Biomedical Engineering Reference
In-Depth Information
development environment, which never proceeded to be implemented in the manufactur-
ing process. Fahrner et al. [33] demonstrated online reverse-phase HPLC for separating
aggregates from recombinant human insulin-like growth factor-I (IGF). Rathore et al.
[43] have shownmany instances inwhich separations were evaluatedwith respect to their
ability to control processes. First, for controlling a protein refolding step, they looked into
the feasibility of designing a control strategy using online monitoring that would allow
refolding operations to end at a time determined by product quality parameters. The data
suggested that while possible it might not be practical due to the “complexity” issues in a
manufacturing environment. Rathore states concerns for using online methods versus the
simplicity of a time-based approach. Concerns such as training operators to make
decision with an online analytical method as well as redundancy for a fail-safe
mechanism would be obstacles in implementing any such technology. In the same set
of studies, they looked at controlling an end point analysis to the UF/DF step based on a
product quality measure. In the final study of the set, they evaluated a control strategy for
ion exchange fraction pooling based on desired product purity. They analyzed the
samples by reverse-phase HPLC for product purity and demonstrated the ability to
control product purity through such a method, while stating similar rationale for not
implementing the strategy in a manufacturing environment.
In a separate body of work, Rathore et al. [11] demonstrated implementation of
online size-exclusion HPLC to control fraction purity pooling from a process step that
used a hydroxyapatite column separation. The systemwas set up to collect sample from a
side stream.
They were able to show the feasibility of implementing the online analysis for
facilitating a real-time decision making for pooling of chromatography column based on
product quality attributes. Experiments were performed with low purity (65.5%),
moderate purity (71.3%), and high purity (76.1%) and feed material. Column eluate
fraction pooling was performed using preset criteria to halt the pooling process when the
fraction purity reached 85%. They were able to achieve a pool purity within 1%variation
despite a 10% variability in the purity of the feed material. A processing paradigm that
Rathore points out is that by targeting consistent pool purity by shifting pooling criteria,
the recovery across the chromatographic step varies. Rathore also points out the fact that
regulatory authorities have historically viewed step yield as a measure of purification
process consistency. If a company were to implement such a tool, a regulatory strategy
would be needed in a filing, of any such process, to address this potential regulatory
concern.
Lanan and McCue [38] demonstrated that online reverse-phase HPLC was useful to
purify low molecular weight impurities from the product of interest during the course of
an ion exchange step. Off-line studies showed that low molecular weight impurities
eluted in early fraction of the ion exchange, so the goal was to monitor the lowmolecular
weight impurity stream to determine when to begin the collection of the protein fraction
of interest. In the online system at pilot scale, a stream splitter was connected to the outlet
tubing of the ion exchange column. The tubing from the splitter was connected to a
rheodyne valve with a fixed volume injection loop. The loop was switched in-line with
the reverse-phase column at predetermined periodic intervals to perform the reverse-
phase separation to separate the low molecular weight impurities. They were able to
