Final Individual Design Report: Column Chromatography ...
Final Individual Design Report: Column Chromatography Sequence ChemE 486 Spring 2013 Kimberlee Sing
Table of Contents 1. Executive Summary 2. Introduction 3. Results 3.1. Column Configuration Specifications- Table 3.2. Flow diagram of Column Configurations 3.2.1. Base Case Flow Diagram- Figure 3.2.2. Continuous Column Rotation Logistic Diagram - Figure 3.2.3. Optimized Process Flow Diagram- Figure 3.3. Percentage of Antibody Lost - Figure 3.4. Remaining CHOP (%)- Figure 3.5. Remaining Antibody Aggregate (%)- Figure
4. Discussion 4.1. Optimization 4.2. Continuous Configuration 4.3. Major Challenges
5. Cost Analysis 5.1. Operating Cost per Year- Figure 5.2. Capital Cost- Figure 5.3. Total Cost Over 30 year Plant Lifetime- Figure 5.4. Overall Cost Analysis of Optimum Configuration Process- Table
6. Appendices 6.1. Optimum Column Configuration Specifications- Table
6.2. Optimum Initial and Final Conditions for the Protein A, IEX and HIC Columns and Viral Inactivation- Table 6.3. Optimum Configuration Summary of Initial and Final Conditions per Batch- Table 6.4. Buffer Cost Analysis of Optimum Configuration Process- Table 6.5. Viral Inactivation Specifications and pH of solutions- Table
7. References
1. Executive Summary In partnership with CellDex, Coho BioSciences will bring an antibody-conjugate cancer drug, glembatumumab vedotin CDX-011, to the worldwide market. To prepare for the manufacturing of the product, a preliminary facility design was requested by the Coho BioSciences Board of Directors to provide a better understanding of the economics and technicality of the production. It is very important that the drug meets delivery and purity specifications to meet FDA requirements and approval. Thus, it is essential to have a rigorous purification process. Initial purification of the antibody through a centrifuge and depth filters provides a rough primary recovery. The harvest fluid is then sent to a column chromatography sequence for more thorough purification. The column chromatography sequence is comprised of a protein A, ion exchange (IEX) and hydrophobic interaction (HIC) column, with a viral inactivation step between the protein A and IEX column. Also, both the operating costs including resin costs of $2000 per liter, and capital costs of the columns, ranging from $25,000 to $1.6 million for 10 to 100 cm diameter columns, respectively, contribute to a large portion of the overall plant costs. [7.1] Thus, it is advantageous to reduce the overall cost of this process, while maintaining delivery and purity specifications. At the request of the Coho Board of Directors, the following document discusses and analyzes the investigation of the best configuration of the column chromatography sequence. The system will be optimized to maximize the capture of antibody, minimize the CHOP and aggregate antibody captured, and minimize the overall cost and time to run the process.
2. Introduction The column chromatography sequence is comprised of three different chromatography columns, protein A, ion exchange (IEX) and hydrophobic interaction (HIC) columns, and a viral inactivation step to purify the harvest fluid after the primary recovery process. This process provides the ultimate purification step of the antibody before it is conjugated to form the product, CDX-011. Each column has a respective resin to capture the antibody, and the linear flow rate of the entering fluid determines the antibody binding capacity, percent antibody yield, percent reduction of CHOP, and percent reduction of aggregate. Resin is an expensive operating cost of $2000 per liter, so it was necessary to determine the best configuration of chromatography columns to reduce the operating and capital costs, maximize the capture of antibody, and reduce the amount of CHOP and undesired aggregate. The amount of resin is determined by the size and number of columns used. The larger the column, the greater the capital costs. However, this may not necessarily mean greater operating costs depending on the column configuration. In addition to the resin, there are 5 different buffers used to sanitize/clean, equilibrate, wash, elute and regenerate the resin for each batch. Each buffer has a respective amount necessary to prepare the column for the batch and thus determined a constraint for the time to run the batch. Investigation of different configurations of columns in series and in parallel provided some solutions to the time constraint. As mentioned earlier, there is also a viral inactivation step in the sequence between the protein A and IEX column. The eluate from the protein A column must be held at a lowered pH at the corresponding holding time to inactivate the virus and raised to 6.0 where the antibody is stable. Then it is further purified in the IEX and HIC columns. The scope of this document focuses on just the chromatography sequence so it was assumed that the optimum pH was 3.6 and held for a minimum of 8 hours. A base case of the chromatography sequence was completed earlier and a discussion and cost analysis to compare it with the optimized case will be presented. The following document will also discuss the investigation of implementing a continuous chromatography processes.
3. Results 3.1. Column Configuration Specifications Table 3.1 is the column configuration specifications for the base case, continuous, and optimum configuration. It includes the diameter, number of columns in parallel and linear load flow rate of each case. Table 3.1 Column Configuration Specifications Protein A IEX HIC Column Column Column Diameter (cm) Base Case Continuous Optimum Configuration Number of columns In parallel Base Case Continuous Optimum Configuration Linear Load Flow Rate (cm/h) Base Case Continuous Optimum Configuration
100 20 65
100 20 30
100 20 30
1 12 2
1 10 4
1 34 4
100 600 100
100 300 100
100 300 100
3.2. Flow diagram of Column Configurations 3.2.1. Base Case Flow Diagram
In Series 1 3 2
Figure 3.2.1 is the process flow diagram for the base case chromatography sequence. There is only one protein A, one IEX and one HIC column for the column configuration. Coincidentally, it also represents the block flow diagram for all of the chromatography configurations. The viral inactivation step is not drawn but is between the Protein A and IEX columns. 3.2.2. Continuous Column Rotation Logistic Diagram
Figure 3.2.2 is a continuous column rotation diagram for one of the column chromatography cycles. Each column goes through a load, wash, elute, regenerate, sanitize and equilibrate step to complete the column cycle. For example, this protein A column cycle has 9 columns that go through the cycle simultaneously. Three columns in series are loaded in series from the UF/DF in columns Load 1, 2 and 3 until Load 1 column reaches binding capacity. Load 1 column is then washed and follows the rest of the cycle in sequential order. Load 2 was not filled to binding capacity and becomes a Load 1 column until it reaches binding capacity. Each subsequent column moves through this sequence and rotates through all the columns in a “continuous” cycle.
3.2.3. Optimized Process Flow Diagram Figure 3.2.3 is a process flow diagram of the optimized batch column configuration. Each column has 2 columns in series, with 2 columns running parallel for the Protein A, 4 for the IEX and 4 for the HIC column with a viral inactivation holding tank, not drawn, between the protein A and IEX column.
3.3. Percentage of Antibody Loss between Column configurations
Percent Antibody Lost (%) 16.7 10.3
2.3
Base Case
Continuous
Optimum Configuration
Figure 3.3 shows the percentage of antibody lost from the base case, continuous, and optimum configuration processes. The base case process has the highest percentage antibody loss of 16.7%, next is the continuous with 10.3% and the lowest is the optimum case with 2.3% loss. 3.4. Remaining CHOP (%)
Remaining CHOP (%) Base Case
Continuous
Optimum Configuration
0.25
6.4E-07
3.125E-08
Remaining CHOP (%)
Figure 3.4 is the percentage of CHOP remaining from the base case, continuous, and optimum configuration processes. The base case process has the highest percentage antibody loss of 0.25%, next is the continuous with 6.4E-07% and the lowest is the optimum case with 3.1E-08% loss.
3.5. Remaining Antibody Aggregate (%)
Remaining Antibody Aggregate (%) Base Case
Continuous
Optimum Configuration
0.13
0.0018
0.0002
Remaining Antibody Aggregate (%)
Figure 3.5 is the percentage of antibody aggregate remaining from the base case, continuous, and optimum configuration processes. The base case process has the highest percentage antibody loss of 0.13%, next is the continuous with 1.8E-03% and the lowest is the optimum case with 2E-04% loss.
4. Discussion The best column chromatography process is one that maximizes the capture of antibody, minimizes the CHOP and aggregate antibody captured, and minimizes the overall cost and time to run the process. 4.1. Optimization The optimization of the process was heavily dependent on the configuration of the chromatography columns. The main optimization variables were the number of columns in series and in parallel, and the size of the columns. Increasing the number of columns in series, by assuming that the first column captured 100% of the antibody bound to the resin, but only lost part of it in the second or third column, drastically reduced the antibody lost, as seen in Figure 3.3. However, 2 columns are optimal for batch system, as the 3rd column had negligible contribution and performed better than the continuous process as seen in Figure 3.4 and 3.5. Smaller columns and varying the sizes of columns between protein A, IEX and HIC steps also optimized the process from the base case. Since the protein A resin had a higher binding capacity than the other column resins, larger columns were optimal for this step. 4.2. Continuous Configuration Another way to optimize the system was investigating a continuous chromatography process. Technically, the system is “semi-continuous” since the columns are run staggered and in parallel, rotating through the chromatography cycle, as in figure 3.2.2. By implementing the continuous cycle, the rotation of the column was contingent on the loading cycle time, which generally takes less time than the limiting, wash step. In order to continually rotate through the columns, the number of columns greatly increased to compensate for the difference in time, which also increases capital costs. However,
reducing the size of the columns also helped to reduce the cost and overall processing time. 4.3. Major Challenges The major challenges encountered in the process were determining a logical configuration for the continuous chromatography sequence and an optimal size to number column ratio for the each case.
5. Cost Analysis 5.1. Operating Cost per Year
Operating Cost per Year Base Case
Continuous
Optimum Configuration
$68,444,849 $52,563,090 $32,151,763
Operating Cost per Year
Figure 5.1 shows the operating cost per year of the base case, continuous and optimum column configuration. The base case process has the highest operating cost at $68.4 million per year, continuous is $52.6 million per year and the optimum configuration is the lowest at $32.2 million per year.
5.2. Capital Cost
Capital Cost Base Case
Continuous
Optimum Configuration
$22,400,000 $11,600,000 $4,800,000
Captial Cost
Figure 5.2 shows the capital cost of the base case, continuous and optimum column configuration. The continuous process has the highest operating cost at $22.4 million, the optimum configuration is $11.6 million and the base case is the lowest at $4.8 million. 5.3. Total Cost Over 30 year Plant Lifetime
30 Year Total Cost Base Case
Continuous
Optimum Configuration
$2,058,145,457 $1,599,292,705 $976,152,888
30 year total cost
Figure 5.3 shows the total cost over the 30 year facility lifetime for the base case, continuous and optimum column configuration. The base case process has the highest operating cost at $2.06 billion, continuous is the next highest at $1.6 billion and the optimum configuration is the lowest at $976 million. Even though the batch case had
much lower capital costs than the continuous case, the high operating cost results in a more expensive process over the 30 year lifetime.
5.4. Overall Cost Analysis of Optimum Configuration Process The protein A column is the most expensive step of the process due to the 4 large, 65 cm diameter column, which accounts for the high resin costs per year. Buffer and resin account for two-thirds of the operating cost of the process due to the sheer volume of buffer used for each batch. Captial cost also surprisingly are very similar between different columns even though there are 4 large columns for the Protein A and 8 smaller 30 cm columns for the IEX and HIC columns. Buffer costs are also surprisingly within a million dollars of each other per year which is opposite to the operating costs for resin. Table 5.4 Overall Cost Analysis Protein A Column IEX Column HIC Column $6,356,652 $4,099,916 $5,791,854
Total $16,248,421
$7,963,937 $70,000
$1,979,203 $140,000
$3,110,177 $140,000
$13,053,317 $350,000
Skid Cost Viral filtration
$500,000 $23.95
$1,000,000
$1,000,000
$2,500,000 $24
Total operating cost per year Captial cost
$14,890,589 $3,600,000
$7,219,119 $4,000,000
$10,042,031 $4,000,000
$32,151,739 $11,600,000
Total Buffer Cost per year Cost of Protein A resin per year QA & QC Testing
6. Appendices 6.1. Optimum Column Configuration Specifications Table 6.1 Column Specifications Protein A IEX Column Column 20 25 bed height (cm) 65 30 Column Diameter (cm) 2 4 Number of columns in parallel 2 2 Number columns in series
HIC Column 25 30 4 2
Linear Flow rate (cm/h) for load and elution antibody binding capacity (g/L) % antibody yield % reduction of CHOP % reduction aggregate
100
100
100
40 95 95 50
30 97 95 95
20 97 99.9 95
# of aliquots per batch # of batches per column of resin # of times replace resin per year
2.0 25 2
5.0 9 7
7.2 6 11
Total Load Time per batch
7.9
20.6
27.8
6.2. Optimum Initial and Final Conditions for the Protein A, IEX and HIC Columns and Viral Inactivation Table 6.2 Initial and Final Conditions for the Protein A, IEX and HIC Columns, and Viral Inactivation
Initial volume (L) Initial antibody (g) Initial CHOP (mg) Initial aggregate (mg)
Protein A Column 1126 11042 4.14E+07 215382
Viral Inactivation 314 11015 2.17E+06 161537
IEX Column 316 11008 2.17E+06 167998
HIC Column 427 10448 1.14E+05 8819
Final Volume (L) Final antibody (g) final CHOP (mg) Final Aggregate (mg)
315 11015 2.17E+06 161537
316 11008 2.17E+06 167998
427 10448 1.14E+05 8820
608.10 10439 119.79 463
Table 6.2 is a breakdown of the initial and final conditions into the Protein A, IEX and HIC columns, and the viral inactivation step. 6.3. Optimum Configuration Summary of Initial and Final Conditions per Batch Table 6.3 Summary of Initial and Final Conditions per Batch Initial volume (L) Initial product antibody (g) Initial CHOP (g) Initial aggregate antibody (mg)
In 1126 11258 41392 215382
# batches per year
70
Out 608 10439 7.3 120
% Out/In 54.0 92.7 0.018 0.056
Table 0 shows the initial volume, antibody, CHOP and antibody aggregate that enters the chromatography sequence from the first UF/DF and the final conditions after the fluid has filtered through all of the columns and viral inactivation per batch. The percentage of product, Out/In multiplied by 100, shows that the process recovers 92.7% of the antibody, while capturing only 0.018% of the CHOP and 0.056% of the aggregate. There are 70 batches total per year. 6.4. Buffer Cost Analysis of Optimum Configuration Process
Sanitize Equilibrate Load Wash Elute Regeneration Clean
Table 6.4 Buffer Cost Analysis Protein A IEX Column Column $14,424 $9,788 $24,127 $16,373 n/a n/a $24,652 $16,729 $3,619 $2,456 $15,407 $10,456 $7,300 $1,944
Total Cost per batch Total cost per year Total Buffer Volume (L)
$89,530 $6,356,652 1,473
$57,745 $4,099,916 392
HIC Column $13,968 $23,365 n/a $23,873 $3,505 $14,921 $1,944 $81,575 $5,791,854 392
6.5. Viral Inactivation Specifications and pH of solutions Table 6.5 a) Viral Inactivation Specifications 3.6 Desired pH of VI 8 Minimum hold time required (hours) Increase of aggregate during hold time 4 (%) Process Vessel (L)
400
volume of HCl (L) volume of NaOH (L)
0.563 0.641
Table 6.5 b) pH of Solutions HCl pH (acid) 0.301 NaOH pH (base) 13.7 Solution pH pre VI 4.5 Solution pH post VI 6
[7.2]
7. References 7.1. “Design Constraints April 6 2013”, Coho BioSciences. 7.2. “Antibody Purification Handbook: Appendix 1. Characteristics of Protein G and Protein A sepharose products”. GE Healthcare, 2002.
Table of Contents 1. Executive Summary 2. Introduction 3. Results 3.1. Column Configuration Specifications- Table 3.2. Flow diagram of Column Configurations 3.2.1. Base Case Flow Diagram- Figure 3.2.2. Continuous Column Rotation Logistic Diagram - Figure 3.2.3. Optimized Process Flow Diagram- Figure 3.3. Percentage of Antibody Lost - Figure 3.4. Remaining CHOP (%)- Figure 3.5. Remaining Antibody Aggregate (%)- Figure
4. Discussion 4.1. Optimization 4.2. Continuous Configuration 4.3. Major Challenges
5. Cost Analysis 5.1. Operating Cost per Year- Figure 5.2. Capital Cost- Figure 5.3. Total Cost Over 30 year Plant Lifetime- Figure 5.4. Overall Cost Analysis of Optimum Configuration Process- Table
6. Appendices 6.1. Optimum Column Configuration Specifications- Table
6.2. Optimum Initial and Final Conditions for the Protein A, IEX and HIC Columns and Viral Inactivation- Table 6.3. Optimum Configuration Summary of Initial and Final Conditions per Batch- Table 6.4. Buffer Cost Analysis of Optimum Configuration Process- Table 6.5. Viral Inactivation Specifications and pH of solutions- Table
7. References
1. Executive Summary In partnership with CellDex, Coho BioSciences will bring an antibody-conjugate cancer drug, glembatumumab vedotin CDX-011, to the worldwide market. To prepare for the manufacturing of the product, a preliminary facility design was requested by the Coho BioSciences Board of Directors to provide a better understanding of the economics and technicality of the production. It is very important that the drug meets delivery and purity specifications to meet FDA requirements and approval. Thus, it is essential to have a rigorous purification process. Initial purification of the antibody through a centrifuge and depth filters provides a rough primary recovery. The harvest fluid is then sent to a column chromatography sequence for more thorough purification. The column chromatography sequence is comprised of a protein A, ion exchange (IEX) and hydrophobic interaction (HIC) column, with a viral inactivation step between the protein A and IEX column. Also, both the operating costs including resin costs of $2000 per liter, and capital costs of the columns, ranging from $25,000 to $1.6 million for 10 to 100 cm diameter columns, respectively, contribute to a large portion of the overall plant costs. [7.1] Thus, it is advantageous to reduce the overall cost of this process, while maintaining delivery and purity specifications. At the request of the Coho Board of Directors, the following document discusses and analyzes the investigation of the best configuration of the column chromatography sequence. The system will be optimized to maximize the capture of antibody, minimize the CHOP and aggregate antibody captured, and minimize the overall cost and time to run the process.
2. Introduction The column chromatography sequence is comprised of three different chromatography columns, protein A, ion exchange (IEX) and hydrophobic interaction (HIC) columns, and a viral inactivation step to purify the harvest fluid after the primary recovery process. This process provides the ultimate purification step of the antibody before it is conjugated to form the product, CDX-011. Each column has a respective resin to capture the antibody, and the linear flow rate of the entering fluid determines the antibody binding capacity, percent antibody yield, percent reduction of CHOP, and percent reduction of aggregate. Resin is an expensive operating cost of $2000 per liter, so it was necessary to determine the best configuration of chromatography columns to reduce the operating and capital costs, maximize the capture of antibody, and reduce the amount of CHOP and undesired aggregate. The amount of resin is determined by the size and number of columns used. The larger the column, the greater the capital costs. However, this may not necessarily mean greater operating costs depending on the column configuration. In addition to the resin, there are 5 different buffers used to sanitize/clean, equilibrate, wash, elute and regenerate the resin for each batch. Each buffer has a respective amount necessary to prepare the column for the batch and thus determined a constraint for the time to run the batch. Investigation of different configurations of columns in series and in parallel provided some solutions to the time constraint. As mentioned earlier, there is also a viral inactivation step in the sequence between the protein A and IEX column. The eluate from the protein A column must be held at a lowered pH at the corresponding holding time to inactivate the virus and raised to 6.0 where the antibody is stable. Then it is further purified in the IEX and HIC columns. The scope of this document focuses on just the chromatography sequence so it was assumed that the optimum pH was 3.6 and held for a minimum of 8 hours. A base case of the chromatography sequence was completed earlier and a discussion and cost analysis to compare it with the optimized case will be presented. The following document will also discuss the investigation of implementing a continuous chromatography processes.
3. Results 3.1. Column Configuration Specifications Table 3.1 is the column configuration specifications for the base case, continuous, and optimum configuration. It includes the diameter, number of columns in parallel and linear load flow rate of each case. Table 3.1 Column Configuration Specifications Protein A IEX HIC Column Column Column Diameter (cm) Base Case Continuous Optimum Configuration Number of columns In parallel Base Case Continuous Optimum Configuration Linear Load Flow Rate (cm/h) Base Case Continuous Optimum Configuration
100 20 65
100 20 30
100 20 30
1 12 2
1 10 4
1 34 4
100 600 100
100 300 100
100 300 100
3.2. Flow diagram of Column Configurations 3.2.1. Base Case Flow Diagram
In Series 1 3 2
Figure 3.2.1 is the process flow diagram for the base case chromatography sequence. There is only one protein A, one IEX and one HIC column for the column configuration. Coincidentally, it also represents the block flow diagram for all of the chromatography configurations. The viral inactivation step is not drawn but is between the Protein A and IEX columns. 3.2.2. Continuous Column Rotation Logistic Diagram
Figure 3.2.2 is a continuous column rotation diagram for one of the column chromatography cycles. Each column goes through a load, wash, elute, regenerate, sanitize and equilibrate step to complete the column cycle. For example, this protein A column cycle has 9 columns that go through the cycle simultaneously. Three columns in series are loaded in series from the UF/DF in columns Load 1, 2 and 3 until Load 1 column reaches binding capacity. Load 1 column is then washed and follows the rest of the cycle in sequential order. Load 2 was not filled to binding capacity and becomes a Load 1 column until it reaches binding capacity. Each subsequent column moves through this sequence and rotates through all the columns in a “continuous” cycle.
3.2.3. Optimized Process Flow Diagram Figure 3.2.3 is a process flow diagram of the optimized batch column configuration. Each column has 2 columns in series, with 2 columns running parallel for the Protein A, 4 for the IEX and 4 for the HIC column with a viral inactivation holding tank, not drawn, between the protein A and IEX column.
3.3. Percentage of Antibody Loss between Column configurations
Percent Antibody Lost (%) 16.7 10.3
2.3
Base Case
Continuous
Optimum Configuration
Figure 3.3 shows the percentage of antibody lost from the base case, continuous, and optimum configuration processes. The base case process has the highest percentage antibody loss of 16.7%, next is the continuous with 10.3% and the lowest is the optimum case with 2.3% loss. 3.4. Remaining CHOP (%)
Remaining CHOP (%) Base Case
Continuous
Optimum Configuration
0.25
6.4E-07
3.125E-08
Remaining CHOP (%)
Figure 3.4 is the percentage of CHOP remaining from the base case, continuous, and optimum configuration processes. The base case process has the highest percentage antibody loss of 0.25%, next is the continuous with 6.4E-07% and the lowest is the optimum case with 3.1E-08% loss.
3.5. Remaining Antibody Aggregate (%)
Remaining Antibody Aggregate (%) Base Case
Continuous
Optimum Configuration
0.13
0.0018
0.0002
Remaining Antibody Aggregate (%)
Figure 3.5 is the percentage of antibody aggregate remaining from the base case, continuous, and optimum configuration processes. The base case process has the highest percentage antibody loss of 0.13%, next is the continuous with 1.8E-03% and the lowest is the optimum case with 2E-04% loss.
4. Discussion The best column chromatography process is one that maximizes the capture of antibody, minimizes the CHOP and aggregate antibody captured, and minimizes the overall cost and time to run the process. 4.1. Optimization The optimization of the process was heavily dependent on the configuration of the chromatography columns. The main optimization variables were the number of columns in series and in parallel, and the size of the columns. Increasing the number of columns in series, by assuming that the first column captured 100% of the antibody bound to the resin, but only lost part of it in the second or third column, drastically reduced the antibody lost, as seen in Figure 3.3. However, 2 columns are optimal for batch system, as the 3rd column had negligible contribution and performed better than the continuous process as seen in Figure 3.4 and 3.5. Smaller columns and varying the sizes of columns between protein A, IEX and HIC steps also optimized the process from the base case. Since the protein A resin had a higher binding capacity than the other column resins, larger columns were optimal for this step. 4.2. Continuous Configuration Another way to optimize the system was investigating a continuous chromatography process. Technically, the system is “semi-continuous” since the columns are run staggered and in parallel, rotating through the chromatography cycle, as in figure 3.2.2. By implementing the continuous cycle, the rotation of the column was contingent on the loading cycle time, which generally takes less time than the limiting, wash step. In order to continually rotate through the columns, the number of columns greatly increased to compensate for the difference in time, which also increases capital costs. However,
reducing the size of the columns also helped to reduce the cost and overall processing time. 4.3. Major Challenges The major challenges encountered in the process were determining a logical configuration for the continuous chromatography sequence and an optimal size to number column ratio for the each case.
5. Cost Analysis 5.1. Operating Cost per Year
Operating Cost per Year Base Case
Continuous
Optimum Configuration
$68,444,849 $52,563,090 $32,151,763
Operating Cost per Year
Figure 5.1 shows the operating cost per year of the base case, continuous and optimum column configuration. The base case process has the highest operating cost at $68.4 million per year, continuous is $52.6 million per year and the optimum configuration is the lowest at $32.2 million per year.
5.2. Capital Cost
Capital Cost Base Case
Continuous
Optimum Configuration
$22,400,000 $11,600,000 $4,800,000
Captial Cost
Figure 5.2 shows the capital cost of the base case, continuous and optimum column configuration. The continuous process has the highest operating cost at $22.4 million, the optimum configuration is $11.6 million and the base case is the lowest at $4.8 million. 5.3. Total Cost Over 30 year Plant Lifetime
30 Year Total Cost Base Case
Continuous
Optimum Configuration
$2,058,145,457 $1,599,292,705 $976,152,888
30 year total cost
Figure 5.3 shows the total cost over the 30 year facility lifetime for the base case, continuous and optimum column configuration. The base case process has the highest operating cost at $2.06 billion, continuous is the next highest at $1.6 billion and the optimum configuration is the lowest at $976 million. Even though the batch case had
much lower capital costs than the continuous case, the high operating cost results in a more expensive process over the 30 year lifetime.
5.4. Overall Cost Analysis of Optimum Configuration Process The protein A column is the most expensive step of the process due to the 4 large, 65 cm diameter column, which accounts for the high resin costs per year. Buffer and resin account for two-thirds of the operating cost of the process due to the sheer volume of buffer used for each batch. Captial cost also surprisingly are very similar between different columns even though there are 4 large columns for the Protein A and 8 smaller 30 cm columns for the IEX and HIC columns. Buffer costs are also surprisingly within a million dollars of each other per year which is opposite to the operating costs for resin. Table 5.4 Overall Cost Analysis Protein A Column IEX Column HIC Column $6,356,652 $4,099,916 $5,791,854
Total $16,248,421
$7,963,937 $70,000
$1,979,203 $140,000
$3,110,177 $140,000
$13,053,317 $350,000
Skid Cost Viral filtration
$500,000 $23.95
$1,000,000
$1,000,000
$2,500,000 $24
Total operating cost per year Captial cost
$14,890,589 $3,600,000
$7,219,119 $4,000,000
$10,042,031 $4,000,000
$32,151,739 $11,600,000
Total Buffer Cost per year Cost of Protein A resin per year QA & QC Testing
6. Appendices 6.1. Optimum Column Configuration Specifications Table 6.1 Column Specifications Protein A IEX Column Column 20 25 bed height (cm) 65 30 Column Diameter (cm) 2 4 Number of columns in parallel 2 2 Number columns in series
HIC Column 25 30 4 2
Linear Flow rate (cm/h) for load and elution antibody binding capacity (g/L) % antibody yield % reduction of CHOP % reduction aggregate
100
100
100
40 95 95 50
30 97 95 95
20 97 99.9 95
# of aliquots per batch # of batches per column of resin # of times replace resin per year
2.0 25 2
5.0 9 7
7.2 6 11
Total Load Time per batch
7.9
20.6
27.8
6.2. Optimum Initial and Final Conditions for the Protein A, IEX and HIC Columns and Viral Inactivation Table 6.2 Initial and Final Conditions for the Protein A, IEX and HIC Columns, and Viral Inactivation
Initial volume (L) Initial antibody (g) Initial CHOP (mg) Initial aggregate (mg)
Protein A Column 1126 11042 4.14E+07 215382
Viral Inactivation 314 11015 2.17E+06 161537
IEX Column 316 11008 2.17E+06 167998
HIC Column 427 10448 1.14E+05 8819
Final Volume (L) Final antibody (g) final CHOP (mg) Final Aggregate (mg)
315 11015 2.17E+06 161537
316 11008 2.17E+06 167998
427 10448 1.14E+05 8820
608.10 10439 119.79 463
Table 6.2 is a breakdown of the initial and final conditions into the Protein A, IEX and HIC columns, and the viral inactivation step. 6.3. Optimum Configuration Summary of Initial and Final Conditions per Batch Table 6.3 Summary of Initial and Final Conditions per Batch Initial volume (L) Initial product antibody (g) Initial CHOP (g) Initial aggregate antibody (mg)
In 1126 11258 41392 215382
# batches per year
70
Out 608 10439 7.3 120
% Out/In 54.0 92.7 0.018 0.056
Table 0 shows the initial volume, antibody, CHOP and antibody aggregate that enters the chromatography sequence from the first UF/DF and the final conditions after the fluid has filtered through all of the columns and viral inactivation per batch. The percentage of product, Out/In multiplied by 100, shows that the process recovers 92.7% of the antibody, while capturing only 0.018% of the CHOP and 0.056% of the aggregate. There are 70 batches total per year. 6.4. Buffer Cost Analysis of Optimum Configuration Process
Sanitize Equilibrate Load Wash Elute Regeneration Clean
Table 6.4 Buffer Cost Analysis Protein A IEX Column Column $14,424 $9,788 $24,127 $16,373 n/a n/a $24,652 $16,729 $3,619 $2,456 $15,407 $10,456 $7,300 $1,944
Total Cost per batch Total cost per year Total Buffer Volume (L)
$89,530 $6,356,652 1,473
$57,745 $4,099,916 392
HIC Column $13,968 $23,365 n/a $23,873 $3,505 $14,921 $1,944 $81,575 $5,791,854 392
6.5. Viral Inactivation Specifications and pH of solutions Table 6.5 a) Viral Inactivation Specifications 3.6 Desired pH of VI 8 Minimum hold time required (hours) Increase of aggregate during hold time 4 (%) Process Vessel (L)
400
volume of HCl (L) volume of NaOH (L)
0.563 0.641
Table 6.5 b) pH of Solutions HCl pH (acid) 0.301 NaOH pH (base) 13.7 Solution pH pre VI 4.5 Solution pH post VI 6
[7.2]
7. References 7.1. “Design Constraints April 6 2013”, Coho BioSciences. 7.2. “Antibody Purification Handbook: Appendix 1. Characteristics of Protein G and Protein A sepharose products”. GE Healthcare, 2002.
