Extreme Scale-Down Approaches for Rapid Chromatography Column Design and Scale-Up During Bioprocess Development - Measurement, Monitoring, Modelling, and Control of Bioprocesses - page 110

3.1 General Considerations............................................................................................. 113

3.2 Empirical Scale-Up Example ................................................................................... 114

3.3 Mechanistic Scale-Up Example ............................................................................... 115

3.4 Other Issues............................................................................................................... 116

4 Microlitre Scale-Down ...................................................................................................... 116

4.1 Overview ................................................................................................................... 116

4.2 General Considerations............................................................................................. 117

4.3 Microlitre Batch Incubation ..................................................................................... 118

4.4 Chromatography Pipette Tips................................................................................... 121

4.5 Miniaturised Packed Columns.................................................................................. 124

4.6 Microfluidic Chromatography .................................................................................. 127

5 Related Issues and Challenges .......................................................................................... 128

5.1 Analytics ................................................................................................................... 128

5.2 Experimental Design Methods ................................................................................. 129

6 Conclusions........................................................................................................................ 133

References................................................................................................................................ 133

1 Introduction

Designing a chromatographic separation for a protein product is one of the most
important steps in biopharmaceutical development for achieving satisfactory levels
of purification [ 22 ], and achieving this capability may require up to a few dozen
column runs during development studies [ 7 ]. Initial studies may employ small col-
umns which are both significantly shorter and narrower than at final manufacturing
scale; For example, pre-packed columns of a few millilitres in size may be used to
determine whether a putative separation space exists and, if so, to identify a smaller
subset of conditions to explore in larger laboratory columns of up to around 20 cm
length (40-50 mL volume). At this stage, the bed height is fixed along with the linear
operating velocity and the feed load expected for larger columns, and the bed
diameter is changed whenever required to accommodate the larger feed challenges
that are representative of later development. Cumulatively, the feed material needed
to supply such work may be considerable (tens to hundreds of millilitres) and may
become a limiting factor if many resins need to be evaluated before finding one that
offers the required capacity and selectivity. Additionally, development timelines are
often compressed, and this prevents lengthy experimentation across a wide param-
eter space using resource-intensive laboratory columns. Such studies are also limited
by the availability of the expensive laboratory equipment required to operate these
columns [ 4 , 34 ]. Although some degree of automated control is possible with these
systems, e.g. using predefined method protocols, there can still be many occasions
where manual intervention is required; e.g. if pre-packed columns of the required
matrix type, volumes or dimensions are unavailable, then column set-up, packing and
HETP/asymmetry testing are needed. As a result of these factors, sample throughput
at laboratory scale tends to be fairly low, and since columns tend to be operated in
series, only one combination of operating conditions can be tested at a time, making
the sequential examination of many parameters time-consuming [ 27 , 31 ].