Dexmedetomidine is an α2-adrenoceptor agonist with sedative, anxiolytic, sympatholytic, and analgesic-sparing effects, and minimal depression of respiratory function. It is potent and highly selective for α2-receptors with an α2:α1 ratio of 1620:1. Hemodynamic effects, which include transient hypertension, bradycardia, and hypotension, result from the drug’s peripheral vasoconstrictive and sympatholytic properties. Dexmedetomidine exerts its hypnotic action through activation of central pre- and postsynaptic α2-receptors in the locus coeruleus, thereby inducting a state of unconsciousness similar to natural sleep, with the unique aspect that patients remain easily rousable and cooperative. Dexmedetomidine is rapidly distributed and is mainly hepatically metabolized into inactive metabolites by glucuronidation and hydroxylation. A high inter-individual variability in dexmedetomidine pharmacokinetics has been described, especially in the intensive care unit population. In recent years, multiple pharmacokinetic non-compartmental analyses as well as population pharmacokinetic studies have been performed. Body size, hepatic impairment, and presumably plasma albumin and cardiac output have a significant impact on dexmedetomidine pharmacokinetics. Results regarding other covariates remain inconclusive and warrant further research. Although initially approved for intravenous use for up to 24 h in the adult intensive care unit population only, applications of dexmedetomidine in clinical practice have been widened over the past few years. Procedural sedation with dexmedetomidine was additionally approved by the US Food and Drug Administration in 2003 and dexmedetomidine has appeared useful in multiple off-label applications such as pediatric sedation, intranasal or buccal administration, and use as an adjuvant to local analgesia techniques.

Background: Dexmedetomidine is a sedative with modest analgesic efficacy, whereas remifentanil is an opioid analgesic with modest sedative potency. Synergy is often observed when sedative-hypnotics are combined with opioid analgesics in anesthetic practice. A three-phase crossover trial was conducted to study the pharmacodynamic interaction between remifentanil and dexmedetomidine. Methods: After institutional review board approval, 30 age-and sex-stratified healthy volunteers were studied. The subjects received consecutive stepwise increasing target-controlled infusions of dexmedetomidine, remifentanil, and remifentanil with a fixed dexmedetomidine background concentration. Drug effects were measured using binary (yes or no) endpoints: no response to calling the subject by name, tolerance of shaking the patient while shouting the name ("shake and shout"), tolerance of deep trapezius squeeze, and tolerance of laryngoscopy. The drug effect was measured using the electroencephalogram-derived "Patient State Index." Pharmacokinetic-pharmacodynamic modeling related the administered dexmedetomidine and remifentanil concentration to these observed effects. results: The binary endpoints were correlated with dexmedetomidine concentrations, with increasing concentrations required for increasing stimulus intensity. Estimated model parameters for the dexmedetomidine EC50 were 2.1 [90% CI, 1.6 to 2.8], 9.2 [6.8 to 13], 24 [16 to 35], and 35 [23 to 56] ng/ml, respectively. Age was inversely correlated with dexmedetomidine EC50 for all four stimuli. Adding remifentanil did not increase the probability of tolerance of any of the stimuli. The cerebral drug effect as measured by the Patient State Index was best described by the Hierarchical interaction model with an estimated dexmedetomidine EC 50 of 0.49 [0.20 to 0.99] ng/ml and remifentanil EC 50 of 1.6 [0.87 to 2.7] ng/ml. conclusions: Low dexmedetomidine concentrations (EC 50 of 0.49 ng/ml) are required to induce sedation as measured by the Patient State Index. Sensitivity to dexmedetomidine increases with age. Despite falling asleep, the majority of subjects remained arousable by calling the subject's name, "shake and shout," or a trapezius squeeze, even when reaching supraclinical concentrations. Adding remifentanil does not alter the likelihood of response to graded stimuli.

Study objectives Dexmedetomidine induced electroencephalogram (EEG) patterns during deep sedation is comparable with natural sleep patterns. Using large scale EEG recordings and machine learning techniques, we investigated whether dexmedetomidine induced deep sedation indeed mimics natural sleep patterns. Methods We used EEG recordings from three sources in this study: 8707 overnight sleep EEG and 30 dexmedetomidine clinical trial EEG. Dexmedetomidine induced sedation levels were assessed using the Modified Observer’s Assessment of Alertness/ Sedation (MOAA/S) score. We extracted twenty-two spectral features from each EEG recording using a multitaper spectral estimation method. Elastic-net regularization method was used for feature selection. We compared the performance of several machine learning algorithms (logistic regression, support vector machine and random forest), trained on individual sleep stages, to predict different levels of the MOAA/S sedation state. Results The random forest algorithm trained on non-rapid eye movement stage 3 (N3) predicted dexmedetomidine induced deep sedation (MOAA/S = 0) with AUC > 0.8 outperforming other machine learning models. Power in the delta band (0-4Hz) was selected as an important feature for prediction in addition to power in theta (4-8 Hz) and beta (16-30Hz) bands. Conclusions Using a large scale EEG data-driven approach and machine learning framework, we show that dexmedetomidine induced deep sedation state mimics N3 sleep EEG patterns.

Background: Sedation indicators based on a single quantitative EEG (QEEG) feature have been criticised for their limited performance. We hypothesised that integration of multiple QEEG features into a single sedation-level estimator using a machine learning algorithm could reliably predict levels of sedation, independent of the sedative drug used. Methods: In total, 102 subjects receiving propofol (N¼36; 16 male/20 female), sevoflurane (N¼36; 16 male/20 female), or dexmedetomidine (N¼30; 15 male/15 female) were included in this study of healthy volunteers. Sedation level was assessed using the Modified Observer's Assessment of Alertness/Sedation (MOAA/S) score. We used 44 QEEG features estimated from the EEG data in a logistic regression algorithm, and an elastic-net regularisation method was used for feature selection. The area under the receiver operator characteristic curve (AUC) was used to assess the performance of the logistic regression model. Results: The performances obtained when the system was trained and tested as drug-dependent mode to distinguish between awake and sedated states (mean AUC [standard deviation]) were propofol¼0.97 (0.03), sevoflurane¼0.74 (0.25), and dexmedetomidine¼0.77 (0.10). The drug-independent system resulted in mean AUC¼0.83 (0.17) to discriminate between the awake and sedated states. Conclusions: The incorporation of large numbers of QEEG features and machine learning algorithms is feasible for nextgeneration monitors of sedation level. Different QEEG features were selected for propofol, sevoflurane, and dexmedetomidine groups, but the sedation-level estimator maintained a high performance for predicting MOAA/S independent of the drug used. Clinical trial registration: NCT 02043938; NCT 03143972.

Because anesthetic drugs exhibit sleep-like patterns during deep hypnosis, can we predict hypnosis level from sleep brain rhythms? • Findings: Deep learning algorithms when trained on nonrapid eye movement stage 3 sleep electroencephalogram can predict dexmedetomidine-induced deep hypnotic level. • Meaning: Anesthetic-induced hypnosis levels can be predicted using sleep electroencephalogram and artificial intelligence techniques, eliminating the need for clinical trials to develop hypnotic level monitors. BACKGROUND: Brain monitors tracking quantitative brain activities from electroencephalogram (EEG) to predict hypnotic levels have been proposed as a labor-saving alternative to behavioral assessments. Expensive clinical trials are required to validate any newly developed processed EEG monitor for every drug and combinations of drugs due to drug-specific EEG patterns. There is a need for an alternative, efficient, and economical method. METHODS: Using deep learning algorithms, we developed a novel data-repurposing framework to predict hypnotic levels from sleep brain rhythms. We used an online large sleep data set (5723 clinical EEGs) for training the deep learning algorithm and a clinical trial hypnotic data set (30 EEGs) for testing during dexmedetomidine infusion. Model performance was evaluated using accuracy and the area under the receiver operator characteristic curve (AUC). RESULTS: The deep learning model (a combination of a convolutional neural network and long short-term memory units) trained on sleep EEG predicted deep hypnotic level with an accuracy (95% confidence interval [CI]) = 81 (79.2-88.3)%, AUC (95% CI) = 0.89 (0.82-0.94) using dexmedetomidine as a prototype drug. We also demonstrate that EEG patterns during dexmedetomidine-induced deep hypnotic level are homologous to nonrapid eye movement stage 3 EEG sleep. CONCLUSIONS:We propose a novel method to develop hypnotic level monitors using large sleep EEG data, deep learning, and a data-repurposing approach, and for optimizing such a system for monitoring any given individual. We provide a novel data-repurposing framework to predict hypnosis levels using sleep EEG, eliminating the need for new clinical trials to develop hypnosis level monitors.

Background Numerous pharmacokinetic models have been published aiming at more accurate and safer dosing of dexmedetomidine. The vast majority of the developed models underpredict the measured plasma concentrations with respect to the target concentration, especially at plasma concentrations higher than those used in the original studies. The aim of this article was to develop a dexmedetomidine pharmacokinetic model in healthy adults emphasizing linear versus nonlinear kinetics. Methods The data of two previously published clinical trials with stepwise increasing dexmedetomidine target-controlled infusion were pooled to build a pharmacokinetic model using the NONMEM software package (ICON Development Solutions, USA). Data from 48 healthy subjects, included in a stratified manner, were utilized to build the model. Results A three-compartment mamillary model with nonlinear elimination from the central compartment was superior to a model assuming linear pharmacokinetics. Covariates included in the final model were age, sex, and total body weight. Cardiac output did not explain between-subject or within-subject variability in dexmedetomidine clearance. The results of a simulation study based on the final model showed that at concentrations up to 2 ng · ml–1, the predicted dexmedetomidine plasma concentrations were similar between the currently available Hannivoort model assuming linear pharmacokinetics and the nonlinear model developed in this study. At higher simulated plasma concentrations, exposure increased nonlinearly with target concentration due to the decreasing dexmedetomidine clearance with increasing plasma concentrations. Simulations also show that currently approved dosing regimens in the intensive care unit may potentially lead to higher-than-expected dexmedetomidine plasma concentrations. Conclusions This study developed a nonlinear three-compartment pharmacokinetic model that accurately described dexmedetomidine plasma concentrations. Dexmedetomidine may be safely administered up to target-controlled infusion targets under 2 ng · ml–1 using the Hannivoort model, which assumed linear pharmacokinetics. Consideration should be taken during long-term administration and during an initial loading dose when following the dosing strategies of the current guidelines. Editor’s Perspective What We Already Know about This Topic What This Article Tells Us That Is New

438 Background: The need for invasive and costly routine bladder biopsies after an induction course of intravesical Bacillus Calmette-Guérin (BCG) in patients with high-grade non-muscle invasive bladder cancer is being questioned. Previous studies focused on differences in tumor detection between cystoscopic, cytologic and histologic results, whereas the clinical impact was disregarded. The aim of this study is to determine the impact of routine bladder biopsies on treatment. Methods: This retrospective multicenter trial included 348 patients from four hospitals in the Netherlands, who had had their BCG therapy evaluated by routine bladder biopsies between 2005 and 2014. The results of urethrocystoscopy (UCS) and/or cytology were compared to pathology outcomes. The influence of malignancies missed by UCS and/or cytology on further treatment was analyzed in 348 patients and in subgroups of patients with and without carcinoma in situ (CIS). A number needed to test was calculated. Secondary outcomes were costs and complications. Results: In patients with a normal UCS, malignancies were detected in 27/228 (11,8%), 19/136 (14%) and 8/92 (8,7%) in respectively all patients, patients with CIS and patients without CIS. This resulted in a radical cystectomy in only six (2,6%) patients, of which five (3,7%) with and one (1,1%) without CIS. Thirty-eight patients with a normal UCS have to undergo routine biopsies to effect therapeutic treatment in one (Number needed to test, NNT = 38). Estimated costs of one change of treatment are $81.300 USD(€72.818). Twenty-three of 348 (6,6%) transurethral biopsies were complicated. Conclusions: Considering the small clinical impact and high costs, routine bladder biopsies to evaluate BCG-induction are not recommended. The decision to perform bladder biopsies should be based on cystoscopic and cytologic results.

Brain monitors which track quantitative electroencephalogram (EEG) signatures to monitor sedation levels are drug and patient specific. There is a need for robust sedation level monitoring systems to accurately track sedation levels across all drug classes, sex and age groups. Forty-four quantitative features estimated from a pooled dataset of 204 EEG recordings from 66 healthy adult volunteers who received either propofol, dexmedetomidine, or sevoflurane (all with and without remifentanil) were used in a machine learning based automated system to estimate the depth of sedation. Model training and evaluation were performed using leave-one-out cross validation methodology. We trained four machine learning models to predict sedation levels and evaluated the influence of remifentanil, age, and sex on the prediction performance. The area under the receiver-operator characteristic curve (AUC) was used to assess the performance of the prediction model. The ensemble tree with bagging outperformed other machine learning models and predicted sedation levels with an AUC = 0.88 (0.81–0.90). There were significant differences in the prediction probability of the automated systems when trained and tested across different age groups and sex. The performance of the EEG based sedation level prediction system is drug, sex, and age specific. Nonlinear machine-learning models using quantitative EEG features can accurately predict sedation levels. The results obtained in this study may provide a useful reference for developing next generation EEG based sedation level prediction systems using advanced machine learning algorithms.Clinical trial registration: NCT 02043938 and NCT 03143972.