Improving Detection of Arrhythmia Drug-Drug Interactions in Pharmacovigilance Data through the Implementation of Similarity-Based Modeling

Identification of Drug-Drug Interactions (DDIs) is a significant challenge during drug development and clinical practice. DDIs are responsible for many adverse drug effects (ADEs), decreasing patient quality of life and causing higher care expenses. DDIs are not systematically evaluated in pre-clinical or clinical trials and so the FDA U. S. Food and Drug Administration relies on post-marketing surveillance to monitor patient safety. However, existing pharmacovigilance algorithms show poor performance for detecting DDIs exhibiting prohibitively high false positive rates. Alternatively, methods based on chemical structure and pharmacological similarity have shown promise in adverse drug event detection. We hypothesize that the use of chemical biology data in a post hoc analysis of pharmacovigilance results will significantly improve the detection of dangerous interactions. Our model integrates a reference standard of DDIs known to cause arrhythmias with drug similarity data. To compare similarity between drugs we used chemical structure (both 2D and 3D molecular structure), adverse drug side effects, chemogenomic targets, drug indication classes, and known drug-drug interactions. We evaluated the method on external reference standards. Our results showed an enhancement of sensitivity, specificity and precision in different top positions with the use of similarity measures to rank the candidates extracted from pharmacovigilance data. For the top 100 DDI candidates, similarity-based modeling yielded close to twofold precision enhancement compared to the proportional reporting ratio (PRR). Moreover, the method helps in the DDI decision making through the identification of the DDI in the reference standard that generated the candidate.     

Direct costs and hospital morbimortality impact from preventable adverse drug events

Introduction. Implementing pharmacovigilance activities consists of monitoring and assessment of activities related to medical attention. However, additional data are necessary to identify conditions where care quality can be improved. Therefore, a focus on adverse drug events analysis from a prevention and economic perspective is needed, with emphasis on its local impact. Objective. Preventable adverse drug events were summarized to establishing their impact on morbidity and mortality, as well as to estimate the ensuing economic burden. Materials and methods. The data were gathered from a level 3 hospital (high complexity), located in Bogotá, Colombia, where specific pharmacovigilance activities were recorded in 2007. Patient charts were reviewed to characterize adverse drug events according to their causality, severity and preventability. Direct costs were estimated by grouping diagnostic tests, length of hospitalization, procedures and additional drugs required. Results. The charts of 283 patients and 448 reports were analyzed. These data indicated that 24.8% of adverse drug events were preventable and that an associated mortality of 1.1% had occurred. The associated direct costs were between USD $16,687 and $18,739. Factors more commonly associated with preventability were drug-drug interactions, as well as inappropriate doses and unsuitable frequencies at which the drugs were administrated. Conclusions. The data recommended that actions be taken to decrease preventable adverse drug events, because of negative impact on patient′s health, and unnecessary consumption of healthcare resources.     

A Pharmacovigilance Approach for Post-Marketing in Japan Using the Japanese Adverse Drug Event Report (JADER) Database and Association Analysis

BackgroundRapid dissemination of information regarding adverse drug reactions is a key aspect for improving pharmacovigilance. There is a possibility that unknown adverse drug reactions will become apparent through post-marketing administration. Currently, although there have been studies evaluating the relationships between a drug and adverse drug reactions using the JADER database which collects reported spontaneous adverse drug reactions, an efficient approach to assess the association between adverse drug reactions of drugs with the same indications as well as the influence of demographics (e.g. gender) has not been proposed.ConclusionsDifferent combinations of adverse drug reactions were noted between the antidepressants. In addition, the reported adverse drug reactions differed by gender. This approach using a large database for examining the associations can improve safety monitoring during the post-marketing phase.     

Machine learning liver-injuring drug interactions with non-steroidal anti-inflammatory drugs (NSAIDs) from a retrospective electronic health record (EHR) cohort

Drug-drug interactions account for up to 30% of adverse drug reactions. Increasing prevalence of electronic health records (EHRs) offers a unique opportunity to build machine learning algorithms to identify drug-drug interactions that drive adverse events. In this study, we investigated hospitalizations’ data to study drug interactions with non-steroidal anti-inflammatory drugs (NSAIDS) that result in drug-induced liver injury (DILI). We propose a logistic regression based machine learning algorithm that unearths several known interactions from an EHR dataset of about 400,000 hospitalization. Our proposed modeling framework is successful in detecting 87.5% of the positive controls, which are defined by drugs known to interact with diclofenac causing an increased risk of DILI, and correctly ranks aggregate risk of DILI for eight commonly prescribed NSAIDs. We found that our modeling framework is particularly successful in inferring associations of drug-drug interactions from relatively small EHR datasets. Furthermore, we have identified a novel and potentially hepatotoxic interaction that might occur during concomitant use of meloxicam and esomeprazole, which are commonly prescribed together to allay NSAID-induced gastrointestinal (GI) bleeding. Empirically, we validate our approach against prior methods for signal detection on EHR datasets, in which our proposed approach outperforms all the compared methods across most metrics, such as area under the receiver operating characteristic curve (AUROC) and area under the precision-recall curve (AUPRC).     

Drug Repositioning by Integrating Known Disease-Gene and Drug-Target Associations in a Semi-supervised Learning Model

Computational drug repositioning has been proven as a promising and efficient strategy for discovering new uses from existing drugs. To achieve this goal, a number of computational methods have been proposed, which are based on different data sources of drugs and diseases. These methods approach the problem using either machine learning- or network-based models with an assumption that similar drugs can be used for similar diseases to identify new indications of drugs. Therefore, similarities between drugs and between diseases are usually used as inputs. In addition, known drug-disease associations are also needed for the methods as prior information. It should be noted that those associations are still not well established due to the fact that many of marketed drugs have been withdrawn and this could affect the outcome of the methods. In this study, we propose a novel method named RLSDR (Regularized Least Square for Drug Repositioning) to find new uses of drugs. More specifically, it relies on a semi-supervised learning model, Regularized Least Square, thus it does not require definition of non-drug-disease associations as previously proposed machine learning-based methods. In addition, the similarity between drugs measured by chemical structures of drug compounds and the similarity between diseases which share phenotypes can be represented in a form of either similarity network or similarity matrix as inputs of the method. Moreover, instead of using a gold-standard set of known drug-disease associations, we construct an artificial set of the associations based on known disease-gene and drug-target associations. Experiment results demonstrate that RLSDR achieves better prediction performance on the artificial set of drug-disease associations than that on the gold-standard ones in terms of area under the Receiver Operating Characteristic (ROC) curve (AUC). In addition, it outperforms two representative network-based methods irrespective of the prior information of drug-disease associations. Novel indications for a number of drugs are also identified and validated by evidences from a different data resource.     

Perceptions of doctors to adverse drug reaction reporting in a teaching hospital in Lagos, Nigeria

Background  

Spontaneous adverse drug reaction (ADR) reporting is the cornerstone of pharmacovigilance. ADR reporting with Yellow Cards has tremendously improved pharmacovigilance of drugs in many developed countries and its use is advocated by the World Health Organization (WHO). This study was aimed at investigating the knowledge and attitude of doctors in a teaching hospital in Lagos, Nigeria on spontaneous ADR reporting and to suggest possible ways of improving this method of reporting.     

Designing adverse event forms for real-world reporting: participatory research in Uganda

The wide-scale roll-out of artemisinin combination therapies (ACTs) for the treatment of malaria should be accompanied by continued surveillance of their safety. Post-marketing pharmacovigilance (PV) relies on adverse event (AE) reporting by clinicians, but as a large proportion of treatments are provided by non-clinicians in low-resource settings, the effectiveness of such PV systems is limited. To facilitate reporting, AE forms should be easily completed; however, most are challenging for lower-level health workers and non-clinicians to complete. Through participatory research, we sought to develop user-friendly AE report forms to capture information on events associated with ACTs.Following situation analysis, we undertook workshops with community medicine distributors and health workers in Jinja, Uganda, to develop a reporting form based on experiences and needs of users, and communication and visual perception principles. Participants gave feedback for revisions of subsequent versions. We then conducted 8 pretesting sessions with 77 potential end users to test and refine passive and active versions of the form.The development process resulted in a form that included a pictorial storyboard to communicate the rationale for the information needed and facilitate rapport between the reporter and the respondent, and a diary format to record the drug administration and event details in chronological relation to each other. Successive rounds of pretesting used qualitative and quantitative feedback to refine the form, with the final round showing over 80% of the form completed correctly by potential end users.We developed novel AE report forms that can be used by non-clinicians to capture pharmacovigilance data for anti-malarial drugs. The participatory approach was effective for developing forms that are intuitive for reporters, and motivating for respondents. The forms, or their key components, could be adapted for use in other low-literacy settings to improve quality and quantity of drug safety reports as new medicines are scaled-up.     

Variable selection for zero‐inflated and overdispersed data with application to health care demand in Germany

In health services and outcome research, count outcomes are frequently encountered and often have a large proportion of zeros. The zero‐inflated negative binomial (ZINB) regression model has important applications for this type of data. With many possible candidate risk factors, this paper proposes new variable selection methods for the ZINB model. We consider maximum likelihood function plus a penalty including the least absolute shrinkage and selection operator (LASSO), smoothly clipped absolute deviation (SCAD), and minimax concave penalty (MCP). An EM (expectation‐maximization) algorithm is proposed for estimating the model parameters and conducting variable selection simultaneously. This algorithm consists of estimating penalized weighted negative binomial models and penalized logistic models via the coordinated descent algorithm. Furthermore, statistical properties including the standard error formulae are provided. A simulation study shows that the new algorithm not only has more accurate or at least comparable estimation, but also is more robust than the traditional stepwise variable selection. The proposed methods are applied to analyze the health care demand in Germany using the open‐source R package mpath .     

A comparison of model selection methods for prediction in the presence of multiply imputed data

Many approaches for variable selection with multiply imputed data in the development of a prognostic model have been proposed. However, no method prevails as uniformly best. We conducted a simulation study with a binary outcome and a logistic regression model to compare two classes of variable selection methods in the presence of MI data: (I) Model selection on bootstrap data, using backward elimination based on AIC or lasso, and fit the final model based on the most frequently (e.g. ) selected variables over all MI and bootstrap data sets; (II) Model selection on original MI data, using lasso. The final model is obtained by (i) averaging estimates of variables that were selected in any MI data set or (ii) in 50% of the MI data; (iii) performing lasso on the stacked MI data, and (iv) as in (iii) but using individual weights as determined by the fraction of missingness. In all lasso models, we used both the optimal penalty and the 1‐se rule. We considered recalibrating models to correct for overshrinkage due to the suboptimal penalty by refitting the linear predictor or all individual variables. We applied the methods on a real dataset of 951 adult patients with tuberculous meningitis to predict mortality within nine months. Overall, applying lasso selection with the 1‐se penalty shows the best performance, both in approach I and II. Stacking MI data is an attractive approach because it does not require choosing a selection threshold when combining results from separate MI data sets     

Suicidal behavior and depression in smoking cessation treatments

Background

Two treatments for smoking cessation—varenicline and bupropion—carry Boxed Warnings from the U.S. Food and Drug Administration (FDA) about suicidal/self-injurious behavior and depression. However, some epidemiological studies report an increased risk in smoking or smoking cessation independent of treatment, and differences between drugs are unknown.

Methodology

From the FDA''s Adverse Event Reporting System (AERS) database from 1998 through September 2010 we selected domestic, serious case reports for varenicline (n = 9,575), bupropion for smoking cessation (n = 1,751), and nicotine replacement products (n = 1,917). A composite endpoint of suicidal/self-injurious behavior or depression was defined as a case with one or more Preferred Terms in Standardized MedDRA Query (SMQ) for those adverse effects. The main outcome measure was the ratio of reported suicide/self-injury or depression cases for each drug compared to all other serious events for that drug.

Results

Overall we identified 3,249 reported cases of suicidal/self-injurious behavior or depression, 2,925 (90%) for varenicline, 229 (7%) for bupropion, and 95 (3%) for nicotine replacement. Compared to nicotine replacement, the disproportionality results (OR (95% CI)) were varenicline 8.4 (6.8–10.4), and bupropion 2.9 (2.3–3.7). The disproportionality persisted after excluding reports indicating concomitant therapy with any of 58 drugs with suicidal behavior warnings or precautions in the prescribing information. An additional antibiotic comparison group showed that adverse event reports of suicidal/self-injurious behavior or depression were otherwise rare in a healthy population receiving short-term drug treatment.

Conclusions

Varenicline shows a substantial, statistically significant increased risk of reported depression and suicidal/self-injurious behavior. Bupropion for smoking cessation had smaller increased risks. The findings for varenicline, combined with other problems with its safety profile, render it unsuitable for first-line use in smoking cessation.     

Esomeprazole covalently interacts with the cardiovascular enzyme dimethylarginine dimethylaminohydrolase: Insights into the cardiovascular risk of proton pump inhibitors

BackgroundProton pump inhibitors (PPIs) are widely prescribed drugs for the treatment of gastroesophageal reflux disease (GERD). Several meta-analysis studies have reported associations between prolonged use of PPIs and major adverse cardiovascular events. However, interaction of PPIs with biological molecules involved in cardiovascular health is incompletely characterized. Dimethylarginine dimethylaminohydrolase (DDAH) is a cardiovascular enzyme expressed in cardiomyocytes, and other somatic cell types in one of two isotypes (DDAH1 and DDAH2) to metabolize asymmetric dimethylarginine (ADMA); a cardiovascular risk factor and competitive inhibitor of nitric oxide synthases (NOSs).MethodsWe performed high throughput drug screening of over 130,000 small molecules to discover human DDAH1 inhibitors and found that PPIs directly inhibit DDAH1. We expressed and purified the enzyme for structural and mass spectrometry proteomics studies to understand how a prototype PPI, esomeprazole, interacts with DDAH1. We also performed molecular docking studies to model the interaction of DDAH1 with esomeprazole. X-ray crystallography was used to determine the structure of DDAH1 alone and bound to esomeprazole at resolutions ranging from 1.6 to 2.9 Å.ResultsAnalysis of the enzyme active site shows that esomeprazole interacts with the active site cysteine (Cys273) of DDAH1. The structural studies were corroborated by mass spectrometry which indicated that cysteine was targeted by esomeprazole to inactivate DDAH1.ConclusionsThe inhibition of this important cardiovascular enzyme by a PPI may help explain the reported association of PPI use and increased cardiovascular risk in patients and the general population.General significanceOur study calls for pharmacovigilance studies to monitor adverse cardiovascular events in chronic PPI users.     

A review of the common properties of drugs with idiosyncratic hepatotoxicity and the "multiple determinant hypothesis" for the manifestation of idiosyncratic drug toxicity

Idiosyncratic drug toxicity is generally believed to be a phenomenon that cannot be readily evaluated experimentally. Reasons for this difficulty include the following: 1. It is a rare event (<1/5,000) and therefore impossible to be studied in clinical trials; 2. It is a human-specific event not detectable in experimental animals. To aid the understanding of idiosyncratic toxicity and to develop an experimental strategy for this phenomenon, a hypothesis is proposed. The hypothesis states that the low frequency of idiosyncratic drug toxicity is due to the requirements for the occurrence of multiple critical and discrete events, with the probability for the occurrence of idiosyncratic drug toxicity as a product of the probabilities of each event. The key determinants of these critical events are proposed to be: 1. Chemical properties; 2. exposure; 3. environmental factors; and 4. genetic factors. Based on this hypothesis, idiosyncratic drug toxicity can be evaluated experimentally via studying these key determinants. The chemical properties critical to idiosyncratic drug toxicity are identified via a review of the common properties of drugs that cause idiosyncratic liver toxicity. These properties include: 1. Formation of reactive metabolites. 2. Metabolism by P450 isoforms. 3. Preponderance of P450 inducers, and 4. Occurrence of clinically significant pharmacokinetic interactions with co-administered drugs. Based on this review, it is proposed that these common properties may be useful experimental endpoints for the prediction and therefore avoidance of the selection of drug candidates with idiosyncratic drug toxicity for further development.     

Joint model of recurrent events and a terminal event with time‐varying coefficients

Joint modeling of recurrent events and a terminal event has been studied extensively in the past decade. However, most of the previous works assumed constant regression coefficients. This paper proposes a joint model with time‐varying coefficients in both event components. The proposed model not only accommodates the correlation between the two type of events, but also characterizes the potential time‐varying covariate effects. It is especially useful for evaluating long‐term risk factors' effect that could vary with time. A Gaussian frailty is used to model the correlation between event times. The nonparametric time‐varying coefficients are modeled using cubic splines with penalty terms. A simulation study shows that the proposed estimators perform well. The model is used to analyze the readmission rate and mortality jointly for stroke patients admitted to Veterans Administration (VA) Hospitals.     

Evaluation of medication side-effects from an economic perspective

Pharmacovigilance is a health sciences discipline devoted to the data collection, data analysis and decision-making related to adverse drug reactions (ADR). It has played an expanded theoretical and practical role since the 1960's. However, few studies have made a careful analysis of the decision-making costs in evaluating ADRs. Herein, the relevant literature is reviewed concerning the costs generated due to the attention of the drug adverse events in medical practice. Examples are taken from international literature which offer by extrapolation of the potential future costs of ADR in Colombia. The objective is to sensitize and generate insights about the need for implementation and development of a national pharmacovigilance system.     

Identification of biomarkers for risk stratification of cardiovascular events using genetic algorithm with recursive local floating search

Conventional biomarker discovery focuses mostly on the identification of single markers and thus often has limited success in disease diagnosis and prognosis. This study proposes a method to identify an optimized protein biomarker panel based on MS studies for predicting the risk of major adverse cardiac events (MACE) in patients. Since the simplicity and concision requirement for the development of immunoassays can only tolerate the complexity of the prediction model with a very few selected discriminative biomarkers, established optimization methods, such as conventional genetic algorithm (GA), thus fails in the high‐dimensional space. In this paper, we present a novel variant of GA that embeds the recursive local floating enhancement technique to discover a panel of protein biomarkers with far better prognostic value for prediction of MACE than existing methods, including the one approved recently by FDA (Food and Drug Administration). The new pragmatic method applies the constraints of MACE relevance and biomarker redundancy to shrink the local searching space in order to avoid heavy computation penalty resulted from the local floating optimization. The proposed method is compared with standard GA and other variable selection approaches based on the MACE prediction experiments. Two powerful classification techniques, partial least squares logistic regression (PLS‐LR) and support vector machine classifier (SVMC), are deployed as the MACE predictors owing to their ability in dealing with small scale and binary response data. New preprocessing algorithms, such as low‐level signal processing, duplicated spectra elimination, and outliner patient's samples removal, are also included in the proposed method. The experimental results show that an optimized panel of seven selected biomarkers can provide more than 77.1% MACE prediction accuracy using SVMC. The experimental results empirically demonstrate that the new GA algorithm with local floating enhancement (GA‐LFE) can achieve the better MACE prediction performance comparing with the existing techniques. The method has been applied to SELDI/MALDI MS datasets to discover an optimized panel of protein biomarkers to distinguish disease from control.     

Regularized estimation for the accelerated failure time model

Summary .  In the presence of high-dimensional predictors, it is challenging to develop reliable regression models that can be used to accurately predict future outcomes. Further complications arise when the outcome of interest is an event time, which is often not fully observed due to censoring. In this article, we develop robust prediction models for event time outcomes by regularizing the Gehan's estimator for the accelerated failure time (AFT) model ( Tsiatis, 1996 , Annals of Statistics 18, 305–328) with least absolute shrinkage and selection operator (LASSO) penalty. Unlike existing methods based on the inverse probability weighting and the Buckley and James estimator ( Buckley and James, 1979 , Biometrika 66, 429–436), the proposed approach does not require additional assumptions about the censoring and always yields a solution that is convergent. Furthermore, the proposed estimator leads to a stable regression model for prediction even if the AFT model fails to hold. To facilitate the adaptive selection of the tuning parameter, we detail an efficient numerical algorithm for obtaining the entire regularization path. The proposed procedures are applied to a breast cancer dataset to derive a reliable regression model for predicting patient survival based on a set of clinical prognostic factors and gene signatures. Finite sample performances of the procedures are evaluated through a simulation study.     

Immunotherapy for Alzheimer's disease: will vaccination work?

Active or passive immunization against the beta-amyloid peptide (Abeta) has been proposed as a method for preventing and/or treating Alzheimer's disease (AD). In addition to lowering brain Abeta and amyloid burden in transgenic mouse models of AD, a beneficial effect of immunization on previously characterized memory impairment(s) has also been reported in these mice. Whether these preclinical data will predict efficacy in AD patients remains to be seen. A clinical trial of active immunization (vaccination) was halted, owing to a serious adverse event (meningoencephalitis), raising questions about the safety of this approach. Two recent reports suggest that immunotherapy-based approaches to treating and preventing AD will require careful antigen and antibody selection, to maximize efficacy and minimize serious adverse events. However, given the potential efficacy of this approach, we believe that immunotherapy for AD should not be prematurely abandoned.     

Assessing the sensitivity of meta-analysis to selection bias: a multiple imputation approach

Evidence synthesis, both qualitatively and quantitatively through meta-analysis, is central to the development of evidence-based medicine. Unfortunately, meta-analysis is often complicated by the suspicion that the available studies represent a biased subset of the evidence, possibly due to publication bias or other systematically different effects in small studies. A number of statistical methods have been proposed to address this, among which the trim-and-fill method and the Copas selection model are two of the most widely discussed. However, both methods have drawbacks: the trim-and-fill method is based on strong assumptions about the symmetry of the funnel plot; the Copas selection model is less accessible to systematic reviewers, and sometimes encounters estimation problems. In this article, we adopt a logistic selection model, and show how treatment effects can be rapidly estimated via multiple imputation. Specifically, we impute studies under a missing at random assumption, and then reweight to obtain estimates under nonrandom selection. Our proposal is computationally straightforward. It allows users to increase selection while monitoring the extent of remaining funnel plot asymmetry, and also visualize the results using the funnel plot. We illustrate our approach using a small meta-analysis of benign prostatic hyperplasia.