10th— 11th - February 2015 – Radisson Blu, Heathrow Airport, Hayes, London, UK Annual BIOINNOVATION | Seamless, Integrated & Innovative BioManufacturing http://www.bioinnovation.gb.com/ Annual PIMS | 2015 |Pharmaceutical Innovation, Operational Excellence & Manufacturing Strategies Summit http://www.pims.gb.com/ 16th & 17th September 2015 | Boston Marriott Cambridge, MA | USA Annual ManuBIO | 2015 Flexible, Integrated & Innovative Biological Manufacturing Excellence USA http://www.manubio.com/ VIDEO WATCH

1. CASE STUDY
EVALUATING
PERFUSION
TECHNOLOGIES
To evaluate the potential benefits and risks associated with
moving to Perfusion fermentation technologies.
• A risk-based model was built to evaluate the benefits of fermentation
perfusion and how this would integrate with an existing plant
• Risk-based approach used backup perfusion vessels to minimize the
effect of disruption to downstream purification processing
• Excess capacity from the production train was used to create a small-scale
‘plant-within-a-plant’ for clinical and small-scale production
• The completed design showed a positive NPV of $40MM over 5 years.
GOAL
Perfusion technologies have shown potential to radically improve cell densities
in fermentation over batch-based technologies, with titers an order of magnitude
higher than those achieved in existing facilities. Such high titers enable much
smaller fermentation vessels to be used, minimizing plant clean-space requirements
and maximizing often limited downstream purification train resources.
One of the big issues with fermentation perfusion remains the radically different
operating characteristics of perfusion technologies. Since the transient time
for perfusion is months longer than a batch-based system, contamination risk
increases considerably. Bioproduction Group’s brief was to create a feasible
perfusion platform that would fit into an existing plant. The goal was to enable
the manufacturer to use both perfusion technologies for new products as well as
the traditional batch-based process for legacy products.
Realizing that contamination risk would be critical to the evaluation of this
technology, Bioproduction Group collected historical contamination data
based on transient time in existing batch-based systems to estimate the
‘per-day probability of failure’. Monte Carlo techniques were then used in
conjunction with Bio-G’s facility simulator to estimate when a perfusion vessel
would become contaminated, and what the effect would be on production.
PROBABILITY OF FAILURE
0%
15%
CHANCE OF
FAILURE AT
60 DAYS
45%
CHANCE OF
FAILURE AT
150 DAYS
75%
CHANCE OF
FAILURE AT
1 YEAR
0 100 200 300 400 500 600 700
CAMPAIGN DAY
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
high level
summary
THE BRIEF
How We Did It
PROBABILITY OF FAILURE VERSUS
PERFUSION CAMPAIGN LENGTH
© Bioproduction Group. All Rights Reserved. 1

2. “Making perfusion work is all about mitigating the effect of contamination”,
comments Principal Rick Johnston. “Making sure that backup perfusion
vessels, innoculum at the correct scale, and even media is available at
all times is critical to success.” After extensive consultation with the
manufacturer, redundant capacity was added to the perfusion train to
allow operations to continue even if a vessel was contaminated.
Bioproduction Group then produced a comprehensive simulation model
of fermentation perfusion technologies to show how they could be used
in ‘semi-batch’ mode with existing (batch-based) purification equipment.
Unused scale-up tanks were re-tasked to produce clinical and small-scale
production, allowing the plant to produce multiple products at very
different scales at the same time.
“Making perfusion work is all
about mitigating the effect of
contamination”
Rick Johnston, Principal
Results Bio-G’s recommendations were to segregate production into a small-scale
batch-based production train and a large-scale perfusion fermentation train,
retaining the ability to operate the plant in large-scale batch mode. The
resulting design increased the plant’s capacity to simultaneously operate
as a launch platform for new products, as well as perform technology
transfer and optimally utilize workers in the plant.
More importantly, the output of the simulations showed that this change
would allow the company to produce significantly more material without
large-scale engineering changes. The NPV of this scenario was $125MM
over 4 years.
production quantity
and contamination risk vs.
campaign length
MORE INFORMATION
BIOPRODUCTION GROUP
CONTACT@BIO-G.COM
www.bio-g.com
100%
75%
PROBABILITY (%)
30 DAY
PERFUSION
CAMPAIGN
THROUGHPUT (KGS)
50%
25%
0%
45 DAY
PERFUSION
CAMPAIGN
60 DAY
PERFUSION
CAMPAIGN
75 DAY
PERFUSION
CAMPAIGN
© Bioproduction Group. All Rights Reserved. 2