Opinion: One Step Forward, Half a Step Back for WHO Biosimilar Guidance

In February 2020, I shared my objections to the World Health Organization (WHO) guidelines for similar biotherapeutic products (SBP). A year later, the WHO proposed to revise these guidelines, accepting several suggestions that I and perhaps others had made but leaving many out that were just as important, which I have described below. Nevertheless, I hope that the WHO will reconsider, because its guidelines are followed worldwide. Furthermore, any flaws in guidance can result in significant safety and efficacy issues with biosimilars.

The WHO suggests that national regulatory authorities (NRAs) decide the prescribing information (PI) to go with the SBPs; this is not correct; the PI should be identical to the product information for the reference product. Both the FDA and the European Medicines Agency (EMA) provide guidance on how to write the PI.

The WHO guidelines continue using the term “comparability” as it pertains to a guideline published by the International Council for Harmonization of Technical Requirements for Registration of Pharmaceuticals for Human Use (ICH). This guidance is known as the ICH Q5E, and many NRAs simply adopt this automatically.

With over 100 biosimilars approved, the choice of critical quality attributes should not be left to the developers but to the regulatory agencies. Any selective choice made by the NRAs can create significant safety problems.

Safety and Efficacy Testing

The WHO suggests that clinical efficacy testing can be reduced based on results of animal testing, but this is not rational; biological drugs act by receptor binding unique to humans. Their toxicity is an extrapolation of their pharmacologic response. Thus, these types of testing are of little value in deciding biosimilarity. Both the FDA and EMA have agreed to exclude animal testing, and the WHO should follow the same path. Furthermore, the WHO suggestion that animal testing should be designed based on shortcomings in structural variability is also faulty; an animal model cannot tell whether a molecule is different.

One piece of advice from the WHO to NRAs is that, to assure that the reference product is safe, they should wait for a few years before allowing SBPs; this view questions the rigor of the regulatory approvals of new products; it also delays the entry of SBPs into the countries that need them most.

The WHO unnecessarily confuses local or nonlocal reference biotherapeutic products (RBPs) and how to match the two; the RBP must be the first product approved in a developed country using the complete dossier; there is no need to confuse the issue.

While in the past, the WHO had declared that no statistical modeling should be done in analytical similarity testing; now it proposes to use mean+/-x*SDoftheRBPbatchdata as the range, basing the choice of the multiplier (x) on the criticality/importance of the attribute. But the multiplier cannot be arbitrary; the FDA learning this the hard way when it was forced to withdraw earlier guidance. The best choice is to use the value of 3, which covers 99% of the range. The WHO further states that if the analytical similarity does not match, then its impact on safety and efficacy should be addressed using functional assays. Unfortunately, the fact is that there is no way to resolve this issue, and the product must be rejected.

The WHO further recommends using in vivo animal studies to resolve relevantdifferences informulation(eg,using excipientsintheSBPnotwidely usedinmedicinalproducts). This is the most irrational suggestion. No animal study can establish the safety of a formulation and its interaction with the active biological entity; a good example was the interaction of erythropoietin with surfactants to cause pure red cell aplasia. Therefore, an SBP should not contain an excipient that is not used in other products.

The WHO suggests that pharmacokinetic (PK)/toxicokinetic (TK) studies in animals should be interpreted by considering neutralizing antibodies formed; the fact is that animal model antibodies have little correlation with human response, and a PK/TK study in animals adds little to ascertain the safety in humans. The only species that might bring some value, but only in the PK studies, are the lower primates, but primate studies are discouraged by the WHO. The toxicokinetic studies are useless because they require a dose range of nonlinear responses, making it impossible to differentiate between an SBP and RBP.

One suggestion of the WHO is that if the SBP contains impurities that are not present in the RBP (eg, because of the use of a novel expression system), the developer should justify it based on additional safety data or a scientific rationale; this suggestion is flawed because, unlike chemical drugs, the toxicology of biological products is related to their pharmacodynamics, which makes such a determination impossible. For this reason, the FDA and EMA discourage using alternate expression systems, even though they are allowed. In addition, no unmatched impurity should be present unless this has already been reported in the reference product's literature.

The WHO takes a nonscientific view of immunogenicity and suggests more investigations if the immunogenicity difference is large; but proving that these differences are not clinically relevant is almost impossible, short of extensive human testing. As suggested by the FDA, differences may be allowed where the immunogenicity does not affect the drug’s disposition, or bodily elimination, profile. Still, if the difference is large in the immunogenicity, the SBP should be rejected, even if there is evidence that immunogenicity does not affect the pharmacokinetics.

Conclusion

The WHO is not a regulatory agency, but its recommendations are widely adopted globally, mainly by developing countries. The new proposal by the WHO to revise its guidelines is a significant step in bringing these guidelines into accord with FDA and EMA recommendations. Still, the WHO guidance continues to miss out on several key concepts of safety and efficacy. For example, the continued insistence by the WHO that animal testing and clinical efficacy studies, both of which are least sensitive, can be used to justify differences in analytical assessment and clinical pharmacology is fallacious. I hope that the WHO will continue to improve its guidelines as they impact most of the world.