Trish Greenhalgh discusses: “Chromatography for virus clearance”

What we are seeing as we work with customers is a drive to simplify and compress processes as much as possible.   Perhaps five years ago, there was a protein A column, low pH step, cation exchange (CEX), anion exchange and a virus filtration step.   What we see now is customers who don’t include, or don’t want to include the cation exchange step or they may be doing direct capture with the CEX resin or a mixed-mode resin rather than a traditional protein A. There is a definite shift to reduce the number of steps, which basically means that the expectations for the virus validation are that you have to have to demonstrate greater clearance across each unit operation from a reduced number of steps than you might have had to before. Instead of validating maybe four steps or five steps in a process, some people are going into a validation study in the hope of getting enough clearance across as few as three steps with the expectation that the chromatographic media are going to deliver greater log reduction at each step. This is particularly challenging for parvoviruses or non-enveloped viruses especially which tend to have less predictable clearance across protein A or CEX.  In Europe, the regulatory guidelines indicate that manufacturing processes for biologics should include two orthogonal steps for virus removal, each capable of reducing virus by at least 4 logs  which is the measure of ‘effective removal’.  It’s not as explicitly stated in the U.S. guidelines, but it’s assumed that there will be more than one effective virus removal step. The virus filter if generally considered an effective step for  parvovirus removal, but if you reduce the number of chromatography steps it could be challenging to robustly deliver at least four logs of parvovirus clearance across one other chromatography step.


There is also a real need to understand the mechanism of action of virus removal for the different purification unit operations. Small differences in the molecule characteristics, buffers, impurity levels, and the preceding purification steps can all have a profound effect on the performance of a subsequent chromatography step and also the levels of virus reduction that can be achieved.   As such, real process understanding is critical particularly as the FDA moves towards  a more holistic approach of processes in terms of  QbD;  there is a far greater expectation on manufacturers today to really understand how small changes in running or manufacturing conditions impacts both the product quality and the safety testing in terms of virus removal.



1 Comment for this entry

November 26th, 2013 on 6:44 am

Interesting comment.
The mentionned reduced steps approach, where the protein A fraction (containing the mAb) would be loaded on an AEX media (without CEX step) is certainly a challenge when looking at the robustness of the process in terms of viral clearance.
Due to an higher content of contaminants (versus a pool that would have gone through a CEX step), we can fear some impact on the capability of the AEX media to reproducibly retain the viruses, and thus ensure that the needed LRVs are obtained.
On that aspect, packed columns AEX media offer more binding capacity, and could provide better performance and robustness, when compared to membrane AEX devices.
For sure a topic on which we will see studies in the coming years !
Best regards to all,

Like or Dislike: Thumb up 1 Thumb down 0