Modified bacterial challenge tests - what to do in case of bactericidal product solution?

Written by Dr. Janet Thode Posted in Filter validation

For pharmaceutical companies that manufacture their products under aseptic conditions, sterile filtration is the last step before filling. The validation of the manufacturing process also includes the step of sterile filtration. Accordingly, it must be demonstrated by filter validation that the filter is able to retain bacteria and thus ensure the sterility of the product. This evidence is provided by a bacterial challenge test according to ASTM F838-20. But, before this bacterial challenge test can be performed, it must be checked by a viability test that the product solution has no negative effect on the bacterium used in the bacterial challenge test. That sounds complicated? It shouldn’t - just imagine what would happen if your product solution is killing the test bacteria...

You wouldn’t know if you get a sterile filtrate in the bacterial challenge test due to the fact that the filter did a good job or due to the fact that your product solution already killed the bacteria.

 

The viability study

So let's get back to the viability study. In this study, the bacterium Brevundimonas diminuta which is used for the bacterial challenge test, is incubated with the product solution for at least the duration of the real filtration process in routine production and under the same temperature conditions. If your product solution is not bactericidal, it will make the bacteria look gorgeous and the filter’s subsequent bacterial challenge test can be designed allowing direct inoculation of the bacteria into the product solution.

But what if the viability test determines that the bacteria are not feeling well? This is our topic for today. There are two scenarios imaginable:

  1. The bacteria are really not quite happy, but they are somehow still alive and bobbing up and down. More scientifically, this is the case when there is a reduction in the number of bacteria, but not more than 1 log.
  2. They are completely dead.

In both cases, the bacterial challenge test can’t be performed as described in ASTM F838-20 with direct inoculation of the bacteria into the product solution as this would lead to false negative results as explained above. Therefore, some alternative test designs for the performance of the bacterial challenge test are described in the PDA technical report No. 26.

 

Alternative test designs of the bacterial challenge test

In the case of 1. “impaired viability”, it is possible to find out how long the bacteria are viable. When this is known, the bacterial challenge test can be set up by a previous "preconditioning" of the filter without bacteria followed by addition of the bacteria for the reduced test time. Let's say the time of filtration in production would be 72 h, but the bacteria are only viable for 14 h in the product solution. Then you would precondition the filter for at least 58 h with the product solution and then directly inoculate the bacteria for the remaining 14 h, resulting in viable bacteria and covering the total process time (58 h + 14 h = 72 h).

In the case of 2. “total antimicrobial effect”, it may be due to the product solution itself or the process conditions applied. Both should be investigated.

In case of bactericidal process conditions, we could imagine a temperature too high, for example. For which reason ever, it might be necessary to perform the sterile filtration at 4°C. Brevundimonas diminuta, which shows optimal growth at 28-30°C, doesn’t grow at all at 4°C (Bergey's Manual of Systematic Bacteriology). In such a case, the filter is preconditioned with product solution applying the real process conditions (in this case 4°C) over the entire process duration, followed by a direct incubation of the bacteria in the product solution under modified test method parameters (in this case 30° C). Of course, the other process conditions do not need to be changed.

If the bactericidal effect is due to the product itself, it is important to identify the bactericidal component. This might be the pH, for example. So let’s imagine an acidic product with a pH of 4, while B. diminuta feels most comfortable in the neutral range (i.e. around pH 7). Thus, to perform the bacterial challenge test, a product with a pH modified to 7 should be used, if possible. This would allow direct inoculation and performance under unchanged process conditions.

If a component of the product composition is the reason for the bactericidal effect, this ingredient should be removed and a surrogate might be used, which should, of course, be as similar as possible to the product solution despite the bactericidal component. With such a modified product composition, the bacterial challenge test is feasible with direct bacterial inoculation and application of the process conditions. Furthermore - if the bactericidal effect only affects B. diminuta – it is allowed to use another test germ instead, such as e.g. a very small isolate of indigenous bioburden.

 

In summary, there are the following possibilities of alternative test designs:

  • Reduced test time in case of impaired / limited viability
  • Change of process conditions
  • Change in product composition
  • Use of another challenge bacterium