CMC Strategies For The Development Of A Bispecific Antibody Platform for Cancer Immunotherapy

This article was created from the author's presentation at the 2018 Bioprocessing Asia conference, with Cytiva as principal sponsor. The BioProcessing Asia Conference series was created to provide a platform to advance the contribution of bioprocessing sciences towards the development and manufacture of affordable biopharmaceutical products in Asia.

Compared to the development of typical monoclonal antibody (mAb) drugs, the development of bispecific antibodies (BsAbs) can present many challenges in product expression, purification, product stability, and scale up of the manufacturing process. Therefore, in addition to efforts made with molecular design, CMC strategies are critical for developing and marketing bispecific drugs.  

A Growing Future For Bispecific Antibodies

Bispecific antibodies is an emerging and potentially highly profitable area of next-generation antibody products. They are engineered immunoglobulins with the ability to bind to two different epitopes on the same or different antigens. Therapy with immunoglobulin G (IgG)-like BsAbs shows active immunization against specific tumors, which in the future could lead to increased use of this format to provide long-lasting antitumor immunity in the organism.¹

There are currently more than 200 BsAbs in clinical trials, with two already available on the market (Hemlibra and Blincyto), bringing in total sales of nearly $1 billion in 2018.² A recent analysis predicts the estimated market opportunity for BsAbs to surpass $8 billion by 2025.² The potential of BsAbs and their promise in cancer immunotherapy is garnering attention from both big and small pharma companies. Wuhan YZY Biopharma, located in Wuhan, China, has two bispecific candidates in Phase I clinical studies and several other candidates in the pre-clinical or discovery stage. Its lead product candidates were developed using YZY’s proprietary bispecific antibody platform, YBODY^®, which uses engineered Fc to promote heterodimer formation. The potency of the antibody can be greatly enhanced by recruitment of T cells.

YBODY^® Bispecific Platform 

YBODY^® is composed of two parts, the half antibody and the single chain for CD3 binding. The half antibody can interact with a tumor-associated antigen, and the single chain can interact with the CD3 on T cells. YBODY® has an IgG-like structure, so it can provide good pharmacokinetics and pharmacodynamics (Figure 1). The knob-in-the-hole and salt-bridge methods were used for production of this antibody, and the CD3 single chain format was used for T cell redirection. The YBODY^® platform currently has a total of nine patents, with four in the U.S.

The images below show two characteristics of the YBODY^® platform. Figure 2 shows that the green fluorescent protein (GFP)-labeled tumor cells are surrounded by lymphocytes when the antibody YBODY® is added into this mixture, indicating YBODY^® can mediate tumor T cell interactions. The high-performance liquid chromatography size-exclusion (HPLC SEC) chromatogram in Figure 3 shows the product can achieve greater than 99 percent purity.  

A computer-modeled 3D structure of YBODY^® is shown in Figure 4. In the single chain, the variable region of the heavy and light chains are combined through a flexible linker.  This domain formed an antigen binding surface, and the full YBODY^® forms an IgG1-like structure.

This molecule can arm the immune cells through the interactions between the single chain and the antigens on the immune cells. Then, the armed immune cells can be redirected onto the tumor cells through the interactions between the half antibody and the antigen on the tumor cells and then attack and kill the tumor cells.

Overcoming YBODY^® Manufacturing Challenges

YBODY^®, like other IgG-like bispecific antibodies, can present several manufacturing challenges, including:

• low expression titer of the target bispecific
• purification difficulties
• instability of the intermediates and the product
• issues in process scale-up.

Several methods can be used during molecule design to overcome these challenges, including mutation of amino acids to decrease aggregation susceptibility, mutation of instability hot spots in CDRs, and selection of Gln residues as the amino-terminal to force pyroglutamylation to decrease the number of charge variants. ³ However, it is also imperative to implement effective CMC strategies, as these will have the most impact on a company’s ability to successfully develop and market bispecific drugs.

The CMC strategies employed by YZY included evaluating the stability of the candidate molecules, choosing clone expression bispecifics with high titer and purity, tailoring purification methods fit for the molecule characteristics, and using sustainable scale-up strategies for large-scale manufacturing.

Evaluating The Stability Of Candidate Molecules

The stability evaluation was performed at a very early stage of process development. Generally, three buffer systems are used for this evaluation. The quality change of the candidate was monitored over the course of two weeks under stress conditions (40 degrees Celsius). Figure 5 and Figure 6 summarize HPLC-SEC and binding affinity results for the stability evaluation of a candidate molecule.

The purity of the candidate (Figure 5) and the binding affinity of the candidate molecule to CD3 and a tumor-associated antigen (Figure 6) did not change significantly at 40 degrees Celsius within two weeks. The results of the stability evaluation provided the confidence to proceed with process development of this molecule.

Choosing The Right Cell For Bispecific Production

At the process development stage, it is important to choose the right cell for bispecific production, as different cell lines may express bispecifics with varied expression and purity levels. For YBODY^®, CHO cells were used as the host cells for expression, with two vectors for expression of the half antibody and the single chain, respectively. Vector A contained a puromycin resistant gene for HL expression, while vector B contained a hygromycin resistant gene for single chain expression; both vectors also contained a dihydrofolate reductase selection marker.

A two-stage pool selection strategy was used to balance the HL and single chain expression before clone selection (Figure 7).

At stage one, a constant selection pressure of 5 micrograms per milliliter (ug/mL) of puromycin and 600 ug/mL of hygromycin was used, resulting in a purity of about 40 percent. In the second stage of pool selection, the half antibody expression and the single chain expression were balanced by increasing the expression of the single chain via increased selection pressure of hygromycin. The bispecific purity of around 88 percent was achieved using 2,400 ug/mL of hygromycin plus 600 ug/mL of methotrexate, but without puromycin, which is a significant increase comparable to its parental pool.

Clones that can produce bispecific antibodies with acceptable purity are then selected from the pool. Figure 8 summarizes the clone selection results for a typical YBODY^®.

As indicated by Figure 8, the titer ranged from 0.2 to 1 gram per liter (L) before process development. The purity ranged from about 20 to 90 percent for different clones.

Because the goal is to manufacture YBODY® at a large scale, it is important to ensure the clone has acceptable stability. Figure 9 shows a clone stability evaluation for a leading clone. The top graph shows the expression titer of the leading clone did not change with the increase of the population doubling level (PDL) to 70. Also, the target gene copy number for the heavy chain, light chain, and single chain did not decrease with increase of the PDL.

The cell growth performance for this leading clone showed its viability and viable cell density did not change significantly with increase of the PDL (Figure 10). The results of the clone stability evaluation indicate the leading clone is suitable for large-scale manufacturing.

Using Sustainable Scale-Up Strategies For Large-Scale Manufacturing

The scale-up strategies for YBODY^® are straightforward. For the upstream process volume-dependent parameters, the volume can be proportionately scaled up. The same setting points are maintained for volume-independent parameters, such as the inoculation density, pH, dissolved oxygen, and feed regime. For the impeller agitation rate, an equivalent P/V is used across scale. For air/oxygen sparging, air sparging is proportionally increased and oxygen is used to control dissolved oxygen. For the downstream process, a constant bed height and linear flow rate were maintained for the column steps, and the column diameter was increased. A constant flux/pressure was maintained for the filtration steps but increased the membrane area at the larger scale.

The YBODY^® is manufactured using a platform process. Figure 11 shows that the production titer and purity by HPLC-SEC of a typical YBODY^® at scale of 200 L and 2 L are very comparable. Figure 12 shows the YBODY^® bispecific has similar charge variants and glycan forms at 2 L and 200 L. Figure13 shows the reproductive production titer of a typical YBODY^® from three 200 L batches.

In summary, manufacturing challenges for IgG-like bispecifics include low expression titer of the target bispecific, difficulty in purification, instability of the intermediates, and the product issues in process scale-up. However, by using molecular design, stability assessment optimizing clone selection, tailored purification method fit for the characteristics of the molecule, and a suitable scale-up method, a high-titer, scalable process for the YBODY^® bispecific antibody can be achieved.

1. Sedykh, Sergey E. et al. (2018, January 22). Bispecific antibodies: design, therapy, perspectives. National Center for Biotechnology Information. 12: 195–208. Retrieved from https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5784585/
2. Medgadget (2019, January 16). Global Bispecific Antibody Market, Drug Sales, and Clinical Pipeline Insight 2025. Retrieved from https://www.medgadget.com/2019/01/global-bispecific-antibody-market-drug-sales-and-clinical-pipeline-insight-2025.html
3. Beck, A., Wurch, T., Bailly, C., et al. Strategies and challenges for the next generation of therapeutic antibodies. Nature Reviews Immunology, 10, 345-352 (2010).

Jianzhong Hu is the director of process development at YZY biopharma. He has nearly ten years of experience in biopharmaceutical process development, tech transfer, and GMP manufacturing. At YZY biopharma, Jianzhong has led cell line, upstream, downstream, and formulation development of innovative bispecific antibodies from pre-clinical stage to clinical stage. Before joining in YZY biopharma, he was the associate director of biomanufacturing at WuXi biologics, where he led tech transfer and GMP manufacturing operation for more than ten biopharmaceuticals from Phase I to commercial launch. Jianzhong Hu got his Ph.D. in biological systems engineering at Virginia Tech, USA and his M.S. and B.S. degree in biochemical engineering at Tianjin University, China. He has published eight peer-viewed journal papers.