Advancements in toxicological risk assessment: integrating Ferguson’s principle, computational models, and drug safety guidelines, a comprehensive framework for improving risk assessment and resource management in toxicology

Abstract

This investigative study examines the transfer of maternal medications into breast milk and their potential impact on breastfeeding infants. Significant factors influencing drug transfer, including physiochemical properties and milk composition, are analysed to corroborate judicious drug administration in nursing mothers. The study investigates, evaluates, and interprets drugs such as: H|chlorpromazine (New England Nuclear [NEN]), diazepam Roche, C|diclofenac (Ciba-Geigy, 6.6 mCi/mmol, K-277), diclofenac (Ciba-Geigy, 0.1317), digoxin (Wellcome, 11725), fluphenazine (Squibb 12240), phenytoin (NEN, 46 Ci/mmol, 2315-061), phenytoin (Parke-Davis 5419972), pirenzepine (Boehringer-Ingelheim-660206), H|prednisolone (Amersham, 67.4 Ci/mmol, 88), warfarin (Amersham, 46 mCi/mmol, 30), outlining and assessing their transferability and perils notably presented. Ferguson’s principle was leveraged to predict drug toxicity, specifically for central nervous system depressants, elucidating drug lethality and safety evaluation. On top of that, advancements in toxicological risk assessment were evaluated, articulated as focusing on naloxone programs, predictive modelling, quantitative structure–activity relationship (QSAR) applications, toxicogenomics, and ordinary differential equation (ODE) models. The comparison between risk assessments and biological monitoring highlights the prominence of evaluating internal dosages. Progress in 3D-QSAR modelling augmented its role in forecasting chemical toxicity, while advancements in toxicogenomics and the application of ODE models have contributed to toxicological research. Hence, the shift toward alternate toxicity assessment methodologies was driven by ethical concerns, budgetary limits, and the demand for more human-relevant data without sacrificing an animal life, which was a concern of the present scientific investigation; fixed by machine algorithms, e.g. random forest, Support Vector Machine (SVM), Ferguson’s principle, etc.; an omics data set for correlation through tactile programmed computational heuristics for decision science.

Background

Vigilance is crucial to addressing the diverse sources of adulterants in the industry.

Recent progress in toxicological research has resulted in the creation of numerous novel methods for assessing the safety and possible risks connected to chemicals and toxic substances that have been developed.¹ Toxicity can present itself in different ways and is often evaluated using quantitative measures such as the lethal dose (LD₅₀) or qualitative classifications like low, moderate, or high toxicity. These evaluations take into account several factors, including the route of exposure, frequency of contact, chemical properties, and biological responses of the test subject. Assessing toxicity is crucial for determining the potential risks that chemicals pose to human health and the environment. Traditionally, toxicity testing has relied on animal studies to detect harmful effects. However, concerns regarding ethical implications, high costs, and the limited applicability of animal data to human physiology have led to increasing scrutiny of this method.² Diligent monitoring is key to managing the wide range of adulterant sources of toxicity in the industry.³ Additivity is a key concept in mixture toxicity, where substances may have additive effects by targeting the same molecular mechanism within a common cell. For instance, dioxin-like compounds can act additively in this manner. Another concept involves an “enhancer substance,” which may amplify the adverse effects of a “driver substance” by increasing its concentration at the target site. Effective risk management of individual substances is crucial for reliably preventing harmful effects in humans under both scenarios.⁴

Using chemical compounds to eliminate weeds in crops is fundamentally a matter of selective toxicity. A selective herbicide must effectively damage the unwanted plants while minimising harm to the crop tissues. Recent scientific advancements of economic significance have largely focused on this challenge of selective toxicity. In the fields of chemotherapeutics and pharmacology, significant success has been achieved in managing diseases and alleviating pain in both humans and animals.⁵

Several studies have explored the relationship between the chemical or physical properties of substances and their toxic effects. Notable examples include Moore’s (1917) research on insect fumigants, Tattersfield and his colleagues’ extensive studies on contact insecticides and fumigants (Tattersfield et al., 1920, 1925, 1926, 1927), and the bacteriological research by Tilley and Schaffer (1926, 1928) as well as others (Coulthard, Marshall, and Pyman, 1930; Dohme, Cox, and Millar, 1926; Klatman, Gatyas, and Shternov, 1931). Additional work focused on fungicides (Morris and Stirk, 1932; Stiles and Rees, 1935). These investigations typically involved assessing a range of chemical compounds to determine the molar concentrations required to achieve equivalent toxic effects on a chosen organism. The toxicity of these compounds was then generally inferred to be inversely related to these equi-toxic concentrations.⁶

The increasing prevalence of breastfeeding emphasises the necessity for evidence-based guidelines, consolidating current knowledge on drug transfer into breast milk to offer practical recommendations for nursing mothers’ safety and their infants’ well-being. Meanwhile, Ferguson’s principle offers a comprehensive framework, emphasising the quantitative relationship between chemical structure and drug toxicity, providing valuable insights into predicting, and preventing drug-induced fatalities.

Toxicokinetics and toxicodynamics are analogous to pharmacokinetics and pharmacodynamics, respectively. Healthcare professionals should recognise that data from pharmacokinetics and pharmacodynamics may not always apply in cases of poisoning or overdose. Complementarily, the typical classification of a substance as toxic or nontoxic does not align with medical toxicology practice. As Paracelsus, a Renaissance physician, famously noted, “There is no substance that is inherently a poison; all things are poison, and nothing is without poison. Only the dose determines whether a substance is poisonous.”⁷

The principle “The dose makes the poison,” attributed to Paracelsus in the 16th century, is a cornerstone of toxicology. Applying this principle to mechanistic toxicology can be challenging because a single compound can influence multiple molecular pathways at varying doses, often resulting in nonlinear dose–response relationships. Typically, mechanistic studies use high doses of xenobiotics to clearly identify disruptions in specific molecular pathways. This approach overlooks the fact that lower doses, more relevant to actual human exposures, might also impact different molecular mechanisms. To refine mechanistic toxicology, it propose adapting Paracelsus’s principle to reflect its core idea: “The dose disrupts the pathway.” This perspective acknowledges that many environmental and occupational xenobiotics affect diverse molecular pathways, that most pathways are sensitive to chemical exposure, and that different doses impact different pathways.^8,9 (Table 1).

Introduction

The use of medications during lactation presents a critical concern due to the potential transfer of drugs into breast milk and their subsequent effects on breastfeeding infants. Understanding the factors that influence drug transfer, including physiochemical properties, milk composition, and maternal pharmacokinetics, is essential for ensuring infant safety. While breastfeeding offers significant health benefits, the inadvertent exposure of infants to medications necessitates careful risk assessment (RA) and evidence-based guidelines for maternal drug use.^10–13

This study aims to evaluate the extent of drug transfer into breast milk and assess the potential risks posed by commonly prescribed medications, including drugs such as chlorpromazine, diazepam, diclofenac, digoxin, fluphenazine, phenytoin, pirenzepine, prednisolone, and warfarin. By applying Ferguson’s principle, the research also explores predictive models for assessing drug toxicity, particularly for central nervous system depressants.^10–13

The findings contribute to toxicological RA by integrating biological monitoring, quantitative structure–activity relationship (QSAR) modelling, toxicogenomics, and ordinary differential equation (ODE) models. This study highlights the need for informed clinical decision-making, providing healthcare professionals with evidence-based recommendations to minimise medication risks while maintaining the benefits of breastfeeding.

Drug toxicity evaluation is a critical aspect of pharmaceutical research, ensuring the safety and efficacy of medications before they reach the market. Comprehensive toxicity assessments, including preclinical studies and clinical trials, are essential for identifying adverse drug reactions (ADRs) and mitigating risks. ADRs range from mild effects to severe complications, such as hospitalisation or death, and are classified based on pharmacological mechanisms, dose–response relationships, and immune-mediated reactions. Regulatory agencies such as the World Health Organisation (WHO) and Food and Drug Administration (FDA) provide guidelines for ADR classification, emphasising the importance of continuous drug safety monitoring.^14–17

Among the various principles used to understand drug toxicity, Ferguson’s principle has played a significant role in assessing drug activity and pharmacokinetics. Originally proposed in the 1930s, Ferguson’s principle explores the relationship between narcotic activity, partition coefficients, and thermodynamic factors. This framework has been instrumental in predicting drug distribution in biological systems, including breast milk, and guiding pharmacokinetic research.^18–20

This study aims to investigate the application of Ferguson’s principle in predicting drug toxicity, particularly for volatile airborne chemicals and central nervous system depressants, while also examining the relationship between drug physicochemical properties and biological effects using QSAR modelling. Over and above that, it evaluates the effectiveness of Ferguson’s principle in toxicology, focusing on drug distribution, narcotic potency, and receptor interactions and explores alternative mechanistic approaches to supplement or replace empirical observations derived from Ferguson’s principle.^15–19

Advancing toxicity prediction: the relevance of Ferguson’s principle in modern toxicology for predicting narcotic potency, toxicity, and drug physicochemical interactions using QSAR modelling

1. Investigate the application of Ferguson’s principle in predicting drug toxicity, particularly for volatile airborne chemicals and central nervous system depressants.

2. Examine the relationship between drug physicochemical properties and biological effects using QSAR modelling.^21–23

3. Evaluate the effectiveness of Ferguson’s principle in toxicology, with a focus on drug distribution, narcotic potency, and receptor interactions.

4. Explore alternative mechanistic approaches to supplement or replace empirical observations derived from Ferguson’s principle.

The findings contribute to advancements in toxicological modelling, refining the understanding of drug safety through interdisciplinary approaches combining pharmacology, biochemistry, and computational techniques. By integrating traditional pharmacological principles with modern predictive tools, this study aims to enhance drug RA, optimise therapeutic interventions, and provide a more precise framework for evaluating drug-induced toxicity.

QSAR models: advancing QSAR models: a computational approach and mathematical approaches to predicting chemical and biological interactions for drug development and environmental RA

Computational molecular modelling has become an indispensable tool in drug discovery, gaining widespread adoption within the pharmaceutical industry. As modern drug researchers navigate the evolving landscape of drug development, they will frequently engage with molecular modelling techniques. Therefore, acquiring a strong foundational knowledge of these tools and methodologies is essential for key success in the fields of chemistry and medicine.²⁴

QSAR models are computational techniques that mathematically link the structural characteristics of chemical compounds to their biological activity or physicochemical properties. By evaluating molecular descriptors, which quantitatively represent molecular features, QSAR models help predict how variations in structure affect a compound’s behaviour. These models play a crucial role in drug discovery and environmental chemistry, enabling the prediction of biological activities for novel compounds based on their molecular attributes. A well-developed QSAR model allows for the estimation of the activity of untested compounds, aiding in the design of new molecules with desirable properties. However, it is essential to define the model’s applicability domain—the range of chemical space within which predictions remain reliable—to prevent inaccurate extrapolations.^21–23

Defining QSAR and computational models, QSAR models

QSAR models are mathematical frameworks that establish correlations between chemical structures and their biological or physicochemical properties. They serve as predictive tools in pharmaceutical research and environmental toxicology by forecasting how new compounds will behave based on known structural attributes. Computational models: Computational models encompass a wide range of mathematical and algorithmic approaches designed to simulate and analyse complex systems. In chemistry and biology, computational models include QSAR modelling, molecular dynamics simulations, quantum mechanics-based calculations, and other predictive techniques. These models help in understanding biological interactions, optimizing chemical properties, and assessing potential toxicity.^21–24

Recommendation

To ensure clarity, these definitions should be introduced in the early sections of a study. For instance: This research employs QSAR models, which mathematically correlate chemical structures with biological activity. Likewise, a variety of computational approaches, including QSAR modelling and other simulation techniques, are used to predict and assess chemical compound properties.^21–23

Advancements in QSAR modelling recent progress in QSAR methodologies has incorporated diverse computational techniques to enhance predictive accuracy and interpretability. The integration of QSAR with molecular docking studies and Absorption, Distribution, Metabolism, Excretion, and Toxicity analysis has significantly improved the in silico evaluation of drug candidates.^21–23

Thus, QSAR models remain indispensable tools in predicting the biological activity of chemical compounds. Their applications extend across drug design, environmental RA, and chemical optimisation, offering valuable insights that aid in developing safer and more effective compounds.^21–23

Adherence to organisation for economic cooperation and development principles

For regulatory acceptance, QSAR models must adhere to the organisation for economic cooperation and development principles, which include a defined endpoint, an unambiguous algorithm, a defined domain of applicability, appropriate measures of goodness-of-fit, robustness, and predictivity, and a mechanistic interpretation, if possible.²⁵

Unraveling missing heritability: the role of systems biology in high-throughput omics data analysis

Despite advancements in high-throughput technologies that facilitate the generation of large omics datasets across various biological disciplines, analytical methods for processing such data remain underdeveloped. The diverse range of available omics data types presents an opportunity to adopt a systems biology approach, which is crucial for unraveling the genomic intricacies of non-Mendelian traits. The concept of “missing heritability” in complex traits may stem, in part, from the predominant reliance on linear models within individual data types, which restricts discoveries to variants that independently correlate with disease.^26,27

Enhancing feature selection in high-dimensional data: evaluating random forest and random forest-recursive feature elimination in omics analysis

Random forest (RF) is a machine tech approach well-suited for high-dimensional data and capable of modelling nonlinear relationships among predictors. However, its performance can be affected when predictors exhibit strong correlations, potentially limiting its ability to identify key variables. To address this limitation, the random forest-recursive feature elimination (RF-RFE) algorithm has been developed, particularly for smaller datasets. However, its effectiveness in large-scale omics datasets remains uncertain.²⁶

A study integrating 202,919 genotypes and 153,422 methylation sites from 680 individuals evaluated the ability of RF and RF-RFE to detect simulated causal associations, including genotype–methylation interactions, linked to triglyceride levels. Findings indicated that RF effectively identified strong causal variables when only a few correlated predictors were present, but it struggled to detect additional causal variables. While RF-RFE reduced the importance of correlated variables, it also diminished the importance of causal ones in highly correlated datasets, making their detection challenging. These results suggest that RF-RFE may have scalability limitations when applied to large, high-dimensional omics data.²⁶

Data can be documented as:

• Individual_ID: Unique identifier for each individual

• Genotype_1, Genotype_2, ..., Genotype_10: Simulated genotype values (binary: 0 or 1)

• Methylation_1, Methylation_2, ..., Methylation_10: Simulated methylation levels (continuous values between 0 and 1)

• Triglyceride_Level: Simulated triglyceride levels (continuous values within a physiological range)

  (Additional supporting data can be found in the supplementary materials.)

Comprehending drug toxicity through the lens of Ferguson’s principle

In the realm of pharmaceutical research and development, the evaluation of drug toxicity is paramount to ensuring the safety and efficacy of medications for human use. Prior to introducing a medicinal product to the market, rigorous assessments are conducted to verify its safety profile, quality, and therapeutic efficacy. This comprehensive evaluation process typically includes preclinical assessments and clinical trials, overseen by regulatory agencies and clinical experts.^18–20

A fundamental aspect of drug toxicity evaluation involves the classification and characterisation of ADRs. ADRs encompass a spectrum of unintended and harmful responses to medications, ranging from mild adverse effects to severe outcomes such as hospitalisation or death. Various classification systems exist to categorise ADRs based on factors such as pharmacological effects, dose relationship, and causal relationship with the drug.^18–20

One notable classification system, introduced by Henri Manasse, distinguishes between different types of ADRs, including predictable type A reactions and unpredictable type B reactions, which are often immune-mediated. Moreover, regulatory bodies such as the WHO and the FDA provide definitions and guidelines for reporting and assessing ADRs, emphasising the importance of monitoring drug safety throughout the drug development process and postmarketing surveillance.^18–20

In parallel with advancements in drug toxicity evaluation, principles such as Ferguson’s principle have contributed to our understanding of drug activity and pharmacokinetics. Proposed by Ferguson in the 1930s, this principle elucidates the relationship between narcotic activity, partition coefficient, and thermodynamics. By considering the equilibrium between drug concentrations in different phases and thermodynamic principles, Ferguson’s principle offers insights into drug activities and distribution in biological systems.^18–20

Report introduces a theoretical framework for the Ferguson principle, specifically focusing on the acute toxic effects of volatile airborne chemicals, such as sensory irritation. Data involving both nonreactive and reactive chemicals are employed to illustrate that the Ferguson principle can be expanded to include reactive chemicals by incorporating descriptors related to chemical reactivity alongside the physicochemical parameters initially used by the Ferguson principle.²⁸

Ferguson’s principle has been instrumental in guiding drug development and pharmacokinetic studies, particularly in understanding the distribution of medications into breast milk and assessing risks to nursing infants. Through empirical research and quantitative analyses, researchers have expanded upon Ferguson’s principle to develop predictive models and computational tools for estimating drug distribution and assessing safety during lactation.

In this context, the integration of traditional pharmacological principles with modern computational techniques emphasises the interdisciplinary nature of drug toxicity evaluation. By leveraging insights from pharmacology, biochemistry, and computational modelling, researchers and healthcare professionals aim to enhance our understanding of drug safety and optimise therapeutic outcomes for patients.^18–20

The Ferguson principle, which posits that the ratio Pnar/PO (where Pnar represents the partial pressure of compounds causing a specific effect through a physical mechanism in a system, and PO denotes the saturated vapour pressure of the liquid narcotic) remains constant for a series of nonreactive narcotics or toxicants in a given system, is scrutinised. Contrary to the claims of Brink and Posternak, it is demonstrated that this principle lacks a thermodynamic foundation. However, conditions under which Pnar/PO might appear relatively constant, as an empirical observation, are outlined. It is shown that this observation aligns with the scenario where the solubilities of liquid narcotics in the receptor area are relatively uniform.²⁹

The success of this approach depends on the identification of specific chemical reactivity mechanisms for the reactive chemicals under investigation. The findings indicate the possibility of replacing the empirical Ferguson principle with formal mechanistic equations, offering a more robust understanding of the mechanisms behind the sensory irritation caused by airborne chemicals.²⁸

An analysis of the relationship between agonist descriptors and biological effects is conducted using three straightforward biological models. The study reveals that the biological activity of nonreactive gases and vapors may be influenced by the equilibrium established between the agonist in the gas phase and the agonist within a receptor, or alternatively, by the equilibrium between the agonist in the gas phase and the agonist within a receptor phase. Besides, using Abraham’s solvation equation, it is demonstrated that specific solvents can be selected to replicate the chemical characteristics of the receptor or receptor phase. In the case of upper respiratory tract irritation in male Swiss OF1 mice, solvents such as N-formylmorpholine, a trialkylphosphate, and wet octanol are identified as mimicking the receptor’s properties, whereas water does not.²⁹

The Ferguson principle has been extensively utilised in toxicology to differentiate and suggest potential mechanisms responsible for the acute toxic effects of chemicals. Despite its widespread use, the principle has not been thoroughly validated due to the unavailability of a comprehensive database containing a sufficient variety of both nonreactive and reactive chemicals. Recently, such a database has been developed.²⁸

A QSAR was developed following Hansch’s methodology to assess the acute toxicity of nine aromatic hydrocarbons (benzene, toluene, ethylbenzene, o-xylene, m-xylene, p-xylene, isopropylbenzene, n-propylbenzene, and butylbenzene) in a freshwater fish species native to Argentina. Solubility, molecular weight, and the octanol–water partition coefficient (K(ow)) were used as key macroscopic molecular descriptors. The slopes from regression analyses were found to be comparable to those for the reference fish species, Pimephales promelas. A new concept of nonpolar narcosis power (NP) was introduced, based on the toxic action mechanism of the tested chemicals. Drawing from Ferguson’s principle and the critical volume hypothesis for nonpolar narcosis, NP was defined as log MW * K(ow). This approach provided a better fit for experimental data compared to regression analysis using only log K(ow). The NP concept improved the quantitative understanding of the relationship between nonpolar narcotic compounds and their toxicity, making it more effective in explaining the physiological mode of action of the tested chemicals.¹⁷

The nonspecific depressant effects of a homologous series of primary alcohols, ranging from methanol to octanol, were studied across four distinct biological systems. The experimental results demonstrated that the log-concentration response curves for alcohols were linear across most of their range, and for any single system, remained parallel throughout the series. As the chain length of the alcohols increased, the concentration needed to achieve a specific effect decreased logarithmically. However, the rate of this decrease varied across the different biological systems and was consistently slower than the decline in solubility with increasing chain length. Notably, in two of the systems, alcohols beyond octanol exhibited no activity, a phenomenon known as the cut-off effect. The implications of these observations are discussed in relation to the mechanism of action for nonspecific depressants. Ferguson’s principle, which suggests using thermodynamic activity instead of concentration as a measure of effectiveness, was applied to the findings. An appendix compares these results with predictions from Mullins’ hypothesis on narcotic action, showing poor alignment between the two.³⁰

Rationale

With the increasing prevalence of breastfeeding, there is a growing need for evidence-based guidelines to inform healthcare decisions for nursing mothers. This study consolidates current knowledge on drug transfer into breast milk and provides practical recommendations to promote the safety of breastfeeding mothers and their infants. By raising awareness of potential risks and offering guidance on medication use during lactation, healthcare professionals can support breastfeeding mothers in making informed decisions about their health and the well-being of their infants.

Ferguson’s principle provides a comprehensive framework for understanding the quantitative relationship between chemical structure and the toxicity of drugs leading to fatality. By considering factors such as solubility and relative saturation, this principle offers practical insights into predicting and assessing the lethal toxicity of drugs in the bloodstream. The correlation between lipid solubility and drug toxicity reinforce the importance of lipid-phase accumulation in determining lethal toxicity, particularly for substances with nonspecific mechanisms of action (MOA) affecting the central nervous system. The application of Ferguson’s principle enables the computation of fatal toxicity levels for various substances, facilitating drug safety evaluation and the development of therapeutic interventions. Overall, Ferguson’s principle serves as a valuable tool in pharmacology, providing insights into drug mechanisms and aiding in the prediction and prevention of drug-induced fatalities.

Understanding and classifying ADRs: a comprehensive overview of monitoring, definitions, and classification systems

Prior to being introduced to the market in one or more member states, a medicinal product for human use typically undergoes thorough examination, including preclinical assessments and clinical trials. This comprehensive process aims to verify its safety, quality, and efficacy for utilisation within the intended population.^31,32

The drug development process involves various stages, starting with the development of drug molecules in research laboratories. These molecules undergo rigorous testing to ensure their quality and efficacy. Clinical studies are then conducted to evaluate their therapeutic effects. Throughout this process, multiple drug regulatory agencies are involved, and ongoing monitoring is essential.³³

Clinical experts such as pharmacists and physicians play a crucial role in monitoring the safety and efficacy of drug molecules or active pharmaceutical ingredients. They follow internationally accepted guidelines, such as those established by the International Council for Harmonisation. The goal of on-site monitoring is to ensure timely access to accurate information.³³

This continuous monitoring ensures that the healthcare industry remains reliable and trustworthy, dedicated to advancing human health and delivering meaningful outcomes.³³

ADRs represent a subset of adverse drug events (ADEs), which encompass various issues related to medication, including errors and other drug-related problems. ADEs denote the unfavourable outcomes resulting from drug-related incidents. Henri Manasse coined the term “drug misadventure” to describe the inherent risks associated with drug therapy when it is deemed necessary.^14,33,34

Both the WHO and Karch and Lasagna offer definitions of ADRs that share similarities. An ADR is characterised as any unintended and harmful response to a drug occurring at doses intended for prevention, diagnosis, or treatment, excluding instances where the intended purpose is not achieved. The FDA focuses on ADRs that are unexpected or result in significant morbidity, including outcomes such as death, life-threatening situations, hospitalisation, disability, congenital anomalies, or interventions to prevent permanent impairment or damage.^33,34

The Joint Commission on Accreditation of Healthcare Organisations (JCAHO) prioritises the reporting of significant ADRs, which encompasses those leading to morbidity, necessitating additional treatment, prolonging hospital stays, causing temporary or permanent disability, or resulting in death. The American Society of Health-System Pharmacists (ASHP) defines significant ADRs as unexpected, unintended, or excessive responses to a drug that may require discontinuation or modification of therapy, hospital admission, prolonged healthcare facility stays, supportive treatment, complicating diagnosis, or negatively impacting patient prognosis, potentially leading to temporary or permanent disability, harm, or death.^33,34 Thus, the data provided and its references in the package leaflet and summary of product characteristics will undergo evaluation within the established framework.^14,32,35,36

There are four main classification systems utilised to categorise ADRs. ADRs can be grouped based on the pharmacological effect of the drug into type A, B, C, and D reactions. Type A reactions involve an exaggerated yet normal pharmacological response to the drug, are predictable, and dose-dependent. On the other hand, type B reactions are not predictable based on the known pharmacological action of the drug and are not dose-related. Many type B reactions are hypersensitive or immune-based and can be further subdivided into type 1 (IgE-mediated), type 2 (IgG or IgM-mediated cytotoxic), type 3 (IgG-mediated immune complex), and type 4 (cell-mediated immune) reactions.³⁴

Type C reactions occur due to the prolonged use of a drug, while type D reactions involve delayed effects such as carcinogenicity or teratogenicity. ADRs can also be classified based on their relationship to the drug dose, categorised as dose-related or dose-unrelated reactions. Another classification system is based on the causal relationship between the reaction and the drug. One widely used classification, based on Naranjo’s descriptions, assesses the causal relationship between the reaction and the drug.³⁴

Evolution of drug safety regulations: harmonisation efforts, regulatory frameworks, and classification standards

In recent years, legislation concerning the safety of chemical substances for human health and the environment has been strengthened. Pharmaceutical and chemical companies must adhere to a range of laws at the national level, with both the United States and Europe having their own regulatory requirements, albeit with some differences in the application of safety parameters. However, there is generally mutual acceptance of safety data between the two regions.³⁷

The safety of chemicals is evaluated through standardised physicochemical, toxicological, and ecotoxicological studies by international evaluation agencies. RA of these chemicals is based on experimental data, chemical usage, and potential exposure to humans and the environment. In Europe, a well-established system has existed since 1981 to evaluate the classification and labeling of chemicals, including new and existing ones. In the United States, safety is regulated by the Toxic Substances Control Act and the Code of Federal Regulations, overseen by the Environmental Protection Agency. The FDA primarily focuses on registering new drugs in the American market.³⁷

One fundamental aspect of handling hazardous chemicals is their classification and labeling, which is detailed in Safety Data Sheets to inform workers about potential hazards and risks associated with the substances. Classification includes categories such as oxidising (O), toxic (T), very toxic (T+), corrosive (C), harmful (Xn), irritant (Xi), and dangerous to the environment (N).

Chemical testing to assess potential hazards and risks to human health primarily relies on data from animal studies, although ethical concerns and challenges in extrapolating animal data to humans are ongoing issues. Despite extensive safety data, many chemicals still lack adequate classification and require further research. Industry associations have initiated programs such as the International Council of Chemical Associations and the High Production Volume initiatives to address gaps in safety information.³⁷

In addition to traditional testing methods, new theoretical approaches like the family approach and QSAR are being validated for regulatory purposes. The current testing package to evaluate chemical hazards includes studies on acute toxicity via oral, dermal, and inhalation routes, dermal and ocular irritation, skin sensitisation, mutagenicity (including Ames and chromosomal aberration tests), repeated dose toxicity (subacute and chronic), reproductive and carcinogenicity effects, ecotoxicological impacts on the environment, identification of endocrine-disrupting chemicals, and physicochemical profiles.³⁷ The advancement of quantitative structure–toxicity relationships (QSTRs), an expanded version of QSAR, is emerging as a valuable tool in predicting and comprehending drug toxicity and metabolism. It serves as an additional resource aiding in the efficient discovery of new drug candidates.¹⁵

By the mid-1960s, there was a shift in attitude toward drug development. The focus moved from merely producing good drugs to striving for better ones, often achievable through minor alterations in the structure of the original or lead compound. If these alterations could be anticipated through the study of structure–activity relationships among derivatives and analogs, significant rewards could be reaped. Advances in biochemistry and pharmacodynamics on the biological front, coupled with advancements in structural and mechanistic chemistry on the physical side, laid the groundwork for a new quantitative approach to drug design. Mathematics acted as the catalyst for this transformation, with computers serving as the indispensable apparatus, facilitated by the pioneering work of organic chemist Corwin Hansch starting in 1962. Structure–activity relationships were quantified by assigning numerical values not only to activity but also to structure and crucially, property.¹⁶

Medicinal chemistry evolved from being purely empirical to a discipline grounded in rational design and mechanistic study. Though practitioners still don the traditional white coat and engage in laboratory work, their habitat has expanded to include computer screens. In 1969, the application of QSAR was first explored in research on antiviral compounds, igniting enthusiasm for the potential of explaining and predicting drug activity through numerical manipulation. Effective drug design necessitates an understanding of biological systems and their modulation, as well as an appreciation for the physicochemical properties of molecules. Armed with this knowledge, suggestions for drug design may stem from biochemical rationale, the testing of bioisosteres, or perhaps QSAR analyses.¹⁶

In 1939, Ferguson, working at the ICI Laboratories, proposed a concept linking narcotic activity, partition coefficient, and thermodynamics. This concept was influenced by Meyer’s statement in 1935, which suggested that identical narcotic effects occur when molar concentrations of narcotics in cell lipids are the same. The partition coefficient determines the equilibrium between the external phase (e.g., water) and the biophase (e.g., nerve tissue). Ferguson reasoned that in equilibrium, thermodynamic principles would govern drug activities, with the relative saturation of substances in the applied phase being a crucial parameter for correlating narcotic activities. This became known as Ferguson’s principle. Thermodynamic arguments have since been applied to various biological systems, yielding relationships with solubility, partition coefficient, surface tension, and parachor. However, these relationships have been of limited use, mainly restricted to certain types of biological activities such as narcotic, anesthetic, or depressant effects and simplified test systems. In whole animals, where kinetics are more relevant than equilibrium, single-parameter correlations have generally not been obtained. One notable exception was the discovery of halothane as an anesthetic, where the application of Ferguson’s principle suggested high potency, justifying further screening and development.¹⁶

Albert fully described Ferguson’s principle in his book “Selective Toxicity.” In 1939, Albert initiated a quantitative study of the antibacterial activity of aminoacridines, revealing the importance of considering the concentration of active species, particularly the cation, in drugs capable of acid–base or other prototropic equilibria. Albert studied over 100 acridines, including proflavine and acriflavine, which were introduced in 1913 and used during World War I for treating sepsis in wounds and burns. Albert’s preliminary results showed that acridines became more active as they became more highly ionised. Compounds fully ionised at pH 7.3 were equally active, while compounds like 2-amino acridine, with a pK_a of 5.8 and <10% ionisation, were 10-fold less active in tests for minimum bacteriostatic activity. Albert concluded that the cation was the active species, but he made further observations. Lowering the pH from 8 to 6 increased the activity of compounds to a certain extent, as more cations were present. However, further decreasing the pH to around five resulted in decreased activity, despite the increasing percentage of cations, suggesting a more complex relationship between pH and activity.¹⁶

Correlation between ethanol volume fraction and solubility profiles of pharmaceuticals: a sigmoidal relationship

The solubility profiles of many semi-polar pharmaceuticals peak at an ethanol volume fraction (f(max)) of between 0 and 1, whereas the solubility patterns of extremely polar and very nonpolar medications are monotonic. The value of f(max) showed a sigmoidal relationship with the log of the solute’s octanol/water partition coefficient (log K(ow)). The value of the volume fraction of ethanol that provides maximum solubility (f(max)) is reasonably predicted by this relationship. The calculation of the overall ethanol/water solubility profile is made possible by combining this sigmoidal relationship with the previously documented linear relationship between the log K(ow) and the starting slope of the plot of log solubility versus ethanol composition.³⁸

Assessing theophylline transfer to breastfed infants: pharmacokinetic insights and risk evaluation

Examining the pharmacokinetics of drugs passing from breastfeeding mothers to their infants, using theophylline as a case study, aimed to assess potential risks to the nursing infant. The investigation involved: (1) Administering theophylline orally to 12 nursing mothers to analyse the translactal passage’s pharmacokinetics through simultaneous serum and milk measurements. (2) Developing a pharmacokinetic model to predict the drug’s uptake by the infant based on translactal passage kinetics and infant drug kinetics. The findings were as follows: (1) Theophylline traverses the blood-milk barrier with a delay, resulting in milk concentration time courses lower than serum concentrations. Milk/plasma ratios ranged from 0.6 to 0.89, decreasing when plasma theophylline peaked. Although elimination half-lives remained consistent, the apparent volume of distribution was higher in milk, correlating well with divergent milk and serum concentration curves. (2) Integrating translactal theophylline passage kinetics and infant theophylline kinetics into the model, assuming constant dosage intervals and drinking amounts, reveal significant drug accumulation, particularly with longer infant half-lives, potentially surpassing therapeutic levels. Considering various pharmacokinetic parameters is essential for estimating drug transfer from breastfeeding mothers to infants via breast milk.³⁹

Only while the mother was taking theophylline did the asthmatic patient’s newborn exhibit irritation related to breast milk consumption. This woman’s and four other individuals’ pharmacokinetic investigations showed that theophylline readily distributed into breast milk. The drug’s average ratio of milk to serum concentration was roughly 0.7, and the time-course of milk concentrations was in line with that of saliva and serum concentrations. A nursing newborn would typically receive <10% of the mother’s theophylline dosage, which is typically an insignificant amount based on relative body weight.⁴⁰

Optimising breastfeeding safety: understanding medication transfer and risks for nursing infants

A drug’s potential for harm to a breastfed newborn depends on how it passes from plasma into milk and how complicated the mechanism of milk synthesis and secretion is.⁴¹ It is reasonable to assume that drugs consumed by a nursing mother will partially emerge in human milk and be consumed by a nursing child. Through passive diffusion, drugs are absorbed by the milk and dispersed throughout its aqueous, protein, and lipid phases. The drug’s physiochemical properties, such as its acid–base properties, relative protein binding in plasma and milk, lipid solubility, and milk composition, will all have an impact on how well it distributes into milk.⁴² Drug entry into milk, variables influencing drug concentrations in milk, mammary cell, and milk synthesis, and milk secretion and composition. Drug dose, amount bound in plasma, molecular weight, lipid solubility, degree of ionisation, and pH differential between plasma and milk are the six main parameters that determine drug concentrations in milk. Compared to weak alkaline medications, weak acid pharmaceuticals exhibit higher levels of ionisation and protein binding.⁴¹ Babies who are breastfed consume medications found in the milk. Every medication crosses the plasma-milk barrier to some degree. The most often used indicator of medication distribution into milk is the milk to plasma concentration ratio (M/P).¹²

Understanding the potential drug content in breast milk is necessary when making decisions regarding the safety of nursing during drug consumption by the mother.¹¹ A more useful measure of an infant’s exposure to a drug is the computation of the daily infant dose of the drug consumed in milk and, from there, the ratio of the dose in milk to the mother dose on a weight-adjusted basis.⁴² This has not been researched for many medications, and mothers are typically advised against nursing.¹¹ This is especially true for medications that have a large volume of distribution, as only a little amount of the mother’s dosage is present in the plasma and can enter the milk.⁴² Since the kid is typically exposed to very little in terms of total dosage, this is definitely not necessary in many circumstances.¹¹ A more accurate measure of a baby’s medication exposure is the steady-state plasma drug concentration in a breastfed child, which is mostly determined by the baby’s oral availability and clearance as well as the dosage rate (via milk). There is less data to suggest that the rate of absorption in neonates differs significantly from that in adults, despite the possibility that it does. However, very little infants have poorer clearance, especially if they are born prematurely.⁴² Knowing, therefore, the approximate amount of medicine that might be present in milk would be very beneficial.¹¹ It was reported that medicines with considerably differing lipophilicity were distributed into human milk lipid.¹⁰ The breastfed infant’s steady-state plasma concentrations would rise proportionately as a result of the reduced clearance. It is possible to estimate the anticipated steady-state plasma concentrations in breast-feeding infants by taking into account the dose taken in milk and the approximate clearance in infants of different ages. Recommendations on the safety of medications during nursing can be derived from these factors.⁴² The milk:plasma unbound ratio, or the distribution of unbound medication, can be estimated thanks to the current theory of pH partitioning. Nevertheless, because whole milk contains proteins and lipids, which medications would distribute in amounts dependent on their unique physicochemical properties, these ratios are not good representations of the concentration ratios for whole milk.¹¹ Medications with dose-independent toxicity or high levels of toxicity should be evaluated independently. Advice concerning “social” drugs, such as alcohol, caffeine, nicotine, and theobromine, is especially challenging because of the unpredictable and varying dosages.⁴²

The majority of medications are known to be eliminated in human breast milk. The levels of the medication in breast milk are often low enough to pose little risk to the baby when the mother takes it in therapeutic doses for brief periods of time. If an infant who is being breastfed has illness or does not thrive and the cause of the morbidity cannot be determined, one of the following actions should be taken: (1) Stop using the medication. (2) Stop nursing your child. Often, this can be achieved temporarily by the mother pumping her breasts to sustain milk while the baby’s reaction is observed. (3) Gather samples of baby plasma, breast milk, and mother plasma for medication testing. Based on the levels of drug detected in the milk or the infant’s plasma, it might be possible to implicate (or exonerate) a drug or one of its metabolites as the cause of the morbidity in circumstances when this is feasible. Despite how time-consuming and unfeasible this method may appear, it would eventually result in the collection of a respectable quantity of data from which fair inferences on the impact of medications on breastfed infants may be derived.¹³

It was reported that 16 medications were dissolved into human milk lipid, each with a notably different lipophilicity. A study revealed that such medications, varying significantly in their lipophilicity, were found to be distributed among the lipid component of human breast milk.¹⁰

Together with the unbound drug distribution, the amount of drug in the protein and lipid phases must be taken into account in order to forecast the milk:plasma concentration ratios for whole milk. Thus, a “phase distribution model” that allows for the calculation of whole milk:plasma concentration ratios has been devised. The unbound drug concentration ratio, the milk lipid:ultrafiltrate partition coefficient, and the plasma and milk unbound fractions must all be known in order to use the model. The model was evaluated by comparing the anticipated values of the whole milk ratio with the milk found in literature. The plasma area under the curve (AUC) ratios showed a tendency towards underprediction for basic medications and overprediction for acidic and neutral drugs.¹¹

Numerous factors, including the solubility of the drug in lipids, the binding of proteins in milk, the maternal plasma protein, and physicochemical parameters influencing the drug’s passage across biological barriers, all influence this ratio. M/P ratios of most medications are <1. From the likely maternal plasma concentration, the likely baby dose is determined using the M/P ratio.¹² By taking logarithms and transforming the phase distribution equation, a relationship is created that can be evaluated using multiple linear regression to create predictive equations for basic and acidic medications that account for the relative contributions of each model component.¹¹

With random distribution of residuals, the independent variables explained 93.1% and 82.9% of the variance in the values for acidic and basic medicines, respectively. The pK_a, octanol:water partition coefficient, and plasma protein binding values of the drug allow the ratio of the milk:plasma AUCs to be estimated for any acidic or basic drug for which the distribution into human milk is unknown. These equations work in conjunction with those that predict unbound fractions of drug in milk and milk lipid:ultrafiltrate partition coefficients.¹¹

A sample of human milk obtained from 10 volunteers. Measured at 37 °C, the association between the octanol–water and milk lipid–ultrafiltrate partition coefficients is strong, as long as digoxin and prednisolone are not present. These steroid-based compounds might disrupt the link between milk lipid and ultraviolet pigment by interacting with free fatty acids in the milk to create micelles. Based on drug pK_a and octanol–water partition coefficients, equations that allow for the estimate of drug distribution into milk for lipid-soluble medicines are provided.¹⁰

The baby dose can then be contrasted with doses administered to mothers or with doses given to newborns for therapeutic purposes. The mean concentrations ultimately reached in the infant’s plasma will depend on how well the drug is cleared by the infant’s plasma in comparison to the mother’s. Determining the safety of breastfeeding while a mother is taking medication is made possible by evaluating these variables.¹²

Because of the pharamcokinetics of most medications consumed by nursing mothers, the least quantity of the drug will be in the milk at the next feeding if the drug is administered at or shortly after the infant nurses. The degree to which a drug or its metabolite is ionised, lipid soluble, and bound to plasma proteins determines how much of the chemical gets into the breast milk. Generally speaking, medications that are well-known to be highly protein bound are eliminated from breast milk less frequently than medications that are poorly linked to plasma proteins. Medications such as pyridoxine (vitamin B6), ergot derivatives, and oral contraceptive hormones can all negatively impact lactation.¹³

Drugs concentrations used in the measurement of milk lipid–ultrafiltrate partition coefficients: the following radiolabeled and nonradiolabeled drugs were used; some of which such as H|chlorpromazine (New England Nuclear [NEN]), H|diazepam NEN, diazepam Roche, C|Diclofenac (Ciba-Geigy, 6.6 mCi/mmol, K-277), Diclofenac (Ciba-Geigy, 0.1317), H|digoxin NEN, 18 Ci/mmol-2228-109; digoxin (Wellcome, 11725), fluphenazine (Squibb 12240), flunitrazepam Roche 2132-066, phenytoin NEN, 46 Ci/mmol 2315-061, phenytoin (Parke-Davis 5419972), pirenzepine NEN, 76 Ci mmol, 2261-155, pirenzepine (Boehringer-Ingelheim-660206), H|prednisolone (Amersham, 67.4 Ci/mmol, 88), prednisolone (Sigma, 74F-0739), I’Hlpropranolol (NEN, 18.5 Ci mmol, 2084-275, propranolol tIC1, ADM 7743083A), 13H| verapainil INEN (70 Ci mmol, 2427-051), verapamil (Sigma, 065F0452), H| warfarin (Amersham, 46 mCi mmol, 30), warfarin (Sigma, 15F-0486) radiochemical purity of drugs was stated to be 97% to 99%.¹⁰

Additional control experiments conducted using human milk showed that transitioning from a milk ultrafiltrate aqueous phase (with a pH of approximately 7.2 ± 0.05) to a phosphate buffer (with a pH of 7.25) led to notably different (with statistical significance indicated by P < 0.05) measured log milk Pap values for pirenzepine, diclofenac, flunitrazepam, phenytoin, imipramine, fluphenazine, and diazepam. Initial analyses and examination of residuals highlighted digoxin and prednisolone as significant outliers from the expected trend, both being neutral drugs with a steroid ring structure.¹⁰

Paediatric Investigation Plan (PIP)

The number of patients concerned about potential negative effects of medication secreted in breast milk on the unborn child has increased due to the rising prevalence of breastfeeding and the frequency of pregnancies among females with chronic medical problems. Patients and practitioners can use the guidelines to help them make sensible decisions in these circumstances. Medication for nursing mothers should not be administered unless there is strong proof the medication will help the ailment for which it is intended. Breastfeeding will be harmed more by the mother’s health deteriorating from an acute or chronic sickness than by the majority of medical therapies for the ailment.¹³

The dangers or risk of some medications to breastfed babies

Alkylating agents, analgesics and anti-inflammatory agents, anticoagulants, anticonvulsants, anti-infective agents, central nervous system stimulants, hormones, laxatives, minerals, provitamins, psychotherapeutic agents, thyroid affecting agents, and vitamins are some drug categories that contain pharmacons that should be limited or avoided by nursing mothers.⁴¹

To yet, pyrimethamine is the only medication found in breast milk that is known to have a positive effect. Corticosteroids, sex hormones, significant tranquillisers, tricyclic antidepressants, antihypertensives, and antituberculars are among the medications that may be harmful to lactation.⁴³

It is advised to take the following safety measures to reduce the possibility of potentially hazardous pharmaceuticals: (1) nursing mothers should refrain from taking any unnecessary medications; (2) if medication is required during lactation, the dosage should be regulated and the baby should be observed for any negative side effects; (3) drugs should be given soon after breastfeeding and the time between feedings should be extended; and (4) bottle feeding is advised if the baby needs to be fed soon after the mother has taken a potentially harmful drug.⁴¹ (Table 2).

Guidance for Canadian take-home naloxone programs: policy recommendations based on 11 yrs of community and scientific evidence

The rising toxicity of opioids in the unregulated drug market has contributed to a growing number of overdoses both in Canada and globally. Take-home naloxone (THN) programs, which provide naloxone kits to community members who might witness an overdose, are a proven intervention. This guidance aims to offer policy recommendations for territorial, provincial, and federal THN programs, drawing on scientific research, gray literature, and community evidence accumulated over 11 yrs of THN distribution in Canada.

Methods

The Naloxone Guidance Development Group, including experts and individuals with direct drug use experience, used the AGREE II tool to develop this guidance. From December 2021 to September 2022, it reviewed published literature and community input, and sought feedback through an External Review Committee and public consultations. The Canadian Institutes of Health Research funded that project, following Guideline International Network conflict of interest principles.

Recommendations

Given the limited quality of existing evidence, the recommendations include offering both intramuscular and intranasal naloxone, and equipping kits with naloxone, a delivery device, personal protective equipment, instructions, and a carrying case. Responders should prioritise rescue breathing and conventional cardiopulmonary resuscitation (CPR).

Interpretation

This guidance provides direction for Canadian THN programs, based on diverse stakeholder input and limited evidence.⁴⁴ (Table 3).

Comparison of psychoactive substance concentrations in fatal and nonfatal opioid overdoses: a case–control study with naloxone correlation analysis in Norway

A study to compare psychoactive substance concentrations in blood samples from nonfatal and fatal opioid overdoses, with a secondary objective of evaluating naloxone levels in nonfatal cases and their correlation with other detected drugs. This case–control study was conducted in Norway, analysing data from fatal overdoses that occurred in 2017 and nonfatal overdoses from February 2018 to September 2019.⁴⁵

The study included 31 nonfatal and 160 fatal opioid overdose cases. Information for the nonfatal cases was obtained from hospital records and blood samples, while data for the fatal cases came from autopsy reports. Blood concentrations of psychoactive substances, including ethanol, were collected upon hospital admission for nonfatal overdoses and during autopsies for fatal cases. The results showed that the median number of substances detected was four in fatal overdoses and five in nonfatal overdoses. Fatal cases had significantly higher concentrations of opioids (188 ng/mL vs. 57.2 ng/mL, P < 0.001), benzodiazepines (5,467 ng/mL vs. 2,051 ng/mL, P = 0.005), and amphetamines (581 ng/mL vs. 121 ng/mL, P < 0.001) compared to nonfatal cases. A linear relationship was found between naloxone levels and pooled opioid concentrations (95% confidence interval = 0.002–0.135, P < 0.05). The study concluded that the overall burden of drug concentrations played a significant role in the fatality of an overdose, while the number of substances detected had less impact on survival. Higher opioid levels were linked to the administration of higher doses of naloxone during treatment.⁴⁵ (Table 3).

Postmortem toxicology of antidepressants and antipsychotics: reference concentrations from fatal poisonings in Finland (2000 to 2020)

Antidepressants and antipsychotics are significant classes of prescription drugs in postmortem (PM) toxicology, as many of these substances are toxic in overdose situations, and the mental health conditions they treat are often linked to suicidal tendencies. Current PM toxicology protocols include a broad range of these medications. However, despite the existence of individual case reports, there are few comprehensive studies in the literature that offer fatal concentration data from large numbers of cases. This study, based on PM investigations conducted in Finland from 2000 to 2020, presents reference concentrations for fatal poisonings where an antidepressant or antipsychotic drug was identified as the primary cause of intoxication.⁴⁶

Summary statistics for drug concentrations in PM femoral blood, including minimum, maximum, mean, and percentiles (10th, 25th, 50th, 75th, and 90th), were calculated for 17 antidepressant drugs (N = 2,007) and 12 antipsychotic drugs (N = 1,161). The proportion of deaths classified as suicide, accident, or of an undetermined manner was noted for each drug. The fatal concentrations identified in this study were compared with previously published data, using both the grouped causes of death approach and the all causes of death approach, which reflect fatal and “normal” PM concentrations, respectively. The study demonstrates that, despite the recognised variability in PM drug concentrations, the carefully generated fatal concentration data for the examined drugs are sufficiently consistent to serve as reliable references during the interpretation of toxicological findings.⁴⁶ (Table 3).

Fatal ropinirole intoxication: blood and urine concentrations from autopsy cases and PM redistribution insights

Ropinirole is an antiparkinsonian drug that has recently been proposed as a potential treatment for amyotrophic lateral sclerosis. With its expected rise in prescriptions, understanding its toxic effects is crucial. Currently, the lethal concentration of ropinirole in blood is not well established. A report presented a fatal case of ropinirole intoxication and compares its blood and urinary concentrations with other autopsy cases involving ropinirole detected through drug screening in a laboratory. Ropinirole was measured in femoral vein blood, cardiac blood, and urine in five autopsy cases. One case involved fatal ropinirole intoxication (the case discussed in this report), while the remaining four cases were nonintoxication deaths where ropinirole was present. The ropinirole concentrations in the fatal case were 100 ng/mL in femoral blood, 160 ng/mL in cardiac blood, and 1,840 ng/mL in urine. In the four nonfatal cases, femoral blood concentrations ranged from 7 to 35 ng/mL (average: 24 ng/mL), cardiac blood concentrations were between 13 and 100 ng/mL (average: 60 ng/mL), and urine concentrations ranged from 140 to 1,090 ng/mL (average: 640 ng/mL). The cardiac-to-peripheral blood ratios fell between 1.6 and 2.1 (average: 1.8).

The results indicated no clear evidence of overdose, though the elevated cardiac-to-peripheral blood ratio suggested PM redistribution. However, the femoral blood concentration of 100 ng/mL in this case was notably higher than the 64 ng/mL previously reported in another fatal ropinirole poisoning case. It was concluded that the cause of death was likely due to shock and fatal arrhythmia triggered by ropinirole poisoning. This case offers critical insights into PM ropinirole levels in blood and urine following fatal intoxication.⁴⁷ (Table 3).

Fatal dextromethorphan poisoning: autopsy findings and public health implications from two case reports

A report details the autopsy of a young man who died from a single overdose of dextromethorphan (DXM), an over-the-counter cough suppressant widely used globally. Recently, DXM has gained popularity among young people for its euphoric, hallucinogenic, and dissociative effects. Although DXM addiction is becoming more prevalent, fatal cases of DXM poisoning remain rare, particularly those involving DXM alone without other substances. In this case, a man in his early 20s was discovered deceased at home, with no external injuries or significant internal abnormalities observed during the autopsy. Toxicological analysis revealed exceptionally high levels of DXM, and no other drugs were detected. To the best of our knowledge, this is the first reported fatality in Japan caused solely by an overdose of DXM. There is a need for increased public awareness regarding the dangers associated with excessive DXM consumption.⁴⁸

A report discusses the case of a woman in her 30s who was receiving treatment for a mental illness with multiple psychotropic medications. She was found unconscious with respiratory arrest at home and was pronounced dead 12 h later. The autopsy revealed symmetrical haemorrhagic necrosis in the putamen of both hemispheres of her brain. While numerous drugs were found in her blood, all substances except DXM were within or below therapeutic levels. The DXM concentrations were 1.73 μg/mL at admission and 1.61 μg/mL at the time of autopsy, levels that fall within the toxic or coma-to-death range. DXM poisoning was determined to be the cause of death. Although DXM can produce hallucinations and euphoria in high doses, it is an over-the-counter medication available at general pharmacies, leading many young people to misuse it under the false belief that it is relatively safe. This case emphasises the need for increased public education about the risks of DXM and stricter regulations on its over-the-counter sale in Japan.⁴⁹ (Table 3).

Comparing traditional RA and biological monitoring: from external exposure to internal dose evaluation

Traditional RA typically focuses on external exposures. It involves using the administered dose identified as either the no-adverse-effect level or lowest-observed-adverse-effect level from animal toxicity studies, adjusted with uncertainty factors to establish a safe exposure level for humans. Daily exposure estimates in human populations are then compared to health-based criteria, such as reference doses, tolerable daily intakes, or minimal risk levels, to determine if they exceed acceptable limits. In contrast, biological monitoring provides a measure of internal exposure, offering a more direct and biologically relevant assessment of actual exposure. This method estimates the concentration of the active chemical at the target site or organ, which is crucial for understanding toxicity and biological response, as opposed to relying solely on external exposure measures.⁵⁰

Additional clinical chemistry tests may be suggested to further investigate potential toxic effects related to a test substance. The choice of specific tests will depend on the observed mechanism of action of the substance. To ensure a thorough toxicological assessment, recommended clinical chemistry analyses may include evaluations of acid/base balance, hormone levels, lipids, methemoglobin, and proteins.

Despite adhering to standard procedures and equipment calibration, significant variation in clinical chemistry results can occur from day-to-day. Ideally, all analyses for different dose groups should be completed within a single day. If this is not feasible, efforts should be made to conduct the analyses in a way that minimises potential variability.⁵¹ (Table 3).

Standardised reporting and QSAR-based toxicity predictions: enhancing chemical RA and resource management in cases of toxin and toxicant exposure

When human illness arises from the accidental or deliberate release of toxins (substances produced by organisms through metabolic processes, such as ricin) or toxicants (natural or synthetic chemicals not produced by organisms, such as nerve agents), standardised reporting is crucial. This uniformity helps allocate appropriate resources, evaluate the impact on health, track affected individuals, and monitor the effectiveness of interventions. Toxins are chemicals generated by organisms as a result of metabolic processes (e.g., marine toxins like saxitoxin or plant toxins like ricin), while toxicants are either synthetic or naturally occurring chemicals not produced through metabolic processes (e.g., nerve agents or arsenic). In cases of illness from such chemical releases, uniform reporting is essential for effective resource management, health assessment, and intervention monitoring. This report from the Centers for Disease Control and Prevention (CDC) provides case definitions to ensure consistent reporting of illnesses caused by both toxins and toxicants.⁵² (Table 3).

Leveraging QSAR models for chemical toxicity prediction: insights from the toxicity estimation software tool

The toxicity estimation software tool (TEST) was created to facilitate the estimation of chemical toxicity using QSARs. QSARs are mathematical models designed to predict toxicity based on the chemical structure’s physical characteristics, known as molecular descriptors. Basic QSAR models determine toxicity using a straightforward linear equation: Toxicity = ax1 + bx2 + c, where x1 and x2 represent the molecular descriptors, and a, b, and c are coefficients derived from the data. Examples of molecular descriptors include molecular weight and the octanol–water partition coefficient.⁵³

Computational toxicology methods, such as QSAR models, facilitate the rapid prediction of chemical toxicity parameters, which are essential for assessing chemical risks.⁵⁴ (Table 3).

Development and validation of QSTR models for predicting toxicity of imidazolium- and pyridinium-based ionic liquids: insights from in vivo and in silico studies

New predictive models for assessing the toxicity of imidazolium- and pyridinium-based ionic liquids (ILs) were developed using the OCHEM platform. These models were validated through cross-validation, yielding a coefficient of determination (q²) between 0.77 and 0.82. They were then used to screen a virtual chemical library to evaluate IL toxicity in Danio rerio and Daphnia magna bioassays. Out of 25 ILs predicted by the models, which were subsequently synthesised and tested in vivo, D. magna proved to be more sensitive than D. rerio. In D. magna assays, 67% of the ILs were classified as extremely toxic, with LC₅₀ values ranging from 0.005 to 0.01 mg/L. Conversely, only one IL, 1-dodecylpyridinium bromide (with an LC₅₀ of 0.08 mg/L) was categorised as extremely toxic in D. rerio tests, while 76% were classified as slightly or moderately toxic. The most toxic ILs (5 and 19) were docked into the human acetylcholinesterase (AChE) active site, showing binding energy values of −9.5 and −9.3 kcal/mol, comparable to the human AChE inhibitor Donepezil, suggesting potential molecular mechanisms of IL toxicity. The QSTR models proved effective for analysing the toxicity of new ILs, demonstrating high predictive accuracy and a strong correlation with in vivo toxicity results.⁵⁵ (Table 3).

Integrating adverse outcome pathways and advanced protein structural prediction tools for enhanced RA of atrazine-induced reproductive toxicity

Adverse outcome pathways (AOPs) are conceptual frameworks that systematically organise scientific knowledge about how stressors interfere with specific biological targets and pathways. An AOP network comprises multiple AOPs that share common key events (KEs), including critical events such as molecular initiating events and adverse outcomes (AOs), facilitating the connection of toxicological pathways. To explore the sequence of KEs in AOP 492, which is triggered by atrazine (ATZ); in which constructed a reproductive toxicity via oxidative stress AOP network using this AOP as a starting point. Sourced individual AOPs from the AOP-Wiki database. Within this network, the KEs “Increased Reactive Oxygen Species” and “Apoptosis” were identified as the most prevalent and highly interconnected, marking them as significant points of divergence. “Increased DNA Damage and Mutations” emerged as a crucial KE, being highly central and also a point of divergence. This suggests that these three KEs hold significant predictive value and could be used to develop or select in vitro assays for evaluating reproductive toxicity. In silico analyses identified the key target proteins for ATZ-induced infertility through oxidative stress in humans as Tp53, Bcl2, Esr1, and Nos3, which interact indirectly with ATZ through intermediary factors like: Mapk3, Mapk1, and Cyp19a1. Gene enrichment analyses further confirmed that these entities are involved in biological processes and pathways directly related to oxidative stress, DNA damage, and apoptosis, reinforcing the validity of the developed network.⁵⁶ (Table 3).

Advancing chemical susceptibility predictions across species through protein structural integration and molecular docking in next-generation risk assessment

Recent developments in protein structural prediction tools, such as AlphaFold and Iterative Threading Assembly Refinement, have significantly improved the ability to predict protein structures across different species using available sequence and structural data. A study presents a novel molecular docking approach that leverages this extensive structural data to improve predictions of chemical susceptibility across various species. The method was applied to the androgen receptor, a key player in endocrine function, demonstrating how integrating protein structures can contextualise species-specific susceptibility within a functional framework. This approach integrates molecular docking into new approach methodologies and supports the next-generation risk assessment paradigm by combining various open-source tools.⁵⁷ (Table 3).

Optimising experimental design in toxicogenomics: the impact of sequencing depth and biological replication on differential gene expression and RA

Transcriptomic data have become a vital tool in toxicology for elucidating cellular MOA and determining points of departure (POD) for hazard assessment and potency comparisons. A MOA describes the sequence of biological events and processes following exposure to a chemical, leading to a specific toxic effect or AO. POD refers to the dose level or reference point on a dose–response curve where a particular toxic effect or response starts to manifest. Recent advancements have greatly enhanced workflows for analysing transcriptomic data, allowing for standardisation in most toxicology applications. Automation in these analytical processes boosts efficiency and throughput, reducing costs while ensuring that best practices are consistently applied. This uniform application of quality control, statistical analysis, and data normalisation helps maintain unbiased comparisons between different compounds and experiments.⁵⁸

In toxicogenomics and RA, sequencing depth, and biological replication are crucial experimental design factors. However, their relative effects on differential gene expression analysis are not well understood. To address this, study analysed an RNA-seq dataset from A549 cells exposed to Prochloraz at eight different doses, systematically varying sequencing depth (from 5% to 100%) and replicates (2 to 4) to assess their impact on detecting differentially expressed genes. Although dose was the main driver of variance, replication had a more significant impact than sequencing depth on detection power. With only two replicates, over 80% of the ~2,000 differential genes were unique to specific depths, reflecting high variability. Increasing the number of replicates to four markedly improved reproducibility, with over 550 genes consistently detected across most depths, representing 30% of the total differential genes. More replicates also improved the overlap of benchmark dose pathways and the precision of median benchmark dose estimates. However, essential gene ontology pathways related to DNA replication, the cell cycle, and cell division were consistently identified even with fewer replicates. This study highlights the trade-offs between sequencing depth and replication in toxicogenomic experimental design. While more replicates enhance reproducibility, the benefits of increased sequencing depth diminish. Prioritising biological replication over depth is a cost-effective strategy that improves data interpretation without compromising the detection of key gene expression patterns. This research offers valuable guidance for designing toxicogenomics experiments.⁵⁸ (Table 3).

An ODE model for endogenous hydrogen peroxide metabolism in hepatocytes: insights into oxidative stress, glutathione depletion, and apoptosis dynamics

An ODE model has been developed to describe endogenous hydrogen peroxide (H₂O₂) metabolism in hepatocytes, and it stands out for its ability to accurately estimate intracellular H₂O₂ levels during oxidative stress. Subsequently, it serves as a valuable tool for constructing physiologically based pharmacokinetic/pharmacodynamic (PBPK/PD) models that assess the impact of reactive oxygen species (ROS)-producing xenobiotics. This model is available for rat hepatocytes in vitro and for mouse liver in vivo. A generalised approach is provided for using the model to generate PBPK/PD models that predict intracellular H₂O₂ concentration and the death of hepatocytes due to oxidative stress. These predictions can be identified using in vitro data sets that track cell death following xenobiotic exposure at different levels. The method is demonstrated using dimethylarsinous acid (DMAIII), a trivalent arsenical generated in the liver as part of arsenic metabolism. This represents the first model of H₂O₂ metabolism in hepatocytes that incorporates endogenous production rates of H₂O₂ from mitochondria and other organelles, as derived from physiological literature. It also includes a detailed and realistic representation of glutathione (GSH) metabolism, achieved through a simplified version of Reed and colleagues’ pioneering model of hepatic GSH metabolism.⁵⁹

Simulations from this model suggest that the immediate trigger for a significant rise in intracellular H₂O₂ concentrations associated with apoptosis is the critical depletion of GSH. Interestingly, this depletion may occur hours after the peak concentration of xenobiotics due to a “delay effect.” Once GSH depletion occurs, the intracellular concentration of H₂O₂ increases rapidly through two distinct phases, the first controlled by the kinetics of GSH peroxidase and the second by catalase activity. The simulations also imply that critical GSH depletion leads to H₂O₂ concentrations that may not directly cause apoptosis but rather indicate an apoptotic environment. This supports the hypothesis that ROS levels, at the concentrations associated with apoptosis, may serve as markers rather than direct causes of cell death. A noteworthy finding of the model, relevant to broader physiological studies, is that the delay effect offers a GSH-based mechanism by which cells can differentiate between temporary H₂O₂ elevations, which may be involved in intracellular signaling, and sustained increases, which indicate pathology or toxin presence and ultimately lead to apoptosis.⁵⁹ (Table 3).

Ferguson’s principle

Overview

The lethality of drugs in the bloodstream is influenced by factors such as solubility and comparative relative saturation, particularly in the context of nonspecific depressants. Ferguson’s principle provides insights into predicting lethal concentrations based on relative saturation, especially for substances with central nervous system depressant properties.

Implications

Understanding Ferguson’s principle aids in predicting and assessing the lethal toxicity of drugs, considering factors such as lipid solubility, relative saturation, and thermodynamic equilibrium. It offers insights into drug mechanisms and assists in drug development and safety evaluation.

Understanding the interplay between solubility, relative saturation, and lethality aids in predicting the toxic effects of drugs in the bloodstream. Ferguson’s principle provides a practical framework for assessing lethal concentrations, particularly for substances with central nervous system depressant properties, contributing to drug safety evaluation and toxicity prediction (Fig. 1).

Ferguson’s principle, when applied, aids in foreseeing the lethal concentrations of drugs present in the bloodstream

Aside from a few rare exceptions (such homologous oxybarbiturates), no clear correlation can be found between any component of molecular structure and lethal drug doses in humans. Nonetheless, it was noted many years ago (Ferguson, 1951) that relative saturation, not concentration, should be taken into account when interpreting drug levels. The principles of thermodynamics, which control equilibrium between immiscible phases, gave rise to this theory. The thermodynamic or chemical potential and the closely related notion of activity are the functions of interest. The chemical potential (μ) of a system with two or more phases at equilibrium is the same in every phase. The ratio C/C°, where C is the solute’s actual concentration and C° is its saturated solution concentration, both represented in mole fractions, determines the solute’s activity in ideal solutions. The Ferguson principle states that when nonspecific medications are present at similar relative saturations, or at similar activities, they elicit quantitatively identical effects. As stated by Ferguson (1951), it is possible to present the case in a maybe stronger way. Thus, according to Ferguson (1951), the argument may be put in a perhaps more convincing form.¹⁸ The implication of the claim that some substances have the same narcotic effect at the same thermodynamic activity is basically this: one molecule of any pure substance found in a reagent bottle on the shelf requires the same amount of work to transport it to its proper location in the critical stage of the narcotised organism. This hypothesis was first applied to volatile insecticides, but it has since been supported by hypnotics, anesthetics, and several simple organic chemicals in a range of experimental settings with a variety of organisms, including man, grain weevils, and astadpoles. Whether anesthesia is caused by inert gases (like xenon) or volatile liquids (like diethyl ether), the impact offers a quantitative knowledge of the process. None of the studies on nonspecific narcosis have examined fatal drug toxicity in people; instead, most have focused on sublethal circumstances.¹⁸

An article claiming that the water solubility (S) of different drug kinds is the only factor that determines their median deadly concentrations (EC₅₀) (fatal concentrations (EC₅₀)/lethal concentration EC₅₀). The EC₅₀ values of the test compounds exhibit a range of 2,000-fold, but the EC₅₀/S ratio remains rather consistent.¹⁸

Numerous medications work through certain processes to achieve their therapeutic effects. These mechanisms may include competitive or noncompetitive inhibition of an enzyme or chemical contact with a receptor. It is now widely acknowledged that the lipid solubility of some substances—particularly hypnotics and volatile anesthetics—is a direct factor in their efficacy, with the finer molecular details being mostly inconsequential. It is believed that these medications function by interfering with physiological processes in the brain’s lipoprotein membranes. These nonspecific drugs’ lethal drug toxicity is also thought to be an extension of their typical therapeutic benefits. As a result, there is a strong association between different indicators of the medications’ lethal toxicity in both people and animals and their lipid solubility, as determined by P, the n-octanol/water partition coefficient.^19,60

This behaviour is characteristic of “Ferguson chemicals” or “Ferguson substances” and allows one to compute the fatal toxicity of related compounds and other medications, including depressants. The data indicate that humans are more sensitive to narcotic drugs than lower species, with a mean relative saturation (EC₅₀/S) of 0.006. A comprehensive paradigm for quantitative structure–activity links in lethal drug toxicity is offered by the Ferguson principle. The Ferguson principle, which is based on the thermodynamic idea of free energy, can be used to obtain the empirical Hansch equation. Thus, the Ferguson principle offers a comprehensive foundation for understanding the quantitative relationship between chemical structure and the toxicity of drugs leading to fatality.¹⁸

The solubility in n-octanol and the n-octanol/water partition coefficient (pH 7.4; 37 °C) for 11 widely used medications were found. The medications that were put to the test included dibucaine, amethocaine, procaine, quinidine, acetylsalicylic acid, amitriptyline, trazodone, dextropropoxyphene, diltiazem, and paracetamol. Using the two factors found above and the median fatal blood level documented in the literature, the relative lipid saturation corresponding to a fatal plasma concentration for eight of the medications was computed. The estimated relative saturation in the lipid phase for five of the eight medications was found to be between 0.001 and 0.004, which is in close proximity to relative saturation values in the aqueous and vapor phases that have previously been reported for substances having a nonspecific or physical mechanism of toxicity. Since their lipid solubility plays a major role in this, it is likely that accumulation in the lipid phase plays a significant role in determining the lethal toxicity of medications and other substances with a nonspecific mechanism of toxicity.⁶¹

There is, however, mounting evidence that this nonspecific mechanism is also responsible for the lethal effects of numerous other medications linked to fatal poisonings. The main molecular characteristics that show up when drugs are ranked according to their index of fatal toxicity in humans are that (lipophilic) aliphatic side-chains are much more common in the most toxic drugs, while (hydrophilic) hydroxyl groups are highly associated with the least toxic drugs.¹⁸

The midpoint concentrations at which drugs and associated substances become lethal in the bloodstream, along with their solubility and comparative saturation, are significant factors to consider (Table 4).

Human poisoning by nonspecific depressants is consistently associated with an unweighted mean value (±1 standard deviation), or EC₅₀/S = 0.006 (±0.0018), even though relative saturations (EC₅₀/S) depend on the circumstances surrounding drug solubility determination. The fact that this number is far lower than what is needed to cause narcosis in lesser animals (Ferguson, 1951) indicates that humans are far more susceptible to the negative effects of narcotics. Assuming that any compound is a Ferguson material, figuring out its median fatal concentration is a simple task. EC₅₀ = (0.006) S, where S is the compound’s solubility under the given conditions, is the result.¹⁸

Chemical potential and relative saturation are concepts that do not apply to “nonnarcotic” poisons, as these substances cause death through very particular ways. Thus, “non-Ferguson substances” include things like carbon monoxide, hydrogen cyanide, paracetamol, and organophosphoric acids. A second drawback to this strategy is when medications have a water partition coefficient (n-octanol/water) greater than 1,000, indicating an extremely high lipid solubility. Since these substances will quickly concentrate in the first accessible lipoidal depot following injection and cannot reach the brain at small quantities, they are unlikely to exert depressive effects. The biological inertness of saturated long-chain hydrocarbons can be explained by this effect. Non-Ferguson compounds are by definition those that are not soluble in lipids.¹⁸

The Ferguson principle seems to be more applicable to the analysis of quantitative structure–activity relations in catastrophic drug toxicity than the Hansch equation. In fact, the idea of chemical potential can be used to create the empirical Hansch equation.²⁰

Many drug kinds function as nonspecific poisons (also known as Ferguson chemicals) when present at deadly concentrations; the sole factor determining fatal toxicity is relative saturation. Because of this tendency, it is possible to determine the lethal blood concentrations of other medications and their related chemicals based only on how soluble they are in water. This strategy is only suitable for substances that have central nervous system depressant properties. Nonetheless, this group includes most medications linked to poisoning deaths in humans.¹⁸ (Fig. 2).

Discussion

This study demonstrates the delicate balance between maternal pharmacokinetics, the properties of the drug, and the potential risk the infant may be exposed to through breast milk. The evaluation of chlorpromazine, diazepam, diclofenac, digoxin, fluphenazine, phenytoin, pirenzepine, prednisolone, and warfarin makes a clear case for the need for effective pharmacovigilance and clinical guidance. The application of Ferguson’s principle within predictive models of drug toxicity helpfully improves understanding of drug distributions, narcotic strengths, and receptor relations. Such assessments aid in evaluating risks and enable better healthcare decision-making when prescribing to lactating mothers.

Effects to consider

This study greatly contributes by focusing on predictive toxicology, especially the application of Ferguson’s principle and QSAR modelling in drug safety assessment. With predictive models at their disposal, researchers and healthcare practitioners can predict the likelihood of drugs transferring to breast milk and harming breastfed infants.

The ability to predict narcotic potency and systemic toxicity through thermodynamic principles represents a paradigm shift in toxicological assessment, reducing reliance on empirical data and animal studies. Notably, the integration of QSAR modelling into this research enhances our understanding of structure–activity relationships and their impact on drug distribution. The identification of molecular descriptors that influence drug behavior in biological systems provides a foundation for designing safer pharmacotherapeutic agents. This study also draw attention to the potential for QSAR-driven approaches in environmental toxicology, where they can be used to assess chemical risks in broader ecological contexts.

The comprehensive evaluation and classification of ADRs outlined in the text highlight the multifaceted nature of drug safety monitoring and management. By categorising ADRs based on their pharmacological effects, relationship to drug dose, and causal relationship, healthcare professionals can better understand and manage the risks associated with medication use.

Viewed from this perspective, the systematic classification of ADRs allows for a more nuanced approach to RA and mitigation in clinical practice. For instance, identifying type A reactions as dose-dependent and predictable helps clinicians anticipate and manage adverse effects that may arise during drug therapy. Conversely, recognising type B reactions as unpredictable and immune-mediated study emphasises the importance of vigilant monitoring and early intervention to prevent serious outcomes.

Moreover, the emphasis on reporting and evaluating significant ADRs by organisations like the JCAHO and the ASHP emphasises the commitment to patient safety and continuous quality improvement within the healthcare industry. The Ferguson principle provides valuable insights into predicting and understanding the lethal toxicity of drugs, particularly those acting as nonspecific depressants. By considering factors such as solubility and relative saturation, this principle offers a practical framework for assessing lethal concentrations in the bloodstream. However, it’s important to recognise that the applicability of Ferguson’s principle is primarily limited to substances with specific mechanisms of toxicity, particularly those affecting the central nervous system with other similar type of toxic drug substances or a Ferguson substance. While it provides a comprehensive foundation for understanding the quantitative relationship between chemical structure and drug toxicity leading to fatality, it may not be universally applicable to all types of poisons or medications. Therefore, while Ferguson’s principle offers valuable insights into drug toxicity prediction, its scope and applicability need to be carefully considered in the context of specific substances and their MOA before the drug is evaluated for clinical toxicity and preclinical studies, as this will save time and money.

Overall, the viewpoint presented the importance of a systematic and multidimensional approach to ADR monitoring, classification, and management of toxic drug substance study by using Ferguson’s principle to ensure the safe and effective use of medications in clinical practice; which is a significant output of this study.

Conclusion

In conclusion, the rigorous process of drug development, involving preclinical assessments to postmarket monitoring, ensures the safety, quality, and efficacy of medicinal products, while continuous monitoring of ADRs, guided by international guidelines and regulatory agencies, is vital for enhancing patient safety and advancing human health, with drug lethality in the bloodstream influenced by factors like solubility and relative saturation, wherein Ferguson’s principle offers insights for predicting lethal concentrations, albeit limited to substances with specific mechanisms of toxicity, emphasising the importance of comprehensive approaches in drug safety evaluation for informed clinical decision-making. The integration of community-based interventions, like naloxone programs, with advanced toxicological and computational tools significantly enhances our ability to assess and manage chemical risks. The studies emphasise the importance of both empirical data and predictive modelling in understanding drug toxicity and improving public health outcomes; as quasi-empirical data. The evidence from THN programs demonstrates the effectiveness of harm reduction strategies, while the toxicological data from fatal poisonings reveal critical insights into substance-specific risks. Advances in QSAR models, AOPs, and protein structural predictions represent significant strides in chemical RA, offering more precise tools for evaluating toxicity across species. On top of that, the findings on experimental design in toxicogenomics and oxidative stress modelling accentuate the need for robust methodologies in toxicological research. Together, these advancements provide a comprehensive framework for improving RA and resource management in toxicology.

Thus, in summary, this study provides significant insights into the application of computational toxicology and pharmacokinetics in assessing drug safety during lactation. The integration of Ferguson’s principle, QSAR modelling, and omics-based analysis presents a promising approach to predictive toxicology, paving the way for more accurate and efficient RAs that is often noninvasive.

Despite offering valuable insights, this study has several limitations. The reliance on existing pharmacokinetic and physicochemical data for model validation may introduce biases due to variations in data quality and sources. In like manner, while Ferguson’s principle and QSAR models provide powerful predictive capabilities, they simplify biological systems and do not fully capture complexities such as active transport mechanisms, enzymatic metabolism, and genetic variability among individuals. Another limitation is the lack of clinical validation, as the computational findings require further real-world studies to confirm their applicability in patient populations. Over and above that, while RF and RF-RFE algorithms are effective in small datasets, their scalability remains a challenge in large-scale omics research, necessitating further algorithmic refinements to enhance their robustness and predictive accuracy.

Hence, the challenges identified highlight the need for further research, validation, and refinement of these methodologies. By addressing these limitations and leveraging emerging technologies, future research can enhance our ability to predict and mitigate drug-induced risks, ultimately improving maternal and infant health outcomes.

Glossary of terms with respect to the above text

• Activity: The effective concentration of a substance in a solution, accounting for its interactions with other molecules.

• ADRs (adverse drug reactions): Unintended or harmful reactions to medications.

• Alkylating agents: Chemicals that add alkyl groups to DNA, often used in chemotherapy.

• American Society of Health-System Pharmacists (ASHP): A professional organisation representing pharmacists who serve as patient care providers in hospitals and health systems.

• Analgesics: Medications used to relieve pain.

• Anticoagulants: Drugs that prevent blood clotting.

• Anticonvulsants: Medications used to prevent or control seizures.

• Antihypertensives: Medications used to lower blood pressure.

• Anti-infective agents: Substances used to treat or prevent infections.

• Anti-inflammatory agents: Substances that reduce inflammation in the body.

• Antituberculars: Drugs used to treat tuberculosis.

• Breastfeeding: The act of feeding an infant with breast milk.

• Central nervous system stimulants: Drugs that increase activity in the central nervous system, often used to treat conditions like ADHD.

• Chemical potential: The thermodynamic quantity representing the potential energy of a system to undergo a chemical reaction.

• Control experiments: Experimental procedures conducted to establish baseline measurements or validate results.

• Corticosteroids: Synthetic drugs that mimic the effects of hormones produced by the adrenal glands.

• Drug action: The effect of a drug on physiological processes or biochemical pathways.

• EC₅₀ (median effective concentration): The concentration of a substance required to produce a response halfway between the baseline and maximum after a specified exposure time.

• Ergot derivatives: Medications derived from ergot alkaloids, used to treat conditions like migraines and Parkinson’s disease.

• Ferguson’s principle: A principle that states that nonspecific drugs at similar relative saturations elicit quantitatively identical effects.

• Hormones: Chemical messengers that regulate various physiological processes in the body.

• Joint Commission on Accreditation of Healthcare Organisations (JCAHO): An independent, not-for-profit organisation that accredits and certifies healthcare organisations and programs in the United States.

• Laxatives: Substances used to promote bowel movements.

• Lethal toxicity: The ability of a substance to cause death.

• Lipid-soluble: Capable of dissolving in lipids or fats.

• Log milk Pap values: Logarithmic values representing the partition coefficients of drugs between milk and other solutions.

• Micelles: Tiny spherical structures formed by the aggregation of molecules in a colloidal solution, often consisting of a hydrophilic shell and a hydrophobic core.

• Minerals: Essential nutrients required for various bodily functions.

• Nonspecific depressants: Substances that depress or slow down physiological processes without targeting specific receptors or enzymes.

• Octanol–water partition coefficients: The ratio of a compound’s concentration in a mixture of octanol and water to its concentration in water, used to determine the compound’s lipophilicity.

• Oral contraceptive hormones: Synthetic hormones used in oral contraceptives to prevent pregnancy.

• Outliers: Data points that significantly deviate from the expected trend or distribution.

• Partition coefficients: Ratios used to describe the distribution of a compound between two immiscible phases, often octanol and water.

• Pharmacokinetics: The study of how drugs are absorbed, distributed, metabolised, and excreted in the body.

• Phosphate buffer: A solution used to maintain a stable pH environment in laboratory experiments.

• pK_a: The negative logarithm of the acid dissociation constant (K_a) of a solution, used to quantify the strength of an acid in solution.

• Plasma proteins: Proteins present in blood plasma that bind to and transport various substances, including drugs.

• Prophylaxis: The preventive treatment or intervention to prevent the occurrence of a disease or condition.

• Provitamins: Precursors to vitamins that can be converted into active forms by the body.

• Psychotherapeutic agents: Medications used to treat mental health conditions.

• Pyridoxine: Also known as vitamin B6, a water-soluble vitamin involved in various metabolic processes.

• Pyrimethamine: A medication used to treat and prevent malaria.

• Radiolabeled drugs: Drugs labeled with a radioactive isotope for tracking and measurement purposes.

• Relative saturation: The ratio of a substance’s concentration to its saturated solution concentration.

• Steroid-based compounds: Chemical compounds containing a steroid ring structure, often associated with hormonal activity.

• Thermodynamic equilibrium: A state in which there is no net change in the properties of a system over time.

• Thyroid affecting agents: Substances that influence the function of the thyroid gland.

• Tricyclic antidepressants: Medications used to treat depression, characterised by a tricyclic chemical structure.

• Ultrafiltrate: The portion of a liquid that passes through a filter under pressure, excluding larger molecules.

• Vitamins: Essential micronutrients required for various metabolic processes.

Funding

No financial support has been extended to back this study. The author alone has covered all costs associated with this research endeavour.

Conflicts of interest: The author declares that there are no conflicts of interest related to this study, its authorship, or its publication. There are no financial, personal, or professional relationships with individuals or organisations that could influence the integrity, objectivity, or outcomes of this research. The study was conducted in adherence to the highest ethical standards, ensuring integrity at every stage of the research study process.

References