Deep learning-enabled natural language processing to identify directional pharmacokinetic drug-drug interactions.

BACKGROUND: During drug development, it is essential to gather information about the change of clinical exposure of a drug (object) due to the pharmacokinetic (PK) drug-drug interactions (DDIs) with another drug (precipitant). While many natural language processing (NLP) methods for DDI have been published, most were designed to evaluate if (and what kind of) DDI relationships exist in the text, without identifying the direction of DDI (object vs. precipitant drug). Here we present a method for the automatic identification of the directionality of a PK DDI from literature or drug labels. METHODS: We reannotated the Text Analysis Conference (TAC) DDI track 2019 corpus for identifying the direction of a PK DDI and evaluated the performance of a fine-tuned BioBERT model on this task by following the training and validation steps prespecified by TAC. RESULTS: This initial attempt showed the model achieved an F-score of 0.82 in identifying sentences as containing PK DDI and an F-score of 0.97 in identifying object versus precipitant drugs in those sentences. DISCUSSION AND CONCLUSION: Despite a growing list of NLP methods for DDI extraction, most of them use a common set of corpora to perform general purpose tasks (e.g., classifying a sentence into one of several fixed DDI categories). There is a lack of coordination between the drug development and biomedical informatics method development community to develop corpora and methods to perform specific tasks (e.g., extract clinical exposure changes due to PK DDI). We hope that our effort can encourage such a coordination so that more "fit for purpose" NLP methods could be developed and used to facilitate the drug development process.

Effect of Paroxetine or Quetiapine Combined With Oxycodone vs Oxycodone Alone on Ventilation During Hypercapnia: A Randomized Clinical Trial.

Importance: Opioids can cause severe respiratory depression by suppressing feedback mechanisms that increase ventilation in response to hypercapnia. Following the addition of boxed warnings to benzodiazepine and opioid products about increased respiratory depression risk with simultaneous use, the US Food and Drug Administration evaluated whether other drugs that might be used in place of benzodiazepines may cause similar effects. Objective: To study whether combining paroxetine or quetiapine with oxycodone, compared with oxycodone alone, decreases the ventilatory response to hypercapnia. Design, Setting, and Participants: Randomized, double-blind, crossover clinical trial at a clinical pharmacology unit (West Bend, Wisconsin) with 25 healthy participants from January 2021 through May 25, 2021. Interventions: Oxycodone 10 mg on days 1 and 5 and the following in a randomized order for 5 days: paroxetine 40 mg daily, quetiapine twice daily (increasing daily doses from 100 mg to 400 mg), or placebo. Main Outcomes and Measures: Ventilation at end-tidal carbon dioxide of 55 mm Hg (hypercapnic ventilation) using rebreathing methodology assessed for paroxetine or quetiapine with oxycodone, compared with placebo and oxycodone, on days 1 and 5 (primary) and for paroxetine or quetiapine alone compared with placebo on day 4 (secondary). Results: Among 25 participants (median age, 35 years [IQR, 30-40 years]; 11 female [44%]), 19 (76%) completed the trial. The mean hypercapnic ventilation was significantly decreased with paroxetine plus oxycodone vs placebo plus oxycodone on day 1 (29.2 vs 34.1 L/min; mean difference [MD], -4.9 L/min [1-sided 97.5% CI, -∞ to -0.6]; P = .01) and day 5 (25.1 vs 35.3 L/min; MD, -10.2 L/min [1-sided 97.5% CI, -∞ to -6.3]; P < .001) but was not significantly decreased with quetiapine plus oxycodone vs placebo plus oxycodone on day 1 (33.0 vs 34.1 L/min; MD, -1.2 L/min [1-sided 97.5% CI, -∞ to 2.8]; P = .28) or on day 5 (34.7 vs 35.3 L/min; MD, -0.6 L/min [1-sided 97.5% CI, -∞ to 3.2]; P = .37). As a secondary outcome, mean hypercapnic ventilation was significantly decreased on day 4 with paroxetine alone vs placebo (32.4 vs 41.7 L/min; MD, -9.3 L/min [1-sided 97.5% CI, -∞ to -3.9]; P < .001), but not with quetiapine alone vs placebo (42.8 vs 41.7 L/min; MD, 1.1 L/min [1-sided 97.5% CI, -∞ to 6.4]; P = .67). No drug-related serious adverse events were reported. Conclusions and Relevance: In this preliminary study involving healthy participants, paroxetine combined with oxycodone, compared with oxycodone alone, significantly decreased the ventilatory response to hypercapnia on days 1 and 5, whereas quetiapine combined with oxycodone did not cause such an effect. Additional investigation is needed to characterize the effects after longer-term treatment and to determine the clinical relevance of these findings. Trial Registration: ClinicalTrials.gov Identifier: NCT04310579.

Predicting potential adverse events using safety data from marketed drugs.

BACKGROUND: While clinical trials are considered the gold standard for detecting adverse events, often these trials are not sufficiently powered to detect difficult to observe adverse events. We developed a preliminary approach to predict 135 adverse events using post-market safety data from marketed drugs. Adverse event information available from FDA product labels and scientific literature for drugs that have the same activity at one or more of the same targets, structural and target similarities, and the duration of post market experience were used as features for a classifier algorithm. The proposed method was studied using 54 drugs and a probabilistic approach of performance evaluation using bootstrapping with 10,000 iterations. RESULTS: Out of 135 adverse events, 53 had high probability of having high positive predictive value. Cross validation showed that 32% of the model-predicted safety label changes occurred within four to nine years of approval (median: six years). CONCLUSIONS: This approach predicts 53 serious adverse events with high positive predictive values where well-characterized target-event relationships exist. Adverse events with well-defined target-event associations were better predicted compared to adverse events that may be idiosyncratic or related to secondary target effects that were poorly captured. Further enhancement of this model with additional features, such as target prediction and drug binding data, may increase accuracy.

Aggregation and analysis of indication-symptom relationships for drugs approved in the USA.

PURPOSE: Drug indications and disease symptoms often confound adverse event reports in real-world datasets, including electronic health records and reports in the FDA Adverse Event Reporting System (FAERS). A thorough, standardized set of indications and symptoms is needed to identify these confounders in such datasets for drug research and safety assessment. The aim of this study is to create a comprehensive list of drug-indication associations and disease-symptom associations using multiple resources, including existing databases and natural language processing. METHODS: Drug indications for drugs approved in the USA were extracted from two databases, RxNorm and Side Effect Resource (SIDER). Symptoms for these indications were extracted from MedlinePlus and using natural language processing from PubMed abstracts. RESULTS: A total of 1361 unique drugs, 1656 unique indications, and 2201 unique symptoms were extracted from a wide variety of MedDRA System Organ Classes. Text-mining precision was maximized at 0.65 by examining Term Frequency Inverse Document Frequency (TF-IDF) scores of the disease-symptom associations. CONCLUSION: The drug-indication associations and disease-symptom associations collected in this study may be useful in identifying confounders in other datasets, such as safety reports. With further refinement and additional drugs, indications, and symptoms, this dataset may become a quality resource for disease symptoms.

PathFX provides mechanistic insights into drug efficacy and safety for regulatory review and therapeutic development.

Failure to demonstrate efficacy and safety issues are important reasons that drugs do not reach the market. An incomplete understanding of how drugs exert their effects hinders regulatory and pharmaceutical industry projections of a drug's benefits and risks. Signaling pathways mediate drug response and while many signaling molecules have been characterized for their contribution to disease or their role in drug side effects, our knowledge of these pathways is incomplete. To better understand all signaling molecules involved in drug response and the phenotype associations of these molecules, we created a novel method, PathFX, a non-commercial entity, to identify these pathways and drug-related phenotypes. We benchmarked PathFX by identifying drugs' marketed disease indications and reported a sensitivity of 41%, a 2.7-fold improvement over similar approaches. We then used PathFX to strengthen signals for drug-adverse event pairs occurring in the FDA Adverse Event Reporting System (FAERS) and also identified opportunities for drug repurposing for new diseases based on interaction paths that associated a marketed drug to that disease. By discovering molecular interaction pathways, PathFX improved our understanding of drug associations to safety and efficacy phenotypes. The algorithm may provide a new means to improve regulatory and therapeutic development decisions.

Updates on the web-based VIOLIN vaccine database and analysis system.

The integrative Vaccine Investigation and Online Information Network (VIOLIN) vaccine research database and analysis system (http://www.violinet.org) curates, stores, analyses and integrates various vaccine-associated research data. Since its first publication in NAR in 2008, significant updates have been made. Starting from 211 vaccines annotated at the end of 2007, VIOLIN now includes over 3240 vaccines for 192 infectious diseases and eight noninfectious diseases (e.g. cancers and allergies). Under the umbrella of VIOLIN, >10 relatively independent programs are developed. For example, Protegen stores over 800 protective antigens experimentally proven valid for vaccine development. VirmugenDB annotated over 200 'virmugens', a term coined by us to represent those virulence factor genes that can be mutated to generate successful live attenuated vaccines. Specific patterns were identified from the genes collected in Protegen and VirmugenDB. VIOLIN also includes Vaxign, the first web-based vaccine candidate prediction program based on reverse vaccinology. VIOLIN collects and analyzes different vaccine components including vaccine adjuvants (Vaxjo) and DNA vaccine plasmids (DNAVaxDB). VIOLIN includes licensed human vaccines (Huvax) and veterinary vaccines (Vevax). The Vaccine Ontology is applied to standardize and integrate various data in VIOLIN. VIOLIN also hosts the Ontology of Vaccine Adverse Events (OVAE) that logically represents adverse events associated with licensed human vaccines.

DNAVaxDB: the first web-based DNA vaccine database and its data analysis.

Since the first DNA vaccine studies were done in the 1990s, thousands more studies have followed. Here we report the development and analysis of DNAVaxDB (http://www.violinet.org/dnavaxdb), the first publically available web-based DNA vaccine database that curates, stores, and analyzes experimentally verified DNA vaccines, DNA vaccine plasmid vectors, and protective antigens used in DNA vaccines. All data in DNAVaxDB are annotated from reliable resources, particularly peer-reviewed articles. Among over 140 DNA vaccine plasmids, some plasmids were more frequently used in one type of pathogen than others; for example, pCMVi-UB for G- bacterial DNA vaccines, and pCAGGS for viral DNA vaccines. Presently, over 400 DNA vaccines containing over 370 protective antigens from over 90 infectious and non-infectious diseases have been curated in DNAVaxDB. While extracellular and bacterial cell surface proteins and adhesin proteins were frequently used for DNA vaccine development, the majority of protective antigens used in Chlamydophila DNA vaccines are localized to the inner portion of the cell. The DNA vaccine priming, other vaccine boosting vaccination regimen has been widely used to induce protection against infection of different pathogens such as HIV. Parasitic and cancer DNA vaccines were also systematically analyzed. User-friendly web query and visualization interfaces are available in DNAVaxDB for interactive data search. To support data exchange, the information of DNA vaccines, plasmids, and protective antigens is stored in the Vaccine Ontology (VO). DNAVaxDB is targeted to become a timely and vital source of DNA vaccines and related data and facilitate advanced DNA vaccine research and development.

Algorithmic Identification of Treatment-Emergent Adverse Events From Clinical Notes Using Large Language Models: A Pilot Study in Inflammatory Bowel Disease.

Outpatient clinical notes are a rich source of information regarding drug safety. However, data in these notes are currently underutilized for pharmacovigilance due to methodological limitations in text mining. Large language models (LLMs) like Bidirectional Encoder Representations from Transformers (BERT) have shown progress in a range of natural language processing tasks but have not yet been evaluated on adverse event (AE) detection. We adapted a new clinical LLM, University of California - San Francisco (UCSF)-BERT, to identify serious AEs (SAEs) occurring after treatment with a non-steroid immunosuppressant for inflammatory bowel disease (IBD). We compared this model to other language models that have previously been applied to AE detection. We annotated 928 outpatient IBD notes corresponding to 928 individual patients with IBD for all SAE-associated hospitalizations occurring after treatment with a non-steroid immunosuppressant. These notes contained 703 SAEs in total, the most common of which was failure of intended efficacy. Out of eight candidate models, UCSF-BERT achieved the highest numerical performance on identifying drug-SAE pairs from this corpus (accuracy 88-92%, macro F1 61-68%), with 5-10% greater accuracy than previously published models. UCSF-BERT was significantly superior at identifying hospitalization events emergent to medication use (P < 0.01). LLMs like UCSF-BERT achieve numerically superior accuracy on the challenging task of SAE detection from clinical notes compared with prior methods. Future work is needed to adapt this methodology to improve model performance and evaluation using multicenter data and newer architectures like Generative pre-trained transformer (GPT). Our findings support the potential value of using large language models to enhance pharmacovigilance.

The expanding universe of NUTM1 fusions in pediatric cancer.

NUT midline carcinoma family member 1 (NUTM1) fusions were originally identified in poorly differentiated and clinically aggressive carcinomas typically located in the midline structures of children and young adults, and collectively known as NUT (midline) carcinomas. Next-generation sequencing later uncovered NUTM1 fusions in a variety of other pediatric and adult cancers of diverse location and type, including hematologic malignancies, cutaneous adnexal tumors, and sarcomas. A vast array of NUTM1 fusions with bromodomain containing 4 (BRD4) or bromodomain containing 3 (BRD3), which are characteristic of NUT carcinoma, and with several other fusion partners have been identified and associated with variable prognosis. These non-kinase fusions are thought to cause epigenetic reprogramming, thereby promoting proliferation, and hindering the differentiation of cancer cells. Many questions about both the function of the naïve NUTM1 protein, which is mostly restricted to the germ cells of the testis and is related to spermatogenesis and the oncogenic mechanisms of the various NUTM1 fusions in both adult and pediatric cancer, are still unanswered. Moreover, whether there is a relationship defined by the presence of NUTM1 fusions between conventional NUT carcinoma and other NUTM1-rearranged neoplasms remains to be elucidated. This review will focus on recent discoveries of NUTM1 fusions found in pediatric cancers, their prognostic impact, and emergence as novel oncogenic drivers.

Rewiring Drug Research and Development through Human Data-Driven Discovery (HD3).

In an era of unparalleled technical advancement, the pharmaceutical industry is struggling to transform data into increased research and development efficiency, and, as a corollary, new drugs for patients. Here, we briefly review some of the commonly discussed issues around this counterintuitive innovation crisis. Looking at both industry- and science-related factors, we posit that traditional preclinical research is front-loading the development pipeline with data and drug candidates that are unlikely to succeed in patients. Applying a first principles analysis, we highlight the critical culprits and provide suggestions as to how these issues can be rectified through the pursuit of a Human Data-driven Discovery (HD3) paradigm. Consistent with other examples of disruptive innovation, we propose that new levels of success are not dependent on new inventions, but rather on the strategic integration of existing data and technology assets. In support of these suggestions, we highlight the power of HD3, through recently published proof-of-concept applications in the areas of drug safety analysis and prediction, drug repositioning, the rational design of combination therapies and the global response to the COVID-19 pandemic. We conclude that innovators must play a key role in expediting the path to a largely human-focused, systems-based approach to drug discovery and research.

Algorithmic identification of treatment-emergent adverse events from clinical notes using large language models: a pilot study in inflammatory bowel disease.

Background and Aims: Outpatient clinical notes are a rich source of information regarding drug safety. However, data in these notes are currently underutilized for pharmacovigilance due to methodological limitations in text mining. Large language models (LLM) like BERT have shown progress in a range of natural language processing tasks but have not yet been evaluated on adverse event detection. Methods: We adapted a new clinical LLM, UCSF BERT, to identify serious adverse events (SAEs) occurring after treatment with a non-steroid immunosuppressant for inflammatory bowel disease (IBD). We compared this model to other language models that have previously been applied to AE detection. Results: We annotated 928 outpatient IBD notes corresponding to 928 individual IBD patients for all SAE-associated hospitalizations occurring after treatment with a non-steroid immunosuppressant. These notes contained 703 SAEs in total, the most common of which was failure of intended efficacy. Out of 8 candidate models, UCSF BERT achieved the highest numerical performance on identifying drug-SAE pairs from this corpus (accuracy 88-92%, macro F1 61-68%), with 5-10% greater accuracy than previously published models. UCSF BERT was significantly superior at identifying hospitalization events emergent to medication use (p < 0.01). Conclusions: LLMs like UCSF BERT achieve numerically superior accuracy on the challenging task of SAE detection from clinical notes compared to prior methods. Future work is needed to adapt this methodology to improve model performance and evaluation using multi-center data and newer architectures like GPT. Our findings support the potential value of using large language models to enhance pharmacovigilance.

Analysis of secondary pharmacology assays received by the US Food and Drug Administration.

Secondary pharmacology studies are a time-efficient and cost-effective method for determining the safety profile of a potential new drug before it enters human trials. The results of these multi-target screens are commonly submitted with Investigational New Drug (IND) applications, but there currently is little guidance on how such information is presented and which targets are chosen for testing. In this study, we expand on our previous analysis of secondary pharmacology reports by manually curating and analyzing all secondary pharmacology results received by the FDA received as part of an IND submission. A total of 1120 INDs submitted by 480 sponsors between 1999 and October 2020 were included in this study. The overall results were largely consistent with previous internal and external studies, showing that the most tested target in our set was the histamine 1 receptor (tested 938 times), the most hit target was sodium channel site 2 (hit 141 times), and the target with the highest hit percentage was the vesicular monoamine transporter 2 (hit 42.2% of the time). Additionally, this study demonstrated that improvements in the secondary pharmacology submission process, such as changes in formatting and nomenclature, could enhance the utility of these assays for regulatory review, including assisting with identifying the safety liabilities of a drug candidate early in development. This updated data set will allow FDA-industry collaborative working groups to continue developing the best methods for regulatory submission of secondary pharmacology data and evaluate the need for a standard target panel.

Ontology-Based Classification and Analysis of Adverse Events Associated With the Usage of Chloroquine and Hydroxychloroquine.

Multiple methodologies have been developed to identify and predict adverse events (AEs); however, many of these methods do not consider how patient population characteristics, such as diseases, age, and gender, affect AEs seen. In this study, we evaluated the utility of collecting and analyzing AE data related to hydroxychloroquine (HCQ) and chloroquine (CQ) from US Prescribing Information (USPIs, also called drug product labels or package inserts), the FDA Adverse Event Reporting System (FAERS), and peer-reviewed literature from PubMed/EMBASE, followed by AE classification and modeling using the Ontology of Adverse Events (OAE). Our USPI analysis showed that CQ and HCQ AE profiles were similar, although HCQ was reported to be associated with fewer types of cardiovascular, nervous system, and musculoskeletal AEs. According to EMBASE literature mining, CQ and HCQ were associated with QT prolongation (primarily when treating COVID-19), heart arrhythmias, development of Torsade des Pointes, and retinopathy (primarily when treating lupus). The FAERS data was analyzed by proportional ratio reporting, Chi-square test, and minimal case number filtering, followed by OAE classification. HCQ was associated with 63 significant AEs (including 21 cardiovascular AEs) for COVID-19 patients and 120 significant AEs (including 12 cardiovascular AEs) for lupus patients, supporting the hypothesis that the disease being treated affects the type and number of certain CQ/HCQ AEs that are manifested. Using an HCQ AE patient example reported in the literature, we also ontologically modeled how an AE occurs and what factors (e.g., age, biological sex, and medical history) are involved in the AE formation. The methodology developed in this study can be used for other drugs and indications to better identify patient populations that are particularly vulnerable to AEs.

A network paradigm predicts drug synergistic effects using downstream protein-protein interactions.

In some cases, drug combinations affect adverse outcome phenotypes by binding the same protein; however, drug-binding proteins are associated through protein-protein interaction (PPI) networks within the cell, suggesting that drug phenotypes may result from long-range network effects. We first used PPI network analysis to classify drugs based on proteins downstream of their targets and next predicted drug combination effects where drugs shared network proteins but had distinct binding proteins (e.g., targets, enzymes, or transporters). By classifying drugs using their downstream proteins, we had an 80.7% sensitivity for predicting rare drug combination effects documented in gold-standard datasets. We further measured the effect of predicted drug combinations on adverse outcome phenotypes using novel observational studies in the electronic health record. We tested predictions for 60 network-drug classes on seven adverse outcomes and measured changes in clinical outcomes for predicted combinations. These results demonstrate a novel paradigm for anticipating drug synergistic effects using proteins downstream of drug targets.

New science, drug regulation, and emergent public health issues: The work of FDA's division of applied regulatory science.

The U.S. Food and Drug Administration (FDA) Division of Applied Regulatory Science (DARS) moves new science into the drug review process and addresses emergent regulatory and public health questions for the Agency. By forming interdisciplinary teams, DARS conducts mission-critical research to provide answers to scientific questions and solutions to regulatory challenges. Staffed by experts across the translational research spectrum, DARS forms synergies by pulling together scientists and experts from diverse backgrounds to collaborate in tackling some of the most complex challenges facing FDA. This includes (but is not limited to) assessing the systemic absorption of sunscreens, evaluating whether certain drugs can convert to carcinogens in people, studying drug interactions with opioids, optimizing opioid antagonist dosing in community settings, removing barriers to biosimilar and generic drug development, and advancing therapeutic development for rare diseases. FDA tasks DARS with wide ranging issues that encompass regulatory science; DARS, in turn, helps the Agency solve these challenges. The impact of DARS research is felt by patients, the pharmaceutical industry, and fellow regulators. This article reviews applied research projects and initiatives led by DARS and conducts a deeper dive into select examples illustrating the impactful work of the Division.

Aggregation and analysis of secondary pharmacology data from investigational new drug submissions at the US Food and Drug Administration.

Secondary pharmacology studies are utilized by the pharmaceutical industry as a cost-efficient tool to identify potential safety liabilities of drugs before entering Phase 1 clinical trials. These studies are recommended by the Food and Drug Administration (FDA) as a part of the Investigational New Drug (IND) application. However, despite the utility of these assays, there is little guidance on which targets should be screened and which format should be used. Here, we evaluated 226 secondary pharmacology profiles obtained from close to 90 unique sponsors. The results indicated that the most tested target in our set was the GABA benzodiazepine receptor (tested 168 times), the most hit target was adenosine 3 (hit 24 times), and the target with the highest hit percentage was the quinone reductase 2 (NQO2) receptor (hit 29% of the time). The overall results were largely consistent with those observed in previous publications. However, this study also identified the need for improvement in the submission process of secondary pharmacology studies by industry, which could enhance their utility for regulatory purpose. FDA-industry collaborative working groups will utilize this data to determine the best methods for regulatory submission of these studies and evaluate the need for a standard target panel.

Target Adverse Event Profiles for Predictive Safety in the Postmarket Setting.

We improved a previous pharmacological target adverse-event (TAE) profile model to predict adverse events (AEs) on US Food and Drug Administration (FDA) drug labels at the time of approval. The new model uses more drugs and features for learning as well as a new algorithm. Comparator drugs sharing similar target activities to a drug of interest were evaluated by aggregating AEs from the FDA Adverse Event Reporting System (FAERS), FDA drug labels, and medical literature. An ensemble machine learning model was used to evaluate FAERS case count, disproportionality scores, percent of comparator drug labels with a specific AE, and percent of comparator drugs with the reports of the event in the literature. Overall classifier performance was F1 of 0.71, area under the precision-recall curve of 0.78, and area under the receiver operating characteristic curve of 0.87. TAE analysis continues to show promise as a method to predict adverse events at the time of approval.

VaximmutorDB: A Web-Based Vaccine Immune Factor Database and Its Application for Understanding Vaccine-Induced Immune Mechanisms.

Vaccines stimulate various immune factors critical to protective immune responses. However, a comprehensive picture of vaccine-induced immune factors and pathways have not been systematically collected and analyzed. To address this issue, we developed VaximmutorDB, a web-based database system of vaccine immune factors (abbreviated as "vaximmutors") manually curated from peer-reviewed articles. VaximmutorDB currently stores 1,740 vaccine immune factors from 13 host species (e.g., human, mouse, and pig). These vaximmutors were induced by 154 vaccines for 46 pathogens. Top 10 vaximmutors include three antibodies (IgG, IgG2a and IgG1), Th1 immune factors (IFN-γ and IL-2), Th2 immune factors (IL-4 and IL-6), TNF-α, CASP-1, and TLR8. Many enriched host processes (e.g., stimulatory C-type lectin receptor signaling pathway, SRP-dependent cotranslational protein targeting to membrane) and cellular components (e.g., extracellular exosome, nucleoplasm) by all the vaximmutors were identified. Using influenza as a model, live attenuated and killed inactivated influenza vaccines stimulate many shared pathways such as signaling of many interleukins (including IL-1, IL-4, IL-6, IL-13, IL-20, and IL-27), interferon signaling, MARK1 activation, and neutrophil degranulation. However, they also present their unique response patterns. While live attenuated influenza vaccine FluMist induced significant signal transduction responses, killed inactivated influenza vaccine Fluarix induced significant metabolism of protein responses. Two different Yellow Fever vaccine (YF-Vax) studies resulted in overlapping gene lists; however, they shared more portions of pathways than gene lists. Interestingly, live attenuated YF-Vax simulates significant metabolism of protein responses, which was similar to the pattern induced by killed inactivated Fluarix. A user-friendly web interface was generated to access, browse and search the VaximmutorDB database information. As the first web-based database of vaccine immune factors, VaximmutorDB provides systematical collection, standardization, storage, and analysis of experimentally verified vaccine immune factors, supporting better understanding of protective vaccine immunity.

Effects of sedative psychotropic drugs combined with oxycodone on respiratory depression in the rat.

Following a decision to require label warnings for concurrent use of opioids and benzodiazepines and increased risk of respiratory depression and death, the US Food and Drug Administratioin (FDA) recognized that other sedative psychotropic drugs may be substituted for benzodiazepines and be used concurrently with opioids. In some cases, data on the ability of these alternatives to depress respiration alone or in conjunction with an opioid are lacking. A nonclinical in vivo model was developed that could detect worsening respiratory depression when a benzodiazepine (diazepam) was used in combination with an opioid (oxycodone) compared to the opioid alone based on an increased arterial partial pressure of carbon dioxide (pCO2 ). The current study used that model to assess the impact on respiration of non-benzodiazepine sedative psychotropic drugs representative of different drug classes (clozapine, quetiapine, risperidone, zolpidem, trazodone, carisoprodol, cyclobenzaprine, mirtazapine, topiramate, paroxetine, duloxetine, ramelteon, and suvorexant) administered alone and with oxycodone. At clinically relevant exposures, paroxetine, trazodone, and quetiapine given with oxycodone significantly increased pCO2 above the oxycodone effect. Analyses indicated that most pCO2 interaction effects were due to pharmacokinetic interactions resulting in increased oxycodone exposure. Increased pCO2 recorded with oxycodone-paroxetine co-administration exceeded expected effects from only drug exposure suggesting another mechanism for the increased pharmacodynamic response. This study identified drug-drug interaction effects depressing respiration in an animal model when quetiapine or paroxetine were co-administered with oxycodone. Clinical pharmacodynamic drug interaction studies are being conducted with these drugs to assess translatability of these findings.

Evaluating kratom alkaloids using PHASE.

Kratom is a botanical substance that is marketed and promoted in the US for pharmaceutical opioid indications despite having no US Food and Drug Administration approved uses. Kratom contains over forty alkaloids including two partial agonists at the mu opioid receptor, mitragynine and 7-hydroxymitragynine, that have been subjected to the FDA's scientific and medical evaluation. However, pharmacological and toxicological data for the remaining alkaloids are limited. Therefore, we applied the Public Health Assessment via Structural Evaluation (PHASE) protocol to generate in silico binding profiles for 25 kratom alkaloids to facilitate the risk evaluation of kratom. PHASE demonstrates that kratom alkaloids share structural features with controlled opioids, indicates that several alkaloids bind to the opioid, adrenergic, and serotonin receptors, and suggests that mitragynine and 7-hydroxymitragynine are the strongest binders at the mu opioid receptor. Subsequently, the in silico binding profiles of a subset of the alkaloids were experimentally verified at the opioid, adrenergic, and serotonin receptors using radioligand binding assays. The verified binding profiles demonstrate the ability of PHASE to identify potential safety signals and provide a tool for prioritizing experimental evaluation of high-risk compounds.