IP Considerations Relating To Antibody Drug Conjugates (ADC) #patent #antibody #drugdevelopment #

Foreign.

My name is ron Najafe. I'm CEO of Emery Pharma. It's a pleasure hosting Dr. Julia Manetti. Julia did her PhD at Brown University, her JD at Santa Clara University. She spent a few years at the Albany Molecular research in Albany, New York, prior to her PhD as a bench chemist. So she is a chemist. She did chemistry methodology, synthetic methodology, which is really my background. And she's been practicing patent law for the last seven, eight, nine years at Wilson Sonsini, very premier law firm in San Francisco Bay area and perhaps in the United States. Julia, today she's going to be talking to us about challenges in patenting antibody drug conjugates, which is really a hot area right now. Several of our team members are actually at a conference on ADCs in Boston. We're obviously involved, many of you are involved in antibody drug conjugate analytical methodologies and bioanalysis. So I thought it's good to learn some of the challenges in patenting ADCs. So here is Julia.

Great. Well, thank you, Ron, for having me here today. I'm gonna talk a little bit about patents, getting patents to antibody drug conjugates. And hopefully this is interesting to all of you. Feel free to ask questions during the talk if you have any. Because I'm a lawyer, I have to give my disclaimer at the beginning. So my comments here do not reflect my firm. They reflect my my own comments and my own opinions. So today I'm going to talk a little about ADCs generally and how we patent them. I'm going to talk a little bit about some of the challenges we're seeing to antibody patents as well as challenges to linker patents. As we all know, antibody drug conjugates have the antibody and the linker bound to the payload. And then we're going to talk about strategies for obtaining ADC patents and finally, some other ADC considerations. Many of you probably know this already, but antibody drug conjugates are molecules that include an antibody, which is referred to often as the targeting moiety, or a portion of an antibody, as well as a covalent linkage system. Now, this could vary between a cleavable linkage system or non cleavable linkage system. But many of them are sort of well established, well known in the field. And finally there's the payload. The payload. In many of the cases of ADCs we're discussing cytotoxic payloads, but we might also be talking about other payloads like immune stimulatory payloads. Or radioisotopes. So other molecules that you want to get to the site that the antibody is drawn to. EDCs are a growing sector of approved drugs, so they're the fastest growing drugs in oncology. And as a September 2023, there's 13 ADCs have been approved by the FDA and seven have been approved by the EMA. So target antigens of approved ADCs include CD33, CD30, CD22 immune cells, HER2, nectin 4, et cetera. The cytotoxic payloads of approved ADCs are commonly recycled. So you've probably heard of MMAE or diroxtican, which is exotican the topoisomerase inhi in that these are very cytotoxic agents that are targeting cancer cells and are often reused. The linkers of approved ADCs are primarily cleavable linkers enzymatically or chemically. Although I will say this area has been growing, I'm seeing a lot more innovation in the linker space over the past couple of years. Just like any other drug, patent protection around ADCs is pretty essential. The cost of developing a drug and getting it to market, as many of you know, is very high. Hundreds of millions to up to $3 billion to get a drug. From early stages to commercialization, we're seeing a lot of investment by VCs in ADC Technologies that's been growing since 2018. And we think there's about $1.8 billion invested in ADCs by the venture capital firms. So. So it's definitely an area of interest and focus. Patent protection is very important to get these investments though. So you want to be able to show that you have exclusivity around your antibody drug conjugate in order to get those investments, grow your small company and eventually get your product to market. I'm going to start sort of simple here. Before we go into the nuts and bolts of how the issues with patenting ADCs and some of the challenges we face and some of the challenges that are facing the patents that are issued in that space, I just want to talk about patentability first. So what's the requirement of a patent? So to get a patent, there are four main statutory requirements. There's Section 102 USC Section 102 that requires that your invention be novel. Section 103 that requires that your invention be not obvious. Section 112 is split into two parts. It's written description and enablement. So written description and these can get confused a little bit. Requires that you describe your invention to show that you were in possession of the invention at the time it was filed. So you need to adequately describe the full scope of your invention. Enablement's a little bit different, although they often get confused, even by the courts, even by the Supreme Court. The enablement requirement says that you have to teach someone of. Of ordinary skill in the art how to make and use the invention without undue experimentation. So as a scientist, I think it makes a little bit of sense to us. When we read a paper, are we able to read that paper and reproduce the invention? That's enablement. Are there enough experimental procedures, is there enough data in the paper to enable us to make and use the invention? Whereas written description is a little bit more of. If you've described a broad scope of antibody drug conjugates, have you adequately described all of the members of that group? Have you described which antibodies you could use? Have you described which payloads you could use? So that's description, not necessarily examples. It's just describing your invention. Okay, sure. Is that review process, I guess, the patent office. Yeah, that's right. Is that someone that has a scientific background or is that. Yes, yeah. So the question is for the evaluation of a patent to determine patentability. Does the person who's evaluating the patent, who's a patent examiner, have a scientific background in order to evaluate it? And yes, with differing degrees of schooling. So often we'll have examiners who have maybe a master's in your area of science. Sometimes there's PhDs in the patent office. I would say it's a bit of a mixture, but they are specialized. And so when you're working with, we call it unit 1600, if you're working with a particular unit at the patent office, the people that you're sending your patent application to will know how to search it, because in order to determine novelty, they need to be able to search the chemical database to determine whether the molecules are novel. So you need to know how to use SciFinder. Right. You need to know how to search the chemical databases. Or if you have a sequence, you need to be able to search the sequence. So you do have to have expertise, enough expertise in the sciences to understand the invention and to search it. I mentioned there's been challenges to antibody patents, so you might have heard, maybe not. But there's been a good amount of case law in the past couple of years around antibody patents. And the issue is, for many years, the patent claims to antibodies were defined by the epitope or by the antigen. So the claim would read something to the effect of an anti CD3 antibody. Right. Which would, you can imagine, cover any antibody that would hit CD3, whether it's hitting the same epitope or a different epitope as your described molecules. And so those claims were traditionally the way patent filers would go after antibodies. Well, that has been largely shut down in the U.S. so the courts and the patent office now have said you cannot get every antibody that hits an antigen. Right. You're just, you haven't now going back to enablement and written description, you haven't shown enough examples of sequence diversity for antibodies that hit that antigen, and you haven't described the full scope of antibodies that hit that antigen. So on neither front either written description or enablement, have you satisfied the requirement for patentability. Probably you may have heard of some of these cases, but I think the one that you might have heard of the most was this most recent Supreme Court case, which was Amgen Sanofi, which was a claim to anti pcsk9 epitope. So the idea was the claim was drawn to any antibody that hit one or two amino acids out of 18 on PCSK 9. And the court had said, you're not entitled to every antibody that hits a, you know, up 1, 1 or more of those residues on PCSK9. You just, you haven't shown enough antibodies that, that fit that description. Right. So as a result, that claim was found invalid for lacking enablement. Now, I think the situation might have been a little bit different if they had disclosed, if Sanofi had disclosed more antibodies in their patent application. So if they had disclosed 100 different antibodies that had a variety of variable regions that hit PCSK9, then they might have been in a better place. But instead they didn't disclose a sufficient number of antibodies. So their case was found to lack enablement and invalidated. Now you can imagine. So why are people claiming this? This is what you're going to ask me. Why do you want to claim an anti a PCSK9 or an anti HER2 antibody rather than claiming just the CDRs or just the variable regions? Right. Well, the reason is claiming just the CDRs is easy to design around. So if another company wants to enter the space and wants the design around your invention, then all they would need to do if you're claiming the CDRs, which I don't everyone know what I mean by CDRs and variable regions. So it's sort of the business end of the antibody. The residues of the antibody that interact with the antigen. If you're only claiming those particular residues, then a third party could design around that by swapping out a residue or two, probably still behaves the same way. It's not going to be that different. If you swap one amino acid in the CDR and you can get around the claims. And so the claims traditionally have tried to get breadth. You want to try to get breadth and not limit yourself to the particular system sequence of the antibody, the particular business end of the antibody sequence. So that's the challenge we've had in antibody patents over the years that the courts are very active in knocking down patents. They're saying essentially you can't get breadth around your antibody. You need to limit your claim to something very narrow and defined by the sequence in either the variable region or the cdr. So the takeaway from this, I'm not going to go into the specific cases, but the takeaway is that epitope or antigen focused claims have been consistently rejected by the Federal Circuit and the US Supreme Court for lacking enablement written description. That carries over to ADCs too. This is a recent decision that came out from the ptab, which is the Patent Trial and Appeal Board. I raised this one because I think it's interesting it concerns linkers rather than the antibody itself. So I don't know. Is anyone familiar with the antibody drug conjugate linkers? Some of the common ones, you can yell it out if you know any. Okay. Some of the common ones are Valcit, you might have heard of that, Val Ala. So these are sort of amino acid linkers that usually have maybe a self immolative group that then attaches to the payload and maybe there's some sort of stretcher unit to attach to the Maleimid that attaches to the antibody. But CGEN was one of the innovators in the space of linkers. CGEN has a number of patents to linkers. In 2008, Daiichi Sankyo and CGEN entered into a collaboration to develop ADCs targeting cancer cells. The collaboration ended. Daiichi sankyo produced in HER2 which you're probably familiar with an anti HER2 antibody covalently linked to Exatecan, which is a topoisomerase inhibitor. And that was for the treatment of breast cancer and non smell cell lung and gastroesophageal. And your HER2 is obviously doing very well as a product. It's projected revenue for is going to be $6 billion per year by 2026. So in her two hits the market Seagen sues Daiichi for patent infringement of their patent number here, the 039 patent, we refer to it as the last three digits. And so this goes to court. Often these cases go to court in the Eastern District of Texas, which it did here. So the case was heard in the Eastern District of Texas and the final judgment was ordered Daiichi to pay an 8% royalty on sales of in Hertu for infringing CGEN's patents. So they found patent infringement. They found that Daiichi Sankyo was indeed infringing CGEN's linker patent. So Daiichi then turns around and challenges the patents at the PTAB using a process known as post grant review. So what this is, it's an interaction with the patent office where you go back to the patent office rather than the courts and you say, I see a problem with this patent. And they challenged it based on lack of enablement due to the breadth of the claim. So CGEN's patent claim is drawn to an antibody, and this is the claim of a patent. If you don't look at claims of patents at all, they're usually one very long sentence that includes all the various parts of the molecule. And this is what's called a Marquish type claim because it has variables in the structure. So there's the structure that's over on the right hand side and there's certain variables that are then defined by the, by the very long sentence. So the sentence says antibody, drug conjugate, having the formula. And the formula is over here on the right hand side, your right hand side. And as you can see, if you're familiar with antibody drug conjugates, they are typically. The antibody is typically bound through a sulfur, a cysteine residue to a succinimide, previously a malamide, some sort of structure unit. There's a W group here, a Y group here, and a D group here. And then you can see there's a little tiny. I don't know what letter that is. I can't really tell. P, I think it's a little P. And that says how many linker drugs are bound to the antibody. That's the dar, if you're familiar with the drug antibody ratio. And so here the antibody is, if you look at the claim, the antibody is an antibody. That's all. It's any antibody. So the claim is drawn to any antibody. There's S is sulfur. So you really need to define that. But they did. W is independently an amino acid unit having the formula denoted below in the square bracket. This W Right here is this. So that's an amino acid. The side chain here, R19 is hydrogen or benzyl. So does anyone know what amino acids those are? Just yell it out. So if you have a hydrogen here, it would be glycine. Yeah, I heard it. Okay. Yeah. And if you have a benzyl here, it would be phenylalanine. You got it. So it basically says you've got W here you can have a repeating unit. It says W is independently amino acid unit having the formula denoted below in the Square bracket where R19 is hydrogen or benzyl. I think it said. Does it say the number of times it can repeat. Y is a spacer unit. Okay. And Then Y is 0, 1 or 2. So that's this group here which is a spacer unit. That's all they've referred to. They haven't said what a spacer unit is. It's anyone's best guess. And then P over here, that's the DAR is 1 to 20. That's quite a range of DARs. Okay, so what I'm trying to point out here is that in some ways this claim is very broad, right? You could have any antibody on the left hand side, you do have to have a tetrapeptide at W, but it can be any glycine, glycine or phenylalanine in any mix and matched order. Why is a spacer unit? That's undefined. And then D is a drug moiety. So any drug can be attached there. So that's the scope of the claim. And then at the end of the claim, there's some functional language, it says, wherein the drug moiety is intracellularly cleaved in a patient from the antibody of the antibody drug conjugate or an intracellular metabolite of the antibody drug conjugate. So there's this sort of limiting functional language that they haven't defined the spacer unit or what attaches the linker to the drug, but they're saying that the molecule has this property that when it's in a patient the drug cleaves intracellularly. So is what it is. Okay, so that was the language of the claim. Speaking of enablement, when the patent office is evaluating enablement, as I mentioned, they're looking can a person reasonably skilled in the art make and use the claimed invention without undue experimentation? So you're looking at the full scope of that claim. That one we just looked at. Can you make and use the full scope of that invention? Can you make any antibody attached to that Linker? Can you make any drug attached to that linker? Do you know what spacer is? Do you know all of the necessary procedures needed to make and use that invention? And enablement turns on the analysis of what's called the wands factors. And there are eight factors that go into the analysis of whether you've enabled your claims. The quantity of experimentation necessary to carry out the invention, the amount of direction or guidance presented in the reference, the presence or absence of examples, the nature of the invention, the state of the prior art, the relative skill of those in the art, the predictability or unpredictability of the art, and the breadth of the claims. Clearly, the breadth of the claims in some ways is quite broad. One problem that the PTAB found with this patent was that this functional assay at the end that I mentioned requires a lot of trial and error. So if you look at that functional assay at the functional language at the end of the claim that said that the drug was cleaved intracellularly, that's not something you could just know by looking at that. You'd have to test that. You'd have to iterate. You'd have to make different molecules and see if they actually fit within the scope of the claim. That could be a lot of work, right? To try and figure out which molecules fall within the scope of the claim, you'd have to run experiments to see if the drug gets cleaved intracellularly. So the PTAB was not, you know, didn't look on that too favorably as they felt that that functional assay required a lot of work. As, you know, undue experimentation is work. Right? That would be. That would not enable the invention. Also, the expert here commented that there would be a lot of experimentation required to create any particular antibody drug conjugate. You can't necessarily just look at that and say, I know how to make any drug bound to this linker and any spacer and any antibody bound together. I think probably the biggest issue they had with this, and this is something I always stress to clients when I'm working with them on patent filings, is that you need more examples. You need more working examples. And what they had here was that every single working example in the patent involved an ADC with a dolostatin or oristatin derivative. So they had a limited number of drugs. They didn't have very many examples, and they didn't have a single example of a tetrapeptide that was a glycine phenylalanine tetrapeptide. In their molecules, so they had no examples to show that they'd actually made a glycine phenylalanine linker. So these are all kind of cautionary tales of writing applications. And it's easy with hindsight, to look at someone's patent and say, you should have put more in there. But moving forward, I think that's going to be the way we have to draft patent applications. We need as many examples in there as we can get, even if some of them are prophetic and you might not perform them for a bit. So the fact that they did not have a single glycine phenylalanine linker in there was an issue, and this patent was found Invalid. So that 8% royalty that Daiichi was paying to CGEN ended after this PTAB decision. Does anyone know what the tetrapeptide is that would have glycine and phenylalanine? I don't know. If you're heavy into the antibody drug conjugate space, it's the GGF.