Prognosis of Patients with Sepsis and Non-Hepatic Hyperammonemia: A Cohort Study

Hyperammonaemia: review of the pathophysiology, aetiology and investigation

Acute hyperammonaemia is a medical emergency as it can progress to cerebral oedema, seizures, coma and death. Hepatic encephalopathy secondary to cirrhotic disease or portosystemic shunting are relatively well-known causes, but non-cirrhotic aetiologies of acute hyperammonaemia are less well-known, especially in the emergency department. However, an elevated ammonia is not required to make the diagnosis of hepatic encephalopathy. Although measurement of plasma ammonia is recommended for patients with acute, unexplained, altered mental status, as early identification allows early effective management which may prevent irreversible brain damage, there is currently reduced awareness among physicians of the non-cirrhotic aetiologies of acute hyperammonaemia. Furthermore, measurement of ammonia in patients with cirrhosis has been shown to have low sensitivity and specificity, and not to have altered management in the majority of cases; thus, measurement of ammonia is currently not recommended in guidelines for management of hepatic encephalopathy. We sought to describe the pathophysiology of hyperammonaemia and review the non-cirrhotic causes. This was achieved by review of MEDLINE, PubMed and Web of Science databases to include published English literature within the last 20 years. We also present a framework for investigating the acute non-cirrhotic causes of hyperammonaemia to assist both chemical pathologists and clinicians managing these often challenging cases.

NHH promotes Sepsis-associated Encephalopathy with the expression of AQP4 in astrocytes through the gut-brain Axis

Sepsis-associated encephalopathy (SAE) is a significant cause of mortality in patients with sepsis. Despite extensive research, its exact cause remains unclear. Our previous research indicated a relationship between non-hepatic hyperammonemia (NHH) and SAE. This study aimed to investigate the relationship between NHH and SAE and the potential mechanisms causing cognitive impairment. In the in vivo experimental results, there were no significant abnormalities in the livers of mice with moderate cecal ligation and perforation (CLP); however, ammonia levels were elevated in the hippocampal tissue and serum. The ELISA study suggest that fecal microbiota transplantation in CLP mice can reduce ammonia levels. Reduction in ammonia levels improved cognitive dysfunction and neurological impairment in CLP mice through behavioral, neuroimaging, and molecular biology studies. Further studies have shown that ammonia enters the brain to regulate the expression of aquaporins-4 (AQP4) in astrocytes, which may be the mechanism underlying brain dysfunction in CLP mice. The results of the in vitro experiments showed that ammonia up-regulated AQP4 expression in astrocytes, resulting in astrocyte damage. The results of this study suggest that ammonia up-regulates astrocyte AQP4 expression through the gut-brain axis, which may be a potential mechanism for the occurrence of SAE.

Inflammatory liver diseases and susceptibility to sepsis

Patients with inflammatory liver diseases, particularly alcohol-associated liver disease and metabolic dysfunction-associated fatty liver disease (MAFLD), have higher incidence of infections and mortality rate due to sepsis. The current focus in the development of drugs for MAFLD is the resolution of non-alcoholic steatohepatitis and prevention of progression to cirrhosis. In patients with cirrhosis or alcoholic hepatitis, sepsis is a major cause of death. As the metabolic center and a key immune tissue, liver is the guardian, modifier, and target of sepsis. Septic patients with liver dysfunction have the highest mortality rate compared with other organ dysfunctions. In addition to maintaining metabolic homeostasis, the liver produces and secretes hepatokines and acute phase proteins (APPs) essential in tissue protection, immunomodulation, and coagulation. Inflammatory liver diseases cause profound metabolic disorder and impairment of energy metabolism, liver regeneration, and production/secretion of APPs and hepatokines. Herein, the author reviews the roles of (1) disorders in the metabolism of glucose, fatty acids, ketone bodies, and amino acids as well as the clearance of ammonia and lactate in the pathogenesis of inflammatory liver diseases and sepsis; (2) cytokines/chemokines in inflammatory liver diseases and sepsis; (3) APPs and hepatokines in the protection against tissue injury and infections; and (4) major nuclear receptors/signaling pathways underlying the metabolic disorders and tissue injuries as well as the major drug targets for inflammatory liver diseases and sepsis. Approaches that focus on the liver dysfunction and regeneration will not only treat inflammatory liver diseases but also prevent the development of severe infections and sepsis.

Detection of Patients at Risk of Multidrug-Resistant Enterobacteriaceae Infection Using Graph Neural Networks: A Retrospective Study

Background: While Enterobacteriaceae bacteria are commonly found in the healthy human gut, their colonization of other body parts can potentially evolve into serious infections and health threats. We investigate a graph-based machine learning model to predict risks of inpatient colonization by multidrug-resistant (MDR) Enterobacteriaceae. Methods: Colonization prediction was defined as a binary task, where the goal is to predict whether a patient is colonized by MDR Enterobacteriaceae in an undesirable body part during their hospital stay. To capture topological features, interactions among patients and healthcare workers were modeled using a graph structure, where patients are described by nodes and their interactions are described by edges. Then, a graph neural network (GNN) model was trained to learn colonization patterns from the patient network enriched with clinical and spatiotemporal features. Results: The GNN model achieves performance between 0.91 and 0.96 area under the receiver operating characteristic curve (AUROC) when trained in inductive and transductive settings, respectively, up to 8% above a logistic regression baseline (0.88). Comparing network topologies, the configuration considering ward-related edges (0.91 inductive, 0.96 transductive) outperforms the configurations considering caregiver-related edges (0.88, 0.89) and both types of edges (0.90, 0.94). For the top 3 most prevalent MDR Enterobacteriaceae, the AUROC varies from 0.94 for Citrobacter freundii up to 0.98 for Enterobacter cloacae using the best-performing GNN model. Conclusion: Topological features via graph modeling improve the performance of machine learning models for Enterobacteriaceae colonization prediction. GNNs could be used to support infection prevention and control programs to detect patients at risk of colonization by MDR Enterobacteriaceae and other bacteria families.

Relationship between Nonhepatic Serum Ammonia Levels and Sepsis-Associated Encephalopathy: A Retrospective Cohort Study

Objectives

Nonhepatic hyperammonemia often occurs in patients with sepsis. Ammonia plays an essential role in the occurrence of hepatic encephalopathy. However, the relationship between nonhepatic serum ammonia levels and sepsis-associated encephalopathy (SAE) remains unclear. Thus, we aimed to evaluate the association between serum ammonia levels and patients with SAE.

Methods

Data of critically ill adults with sepsis who were admitted to the intensive care unit were retrieved from the Medical Information Mart for Intensive Care IV (MIMIC IV) between 2008 and 2019 and retrospectively analyzed. Data of patients with sepsis patients and serum ammonia not related to acute or chronic liver disease were not included.

Results

Data from 720 patients with sepsis were included. SAE was found to have a high incidence (64.6%). After adjusting for other risk factors, a serum ammonia level of ≥45 μmol/L (odds ratio (OR): 3.508, 95% confidence interval (CI): 2.336-5.269, p < 0.001) was found to be an independent risk factor for patients with SAE; moreover, as the serum ammonia level increased, the hospital mortality of SAE gradually increased in a certain range (serum ammonia <150 μmol/L). Serum ammonia levels of ≥45 μmol/L were associated with higher Simplified Acute Physiology Score II and Sequential Organ Failure Assessment (SOFA) scores in patients with SAE. Besides, our study found that patients with SAE used opioid analgesics (OR:3.433, 95% CI: 1.360-8.669, p = 0.009) and the SOFA scores of patients with SAE (OR: 1.126, 95% CI: 1.062-1.194, p < 0.001) were significantly higher than those without SAE.

Conclusions

Nonhepatic serum ammonia levels of ≥45 μmol/L evidently increased the incidence of SAE. Serum ammonia levels should be closely monitored in patients with sepsis.

Tissue-specific reprogramming of glutamine metabolism maintains tolerance to sepsis

Reprogramming metabolism is of great therapeutic interest for reducing morbidity and mortality during sepsis-induced critical illness. Disappointing results from randomized controlled trials targeting glutamine and antioxidant metabolism in patients with sepsis have begged a deeper understanding of the tissue-specific metabolic response to sepsis. The current study sought to fill this gap. We analyzed skeletal muscle transcriptomics of critically ill patients, versus elective surgical controls, which revealed reduced expression of genes involved in mitochondrial metabolism and electron transport, with increases in glutathione cycling, glutamine, branched chain, and aromatic amino acid transport. We then performed untargeted metabolomics and 13C isotope tracing to analyze systemic and tissue specific metabolic phenotyping in a murine polymicrobial sepsis model. We found an increased number of correlations between the metabolomes of liver, kidney, and spleen, with loss of correlations between the heart and quadriceps and all other organs, pointing to a shared metabolic signature within vital abdominal organs, and unique metabolic signatures for muscles during sepsis. A lowered GSH:GSSG and elevated AMP:ATP ratio in the liver underlie the significant upregulation of isotopically labeled glutamine's contribution to TCA cycle anaplerosis and glutamine-derived glutathione biosynthesis; meanwhile, the skeletal muscle and spleen were the only organs where glutamine's contribution to the TCA cycle was significantly suppressed. These results highlight tissue-specific mitochondrial reprogramming to support liver energetic demands and antioxidant synthesis, rather than global mitochondrial dysfunction, as a metabolic consequence of sepsis.

Prognostic role of serum ammonia in patients with sepsis-associated encephalopathy without hepatic failure

Objectives

Our previous study shows that serum ammonia in sepsis patients without hepatic failure is associated with a poor prognosis. The relationship between serum ammonia level and the prognosis of sepsis-associated encephalopathy (SAE) patients without hepatic failure remains unclear. We aimed to explore the relationship between serum ammonia levels and the prognosis of patients with SAE.

Materials and methods

This study is a retrospective cohort study. We collected 465 patients with SAE admitted to the intensive care unit (ICU) from Medical Information Mart for Intensive Care IV (MIMIC IV) from 2008 to 2019. Patients with SAE were divided into a survival group (369 patients) and a non-survival group (96 patients). We used the Wilcoxon signed-rank test and the multivariate logistic regression analysis to analyze the relationship between serum ammonia levels and the prognosis of patients with SAE. R software was used to analyze the dataset.

Results

The primary outcome was the relationship between serum ammonia level and hospital mortality of SAE. The secondary outcomes were the relationship between serum ammonia level and hospital stays, simplified acute physiology score (SAPS II), Charlson, Glasgow coma scale (GCS), sequential organ failure assessment (SOFA), and lactate level of SAE. The mortality of patients with SAE was 20.6%. The serum ammonia level was not significantly associated with hospital mortality, longer hospital stays, higher SAPS II and Charlson scores, and lower GCS of patients with SAE. The serum ammonia level was associated with higher SOFA scores and lactate levels in patients with SAE. The SAPS II and Charlson scores were independent risk factors for death in patients with SAE.

Conclusion

Serum ammonia level was associated with higher SOFA scores and lactate levels in patients with SAE. In addition, the SAPS II and Charlson scores can be used to assess the prognosis of patients with SAE. Therefore, we should closely monitor serum ammonia, SAPS II, and Charlson levels in patients with SAE.

Non-hepatic Hyperammonemia: A Potential Therapeutic Target for Sepsis-associated Encephalopathy

Sepsis-Associated Encephalopathy (SAE) is a common complication in the acute phase of sepsis, and patients who develop SAE have a higher mortality rate, longer hospital stay, and worse quality of life than other sepsis patients. Although the incidence of SAE is as high as 70% in sepsis patients, no effective treatment is available for this condition. To develop an effective treatment for SAE, it is vital to explore its pathogenesis. It is known that hyperammonemia is a possible factor in the pathogenesis of hepatic encephalopathy as ammonia is a potent neurotoxin. Furthermore, our previous studies indicate that non-hepatic hyperammonemia seems to occur more often in sepsis patients; it was also found that >50% of sepsis patients with non-hepatic hyperammonemia exhibited encephalopathy and delirium. Substatistical analyses indicate that non-hepatic hyperammonemia is an independent risk factor for SAE. This study updates the definition, clinical manifestations, and diagnosis of SAE; it also investigates the possible treatment options available for non-hepatic hyperammonemia in patients with sepsis and the mechanisms by which non-hepatic hyperammonemia causes encephalopathy.

Case Report: Hyperammonemic Encephalopathy Linked to Ureaplasma spp. and/or Mycoplasma hominis Systemic Infection in Patients Treated for Leukemia, an Emergency Not to Be Missed

Background

Hyperammonemic encephalopathy caused by Ureaplasma spp. and Mycoplasma hominis infection has been reported in immunocompromised patients undergoing lung transplant, but data are scarce in patients with hematological malignancies.

Case Presentation

We describe the cases of 3 female patients aged 11–16 years old, developing initially mild neurologic symptoms, rapidly evolving to coma and associated with very high ammonia levels, while undergoing intensive treatment for acute leukemia (chemotherapy: 2 and hematopoietic stem cell transplant: 1). Brain imaging displayed cerebral edema and/or microbleeding. Electroencephalograms showed diffuse slowing patterns. One patient had moderate renal failure. Extensive liver and metabolic functions were all normal. Ureaplasma spp. and M. hominis were detected by PCR and specific culture in two patients, resulting in prompt initiation of combined antibiotics therapy by fluoroquinolones and macrolides. For these 2 patients, the improvement of the neurological status and ammonia levels were observed within 96 h, without any long-term sequelae. M. hominis was detected post-mortem in vagina, using 16S rRNA PCR for the third patient who died of cerebral edema.

Conclusion

Hyperammonemic encephalopathy linked to Ureaplasma spp. and M. hominis is a rare complication encountered in immunocompromised patients treated for acute leukemia, which can lead to death if unrecognized. Combining our experience with the few published cases (n=4), we observed a strong trend among female patients and very high levels of ammonia, consistently uncontrolled by classical measures (ammonia-scavenging agents and/or continuous kidney replacement therapy). The reversibility of the encephalopathy without sequelae is possible with prompt diagnosis and adequate combined specific antibiotherapy. Any neurological symptoms in an immunocompromised host should lead to the measurement of ammonia levels. If increased, and in the absence of an obvious cause, it should prompt to perform a search for Ureaplasma spp. and M. hominis by PCR as well as an immediate empirical initiation of combined specific antibiotherapy.

Association Between Continuous Kidney Replacement Therapy Clearance and Outcome in Pediatric Patients With Hyperammonemia Not Due to Inborn Error of Metabolism

OBJECTIVES

To describe a single-center experience of pediatric patients with hyperammonemia not due to inborn errors of metabolism and determine the association between use of continuous kidney replacement therapy (CKRT) treatment and outcomes.

DESIGN

Retrospective cohort study.

SETTING

Tertiary-care children's hospital.

PATIENTS

All children less than 21 years old admitted to the hospital with hyperammonemia defined as an elevated ammonia levels (>100 µmol/L) not due to inborn error of metabolism.

INTERVENTIONS

None.

MEASURES AND MAIN RESULTS

Of 135 children with hyperammonemia, the most common reason for admission was infection in 57 of 135 (42%), congenital heart disease in 20 of 135 (14%), and bone marrow transplantation in 10 of 135 (7%). The overall mortality was 61% (82 of 135), which increased with degree of hyperammonemia (17 of 23 [74%] in those with ammonia >250 µmol/L). After multivariable regression, hyperammonemia severity was not associated with mortality (aOR, 1.4; 95% CI, 0.92-2.1; p = 0.11). Of the 43 patients (32%) receiving CKRT, 21 were prescribed standard clearance and 22 high clearance. The most common indications for CKRT were fluid overload in 17 of 43 (42%) and acute kidney injury or uremia in 16 of 43 (37%). Mean CKRT duration was 13 days. There was no difference between standard and high clearance groups in risk of death (76% vs 86%; p = 0.39), cerebral edema on CT scan (19% vs 27%; p = 0.52), nor decrease in ammonia levels after 24 or 48 hours of CKRT ( p = 0.20, p = 0.94). Among those receiving CKRT, we failed to find an association between high clearance and decreased risk of death in multivariable analysis (aOR, 1.2; 95% CI, 0.64-2.3; p = 0.55).

CONCLUSIONS

In our single-center retrospective study, we failed to find an association between clearance on CKRT and improved survival nor decreased cerebral edema on head imaging. In fact, we failed to find an association between ammonia level and mortality, after controlling for illness severity.

Epidemiology, Clinical Presentation and Treatment of Non-Hepatic Hyperammonemia in ICU COVID-19 Patients

(1) Background: Although COVID-19 is largely a respiratory disease, it is actually a systemic disease that has a wide range of effects that are not yet fully known. The aim of this study was to determine the incidence, predictors and outcome of non-hepatic hyperammonemia (NHH) in COVID-19 in intensive care unit (ICU); (2) Methods: This is a 3-month prospective observational study in a third-level COVID-19 hospital. The authors collected demographic, clinical, severity score and outcome data. Logistic regression analyses were performed to identify predictors of NHH; (3) Results: 156 COVID-19 patients were admitted to the ICU. The incidence of NHH was 12.2% (19 patients). The univariate analysis showed that invasive mechanical ventilation had a 6.6-fold higher risk (OR 6.66, 95% CI 0.86–51.6, p = 0.039) for NHH, while in the multiple regression analysis, there was a 7-fold higher risk for NHH—but it was not statistically significant (OR 7.1, 95% CI 0.90–56.4, p = 0.062). Demographics, clinical characteristics and mortality in the ICU at 28 days did not show a significant association with NHH. (4) Conclusions: The incidence of NHH in ICU COVID-19 patients was not low. NHH did not appear to significantly increase mortality, and all patients with non-hepatic hyperammonemia were successfully treated without further complications. However, the pathogenesis of NHH in ICU patients with COVID-19 remains a topic to be explored with further research.