(1) Antibiotics are an important weapon in the fight against serious bacterial infections and are considered a common cause of drug-induced liver injury (DILI). The hepatotoxicity of many drugs, including antibiotics, is poorly analyzed in human in vitro models. (2) A standardized assay with a human hepatoma cell line was used to test the hepatotoxicity of various concentrations (Cmax, 5× Cmax, and 10× Cmax) of antibiotics. In an ICU, the most frequently prescribed antibiotics, ampicillin, cefepime, cefuroxime, levofloxacin, linezolid, meropenem, rifampicin, tigecycline, and vancomycin, were incubated with HepG2/C3A cells for 6 days. Cell viability (XTT assay, LDH release, and vitality), albumin synthesis, and cytochrome 1A2 activity were determined in cells. (3) In vitro, vancomycin, rifampicin, and tigecycline showed moderate hepatotoxic potential. The antibiotics ampicillin, cefepime, cefuroxime, levofloxacin, linezolid, and meropenem were associated with mild hepatotoxic reactions in test cells incubated with the testes Cmax concentration. Rifampicin and cefuroxime showed significantly negative effects on the viability of test cells. (4) Further in vitro studies and global pharmacovigilance reports should be conducted to reveal underlying mechanism of the hepatotoxic action of vancomycin, rifampicin, tigecycline, and cefuroxime, as well as the clinical relevance of these findings.

A hepatocyte cell line was used to determine the hepatotoxicity of sedatives and opioids, as the hepatotoxicity of these drugs has not yet been well characterized. This might pose a threat, especially to critically ill patients, as they often receive high cumulative doses for daily analgosedation and often already have impaired liver function due to an underlying disease or complications during treatment. A well-established biosensor based on HepG2/C3A cells was used for the determination of the hepatotoxicity of commonly used sedatives and opioids in the intensive care setting (midazolam, propofol, s-ketamin, thiopental, fentanyl, remifentanil, and sufentanil). The incubation time was 2 × 3 days with clinically relevant (Cmax) and higher concentrations (C5× and C10×) of each drug in cell culture medium or human plasma. Afterward, we measured the cell count, vitality, lactate dehydrogenase (LDH), mitochondrial dehydrogenase activity, cytochrome P 450 1A2 (CYP1A2), and albumin synthesis. All tested substances reduced the viability of hepatocyte cells, but sufentanil and remifentanil showed more pronounced effects. The cell count was diminished by sufentanil in both the medium and plasma and by remifentanil only in plasma. Sufentanil and remifentanil also led to higher values of LDH in the cell culture supernatant. A reduction of mitochondrial dehydrogenase activity was seen with the use of midazolam and s-ketamine. Microalbumin synthesis was reduced in plasma after its incubation with higher concentrations of sufentanil and remifentanil. Remifentanil and s-ketamine reduced CYP1A2 activity, while propofol and thiopental increased it. Our findings suggest that none of the tested sedatives and opioids have pronounced hepatotoxicity. Sufentanil, remifentanil, and s-ketamine showed moderate hepatotoxic effects in vitro. These drugs should be given with caution to patients vulnerable to hepatotoxic drugs, e.g., patients with pre-existing liver disease or liver impairment as part of their underlying disease (e.g., hypoxic hepatitis or cholestatic liver dysfunction in sepsis). Further studies are indicated for this topic, which may use more complex cell culture models and global pharmacovigilance reports, addressing the limitation of the used cell model: HepG2/C3A cells have a lower metabolic capacity due to their low levels of CYP enzymes compared to primary hepatocytes. However, while the test model is suitable for parental substances, it is not for toxicity testing of metabolites.

Immunoadsorption (IA) has proven to be clinically effective in the treatment of steroid-refractory multiple sclerosis (MS) relapses, but its mechanism of action remains unclear. We used miniaturized adsorber devices with a tryptophan-immobilized polyvinyl alcohol (PVA) gel sorbent to mimic the IA treatment of patients with MS in vitro. The plasma was screened before and after adsorption with regard to disease-specific mediators, and the effect of the IA treatment on the migration of neutrophils and the integrity of the endothelial cell barrier was tested in cell-based models. The in vitro IA treatment with miniaturized adsorbers resulted in reduced plasma levels of cytokines and chemokines. We also found a reduced migration of neutrophils towards patient plasma treated with the adsorbers. Furthermore, the IA-treated plasma had a positive effect on the endothelial cell barrier’s integrity in the cell culture model. Our findings suggest that IA results in a reduced infiltration of cells into the central nervous system by reducing leukocyte transmigration and preventing blood–brain barrier breakdown. This novel approach of performing in vitro blood purification therapies on actual patient samples with miniaturized adsorbers and testing their effects in cell-based assays that investigate specific hypotheses of the pathophysiology provides a promising platform for elucidating the mechanisms of action of those therapies in various diseases.

HiPSC-derived blood–brain barrier (BBB) models are valuable for pharmacological and physiological studies, yet their translational potential is limited due to insufficient cell phenotypes and the neglection of the complex environment of the BBB. This study evaluates the plasma compatibility with hiPSC-derived microvascular endothelial-like cells to enhance the translational potential of in vitro BBB models. Therefore, plasma samples (sodium/lithium heparin, citrate, EDTA) and serum from healthy donors were tested on hiPSC-derived microvascular endothelial-like cells at concentrations of 100%, 75%, and 50%. After 24 h, cell viability parameters were assessed. The impact of heparin-anticoagulated plasmas was further evaluated regarding barrier function and endothelial phenotype of differentiated endothelial-like cells. Finally, sodium-heparin plasma was tested in an isogenic triple-culture BBB model with continuous TEER measurements for 72 h. Only the application of heparin-anticoagulated plasmas did not significantly alter viability parameters compared to medium. Furthermore, heparin plasmas improved barrier function without increasing cell density and induced a von Willebrand factor signal. Finally, continuous TEER measurements of the triple-culture model confirmed the positive impact of sodium-heparin plasma on barrier function. Consequently, heparin-anticoagulated plasmas were proven to be compatible with hiPSC-derived microvascular endothelial-like cells. Thereby, the translational potential of BBB models can be substantially improved in the future.