Post-screen (grey)A systematic comparison of hepatobiliary adverse drug reactions in FDA and EMA drug labeling reveals discrepancies
Introduction
Ensuring the safe use of medicines and protecting public health from ADRs is the mission of regulatory agencies, such as the FDA and EMA.1 Drug labeling, which is drafted by manufacturers and approved by the agencies, is issued to inform the public about the safe and effective use of drugs. For each drug approved by the agencies, drug labeling is a reference guiding healthcare professionals in selecting appropriate drugs for the right patients to avoid ADRs.2, 3, 4 The drug labeling is referred to as ‘prescribing information’ in the USA and ‘summary of product characteristics’ in Europe and is the result of fine-tuning between regulatory agencies and pharmaceutical companies. Labels are legal documents that provide comprehensive information for prescribers. By contrast, a simplified version of labeling, ‘patient information leaflet’, is used to convey concise and general information to patients.
In the drug labeling, several sections are designed to communicate relevant safety information. For example, the ‘Warnings and Precautions’ and ‘Boxed Warning’ sections5 in FDA labeling and the ‘Special warnings and precautions for use’ section6 in EMA labeling inform about potential serious or clinically significant ADRs associated with the use of medications, whereas the ‘Contraindications’ section5, 6 aims to prevent the use of such drugs among high-risk patients. Considering the importance of the proper use of drugs, both agencies have published guidance documents for industry on drafting these sections of the drug labeling for human prescription drugs and biological products.5, 6
Remarkably, a joint FDA/EMA analysis showed high concordance (91–98%) in marketing authorization decisions for new medicines, although both agencies evaluate applications independently of each other.7 By contrast, several reports have highlighted the fact that summarized safety information in FDA and EMA labeling differs in parts.8, 9, 10 This has raised concerns that drug labeling with discrepancies might not provide healthcare professionals with accurate information needed to prescribe drugs appropriately. Inaccurate information also could increase the risk of ADRs, estimated to cause over 106,000 deaths in the USA per year and result in hundreds of billions dollars in hospital expenses annually.11, 12, 13, 14
Specifically, drug-induced liver injury (DILI) is a common ADR and a top-ranking cause of acute liver failure in the USA.15, 16, 17 Bjornsson et al. conducted a pioneering study8 to assess the labeling discrepancies of contraindications and cautions for liver disease in the package inserts of the FDA and EMA for the top 200 (in sales) US-marketed drugs. The analysis found 18 of 150 (or 12%) of US drugs and 43 of 139 (or 31%) of European drugs were contraindicated in patients with liver disease. Labeling for 33 of these drugs showed discrepancies between the FDA and EMA in contraindications for liver disease. The findings were significant, which led to important questions about the extent of labeling discrepancies in drug labeling, the possible reasons, and clinical implications for the discrepancies.
Given the differences in labeling between the FDA and EMA, here, we systematically evaluated the extent of labeling discrepancies between the package inserts of the FDA and EMA. We compared side-by-side 549 drugs approved by both agencies based on their hepatic safety information and considered the ADR warning- and contraindication-related sections. Next, we calculated the labeling discrepancies between the two agencies in terms of warnings for hepatic ADRs and contraindications for liver disease. Potential factors contributing to the discrepancies were explored. Moreover, we examined the association of labeling discrepancies with hepatic ADRs using published clinical cases included in the LiverTox database. Additionally, we also explored drugs with requirement of dose adjustment.
Section snippets
Data sources for drugs
Drug labels were retrieved from DailyMed (www.dailymed.nlm.nih.gov) and the Electronic Medicines Compendium (EMC). DailyMed is a public database that contains up-to-date drug labeling approved by the FDA. The FDA provides contents for DailyMed and updates it daily. Meanwhile, because the EMA issues standardized package inserts for the drugs approved through a centralized procedure, we used UK-marketed drugs as representatives of drugs authorized in Europe. The EMC (www.medicines.org.uk)
Drug selection criteria
We selected prescription drugs based on two criteria: (i) with a single active ingredient; and (ii) for either oral or injection use (Fig. S1 in the supplemental information online). Over-the-counter (OTC) drugs were removed because of their different labeling format and requirements compared with prescription drugs. We retrieved a total of 9135 drug labels from the EMC. Drugs not of a single active ingredient, and those not approved for oral or injection use were excluded, yielding 881 unique
Comparison of drug labeling in term of warnings for hepatic ADRs and contraindications for liver disease
Drugs were categorized into two groups: (i) with and (ii) without warning for hepatic ADRs. We searched for hepatic ADRs using selected terms17 from the Medical Dictionary for Regulatory Activities (MedDRA) (Table S1 in the supplemental information online) in the labeling sections of ‘Boxed Warning’ (FDA), ‘Warnings and Precautions’ (FDA), and ‘Special warnings and precautions for use’ (EMA). If a drug carried a description of hepatic ADRs in these warning sections, it was categorized as having
Clinical severity and frequency of hepatic ADRs
Severity of adverse reactions is a crucial factor in risk assessment. In this study, the severity of hepatic ADRs associated with a single drug was determined by considering both the Common Terminology Criteria for Adverse Events v5.0 from the US National Cancer Institute and the clinical severity grading scale (score 1–5) defined by the US DILI Network (DILIN).25 Specifically, the severity of liver injury was graded based on the hepatic ADRs listed in the sections ‘Boxed Warning’, ‘Warnings
LiverTox data
The LiverTox database (www.ncbi.nlm.nih.gov/books/n/livertox/) contains published DILI cases evaluated by the Expert Review Committee members and external reviewers for over 1000 drugs. Likelihood scores of hepatotoxicity were assigned to the drugs based on the frequency of well-documented published cases.27, 32 Among the 549 investigated drugs in this study, 349 drugs were found with records in the LiverTox database and, thus, were included in the LiverTox data analysis.
The Likelihood scores
Data analysis
Fisher’s exact, Chi-square, and t-tests were performed using the GraphPad Prism version 6.07 for Windows (GraphPad Software, San Diego, California USA, www.graphpad.com). All bar charts and heatmaps were constructed using custom scripts by ggplot2 version 3.2.0 in R.
A positive likelihood ratio (LR) indicated the ratio of the probability of drugs with labeling discrepancies causing liver injury over that of drugs without labeling discrepancies causing liver injury, whereas as a negative LR
Labeling discrepancy in terms of warnings for hepatic ADRs and contraindications for liver disease between FDA and EMA drug labeling
To investigate the extent of labeling discrepancy in hepatic safety information between FDA and EMA labeling, we compared side-by-side 549 drugs approved by the two agencies mainly from: (i) warning for hepatic ADRs; and (ii) contraindications for liver disease (Fig. 1).
By using a MedDRA term-based text-mining approach, the investigated drugs were classified into two categories (with or without warning for significant hepatic ADRs) according to the sections for warning and precautions in the
Factors contributing to labeling discrepancies
Incidence is a crucial factor for assessing risk of adverse reactions. Among the 167 drugs with warnings for hepatic ADRs in either FDA or EMA labeling, 134 drugs were found with available hepatic ADR frequency information, which was consistent between FDA and EMA labeling. Meanwhile, 169 drugs were contraindicated by either agency, among which 99 had hepatic ADR frequency information in the labeling. The labeling discrepancies in warnings for hepatic ADRs showed no statistical association with
Association between labeling discrepancies and occurrence of liver injury events
To address the possibility of clinical relevance, we investigated the association between labeling discrepancies and the frequency of hepatic adverse events by exploring the likelihood scores of drugs collected in the LiverTox database.
Our analysis of the LiverTox database suggested that discrepancies in warnings for hepatic ADRs were not statistically associated with drugs commonly associated with DILI or a higher number of well-documented case reports for DILI (Table 4). Next, we investigated
Discussion
In this study, we systematically compared the hepatic safety information in FDA and EMA drug labeling sections describing warnings, precautions, and contraindications. We considered two scenarios: discrepancies in labeling with hepatic ADR warnings and contraindications because of liver disease. We also considered ADR frequency (Table 2) and ADR severity (Table 3). Specifically, we found that 8.7% of the 549 drugs evaluated differed in hepatic ADR warnings, whereas the discrepancy in labeling
Concluding remarks
In conclusion, we found limited discrepancies regarding hepatic safety information in FDA- and EMA-approved drug labeling. Discrepancy in hepatic ADR warnings was calculated to be 8.7%. Discrepancy in contraindications for liver disease was estimated to be 21.3%, including 4.9% or a quarter with hepatic ADR warning and 16.4% or three-quarters without hepatic ADR warning, respectively. By considering combined ADR warnings and contraindications, the FDA and EMA had concordance of hepatic safety
CRediT authorship contribution statement
Yue Wu: Data curation, Methodology, Writing – review & editing. Wenzhong Xiao: Writing – review & editing. Weida Tong: Funding acquisition, Writing – review & editing. Jürgen Borlak: Methodology, Writing – review & editing. Minjun Chen: Conceptualization, Data curation, Methodology, Funding acquisition, Writing – review & editing.
Acknowledgements
We thank Peng Li and Chloe Cheng for help in data analysis and Joanne Bergerfor manuscript editing assistance. The research is internally funded by the project (E0767701) at National Center for Toxicological Research, US FDA. The views expressed in this manuscript do not necessarily represent those of the FDA.
Competing interests
The authors have no conflicts to disclose.
References (39)
- et al.
Discrepancies in liver disease labeling in the package inserts of commonly prescribed medications
Gastroenterology
(2015) - et al.
Assessment of PRO label claims granted by the FDA as compared to the EMA (2006–2010)
Value Health
(2013) - et al.
Differences in product information of biopharmaceuticals in the EU and the USA: implications for product development
Eur J Pharm Biopharm
(2005) - et al.
Drug-related morbidity and mortality: updating the cost-of-illness model
J Am Pharm Assoc
(2001) - et al.
‘Black box’ 101: how the Food and Drug Administration evaluates, communicates, and manages drug benefit/risk
J Allergy Clin Immunol
(2006) - et al.
DILIrank: the largest reference drug list ranked by the risk for developing drug-induced liver injury in humans
Drug Discovery Today
(2016) - et al.
Comedications alter drug-induced liver injury reporting frequency: data mining in the WHO VigiBase™
Regulat Toxicol Pharmacol: RTP
(2015) - et al.
Drug-induced liver injury in patients with preexisting chronic liver disease in drug development: how to identify and manage?
Gastroenterology
(2016) - et al.
Risk factors for idiosyncratic drug-induced liver injury
Gastroenterology
(2010) - et al.
Are patients with elevated liver tests at increased risk of drug-induced liver injury?
Gastroenterology
(2004)