Global Perspective on the Effect of Biosimilars - Episode 3
Hope S. Rugo, MD: The approval process for biosimilars is really interesting. People usually show a slide that has a triangle that is upside down, so there’s a point, and the big part of the triangle is at the top. It’s like an upside-down triangle. The bottom sort of represents that you’re developing the biologic agent and are doing some purification processes. Of course, that takes a lot. It looks like it’s the plane of a triangle, but it’s a huge amount of work, right?
Then you go to the meat of the whole matter, which is showing that these agents are safe and effective in a clinical setting. Not only do you have to show your original clinical setting, but then you have to test it in other settings. A nice example of that is with trastuzumab. It took a long time to develop trastuzumab and show that it was safe. You started doing clinical trials. You gave it in 1 trial with 1 combination. Another showed that the survival was better. You moved it into early-stage cancer and gave it in combination with different targeted agents. We’re still using trastuzumab as a backbone for studies looking at ways to expand on treatment options and improve outcomes for patients with [human epidermal growth factor receptor 2, HER2]—positive breast cancer.
So the clinical part is the biggest part of the triangle, because it lasts for a long time. The drug is tested in many different ways. That’s a reference biologic.
When you get to the biosimilar, the triangle is reversed. That reference biologic has done all of the work at the top of the triangle. The biosimilar is really showing that it’s biologically similar. So the triangle is reversed. You need to show that it’s clinically similar. Here, it’s really sitting on top of a much bigger weight of evidence that shows that it’s similar in the laboratory, as well as biologically.
You have to show that it has a similar structure, which is a complicated process, because these are antibodies made in living cells. You have to show that there is similar biologic characterization. The caveat to that is that we know that any biologic test that you do in vitro doesn’t correlate very well with what the agent does in vivo, in a patient. There are still tests that are done. One example is that antibody-dependent cell-mediated cytotoxicity, or ADCC, is something that can be done in a laboratory. You can use that as a way to compare and contrast the reference and the proposed biosimilar to see if there are major differences, but you know that’s not the end-all, by any means.
Then you look at a lot of other structural characteristics. You look to see, for example, if there’s a difference in the number of sugars. It turns out that the differences are mostly in the amount of glycosylation, which doesn’t seem to make a big difference, actually. In animal models, you look at pharmacokinetics, pharmacodynamics, and immunogenicity. You can imagine that the bottom of the triangle is really the big part where you’re doing lots and lots of studies—many more than the number that were required for the reference biologic to show that it’s biologically similar.
And then at the very top of the triangle are clinical studies. The idea here, as defined by the different regulatory agencies and the World Health Organization, is to show that there are no biologically meaningful or clinically meaningful differences between the reference product and the proposed biosimilar. You have to have shown all of that sort of prework, such as working in the laboratory or clinic.
What’s interesting is that through the initial work of figuring out how to do the clinical study, an idea came about: If the reference biologic was made in the United States versus in Europe, did you need to show whether or not they were different products?
It turns out that when you move the production of a reference biologic or you simply change the production process, you can change the reference biologic enough so that it is actually as much changed as a biosimilar would be. Where you make the biologic and how you make it really makes a very big difference when you’re making something in living cells.
One of the things that needed to be done, for example, for the biosimilar trastuzumab—not in people with cancer but in healthy subjects—was that you had to show that the pharmacokinetics and pharmacodynamics were similar. Obviously, through very short exposure, they compared the proposed biosimilar with the European and United States trastuzumab because the 2 of them could be different. This is very, very interesting because we hadn’t thought about that. Before there were biosimilars in the clinic, we never thought about that.
There’s also been some really interesting work that has shown that the potency of ADCCs may vary. We don’t know if it affects the clinical efficacy, but from lot to lot of produced trastuzumab, there were variations in ADCC and the reference biologic, which is also quite interesting. It has been looked at in some of the reference biologics used as a control in the biosimilar trial, so that’s quite intriguing. Basically, you have to show that all of these things that are in the preclinical area and in healthy individuals are the same.
Then you get to the actual efficacy and safety in patients. The regulations have noted that you want to study your biosimilar in a clinically sensitive area. For example, you would go back and look at the development of that reference biologic and you would say, “Where have we seen a big, reproducible impact from that agent that we would trust to actually tell us that the agent worked?”
That’s brought about some controversies. It’s differed based on the agent, etc. But in general, you pick a sensitive area—the indication for the drug. Then you have to pick an endpoint. For most of these reference biologics, when they got approved, they need to show an endpoint of disease control and, for many of them, overall survival, if they’re cancer drugs. Things are a bit different in the rheumatologic area and for supportive care, where we also have biologics.
We talk about cancer drugs. They’ve really been showing efficacy. Many of the reference biologics have actually shown survival differences. If you made the proposed biosimilar do the exact same thing, you would never get them on the market. That’s part of this whole idea of competition. You want to use a short-term clinical endpoint that correlates with the long-term end point, and you want to use that for your approval.
Response rate has been used as a short-term clinical endpoint with the idea that, in most cases, although you’re not obliged by regulatory guidelines, you would look at the longer-term endpoints over time, if you’re able to. But every time you extend out your study you add cost, right? So if you’re going to follow people for 10 years, that’s going to be a big cost added to the development of that agent.
Then you have the potential approval, right? You’ve done your clinical study. You’ve shown that your short-term end point is similar. Maybe it’s the response rate, for example, for a cancer agent like trastuzumab in the neoadjuvant setting or in the first-line metastatic setting. Then the question is, what do you do with that data? That brings up 2 other areas where the regulators have had comments, and 1 is interchangeability and extrapolation.
We talk about extrapolation as the idea that if these biologic agents worked just as well in combination with different chemotherapy drugs, hormone agents, other antibodies, or different oral biologic targeted agents like tyrosine kinase inhibitors, you could extrapolate from your approval to all of these different combinations. If it worked just as well in the metastatic and early-stage settings, you could also extrapolate freely between the 2 settings. And then, lastly, if it works for more than 1 disease, like trastuzumab for breast cancer or gastric cancer, you could extrapolate between those indications, as well.
So that is the entire approval process, from beginning to end. And right now, in terms of what’s been approved so far, the United States FDA and [European Medicines Agency, EMA] have approved these agents with the idea that extrapolation will be used.