Somewhere in a lab, someone looked at a chemotherapy drug and thought, "What if we strapped this to a homing missile instead of just carpet-bombing the entire body?" That someone was onto something. Antibody-drug conjugates - ADCs for short - are basically the cancer treatment equivalent of a GPS-guided delivery drone that drops a tiny bomb on exactly the right doorstep, then flies away.

A sweeping new review in Physiological Reviews by Cheung, Liu, and colleagues maps the entire ADC landscape: from the 14 FDA-approved drugs to the 430+ candidates now in clinical trials, plus the emerging role of AI in designing the next generation of these molecular weapons (Cheung et al., 2026). It's basically the encyclopedia of ADCs, and it paints a picture of a field that's sprinting.

The Three-Part Recipe

Every ADC has three ingredients: an antibody (the homing device), a linker (the leash), and a payload (the poison). The antibody locks onto a specific protein sitting on the surface of cancer cells like a bouncer checking IDs at a club. Once it binds, the whole package gets swallowed by the cancer cell. Inside, the linker breaks apart, the cytotoxic payload is released, and the cell's internal machinery gets thoroughly wrecked - either by shredding its DNA or jamming its ability to divide (Fu et al., 2025).

Simple enough, right? Except every piece of that chain took decades of engineering to get right. Early ADCs had unstable linkers that would release their toxic cargo too early - imagine your delivery drone dropping packages over random neighborhoods. Modern cleavable linkers respond specifically to the acidic, enzyme-rich environment inside tumor cells, while non-cleavable versions require complete cellular digestion before releasing their payload. The payloads themselves have gotten nastier too: topoisomerase I inhibitors (the class used in Enhertu, the ADC everyone's talking about) are 100-1,000 times more potent than standard chemo drugs.

The Bystander Who Showed Up to Help

Here's a twist the original ADC designers didn't fully expect: the bystander effect. When a targeted cancer cell dies from its ADC payload, the released drug can leak out and kill neighboring tumor cells that might not even express the target antigen. It's like ordering one pizza and the delivery person accidentally feeds the whole block. This has turned out to be therapeutically useful, especially in solid tumors where antigen expression is uneven - not every cancer cell wears the same molecular name tag.

Enhertu Changed the Game

If ADCs had a celebrity, it would be trastuzumab deruxtecan (Enhertu). This HER2-targeting ADC reduced the risk of breast cancer recurrence or death by 53% compared to the previous standard in high-risk patients, earning Breakthrough Therapy Designation from the FDA in 2025 (AstraZeneca/Daiichi Sankyo, 2025). Even wilder: the DESTINY-PanTumor02 trial showed it works across multiple cancer types, not just breast. Enhertu essentially proved that a well-designed ADC with the right linker-payload combination could rewrite treatment guidelines overnight.

The ADC market has noticed. Revenue jumped from $1.6 billion in 2017 to $7.9 billion in 2024, with projections hitting $64.7 billion by 2030 - a compound annual growth rate that would make most tech stocks jealous (Li et al., 2025).

AI Enters the Chat

Perhaps the most forward-looking section of Cheung et al.'s review covers AI-driven ADC design. Machine learning models are now predicting optimal drug-to-antibody ratios, identifying promising conjugation sites, and even screening antigen targets from pathology data before a single molecule is synthesized (Zhang et al., 2025). Deep learning can model three-dimensional antibody structures to predict binding affinity, essentially letting researchers test millions of molecular combinations in silico before committing to expensive wet-lab experiments. If you enjoy watching AI tackle complex spatial problems, mapb2.io applies similar visual-spatial thinking to mapping and diagramming - though the stakes are admittedly lower than curing cancer.

Next-generation designs getting the AI treatment include bispecific ADCs (antibodies that recognize two different targets), immunostimulatory ADCs that recruit the immune system alongside delivering their payload, and degrader-antibody conjugates that hijack the cell's own protein disposal system to eliminate cancer-driving proteins.

What's Still Hard

ADCs aren't magic. Resistance develops when cancer cells downregulate the target antigen, pump out the payload, or reroute their survival pathways. Toxicity remains a problem - some payloads still cause peripheral neuropathy, eye damage, or lung inflammation. And manufacturing these three-component molecules at consistent quality is a headache that would give any process engineer nightmares.

The review highlights combination strategies as a promising solution: pairing ADCs with immune checkpoint inhibitors like pembrolizumab has shown synergistic effects in early trials, where the ADC kills tumor cells and the released antigens prime the immune system for a broader attack. It's teamwork between a sniper and the cavalry.

The Bottom Line

With 15 FDA-approved ADCs and hundreds more in the pipeline, this field has moved from proof-of-concept curiosity to a pillar of modern oncology in about two decades. Cheung and colleagues provide the roadmap for where it's headed: smarter targets, more stable linkers, more potent payloads, AI-optimized design, and precision biomarkers to match the right ADC to the right patient. The carpet-bombing era of chemotherapy isn't over, but its replacement is being assembled one molecular Trojan horse at a time.

References

1. Cheung, A., Liu, Y., Chenoweth, A. M., Montaseri, H., Esapa, B., Chudasama, V., Baker, J. R., Thurston, D. E., & Karagiannis, S. N. (2026). Antibody-drug conjugate design and mechanisms of action for cancer treatment: state of the art and beyond. Physiological Reviews. DOI: 10.1152/physrev.00039.2025 | PubMed: 41855081

2. Fu, Z., Li, S., Han, S., Shi, C., & Zhang, Y. (2025). Antibody-drug conjugates in cancer therapy: applications and future advances. Frontiers in Immunology, 16, 1516419. DOI: 10.3389/fimmu.2025.1516419 | PMC12133739

3. Zhang, Y., et al. (2025). Leveraging artificial intelligence in antibody-drug conjugate development: from target identification to clinical translation in oncology. npj Precision Oncology. DOI: 10.1038/s41698-025-01159-2

4. Li, J., et al. (2025). Antibody-drug conjugates in cancer therapy: current advances and prospects for breakthroughs. Frontiers in Cell and Developmental Biology, 13, 1669592. DOI: 10.3389/fcell.2025.1669592 | PMC12540542

5. AstraZeneca/Daiichi Sankyo. (2025). ENHERTU reduced the risk of disease recurrence or death by 53% vs. T-DM1 in DESTINY-Breast05. Press Release

Disclaimer: This blog post is a simplified summary of published research for educational purposes. The accompanying illustration is artistic and does not depict actual model architectures, data, or experimental results. Always refer to the original paper for technical details.