 [1.10; 1.66] . In a single-center retrospective study including 214 patients, Caluwaerts et al. showed that maximum vasopressor dose, change in lactate during the first 24 hours in the ICU and anticoagulation were the three predictors of mortality in AMI . The AMI mortality reported by the authors of this study was close to our ICU mortality. Moreover, the ability of serum lactate at D0 to predict mortality was also in line with their findings (AUC ROC curve of 0.72 [0.64–0.80] vs. 0.74 [0.65; 0.82]). However, we did not observe the same results for D1 serum lactate levels and the difference between D0-D1. Indeed, they found that the difference D0-D1 was the best predictor of mortality (AUC 0.89 [0.84–0.94]) whereas we reported a low AUC (0.59 [0.47; 0.71]). In our study, we did not include patients with AMI diagnosed after ICU admission. This probably led to the selection of more cases of AMI with a vascular cause. Indeed, non-occlusive mesenteric ischemia (NOMI) often occurs during the ICU stay and not at ICU admission. NOMI is typically associated with a form of circulatory failure with low cardiac output. Although NOMI is considered to be more severe than AMI with a vascular cause , correction of low cardiac output may normalize serum lactate and improve digestive circulation . In this situation, a decrease in serum lactate probably reflects the overall improvement of the patient rather than specifically the perfusion of the bowel.
Our results confirm that serum lactate is indeed predictive of mortality in AMI but also shows that this is not specific to AMI.
Even if serum lactate is not a specific marker of AMI prognosis, once AMI is diagnosed, serum lactate can be useful for patient management. Nuzzo et al. found in 67 patients that lactate was one of the three factors predicting irreversible AMI called irreversible transmural intestinal necrosis and could therefore be used to decide on surgical intervention . However, in this study the predictive threshold for irreversible transmural intestinal necrosis was 2 mmol/L, which can be considered a low serum lactate level. Furthermore, the model was accurately predictive of irreversible transmural intestinal necrosis when at least two risk factors were present. Thus, if hyperlactatemia was the only risk factor, AMI was irreversible in only 38% of cases. Emile et al. did not find that serum lactate was predictive of intestinal necrosis in a cohort of 101 patients including a majority with venous thrombosis . Hyperlactatemia may be more indicative of systemic hemodynamic impairment or multiorgan failure than specific bowel failure. Ambe et al. did not find a relation between intraoperative serum lactate and the extent of intestinal necrosis but the group of patients with the highest serum lactate were those with multiorgan failure . Serum lactate is therefore a marker of severity, signifying the presence of systemic complications in a population of AMI patients .
This multicenter and matched study has several limitations. As data were collected retrospectively, the causes of death were not determined. Specifically, we did not collect data on the “do not resuscitate” orders, which were probably frequent in this population. The type of AMI was poorly reported, although this feature may be of interest in terms of outcomes. Specifically, we did not distinguish AMI due to arterial or venous occlusions and NOMI as clear guidelines defining NOMI were not available at the time of data collection, . We did not collect data on several confounding factors such as history of hypertension, stroke, and coronary artery disease and our data cannot discriminate if atrial fibrillation was an acute event or a chronic disease. We did not collect on other relevant variables such as blood pressure or duration of hypotension although they are included to some degree in the SOFA score. Finally, we did not determine the delay between AMI diagnosis and surgery although this delay is probably one of the critical endpoints in patients with AMI as it was reported elsewhere as an independent risk factor of death [6, 34].