Risk-Based Cleaning Validation: Applying ICH Q9 Principles in Practice

Cleaning validation remains one of the most frequently cited areas during FDA and EMA inspections. Despite decades of guidance — from the FDA's 1993 Cleaning Validation Guide to EMA's 2014 Guideline on Setting Health-Based Exposure Limits — many organisations continue to struggle to translate risk principles into a defensible, inspection-ready programme. ICH Q9(R1), published in January 2023, brought important refinements to pharmaceutical risk management. While Q9 is not a cleaning validation guideline per se, its revised principles have direct implications for how we justify and scope cleaning validation activities.

Why Cleaning Validation Keeps Getting Cited

The persistence of cleaning validation findings at inspection stems from several root causes that are worth naming directly:

• Legacy acceptance criteria: Many programmes still use visually determined limits or the historical "10 ppm / 0.001 dose" rules as primary criteria — approaches that regulators no longer accept as scientifically defensible standalone positions.
• Worst-case selection without documentation: Identifying a "worst-case" product or surface without documented, scientifically supported rationale is routinely challenged during inspection.
• Static validation approaches: Validating at a single cleaning cycle without understanding the method's operating window creates fragility the moment process conditions change.
• Incomplete HBEL framework adoption: Some organisations have adopted Health-Based Exposure Limits for acceptance criteria but have not fully integrated them into risk assessment, equipment grouping, or bracketing strategies.

The HBEL-Anchored Risk Assessment

The starting point for a modern cleaning validation programme is the derivation of a Health-Based Exposure Limit — specifically the Permitted Daily Exposure (PDE) or Acceptable Daily Exposure (ADE) — for each active pharmaceutical ingredient in your shared equipment train. The PDE/ADE, derived by qualified toxicologists using established methodology (EMA/CHMP guideline EMA/CHMP/CVMP/SWP/169430/2012), provides the scientifically defensible benchmark for "how clean is clean enough."

From the PDE/ADE, you calculate a Maximum Allowable Carryover (MACO) into the next product, accounting for:

• Batch size and minimum therapeutic dose of the subsequent product
• Product contact surface area of the equipment train
• Recovery factor validated from your swab or rinse sampling method

This calculation replaces arbitrary limits with health-protective, product-specific thresholds. Regulators expect this framework — and if you cannot provide it, you will be asked why.

Risk Assessment Drives Scope, Not Just Criteria

Beyond acceptance criteria, the risk assessment should drive the design and scope of your validation programme. This is where ICH Q9(R1)'s emphasis on reducing subjectivity becomes most practically relevant.

Equipment grouping: Use documented product properties — solubility profile, physical form, toxicological classification — and equipment characteristics to group items and justify representative worst-case selection. A risk matrix with pre-defined, weighted scoring criteria is far more defensible than a narrative statement asserting "Product X was selected as worst-case." The matrix forces explicit reasoning and enables independent review.

Campaign limits: Risk assessment should establish scientifically based campaign limits — the number of batches or manufacturing time permitted before mandatory cleaning — rather than relying on fixed calendar intervals. The scientific basis must account for residue build-up kinetics and equipment design.

Degradation products: Consider whether active pharmaceutical ingredients degrade under cleaning conditions into compounds with greater toxicity or reduced solubility. This is an area of increasing regulatory scrutiny, particularly for oncology and highly potent compounds where the parent HBEL may be already stringent.

Visual Inspection: Is It Sufficient?

FDA's 1993 guidance established "visibly clean" as the minimum standard. Visual inspection today is best understood as a limit test — a floor below which you must not fall — rather than a standalone release criterion for potent or sensitising compounds.

For highly potent APIs (PDE less than 10 µg/day), visual inspection alone is essentially never sufficient. Analytical testing, typically reversed-phase HPLC or TOC measurement, must confirm residue levels below the MACO. The risk assessment should explicitly address this and document the decision logic, not simply assume visual inspection is adequate because the surface appears clean.

For lower-hazard compounds, a risk-based argument for visual inspection as the primary criterion remains possible but must be clearly justified in documentation — referencing the HBEL, the calculated MACO, and the visual detection limit relevant to your specific equipment geometry and surface finish.

Lifecycle Approach: Beyond One-Time Validation

Just as analytical procedures and manufacturing processes are now expected to be managed across a lifecycle, cleaning validation programmes should evolve from one-time compliance exercises to continuous performance verification frameworks.

A lifecycle-structured programme has three stages:

1. Cleaning Development (Design): Understand your cleaning agent's efficacy against the target residues, identify critical cleaning variables — contact time, temperature, concentration, flow dynamics — and define the cleaning operating space. This stage generates the scientific rationale that underpins everything that follows.
2. Cleaning Qualification (Validation): Demonstrate that the cleaning procedure, performed within its defined operating space, consistently achieves acceptance criteria under worst-case conditions. Three cleaning cycles under worst-case conditions remains the regulatory expectation for traditional validation approaches.
3. Continued Cleaning Verification: Ongoing monitoring through periodic rinse sampling, trend analysis of TOC or HPLC data, and periodic re-evaluation of the risk assessment detects performance drift before it becomes a compliance issue.

ICH Q9(R1) in Practice

The 2023 revision of ICH Q9 sharpened focus on two principles especially relevant to cleaning validation. First, reducing subjectivity: Q9(R1) explicitly acknowledges that poorly structured risk assessments — where the outcome is predetermined by evaluator bias or organisational pressure — undermine the entire value of risk management. Structured risk matrices, defined scoring criteria, and multi-disciplinary cross-functional review are the practical antidotes.

Second, proportionality: risk management effort should match the actual risk. Not every equipment item or API warrants the same depth of validation. A dedicated single-product train manufacturing a solid-dose vitamin supplement and a multi-product facility manufacturing hormonal APIs require fundamentally different programme architectures. Document the reasoning, not merely the conclusion.

Building an Inspection-Ready Programme

A cleaning validation programme that holds up under rigorous inspection has three characteristics. It is scientifically anchored — in HBELs, validated sampling methods, and documented risk assessments. It is operationally integrated — cleaning parameters are controlled, monitored, and linked to the quality system. And it is documented with auditable rationale — not simply records of compliance, but records of the thinking that led to each key decision.

Regulators are not expecting perfection at first inspection. They are expecting a credible programme with documented scientific rationale, honest identification of gaps, and a clear, time-bound remediation trajectory. Starting from scratch or overhauling a legacy programme? Begin with a systematic risk assessment across your entire equipment portfolio and API list. Prioritise by risk. Then remediate in a structured programme with measurable milestones.