A Smarter Way to Derisk Drug Targets in Early Drug Development

Drug development is a long, costly, and high-risk journey: one that often ends far too soon. In oncology, as little as 3% of small molecule drug projects reach the market (1). Lack of efficacy and safety concerns remain the top two causes for failure in clinical trials (2). Despite multiple layers of pre-clinical screening, from in silico predictions to in vitro assays, the first red flags often still appear in animal studies where a majority of small molecule preclinical drug discovery projects are abandoned (3). Not only does this raise ethical concerns but highlights the need for the optimization of predictive and translational pre-clinical models.

What if we could spot the danger earlier?

Imagine two scenarios in drug discovery. In the first, a target looks promising through years of validation, medicinal chemistry, and disease modelling only to be derailed by a safety signal during late-stage animal testing. In the second scenario, that same safety issue is flagged before medicinal chemistry even begins, thanks to a robust, predictive model. The implications are clear: early insights into a drug target’s risk profile could save years of efforts and millions in investment.

We wanted to test this hypothesis: could we find better ways to derisk drug targets early on, before entering the traditional preclinical gauntlet?

The DCPS-FHIT Axis: A Case Study in Predictive Target Profiling

Our recent study explored this concept through the lens of DCPS (Scavenger decapping enzyme) a novel therapeutic target for Acute Myeloid Leukemia (AML).

In fact, we first identified DCPS in an unbiased screening approach where we first examined the safety profile of potential targets in healthy tissue. DCPS emerged as a standout candidate—one that not only passed early toxicology filters but had also already been evaluated in a Phase I clinical trial for a different disease indication using a small molecule inhibitor, RG3039. Even more compelling, emerging preclinical data pointed to DCPS as a promising therapeutic target in AML (4,5).

Building on this foundation, we hypothesized that expression of another mRNA cap-degrading enzyme, FHIT, could modulate sensitivity to DCPS inhibition. In other words, could FHIT serve as a biomarker for identifying patients most likely to benefit from DCPS-targeted therapy?

Using a panel of AML cell lines and patient-derived xenograft (PDX) models, we discovered a synthetic lethal interaction: cells with low FHIT expression and mutations in DNMT3A and FLT3 were particularly vulnerable to DCPS inhibition. These same genetic markers are associated with poor prognosis, making DCPS targeted therapy especially meaningful to such a patient cohort.

Further, we found that DCPS inhibition triggered an RNA-mediated interferon response, driving cell cycle arrest and differentiation in AML cells. Altogether, these findings suggest not only a promising therapeutic avenue, but also a clear strategy for patient stratification based on FHIT, DNMT3A, and FLT3 status.

Towards a derisked and hypothesis driven Drug Discovery Pipeline

Our work with DCPS and FHIT highlights the value of early risk stratification in drug discovery. If we can use hypothesis driven science to identify biomarkers and model systems that accurately predict adverse effects and identify responsive patient subgroups, we stand to make the entire pipeline more efficient, ethical, and effective.

This research is part of a growing collective effort to improve how we assess target safety before committing vast resources to drug development. Our approach shows that early risk profiling isn’t just possible—it’s essential.

What’s Next

We believe our findings open the door to more personalized AML therapies and a shift towards smarter drug development. Next steps include further pre-clinical validation of the FHIT biomarker strategy and exploration of DCPS-targeted agents in high-risk AML subsets.

Drug discovery will always carry risk—but with the right approach, we can make better decisions of how to spend resources, and bring safer, more effective therapies to patients in need.

References

1. Wong CH, Siah KW, Lo AW. Estimation of clinical trial success rates and related parameters. Biostatistics. 2019 Apr 1;20(2):273–86.
2. Hwang TJ, Carpenter D, Lauffenburger JC, Wang B, Franklin JM, Kesselheim AS. Failure of investigational drugs in late-stage clinical development and publication of trial results. JAMA Intern Med. 2016 Dec 1;176(12):1826–33.
3. Cook D, Brown D, Alexander R, March R, Morgan P, Satterthwaite G, et al. Lessons learned from the fate of AstraZeneca’s drug pipeline: A five-dimensional framework. Vol. 13, Nature Reviews Drug Discovery. Nature Publishing Group; 2014. p. 419–31.
4. Yamauchi T, Masuda T, Canver MC, Seiler M, Semba Y, Shboul M, et al. Genome-wide CRISPR-Cas9 Screen Identifies Leukemia-Specific Dependence on a Pre-mRNA Metabolic Pathway Regulated by DCPS. Cancer Cell. 2018 Mar 12;33(3):386–400.
5. Swartzel JC, Bond MJ, Pintado-Urbanc AP, Daftary M, Krone MW, Douglas T, et al. Targeted Degradation of mRNA Decapping Enzyme DcpS by a VHL-Recruiting PROTAC. ACS Chem Biol. 2022 Jul 15;17(7):1789–98.