Upload
duongtram
View
214
Download
0
Embed Size (px)
Citation preview
0
Implementation of UPLC as new technology for Biomolecules in Pharma
Mario HellingsAssociate Director Analytical Sciences & Centers of Excellence Janssen Pharmaceutical Companies of Johnson & Johnson, Beerse, Belgium
17th October 2012 – Waters BioLC Summit
Historical
2
Before 2010:
• SM & LM separate
• Mixture of binary UPLC and quaternary H-Class system for SM applications
• No harmonization between SM applications (cleaning, assay/purity, CU, dissolution,...)
• No harmonization between R&D and commercial
• No use of UPLC for LM applications
After 2010:
• Merge of SM & LM
• Focus on cost/time reduction & efficiency improvement
• Platforming
• Standardization/Harmonisation of approaches
• Focus on easy transferability between sites
• New technology introduction team
• Explore UPLC for LM as new technology
J&J UPLC Implementation plan
Application for LM: Protein/MAb Characteristics
3
• Is the biomolecule what I think it is?• Is it biologically active?• Is the quality ok?• Is product consistent for release?
Analytical Techniques used for biomoleculesTechnique Use
SDS-page / cSDS Identity, Purity (Proteins)
cIEF Identity, Purity, Isoform distribution
Size Exclusion (HPLC) Identity and purity (aggregation, size/shape variants)
Oligosaccharide mapping (HPLC) Characterization and quantification of N-, O-linked glycans
Peptide mapping (HPLC) Identity, characterization of post-translational modifications (e.g. oxidation, deamidation, etc.)
Cell-based bioactivity Bioactivity
ELISA Bioactivity and Impurity (Host cell protein, Residual Protein A, BovIgG, BSA)
qPCR Impurity (DNA)
Microbiology Bioburden, Mycoplasma, Endotoxin, etc.
Light Scattering, Circular Dichroism, LC-MS-MS, ultracentrifugation, Karl-Fisher,..
Characterization, purity, stability
4
Analytical Techniques used for biomoleculesTechnique Use
SDS-page / cSDS Identity, Purity (Proteins)
cIEF Identity, Purity, Isoform distribution
Size Exclusion (HPLC) Identity and purity (aggregation, size/shape variants)
Oligosaccharide mapping (HPLC) Characterization and quantification of N-, O-linked glycans
Peptide mapping (HPLC) Identity, characterization of post-translational modifications (e.g. oxidation, deamidation, etc.)
Cell-based bioactivity Bioactivity
ELISA Bioactivity and Impurity (Host cell protein, Residual Protein A, BovIgG, BSA)
qPCR Impurity (DNA)
Microbiology Bioburden, Mycoplasma, Endotoxin, etc.
Light Scattering, Circular Dichroism, LC-MS-MS, ultracentrifugation, Karl-Fisher,..
Characterization, purity, stability
5
Potential for UPLC
What is so special about Ultra Performance Liquid Chromatography (UPLC)?
6
• Column has smaller particles:• Provide increased efficiency (lower pate
height H)• Maintain efficiency over a wider linear
range• Ability to increase resolution and speed
of separation• Need for a system with ultra-low
dispersion!
Well-known & implemented for SM applications
A cross-functional team created
Aim: Proof of concept of UPLC for LM, increase resolution (i.e better separation) & speed in robust manner in the LM applications
A well-known compound selected as test subject and DOE principles applied
HPLC UPLC
Size Exclusion Chromatography (SEC)
• Analysis of aggregates, monomer (IgG) and fragments
• Release and stability testing of biomolecules according platform method
• DOE:– Mobile phase
0.2M Sodium Phosphate pH 6.8 or pH 6.0;50mM Sodium Phosphate, 200mM Sodium Chloride pH 6.0
– Temperature30°C, 45°C, 60°C
– Flow rate0.2, 0.4 or 0.6 mL/min
– Sample concentration0.15, 0.5 or 0.75 mg/mL
– Sample volume3, 6 or 10 µL
7
SEC – Mobile Phase
8
Fragment A
A. 0.2M NaPi pH 6.8B. 0.2M NaPi pH 6.0C. 50mM NaPi, 0.2M NaCl pH 6.0
Fragment B
Histidine
Zoomed 4x
SEC DOE – Resolution Aggregate - Monomer(Mobile Phase A)
9
604530
1.80
1.75
1.70
1.65
1.60
0.60.40.2
0.750.500.15
1.80
1.75
1.70
1.65
1.60
1063
Temperature
Mea
nFlow rate
Concentration Volume
Main Effects Plot for Raggr-monData Means
Optimized conditions
• Based on main effect plots:
– Mobile phase: A 0.2 M Sodiumphosphate pH =6.8
– Temperature: 30°, Higher temperature gives temperature induced fragments
– Flow: 0.4 mL a higher flow gives problems with the maximum pressure of the system.
– Injection volume: 3 µL gives the best results in resolution both between aggregates and Monomer and between Monomer and fragments.
• Method parameters tested: LOQ, Linearity, Reproducibility
10
SEC UPLC - Reproducibility
Reproducibility UPLC vs HPLC
11
UPLC HPLC differencesample aggregate monomer fragment aggregate monomer fragment aggr mono fragBULK 0.15 99.83 0.02 0.09 99.91 0.00 0.06 -0.08 0.02BULK 0.13 99.85 0.02 0.07 99.93 0.00 0.06 -0.08 0.0212M/5°C 0.32 99.63 0.04 0.23 99.76 0.02 0.09 -0.13 0.0224M/5°C 0.29 99.62 0.09 0.20 99.75 0.04 0.09 -0.13 0.054M/5°C 0.38 99.54 0.08 0.34 99.62 0.04 0.04 -0.08 0.048M/5°C 0.20 99.72 0.08 0.11 99.85 0.04 0.09 -0.13 0.04BULK 0.11 99.88 0.01 0.06 99.94 0.00 0.05 -0.06 0.0117M/35°C 4.84 89.70 5.46* 3.89 93.35 2.76 0.95 -3.65 0.4431M/5°C 0.53 99.33 0.14 0.41 99.51 0.08 0.12 -0.18 0.0624M/25°C 1.12 97.68 1.20 0.89 98.14 0.97 0.23 -0.46 0.23
* = sum of 2 fragmentsfragment A = 2.24%fragment B = 3.20%
Change in trend : UPLC shows more aggregate and fragmentExplanation: less monomer tailing and better resolutionNo impact on specs
SEC UPLC – ReproducibilityReproducibility UPLC vs HPLC
12
UPLC HPLC
UPLC HPLCsample aggregate monomer fragment aggregate monomer fragment17M/35°C 4.84 89.70 5.46* 3.89 93.35 2.76
* = sum of 2 fragmentsfragment A = 2.24%fragment B = 3.20%
Overlay GF Standard UPLC and HPLC
13
SEC UPLC – Column Lifetime
14
After ± 560 injections After ± 700 injections
UPLC SEC POC: Conclusions
• UPLC method gives better resolution and is more sensitive than HPLC
• Less difficult to integrate
• Duration of 1 run = 6 min instead of 20 min HPLC
15
Peptide mapping
• The Identity Peptide Mapping assay is Final Bulk release assay requested by the Japanese Agency for some products
• The current method focuses on the Peak Area Ratio (PAR) and Relative Retention Time (RRT) for two product specific peptides for identity. One Reference peptide peak is used for the calculation.
• The peptides used are:
– HC60-65 (Identity Peptide)
– HC66-76 (Identity Peptide)
– HC284-297 (Reference Peptide)
• The goal for the POC for the Identity Peptide Map is to decrease the runtime of the assay with comparable resolution and same number of identified peaks.
16PDMS-AD-LMMD Leiden
Peptide Mapping HPLC with Vydac C18 vsUPLC with BEH300 column
17
m in2 0 4 0 6 0 8 0 1 0 0 1 2 0 1 4 0
m A U
0
5 0
1 0 0
1 5 0
2 0 0
2 5 0
D A D 1 A , S i g = 2 1 4 ,4 R e f = 5 5 0 , 1 0 0 ( M A R T I J N T E N . .. T O 1 2 7 5 _ 1 4 8 L Y S C _ 0 7 D E C 1 0 2 0 1 0 -1 2 -0 8 1 7 -3 2 -0 9 \ M A 0 7 D E C 1 0 0 0 0 0 0 9 . D )
HC 60-65 (25min)
HC 66-76(38min)
HC 284-297Reference Peak (65 min)
HC 60-65
(13 min)
HC 66-76 (20min)
HC 284-297Reference Peak(31min)
HPLC
UPLC
Gradient optimization using LC simulation software
• To determine the gradient for the optimal separation of some of the peptides LC simulation software (LC simulator) was used.
• The software determined the optimal gradient to separate all the peptides based on different parameters (Retention time, Peak Area and Peak Height) from the first run using a linear gradient.
• The proposed optimal gradient by the LC simulation software was 3-8 %B (15 min); 8-18 %B (15min); 18-22%B (10min) and 22-50%B (50min).
• The proposed gradient by the simulation software was executed on a UPLC H-class using a flow at 0.2 mL/min and column temperature of 45°C.
18
Gradient optimization using LC simulation software
Acquity BEH 300 C18 1.7um 2.1*150mm17:54:42
30-May-2011
ESI+ 50-2000 - MSE (high energy CE 20-60eV)Synapt HDMS G2Lys C digested CNTO 148 (Simponi)
Time10.00 15.00 20.00 25.00 30.00 35.00 40.00 45.00 50.00 55.00 60.00 65.00 70.00 75.00
AU
0.0
2.0
4.0
6.0
8.0
1.0e+1
CNTO 148_110530_MS011_03 4: Diode Array Range: 1.393e+17.30
64.62
42.84
41.16
21.38
12.35 13.2714.14
30.56
22.8725.23 29.06
28.5038.6134.55
37.96
46.11
45.79
56.2955.6149.35
48.94
50.77
63.56 68.4565.60 69.02
UV chromatogram @ 214nm
HC 66-76(25.2min)
HC 284-297Reference Peak(41.2min)
7472706866646260585654525048464442403836343230282624222018161412108642
Time, min
0.5
1.0
1.5
2.0
2.5
HC344-347LC185-189 HC223-227
HC60-65
LC209-215HC59-65 LC105-108
HC419-423LC147-150
LC151-170
HC66-76HC449-456
HC449-455
HC336-343LC192-208
HC143-156
HC350-369
HC131-142
HC370-379
LC171-184
HC424-448HC380-401
HC402-418
LC109-127
HC258-283
LC128-146
HC232-257
HC232-255
LC1-39HC298-326 Glycans
HC298-326 Glycans
LC40-104
HC157-219HC157-214 LC1-104
HC1-59 (N43 IsoD)HC1-58 (N43 IsoD)HC1-59
HC1-58HC1-59 (N43 D)
HC1-58 (N43 D)
MS TIC50 - 2000
HC 60-65(12.4min)
HC 66-76 (24.9min)
HC 60-65(12.3min)
HC 284-297Reference Peak(40.5min)
The simulated peptide Retention times and profile were very comparable to the determined Retention times and profile of the peptides from the experiment.
Peptide Mapping UPLC with BEH300 column and increased flow of 0.3mL/min
20
Acquity BEH 300 C18 1.7um 2.1*150mm19:34:17
30-May-2011
ESI+ 50-2000 - MSE (high energy CE 20-60eV)Synapt HDMS G2Lys C digested CNTO 148 (Simponi)
Time10.00 20.00 30.00 40.00 50.00 60.00 70.00 80.00 90.00
AU
0.0
2.0e-1
4.0e-1
6.0e-1
8.0e-1
10.00 20.00 30.00 40.00 50.00 60.00 70.00 80.00 90.00
AU
0.0
2.0e-1
4.0e-1
6.0e-1
8.0e-1
CNTO 148_110530_MS011_04 4: Diode Array 214
Range: 2.3291.42
2.45
4.9143.56
29.0928.0120.5814.35
8.965.22 15.40 26.2323.29
31.3433.44 36.6442.87 46.12
CNTO 148_110530_MS011_03 4: Diode Array 214
Range: 2.3292.85
3.59
5.17
7.30 64.62
42.84
41.1630.5621.3813.277.66
22.8725.23 38.6134.55
46.1156.2949.35 54.37
50.76
63.56 68.45
m in2 0 4 0 6 0 8 0 1 0 0 1 2 0 1 4 0
m A U
0
5 0
1 0 0
1 5 0
2 0 0
2 5 0
D A D 1 A , S i g = 2 1 4 , 4 R e f = 5 5 0 , 1 0 0 ( M A R T I J N T E N . . . T O 1 2 7 5 _ 1 4 8 L Y S C _ 0 7 D E C 1 0 2 0 1 0 - 1 2 -0 8 1 7 - 3 2 - 0 9 \ M A 0 7 D E C 1 0 0 0 0 0 0 9 . D )
UPLC at 0.3mL/min runtime 60 min
UPLC at 0.2mL/min runtime 90 min
HPLC at 0.2mL/min runtime 180 min
Peptide map POC: Conclusions
• Runtime can be shortened from 180 min to 60 min.
• All peaks have been successfully identified by MS
• All peaks used for the release identity peptide map test calculation, are baseline separated and elute in the same order as in the HPLC peptide map platform method.
• The UPLC method is a suitable platform to replace the current HPLC Identity Peptide Map method
• Waters tests peptide mapping columns (PST) with peptide map mixture for reliable batch-to-batch reproducibility
21
Oligo Saccharide map
• Analysis of neutral glycans, and charged glycans.
– Release of Glycans with PNGase F
– Label Anthranilic Acid (2-AA)
• Detection at Fluorimetric
– Emission : 425 nm
– Excitation: 360 nm
22
Result Glycan standard
UPLC MethodBaseline separation between M5 and G2FG1F split into two peaksTotal run time 45 min.
WAX methodNo good separation between M5 and G2FTotal run time 120 min.
SA2F
SA1F
G2F
M5
G1F
G0F
Test substance reference standard
UPLC methodG1F split into two peaksSA2F 30.6 min.
WAX methodSA2F 84.4 min.
SA2F
SA1F
G2F
G1F
G0F
POC UPLC Oligo: Conclusions
• Faster and gives a better (baseline) resolution than the WAX method (30 minutes compared to 120 minutes).
• Results are comparable.
25
Overall summary UPLC for biomolecules
• UPLC Proof Of Concept was successful for:
– SEC
– Identity Peptide Map
– Oligo Saccharide Map
• Compared to classical HPLC UPLC provides:
– equal or better resolution
– faster analysis for all three methods
• UPLC could be a suitable new platform for SEC, Identity Peptide Map and Oligo Sacharride Map for future NME’s
Furter implementation
27
Succesfull proof of concept, how to push technology further?:
• Implementation on R&D portfolio • NME’s easy, late phase needs check for spec setting and comparison• Low impact and can be ran parallel with LC samples • Test more products to gain experience => 5 R&D compounds selected
• Implementation on legacy portfolio• Filter interesting products with customer, typically those were method issues
or improvement is desired => 2 extra compounds selected• Perform testing and check if spec changes• Check impact with regulatory
• Demonstrate System/Column/Vendor reliability• Extensive column/system testing => high volume characterization groups
ideally suited for this as they can perform parallel testing• Vendor needs to give insight in manufacturing process and associated quality
control => Waters provided this information
Outcome
28
• POC SEC method could be applied for all 8 compounds (3 legacy, 5 R&D)
• For legacy although excellent results and no spec impact several pitfalls 1. Batches are not split and products were global
• Use of technology only applicable if all countries accepted (US, EU, ROW, Canada, Japan)
• Takes up to 3 years• Parallel testing is useless and expensive for commerial sites
2. Just no business benefit (e.g. product sales decrease, filing cost to high)• E.g. products sales decrease, filing cost to high to do effort
3. No spec changes but trend breaches• Related to better separation, better integration • Hesitance as this might trigger more questions, comparison etc.
• Technology implemented within the R&D functional groups responsible for the LM test package
Non-scientific part
29
• Platform selection throughout entire J&J • Bio-H-Class for LM applications and mixed groups
• Absolute requirement for SEC
• H-class for SM applications • Quaternary system provided required flexibility• Can run traditional LC methods
• Available binary systems• Used for non-gmp characterization purposes only and if broken replaced
by I-Class
Non-scientific part
30
• Define global site roll-out map • # units/site/year• Replacement program• SIPOC’s for ownership/maintenance/training etc.
• Price negotiation with vendor Platform prices Exchange of old systems / Trade-in’s (max 2 for 1) Training packages, consumables, support, ....
Lessons learned
31
• Global Capital Request hard to achieve• Different organizations (R&D – Commercial) => Finance pushed to take up
into CAPEX• Not taken up in CAPEX
• System Lease => loss of all benefits because third party interaction, also not desired by internal finance
• Get now, pay later => very interesting, but book values need to be captured upon delivery so not a solution for finance
• Reprioritize and free up CAPEX
• Numbers fluctuate• Remain ball-park figures, Pharma is just not predictable (budget cuts, new
licensed projects, issues, etc.)• Incorporate in site master plans
• Strong alignment internally needed + alignment in vendor’s organization• SIPOC’s for J&J (Project lead + Procurement) and SIPOC’s for Waters• Strong internal communication needed• Follow-up of delivery and installation (problems)
General Conclusion
32
Benefits by selecting platform
• Global Alignment• Easy and fast transfer of methods• Training more effective, less confusion with analysts• Standard for new compounds, switch for old methods with quality or
efficiency issues• Flexibility (SM/LM, quaternary/binary, LC/UPLC)
• Cost saving• Operational cost related to transfers or analyst mistakes• 2 for 1 trade • Maintenance• Energy [HPLC 56W vs (Bio)-H-Class 29W]• Increased throughput
• Space saving• HPLC 58.4cm vs (Bio)-H-Class 34.3cm (with 30cm oven 54cm)
Acknowledgements
33
• Werner Claes• Tom Vennekens• Hein Van Nunen• Martijn Ten Asbroek• Herman Bentlage• Jan Plasse• Koen Demoen• Ivan Somers• Koen Vanhoutte• Els Poupeye• Davy Petit• Kim De Kock• Marleen van Wingerden• ....