.webp)
1. Herb-Drug Interactions Database: Background
Herbal remedies continue to see rapid growth in usage worldwide (1). The World Health Organization (WHO) reports that approximately 60% of the world's population relies on traditional herbal remedies as complementary or alternative medicine. In developing countries, this reliance is even more significant, with 80% of the population depending almost entirely on herbal remedies for their health care needs.
In the United States, dietary supplements, including botanicals, are commonly used. The latest National Health Interview Survey reveals that around 58% of American adults utilize some form of dietary supplement. Among those aged 40–59, botanicals rank as the fourth most common supplement at 8.3%, and fifth for those aged 20–39 at 5.1%. (2)
As a result, there is growing concern about the safety of potential Herb-Drug Interactions (HDIs) due to a limited understanding of the mechanisms of plant compounds and mixtures.
A More Efficient Path to Successful Medication Therapy
Supplements are often utilized for health maintenance, or to delay or enhance the effect of pharmaceutical drugs. However, despite their potential effectiveness, supplements, such as herbs, can have adverse side effects and may interact with drugs if taken simultaneously. To fulfill their intended purpose, supplements must demonstrate both efficacy and safety.
A database that anticipates harmful interactions between a plant and a drug can help in achieving this goal. When such an interaction is detected, healthcare practitioners can make adjustments to the dosage or formulation, or suggest alternative supplements that do not interact with the prescribed medication.
Currently, there is a shortage of comprehensive databases documenting interactions between plants and modern allopathic drugs. Therefore, we have created an accessible and well-organized database intended for use by healthcare practitioners and manufacturers of phyto-drugs or dietary supplements, featuring validated HDIs. This database will support HDI analysis within Synapse's prescription assistance platform, which is used by healthcare practitioners.
The new database we propose is a global tool that focuses on HDIs and is presented in a structured format. It is named PHYDGI, which stands for PHYtotherapy DruG Interaction. PHYDGI evaluates data based on criteria for reliability in predicting potential adverse events (AEs) and the strength of pharmacokinetic interactions.
It is important to note that we are currently developing PHYDGI for France, with plans to make it available for the U.S. market in the near future.
2. Inside the PHYDGI Database: Understanding the Scientific Validation Process
PHYDGI Database: A Comprehensive Look Inside
The PHYDGI database offers information on herbal entities and their interactions, along with the degree of pharmacokinetic interaction strength. Similar to other databases, HDI information is extracted from scientific literature, with the additional inclusion of case reports from the French pharmacovigilance database. Comprehensive details about the experimental conditions are provided, and references to the original sources of HDI information are consistently included.
The PHYDGI database provides two distinct scales: one for assessing the strength of HDIs based on pharmacokinetic data, and another for evaluating the quality of evidence from the information source.
Noteworthy is the absence of specific clinical recommendations, given that the clinical significance of an HDI is contingent upon the patient's medical history and comorbidities. Consequently, the assessment of such a multifaceted interaction necessitates the application of medical judgment.
HDI data are structured in a format that readily integrates into a clinical decision support system’s drug interaction analyzer.
Ongoing evaluation and augmentation of content are carried out by domain experts, with the inclusion of newly discovered HDIs from recent studies. It is imperative to highlight that the database undergoes annual updates to ensure currency and relevance.
Analysis of Existing Herb-Drug Interaction Databases
Before the creation of PHYDGI, we examined existing HDI tools, drawing upon validated information retrieved from PubMed, and analyzed them for content. We identified fifteen databases that reported HDIs, and subjected them to a comprehensive evaluation.
The content of these databases exhibited considerable diversity. Some tools concentrated on plants of specific origins, such as Traditional Chinese Medicine (TCM), while others included only commonly used plants like St. John’s Wort, Milk Thistle, Garlic, Ginkgo, among others. Examination of the data sources in these databases revealed that the information was derived from a variety of sources, including original scientific articles, agency reports, monographs, and other HDI databases.
Most of these databases did not employ organized data structures but instead presented data through unstructured text. We observed that e-tools based on these databases were primarily used by medical doctors, with some specialized specifically for oncologists or nephrologists. In terms of accessibility, tools available to both health practitioners and patients were mostly found on free websites.
One critical observation was the inconsistent approach to assessing the clinical relevance of an HDI. For example, this inconsistency manifested in the adoption of either a three-level risk classification (high, intermediate, and low), or a five-level risk classification (high, medium, weak, none, unknown). Given the absence of a comprehensive and standardized classification for herb-drug interaction risks, HDI risk assessment was found to be highly dependent on the specific database being used. After a thorough review, we concluded that no existing online resource comprehensively represented HDI knowledge in a structured and standardized format. In response to this observed inconsistency in HDI risk assessment, we initiated the development of PHYDGI, aimed at providing a consistent and standardized solution.
Data collection, organization, and grading
Data collection
Research articles containing relevant data on HDIs are manually extracted from the scientific literature by Master’s students specializing in Health Sciences at Bordeaux University. Two phytotherapy experts review the extracted data: a pharmacist and a senior scientist.
Collection and Analysis of HDI Case Reports from French Pharmacovigilance Centers
We’ve established a collaboration with the French Pharmacovigilance Centers (CRPV) to integrate case reports concerning interactions between plants or food supplements and drugs. This retrospective analysis of the national pharmacovigilance database will be updated annually. The inclusion criteria encompass cases where plants or food supplements have been categorized as ‘Suspect’ or ‘Interaction.’ Plant/food supplements mentioned solely in case commentary and not receiving a categorization are not included in the data.
Data Organization
The PHYDGI database is organized into four major sections: Herbs, Drugs, Interactions, and Sources. Each section is further divided into subcategories.
Herbs: Vernacular name, Latin name, family, active ingredients (identifying the specific plant parts and molecules involved), active dose, and duration of the active dose associated with the interaction are all present in the database. Particular attention is paid to delineating the nature of the plant extract used and, when available, specifying the substance involved in the HDI.
Drugs: We include international non-proprietary names, dosage information, and the duration of the active doses.
HDI: Included is the mechanism of interaction (i.e. whether it’s pharmacokinetics or pharmacodynamics) and its effects. We also included pharmacokinetics values (Cmax and AUC) and a strength scale for pharmacokinetic interactions based on the pre-cited value.
Source: We provide access to both the reference and the quality of the evidence, ensuring that users have access to reliable and well-documented sources of information.
Grading
We think it’s important to distinguish several parameters when stratifying the clinical risks associated with herb-drug interactions: Firstly, the quality of the evidence supporting the interaction is considered. Secondly, the strength of the pharmacokinetic interactions. Third, the incidence of these adverse reactions, and lastly, the presence of risk factors that may heighten the severity and/or frequency of adverse reactions. To comprehensively address these factors, the PHYDGI database provides 2 grading scales for each HDI: One for assessing the strength of pharmacokinetic HDIs and another for determining the reliability of the information source.
Quality of Evidence
Evidence is graded on a scale from 0 to 4, and descriptive indicators are included, with the goal of providing a consistent framework for stratifying HDIs. The below table highlights the grading scale:
Pharmacokinetic Herb-Drug Interaction Strength
Since there is no consensus on how to precisely characterize the strength of an HDI, we’ve devised a method to quantify the strength of pharmacokinetic interactions using mathematical formulas. This approach is designed to be user-friendly and easily comprehensible.
The strength scale we’ve developed is grounded in the extent of exposure to a drug. Consequently, we’ve established three distinct levels: low, moderate, and high, primarily based on alterations in the AUC value of the drugs involved..5
Low: When the HDI results in a modification of the drug's AUC by less than 50%, the interaction is categorized as having low strength.
Moderate: If the HDI leads to a modification of the drug's AUC value between 50% and 100%, it falls into the moderate strength category.
High: When the HDI significantly alters the drug's AUC value by more than 100%, it is classified as a high-strength interaction.
Note: The strength code should not be mistaken for a severity code that directly correlates with the clinical significance of an HDI. Indeed, even a slight increase in AUC may cause serious adverse effects, while a substantial increase in AUC may not necessarily result in clinically relevant adverse effects. In addition, an HDI may not influence the AUC, but could impact the interaction between the drug and the receptor, thereby affecting the drug’s efficacy. Therefore, it is essential to consider the strength scale in conjunction with other critical factors, including the mechanism of interaction, the specific drug involved (i.e. drugs with narrow therapeutic indices), and the patient (i.e. those at higher risk), to mitigate the risk of severe/acute clinically relevant HDIs.
Moreover, when evaluating pharmacokinetic HDIs and their strength codes, it's imperative to take into account their incidence, as determined by the number of reports regarding interactions between identical pairs of plants and drugs. Essentially, the clinical impact of an HDI at the population level hinges not solely on its strength but also on its potential frequency of occurrence. The database maintains records of the repetition of HDI reports but does not merge this data with the HDI's strength.
In summary, decisions concerning the therapeutic management of an HDI should be made through informed medical judgment. Conversely, it is not feasible to establish simple strength levels for pharmacodynamic interactions due to their significant heterogeneity and multifaceted nature.
3. Combining the PHYDGI Database with Synapse Medicine’s Clinical Decision Support Technology
Organic Does Not Equal Safety
Clinical manifestations of HDIs can vary greatly, potentially leading to increased toxicity or reduced treatment efficacy, thus posing a risk of treatment failure.. In cases involving anticancer treatments, such treatment failures can result in a significant loss of survival opportunities.
Pharmacokinetic-based drug interactions may arise due to alterations in the absorption, distribution, metabolism, and/or excretion (ADME) pathways. Among these, the most prevalent mechanism underlying pharmacokinetic interactions is the inhibition of drug-metabolizing enzymes by other substances.
Grapefruit juice (GFJ), a popular breakfast juice in the US, is a classic example of an enzyme-mediated botanical herb-drug interaction. Numerous interaction studies have demonstrated that the consumption of GFJ can substantially increase drug exposure, with as much as an 85% to 300% increase in AUC (Area Under the Curve), for drugs such as simvastatin, nisoldipine, saquinavir, and cyclosporine. This effect is primarily attributed to the inhibition of the CYP3A4 enzyme.
HDIs may also occur because of increased expression/upregulation of a protein involved in drug transport, a phenomenon referred to as induction. The most common mechanism of induction is a ligand-dependent binding and activation of nuclear receptors that function as gene transcription factors, such as PXR (pregnane X receptor). Notably, hyperforin, the principal active constituent of St. Johnʼs Wort, is the most potent agonist for human PXR.
Schematic representation of selected examples of possible mechanisms for enzyme- and transporter-mediated botanical–drug interactions in the intestinal epithelial (a) and liver cells (b) with effect on victim drug exposure.6
Preventing Herb-Drug Interactions with Synapse’s Drug Interaction Component
It is important to keep in mind the substantial challenges associated with preventing HDIs. Caregivers often do not readily suspect that food supplements may be responsible for AEs. This is primarily due to their lack of awareness regarding their patients' use of herbal remedies and the possibility of interactions with conventional drugs. The intricate nature of these interactions further compounds the issue.
To enhance the effectiveness of medication analysis, there is a pressing need for the development of sensitive and highly specific software programs dedicated to interaction detection. This is where the integration of the PHYDGI database into Synapse's Drug Interaction component proves invaluable, as it serves as a rich source of information on HDIs. This integration is poised to significantly optimize medication safety and contribute to more informed and accurate healthcare decisions.
4. Conclusion: Improving Prescribing and Dispensing Practices with PHYDGI
The PHYDGI database at-a-glance:
- Structured scientific data on HDIs, complete with scientific source references.
- In-depth insights into herbal entities and their interactions, including pharmacokinetic interaction strength
- Integrated within Synapse’s Drug Interaction component.
- The combination of the PHYDGI database and Synapse’s Drug Interaction component contributes to reducing potential HDIs and promoting the safe use of plant-based remedies or supplements.
In the face of the growing reliance on herbal remedies and dietary supplements, the need for a profound understanding of HDIs has never been more urgent. While natural remedies offer promising therapeutic benefits, they are not devoid of risks. The intricate interactions between herbs and conventional drugs can lead to increased toxicity, reduced efficacy, and, in severe cases, treatment failure.
The PHYDGI Database, as outlined in this article, represents a significant advancement in the realms of herbal medicine and pharmacology. By providing well-structured scientific data on HDIs, including robust referencing and detailed information on herbal entities, it bridges the knowledge gap that has long existed in this area. By fostering collaboration amongst researchers, pharmacists, and healthcare providers, and by harnessing the power of Synapse’s cutting-edge clinical decision support technology, the PHYDGI Database serves as a beacon of scientific rigor and innovation.
Sources
- Ahmad Khan, M.S., Ahmad, I., 2019. Chapter 1 - herbal medicine: current trends and future prospects. In: Ahmad Khan, MS, Ahmad, I, Chattopadhyay, D (Eds.), New Look to Phytomedicine. Academic Press, pp. 3–13. https://doi.org/10.1016/B978-0- 12-814619-4.00001-X.
- Mishra S, Stierman B, Gahche JJ, Potischman N. Dietary supplement use among adults: United States, 2017–2018. NCHS Data Brief, no 399. Hyattsville, MD: National Center for Health Statistics. 2021. DOI: https://doi.org/10.15620/cdc:101131
- Anses, 2011. relatif `a la construction d’une m ́ethode d’imputabilit ́e des signalements d’effets ind ́esirables de nutrivigilance. Saisine n◦2010-SA-0195.
- De Smet, et al. (2007) : De Smet, P.A.G.M., 2007. Clinical risk management of herb-drug interactions. Br. J. Clin. Pharmacol. 63 (3), 258–267. https://doi.org/10.1111/j.1365-2125.2006.02797.x.
- J. Perrot1 , C. Bennetau-Pelissero2,3,4 ,G. Miremont-Salame5,6, F. Petitet7, S. Cluzet8, H. Peyrouzet6, L. Letinier1 (1) Synapse Medicine, France ; (2) University of Bordeaux, Bordeaux, 33070 France ; (3) U1212 Inserm, UMR Inserm U1212, CNRS 5320, France ; (4) Bordeaux Sciences Agro, Gradignan, 33175 France ; (5) INSERM, BPH, U1219, Team Pharmacoepidemiology, Univ. Bordeaux, Bordeaux, France ; (6) CHU de Bordeaux, Pole de Sante Publique, Service de Pharmacologie Medicale, Centre de Pharmacovigilance de Bordeaux, Bordeaux, France ; (7) Herbeo Bordeaux, 33000 France ; (8) Univ. Bordeaux, INRAE, Bordeaux INP, Bordeaux Sciences Agro, OENO, UMR 1366, ISVV, Villenave d’Ornon, 33140 France
- Grimstein, Manuela, and Shiew-Mei Huang. “A regulatory science viewpoint on botanical-drug interactions.” Journal of food and drug analysis vol. 26,2S (2018): S12-S25. doi:10.1016/j.jfda.2018.01.013
