Highly Similar

Last month, I noted that the patent provisions in the US healthcare reform legislation create a significant disincentive for companies to file for the approval of biosimilars. Should companies indeed file, however, by what standards will their applications be judged?

Laura Bush

According to the new legislation, a biosimilar product must be "highly similar" to the reference product. The term is not defined in the bill, so the FDA's interpretation will be critical. Innovators may hope that a very high hurdle is established. They should be careful what they wish for, however.

For example, must a biosimilar and its reference product have identical amino acid sequences, including irrelevant residues? If a company wanted to gain approval of a biosimilar monoclonal antibody, would it have to forego current technology that enables the production of fully human MAbs, and develop a mouse or chimeric MAb that more closely matches the older reference product? Of course, some variation in glycosylation still would be expected, even if within the range of variation seen in the innovator product or shown not to be clinically relevant. Would that prevent the biosimilar from being approved?

If the standards for similarity are set to an extreme, they could be applied to innovators, too, when they file comparability protocols for manufacturing changes. If we define "comparable" essentially to be "indistinguishable," manufacturers will be trapped in legacy systems, afraid to seek regulatory permission to update processes lest they be held to an impossible standard of comparability. Likewise, companies will be trapped in legacy knowledge, reluctant to use new analytical technologies that tell us more. Continuous improvement and innovation in manufacturing will become impossible.

That is not a desirable path. Further, it is at odds with the FDA's recent efforts to improve manufacturing knowledge and quality and to increase opportunities for continuous improvement through the process analytical technology and Quality by Design initiatives.

To establish a framework for comparing biosimilars to their reference products, the FDA should refer to the only other guideline that uses the term "highly similar," the ICH Q5E guideline on comparability following manufacturing changes. Q5E defines comparable as:

A conclusion that products have highly similar quality attributes before and after manufacturing process changes and that no adverse impact on the safety or efficacy, including immunogenicity, of the drug product occurred. This conclusion can be based on an analysis of product quality attributes. In some cases, nonclinical or clinical data might contribute to the conclusion.

ICH Q5E was adopted in 2005, after a full public comment process, and the FDA has extensive experience applying it. More importantly, it makes sense to apply one standard for all comparability exercises for biologics, be they biosimilars or manufacturing changes at a single manufacturer. A single standard does not preclude a case-by-case evaluation. For each biosimilar application, the FDA will still have to apply the standard and see if the data measure up.

From the case of Genzyme's Myozyme, we know that the FDA examines comparability very carefully. So there is no reason to fear that applying the comparability standard used for manufacturing changes will create an easy test for biosimilars. Rather, it will establish a level playing field for biosimilars and innovator products, provide a consistent, data-driven, and science-based rationale that protects the FDA as well as all sponsors, and will keep open the door for manufacturing improvements.

Laura Bush is the editor in chiefof BioPharm International, lbush@advanstar.com