Filter validation: Information from the filter supplier


When buying filters intended to be used for sterile filtration during the production of liquid parenterals, the filter manufacturer can generally provide the pharmaceutical company with a detailed filter validation documentation besides the certificate of release for the filter. This documentation and the studies performed respectively are going to be briefly explained in the following paragraphs.

A bacterial retention test / bacterial challenge test (BCT) is generally performed according to the standard method ASTM F838-20 and has to provide proof that the filter can retain bacteria of a certain size. For this, the filter is “incubated” with a defined number of bacteria from a strain of known size under certain conditions. Finally, the filtrate is checked for the absence of bacteria.

The filter integrity test can show critical damage to the filter that can have already had a negative impact on the bacterial retention capability, but hasn’t yet lead to visible damage. Ideally it shows that the filter was integer at the time of the test, meaning that the given pore size was small enough to retain bacteria under the applied process conditions. There are different kinds of integrity tests that can be used, for example the bubble point test, diffusion tests (including the forward flow test and the pressure hold test) and the water intrusion test for hydrophobic membranes. These non-destructive tests are all based on the same principle: The flow of a gas / liquid through a wetted membrane in relation to the applied pressure. The manufacturer specifies a level for the pressure above which the filter produces a sterile filtrate. Correspondingly, reduced pressure levels indicate enlarged pores and thus a compromised integrity.

The influence of the sterilization method on filter integrity is another aspect generally checked by the manufacturer. Proof of filter sterilization capability can be covered by data on repeated sterilization cycles (for example x times autoclaving or x SIP rounds). Or there is information on gamma sterilization.

By “Chemical compatibility” the compatibility of the product composition with the filter material is means. No chemical reaction should take place that influences the integrity of the filter on one side and the product composition on the other side. A compatibility table provided by the filter supplier can be used to check for chemical compatibility. In this table, different chemical substances (for example acids, bases, solvents, common buffers) are contrasted with the filter / filter materials, including data about compatibility. Extractables as extractable components can be a part of the validation documentation by the manufacturer as well.

Extractables are substances that are released by the material under exacerbated conditions (higher temperature, longer time, stronger solvents). Leachables on the other hand migrate from the contact surface into the liquid under normal process conditions. If extractables data by the manufacturer is sufficient, or if it needs to be complemented by product-specific leachables data by the pharmaceutical company has to be decided on an individual basis. In its published document “Guideline on the sterilization of the medicinal product, active substance, excipient and primary container”, the EMA points out that leachables studies are only relevant if the extractables data indicate that toxic components from the filter might leach into the solution to be filtered. Specifying gravimetric extractables (measured as non-volatile residue, NVR) alone doesn’t necessarily comply with the recommendations made by the PDA 26 report.

The studies about release of particles and fibers as well as about conductivity, total organic carbon content (TOC) and oxidable substances if necessary are meant to define a suitable flushing volume in order to avoid exceeding the limit values defined by the pharmacopoeia.

A potential release of particles and fibers by the filter is checked by rinsing the filter with water. Per flushing volume used, the number of detected particles is listed by their size. From that, the necessary flushing volume can be derived. It’s compliant if it conforms to the compendial guidelines for “particulate matter in injections” (USP <787> / <788>, Ph. Eur. 2.9.19 and JP 6.07) of less than 6000 ≥ 10 µm large and less than 600 ≥ 25 µm large particles. The same criteria are classified as “non-fiber releasing”.

The flush tests for TOC and conductivity are used as a recommendation for a suitable flushing volume. The flushing volume determined by the manufacturer is supposed to guarantee a TOC content of less than 500 ppb (= 0,5 mg/L) and a conductivity of less than 1,3 µS/cm at 25°C, complying with compendial requirements (USP, Ph. Eur. and JP) for water for injection.

For the same purpose, the content of oxidable substances can be determined as well. However, this test is outdated. The USP only lists it in the requirements for sterile WFI. For “normal” WFI, it has been replaced by TOC since USP 23 because of its low sensitivity.

The tests for the content of bacterial endotoxins as well as toxicity are for safety aspects. It needs to be ensured that the filter doesn’t have any toxic effects. Neither should it have been contaminated by bacteria and their endotoxins during manufacture, nor should the filter material release substances with possible toxic effects upon contact with liquids.

Bacterial endotoxins are generally determined using the LAL test (Limulus Amebocyte Lysate; USP <85>, Ph. Eur. 2.6.14) and must be lower than the value of max. 0,25 EU/mL approved by the pharmacopoeia (USP, Ph. Eur. and JP) for WFI (water for injection).

Compendial methods can be used to test the toxicity of filter extracts. These include, for example, the „General (Mouse) Safety Test“ (USP <88>, Ph. Eur. 2.6.9). The test is passed when the animals don’t show signs of toxicity and don’t die.