Selecting the Most Sensitive Population for Confirmatory Trials

Gillian Woollett, MA, DPhil: So it’s fundamental. I think the greatest value for sponsors is this ability to extrapolate between indications once they’ve established high similarity. We often hear this discussion, and Carlos, I’m going to ask you this question about the most sensitive population, that you are only studying the biosimilar limited confirmatory clinical studies in the most sensitive population. Can you help us understand what the most sensitive population means and how you choose which studies you do clinically for a biosimilar?

Carlos Sattler, MD: Sure. I just wanted to highlight a couple of things about extrapolation and I'll get to the question. It will lead to that. I want to emphasize a couple of things. Extrapolation has to be scientifically justified by the sponsor. It’s not automatic. And as Gillian mentioned, it is based on the totality of the data that show that a biosimilar is highly similar to a reference product and that you don’t expect any clinically meaningful differences between one and the other. Other aspects also taken into account when determining extrapolation, for example, include the mechanism of action. But the essential component here is that if the molecules are highly similar to the biosimilar reference product, then they should behave the same across all indications. That is number 1.

Number 2: I think it is absolutely understandable why clinicians would feel uncomfortable prescribing a biosimilar if there are no clinical data because, as Gillian mentioned, they are used to clinical data. That’s how they make determinations. So here it is incredibly important that providers, clinicians, prescribers understand the biosimilar concept because without understanding it and understanding how that leads to extrapolation, there will be this discomfort and resistance.

I also agree that the FDA is in a really special position, because they granted the extrapolation of indications, to be able to educate about that.

I think the last thing I’ll say and I’ll get to your question, I promise, is that for the biosimilar paradigm to be successful and for sponsors of biosimilars to be able to bring these products to market at a lower cost, the development costs have to be lower. And with new biologics, the majority of development costs are spent in clinical trials, and we’re talking about sometimes hundreds of millions of dollars, depending on the product and the size of the trial. By scientifically justifying extrapolation, then a manufacturer can do a study in 1 population, and this is where we’re getting towards the most sensitive population. The Sandoz approach to this is we try to identify a population where a few things happen.

Number 1: We see the greatest effect of the particular biologic on that particular disease stage, so the biggest delta. And that allows us to see that if there are any clinical differences between the biosimilar and the reference product, they will be identified. So that’s one concept.

Secondly, we’re looking at populations that are not highly variable. So you can actually, when you measure the impact, attribute it to the biosimilar or the reference product. And we’re also looking at populations that are not necessarily using concomitant medications that could obscure some of the measurements that you’re looking at. So, for example, patients who are receiving other immunosuppressants may be difficult to evaluate immunogenicity on, or compare the immunogenicity between products, between the biosimilar and the reference product. So that is how we choose the most sensitive clinical population in order to determine biosimilarity. But it’s also important to note that is the last step in biosimilar demonstration.

It hinges on having demonstrated that, analytically, the reference product and the biosimilar are highly similar, that it’s a structural and functional perspective, that you’ve demonstrated, if applicable, in preclinical models that there’s no difference from a clinical performance perspective, and that you’ve evaluated bioequivalence to pharmacokinetic, pharmacodynamic evaluation in patients or in human volunteers, depending on the situation. So this is the last step in the biosimilar development paradigm, if you may, and hence, why the ability to extrapolate from that indication hinges on all of the other information and data.

Gillian Woollett, MA, DPhil: I think you’re making a good point there. These are confirmatory studies.

Carlos Sattler, MD: Correct.

Gillian Woollett, MA, DPhil: And if you haven’t established high similarity analytically in the first place, you can do all of the clinical studies in the world and you’re never going to be a biosimilar. So I think it’s very important that the clinical studies address what is variously called “residual uncertainty.”