ADVANCING THE ROBUST MANUFACTURE OF T-CELL THERAPIES THROUGH THE

APPLICATION OF STIRRED TANK BIOREACTORS

Alex Klarer1,Matthew Marsh1, Shaun Ranade1, David Smith1

1Hitachi Chemical Advanced Therapeutics Solutions, Allendale, NJ

Introduction

With the rise of T-Cell immunotherapies cell therapy developers have

encountered the difficulty of producing consistent, quality products.

Approved therapies Kymria™ and Yescarta™ as well as late phase

clinical products BB2121 and UCART19 have been critiqued for

potential obstacles to patient access due to high sticker prices and

variable manufacturing outcomes. Manufacturing processes need to

be developed to allow for increased control over each unit step to

improve the uniformity of the cell journey over the life of the product.

Bioreactor culture systems provide a pathway for a higher level of

process control over cell expansion.

With the high cost of developing and manufacturing cell therapies,

manufacturing techniques that allow for rapid translation to the clinic

and reduce operating costs need to be investigated. Stirred-tank

bioreactors have been heavily adopted by the biologics industry and

have well defined characteristics that facilitate the scale-up from small

scale development work to full scale clinical and commercial

manufacture. Established modelling of the fluid dynamics present

within these bioreactors may also reduce the reagent costs of cell

culture with more efficient mass transfer.

Conclusion

T-Cell immunotherapies have proven efficacy as viable medications for

previously untreatable diseases. However, manufacturing options for these

transformative therapies threaten patient access due to inconsistent

manufacturing, long processing times, and high cost. Stirred-tank bioreactors

present an option for the expansion of primary human T-Cells that has

previously not been pursued due to concerns about shear stress. Small scale

studies have shown that impeller driven mixing in stirred-tank bioreactors does

not increase cell death or limit expansion potential. Translating the learnings

from those small scale studies to the autologous manufacturing scale has

shown that the BIOne stirred-tank bioreactor produces a consistent culture

environment and high cell yields. The BIOne significantly outperformed rocking

bioreactor systems in both total yield and production efficiency. These

improvements could impact patient access by reducing the manufacturing

timeline and reducing the cost from both reagents and labor. The high yield of

the BIOne was only possible by integrating it with the Lovo spinning membrane

system to allow for media exchange; nutrient and metabolite analysis

throughout the culture reveals that increasing the exchange rates would further

prevent lactate accumulation and potentially further increase yields.

0.0

5.0

0 1 2 3 4 5 6 7 8 9 10 11 12 13

mg/

dL

Day

Lactate Concentration Over Time

ambr - Low Perfusion

ambr - High Perfusion

Batch Fed

Static

Figure 2. The change in key metabolite

concentrations over the course of the thirteen

day culture (N=3).

0.0

2.0

4.0

6.0

0 1 2 3 4 5 6 7 8 9 10 11 12 13

mg/

dL

Day

Glucose Concentration Over Timeambr - Low Perfusionambr - High PerfusionBatch FedStatic

Establishing Key Parameters of Stirred-Tank Bioreactor Cultures

Autologous Scale Expansion of Primary T-Cells

0.0E+00

2.0E+07

4.0E+07

6.0E+07

8.0E+07

1.0E+08

0 2 4 6 8 10 12

Cel

ls

Day

ambr - Low Perfusion

ambr - High Perfusion

Batch Fed

Static

0.0E+00

2.0E+07

4.0E+07

6.0E+07

8.0E+07

1.0E+08

0 2 4 6 8 10 12

Cel

ls

Day

ambr - Low Perfusion

ambr - High Perfusion

Batch Fed

Static

Figure 1. Growth curves from two donors are compared independently to

their control cultures (N=3). The largest sample to sample increase in

growth occurs between day five and seven. Bioreactor cultures performed

markedly better in two of the three donors, but more development is

necessary to improve consistency across various donor types.

The results of the small

scale study showed that the

shear stress of a stirred

tank bioreactor does not

inhibit cell proliferation and

may provide additional

efficiencies in mass transfer.

The experimental conditions

outperformed the static

controls and were able to

achieve cell densities

comparable to continuously

perfused systems using a

rudimentary mock batch-fed

process. The results

indicate that high

concentrations of lactate

prevented further expansion

of the cultures.

References

➢ Marenghi, Angela, et al. “Perfusion’s Role in Maintenance of High-Density T-Cell Cultures.”BioProcess International, 13 Jan. 2015, www.bioprocessintl.com.

➢ Nienow, Alvin W., et al. “The Physical Characterisation of a Microscale Parallel Bioreactor Platform with an Industrial CHO Cell Line Expressing an IgG4.” Biochemical Engineering Journal, vol. 76, 2013, pp. 25–36., doi:10.1016/j.bej.2013.04.011..

➢ Rameez, Shahid, et al. “High-Throughput Miniaturized Bioreactors for Cell Culture Process Development: Reproducibility, Scalability, and Control.” Biotechnology Progress, vol. 30, no. 3, May 2014, pp. 718–727.,

doi:10.1002/btpr.1874.

➢ Molleryd, Carin, et al. “Scaling up Clinical T Cell Expansion in a Xuriâ„¢ Cell Expansion System.” Cytotherapy, vol. 17, no. 6, 2015, doi:10.1016/j.jcyt.2015.03.379.

➢ Li, F., et al. “A Systematic Approach for Scale-Down Model Development and Characterization of Commercial Cell Culture Processes.” Biotechnology Progress, vol. 22, no. 3, Feb. 2006, pp. 696–703., doi:10.1021/bp0504041.

Medium Exchange Using Spinning Membrane Filtration

Figure 7. Nutrient and metabolite analysis shows that media exchange

rates were not sufficient to prevent lactate accumulation and cell death

late in the culture. Lactate concentrations and LDH, a measure of cell

lysis, have a linear coefficient of determination of 0.72.

y = 2000e-0.236x

0

500

1000

1500

2000

2500

0 5 10 15 20 25 30Vo

lum

e o

f Sp

ent

Med

ia [

mL]

Hour

Figure 6. Assuming the system functions as a true continuous stirred

tank reactor, spent media removal is modelled as an exponential decay

function to determine the true media exchange rate at a set perfusion

rate.

Configuration

T-Cell Isolation and Activation: Lymphocytes were isolated from apheresis collections by elutriation and incubated

with CD3/CD28 Dynabeads™ at room temperature with light agitation. The cell-bead conjugates were incubated at

37°C and 5% CO2 in static Vuelife culture bags for 72 hours to fully activate. The cells were partitioned into the

various culture conditions after the activation period.

Culture Media: TexMACS media (Miltenyi) was supplemented with 5% human AB serum, 0.1% Pluronic F-68

(Thermo-Fisher), 166 ppm Antifoam-C (Sigma Aldrich), and 30 U/mL IL-2 (R&D Systems). Media exchange, when

implemented, was set to 25% for cell concentrations >1.5x106 cells/mL, 50% for cell concentrations >2.0x106

cells/mL, 75% for cell concentrations >3.0x106 cells/mL, and 100% for cell concentrations >4.0x106 cells/mL.

Culture Devices: The Distek BIOne bioreactor was used as the stirred-tank culture vessel. Media exchange was

performed on the BIOne through integration with the Lovo (Fresenius Kabi) to remove media through the spinning

membrane and replacing removed media directly into bioreactor. Rocking motion cultures used the Xuri Cell

Expansion System (GE Healthcare) and exchanged media through the internal media filter. Cultures were controlled

to 50% dissolved oxygen and a pH of 7.15 with N2, CO2, and AIR via sparging in the BIOone and overlay in the Xuri.

Figure 3. Activated T-Cells expanded in Distek BIOne

bioreactors showed the capacity for significant expansion

across four experiments and two donors. In all cases the

cultures yielded greater than 15 x 109 cells prior to eight

days of total culture time.* The four experiments show

consistent growth profiles within donors which indicates that

the BIOne is able to create a culture environment that is

highly reproducible lot to lot.

*BIOne Run 2 was terminated after day five due to a

malfunction of the perfusion system.

-1.00

0.00

1.00

2.00

3.00

4.00

5.00

6.00

0 1 2 3 4 5 6 7 8 9

Pop

ula

tio

n D

ou

blin

gs

Days in Culture

BIOne Run 1

BIOne Run 2

BIOne Run 3

BIOne Run 4

-1.00

0.00

1.00

2.00

3.00

4.00

5.00

6.00

0 1 2 3 4 5 6 7 8 9

Pop

ula

tio

n D

ou

blin

gs

Days in Culture

BIOne Run 3

BIOne Run 4

Rocking Bioreactor

0.0E+00

5.0E+05

1.0E+06

1.5E+06

2.0E+06

2.5E+06

3.0E+06

3.5E+06

0 1 2 3 4 5 6 7 8

Cel

ls p

er M

illili

ter

of

Med

ia C

on

sum

ed

Days in Culture

BIOne

Rocking Bioreactor

0

0.5

1

1.5

2

2.5

3

0 2 4 6 8

Lact

ate

Co

nce

ntr

atio

n [

g/L]

Days in Culture

BIOne Run 3

BIOne Run 4

Rocking Bioreactor

0

20

40

60

80

100

120

140

160

180

200

0 2 4 6 8

LDH

Co

nen

trat

ion

[U

/L]

Days in Culture

BIOne Run 3

BIOne Run 4

Rocking Bioreactor

Figure 4. Comparing a single donor across the

BIOne stirred tank system and the rocking

bioreactor system indicates that the BIOne enables

increased cell expansion toward the later stage of

the cell culture. The average increase in expansion

was 1.22 population doublings. This difference in

expansion shown in this experiment could

correspond to a 24-48 hour reduction in total

culture length depending on the target yield.

Culture is typically the limiting factor in production

speed and reducing the length of time required to

achieve a therapeutic yield will result in faster

product release.

Figure 5. In addition to higher total yields, the

BIOne was shown to expand cells more efficiently

than the rocking bioreactor. 2.99 x 106 cells were

produced per milliliter of media consumed in the

stirred-tank bioreactor which is a 58% increase

over the 1.88 x 106 cells produced per milliliter in

the rocking bioreactor. Accounting for the high cost

of media components in standard T-Cell media,

more efficient cell expansion would help reduce the

cost of the resulting therapy.