The package insert: who writes it and why, what are its implications, and how well does medical school explain it?

Life-threatening Rash Due to Lamotrigine and a Failure to Understand Its Pharmacology: How Forensic Detective Work Uses Medical Knowledge and Clinical Pharmacology to Solve Cases

This column is the second of a 3-part series describing cases where general medical knowledge, including psychiatric and clinical pharmacology, were instrumental in determining dereliction and direct cause in a malpractice suit. This case summarizes how lamotrigine can cause dangerous consequences if its pharmacology is not properly understood. The case also illustrates how the 4 Ds of a forensic malpractice suit were met in this case. First, there was duty on the part of the prescriber which, if followed, would have prevented or minimized the damages experienced by the patient. Dereliction in the performance of a patient-physician treatment contract was a direct cause of the development of Stevens-Johnson syndrome/toxic epidermal necrolysis (SJS/TEN) in this patient. An immune-mediated reaction to lamotrigine or one of its metabolites has been extensively reported in the literature, with the risk of this reaction increasing at higher doses and with more rapid titration, fulfilling the elements of direct cause. Dereliction implies a deviation from the standard of care. On the basis of the clinical information from the package insert, more likely than not a deviation from the standard of care occurred in this case when lamotrigine was titrated faster than recommended by the package insert.

Automatic text classification of drug-induced liver injury using document-term matrix and XGBoost

Introduction

Regulatory agencies generate a vast amount of textual data in the review process. For example, drug labeling serves as a valuable resource for regulatory agencies, such as U.S. Food and Drug Administration (FDA) and Europe Medical Agency (EMA), to communicate drug safety and effectiveness information to healthcare professionals and patients. Drug labeling also serves as a resource for pharmacovigilance and drug safety research. Automated text classification would significantly improve the analysis of drug labeling documents and conserve reviewer resources.

Methods

We utilized artificial intelligence in this study to classify drug-induced liver injury (DILI)-related content from drug labeling documents based on FDA's DILIrank dataset. We employed text mining and XGBoost models and utilized the Preferred Terms of Medical queries for adverse event standards to simplify the elimination of common words and phrases while retaining medical standard terms for FDA and EMA drug label datasets. Then, we constructed a document term matrix using weights computed by Term Frequency-Inverse Document Frequency (TF-IDF) for each included word/term/token.

Results

The automatic text classification model exhibited robust performance in predicting DILI, achieving cross-validation AUC scores exceeding 0.90 for both drug labels from FDA and EMA and literature abstracts from the Critical Assessment of Massive Data Analysis (CAMDA).

Discussion

Moreover, the text mining and XGBoost functions demonstrated in this study can be applied to other text processing and classification tasks.

Comparative Pharmacology of the 3 Marketed Dual Orexin Antagonists-Daridorexant, Lemborexant, and Suvorexant-Part 2. Principal Drug Metabolizing Enzyme, Drug-Drug Interactions, and Effects of Liver and Renal Impairment on Metabolism

This column is the second in a 2-part series presenting the comparative pharmacology of the 3 Food and Drug Administration-approved dual orexin receptor antagonists, daridorexant, lemborexant, and suvorexant. Both of the columns in this series emphasize the pharmacokinetics of these drugs as they are relevant to their use as sleep medications. Although other classes of sleep medications are not discussed, the same pharmacokinetic principles also apply to them in terms of endeavoring to match the pharmacokinetics of an agent to the individual's usual sleep cycle. This second column in the series focuses on the metabolism of each of the 3 drugs by the cytochrome P450 enzyme CYP3A, guidance for using these agents in combination with drugs that are CYP3A inhibitors or inducers, and how to adjust dosing in patients with comorbid conditions such as hepatic or renal impairment.

Comparative Pharmacology of the 3 Marketed Dual Orexin Antagonists-Daridorexant, Lemborexant, and Suvorexant: Part 1: Pharmacokinetic Profiles

This first column in a 2-part series focuses on the pharmacokinetics of the 3 Food and Drug Administration-approved dual orexin receptor antagonists, daridorexant, lemborexant, and suvorexant, specifically as they relate to their use as sleep medications. Although other classes of sleep medications are not discussed, the same pharmacokinetic principles also apply to them.

Treatment of Agitation in Individuals With Bipolar Disorder or Schizophrenia: Lessons Learned for Clinical Psychiatry and Psychiatric Drug Development

Six lessons can be learned from the pivotal registration trials for sublingual dexmedetomidine (SLD) for the treatment of agitation in individuals with bipolar disorder or schizophrenia: (1) Knowing the function of a well-defined circuit in the brain, such as the locus coeruleus (LC), facilitates the development of central nervous system drugs. (2) Agitation can be conceptualized both clinically and physiologically. From both perspectives, agitation can present and escalate along a spectrum from mild, characterized as mainly hyperaroused (possibly only a subjective experience with no observable manifestations in its mildest form), to moderate to severe. In the severe state, the patient poses a potential danger to self and others. The level of agitation a patient is experiencing can determine the most appropriate treatment. Behavioral techniques may be sufficient for the mild state. As agitation progresses beyond mild severity, medication intervention becomes needed. SLD can be effective when agitation is moderate or even more severe. At this stage, patients can recognize and be distressed by their symptoms and participate in treatment. When agitation has escalated to such a severe state that patients can no longer participate in treatment, then intramuscular or intravenous medication may be needed. In quite severe cases, physical restraint as well as medication may be required. The Positive and Negative Syndrome Scale-Excited Component (PANSS-EC or PEC), a subscale of the PANSS, is a helpful instrument to assess where an individual is along the agitation spectrum. The PEC has been used in studies of pharmacological treatments for agitation, and it is accepted by the US Food and Drug Administration as the primary rating instrument in pivotal efficacy studies of treatments for agitation. (3) Where the patient is on the agitation spectrum is a function of the activity of the LC, which can be one factor in determining the SLD dose that will optimize the patient's clinical outcome. Clinical outcome is optimized when complete resolution of agitation is rapidly achieved, and adverse effects either do not occur or are not clinically meaningful. The adverse effects of greatest interest with SLD are decreases in resting systolic and diastolic blood pressures, reductions in these blood pressures under orthostatic stress, and lower resting heart rate. (4) To ensure safety, the subjects in 2 healthy volunteer studies were not administered doses equivalent to those used to treat agitated patients. The highest dose which a healthy volunteer tolerated in those studies was 40 µg. Agitated patients were treated with 120 and 180 µg doses. Thus the difference in doses was 3- to 4.5-fold. Agitated patients could also receive 2 additional half doses with an interval of 2 hours between the first and second administrations. For context, there are other examples of situations in which the dose of a drug that is well tolerated by healthy volunteers is lower than the dose that is well tolerated by patients. For example, it has long been accepted that patients with an acute relapse of schizophrenia can tolerate and need higher doses of D2 antagonists for efficacy than healthy volunteers can tolerate who will generally experience substantial sedation if given what is a clinically effective dose in such patients. (5) Agitation is a state phenomenon that may not recur when it is effectively treated, so that the treatment effect can persist for 24 hours despite the plasma half-life of the drug being 2 to 3 hours. (6) Given the established function of the LC, the fact that the dose response and the time curve of the effect are virtually identical in agitated individuals with bipolar disorder or schizophrenia supports the conclusion that the drug is not treating the syndromic diagnoses of bipolar disorder and schizophrenia but rather the state of being agitated because of overactivity of the LC. These 6 lessons are consistent with the discussions in numerous earlier columns in this series and are critical for both the practice of clinical psychopharmacology and psychiatric drug development.

BERT-Based Natural Language Processing of Drug Labeling Documents: A Case Study for Classifying Drug-Induced Liver Injury Risk

Background & Aims: The United States Food and Drug Administration (FDA) regulates a broad range of consumer products, which account for about 25% of the United States market. The FDA regulatory activities often involve producing and reading of a large number of documents, which is time consuming and labor intensive. To support regulatory science at FDA, we evaluated artificial intelligence (AI)-based natural language processing (NLP) of regulatory documents for text classification and compared deep learning-based models with a conventional keywords-based model.

Methods: FDA drug labeling documents were used as a representative regulatory data source to classify drug-induced liver injury (DILI) risk by employing the state-of-the-art language model BERT. The resulting NLP-DILI classification model was statistically validated with both internal and external validation procedures and applied to the labeling data from the European Medicines Agency (EMA) for cross-agency application.

Results: The NLP-DILI model developed using FDA labeling documents and evaluated by cross-validations in this study showed remarkable performance in DILI classification with a recall of 1 and a precision of 0.78. When cross-agency data were used to validate the model, the performance remained comparable, demonstrating that the model was portable across agencies. Results also suggested that the model was able to capture the semantic meanings of sentences in drug labeling.

Conclusion: Deep learning-based NLP models performed well in DILI classification of drug labeling documents and learned the meanings of complex text in drug labeling. This proof-of-concept work demonstrated that using AI technologies to assist regulatory activities is a promising approach to modernize and advance regulatory science.

A systematic comparison of hepatobiliary adverse drug reactions in FDA and EMA drug labeling reveals discrepancies

Drug labeling informs physicians and patients on the safe and effective use of medication. However, recent studies suggested discrepancies in labeling of the same drug between different regulatory agencies. Here, we evaluated the hepatic safety information in labeling for 549 medications approved by the US Food and Drug Administration (FDA) and the European Medicines Agency (EMA). Limited discrepancies were found regarding risk for hepatic adverse drug reactions (ADRs) (8.7% in hepatic ADR warnings and 21.3% in contraindication for liver disease), while caution should be exercised over drugs with inconsistencies in contraindications for liver disease and evidence for hepatotoxicity (4.9%). Most discrepancies were attributable to less-severe hepatic events and low-frequency hepatic ADR reports and had limited implication on clinical outcomes.

Two Clinically Important but Underutilized and Misunderstood Tools: Formulas to Estimate Creatinine Clearance and Therapeutic Drug Monitoring

This column first reviews 2 key equations that are central to clinical pharmacology. Clinicians can use the first equation to predict the effect of a specific dose of a specific drug in specific circumstances on the basis of 3 variables: (1) the drug's pharmacodynamics, (2) the drug's pharmacokinetics, and (3) biological variance in the individual patient. Clinicians can use the second equation to determine the concentration of a drug that a patient will achieve on a given dose depending on the patient's ability to clear the drug from the body. These 2 equations allow prescribers to predict whether the dose of a drug a patient is receiving is likely to achieve the desired clinical response (not so low that it is clinically ineffective or so high that it causes adverse effects that interfere with the patient's ability to tolerate or benefit from the treatment). The author then describes 2 tools clinicians can use to determine a patient's ability to clear a drug from the body, and thus calculate the concentration of the drug using Equation 2. These tools are: (1) estimation of creatinine clearance and (2) therapeutic drug monitoring.

Legal regulations, ethical guidelines and recent policies to increase transparency of clinical trials

Timely and accurate dissemination of outcomes is essential to accomplish main benefits of scientific research including clinical trials. Clinical trial results can be disseminated in two main ways: by publication in a peer-reviewed journal and by posting on a publicly available clinical trial register. The credibility of the literature on clinical trials is significantly diminished because a high percentage of trials is not published. While current legal regulations both in the European Union (EU) and the USA impose a duty to submit summary results of clinical trials to a respective register (EU Clinical Trial Register and ClinicalTrials.gov, respectively), the compliance with this requirement has been generally inadequate. Trial outcomes can be also made accessible by data sharing. However, in spite of the wide promotion of this idea, the access of investigators to participant-level datasets remains limited. The main objective of this review is to discuss current legal regulations, international standards, ethical guidelines and recent policies pertaining to dissemination of clinical trial results.

CNS Drug Development, Lessons Learned, Part 5: How Preclinical and Human Safety Studies Inform the Approval and Subsequent Use of a New Drug-Suvorexant as an Example

This column is the fifth in a series examining the advances being made in central nervous system drug development because of advances in molecular pharmacology and an improved understanding of the neurobiology underlying disturbances in brain function including psychiatric illnesses. This column covers the special animal and human studies conducted as part of the development of suvorexant, which is the first in the class of dual orexin 1 and 2 receptor antagonists to be approved; it has an indication for the treatment of disturbances in sleep onset and maintenance. The animal studies included determination of the therapeutic index of the drug (ie, lethal dose 50 which is the dose at which 50% of animals die following administration of the drug), adverse effects on fertility, teratogenicity, carcinogenicity, and ability to cause narcolepsy. The human studies included investigation of the effects of the drug on balance, memory, driving performance, and propensity to cause respiratory depression in normal volunteers and individuals with mild to moderate chronic obstructive pulmonary disorder or mild to moderate obstructive sleep apnea. This column illustrates how targeting the drug to one mechanism out of hundreds yields increased safety and highlights the importance of the package insert which summarizes the results of all of the studies from the drug's development program.