Bacterial retention test (BRT): how sterilizing filtration is validated with your product

The bacterial retention test (BRT) challenges a filter membrane with a very high load of a bacterium of known size, typically Brevundimonas diminuta: at least 10⁷ colony-forming units (CFU) per cm² of filtration area. If the filtrate comes out sterile, the membrane is considered sterilizing-grade. It is a destructive test: the challenged filter is not used again.

A cartridge’s retention certificate comes from that test, run at the factory with water or saline solution: it describes the filter, not your process. Product-specific validation repeats the challenge with your product and under your actual conditions.

Process validation and filter validation: what each one covers

• Process validation: covers the entire manufacture of the product, including each compounding and filling step. It may include filtration, but it is generally limited to specifying the filter.
• Filter validation: studies that step in depth through several tests: bacterial retention, compatibility, adsorption, extractables, and leachables.

Filter validation can be performed under standard or product-specific conditions. ASTM F838-20 draws the line in its own scope: its method measures retention under standard conditions and is not intended for product-specific validation, which has to use the product, or a surrogate fluid, and the process parameters. The FDA’s 2004 guidance on aseptic processing states that filter validation data should apply to the filter user’s products and conditions of use, even when someone else runs the test.

Aspect	Standard conditions	Product-specific
Who performs it	The filter manufacturer	The plant, with the manufacturer or a laboratory
Fluid	Water, saline solution, ethanol, or another reference solvent	The product or a justified surrogate
Conditions	Those of the reference method	Those of the process, at worst case
What it demonstrates	That the membrane is sterilizing-grade	That retention holds with that product and that process
Integrity limit	With the reference wetting liquid	With the product, if the filter is tested wetted with it

How a bacterial retention test is run with your product

The basis is the ASTM F838 method, which we review in what 0.2 µm means and how a sterilizing-grade filter is validated. What changes is everything around the challenge:

1. Define the worst case and, from it, the test conditions.
2. Study the viability of the bacterium in the product.
3. Measure the integrity of the filters to be challenged.
4. Challenge the filters with the inoculated product, or the fluid that replaces it, at the defined pressure, flow rate, time, and temperature.
5. Recover and count: all of the filtrate passes through an analysis membrane, which is incubated, and the colonies are counted.
6. Close out: the integrity test is repeated and the controls are reviewed.

The acceptance criterion does not change: no colonies of the challenge organism in the filtrate, with the controls performing as expected.

What happens if the product affects the challenge bacterium

If the product kills or inhibits the bacterium, the filtrate comes out sterile even if the membrane let cells through. That is why a viability study is done first: B. diminuta is inoculated into the product, and the population is counted at the beginning and end of the maximum contact time, at the process temperature. If the population holds, the bacterium is inoculated directly into the product.

The FDA’s 2004 guidance prefers that route because it assesses the effect of the product on both the membrane and the bacterium at the same time, and it warns that, with bactericidal or oil-based products, direct inoculation can lead to erroneous conclusions. EU GMP Annex 1 follows the same line: the product whenever possible and, if it is not suitable, a justified surrogate.

Typical cases are products containing preservatives, antibiotics, alcohol, or solvents, or with an extreme pH. PDA Technical Report 26 addresses these fluids, and the usual approaches are:

• Adjust the method, for example by shortening the bacterium’s exposure to the product.
• Modify the formulation: remove the preservative or the antimicrobial component and use it as the vehicle.
• Use a surrogate solution with viscosity, surface tension, pH, and ionic strength similar to those of the product, but harmless to the bacterium.
• Separate exposure and challenge: first filter the actual product under worst-case conditions and then, through the same filter, the challenge in a modified fluid.
• Use a bioburden isolate that is resistant to the product, if it is justified as an equivalent or greater challenge.

Any difference from the actual product is justified in the protocol. With alcohol or solvents, the filter also has to be compatible: the nylon 6,6 UltraPure Plus Pharma is indicated in its catalog for hydroalcoholic mixtures and solutions containing solvents.

How the worst case is chosen

When a plant filters several products, a worst-case approach is usually adopted: grouping the products into families and testing the least favorable one. Annex 1 provides for grouping the validation by strengths or variants of a product, provided it is performed under worst-case conditions and the grouping is justified. PDA Technical Report 26 frames retention studies on the basis of a risk assessment. Some parameters to consider:

Parameter	What is assessed	Usual worst case
Batch size	Volume filtered per batch	The largest batch
Prefiltration	Step upstream of the sterilizing-grade filter	No prefiltration
Filter loading	Liters per cm² of filtration area	The highest
Reuse	Batches with the same filter	The maximum number of batches
Differential pressure	Pressure drop across the filter	The maximum operating value
Flow rate per unit area	Flow rate divided by filtration area	The maximum
Filtration time	Product contact with the filter	The longest

Temperature and product properties also count: viscosity, surface tension, pH, and osmolality. If the filter is reused, the FDA’s 2004 guidance states that the maximum number of batches should be validated; the usual practice is to discard it after each one.

These parameters are used to build a scoring system: each product receives a value per parameter—for example, from 1 to 3 according to risk—the values are added up, and the product with the highest score is the worst case for the family. For the prefiltration parameter, the stage-by-stage overview of filtration for pharmaceuticals and biotech shows the prefilters that protect the sterilizing-grade filter.

Positive and negative controls, and good microbiological practices

A count of zero is only valid if the test was capable of detecting bacterial passage. That is why controls are run in parallel:

• Positive control: a more open membrane, usually 0.45 µm, challenged with the same suspension. It has to let bacteria through: this confirms that the cells arrived small and dispersed. According to the FDA’s 2004 guidance, B. diminuta has a mean diameter of 0.3 µm.
• Negative control: the same flow path, assembled and sterilized the same way, with uninoculated product. There should be no growth.
• Challenge count: the suspension is titered to confirm the load the filter received and that the bacterium remained viable.

Good microbiological practices prevent the opposite error, downstream contamination that is read as a filter failure: the filtrate flow path is assembled as a closed system and sterilized, the connections are aseptic, and the challenge suspension does not touch the filtrate side. If the product leaves residues that inhibit growth, the analysis membrane is rinsed before it is incubated.

How the test correlates with filter integrity

The challenge cannot be repeated on the filter that goes on the line. What extends its result to each production filter is a non-destructive physical test—bubble point, diffusive flow or, for hydrophobic membranes, water intrusion—whose results, according to Annex 1, should be correlated with the retention established in the validation. Each method is covered in detail in filter integrity test: bubble point and diffusive flow.

Product-specific validation adds one more consideration: the wetting liquid. If the plant tests the membrane wetted with product, the limit the manufacturer set with water does not apply as is, because a product with lower surface tension lowers the bubble point. Annex 1 includes that liquid and its specification value among the validation parameters, and PDA Technical Report 26 covers product-wetted tests in detail, for example through the ratio between the bubble point measured with product and with the reference liquid.

On the line, the SaniFlow STU-01 sanitary housing comes with a vent and drain that make it easier to test the installed cartridge.

How we do it: membrane, lot, and challenge under the same control

We developed our own method based on ASTM F838-20, aligned with the product-specific approach of Annex 1 and PDA Technical Report 26. It is run with your product and under your actual process conditions, on cartridges and capsules supplied by FluidsControl:

1. Challenge suspension: a pressurized 20 L vessel with Brevundimonas diminuta ATCC 19146, for a challenge of 1 × 10⁷ CFU/cm².
2. Sterilizing-grade filter: a 0.2 µm membrane cartridge in a 10″ housing.
3. Analysis membrane: the filtrate is collected in a 47 mm membrane holder for counting. A retentive filter gives zero CFU.

The challenge is performed on the complete filter; the FDA’s 2004 guidance notes advantages in using production filters in these studies. Because we manufacture the cartridge and run the test, the membrane, the lot, and the challenge are under the same control: pharmaceutical-grade cartridges are engraved with the lot, model, pore size, and part number.

Depending on the product, the filter can be PES, like the LifePure Plus Pharma; nylon 6,6, like the UltraPure Plus Pharma; or dual-layer for heavily loaded fluids, like the LifePure Protec Pharma and the UltraPure Protec Pharma. All four declare B. diminuta retention per ASTM F838-20 and have a validation guide: the starting point for the study with your product.

Tell us what product you filter

We don’t tell you how to run the test: we run it for you. If you are putting together the validation of a sterilizing filtration process, write to our engineering team with the product, the volume per batch, the filter you use, and the process conditions—pressure, flow rate, time, and temperature—and we will look at your specific case.