Cancer drug discovery is entering a more disciplined era. Investors, partners, and clinicians are increasingly focused on therapeutic approaches that can demonstrate clear clinical relevance, strong translational rationale, and practical paths to development. At the same time, advances in peptide engineering, targeted delivery systems, and computational drug discovery are expanding the range of previously inaccessible oncology targets. In this conversation with BioSpectrum Asia during BIO International Convention 2026, Karsten Eastman, Co-founder and CEO of Sethera Therapeutics, shares his perspectives on the evolution of precision oncology, the growing promise of constrained peptide therapeutics, and the key trends shaping the next generation of cancer innovation.
Precision oncology continues to evolve rapidly. Which emerging therapeutic approaches do you believe will have the greatest clinical impact over the next five years?
The greatest impact will likely come from therapeutic approaches that improve selectivity without sacrificing developability. Oncology has already moved well beyond the simple idea of “cytotoxic versus targeted therapy.” The field is now asking more specific questions: Can we reach the right tumor biology? Can we modulate the right protein interaction? Can we deliver the right payload? Can we identify the patients most likely to benefit?
I think several areas will be especially important. We will continue to see progress in modalities that can reach historically difficult targets, including protein-protein interactions, intracellular targets, transcriptional regulators, and tumor-specific conformational states. Targeted delivery approaches, including antibody-drug conjugates, radioligand approaches, and ligand-directed therapeutics, will remain central because they create therapeutic index where small molecules or systemic biologics may struggle. Finally, I believe constrained peptides and macrocyclic peptides are moving into a particularly important period.
Peptides sit in a valuable space between small molecules and antibodies. They can provide high-affinity, high-selectivity recognition while also being engineered for stability, tissue exposure, target engagement, and conjugation. Historically, peptides were viewed as limited by stability, permeability, and delivery. That view is becoming outdated. As the field develops better ways to stabilize peptide structures, screen very large chemical spaces, and build drug-like properties into peptide scaffolds earlier, peptides are becoming a serious therapeutic class for oncology targets that are not easily addressed by conventional modalities.
What are the biggest barriers biotech companies face when translating promising cancer research into successful clinical programs?
The largest barrier is not usually the absence of interesting biology. There is a tremendous amount of compelling cancer biology. The harder question is whether that biology can be converted into a product with a clear clinical path.
For biotech companies, the main translational barriers are target validation, modality fit, biomarker strategy, manufacturability, pharmacology, and capital efficiency. A target may be scientifically compelling but still not be therapeutically tractable. A molecule may bind well in vitro but fail because of exposure, stability, delivery, toxicity, or lack of functional effect in the relevant cellular context. A biomarker may be useful scientifically but not practical for patient selection. These are the gaps that often separate promising research from a successful clinical program.
Another major issue is that platforms can create the impression of progress faster than they create clinical evidence. This is particularly true in areas where computational design, machine learning, or large-scale screening generates many apparent candidates. These tools can be very powerful, but only when they are connected to high-quality experimental data and a disciplined translational strategy. In oncology, the patient does not benefit from a predicted binder or a beautiful model. The patient benefits from a molecule that reaches the right tissue, engages the right target, produces the right biological effect, and has a tolerable safety profile.
The companies that will translate best are the ones that combine discovery scale with experimental discipline. The winning platforms will not just generate more molecules; they will generate better decision-making.
How is the current funding and partnership environment influencing oncology innovation and development priorities?
The funding environment is forcing more discipline. That is not entirely negative. In a more selective market, companies have to explain not only what their technology can do, but why it matters clinically, what the first product path looks like, and what data will de-risk the program.
For oncology innovation, this means investors and partners are prioritizing clearer development paths, stronger translational rationale, and assets that can produce meaningful inflection points before requiring very large amounts of capital. Platform companies have to be especially thoughtful. A broad platform story is valuable, but it is not enough. The platform needs to produce product candidates, and those candidates need to address real limitations in existing treatment paradigms.
Partnerships are also becoming more focused. Large pharmaceutical companies still want differentiated science, but they are looking carefully at whether the modality can be developed, manufactured, protected, and advanced through clinical testing. In that environment, peptide therapeutics are becoming more attractive because they can offer a practical balance of novelty and developability. A constrained peptide that reaches a difficult oncology target, demonstrates strong binding or functional modulation, and can be modified for stability, delivery, or payload attachment creates several partnership paths.
I also think the market is becoming more discerning around AI and machine learning. These tools are useful, and they will remain important, but the strongest opportunities are likely to be those where computational methods are tied to proprietary experimental systems. AI without a high-quality data engine is less compelling. AI connected to real selection data, chemistry, binding measurements, and biological assays can be extremely valuable.
What unmet needs in difficult-to-treat cancers remain insufficiently addressed by existing therapies?
The most important unmet needs remain in cancers where current therapies do not produce durable responses for enough patients, where resistance emerges quickly, or where key disease biology remains difficult to drug.
Several categories stand out. Tumors with complex or heterogeneous driver biology remain difficult. Immunologically cold tumors remain difficult. Cancers driven by intracellular interactions or loss-of-function tumor suppressor biology remain difficult. Metastatic and recurrent disease remains a major problem, even in cancer types where initial responses have improved. There are also major needs in cancers where the target is known, but the modality needed to reach that target has not been good enough.
This is where I think new peptide modalities may be particularly useful. Many oncology targets are not simple enzyme active sites. They are surfaces, complexes, conformational states, or interactions. Antibodies can be powerful when the target is extracellular. Small molecules can be powerful when there is a well-defined pocket. But there is a large space between those two categories where constrained peptides may have an advantage.
A peptide can be engineered to recognize a protein surface with high specificity. It can be stabilized through macrocyclization. It can be linked to payloads, imaging agents, or delivery systems. It can be optimized for potency, selectivity, and pharmacology. That flexibility is valuable for difficult cancers, particularly where the therapeutic hypothesis requires more than simple inhibition of an enzyme pocket.
What key trends should industry leaders and investors be watching at BIO 2026?
The field is moving from platform breadth to product credibility. Investors and partners want to see whether a platform can generate real therapeutic candidates, not just interesting discovery outputs.
Peptide therapeutics are becoming more central. The success of peptide drugs in other therapeutic areas has changed how the industry thinks about the modality. In oncology, the next step is to apply increasingly sophisticated peptide chemistry and screening technologies to targets that have been difficult for both small molecules and antibodies.
Data quality will matter more than data volume. Large datasets, computational models, and AI-enabled workflows are useful, but the differentiator will be whether the data are experimentally grounded and biologically relevant. The industry should be cautious about confusing prediction with validation.
Targeted delivery and conjugation strategies will continue to expand. Whether through radiopharmaceuticals, ADCs, peptide-drug conjugates, or other ligand-directed systems, the central goal is the same: improve therapeutic index by putting activity where it is needed.
Partnerships will increasingly form around clear translational inflection points. Companies that can show differentiated chemistry, a defined clinical use case, strong early pharmacology, and a credible path to development will be in a much stronger position.
I think oncology innovation is entering a more disciplined phase. That is good for the field. The next generation of successful companies will be those that combine ambitious science with practical therapeutic execution.
