Application of machine learning to predict reduction in total PANSS score and enrich enrollment in schizophrenia clinical trials.

Clinical trial efficiency, defined as facilitating patient enrollment, and reducing the time to reach safety and efficacy decision points, is a critical driving factor for making improvements in therapeutic development. The present work evaluated a machine learning (ML) approach to improve phase II or proof-of-concept trials designed to address unmet medical needs in treating schizophrenia. Diagnostic data from the Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE) trial were used to develop a binary classification ML model predicting individual patient response as either "improvement," defined as greater than 20% reduction in total Positive and Negative Syndrome Scale (PANSS) score, or "no improvement," defined as an inadequate treatment response (<20% reduction in total PANSS). A random forest algorithm performed best relative to other tree-based approaches in model ability to classify patients after 6 months of treatment. Although model ability to identify true positives, a measure of model sensitivity, was poor (<0.2), its specificity, true negative rate, was high (0.948). A second model, adapted from the first, was subsequently applied as a proof-of-concept for the ML approach to supplement trial enrollment by identifying patients not expected to improve based on their baseline diagnostic scores. In three virtual trials applying this screening approach, the percentage of patients predicted to improve ranged from 46% to 48%, consistently approximately double the CATIE response rate of 22%. These results show the promising application of ML to improve clinical trial efficiency and, as such, ML models merit further consideration and development.

Inhibitory Effects of Probenecid on Pharmacokinetics of Tenofovir Disoproxil Fumarate and Emtricitabine for On-Demand HIV Preexposure Prophylaxis.

In a randomized, crossover pharmacokinetic study in healthy volunteers (N = 14), a single dose of 2 g probenecid (PRO)-boosted 600 mg tenofovir disoproxil fumarate (TDF)/400 mg emtricitabine (FTC) (test (T) +PRO) was compared with the current on-demand HIV preexposure prophylaxis from the IPERGAY study (a 600 mg TDF/400 mg FTC on day 1 and 300 mg TDF/200 mg FTC on days 2 and 3) (control, C IPERGAY). PRO increased mean single-dose area under the plasma concentration-time curve extrapolated to infinity (AUC0-∞,SD ) of tenofovir (TFV) and FTC by 61% and 68%, respectively. The TFV-diphosphate (TFV-DP) concentrations in peripheral blood mononuclear cells were higher (~30%) at 24 hours in T +PRO but then fell significantly lower (~40%) at 72 hours compared with C IPERGAY. The interaction between FTC and PRO was unexpected and novel. Further study is needed to determine if this PRO-boosted TDF/FTC regimen would be clinically effective.

Prediction of Nephropathy in Type 2 Diabetes: An Analysis of the ACCORD Trial Applying Machine Learning Techniques.

Applying data mining and machine learning (ML) techniques to clinical data might identify predictive biomarkers for diabetic nephropathy (DN), a common complication of type 2 diabetes mellitus (T2DM). A retrospective analysis of the Action to Control Cardiovascular Risk in Diabetes (ACCORD) trial was intended to identify such factors using ML. The longitudinal data were stratified by time after patient enrollment to differentiate early and late predictors. Our results showed that Random Forest and Simple Logistic Regression methods exhibited the best performance among the evaluated algorithms. Baseline values for glomerular filtration rate (GFR), urinary creatinine, urinary albumin, potassium, cholesterol, low-density lipoprotein, and urinary albumin to creatinine ratio were identified as DN predictors. Early predictors were the baseline values of GFR, systolic blood pressure, as well as fasting plasma glucose (FPG) and potassium at month 4. Changes per year in GFR, FPG, and triglycerides were recognized as predictors of late development. In conclusion, ML-based methods successfully identified predictive factors for DN among patients with T2DM.

Pharmacokinetics of olanzapine long-acting injection: the clinical perspective.

Olanzapine long-acting injection (OLAI) is a sustained-release depot antipsychotic for the treatment of schizophrenia in adults. Our objective was to explain the pharmacokinetics of OLAI to provide clinical insight. Simulation models and data from clinical trials are presented. Olanzapine concentrations were observed immediately upon injection. Half-life was ∼30 days, controlled by the slow rate of intramuscular absorption rather than the 30-h elimination rate-based half-life of oral olanzapine. As each injection builds on the drug still being released from previous injections, concentrations increase gradually until a steady state is reached after ∼3 months. Concentrations were similar to oral olanzapine and proportional to the dose; the average steady-state concentrations (10th-90th percentile) for the 150, 210, and 300 mg/2-week doses were 16-32, 15-55, and 20-67 ng/ml, respectively, and those for the 300 and 405 mg/4-week doses were 19-48 and 19-62 ng/ml, respectively. Peak concentrations most often occurred at 2-4 days after injection. Peak-to-trough fluctuation was greater for the 4-week dosing interval than the 2-week one, with no apparent clinical ramifications for these differences. Trough concentrations were above the lower end of the therapeutic range, even at the first injection. Long-term use up to 6 years indicated no additional accumulation. The impact of smoking and sex was similar, but less pronounced than for oral olanzapine.

Post-injection delirium/sedation syndrome in patients with schizophrenia treated with olanzapine long-acting injection, II: investigations of mechanism.

BACKGROUND: Olanzapine long-acting injection (LAI) is a salt-based depot antipsychotic combining olanzapine and pamoic acid. The slow intramuscular dissolution of this practically insoluble salt produces an extended release of olanzapine lasting up to 4 weeks. However, in a small number of injections (< 0.1%), patients experienced symptoms suggestive of olanzapine overdose, a phenomenon that has been termed "post-injection delirium/sedation syndrome" (PDSS). The authors conducted a series of parallel investigations into the possible reasons PDSS events occur. METHODS: Healthcare providers involved in the PDSS cases were queried for clinical information around the events. Plasma samples from patients experiencing PDSS were collected when possible (12/30 cases) and olanzapine concentrations compared with the known pharmacokinetic profile for olanzapine LAI. Product batches and used vials from the PDSS cases were evaluated for compliance with established manufacturing standards and/or possible user error. Because this depot formulation depends upon slow dissolution at the intramuscular injection site, in-vitro experiments were conducted to assess solubility of olanzapine pamoate in various media. RESULTS: Injection administrators reported no unusual occurrences during the injection. No anomalies were found with the product batches or the remaining suspension in the used vials. Olanzapine concentrations during PDSS events were higher than the expected 5-73 ng/mL range, with concentrations exceeding 100 ng/mL and in some cases reaching >600 ng/mL during the first hours after injection but then returning to the expected therapeutic range within 24 to 72 hours. Solubility and dissolution rate of olanzapine pamoate were also found to be substantially greater in plasma than in other media such as those approximating the environment in muscle tissue. CONCLUSIONS: Manufacturing irregularities, improper drug reconstitution, and inappropriate dosing were ruled out as possible causes of PDSS. In-vitro solubility and in-vivo pharmacokinetic investigations suggest that PDSS is related to exposure of the injected product to a substantial volume of blood. This exposure is most likely the result of unintended partial intravascular injection or blood vessel injury during the injection (occurring even with proper injection technique) with subsequent seepage of the medication into the vasculature, which would produce higher than intended olanzapine concentrations and symptoms consistent with PDSS. TRIAL REGISTRATION: ClinicalTrials.gov ID; URL: http://http//www.clinicaltrials.gov/: NCT00094640, NCT00088478, NCT00088491, NCT00088465, and NCT00320489.

Oral olanzapine disposition in adolescents with schizophrenia or bipolar I disorder: a population pharmacokinetic model.

BACKGROUND: Olanzapine is an atypical antipsychotic approved for the treatment of adults and adolescents (aged 13 -17 years) with schizophrenia or bipolar I disorder (manic or mixed episodes). OBJECTIVES: To characterize the pharmacokinetics of olanzapine in adolescents, to estimate the sources of variability, and to identify significant co-variates. In addition, olanzapine pharmacokinetic parameters in adolescents were compared with those in adults to guide appropriate dosing recommendations for adolescent patients. METHODS: A population pharmacokinetic modeling study was performed. The majority of pharmacokinetic data for the model came from a multicenter, open-label study in which 4.5 weeks of oral olanzapine 2.5-20 mg once daily was administered to a total of 105 patients aged 13-17 years (41.1-148 kg) who had a diagnosis of schizophrenia or bipolar I disorder. Four blood samples at steady state were obtained from each patient. Olanzapine concentrations in plasma were determined using a validated high-performance liquid chromatography method with electrochemical detection. Similar data from 11 adolescents from three previous studies were also included. A pharmacokinetic model was developed and the potential effects of patient characteristics (sex, bodyweight, age, ethnic origin) were investigated using a nonlinear mixed effects modeling program. The distributions of pharmacokinetic parameters for olanzapine in adolescents were compared with those previously reported in adults (n = 912, diagnosis of schizophrenia, olanzapine 5-20 mg/day) using the Kolmogorov-Smirnov 2-sample test. A visual predictive check was performed using Monte Carlo simulations on an external validation dataset. RESULTS: The pharmacokinetics of oral olanzapine in adolescent patients were described by a one-compartment pharmacokinetic model. The typical model estimates were 13.6 L/h (70 kg female patient) for oral clearance (CL/F) and 899 L for oral volume of distribution (V/F). Interpatient variability (40.5% for CL/F, 65.4% for V/F) and residual error (27%) were moderate. Bodyweight and sex had a significant influence on CL/F, which was lower in patients with lower weights and approximately 30% higher in males than females. Olanzapine exposure was typically 27% higher in adolescents versus adults. Approximately 77% of adolescents and adults had comparable CL/F values and 69% had comparable V/F values. CONCLUSIONS: The pharmacokinetics of oral olanzapine in adolescent patients are similar to those in adults, and are linear in the dosage range of 2.5-20 mg/day. Given the small magnitude of co-variate effects and the interpatient variability, dose adjustments based on bodyweight or sex are not necessary in adolescents.

Olanzapine long-acting injection: a 24-week, randomized, double-blind trial of maintenance treatment in patients with schizophrenia.

OBJECTIVE: The purpose of the present study was to evaluate the efficacy and tolerability of olanzapine long-acting injection for maintenance treatment of schizophrenia. METHOD: Outpatients with schizophrenia who had maintained stability on an oral regimen of olanzapine (10, 15, or 20 mg/day) for 4 to 8 weeks were randomly assigned to 24 weeks of double-blind treatment with "low" (150 mg every 2 weeks; N=140), "medium" (405 mg every 4 weeks; N=318), or "high" (300 mg every 2 weeks; N=141) doses of olanzapine long-acting injection; a very low reference dose of olanzapine long-acting injection (45 mg every 4 weeks; N=144); or their stabilized dose of oral olanzapine (N=322). Rates of and time to psychotic exacerbation were estimated using Kaplan-Meier methodology. RESULTS: At 24 weeks, the majority of oral olanzapine-treated patients (93%), as well as most olanzapine long-acting injection-treated patients receiving high (95%), medium (90%), low (84%), and very low doses (69%), remained exacerbation free, with the therapeutic 4-week regimen (medium dose) and pooled 2-week regimen (low and high doses) demonstrating efficacy similar to that of oral olanzapine as well as to each other. The three standard long-acting doses were superior to the very low reference dose based on time to exacerbation. Incidence of weight gain > or = 7% of baseline was 21% for oral olanza-pine compared with 21%, 15%, 16%, and 8% for the high, medium, low, and very low olanzapine long-acting treatment groups, respectively. No clinically significant differences were observed between the long-acting injection and oral olanzapine groups in general safety parameters. Few injection-site reactions occurred (3%). Two patients experienced sedation and delirium consistent with olanzapine overdose following possible accidental intravascular injection. CONCLUSIONS: Olanzapine long-acting injection was efficacious in maintenance treatment of schizophrenia for up to 24 weeks, with a safety profile similar to oral olanzapine except for injection-related adverse events.

Olanzapine plasma concentrations after treatment with 10, 20, and 40 mg/d in patients with schizophrenia: an analysis of correlations with efficacy, weight gain, and prolactin concentration.

Objectives of the study were to evaluate the relationship between olanzapine plasma concentrations and efficacy, prolactin, and weight and to assess effects of smoking, sex, and race on the pharmacokinetic characteristics of oral olanzapine up to 40 mg/d. Patients were randomly allocated to olanzapine 10, 20, or 40 mg/d for 8 weeks. Olanzapine concentrations in 634 samples from 380 patients were analyzed. Mean sample collection time was approximately 15 hours after dose for all groups. Mean olanzapine concentrations were 19.7 +/- 11.4, 37.9 +/- 22.8, and 74.5 +/- 43.7 ng/mL for 10-, 20-, and 40-mg doses, respectively. Olanzapine concentration and Positive and Negative Syndrome Scale improvement were not significantly correlated. Change in both weight and prolactin showed significant dose response. Prolactin concentration was correlated with olanzapine concentration (r = 0.46, P < 0.001). No significant correlation between olanzapine concentration and weight change was observed. Olanzapine concentrations were lower in self-reported smokers (16.5 +/- 9.6, 34.2 +/- 20.8, and 60.9 +/- 34.6 ng/mL) than in self-reported nonsmokers (25.6 +/- 12.3, 43.4 +/- 24.7, and 113.2 +/- 44.0 ng/mL) for 10-, 20-, and 40-mg doses, respectively (P

In vitro and in vivo evaluations of cytochrome P450 1A2 interactions with duloxetine.

OBJECTIVE: To determine whether duloxetine is a substrate, inhibitor or inducer of cytochrome P450 (CYP) 1A2 enzyme, using in vitro and in vivo studies in humans. METHODS: Human liver microsomes or cells with expressed CYP enzymes and specific CYP inhibitors were used to identify which CYP enzymes catalyse the initial oxidation steps in the metabolism of duloxetine. The potential of duloxetine to inhibit CYP1A2 activity was determined using incubations with human liver microsomes and phenacetin, the CYP1A2 substrate. The potential for duloxetine to induce CYP1A2 activity was determined using human primary hepatocytes treated with duloxetine for 72 hours. Studies in humans were conducted using fluvoxamine, a potent CYP1A2 inhibitor, and theophylline, a CYP1A2 substrate, as probes. The subjects were healthy men and women aged 18-65 years. Single-dose duloxetine was administered either intravenously as a 10-mg infusion over 30 minutes or orally as a 60-mg dose in the presence or absence of steady-state fluvoxamine (100 mg orally once daily). Single-dose theophylline was given as 30-minute intravenous infusions of aminophylline 250 mg in the presence or absence of steady-state duloxetine (60 mg orally twice daily). Plasma concentrations of duloxetine, its metabolites and theophylline were determined using liquid chromatography with tandem mass spectrometry. Pharmacokinetic parameters were estimated using noncompartmental methods and evaluated using mixed-effects ANOVA. Safety measurements included vital signs, clinical laboratory tests, a physical examination, ECG readings and adverse event reports. RESULTS: The in vitro results indicated that duloxetine is metabolized by CYP1A2; however, duloxetine was predicted not to be an inhibitor or inducer of CYP1A2 in humans. Following oral administration in the presence of fluvoxamine, the duloxetine area under the plasma concentration-time curve from time zero to infinity (AUC(infinity)) and the maximum plasma drug concentration (C(max)) significantly increased by 460% (90% CI 359, 584) and 141% (90% CI 93, 200), respectively. In the presence of fluvoxamine, the oral bioavailability of duloxetine increased from 42.8% to 81.9%. In the presence of duloxetine, the theophylline AUC(infinity) and C(max) increased by only 13% (90% CI 7, 18) and 7% (90% CI 2, 14), respectively. Coadministration of duloxetine with fluvoxamine or theophylline did not result in any clinically important safety concerns, and these combinations were generally well tolerated. CONCLUSION: Duloxetine is metabolized primarily by CYP1A2; therefore, coadministration of duloxetine with potent CYP1A2 inhibitors should be avoided. Duloxetine does not seem to be a clinically significant inhibitor or inducer of CYP1A2; therefore, dose adjustment of CYP1A2 substrates may not be necessary when they are coadministered with duloxetine.

Pharmacokinetics of duloxetine in breast milk and plasma of healthy postpartum women.

OBJECTIVE: The purpose of this study was to characterize duloxetine pharmacokinetics in the breast milk and plasma of lactating women and to estimate the duloxetine dose that an infant might consume if breastfed. METHODS: This open-label study included six healthy women aged 22-35 years who stopped nursing during and after the study. Duloxetine 40 mg was given orally every 12 hours for 3.5 days; seven plasma and milk samples over 12 hours were obtained after the seventh dose. Plasma and milk samples were analysed using validated liquid chromatography-tandem mass spectrometry methods. Safety measures included vital signs, ECGs, laboratory tests, adverse event monitoring and depression rating scales. RESULTS: The mean steady-state milk-to-plasma duloxetine exposure ratio was 0.25 (90% CI 0.18, 0.35). The amount of duloxetine in the breast milk was 7 microg/day (range 4-15 microg/day). The estimated infant dose was 2 microg/kg/day (range 0.6-3 microg/kg/day), which is 0.14% of the maternal dose. Dizziness, nausea and fatigue were commonly reported adverse events. No clinically important changes in safety measures occurred. CONCLUSION: Duloxetine is detected in breast milk, and steady-state concentrations in breast milk are about one-fourth of those in maternal plasma. As the safety of duloxetine in infants is unknown, prescribers should carefully assess, on an individual basis, the potential risks of duloxetine exposure to infants and the benefits of nursing an infant when the mother is on duloxetine therapy.

Olanzapine pharmacokinetics are similar in Chinese and Caucasian subjects.

AIM: To compare the pharmacokinetic profiles and dose proportionality of olanzapine in Chinese and Caucasian subjects. METHODS: Randomized, three-period study with 12 Chinese and 12 Caucasian, healthy, male subjects administered 2.5, 5 and 10 mg olanzapine. Noncompartmental pharmacokinetic parameters were derived. RESULTS: No statistically significant racial differences in the weight-normalized pharmacokinetic parameters were observed except for Vz/Fnorm, which was 17% lower at the 5- and 10-mg dose in the Chinese group (95% confidence interval 8.49, 10.1 and 8.05, 9.73, respectively), compared with the Caucasian group (9.53, 12.8 and 9.39, 12.0, respectively). Olanzapine's pharmacokinetics were linear and dose proportional in both racial groups. CONCLUSION: The pharmacokinetics of olanzapine are similar in both Chinese and Caucasian racial groups.