May 15, 2015 (Vol. 35, No. 10)

Rong-Rong Zhu Senior Scientist EMD Millipore
Jennifer Campbell Director of Worldwide Biosimilar Market EMD Millipore

The Crucial Role that Analytics Play in Product Comparability, Continuous Quality, and Assessment

As more biosimilar molecules are approved for use, development and manufacturing processes are becoming more defined. Success depends on the ability to characterize the biosimilar and demonstrate its similarity to the originator molecule.

One key component in biosimilar development is analytics, which are used to demonstrate comparability of the biosimilar to the originator. Having a closer structure and functionality to the original can lower the potential requirement for costly clinical trials.

Analytics are also used to establish critical quality attributes (CQAs) of the biosimilar molecule, enabling the development of a well-controlled, robust manufacturing process that helps ensure product safety and efficacy.

Here we describe how analytical methods are applied in upstream process development and the impact of process changes on the CQAs of monoclonal antibodies (mAbs). Our studies show that thorough characterization plays a key role in bringing biosimilars to market efficiently and safely.

Assessing CQAs

One challenge in comparability analysis of biotherapeutics is their large size and potential degradation pathways. Production from living cells adds further complexity to the development and characterization. In manufacturing, product quality encompasses physicochemical properties of products and impurities. There are product-related impurities, such as aggregates, and process-related impurities, including cell culture media components, host cell proteins, DNA, and leachables from processing equipment.

Orthogonal analytical methods are critical to fully characterize biosimilars through the development process, and confirm the manufacturing process did not change the critical attributes of the molecule. Comparability testing of pre- and post-process changes is an essential step to demonstrate that a process change does not affect product quality, safety, and efficacy.

Determination of product comparability is based on the combination of analytical and biologic test results, with the goal of understanding the correlation between process changes and product quality attributes. Understanding the link between process and product quality helps result in better manufacturing processes by enabling the identification of critical process parameters and their impact on CQAs.

Cell-Line Selection

From the early stages of process development, analytics help determine the optimal cell line and clone, through screening the impact of various cell lines and clones on the resultant molecule. For example, it is well known that glycosylation can affect protein stability, half-life in serum, drug efficacy, and immunogenicity.

Differences in glycosylation profiling are often due to upstream process changes such as cell lines, cell culture media, bioreactor scaling, and bioreactor conditions (pH, temperature, feeding, and harvest time). The glycosylation profile is considered a CQA, and it should be monitored and quantified to ensure product quality and manufacturing process consistency.

Our studies compared the glycan profiles of a mAb generated from either CHO-S or DG44 cell lines, and found significant differences (data not shown). The glycan profile of mAb04 produced in CHO-S is 85.3% (G0F), 9.7% (G1F), and 4.9% (G2F). However, the G0F, G1F, and G2F of mAb04 generated from DG44 are 42.2%, 40.2%, and 17.5% respectively. Therefore, mAbs produced in CHO-S cell lines would not be considered a biosimilar to those generated in DG44 cell lines.

After cell line and clone selection, analytics are used to determine the optimal cell culture media for the clone expression. Basal cell culture media can affect cell growth, protein titer, and some product quality attributes. For the biosimilars manufacturer, optimal media can ensure higher similarity in glycan profiles.

Deamidation, oxidation, C terminal lysine, N terminal cyclization, and sialylation are some of the common charge variants for mAbs. Changes in the upstream and downstream process can result in changes in the charge profile, which can have significant impact on drug efficacy and stability.

Accurately monitoring the charge distribution of biopharmaceuticals is one of the most important steps to ensure biosimilars are the same as originator molecules. Weak cation exchange HPLC and Image cIEF are the most common methods used to monitor charge variants.

We produced mAbs in two different basal media (S and M), under similar upstream process conditions. Our results showed no difference in aggregates and fragment profiles when the basal media changed (data not shown). We also found no significant change in acidic peak populations.

There was a slight increase in basic peak 2 from 11.9% in basal media M to 16.7% in basal media S (data not shown). However, basic peak 2 is due to C terminal lysine variation of the heavy chain, and it is not considered a CQA.

In contrast, oligosaccharide distribution is very sensitive to cell culture media changes. Figure 1 compares the oligosaccharide profiles of a mAb produced in basal media M versus basal media S; significant changes in glyco profiles are observed. The glycan distribution of mAb05 in basal media M is 79.0% (G0F), 19.2% (G1F), and 1.9% (G2F). Comparatively, the glycan composition of mAb05 produced in basal media S is 50.1% (G0F), 42.7% (G1F), and 7.2% (G2F) instead.

It is critical to understand the correlation between glycan profile distribution and basal media selection during early biosimilar process development stages to ensure the biosimilars have similar glyco profiles to the originator molecules.


Figure 1. Glycan Profile (via 2AB-UPLC) of mAb05 produced in either basal media M (A) or basal media S (B).

Impact of Harvest Cell Viability

It is well known that cell viability can affect the quality and quantity of host cell proteins. A study was performed to investigate if harvest cell viability can affect product quality as well.

Three lots of mAbs were manufactured with the same cell lines and cell culture media, but harvested at three different cell viabilities: 98%, 75%, and 50%. No significant difference was observed on aggregation and fragmentations, but differences were observed in the charge profiles.

As shown in Figure 2, the lower the harvest cell viability, the lower the percentage of the basic peaks. The basic peaks are due to C terminal lysine variation of heavy chains, which are cleaved by carboxypeptidase B, one of the host cell proteases in CHO cells.

At lower cell viability, more carboxylpeptide B is released in the soluble supernatant of the cell culture and results in less basic peaks. These data clearly demonstrate that a shift in cell culture viability at harvest can have a profound impact on the resultant charge profile.

High-quality analytics not only demonstrate comparability of a biosimilar molecule to the reference medicinal product, but also enable process development and the establishment of biosimilar CQAs. During process development, analytics are essential to understand the resultant molecule produced as different parameters are investigated.

Changes in cell line and cell culture media have a marked impact on mAb glycan profiles. In addition, shifts in harvest cell viability result in differences in the acid and basic peak populations. Thorough analytics provide a better understanding of these process parameters, enabling development of a well-controlled process.

As analytics increase in their ability to discern subtle molecular changes, the application of advanced analytics can provide a higher degree of patient safety.


Figure 2. Charge variant profile (via WCX-HPLC) of mAb05 harvested at different cell viabilities. Chromatographs off-set for easier comparison.

Rong-Rong Zhu ([email protected]) is senior scientist and Jennifer Campbell ([email protected]) serves as director of the worldwide biosimilar market for EMD Millipore.

Previous articleHuman Longevity and Cleveland Collaborate on Heart Disease Genomics
Next articleBlood Holds Clues to Breast Cancer Risk