There’s an old saw about bacon and eggs: the chicken, it is said, is involved in the dish, but the pig – he’s committed.
This framing may offer an apt description of pharma’s relationship with software and biotech engineers, as R&D leaders appear keen to leverage data science and digital technology, but seem to be placing their most substantive bets on emerging biological technologies like cell therapy.
In the world of early-stage life science, where I sit, it is positively staggering to contemplate what passes for commonplace in today’s world of engineered biology. I am reminded of this regularly as my venture team* & I review the latest opportunities, or when I’m at a conference (like the annual meetings of the American Society for Cancer Research [AACR] or the American Society of Gene and Cell Therapy [ASGCT]), or when I’m fortunate enough to listen to Janelle Anderson on a podcast. This used to be easy enough, when she hosted the outstanding “Human Proof of Concept” podcast, but has been a challenge since she put this must-listen show on hold and disappeared into a VC firm to serve as an entrepreneur-in-residence.
Happily, Anderson has resurfaced – not only as the Chief Strategy Officer of the recently-announced Century Therapeutics (I am not an investor), a cancer cell therapy startup that launched this summer with an eye-popping $250M in the bank, but also as a guest on Luke Timmerman’s always-excellent Long Run podcast.
The whole podcast is captivating, as Anderson describes her journey from Manitoba to McGill to Harvard to BCG to Merck to Versant to Century (and I’m sure I’ve left out a few stops along the way). As I listened to her description of the bioengineered therapeutics envisioned at Century, it occurred to me just how far we’ve come, and how much we now routinely take for granted, and I thought it might be helpful to go over (at a high level) some of these concepts and approaches.
The I/O Scientific Revolution
First some context (see also this fairly recent Wall Street Journal review and the two books I discuss therein): after years where cancer therapeutics focused almost exclusively on removing the tumor (where feasible) with surgery, or zapping it with chemicals or radiation – the so-called cut/poison/burn palate of options -- recent work reawakened medicine to the possibility that the body has the ability to attack cancer using the cells of the immune system, insight that ushered in the immuno-oncology (I/O) revolution of today. Our immune system, it turns out, is engaged in a constant battle against cancerous cells that are continuously popping up, and which the immune system (mercifully) is usually able to destroy. But sometimes – for a range of reasons that are the subject of intensive research – the cancer seems to escape control, in part by doing whatever it can to avoid or mitigate immune system attack. The essence of multiple recent insights is the discovery of ways to help restore the balance. Much of this focus has been on ways to reengineer one of the key elements of the immune system, the “effector T cell,” which is responsible for killing cells it recognizes as being cancerous.
The gist of the approach Anderson described at Century – and of related approaches researchers and startups are exploring globally – is to turn a cell that already is predisposed to assassination and equip it to be a Delta Force commando. The amazing thing is that you can actually engineer a cell this way (or at least plausibly aspire to), and contemplate dialing-in so many different functionalities, as Anderson almost casually outlines. Her description of Century offers a prismatic example of what state of the art in immune-oncology cell therapy looks like today.
The Commercial I/O Landscape
Let’s start with a quick review of the commercial history of immune-oncology. The recognition that our immune system has remarkable potential to fight cancer – research recognized by the Nobel Assembly last year -- led to the development of FDA-approved therapeutics that work by juicing our existing cells, essentially by blocking the “stand down” order that ordinarily restrains the attack. This category of drug, called “checkpoint inhibitors,” includes: ipilimumab (Yervoy, first approved in 2011, marketed by BMS); pembrolizumab (Keytruda, 2014, Merck --see here for a history of how this drug was discovered and developed); nivolumab (Opdivo, 2014, BMS); atezolizumab (Tecentriq, 2016, Genentech/Roche); avelumab (Bavencio, 2017, EMD Serono), durvalumab (Imfinzi, 2017, AstraZeneca); and cemiplimab-rwlc (Libtayo, 2018, Sanofi/Regeneron). All of these drugs are engineered antibodies, examples of the category of medicines called “biologics,” in contrast to small molecules, like statins.
A second category of therapeutics takes a far more radical approach to revving up the immune system, and involves removing the relevant immune cells (T cells, specifically), reengineering them in some way, then placing them back in the body. The most prominent commercial examples in this category are tisagenlecleucel (Kymriah, 2017, Novartis), and axicabtagene ciloleucel (Yescartia, 2017, Gilead* following acquisition of Kite). Both of these marketed products involve a version of genetic engineering in which the DNA required to make a new kind of molecule is inserted in T cells; this molecule, called a “chimeric antigen receptor” or simply, “CAR” (the engineered T-cell is called a “CAR-T”) is designed to sit on the surface of T cells, recognize a unique target on a tumor cell, and when it sees it, to activate the T cell to strike.
There are many detailed reviews of CAR-Ts available; I stumbled across this particularly useful one from Nisarg Patel. Also, although not the focus of this post, there’s considerable research – and investment -- into the engineering of other types of immune effector cells, including NK cells and macrophages, for the treatment of cancer. Modulation of another type of T cell, called Tregs, which depress effector T cell activity, is also the subject of intensive investigation – not only in cancer, but also in a range of autoimmune indications (useful reviews here, here, here).
Next-Gen I/O Problems To Be Solved
With this history in mind, we can begin to better appreciate Anderson’s discussion of Century. The general goal – of both Century and so many in the field – is to build on what works with first-generation CAR-T approaches, and to fix, or at least mitigate, some of the challenges.
For starters, a challenge with first-generation CAR-T approaches is the need to engineer custom cells for each patient, the inevitable consequence of “autologous” cell therapy that uses the patients’ own cells. An often-discussed goal among many in the field is to generate so-called “off the shelf” cells, cells that could be compatible with any patient, and so ideally, you could think of them like any other drug product, where you manufacture a large batch in advance, parcel it out, and use as needed. Easy to say, of course, but much harder when the drug product is a living cell.
The approach Century (and others) are using involves a type of stem cell called an “induced pluripotent stem cell,” or “iPSCs.” The power of iPSCs is that they have the potential to differentiate into virtually any cell type; yet, in contrast to embryonic stem cells (ESCs) which have essentially the same property, iPSCs are not derived from embryos, and thus are free from many of the legal, ethical, and political considerations with which ESCs are freighted. iPSCs instead come from specialized cell types that have been returned to the stem cell-like state by the addition of a specific chemical cocktail; Shinya Yamanaka* shared the Nobel Prize in 2012 for figuring this out.
As Anderson explains it, iPSCs are a renewable cell source – meaning you can grow them indefinitely in culture, in contrast to typical differentiated cells, which tend to peter out. This is critical for Century, according to Anderson, because it means that you can serially and precisely engineer these cells, introducing a seemingly limitless number of modifications to enable the iPSC, once differentiated into a T cell, to become, at least in theory, a highly specific, highly effective tumor killing machine. Moreover, this serial engineering can be done through highly-specific gene editing approaches, like CRISPR, so you can add new genetic material and remove unwanted genetic material in a molecularly precise fashion.
After each genetic modification, Anderson says, you can select the cells you want, and expand them to generate a large number. This population, Anderson continues, is “consistent, and can be fully sequenced and characterized so you know what you have.” This contrasts with the approach used for autologous approaches, where you need to need to repeat your (typically far more limited) engineering for each patient, and there can be a lot more variability. Such “heterogeneity affects product potency,” Anderson points out.
Through sequential editing, researchers can address a range of potential cell therapy challenges. For example, if introduced cells still have their own T cell receptors (in addition to the CAR), they could potentially attack a patient’s own healthy cells – so-called graft-vs-host disease (GvHD) . Gene editing approaches could enable the CAR to be inserted in place of the endogenous TCR, for example. Similarly, the recipient’s own T cells could potentially attack the introduced cells – the “host versus graft” (HvGR) reaction. This could be managed by deleting key molecules on the cell surface critical for such recognition; removal of these structures could enable the introduced cell to largely escape detection.
Cell-based cancer therapy approaches confront a range of additional challenges; many tumors – especially many solid tumors – create a milieu, known as the “tumor microenvironment,” or “TME,” that’s hostile to immune cells, and are filled with factors that suppress the immune reaction (like the poppies in the Wizard of Oz). Attacking T cells also suffer from what’s actually known as “T cell exhaustion,” meaning they may still be present but operate much less effectively, as if they are just tired out. A third challenge for cell therapies is achieving specificity – many tumors don’t have unique markers, so attacking them without injuring normal cells is a significant problem; there are a range of approaches now under development to address this (such as requiring the introduced T cell to recognize some combination of markers or factors in order to respond; this is addressed nicely in Patel’s overview); presumably this capability is something that could also be engineered into the cells Anderson describes. Finally, a fourth opportunity is enhancing the lethality of the cells you are introducing, so that they are even more effective at obliterating the tumor.
As Anderson emphasizes, the key advantage to this approach is that researchers “can edit almost without limitation. If you start with donor derived cells [whether autologous or allogenic], you are much more limited with respect to every gene edit.” But if you use iPSCs, she argues, “the sky’s the limit because of the number of gene edits one can make because of the base technology,” adding, “there’s a lot you can pack in there." As Century's Chief Scientific Officer Luis Borges subsequently elaborated, "One reason we can do editing without much limitation is because we have the luxury to check for any off-target effects of the editing procedure through genome sequencing, and can eliminate any cells that have unwanted changes. We can check the engineered cells and then do single cell cloning to expand a clone that has just the precise engineering we intended to introduce."
Caveats
Of course, there’s a long way between theory and clinical practice. My most significant concern regarding iPSC-based approaches reflects my own experience as a post-doc in a stem cell lab; there, I was frequently struck by the impact of what’s known as “passage number” – essentially reflecting the number of times a particular cell line has been expanded. While in theory, stem cells can replicate endlessly, in practice, each successive generation may be slightly different than the one before, and may have accumulated new mutations, or new modifications that impact how the cell behaves. I instinctively worry about the length of time a cell must be cultured and expanded, and I worry that a cell line after successive deliberate modifications may have accumulated a series of imperceptible additional modifications that could impact the behavior of the cells – though it’s true that a population of cells developed in this fashion at least should be more homogenous than populations of cells engineered anew for each recipient, if perhaps not as “fresh.”
Beyond this iPSC-focused concern, many of the other engineering approaches that Century and others are contemplating are attractive but largely unvalidated clinically; you can understand why they make sense, but that’s not the same thing as working in afflicted patients.
Audacious Biological Engineering
At another level, though, what’s apparent from Anderson’s description of Century, and from the work of other companies in the cell therapy space, is how incredibly audacious – and routinely audacious -- biological engineering has become. Even the first-generation CAR-T approaches are astonishing, in that they introduce a genetically engineered fragment into a patient’s own cells – arguably an example of gene therapy, or at least a gene therapy-style technique. Then consider the approaches that Anderson describes – such as precise gene editing using CRISPR or similar techniques, as well as the use iPSCs; and it’s not just Century -- aspects of these approaches are typical features of many company proposals my team and I evaluate. Any one of these elements would have been considered beyond fanciful back when I was training, or perhaps at best the sort of zany thing a radical biotech might contemplate. Today, these are the techniques and therapeutic approaches that most large pharmas are pursuing -- aggressively. Virtually all major drug makers seem to be going after opportunities in this space; in addition to the manufacturers of FDA-approved products cited above, others jumping in include J&J (here, here), Pfizer (here, here), Bayer (here, also a key partner in Century as Anderson discusses), and many others (including, Takeda*, as was recently discussed here and here).
While it’s still very early days for cell therapy, and successful financings in this space are far more abundant than FDA-approved therapies, you can certainly appreciate why so many biological engineers, in academia and industry alike, are so excited – and why even some software engineers, like Sean Parker, are starting to think the world of biology may represent “a much more exciting place to be.”
The reality is that it’s both breathtaking and humbling to be at the cusp of two profound engineering revolutions, one involving biology, the other, software. Individually, each has the potential to profoundly reshape the way new medicines are designed, developed, deployed, and evaluated; if thoughtfully integrated, the prospects are staggering. Biopharma leaders increasingly will need to have a solid grounding in both of these spaces, and understand the possibilities, limitations, and implementation hurdles if this remarkable potential is to be translated effectively into the elusive outcome that matters most – improved health for patients.
*Disclosures
- Yamanaka partners with Takeda as described here. 2. TVI, where I work, is the corporate venture arm of Takeda. As always in this column, the views expressed are entirely my own and may not reflect the views of any entity with which I’m affiliated. 3. My wife is a senior R&D executive at Gilead but is not involved in the oncology portfolio.
