Despite these uncertainties, the first step for manufacturers is to develop a clear analytical plan for demonstrating biosimilarity
of a follow-on to a reference product. A "step-wise" approach for doing this was laid out by CDER officials at the DIA meeting
and subsequently by Steve Kozlowski, director of CDER's Office of Biotechnology Products (OBP), at the Generic Pharmaceutical
Association's (GPhA) fall 2012 conference. Kozlowski observed that manufacturing and product analysis are usually at the
"low end of the totem pole" in drug development, but that a "totality of evidence," starting with structural and functional
characterization, provides a foundation for determining the scope of preclinical and clinical studies for a biosimilar. "Once
you know how close you are," Kozlowski explained, "the rest follows."
To make the most of an initial advisory meeting with FDA, sponsors should have a clear rationale for product development,
with early characterization data for the proposed biosimilar and reference product lots and justification for as much of the
analytical approach as possible. In-depth characterization assay development is desirable, as are preliminary analytical functional
similarity studies and formulation studies. It also may help to have validated research and stability assays to support an
"Know your protein" is a main FDA theme for biosimilar development. Selection of analytical test methods should be based on
the nature of the protein and knowledge of its structure. Agency staff wants to know which attributes are important and how
the relationship between protein attributes and the clinical safety and efficacy profile can predict "clinical similarity."
Also important is whether differences between a chosen expression system and that of the reference product will significantly
affect process- and product-related substances and impurities, how differences in the impurity profile or in excipients will
affect safety, and the strengths and weaknesses of each analytical method.
FDA's draft guidance on quality considerations for biosimilars describes a broad range of analytical studies that may be relevant,
including assessment of expression system, physicochemical properties, functional activities, receptor binding, immunochemical
properties, impurities, reference product, and stability. Protein evaluation stands to benefit from a growing number of analytical
tools, including mass spectrometry, peptide mapping and chromatography to assess amino acid sequence, protein folding, subunit
interactions, heterogeneity, glycosylation, PEGylation, bioactivity, and other methods.
Advances in manufacturing science and adoption of quality-by-design approaches may support comparative assessment using a
fingerprint-like analysis or "super characterization" approach that involves evaluating combinations of attributes using orthogonal
methods. It is not clear whether biotech manufacturing processes are too variable to allow for a fingerprinting approach,
Kozlowski explained at the GPhA conference, but he anticipates that upfront efforts to select appropriate cell lines will
permit manufacturers to deliver an exact desired product.
Transparency also is important. It is better for a manufacturer to acknowledge uncertainties from test results and to seek
advice from FDA for dealing with these issues than to ignore or hide discrepancies. FDA advises manufacturers to carefully
consider the importance of differences in expression systems and the need to justify minor modifications in an amino acid
At the same time, FDA is trying to be flexible, Kozlowski emphasized, noting that differences in formulation from the reference
product may be acceptable, as well as alternative delivery devices or container-closure systems. Applicants may market a proposed
biosimilar for fewer than all conditions of use and presentations for which the reference product is licensed.
"The more analytical data one has up front, the more targeted the rest of development can be," according to Koslowsky. Although
FDA expects that sponsors will have to conduct at least one immunogenicity study, Koslowski says that "the door is open,"
to the concept that "added studies may not be necessary."
Jill Wechsler is BioPharm International's Washington editor, Chevy Chase, MD, 301.656.4634, firstname.lastname@example.org