Introduction
Extractables and leachables (E&Ls) represent a critical class of drug product impurities originating from materials used in manufacturing systems, packaging, and delivery devices. These compounds have the potential to migrate into drug products and must be controlled within safe exposure limits to ensure patient safety.
Regulatory institutions recognize the complexity and diversity of E&Ls and the need for a structured, science-driven approach to their assessment. The draft ICH Q3E guideline establishes a harmonized framework for evaluating E&Ls using risk-based principles.
The recent peer-reviewed research study “Exploring the chemical space of pharmaceutical extractables and leachables”, demonstrates how advanced computational approaches support this evolution by transforming a highly complex chemical space into an organized and assessable framework.
The Importance of the ICH Q3E Guideline
ICH Q3E represents a significant advancement in the safety assessment of E&Ls by introducing a unified risk-based strategy. At its core is the concept of the Safety Concern Threshold (SCT), defined as the exposure level below which toxicity concerns are considered negligible.
This framework integrates multiple scientific principles:
- Threshold of Toxicological Concern (TTC) for mutagenic risk
- Qualification Thresholds (QT) for non-mutagenic effects
- Consideration of route of administration and exposure duration
The guideline emphasizes that the lowest applicable threshold, whether mutagenic, non-mutagenic, or local toxicity, should guide decision-making.
Importantly, ICH Q3E encourages the application of read-across approaches when experimental data is limited. This reflects the practical reality that a significant proportion of E&Ls lack comprehensive toxicological datasets. In the referenced study, approximately 50% of substances had no experimental toxicity data, underscoring the need for predictive and read-across methodologies.
By formalizing these concepts, ICH Q3E provides a consistent scientific foundation for evaluating E&Ls across pharmaceutical development programs.
The Challenge: Complexity of the E&L Chemical Space
E&Ls originate from diverse materials, including polymers, elastomers, adhesives, and excipients. Consequently, the resulting chemical space is highly heterogeneous, encompassing a wide range of structures, physicochemical properties, and toxicological profiles.
The study compiled a dataset of over 1200 potential E&L compounds and demonstrated that:
- The dataset spans a broad range of chemical classes
- Structural diversity is significant, with both highly populated and sparsely populated groups
- Only a small proportion of compounds exhibit concerning toxicological properties, such as mutagenicity (~9%) or strong sensitization (~3%)
This complexity presents a major barrier to traditional assessment approaches, which rely heavily on compound-specific data.
Leadscope’s Role: Structuring Chemical Complexity
Leadscope played a central role in enabling the analytical framework used in this research. Its computational toxicology and cheminformatics capabilities supported several critical steps:
1. Systematic Chemical Clustering
Leadscope’s structure-based clustering algorithms were used to organize the dataset into:
- 78 Tier-1 clusters representing broad chemical classes
- 323 Tier-2 subclusters providing fine-grained structural distinctions
This two-tier system transformed an unstructured dataset into a hierarchical framework that reflects both chemical similarity and diversity.
2. Physicochemical Characterization
The platform enabled the calculation of key descriptors such as molecular weight and logP, which are essential for understanding behavior, exposure potential, and relevance to safety thresholds.
The analysis showed that most E&Ls fall within a consistent physicochemical range, indicating that the compiled dataset represents the broader E&L chemical space.
3. In Silico Toxicity Prediction
Leadscope Model Applier was used to generate predictive assessments of:
- Mutagenicity
- Skin sensitization
These predictions, combined with available experimental data, provided a comprehensive hazard profile for each compound.
Learn more about the capabilities of Leadscope Model Applier.
Enabling Risk-Based Strategies Through Classification
The systematic classification enabled by Leadscope directly supports the goals of ICH Q3E by making the E&L chemical space more interpretable and actionable.
Representative Sampling of Chemical Space
By organizing compounds into clusters, the study demonstrated that a curated dataset can effectively represent the broader E&L universe, with minimal gaps in chemical coverage.
This is critical for deriving safety thresholds, as representative sampling ensures that conclusions are scientifically justified across the entire space.
Read-Across for Data-Poor Substances
Cluster-based grouping allows toxicological information from data-rich compounds to be applied to structurally similar, data-poor substances.
Given that roughly half of the compounds in the dataset lack sufficient experimental data, this capability is essential for efficient and ethical risk assessment.
Identification of Priority Clusters
The long-tail distribution observed in the dataset highlights that:
- A small number of clusters contain a large proportion of compounds
- Many clusters contain only a few substances
- The chemical space is diverse
Benefits to the Pharmaceutical Community
The integration of Leadscope methodologies into E&L assessment delivers several tangible benefits aligned with ICH Q3E:
1. Strengthened Scientific Foundation
The combination of predictive toxicology, toxicity profiling, physicochemical analysis, and structural clustering provides a robust, evidence-based framework for decision-making.
2. Efficient Handling of Data Gaps
In silico methods and read-across reduce reliance on new experimental studies while maintaining scientific rigor, particularly for compounds lacking toxicological data.
3. Improved Consistency in Risk Assessment
Defined read-across processes ensure that similar compounds are evaluated using consistent criteria, reducing variability across programs.
4. Accelerated Development Timelines
Computational approaches streamline the evaluation process, supporting faster and more efficient product development.
5. Alignment with Regulatory Expectations
The methodologies demonstrated in this research directly support the principles outlined in ICH Q3E, positioning organizations for compliance with emerging global standards.
Conclusion
The draft ICH Q3E guideline marks a pivotal shift toward risk-based E&L safety assessment. However, the successful implementation of this framework depended on the ability to manage the inherent complexity of the E&L chemical space.
The referenced research demonstrates that Leadscope provided the necessary tools to achieve this transformation. Through systematic chemical classification, predictive modeling, and data integration, it enables a structured and scientifically defensible approach to E&L evaluation.
For pharmaceutical teams, this translates into more confident decision-making, particularly when addressing data-poor substances and applying read-across methodologies. Ultimately, these capabilities support the development of safer products through evidence-based strategies aligned with modern regulatory expectations.
Request a demo of Leadscope Model Applier to explore how predictive toxicology and systematic read-across can strengthen your E&L safety assessment strategy.
