No time to read the update? Watch/listen for free:
Edited by Brian Birnbaum.
Abcellera made a key discovery in Car-T therapy in Q1 2024 and the company needs a digital twin of your proteome to become a FCF machine. These two things are actually related.
Abcellera is building a platform that could in theory synthesise any molecule in the human body, thus potentially being capable of synthesising a drug for any condition. The human body is simply a large and complex lego puzzle, made up of molecules with a specific shape and the shape of each molecule determines its function. A molecule of a given shape will eventually bind to a molecule with an inverse shape, due to weak electromagnetic forces. Life is sustained by many molecules binding to each other, per their respective shapes, thus executing all biological processes required for a human body to stay alive and healthy, for example.
In the graph below, you can see antibodies binding onto receptor sites of Covid-19’s spike protein. Antibodies basically have the inverse shape of the target they’re trying to take out. By binding on these receptors, the antibodies ultimately modify the shape of these receptors and render them functionless. Abcellera started off with a platform that could identify and print antibodies a la carte and by acquiring new platforms that it has layered on top of its platform, Abcellera has gradually acquired the ability to produce molecules of any given shape, at least in vitro. The theory is that by producing these molecules via biological processes, as Abcellera does, we can eventually produce more effective and safer drugs at scale, at a lower cost.
But these unit economics are yet to materialise.
For future reference, in-vitro means within a petri dish and in-vivo means within a living organism, such as an animal or a human. The distinction is important because a drug can work well in-vitro but no in-vivo, for reasons that are often times not apparent to researchers.
Dysfunction in the human body is the result of one of these interactions going wrong and so in theory, illness can be fixed by putting one or more molecules of the right shape into the body. Abcellera uses its platform to discover new molecules with specific shapes, that solve conditions with an appealing addressable market size. Abcellera thus becomes investable the moment they regularly increase the number of molecules taken to the clinic every quarter, thus driving revenue growth.
At the moment this is not happening, with no molecule advanced to the clinic in Q1 2024 and just one undisclosed molecule (meaning they haven’t disclosed its function and target) in Q2 2024, named ABD-147. Having said that, the number of molecules in the clinic continues to trend in the right direction since 2020, with cumulative molecules in the clinic coming in at a record 14 in Q2 2024. AbCellera uses the term "molecules in the clinic" to refer to drug candidates or therapeutic antibodies that have been developed and advanced to clinical trials. These are molecules that originated from their drug discovery platform and are now being tested in human patients to evaluate their safety, efficacy, and potential as treatments for various diseases.
While Abcellera is therefore not yet investable for me, I qualitatively see the company’s platform evolve correctly, per the following advancement. Abcellera has found a solution to the biggest problem in Car-T therapy, which is cytokine storms. Before we get into the latter, however, bear in mind that Car-T therapy simply consists in using genetically engineered T-cells to target and destroy cancer cells. The patient’s T-cells (a type of immune cell) are collected and they are then modified so they can bind onto the target tumor. Once put back into the body of the patient, these cells go straight to the tumor and kill it, in simplified terms.
A year ago I had dinner with a Car-T therapy expert and he explained how the side effects were intolerable in many cases. These side effects stem in part from cytokine storms, whereby the T cells in question release too many cytokines thus damaging patient tissue via excessive inflammation. At the time, this doctor was enthused with Car-T therapies but saw no way around the excessive cytokine release. In Q1 2024 I was fascinated to see that Abcellera had reproducibly generated TCEs (t-cell engagers) with low cytokine release, thus outperforming Car-T in vitro.
In the Q2 2024 call, Abcellera CEO Carl Hansen explained how they initially thought the cytokine storm problem could be fixed by identifying CD3 binders, that would latch onto the CD3 region in T-cells and thus down-regulate T-cell activity, releasing less cytokines. Indeed, they’ve found three families of CD3 binders within their existing molecule library that yield high tumor cell killing and low cytokine release. Going forward, Abcellera will be testing whether this approach works in animals and then in humans.
Regardless of how this particular advancement performs in-vivo, I continue to be very interested in Abcellera because I believe the future of healthcare consists in a personalised digital twin of your proteome. The digital twin reveals the stuff that’s going wrong in your body and the molecules you put into it are the buttons you “lock and key” to fix it: molecules with the right shape will fit and initiate the function. This combination would exponentiate humanity’s ability to detect and prevent illness across the board. Abcellera is evolving into the “lock and key” factory and if you think about it at a high level, it only needs the digital twin component to become a cash machine. In essence, Abcellera is currently looking for issues that its factory can fix.
This brings to mind the main lesson that I’ve taken away from my mistakes - that investing in a company that does not solve a growing volume of acute customer pains, in a way that’s increasingly harder to imitate while producing cash tends to end in tears. Yet, the fact that Abcellera found an in-vitro solution to cytokine storms in Car-T within its library suggests that they are building a digital twin of sorts, just not a personalised one yet. Over time and especially as the company leans on developing its own drugs and thus experiment on animals and patients directly, a digital twin could emerge. Just because Abcellera isn’t solving a growing volume of acute customer pains today, it doesn’t mean that it won’t.
A personalised digital twin of your proteome becoming a reality or not is a direct function of how AI evolves over the coming decades. Indeed, a real time map of your proteome, together with a lock and key factory is not enough to yield a cash machine. You also need to run real-time simulations so you know what buttons will do exactly what, in order to not only terminate illness but also minimise side effects or even eliminate them completely. The human body has approximately 37 trillion cells, which means that the total volume of data generated by the body could be millions of terabytes per day. So we will need much more powerful smartphones than at present, because in order to run a real-time digital twin of our proteomes will need so much more compute than we carry in our pockets now.
Regardless, Abcellera does not have to develop a full digital twin of the human proteome in order to grow revenue and print profits. I was pleased to see Abcellera announce in the Q2 2024 earnings call that both ABCL635 and ABCL575 will be submitted to clinical trial applications in Q2 2025, as the company had announced when I wrote my original deep dive. How these two molecules perform in clinical trials will enable insight into how well Abcellera’s platform performs. Remember that Abcellera has allegedly built a platform that can synthesise any molecule, by layering acquired technology on top of its antibody platform.
Thus, since I wrote my deep dive the major signal Abcellera has emitted since is that their platform has been capable of circumventing the primary issue with Car-T therapies (in-vitro), which is cytokine storms. I also found it interesting to see Abcellera pointing their platform to potentially solving atopic dermatitis, by targeting the OX40 ligand. The OX40 ligand (OX40L) binds to the OX40 receptor (OX40) on T-cells, which enhances the proliferation and survival of memory T-cells. This interaction leads to sustained immune responses, including the release of inflammatory cytokines.
Atopic dermatitis is therefore chronic inflammation in the skin and by neutralising the OX40 ligand, Abcellera can theoretically solve the condition, by interrupting the chain of causality that ultimately leads to the excessive release of inflammatory cytokines. For a review on how ligands work, refer to my original Abcellera deep dive. But essentially, as depicted in the graph below, a ligand of a specific shape will simply bind to a receptor of a specific shape sitting in the cell membrane. Once the two engage, this sets off downstream changes in the cell. Enhanced memory-T cell activation is simply one of the many changes that can be triggered.
My view on Abcellera remains that the platform has the potential to greatly enhance humanity’s ability to cure (physical) illness. But, it is still not sending molecules to the clinic at a rate that would be conducive to revenue growth. While the platform continues to exhibit its optionality, via the aforementioned circumvention of cytokine storms in vitro and via the potential solution of atopic dermatitis, I believe this company needs a far more accurate and computable map of the human proteome in order to be truly useful. The world is definitely trending in that direction, with AI and compute trending up and to the right, so I will continue tracking this company going forward.
For a full-on deep dive on the science of Abcellera’s platform, check out my original deep dive. Biotech is a treacherous space at present, but the space is evolving to world’s primary platform of value creation, as are digital technologies at present. As humanity’s ability to co-create with biology matures, learning more about biology will position us to capitalise on opportunities as they emerge. My original Abcellera deep dive will greatly enhance your biotech education.
Until next time!
⚡ If you enjoyed the post, please feel free to share with friends, drop a like and leave me a comment.
You can also reach me at:
Twitter: @alc2022
LinkedIn: antoniolinaresc