At the end of July, the FDA granted marketing approval to an anticoagulant, the low-molecular-weight heparin (LMWH) enoxaparin sodium injection. The product, co-developed by Momenta Pharmaceuticals and Sandoz, was approved under generics regulations and was designated equivalent to and substitutable for Lovenox, manufactured by Sanofi-aventis. The FDA's decision to consider enoxaparin under the Abbreviated New Drug Application (ANDA) pathway has raised some eyebrows. It not only runs counter to the established European regulatory framework for biosimilars but also suggests that the FDA might, in some cases at least, not require extensive clinical trials for follow-on biologics.

Enoxaparin is not the first generic biologic approved in the United States. Six other generic biologics have been approved under ANDAs, a major advantage of which is that no clinical trial data are required and products can be designated therapeutically equivalent to a brand. Under this path, FDA can also designate a generic as automatically substitutable for the brand in the pharmacy. Of the biologics, only Lovenox and calcitonin generics have been approved under an ANDA and received brand substitutability status. Sandoz's Omnitrope (human growth hormone) was licensed using a New Drug Application (NDA) filed under 505(b)2, a regulatory pathway that allows submission of clinical data demonstrating safety and efficacy but does not offer substitutability.

When Momenta filed its ANDA for enoxaparin in August 2005, the challenge facing the FDA was how to demonstrate its therapeutic equivalence to Lovenox. The problem, ostensibly, is that enoxaparin is complex and difficult to define chemically. Its manufacture involves the alkaline depolymerization of heparin from pig intestinal mucosa. Heparin itself is a mixture of linear polysaccharide chains consisting of repeating disaccharide units composed of glucuronic or iduronic acid and N-sulfated or N-acetylated glucosamine; during its depolymerization to enoxaparin, additional distinctive chemical modifications may occur. The resulting final product has a mean molecular mass of 4.5 kDa but consists of polysaccharide chains varying in length, composition and distribution. The brand drug has never been fully characterized, nor has the contribution of its components to therapeutic efficacy been established.

In information that accompanied the approval, the FDA indicated how it compared the brand and generic versions, an assessment with possible implications for other biologics. The agency's analysis was based on five criteria.

First, the comparison involved gross physicochemical properties (molecular mass distribution and overall chemical composition): the FDA wanted to see that similar oligosaccharide chain lengths were present in the same relative abundance in generic and brand enoxaparin. Second, the drug sources were compared: the source heparin had to have a similar distribution of disaccharide building blocks, and beta-elimination of the heparin benzyl ester had to be shown during depolymerization. A third comparison looked at the products' molecular nature: besides confirming the presence of a pharmacologically important 1,6-anhydro ring at the ends of 15–25% of enoxaparin chains, the FDA looked at the spectrum of disaccharide building blocks and the oligosaccharide lengths and sequences—factors affected by the temperature, depolymerization time and other conditions of preparation.

Besides these physiochemical analyses, the FDA required not only biological and biochemical assay data to demonstrate equivalent anticoagulant activity in vitro and in vivo but also pharmacodynamic data from healthy human volunteers. Taking all the data together, FDA concluded “that generic enoxaparin will have the same active ingredient components as those of Lovenox's enoxaparin (within the context of its variability), even though the contribution of each component has not been fully elucidated.”

As immune reactions, such as pruritus, urticaria and anaphylactic and anaphylactoid responses, have been observed with Lovenox, the agency also requested that manufacturers demonstrate equivalent immunogenicity for generic enoxaparin. Again, it stopped short of requiring clinical studies, accepting data from in vitro and ex vivo assays and from animal studies.

Overall, the generic enoxaparin approval indicates that current analytical technology and integrated, multivariate data analysis can convince the FDA of the equivalence of two complex, biologically derived preparations. The FDA has determined that the product is the same if it meets its criteria of identity (even if the product is not identical)—a substantial departure from European guidelines for LMWH biosimilars designed for products that contain a similar active ingredient. The good news, it seems, is that as long as the FDA is satisfied that the data package sufficiently establishes the sameness of the active ingredient, the need for clinical data diminishes.

The billion-dollar question—as yet unanswered—is whether the FDA will consider similar supporting data for complex biologics approved under the Biologic License Application pathway as sufficient to demonstrate therapeutic equivalence without large clinical trials. As the complexity of the biologic being reproduced becomes greater—from peptides, to hormones and growth factors, and all the way to monoclonal antibodies—the capacity of current technology is likely to approach its limits. In this respect, another generic product may soon provide an answer.

An ANDA for a generic of Copaxone, Teva's treatment for multiple sclerosis, has been before the FDA since July 2008. Copaxone is perhaps the quintessential complex peptide drug, a heterogeneous mixture containing a huge number of synthetic polypeptides. Its analysis will certainly push the envelope for current technology and illustrates how important sophisticated technical capability will be to sponsors wishing to work in this area.

Even the Copaxone case, though, may still not provide much guidance for recombinant biologics. Like enoxaparin, Copaxone's complexity largely stems from the active ingredient. In contrast, variation in most recombinant products—and thus the analytical challenge—arises not in the active ingredient but in post-translational modifications, proteolysis, oxidation and aggregation that occur during manufacture, formulation and storage. All of which adds up to a different challenge again.