Dr. Niazi, Adjunct Professor of Pharmaceutical Sciences at the University of Illinois and the University of Houston, founded the first US biosimilar company, Therapeutic Proteins (later Adello and Kashiv), and is the founder of Novel351k, as well as a biosimilar advisory company PharmSci. In part 2 of this two-part interview, we talked with Dr. Niazi about the role of artificial intelligence in clinical trials and his views on how the Inflation Reduction Act can improve the prospects for future biosimilar development.
Biosimilars Review & Report: I read a very interesting paper you wrote about artificial intelligence’s role in streamlining the clinical process. Can you give us a little bit of background on that?
Sarfaraz Niazi, PhD: Yes, that’s one of the most exciting developments. About two years ago, Google introduced an AI program called AlphaFold2.
Every time a new batch of reference product or the biosimilar is produced, it can have slight structural differences. But there are two types of differences. One is a posttranslation difference, which is of lesser concern because these attributes are readily tested for similarity. But the other variations coming from pretranslation modifications are difficult to assess. These differences arise due to the inherent thermodynamic instability driven by the atoms’ intramolecular interactions.
If I can show that a molecule has very little instability and if I have the same primary sequence, which is easy, I can ask, “Why would the three-dimensional structure be different?” And, since it is the 3D structure that produces the pharmacology and toxicology, theoretically, there should be no difference between a biosimilar and a reference product. Unfortunately, however, this novel approach goes against the common wisdom; it will take a while to settle it down.
The challenge was identifying the molecules likely to have more structural variability than others. I’m researching a thermodynamic stability demonstration of biosimilars for which the FDA has provided funding; this is in collaboration with the University of Michigan. Additionally, I am working on creating a new paradigm to establish biosimilarity by modeling all approved therapeutic recombinant DNA products(more than 200). We seek to predict structural variability with several algorithms, including AlphaFold2, based on a neural network, and ESMFold, which is model-based. We were able to classify molecules at high risk and low risk for structural variability.
Before running the evaluations, I perceived that the larger molecules would be inherently at higher risk. But, to my surprise, they are much more stable than simple peptides. So, now we are taking that one step further, testing for similarity in the domain structure.
BR&R: Can you elaborate?
Dr. Niazi: Let me put it this way: The protein structure may vary. The side chain may differ but does not interact with its binding properties. So now we are trying to identify the domains and binding and show their similarity. If you can do that, it means additional testing is unnecessary. How can the body differentiate between molecules if the domains have the same structure? It will make a big difference; I’m glad you brought this up.
The Benefits of the Inflation Reduction Act for Biosimilars
BR&R: Let’s switch gears a little bit. You have a more positive take on the effect of the Inflation Reduction Act on biosimilars than many do, including me. Why do you think the IRA will be more beneficial for biosimilar developers?
DR. NIAZI: I had the privilege of helping draft the IRA, so let me quickly summarize what it means for biologics. Suppose the originator of a biologic product has monopoly on that biologic for 12 years or more, and no biosimilar is presently competing with it or forthcoming in the near future. In that case, the list price will be reduced, not controlled, by 35% until a biosimilar arrives. This is how this act is written.
Now, the greatest misunderstanding has been around how the IRA would affect the entry of biosimilars. First, this applies only to CMS reimbursements, which are 30% of the market. It doesn’t have any effect on the rest of the market. What if the company producing the reference product reduces the price by 50%? The argument in opposition is that if the reference product price is reduced, then there is less incentive for biosimilar developers to compete for the business. It is not unusual for the reference product to drop its price once a biosimilar arrives; the current argument against the IRA concludes that this does not favor biosimilars. It makes no sense and has little to do with the IRA.
The IRA applies to part D and B drugs in a staggered manner where only the top 10 drugs with the highest reimbursement are chosen; none of the biological drugs fall in this category now. The drugs reimbursed by the CMS do not have the same market structure as the general public, which is completely ignored by those opposing the IRA.
Congress also studied the impact of the IRA on new drugs coming to market, finding that it will affect the entry of three drugs over 15 years. Unfortunately, this, too, was misrepresented in many statements made by those opposing the bill.
Another important aspect is that, for the first time, it will be in the interest of the reference drug maker to let the biosimilars walk in because then they won’t be subject to Medicare price negotiation. So that could reduce the risk of a patent dance.
If a biosimilar cannot compete with the reference product at a 35% lower cost, it should not be in this business. For example, I made antibodies and cytokines. I can sell them at an 80% to 90% discount and still make a 90% profit margin.
In my opinion, this argument that the IRA will discourage manufacturers from developing new biosimilar products is ill-founded.
BR&R: Well, your point about the cost of goods being far lower than what people perceive them to be is true. On a recent manufacturer’s fourth-quarter earnings report, they cited the cost of goods for producing adalimumab as something on the order of 1% of the total budget.
DR. NIAZI: Exactly! The cost of producing antibodies, regardless of nature, is between $50 and $100 per gram.
I calculated drug reimbursements based on cost per gram. CMS now paid the lowest price, around $3,000 per gram. The highest reimbursement was around $25 million per gram for something that costs $100 per gram to make.
We must break this barrier and get all stakeholders to accept that 12 years of a monopoly is more than enough time. Europe doesn’t even allow that much. When the BPCIA was brought to the debate, the Obama Administration had proposed only seven years of exclusivity. Then all hell broke loose, and the Administration fell under pressure, increasing the duration of exclusivity to 12 years.
BR&R: One problem is that 12 years of exclusivity gives a manufacturer time to introduce new formulations with new patents.
DR. NIAZI: There you go! That’s the other part I’m fighting!
Grow Up and Get Creative
BR&R: Excellent. I want to close the discussion with a quote you wrote in Biologics last year. You stated, “Biosimilars have come of age; now it is the developers’ turn to grow up.” What did you mean by that?
DR. NIAZI: Oh, my goodness. That quote got me a lot of comments. I do believe in it, though.
Eighteen years have passed since the EMA approved the first biosimilar product in 2005 (somatotropin). Right now, in the US at least, biosimilars are mainly in the hands of big pharma. They have no issue spending $100 to $300 million, most likely because this is the allocated funding. Still, it has set a poor example for smaller developers.
It is up to the biosimilar developers now to challenge the regulatory agencies. I always teach one thing: Every agency guidance has on the first page the line, “The FDA is not bound by it.” There should also be another line: “Neither are you.”
I recommend that developers take a more creative approach to it. One of the things I’ve learned over the years is not to follow others—they may have set a bad example. Intelligent biosimilar developers can streamline the development costs quite a bit. That’s what I meant by growing up: Biosimilar developers must raise the science to lower the costs. It’s all based on science.
