When Does a Biomarker Assay Need Formal Bioanalytical Validation?

A biomarker assay can be scientifically useful long before it is ready to support a consequential clinical or regulatory decision.

The important question is not simply:

Does the method work?

It is:

Is the evidence supporting this method strong enough for the decision we intend to make with its results?

That distinction should drive the validation strategy.

A method being used in a clinical trial does not automatically require the same validation package as every other clinical assay.

Likewise, calling an assay “exploratory” does not make analytical performance unimportant.

The appropriate level of validation depends on what the biomarker is being used for, what decisions depend on the result, and what the consequences of analytical error would be.

FDA’s April 2026 guidance on Bioanalytical Method Validation for Biomarkers makes this principle explicit: the appropriate extent of method validation should follow a fit-for-purpose approach.¹

Start with intended use

Before selecting validation parameters or drafting a protocol, define what the biomarker result is supposed to accomplish.

FDA describes a biomarker’s context of use as its biomarker category together with its specific use in drug development.² Examples include supporting dose selection, defining trial enrollment, establishing proof-of-concept, or evaluating treatment response.

For the analytical team, that means starting with questions such as:

• Is the assay supporting exploratory research?
• Is it being used for candidate selection or an internal go/no-go decision?
• Will it provide pharmacodynamic or proof-of-concept information?
• Will it influence dose selection?
• Will it determine whether patients enter or remain in a study?
• Will the result contribute to an important safety or efficacy assessment?
• Will FDA rely on the biomarker data in making a pivotal regulatory decision?

Those are very different uses.

And they can justify very different amounts of analytical evidence.

An LC-MS method used to explore a biological mechanism is not evidentially equivalent to an assay whose result contributes to a pivotal determination of effectiveness—even if the instrument, analyte, and laboratory are identical.

The platform does not determine the validation strategy by itself.

The intended use does.

Think in terms of consequence

One useful way to determine validation rigor is to ask:

What happens if this measurement is wrong?

At one end of the spectrum, analytical error may weaken an exploratory conclusion or cause the team to repeat an experiment.

At the other end, an unreliable measurement could contribute to:

• an incorrect dose-selection decision;
• misclassification of subjects;
• a misleading safety conclusion;
• an unreliable efficacy assessment;
• or a regulatory conclusion the data cannot adequately support.

As the consequence of analytical uncertainty increases, the evidence supporting the method should generally become more rigorous and more traceable.

That does not mean there is one universal regulatory ladder in which every biomarker neatly moves through predefined validation tiers.

There is not.

But for development planning, it is useful to think about increasing levels of commitment.

Level 1: Fit-for-purpose characterization

Early in development, the immediate need may simply be to understand whether the assay is reliable enough to answer an internal scientific question.

That could include evidence around:

• selectivity;
• precision;
• sensitivity;
• analytical range;
• stability;
• matrix effects;
• reproducibility.

The exact work should reflect the question being asked.

The objective is not to claim that the method has undergone comprehensive regulatory validation.

It is to establish enough analytical understanding to make the intended internal decision responsibly.

FDA specifically recognizes this flexibility for biomarker assays used only for internal pharmaceutical decision-making, such as candidate selection, go/no-go decisions, or proof-of-concept.¹ In those situations, the sponsor can determine the extent of method validation appropriate to the use.

Level 2: Fit-for-purpose analytical validation

As the biomarker becomes more important to a clinical-development program, a more structured validation effort may be appropriate.

At this stage, the team may need predefined protocols and acceptance criteria addressing the particular analytical risks associated with:

• the intended biological matrix;
• clinically relevant concentrations;
• expected sample handling;
• the analytical platform;
• the specific decision the data will support.

The key word is still purpose.

A useful validation program is not defined by how many experiments appear in the protocol.

It is defined by whether those experiments establish the reliability of the method for its intended use.

This is where copying a generic validation template can become particularly dangerous.

A parameter that matters for one platform or context may be less relevant for another.

Conversely, an important analytical risk can be missed simply because it was not included in the template.

Level 3: Full bioanalytical method validation

When biomarker data will support consequential regulatory decision-making, the expectations become stronger.

FDA specifically states that when biomarker data will support regulatory decisions such as a pivotal determination of safety or effectiveness supporting approval, or dosage instructions in product labeling, the assay should be fully validated.¹ FDA identifies ICH M10 as the starting point for that validation approach, particularly for chromatography- and ligand-binding-based assays.^1, 3

Importantly, M10 itself was developed primarily for measurement of drug and active-metabolite concentrations and excludes biomarkers from its direct scope.³ FDA’s 2026 biomarker guidance addresses that gap by recommending that M10 principles serve as the starting point for biomarker validation while recognizing that some M10 characteristics or criteria may not apply to every biomarker or analytical platform.¹

That nuance matters.

Full validation does not mean blindly copying every M10 experiment and acceptance criterion into a biomarker protocol.

It means developing a comprehensive and scientifically justified body of evidence appropriate to the method and its intended regulatory use.

FDA encourages sponsors to discuss differing approaches with the appropriate review division early in development and to justify those differences in the validation report.¹

What should validation actually establish?

Regardless of the exact validation package, several underlying questions are remarkably consistent.

Does the method measure what we think it measures?

Consider:

• specificity;
• selectivity;
• interference;
• endogenous background;
• structurally related analytes.

How variable is the measurement?

Consider:

• accuracy;
• precision;
• within-run variability;
• between-run reproducibility.

Across what range is the method reliable?

Consider:

• sensitivity;
• analytical range;
• lower and upper limits of quantitation where applicable;
• performance around concentrations that matter to the intended use.

Can sample handling change the answer?

Consider:

• collection;
• processing;
• shipment;
• storage;
• freeze-thaw cycles;
• benchtop handling;
• processed-sample stability.

FDA identifies accuracy, precision, sensitivity, selectivity, parallelism, range, reproducibility, and stability among the important characteristics that may define biomarker-assay performance.¹ It also explicitly highlights sample collection, handling, and storage as potential sources of unreliability.¹

The validation plan should turn those general questions into experiments appropriate to the assay.

Formal validation is more than a laboratory event

Teams sometimes think of validation as something the laboratory performs.

That framing is incomplete.

A robust validation program is also a decision and documentation system.

Before execution, the team should make explicit:

• what is being measured;
• why it is being measured;
• the matrix and relevant concentration range;
• the analytical characteristics that matter;
• the experimental design;
• predefined acceptance criteria;
• run-acceptance rules;
• treatment of deviations;
• rules for repeats or exclusions;
• data-review procedures;
• and the conclusions the study is intended to support.

The laboratory generates the data.

The validation program determines what those data mean.

That distinction becomes especially important when work is outsourced.

The CRO should not define the intended use for you

A CRO or external laboratory can be an excellent technical partner.

It can recommend methods.

It can propose validation experiments.

It can execute a protocol.

It may have far more experience with the analytical platform than the sponsor.

But the sponsor still owns a fundamental question:

What does this result need to support in our development program?

Without that context, a vendor can execute a technically excellent study that does not fully answer the sponsor’s actual question.

The validation strategy should therefore be developed from the sponsor’s intended use and development requirements, then translated into an executable laboratory program.

Outsourcing the assay does not outsource the scientific intent.

Avoid validating too early

Full validation can be expensive.

More importantly, it freezes assumptions.

If fundamental assay characteristics are still changing during the validation study, the team may end up producing an impressive body of evidence for a method it no longer intends to use.

Signs that the method may still be in development include continuing changes to:

• extraction or sample preparation;
• critical reagents;
• calibration strategy;
• assay conditions;
• instrumentation;
• data-processing methods;
• reportable range.

Some changes during development are inevitable.

But before investing in formal validation, ask whether the method is sufficiently stable that the resulting evidence will remain relevant.

Otherwise, more characterization or method development may be the better next investment.

Avoid validating too late

The opposite problem can be equally expensive.

Sometimes a biomarker assay begins producing important clinical data while its analytical evidence remains scattered across:

• laboratory notebooks;
• feasibility experiments;
• qualification runs;
• emails;
• slide decks;
• SOPs;
• and institutional memory.

The assay may perform well.

But when someone later asks:

What evidence demonstrates that this method was fit for this use?

the answer has to be reconstructed retrospectively.

That is much harder than building the evidence trail deliberately.

Validation planning should occur before the point at which analytical uncertainty becomes a threat to the usefulness of irreplaceable study samples.

Do not borrow the wrong validation framework

Templates are useful.

They are also dangerous when they substitute for thinking.

Common mismatches include:

A drug-bioanalysis template applied mechanically to a biomarker assay.

The general principles may be useful, but biomarker-specific issues such as endogenous analyte, parallelism, or platform differences may require adaptation.

An exploratory research SOP stretched into clinical bioanalysis.

The method may have been perfectly adequate for discovery work while lacking the evidence or traceability needed for a more consequential use.

An IVD framework confused with drug-development biomarker bioanalysis.

These can overlap scientifically but may have different regulatory objectives and requirements.

The right framework begins with what the result will actually be used to support.

Method validation is not the same as biomarker qualification

This distinction is particularly important.

Analytical method validation asks whether a method reliably measures the biomarker for its intended use.

Biomarker qualification is a separate FDA process establishing that a biomarker can be relied upon for a specified interpretation and context of use across drug-development programs.⁴

FDA explicitly notes that a biomarker may be qualified while the specific assay used to measure it is not itself thereby “qualified.”⁵

An analytically excellent method does not by itself prove that the biomarker has the biological or clinical meaning being claimed.

Conversely, a scientifically useful or qualified biomarker still needs a reliable measurement method.

These are related problems.

They are not the same problem.

And an IVD may introduce another regulatory pathway

There is another boundary worth making explicit.

FDA’s 2026 biomarker bioanalytical guidance addresses methods used to measure in vivo biomarker concentrations in biological matrices in drug-development settings.¹

If the assay itself is being developed as a clinical diagnostic, companion diagnostic, or other in vitro diagnostic device, additional device-specific regulatory considerations may apply.^1, 6, 7

A bioanalytical validation strategy should therefore not automatically be treated as the complete validation strategy for an IVD product.

Again:

Start with intended use.

A practical decision framework

Before deciding whether to move into formal validation, I would ask six questions.

1. What exactly is the assay being used for?

Write it in one sentence.

If the answer is vague, the validation strategy will probably be vague too.

2. What decision depends on the result?

Internal scientific decision?

Dose selection?

Proof-of-concept?

Patient selection?

Pivotal safety or effectiveness conclusion?

Labeling?

3. What is the consequence of analytical error?

The more consequential the error, the stronger the analytical evidence should generally be.

4. Is the method sufficiently stable to validate?

If the method is still undergoing fundamental development, characterize and stabilize it first.

5. What evidence already exists?

Do not automatically rerun experiments simply because they appear in a validation template.

Inventory what has already been demonstrated and determine what remains missing.

6. What will a future reviewer need to reconstruct?

Can another qualified scientist understand:

Requirement → Experiment → Criterion → Result → Conclusion?

If not, the problem may be as much documentation and traceability as laboratory performance.

The useful sequence

For most early-stage teams, I prefer this sequence:

1. Define intended use

↓

2. Identify the decision and consequence of error

↓

3. Assess the analytical risks

↓

4. Inventory existing evidence

↓

5. Determine the appropriate validation rigor

↓

6. Define experiments and acceptance criteria

↓

7. Execute under a predefined protocol

↓

8. Document what the evidence supports

That sequence produces something very different from:

Find a validation template and fill in the blanks.

It produces an evidence strategy.

The bottom line

A biomarker assay should not be validated simply because “validation is what comes next.”

Nor should validation be postponed simply because a biomarker is labeled exploratory.

The appropriate question is:

What are we asking this result to support, and what analytical evidence is necessary to trust it for that purpose?

FDA’s current biomarker guidance reflects exactly that fit-for-purpose principle.¹ Internal decision-making may justify a sponsor-defined extent of validation, while biomarker assays supporting pivotal regulatory determinations or dosage information should be fully validated, with M10 principles serving as an important starting point.^1, 3

The objective is not maximum validation.

It is the right evidence, generated at the right time, for the decision that actually matters.

References

1. U.S. Food and Drug Administration. Bioanalytical Method Validation for Biomarkers: Guidance for Industry. April 2026.
2. U.S. Food and Drug Administration. Context of Use. Biomarker Qualification Program.
3. U.S. Food and Drug Administration. M10 Bioanalytical Method Validation and Study Sample Analysis: Guidance for Industry. November 2022.
4. U.S. Food and Drug Administration. Biomarker Qualification Program.
5. U.S. Food and Drug Administration. About Biomarkers and Qualification. Biomarker Qualification Program.
6. U.S. Food and Drug Administration. In Vitro Companion Diagnostic Devices: Guidance for Industry and Food and Drug Administration Staff. August 2014.
7. U.S. Food and Drug Administration. In Vitro Diagnostics.