Integrated in Silico Assessment of Environmental and Human Toxicity of Carbamazepine as a Pseudo-persistent Emerging Contaminant

Introduction: Carbamazepine is one of the most frequently detected pharmaceutical residues in aquatic environments due to its extensive use, incomplete removal in wastewater treatment plants, and resistance to biodegradation. Therefore, it is considered a pseudo-persistent emerging contaminant. This study aimed to provide an integrated in silico evaluation of its environmental persistence, bioaccumulation, and potential human toxicity.

Methods: Carbamazepine and its metabolites/degradation products (n = 16 compounds) were evaluated using VEGA QSAR (v1.2.4) and EPA TEST (v5.1.2). Environmental persistence was assessed through water and sediment persistence and ready biodegradability models. Bioaccumulation potential was evaluated using bioconcentration factor (BCF), logKow, and logKoc. Toxicity assessment focused on mutagenicity, developmental toxicity, carcinogenicity, hepatotoxicity, and endocrine-disrupting activity involving estrogen and androgen receptor pathways.

Results: Persistence predictions generally showed low reliability, limiting confidence in water- and sediment-persistence estimates. However, 9,10-dihydro-9-oxoacridine was identified as a reliably non-biodegradable compound with good reliability (GR), whereas iminodibenzyl showed moderate reliability. Hydrophobic derivatives, particularly acridine and iminodibenzyl, demonstrated the highest bioaccumulation potential, and in silico physicochemical predictions were consistent with experimental LogP and solubility data. Most compounds were predicted to be non-mutagenic; however, acridine exhibited the strongest genotoxic profile, supported by experimental evidence. Additionally, 10,11-epoxycarbamazepine, 2-hydroxycarbamazepine, and 3-hydroxycarbamazepine showed predicted mutagenic potential. Almost all investigated derivatives triggered alerts in the carcinogenicity and developmental toxicity models. Ten compounds, including carbamazepine, oxcarbazepine, eslicarbazepine, licarbazepine, and 10,11-epoxycarbamazepine, were identified as developmental toxicants with GR. Hepatotoxicity was experimentally confirmed for carbamazepine and oxcarbazepine and predicted for most derivatives. Endocrine-disrupting activity was generally absent; however, 2- and 3-hydroxy-carbamazepine exhibited possible estrogen receptor interaction signals with GR.

Conclusions: Carbamazepine and several transformation products may pose human health concerns, particularly through persistence, genotoxicity, and developmental toxicity-related signals. These findings demonstrate the utility of integrated in silico approaches for prioritizing pharmaceutical contaminants and their metabolites.