While oncogenes are notorious for causing cancer when mutated, they are not the only genes that keep cancer alive and spreading. Cancer fitness genes don’t cause cancer, but they help cancer cells avoid death through a variety of different mechanisms including by keeping oncogenes functioning, promoting certain signaling pathways, or sustaining an immunosuppressive environment around the tumor.

Cancer fitness genes encode “enzymatic pathways that cancer cells use in settings of stress to give them multiple advantages,” said Catherine Sabatos-Peyton, who is the Chief Executive Officer of Larkspur Biosciences, a biotechnology company developing small molecule drugs that target cancer fitness genes. “We've discovered that knocking out or degrading these targets makes a cancer cell vulnerable to multiple mechanisms of death.”

Without protection from cancer fitness genes, tumors become vulnerable to apoptosis, become less invasive, and can become susceptible to killing by the immune system.

At the recent American Association for Cancer Research meeting (AACR 2025), Sabatos-Peyton and her team reported preclinical data on two of their small molecule degraders that target the cancer fitness enzymes phosphatidylinositol 5-phosphate 4-kinase, type II, gamma (PIP4K2C) and peptidylprolyl cis/trans isomerase, NIMA-interacting 1 (Pin1) (1,2). By targeting cancer’s response to stress, the Larkspur team plans to weaken cancers — starting with colorectal cancer — enough to destroy them altogether.

How did the Larkspur team decide to focus on PIP4K2C and Pin1 as cancer fitness targets?

There's significant genetic validation for both targets, and historically, both were difficult to drug. While PIP4K2C is a lipid kinase, it’s actually the least active human protein kinase. People tried targeting it with inhibitors, but they did not see an effect. What our founders — Lewis Cantley, Nathanael Gray, and Vijay Kuchroo — discovered was that knocking out or degrading the target is what's required. Its important function is in a scaffolding role. Inhibition is insufficient; we needed to therapeutically degrade it and remove it from the system.

Similarly, Pin1 has been known to be involved in cancer fitness for a long time. Researchers have studied it in cancer for probably more than 20 years, largely through genetic strategies. People made lots of attempts to inhibit it, which turned out to be quite difficult. We started with an initial lead from our founder Nat Gray and our Scientific Advisory Board member Nir London. They had developed a very interesting covalent inhibitor for Pin1. From there we moved to a degrader strategy because we discovered that that was a good way of targeting what is a difficult-to-drug target. As we've done our work on Pin1, we've uncovered aspects of that target that were simply unappreciated in attempts to inhibit it previously, which I think is really exciting. It's really this marriage of important targets that are these cancer fitness genes and the chemical strategy of degrading that has opened up a space that historically was intractable.

How do your small molecule degraders work?

They are hetero-bifunctional degraders that are based on cereblon (a protein that, with other proteins, forms an E3 ubiquitin ligase complex). Most commonly people would call them PROTACs. We have a targeting ligand, a linker, and then the degrader component. We have used cereblon as the degrader in both of them, which I think has the best clinical translational data to date. There are obviously other flavors of degraders, but cereblon degraders have become the most advanced. Our lead drug, Lark-A, which targets PIP4K2C, is a really potent, fast-acting degrader that’s orally bioavailable, which is exciting. We're developing it as a once-daily oral pill for the clinic, and it has terrific specificity as well as efficacy in our preclinical studies.

Can you tell me more about the preclinical work for targeting PIP4K2C?

Absolutely, it's super exciting! We know that the PIP4K family in general — PIP4K2C is one of the isoforms — is important for lipid remodeling. Building on the publicly available data and some proprietary work from our founders, we discovered that when we degrade PIP4K2C we do a few critical things. One is that we make the cancer cells less invasive. We've also discovered that when cancer cells are exposed to stress and we treat with our drug, we see increased apoptosis. In those stress settings and in a dose dependent way, Lark-A increases the apoptotic response; it exposes that vulnerability.

PIP4K2C is also involved in intracellular vesicle trafficking. When we treat cancer cells with Lark-A, we see an increase in the levels of NKG2D ligands to the surface of cancer cells. These are the natural killer (NK) cell “kill me” ligands that decorate a cancer cell for NK-mediated killing. We get upregulation of those “kill me” ligands, and we get increased killing by NK cells.

The third component is a cooperative mechanism between the cancer cell and phagocytes, which we modeled using dendritic cells. When we treat with Lark-A, we get increased uptake of dying tumor cells. This exposes them to the immune system, and we also see increased T cell-mediated killing. It really is the complementarity and the components of the mechanism that make it so potent.

What makes Lark-A a good approach for treating colorectal cancer?

Our first indication is metastatic, microsatellite stable (MSS) colorectal cancer (CRC). There is a massive unmet medical need for patients. Eighty-five percent of patients with CRC don't respond to immunotherapy, and when we look at third-line plus therapies, there are approved drugs with response rates of 1.5 to 6 percent. CRC is also on the rise in young people. It’s quite a devastating disease.

We've discovered that the knockout or the degradation of PIP4K2C is particularly important in MSS CRC and not in other forms of CRC. We also have some really interesting emerging biomarker data that's showing us why that is mechanistically. We're able to show that those “kill me” NK ligands only come up in the MSS cells. When we treat with drug, we get a preponderance of responses both in in vivo models and in ex vivo patient spheroid samples. We will discover more in the clinic, but the preclinical data suggest that this is potentially important for these specific patients.

What is your LarkX platform, and how does it fit into your preclinical research?

This is a bioinformatics and machine learning approach that we have developed at the company to discover other pathways that are involved in cancer fitness. Our focus is on uncovering other vulnerabilities in a cancer cell that are intrinsic while also bringing in the immune system. It does two things. One, it has helped us position the targets that we're already pursuing translationally in patient populations where we think that these adaptations are particularly used and important. Therefore, we can really enrich for what we hope will be a patient signal when we transition to the clinic. Then, two, we have also used it to uncover other mechanisms that are complementary to the targets we're already pursuing, and we hope to further expose these cancer adaptations to stress.

What’s next for Larkspur?

We are on track to file for our first-in-human study later this year. We are intending to get into the clinic before the end of the year, so that's really exciting. Being able to make that transition to become a clinical-stage company will be a huge milestone for us, so we're really concentrating all of our efforts in that way. For the second program in Pin1, we intend to deliver a candidate this year, and I think that's a great opportunity for the company. As we're bringing the lead candidate to clinical testing, we’re then coming in with the second candidate that can move into IND-enabling studies and really build out the pipeline in that way. Those are two key inflection points that we're looking forward to.

What has been the most exciting part about developing these small molecule degraders for cancer treatment so far?

So many things! I'm an immunologist by training. I'm a data junkie. Our team is wonderful in that as they uncover new things, they're always excited to share them with me, and I really appreciate that they are willing to keep me close to the science as I do what I need to do to grow the company. I really love thinking about the data and trying to understand what it means to help us create the best translational program at this stage and to deliver the best for patients in the clinic. The team made me a printout of some of our most exciting preclinical data because I loved it, and now it lives on the wall outside my office so I can look at it. It's been wonderful!

This interview has been condensed and edited for clarity.

References

1. d'Hennezel, E. et al. Abstract 3963: Tumor-intrinsic impact of PIP4K2C degradation drives preclinical efficacy in colorectal cancer. Cancer Res  85, 3963 (2025).
2. Chow, M. et al. Abstract 2709: Disrupting cancer fitness and pro-tumorigenic fibroblast functions with a first-in-class Pin1 degrader. Cancer Res  85, 2709 (2025).