COMPARABILITY FILTERABILITY TRIALS: THE UTILIZATION OF APPROPRIATE, EFFECTIVE FILTRATION AREAS

BioPharm E-newsletter Case Study #2

COMPARABILITY FILTERABILITY TRIALS: THE UTILIZATION OF APPROPRIATE, EFFECTIVE FILTRATION AREAS

BACKGROUND

Filterability trials with 47 mm disc composites are not unusual. They are routinely used to evaluate optimal filter combinations, polymer, and required retentivity. However these trials, performed at such small-scale filtration areas and not in a pleated configuration, can only be used as “indicator trials” and not for scaling or sizing purposes. Once the optimal combination has been indicated, pleated device trials must be performed to verify initial results and to size the filtration system appropriately (verification and assurance trials).
Most of the time 47 mm disc composites are used for such filterability trials. Composites should be actual 47 mm cut outs of the filter cartridge used in the final process otherwise comparability is not achieved. These composites are either installed in a 47 mm stainless steel holder or found welded into a plastic housing as a disposable device. However, if a combination of either is utilized during the trials, the effective filtration area must be considered when comparing the performance of these devices.
This paper discusses the effect of the differences among devices and explains the necessary action required to achieve comparability of filterability test trials with 47 mm discs.

FINDINGS

The effective filtration area of common 47 mm discs within a stainless steel holder and disposable device were measured by coloration of the effectivefiltration area and measurement of the diameter of the coloration.

The effective filtration area of the disposable device is 36% larger than the 47 mm disc utilized in a stainless steel holder. The main reason for this discrepancy is that the stainless steel device requires an O-ring sealing. The O-ring diameter and seal will reduce the effective filtration area. The disposable device utilizes a heat welding, which is a very thin ring aroundthe membrane.

CONCLUSION

47 mm disc filterability trials require thorough observation of the effective filtration area of small-scale devices, especially when results of different devices are compared. The data provided by manufacturers’ literature may not be tightly specified, and they may result in the use of wrong correction factors and test data. The example in Figure 1 demonstrates what could happenwhen devices with different filtration areas are wrongfully compared.

PES* is a disposable device used to measure total throughput, a typical test for filterability trials. The device shows far higher throughput performance than the other filter discs of other manufacturers. Yet, the other filter performances were tested with stainless steel holders, i.e. a 36 % lower effective filtration area (EFA).
Once a correction of the measured value has been performed by taking the higher filtration area into account, the performance is comparable to another PES. The first measured PES filter now has a comparable performance of PES#.
If the effective filtration area had not been considered, none of the other filters would have been chosen for additional filterability studies with pleated devices (verification and assurance trials). This would have eliminated the chance of a second vendor or supplementary tests like unspecific adsorption or mechanical stability. The end user would have relied on only one filter, which limits required flexibility and choice.

Summary
Filterability trials are an essential tool to evaluated total throughput and only total throughput of single filters or pre-/final filter combinations. However, such filterability trials should be comparable and cannot be solely relied upon when filter sizing is required. Filterability trials with 47 mm discs are indicator trials and require examination of the effective filtration area of the devices used. A pleated device trial (verification trial) should always be performed after such initial investigation; the filtration areamust be comparable in these tests.

Maik Jornitz, Sartorius