2. • Dasiprotimut-T
• Therapeutic cancer vaccine
• Treatment for non-
Hodgkins lymphoma
• Trains immune system
against BCR
• Located in Coon Rapids,
MN
Biovest International
3. ELISAs for Dasiprotimut-T
1. Qualitative
a) Fusion Screening
2. Quantitative
a) Antibody concentration
• IgM/IgG
b) Impurities
• Host cell proteins
• Bovine proteins
6. Fusion Screening ELISAs
• Qualitative so development of assay
was faster
• Fewer requirements for validation
• Developed in house so cheaper
reagents and you can buy them in bulk
• Needed to be completely developed
• Find vendors for appropriate antibodies
and buffers
• Performed qualification study to determine
concentrations
• Precious samples, only had enough each
time for one ELISA
• Harder to be prepared for, need at least 20
ELISA plates coated at anytime
• Does not have a standard curve, harder to
monitor assay performance over time
9. Quantitative ELISAs
• Qualifications were completed
• Developed in house so cheaper
reagents and you can buy them in bulk
• Quantitative ELISA so additional
requirements are needed for validation
(accuracy, repeatability)
• Concentration of sample is unknown
• Analysis may need to be repeated until
the sample quantitates on the standard
curve
11. Impurity ELISA - HCP
TMB
TMB
HRP
Anti-K6H6/B5
Host cell
proteins
Biotinylated anti-K6H6/B5
B
B
B TMB
TMB
HRP
TMB
TMB
HRP
Streptavidin HRP
12. Impurity ELISA - HCP
• Assay antibodies were well
characterized
• Very low background
• Tailor-made for our product
• Very limited quantity without another
rabbit study
• Variability with biotinylated antibody
• Depending on # biotins per antibody
the working dilution changed per batch
• Analysis may need to be repeated until
the sample quantitates on the standard
curve
13. Impurity ELISAs –
Bovine proteins
Anti-bovine protein
Bovine
protein
Anti-bovine protein HRP
TMB
TMB
HRP
14. Impurity ELISAs -
Bovine proteins
• Plates are pre-coated and pre-blocked
• Plates don’t need to be coated ahead
of time
• Analysis time is shorter
• Each kit has been tested as a lot
• Plate strips can be separated
• Very expensive ($500-750)
• Large lot to lot variability
• Very sensitive to timing on an
automated system
17. Tecan Automated Systems
• Can process samples overnight
• Process more plates at one time
• Automation results in a higher
reimbursement from EMA
• Need fewer technicians to run ELISAs
• Timing of sample addition
• Need a programmer for plate maps
• May freeze
21. Quality Control
• Documentation
• Veeva Vault
• Writing of validation protocols
• ICH Q2(R1), FDA
• Preparation of final reports for EMA submission and protocol issuance
• Stability studies
• Lot to lot comparison
• Reagent inventory and budget
22. Tecan Automated Systems
1 2 3 4 5 6 7 8 9 10 11 12
A PTT
Blank
Control
RMQ013 Lambda
Vial 1
RMQ013 Kappa
Vial 1
B RMQ013 Lambda
Vial 21
RMQ013 Kappa
Vial 21
C Media
Blank
Control
RMQ013 Lambda
Vial 41
RMQ013 Kappa
Vial 41
D RMQ013 Lambda
Vial 61
RMQ013 Kappa
Vial 61
E IgM
Lambd
a
Control
RMQ013 Lambda
Vial 81
RMQ013 Kappa
Vial 81
F RMQ013 Lambda
Vial 100
RMQ013 Kappa
Vial 100
G IgM
Kappa
ControlH
Good morning. This morning I’m going to talk to you about the most recent project I was involved in.
I was part of a team to bring Dasiprotimut-T to market approval in the European Union. Dasiprotimut-T is a therapeutic vaccine for patients with non-Hodgkins lymphoma that works by eliciting a cytotoxic T-cell response against the patient’s B-cell receptor (BCR) on the surface of the lymphoma cells. The treatment was originally developed at the National Cancer Institute in collaboration with Biovest using Biovest’s AutovaxID hollow-fiber biorectors, pictured here.
During the vaccine manufacturing process there are different checkpoints that are monitored with the help of 8 different ELISAs. While I was at Biovest, a colleague and myself optimized and validated all 8 of these for use in Biovest’s facility. I’ve divided them into 2 major categories: qualitative and quantitative with the quantitative ELISAs being divided further into testing for antibody concentration and the presence of protein impurities. These ELISA assays needed to be validated for in-house use but also to support the in-process validation for vaccine manufacturing. Some of the ELISAs needed to be completely developed, some were qualified and some were kits that needed to be optimized for use at our facility.
Biovest receives a lymph node biopsy. The biopsy is macerated and a single cell suspension is generated. The single cell suspension is cryopreserved and an aliquot of cells is analyzed to determine the antibody isotype (IgM or IgG) utilizing flow cytometry.
K6H6/B5 fusion partner cell line is grown to mid-log phase (this is a non-secreting myeloma cell line). Patient tumor cells are thawed, washed and counted. Patient tumor cells are combined with K6H6/B5 cells in a polyethylene glycol fusion reagent. Cells are resuspened in culture medium and aliquoted into 96-well plates at various densities. Selection media is added, cells are grown and monitored for an average of 5 weeks. Fusion screening ELISAs are performed to identify candidate hybridoma clones.
The two qualitative ELISAs test for successful fusion of a patient’s biopsy cells with K6H6/B5. The plate is coated with either anti-IgM or IgG whichever was predetermined by flow cytometry. Since K6H6/B5 cells are non-secreting, successful fusions will express an antibody. In the beginning there are 5-15, 96-well plates per patient which is a ton of clones so we needed to develop a way to easily pick more successful clones.
So we developed a macro to process the qualitative data for the fusion screening ELISAs. A ratio is calculated using the positive control OD and sample OD. Anything over a ratio of 1, shows up red on the map, between 0.5 and 1 is orange and anything lower than 0.5 is in yellow. This is a clear visual for the technician of which fusion clones are currently secreting antibody the best for that patient. In the beginning many clones are picked per patient but as they are expanded, fewer are allowed to continue growing. After sequencing confirmation, up to 10 clones are chosen per patient.
When talking about it needing to be completely developed mention that it was also the first assay that had to be validated to support in process validation (vaccine manufacturing)
3. Fusions that are positive for secretion of the correct heavy and light chains are expanded. The fusion screening ELISA is continually performed to monitor for clones that continue to be positive for immunoglobulin secretion. In the end up to 10 positive clones are selected for further analysis/expansion and aliquots are cryopreserved. Immunoglobulin production of each cell line is measured by HPLC or quantitative ELISA. The IgH variable region of the candidate production banks are amplified by RT-PCR and sequenced. Production banks are chosen for scale-up based on immunoglobulin secretion, growth rate and variable region identity matching to the patient’s biopsy.
4. One of the five vials of the production bank are thawed, washed and counted. Cells are grown and expanded four to six times over two to three weeks depending on doubling time. Then the cells are transferred to a bioreactor and the hybridoma clones grows for an average of 4-6 weeks. During this time antibody concentration is monitored.
The quantitative ELISA for the antibody concentration is very similar to the fusion screening ELISAs. However they utilize a standard curve so the amount of antibody can be determined.
5. Once the cells begin dying, the contents of the bioreactor is harvested and purification begins. Over 5 different steps, the patient idiotype is purified from the cell culture supernatant. These steps vary depending on whether the isotype was IgM or IgG. Using different salt buffers, the harvest undergoes a round of affinity chromatography, then anion exchange, another round of chromatography followed by virus reduction and finally 0.2 micron filtration. During each stage of purification certain threshold need to be met for protein impurities and that is monitored using ELISA.
The first protein impurity I’m going to talk about is the host cell protein quantification ELISA. K6H6/B5 cells were weaned from serum, pelleted and sheered to make a pool of host cell proteins. These proteins were used as the reference standard in the ELISA and also used to immunize rabbits to produce anti-host cell protein antibodies. This was probably the most important ELISA because it’s unique to the product and you have to make sure the host cell proteins are removed during purification so you’re not injecting foreign proteins into the patient. The purification process effectively eliminated these proteins so streptavidin was needed to amplify the signal.
The remaining three assays share two things in common: they are all bovine proteins and they are all related to bovine proteins found in the serum added to the tissue culture during the cell expansion process. These assays detected bovine serum albumin, bovine IgG and bovine plasma proteins. Bethyl Laboratories manufactures the BSA kit and Cygnus Technologies produced the BSA and bovine IgG.
The automated systems we used were made by Tecan. The EVO 100 made dilutions and transferred samples to the ELISA plate. There are slots on the bay for sample diluent, 1 mL, deep well plates for making dilutions, 96-well ELISA plates and conductive tips.
Once the samples were loaded on the ELISA plate, the plate is transferred to the EVO 150 to complete the ELISA assay. In addition to having the same reagent pipetting capabilities of the EVO 100, the 150 also has an integrated 96-head plate washer. “Hotels” which hold the ELISA plates protecting them from light during incubation samples, are capable of shaking the plate at a specific speed or holding the plate at a certain temperature. And finally there is an integrated plate reader to measure the sample OD when the assay is finished.
So once all of the ELISAs are completed and the patient idiotype was successfully purified, conjugation of the product can begin. Vials for the vaccine product and KLH are sterilized in a glove box. Purified idiotype, KLH and buffer are mixed. Glutaraldehyde is added and vaccine is mixed for 2-4 hours. Glycine is added to stop the reaction and final product is dialyzed against saline. Vials are then filled with patient vaccine to await release to clinic.
The Committee for Medicinal Products for Human Use adopted a negative opinion for dasiprotimut T for the treatment of Follicular non-Hodgkin’s lymphoma in European Union and requested additional clinical trials. Since we were privately funded, the project was discontinued at the end of March of this year.
The instrument division of Biovest does still exist. And for comparison this is how the bioreactors perform compared to other cell culture methods so you can generate large amount of antibody in a relatively short time in a cost-effective manner compared to other methods.
Once you receive samples you needed to submit a plate map, similar to this one. Indicating where on the 96-well ELISA plate blank, standard curve, controls and samples are located. You also needed to indicated whether any dilutions are necessary. Which can be very difficult especially in actual practice because at best you only have a ballpark for the concentration of the samples. So during the in-process validation where we made the vaccine from beginning to end for 3 different patients, we tracked concentrations and were able to determine set dilutions for the different stages on the vaccine product so the samples would quantitate on the standard curve.