Chemotherapy has been our main weapon against late-stage cancers for decades, bringing with it wide-ranging side effects caused by the subsequent damage to healthy, growing cells. As research into targeted therapies has advanced, pharmaceutical companies have begun pairing both approaches in one, delivering powerful chemotherapy agents in combination with monoclonal antibodies (mAbs) that bind to antigens on the surface of cancer cells.

Antibody drug conjugates (ADCs), as they are known, now represent an exciting class of emerging cancer therapies, and according to Jerzy Woźnicki, PhD, Scientific Project Lead at Turbine, “it’s the story of a niche idea becoming one of the most dynamic pillars of modern oncology.”

“Over the past five to seven years, we have reached an inflection point,” Dr Woźnicki says. “Approvals like Enhertu, which has transformed outcomes in HER2-positive and HER2-low breast cancer, and Trodelvy, which expanded treatment options for triple negative breast cancer, a disease where TROP2 is broadly expressed, have redefined our expectations of what ADCs can do in solid tumours.”

ADCs in 2025

For 2025, the biggest success story came from Daiichi Sankyo, which brought Datroway to market for previously treated, locally advanced EGFR-mutated non-small cell lung cancer via the FDA’s accelerated approval pathway. Datroway demonstrated an overall response rate (ORR) of 45% and a median duration of response of 6.5 months in clinical trials, placing it above many historic ADCs (which often have ORRs <30%).

Meanwhile, in late 2025, GlobalData’s Drugs database shows 834 ADCs in preclinical development around the world, as well as 131 with an IND/CTA filed, 252 in Phase I, 184 in Phase II, and 68 in Phase III.

“ADCs are moving decisively into earlier lines and even to curative settings,” notes Dr Woźnicki. “We also see that ADCs are expanding horizontally into new tumour types, showing that this modality is truly an adaptive one. Across meetings like AACR, ESMO and World ADC, the conversation has shifted from whether ADCs can work, to how do we broaden what ADCs can be?”

But while 2025 was a defining year for ADCs, challenges are still front and centre.

Bottlenecks persist

“Even with all the momentum that we’re seeing, ADCs still run into the same biological bottlenecks,” says Dr Woźnicki. “Many ADC targets also show low levels of expression in healthy tissue, so even the most successful ADCs still bring interstitial lung disease, neutropenia or gastrointestinal toxicity – and as we move into earlier-stage, curative settings, the tolerance for toxicity becomes even narrower.”

Another challenge is tumour heterogeneity. “Single payload ADCs only hit what they see, meaning that low-expressing or antigen-poor subclones survive and can become the root of relapse,” adds Dr Woźnicki. “Bystander effect can indeed help, but increasing it also increases off-tumour exposure, so you end up trading gaps in coverage for higher systemic toxicity.”

Resistance is another issue. Most approved ADCs rely on just two payload families (TOPO1 inhibitors and microtubule disruptors) which tumours can adapt to, leading to shorter and less durable responses.

Combining ADCs with other drugs such as immunotherapies or hormone inhibitors has been looked at as a potential solution to some of these challenges. However, while many combinations looked promising in preclinical studies, increased toxicity became a major hurdle in clinical trials.

The dual payload promise

Dual payload ADCs use a single monoclonal antibody to deliver two cytotoxic payloads with distinct mechanisms of action to the tumour site. In light of the pitfalls of conventional ADCs, it’s easy to see why two payloads could be much better than one.

“By default, dual payload ADCs are designed for the biological realities that single payload ADCs struggle with,” Dr Woźnicki explains. “The first promise here is addressing heterogeneity directly, since one payload no longer fits all. The second promise is overcoming resistance. Tumours can adapt quickly to TOPO1 inhibitors and microtubule disruptors, and if your ADC carries two distinct mechanisms, you can make this escape much harder.”

Dual-payload ADCs are also enabling researchers to balance potency and safety in a “more intelligent way”, Dr Woźnicki adds. Many teams across the ADC discovery field are currently designing ADCs where a highly potent payload is paired with a more forgiving one, enabling them to broaden tumour coverage without pushing toxicity too high.

New possibilities continue to expand, with companies exploring pairing degraders with immune regulators and micro-environment modulators, potentially reshaping how ADCs drive anti-tumour immunity.

In 2025, the first dual-payload ADCs entered clinical trials, including Chengdu Kanghong Biotech’s KH815 and Innovent Biologics’ IBI3020. Chengdu Kanghong’s Phase I trial is set for completion in June 2027, while Innovent’s is set for the following December. According to GlobalData estimates, IBI3020 has a 50% likelihood of progressing to Phase II in colon cancer.

Sutro Biopharma is also at the fore, working on immunostimulatory candidates like STRO-227, which is targeting an IND submission in 2026 or 2027.

Double the complexity

Dual-payload ADCs have been recognised as the next frontier in cancer research, but the development challenges are significant. “If you think that a single payload ADC requires a very careful alignment of biology and chemistry, dual-payload ADCs double that complexity. You’re essentially designing two ADCs that must function as one.”

The first challenge is biology. Researchers need to know which two payloads belong together, and this depends on whether they are meant to target heterogeneity, resistance, or both. Without this biological rationale, the molecule lacks translational value.

Chemistry and design are also challenging, with different payloads requiring different linkers and conjugation methods. Unfortunately, safety is harder to predict, since the combination of two payloads can lead to unexpected toxicities depending on factors such as ADC internalisation rate, trafficking, or processing.

Finally, the bar is raised when it comes to outcomes. “To justify their complexity, dual-payload ADCs must outperform what single-payload ADCs can achieve,” explains Dr Woźnicki. “They should bring deeper responses, longer durability, or a clear benefit in heterogenous or resistant tumours.”

In essence, while dual-payload ADCs could allow the industry to solve some of the challenges that single payloads simply aren’t built to do, they also amplify many of the development challenges with conventional ADCs. So, how do we solve for that?

Can simulated assays help?

When ADCs succeed, it is because their design is truly fit for the patient’s underlying biology rather than utilising clever linkers or novel designs arbitrarily. To help the industry achieve this, Turbine has developed the world’s first Virtual Lab platform, augmented by their class-leading virtual cell technology and over ten years of practical experience. The platform simulates millions of therapeutic perturbations of drug effects and genetic changes on virtual disease models.

“We enable our collaborators and internal teams to explore payload biology at scale through virtual screens across 1,200+ cancer cell line models readily, with the capacity to extend this by hundreds of patient-derived xenograft (PDX) models,” explains Dr Woźnicki. “This breadth of different genetic backgrounds, underlying signaling states and drug vulnerabilities lets you see patterns very quickly.”

Within hours, it becomes possible to understand which tumours might respond to, for example, TOPO1 inhibition, which are sensitive to microtubule disruptors, versus which are intrinsically resistant to both those classical mechanisms. As opportunities for dual-payload ADCs emerge, Turbine can also highlight which alternative or novel mechanisms would be biologically relevant in such specific tumour contexts.

Dr Woźnicki adds that these predictions don’t exist in isolation but, crucially, are part of a “continuous lab-in-the-loop system that learns and improves with every iteration.” With each experiment feeding back into the continually improving model, its predictions are constantly growing smarter, more accurate, and broader in scope.

Overcoming resistance with AI

“We also go beyond predicting responses,” adds Woźnicki. “Our modelling can separate true drivers of response from just trivial correlations, allowing us to mechanistically trace the underlying biology. Because we can also simulate drug modifier screens (essentially CRISPR-like perturbations run in silico), we can map the resistant routes a tumour might use.”

Taken together, these layers of information can reveal which additional mechanisms have the greatest chance of successfully closing the resistance route, rather than simply adding toxicity or hitting the same vulnerability twice.

Positioning for synergy is another core strength of the platform. This means the technology not only predicts whether two distinct mechanisms of action could synergise, but also identifies in which cancer subtypes this synergy is likely to be strongest.

“In practice, that means that teams can explore dozens of payload bearings across the virtual library of cancer samples and PDX models, rank them by synergy strength and indication fit, and then only take the most promising constructs into the lab,” explains Dr Woźnicki. “It’s truly the essence of a fit-for-biology approach.”

As this foundation strengthens, the company is also developing towards offering full ADC modelling in the first half of 2026, bringing in other biological features that shape drug responses, from antigen presentation and downstream signaling to resistance mechanisms like payload efflux.

“Once you model an ADC end-to-end, you can start asking deeper questions,” says Dr Woźnicki. “What happens when two ADCs are combined, or when an ADC is paired with a small molecule inhibitor that hits a complementary pathway, for instance?”

Turbine’s ADC Simulation Suite marks a new era in ADCs, accurately predicting drug responses, matching novel payloads with responder indications, and enabling dual-payload ADCs to be designed rationally rather than by trial-and-error. Download the poster below to learn more.