October 15, 2012 (Vol. 32, No. 18)

Michael Davies
Xing Wang, Ph.D.

Comparability Testing Platform for Biologics at Molecular Level Now Available

Recent healthcare legislation in the U.S. has created a pathway for biosimilar approval and commercialization, potentially unleashing a wave of competition to the current class of blockbuster biologics now on the market. However, reconciling the development of these biosimilars with the innovator biologics they are designed to replace has created numerous challenges, a situation which might be summed up as “how similar is similar enough?”

In an August 2011 New England Journal of Medicine Perspective, and draft guidance issued in February 2012, the FDA outlined the challenges facing the biosimilars approval process and suggested that a “meaningful finger-print-like analysis” would streamline the process and speed the approval process. In this tutorial we suggest one method of accomplishing the fingerprint-like analysis that the FDA has proposed.

Over the past three decades, biologics (therapeutic proteins) have made up an increasing volume of pharma sales. Within the next decade, patent expiration will affect many of these innovator biologics and both generic and innovator manufacturers are gearing up programs to produce biosimilar and biobetter drugs.

The term biosimilar is applied to products that have been shown to be similar to the innovator biologic through head to head tests of quality and appropriate comparative studies. If these criteria are met, then the biosimilar can undergo an abbreviated pathway for approval under the Biologics Price Competition and Innovation (BPCI) Act of 2009.

Biobetter drugs are intended to be superior to the innovator product and as such are considered new molecular entities and must go through the full development and approval process. For the remainder of this article, we focus on biosimilar drugs and the innovator biologics that they seek to replace.

Unlike generic small molecule drugs, biologic drug production is complex, meaning that biosimilars will always be different from the original innovator drug. Even if the biosimilar uses the same gene as the innovator, differences in production, including cloning vector, expression system, fermentation, and purification will generally always result in a biosimilar that is slightly different from the original. The question facing the FDA and biosimilar applicant is “how close is close enough?”

While the BPCI Act outlined a pathway for biosimilar approval in the U.S., it left unanswered many questions surrounding the specific scientific quality criteria addressing how “similar” a biosimilar should be to the innovator drug in order to be approved. Draft guidance documents further clarifying this were issued in February 2012, with the FDA suggesting that a “meaningful fingerprint-like” comparison of a large number of product attributes in the innovator and biosimilar products would be very helpful in streamlining the approval process.

This appeared to be a worthy goal, provided an applicant was able to recognize which product attributes were most critical to compare to a safe and effective biosimilar drug.

Biologic “Fingerprinting” Using the PCA-ELISA

At Array Bridge we asked ourselves this same question and concluded that a biologic’s three-dimensional (3D) epitope distribution would be a sensitive marker for structural changes in the molecule. Furthermore, if an easy method for following the 3D epitope distribution existed it would provide a powerful tool for “meaningful finger-print-like” comparisons of biosimilars and innovator drugs.

With this in mind we have developed a series of Protein Conformation Array ELISA (PCA-ELISA) kits for three-dimensional structural comparability analyses of biologics and biosimilars. These PCA-ELISA kits can provide valuable information on the 3D structure and heterogeneity of biologics and can be used at many stages of biologics/biosimilars development including cell-line selection, process development, formulation development, and product release testing.

The basis of the kit is a series of polyclonal antibodies to overlapping peptides spanning the entire length of the amino acid sequence of the biologic (Figure 1). When used in an ELISA format, with a separate family of polyclonal antibodies in each well of a 96-well plate, one can interrogate the entire surface of the biologics.


Figure 1. Schematic of PCA-ELISA

One key feature of the system is that, in its native form, few of the epitopes will be exposed on the surface of the biologic, but if the higher-order structure of the biologic changes slightly, additional epitopes will be exposed, resulting in a signal increase in the well containing the antibodies to that particular epitope.

Effectively, the kit provides a “fingerprinting” technique for the native biologic that is also primed to detect very small changes in structure due to the array of antibodies made to all the buried epitopes. It is highly sensitive to changes in structure or denaturation of the protein and is able to detect as little as 0.1% denaturation of a protein sample (Figure 2).

The assay is a robust sandwich-type ELISA and other than a colorimetric plate reader and multichannel pipettes, no specialized lab equipment is needed. Each assay kit consists of three 96-well plates coated with an array of 30–31 polyclonal antibody families, distributed column-wise across the plates with each polyclonal family represented six times on the plates.

One biosimilar can be compared to innovator in triplicate, or two biosimilars can be compared to innovator in duplicate.


Figure 2. Detection of new epitope exposure in a forced degradation study

An Early Predictor of Changes in Higher Order Structure?

In another example (Figure 3—top panel) we show the conformational array ELISA result of seven commercially successful monoclonal antibody drugs, with only the Fc (constant region) results depicted. The point to emphasize here is that the results are fairly similar across all seven biologics, as would be expected since this region is common among all antibodies in this class.

Compare this result to the lower panel of the figure that shows three candidate monoclonal antibody drugs that all failed in clinical trials. It becomes obvious that two of these candidates showed significant additional epitope exposure to Ab17 and Ab18 (near the hinge region), while the third showed significant epitope exposure to Ab23 and Ab24 (near the glycosylation site). Whether this additional epitope exposure actually caused the clinical trial failure is not yet understood, but the linkage is intriguing and we feel warrants further study.


Figure 3. Comparison of mAb higher order structure

Conclusions

Biologics and biosimilar drug development is likely to increase greatly in the coming years, both due to the promise that biologics bring to unmet medical needs and the price competition that biosimilars bring. As the FDA suggests, having “fingerprint-like” analysis tools will streamline development and improve safety; we believe that the technology offered by PCA-ELISA is one way of achieving this.

Kits are currently offered for seven of the most popular mAb-based innovator biologics (Rituximab, Trastuzumab, Bevacizumab, Adalimumab, Cetuximab, Alemtuzumab, and Palivuzumab, as well as a kit designed for novel mAb-based biologics). In each case the kit is designed to perform a PCA-ELISA comparison between one innovator standard and one biosimilar sample (in triplicate), or between one innovator standard and two biosimilar samples (in duplicate).

In each case the customer will need to supply samples of the biosimilar samples and innovator drug being compared to. Array Bridge can perform the PCA analysis in-house from supplied samples, or can supply kits for customer use.

Michael Davies ([email protected]) is senior scientist and Xing Wang, Ph.D., is president at Array Bridge.

Previous articleEnter the GEN Ten Survey for Your Chance to Win an iPad Mini!
Next articleCE Mark for Rapid Acute Kidney Injury Test