Delivering a direct hit to tumours

The effective delivery of chemotherapy drugs directly to tumours is a long-held goal. While many drug-delivery systems for cancers are successful, there is significant room for improvement, particularly for treating aggressive cancers.

Yasuhiro Matsumura, former chief of developmental therapeutics at the National Cancer Center in Kashiwa, Chiba, Japan, has dedicated his career to designing and improving drug-delivery systems for targeted cancer therapies.

“Cancer drugs are, by necessity, highly potent and toxic, which is why people often feel so unwell when undergoing chemotherapy,” says Matsumura, who is also the founder of RIN Institute Inc. in Tokyo, Japan. “To minimize damage to healthy cells and the resulting side effects, we have to find delivery methods that ensure drugs accumulate only in tumour cells.”

Scientists now know that healthy dividing cells share growth factors with cancer cells, so molecular-based targeting agents that inhibit cell-proliferation signals have limitations as cancer treatments.

“It’s vital that drug development for cancers takes a step back from the microscopic level and considers the entire tumour environment,” says Matsumura. “A full understanding of macroscopic tumour tissues is vital for future cancer therapies to realize their full potential.”

Exploiting a natural effect

In 1986, Matsumura and his former boss, the late professor Hiroshi Maeda of the Kumamoto University School of Medicine, published a paper revealing that molecules of certain sizes were more likely to accumulate in large numbers in tumour tissues than in healthy tissues¹. This ‘enhanced permeability and retention’ (EPR) effect is thought to arise from the over-stimulation of blood vessel growth by tumour tissues, leading to abnormal vessel structures and unusual fluid dynamics.

This effect can help distinguish cancerous tissues from normal tissues, and many researchers have used it to deliver antibody-based drugs and other therapies directly to tumours.

While this approach works well in mouse models, problems have arisen in humans. “We began to notice that too few drug molecules were reaching their target tumour cells within solid tumour tissues in humans, which can reduce the antitumour effect,” explains Matsumura. “The path of the antibodies was blocked by structural barriers in human solid tumours called stroma, which are not present to the same extent in xenografted experimental tumours in mice.”

When a malignant tumour forms, the cancerous activity induces blood clotting, along with the associated deposition of insoluble fibrin, collagen and inflammatory cells. These components form a strong structural barrier, or stroma, around the tumour cells, protecting them and helping to hold the solid tumour together. Stroma are common across all cancer types, but their extent and thickness varies widely. Indeed, there is a direct link between stroma extent and malignancy, with more aggressive cancers having the strongest stroma presence.

“Stroma can block treatments from reaching and killing cancer cells,” says Matsumura. “We decided to focus on one component of stroma, insoluble fibrin, and see if we could unlock a way through stroma to reach the cancer cells. Crucially, these insoluble fibrin molecules are found only in malignant tissues, not in healthy tissues or in the chronic phase of non-malignant diseases.”

A weak spot in stroma

Matsumura developed a monoclonal antibody specific to insoluble fibrin in January 2006. He and his team then searched for the epitope of the antibody on the insoluble-fibrin molecule. They identified a hollow on the side of the insoluble fibrin molecule with an amino-acid sequence of the epitope that was only exposed when insoluble fibrin is fully formed². It was also clarified that in its precursor or degraded forms, epitope peptides are tightly closed by hydrophobic bonds with their neighbouring peptide.

“This epitope is fully conserved across all species, from fish to humans — I hazard a guess that even the dinosaurs had it,” says Matsumura. “We’re confident that targeting this epitope will work in humans, though we have yet to reach clinical trials.”

This discovery enabled the team to create an antibody–drug conjugate (ADC) that can carry and release a chemotherapy drug molecule (monomethyl auristatin E, or MMAE) that is small enough to slip through the stroma to reach the tumour cells³.

The ADC is designed to behave in a certain way at each step of the delivery process. When it arrives at the insoluble fibrin site, the antibody binds to the epitope and the linker molecule holding MMAE is degraded by an enzyme called plasmin, which is activated only where insoluble fibrin is present. MMAE then passes through the stroma and attacks the cancer cells.

“Only insoluble fibrin releases plasmin, meaning that MMAE is released only by the interaction with the insoluble fibrin epitope, and nowhere else,” says Matsumura. “This makes our ADC highly specific and potentially safer than other delivery systems for chemotherapy drugs.”

This cancer stromal targeting, or CAST therapy, holds great potential for multiple cancer types, particularly those aggressive cancers that are currently difficult to treat. To optimize the effectiveness of the ADC, the team has recently developed clones of the original antibody⁴.

From mice to humans

Matsumura’s team is conducting extensive experiments using mice models and human cell cultures to verify that their ADC is stable and safe. They developed a probe that enables them to visualize insoluble-fibrin-rich stroma in mice and observe how their ADC performs. Recently, they have succeeded in producing an ADC that exerts a higher antitumour effect compared to the original ADC by improving the linker part.

Mouse model trials using the ADC against glioblastoma, a brain cancer which develops fast growing, highly aggressive tumours, have shown promise. Different from mouse xenografts, the researchers took care to ensure that the patient-derived-xenografted tumours were stroma rich, mimicking those in humans. The ADC remained stable in the bloodstream, accumulated specifically at the insoluble-fibrin site, where the drug was released and passed through stroma. Their treatment had marked antitumour effects, but the dosage had to be carefully monitored to prevent the mice losing weight. They didn’t observe tissue damage or haemorrhaging even at thrombin-formed thrombus site during the trial. Further toxicity studies are planned, together with investigations into optimal dosages.

CAST therapy requires further animal trials and safety profile tests before it can be tested in humans. Matsumura and his team are aiming to bring CAST therapy to human trials in three years, funding permitting.

“I’m also intrigued to see whether CAST could work well in combination with immune checkpoint inhibitors whose targets exist in the stroma,” says Matsumura. “I believe our unique therapeutic system could transform future cancer treatments, as well as open doors for treating other diseases involving blood clotting.”