Biologics CMC Series, Part 6: Biologics Process Validation — Are Three Batches Enough for Commercial Manufacturing?
Executive Summary
A batch that meets specifications provides evidence about that batch's acceptability. Process validation builds confidence that the process can consistently deliver a product meeting its quality requirements.
This article addresses four questions: why validation begins during development, why the number of PPQ batches needs justification, what to monitor after commercialization, and how manufacturing changes connect back to control strategy and lifecycle management.
Monoclonal antibody manufacturing is the main example. All cases and numerical values are hypothetical teaching examples, not company data or universal acceptance criteria.
Introduction: Making the Product and Establishing a Reliable Process
Imagine a team has scaled up an antibody process. Purity and potency meet their criteria, and output is on target. Management then asks: “If we make another three batches, can we get ready for launch?”
It is a practical question. Three batches represent materials, facility time, staffing, and money. They also offer a visible endpoint for the project.
My first question would be: What are those three batches intended to demonstrate?
Suppose all three use the same medium lot, the same operators, and relatively new chromatography resin. Consistent results are encouraging. But how much confidence do they provide when a material lot changes, another shift takes over, or the resin approaches the end of its usable life?
The point is not to force every possible scenario into three production runs. Successful batches and sufficient evidence need to be evaluated separately.
The Biologics Process Validation Framework: From Design to Continued Monitoring
FDA's Process Validation: General Principles and Practices describes three stages: Process Design, Process Qualification, and Continued Process Verification. Process Performance Qualification (PPQ) is part of Stage 2; it does not represent the entire lifecycle. FDA process validation guidance
In project terms, I think of these as three questions:
Why do we believe this process design is appropriate?
Can it be executed reproducibly under the intended commercial manufacturing conditions?
Does that conclusion remain valid as routine production continues?
This follows the earlier articles in the series. Specifications define acceptance criteria, the control strategy organizes the controls, and process validation provides evidence that those arrangements work.
Stage 1: Development Data Must Address Commercial Manufacturing Questions
Process development can easily become focused on output. For an antibody program, a higher titer or better purification recovery can directly affect cost and supply capacity.
But if increased output comes with changes in aggregates, charge heterogeneity, or other quality attributes relevant to product function, the team needs to understand whether the change is acceptable and which conditions require control.
ICH Q11 connects understanding of material attributes, process parameters, and drug substance quality with process development and the control strategy. A development report should explain why conditions were selected, as well as record which conditions worked. ICH Q11
For a hypothetical CHO antibody process, I would ask the team to organize three groups of questions:
Cell culture: Which quality attributes might be affected by pH, temperature shifts, and feeding schedules?
Purification: How might loading, residence time, and material differences affect separation performance?
Between operations: What risks arise from the temperature and duration of intermediate holds?
The purpose is not to classify every parameter as a CPP. It is to make the basis for important decisions traceable.
What Does a Small-Scale Model Represent About the Commercial Process?
Not every biologics study can be performed at commercial scale, which makes small-scale models essential. EMA's guideline for biotechnology-derived active substances brings process characterization, model suitability, and commercial-scale evidence into a common framework. EMA guideline on process validation for biotechnology-derived active substances
One easily overlooked issue is the model's intended use.
For example, a small chromatography column used to study impurity clearance does not also demonstrate consistent packing of the commercial column. Likewise, a small culture system used to study the effect of feeding on quality may not fully reproduce the mixing environment in a large vessel.
I would expect each model to explain three things: the question it addresses, the commercial process characteristics it needs to preserve, and the differences that limit extrapolation.
A model can have limitations and still be useful. Leaving those limitations unexplained makes its results harder to defend during technology transfer or regulatory review.
Stage 2: Qualified Equipment Is the Starting Point for Evaluating the Process
Equipment qualification and PPQ evaluate different things. The former addresses whether equipment, facilities, and utilities are suitable for their intended use. The latter evaluates the process together with materials, personnel, and controls. FDA includes both within Process Qualification. FDA process validation guidance
For example, a bioreactor's ability to control temperature accurately is necessary, but does not by itself establish that the antibody process can consistently achieve the intended quality throughout culture.
Before PPQ, I would hold a readiness review centered on evidence gaps, rather than just completion percentages on the project schedule.
The discussion should address unexplained observations, whether measurement methods support the intended decisions, whether sampling can detect the variability of concern, and who will investigate unexpected results and assess their effect on the overall conclusion.
These questions are best resolved before manufacturing begins. Once expensive batches have been completed, schedule pressure can easily influence discussions about how to judge the results.
Must PPQ Include Three Batches? Understand the Guidance and Its Scope
FDA's lifecycle approach does not establish a fixed three-batch rule as sufficient evidence for every process. Validation design needs to account for process understanding and variability. FDA process validation guidance
Under the traditional approach, EU GMP Annex 15 states that a minimum of three consecutive batches manufactured under routine conditions is generally considered acceptable, while also requiring justification of batch numbers and allowing a justified alternative. Additional ongoing verification data may still be needed. The relevant passage concerns finished-product manufacturing; it should not be treated as a universal rule for every biologic drug substance. EU GMP Annex 15, sections 5.18–5.20
The practical question is whether the proposed number supports this product, this process, and this regulatory submission.
“Everyone uses three batches” is not a justification. Nor does a risk-based approach mean batch numbers can be reduced arbitrarily.
A Hypothetical Antibody Program: Assigning Questions to the Right Evidence
Consider a CHO monoclonal antibody program preparing for commercial manufacturing at 2,000 L. This scale is a teaching assumption.
The team identifies three areas of concern: the relationship between culture conditions and quality attributes, changes in chromatography performance with use, and intermediate hold times.
I would assign these questions to different sources of evidence, rather than expect PPQ to answer everything.
Development and characterization studies clarify parameter effects and appropriate ranges. Suitable model studies investigate specific risks. Commercial-scale execution establishes whether the actual equipment, people, and procedures work together as intended.
For example, a few production batches using new resin cannot resolve the resin-lifetime question. Conversely, even if a model shows stable resin performance, commercial use, cleaning, and storage records still matter. EMA's discussion of process evaluation and verification provides context for combining these sources of evidence. EMA guideline on process validation for biotechnology-derived active substances
The final report should explain which data support each risk assessment, what uncertainties remain, and how those uncertainties will be followed after commercialization.
Stage 3: Results Within Specification Can Still Warrant Investigation
Continued Process Verification addresses whether the process remains in a state of control during routine production. This is the core of FDA's third stage. FDA process validation guidance
Suppose an impurity has a release limit of 100 units. Historical results are around 20–30, but recent consecutive batches give 45, 55, and 65. These values are entirely hypothetical and must not be used as a product specification.
All three may meet that release criterion. Yet the team should still ask whether the pattern reflects measurement differences, material changes, equipment condition, or process drift.
This illustrates the different roles of specifications and trend monitoring. Specifications establish batch acceptance criteria; trend analysis helps identify changes. Neither replaces the other.
In practice, I would avoid starting with a dashboard crowded with metrics. First select indicators relevant to product quality and known risks. Then define when assessment is needed, who is responsible, and what follows that assessment.
Charts without responsibility for action can easily become another page in an annual report.
Continued and Continuous Process Verification: Spell Out the Terms
FDA's Continued Process Verification refers to the third lifecycle stage. Continuous Process Verification in EU Annex 15 is an approach that may replace traditional process validation; the Annex also includes ongoing process verification requirements across the lifecycle. Documents should spell out the terms so that teams do not assume every use of “CPV” means the same thing. FDA guidance, EU GMP Annex 15, sections 5.23–5.32
Working With a CDMO: Plan Data Deliverables Early
For teams outsourcing manufacturing, one situation I would particularly want to avoid is receiving a report without being able to understand the reasoning behind it.
The sponsor may see final results but lack access to trend data. It may see that a deviation was closed without understanding the evidence for the root cause or the follow-up. That makes prior knowledge harder to use when the next manufacturing change arises.
ICH Q10 places oversight of outsourced activities, process performance and product quality monitoring, CAPA, change management, and management review within the pharmaceutical quality system. ICH Q10
My recommendation is to define data formats, review responsibilities, deviation notifications, and long-term monitoring roles early in the project. These arrangements affect technology transfer and problem-solving capacity as well as contract management.
Jason’s Take Away: Know What Supports Your Confidence in the Process
On a project schedule, PPQ can look like a gate to pass before launch.
Scientifically and operationally, it is an opportunity to clarify the basis for confidence: which conclusions are supported by data, which conditions have been evaluated, and which risks still require observation.
What matters most to me is whether the team can explain a difference in the next batch and knows what to do about it.
A good validation strategy allows process development, analytical, manufacturing, QA, and regulatory teams to discuss questions using a shared body of evidence. That is how validation data remain useful after commercialization.
LuTra Studio: Turning Process Knowledge Into an Actionable CMC Plan
If your team is moving from early development toward technology transfer or commercialization, LuTra Studio can help organize CMC evidence gaps, connect process development with control strategy, and frame technical discussions with CDMOs.
Our aim is to help teams clarify which data they still need, why they need them, and when to obtain them, so scientific decisions translate into an executable plan.
Contact LuTra Studio to discuss your product's stage and project needs.
Comments