General anesthesia and depth of anesthesia (DoA) evaluation methods in laboratory animals: a comprehensive review

In preclinical research, general anesthesia is essential for humane and feasible procedures but profoundly modulates autonomic, cardiovascular, neurological, and biochemical systems, risking bias in experimental outcomes. In this review our aim was to synthesize current knowledge on commonly used general anesthetics in laboratory animals, their mechanisms and protocols across species (mice, rats, rabbits, pigs), and evidence-based methods to evaluate depth of anesthesia (DoA). Injectable agents (ketamine with α 2 -agonists, barbiturates, propofol) and inhalational agents (isoflurane, sevoflurane, desflurane) act primarily via NMDA antagonism or GABA-A/glycine modulation, with distinct profiles for analgesia, hemodynamics, respiration, and recovery. Species-specific dosing and routes are summarized for small rodents, rabbits, and pigs, including practical considerations (e.g., thermoregulation, airway management, malignant hyperthermia risk). DoA assessment spans traditional reflex-based scoring and advanced monitoring. EEG, raw and processed indices (e.g., BIS), offers continuous cortical information but requires species-specific validation and cautious interpretation. Autonomic indicators (heart rate, blood pressure, respiratory patterns) and heart-rate variability provide complementary, noninvasive signals yet are confounded by drugs, ventilation, and surgical stimuli. Anesthetic choice is a major experimental variable. Reliable practice demands multimodal DoA monitoring that integrates reflexes, physiologic trends, and, where feasible, EEG, alongside rigorous, species-adapted protocols and transparent reporting. Priorities include validated EEG algorithms for nonhuman species, standardized autonomic indices, and broader adoption of awake models when compatible with scientific aims. These strategies will improve animal welfare and enhance the reproducibility and interpretability of preclinical findings.