Aspis-cluster

Integrated Testing Strategies: One Decision, Many Data Sources

Why One Test Is Rarely Enough

Regulatory decisions about chemical safety seldom rest on a single experiment. A skin sensitisation assessment, for example, may draw on in chemico reactivity data, in vitro cell-based assays, and historical in vivo observations. Each source answers a different question, and together they describe a biological pathway rather than a single endpoint. That is the core idea behind Integrated Testing Strategies, often abbreviated as ITS.

An ITS is not a random collection of tests. It is a structured plan that defines which evidence streams are needed, in what order, and how their outputs will be combined. The strategy is designed before the testing begins, so that every assay has a defined role in the final decision.

Many Data Sources, One Decision Logic

Different data sources carry different weights. Mechanistic assays can clarify whether a substance interacts with a key molecular target. In vitro assays can show cellular responses under controlled conditions. Physicochemical and exposure information can determine whether a hazard is likely to be realised in practice. Computational models can fill gaps where testing is impractical.

The decision logic of an ITS connects these streams through defined rules. A common structure is a tiered workflow:

  1. Gather existing data and structural alerts.
  2. Run targeted in vitro or in chemico assays.
  3. Apply computational predictions for missing properties.
  4. Integrate all evidence through a weight-of-evidence assessment.

Building Blocks of a Combined Plan

Several elements appear repeatedly in well-designed strategies. They help ensure that the plan is transparent, reproducible, and acceptable to regulators.

  • Clearly stated decision context and regulatory question.
  • Defined applicability domain for each method.
  • Criteria for positive, negative, and inconclusive results.
  • Rules for combining results across methods.
  • Documentation of uncertainty and data gaps.

These building blocks turn separate measurements into a coherent argument. Without them, data remain isolated and difficult to interpret.

Integrating Evidence in Practice

A typical integrated assessment can be summarised as a matrix that maps evidence sources to decision criteria. The table below illustrates a simplified example for a hypothetical substance.

Evidence sourceType of dataContribution to decision
In chemico assayReactivitySupports mechanistic concern
In vitro assayCellular responseConfirms biological activity
QSAR modelPredicted propertyFills data gap
Exposure estimateReal-life scenarioRefines risk context

As one toxicologist involved in the ASPIS project noted, the value of an ITS lies in the reasoning: "The decision is not made by one test, but by a transparent chain of evidence." This quotation captures why regulators increasingly accept integrated approaches.

Advantages and Open Questions

Integrated Testing Strategies can reduce animal testing, save resources, and improve mechanistic understanding. They also make uncertainty explicit, which supports more defensible decisions. However, they require robust validation, clear reporting standards, and trained assessors who can interpret conflicting results.

The ASPIS cluster works to address these challenges by sharing methods, data, and case studies across projects. The goal is not to replace expert judgement, but to give experts better structured evidence. When one decision draws on many data sources, the quality of the integration determines the quality of the outcome.

Latest articles