Purpose: In general, monoclonal antibodies may be prone to self-association leading to formation of dimers, trimers or high molecular weight species (HMWs) during bio-processing. This self-association is more pronounced under aqueous conditions and could be exacerbated under stress conditions as typically encountered during bio-processing. In order to implement appropriate manufacturing and control strategies during bio-processing, it is important to understand various real life bio-processing conditions where such self-associations may manifest. One such case study is presented here of increase in dimer content for a monoclonal antibody during commercial scale bio-processing and the approach taken to understand the under-lying mechanism. In this example, a therapeutic monoclonal antibody (mAb) being commercialized demonstrated a consistently higher dimer values (~0.5% higher) in the drug product during release when compared to the same value measured in the corresponding drug substance lot. This observation was interesting since the DS was supplied frozen, and the DS and DP share the same formulation composition. Therefore, the purpose of this study was to systematically investigate each step of the drug product manufacturing process to characterize and control the increase in dimer formation. As a control, dimerization kinetics of monoclonal antibody was monitored at 2-8°C and room temperature under static conditions.
Methods: Size exclusion chromatography (SEC) was used to quantitate %dimer in purified DS and in aqueous DP. In order to characterize 0.5% increase in dimer in final drug product compared to drug substance, in-process samples after each fill-finish unit operation were subjected to SEC analysis to test for %monomer, %dimer and HMWs. For comparison, sample immediately after thawing of bulk drug substance was aliquoted and subjected to static incubation at 2-8°C and at room temperature. Another set of aliquots was stored frozen at -80°C to be used as control.
Results: The average %dimer in purified bulk DS was quantitated around 1.3% which was stored frozen until further bio-processing into final DP. For drug product manufacturing, the frozen bulk drug substance was thawed at room temperature and subjected to fill-finish manufacturing process steps which included compounding, filtration and filling into final container closure. After each unit operation, the in-process samples tested for %dimer showed a gradual increase in dimer as a function of time over a period of 13 days accumulating to 1.8%dimer at the end of DP manufacturing process. As a control, sample immediately after thaw of bulk drug substance was subjected to static incubation at 2-8°C and room temperature. Both samples showed a gradual increase in dimer over a period of 13 days, however, the rate of increase in dimer at room temperature was higher compared to sample stored at 2-8°C. At the end of 13 days, %dimer in sample at room temperature had an average dimer of 1.8% identical to the sample that underwent the fill-finish manufacturing process. In comparison, %dimer in sample stored at 2-8°C was quantitated at around 1.5%. The results from this study demonstrate that self-association of this particular antibody can be explained by temperature dependent reaction kinetics of monomer to form dimer over a period of time.
As an outcome of this investigation, we propose that bio-processing of drug product at 2-8°C could lower the self-association rate of monoclonal antibody and therefore 2-8°C processing conditions could be applied for the entire drug product manufacturing train.
Srishty Subramanian– Teva Pharmaceutical Industries, Inc.
Rowena DMello– Teva Pharmaceutical Industries, Inc.
Charlene Brisbane– Teva Pharmaceutical Industries, Inc.
Shouvik Roy– Teva Pharmaceutical Industries, Inc.