Trastuzumab deruxtecan kills skin and uveal melanoma in vivo by HER2-independent payload release

Cancer treatment often relies on the idea of a "smart bomb": a drug designed to seek out a specific marker on the surface of a tumor cell, attach itself, and then deliver a lethal payload directly inside. One such marker is a protein called HER2, which acts like a switch on the cell surface, telling the cell to grow and divide. When this switch is stuck in the "on" position, as it is in certain breast and stomach cancers, it drives the disease forward. Scientists have developed powerful medicines that target this switch, including a type of drug known as an antibody-drug conjugate. These medicines work like a guided missile; the antibody part finds the HER2 switch, and once attached, the drug releases a toxic chemical inside the cell to kill it. For years, the prevailing belief was that for this strategy to work, the tumor needed to have a lot of these switches on its surface. If a cancer had very few or no HER2 switches, doctors assumed the smart bomb would simply fly past, unable to find its target.

This assumption is now being challenged by new research involving a particularly difficult-to-treat cancer called melanoma. While melanoma is well known for appearing on the skin, it can also develop in the eye, specifically in the uvea, the middle layer of the eye. Uveal melanoma is aggressive and often spreads to the liver, where it is notoriously hard to treat. Researchers in Gothenburg, Sweden, set out to test whether a specific HER2-targeting drug, known as trastuzumab deruxtecan, could work against these eye and skin cancers, even though these tumors typically have very low levels of the HER2 protein. They wanted to see if the drug could kill the cancer cells by latching onto the few switches present, or if something else entirely was happening.

The team began by growing tumors from human patients inside mice. They took samples from patients with uveal melanoma and implanted them under the skin of the mice, creating a living model of the human disease. When they treated these mice with the drug, the results were striking. The tumors in the mice shrank dramatically, and in some cases, they disappeared completely. Encouraged by this success in the mice, the researchers treated the original human patient who had provided the tumor sample. The patient had exhausted all other treatment options and was receiving the drug through a compassionate use program. While the drug did not completely eliminate the patient's tumors, it did slow their growth and lowered a marker of disease activity in the blood, suggesting the drug was having an effect, even if the high volume of cancer in the liver made it harder to see a full recovery.

However, the story took a surprising turn when the scientists looked closer at how the drug was working. They noticed that the drug seemed to work much better on tumors growing under the skin than on those growing in the liver, hinting that the location of the cancer mattered. More importantly, they tested whether the drug actually needed the HER2 switch to work. They created melanoma cells that had been genetically edited to completely remove the HER2 protein. Logic suggested that without the target, the drug should be useless. Yet, when these "switch-less" tumors were grown in mice and treated with the drug, they initially shrank just as much as the tumors that still had the switch. The tumors without the switch eventually started growing back sooner, but the initial killing power of the drug did not depend on the presence of HER2. This finding ruled out the idea that the drug was simply finding and destroying the few HER2 switches on the surface of the cells.

To understand what was actually happening, the researchers looked at the environment inside the tumor. They discovered a specific enzyme, a type of protein that acts like a pair of molecular scissors, called cathepsin L. In healthy cells, this enzyme is usually locked away inside a storage compartment. But in these melanoma tumors, the researchers found that the enzyme was leaking out into the space between the cells. This enzyme is known to be able to cut the chemical link that holds the toxic drug to its antibody carrier. The researchers suspect that this enzyme, floating freely outside the cancer cells, is snipping the drug loose from its guide. Once the drug is free, it can enter nearby cancer cells and kill them, regardless of whether those cells have the HER2 switch or not. It is as if the drug, instead of needing a specific key to enter a house, is being dropped off by a delivery person who cuts the package open right on the doorstep, allowing the contents to seep into the house through the open door.

The researchers confirmed this theory by testing the cells in a dish, outside of the body. In a controlled environment with a neutral pH, the melanoma cells refused to take up the drug and were completely resistant to it. They did not absorb the antibody, and the drug did not kill them. This contrasted sharply with breast cancer cells that have high levels of HER2, which eagerly absorbed the drug and died quickly. The difference was that inside the living mouse, the tumor environment provided the enzyme needed to release the drug, but the isolated cells in the dish did not. This explains why the drug worked in the mice but failed in the lab dishes: the living tumor creates a specific chemical environment that activates the drug in a way that the simple lab setup cannot replicate.

The study also compared this drug to another similar medicine that targets HER2 but uses a different delivery mechanism. That other drug, which relies more heavily on the cell absorbing the whole package, showed very little effect on the melanoma tumors. This suggests that the unique ability of the first drug to release its payload outside the cell, driven by the enzyme, is the key to its success in these specific cancers. The findings suggest that for this type of melanoma, the amount of HER2 on the surface of the cell might not be the most important factor in deciding if a patient will respond to treatment. Instead, the presence of the enzyme that cuts the drug loose might be the true indicator of success.

While the results offer a new way to understand how these drugs work, the researchers caution that the story is not fully finished. The drug did not work as well in the liver models as it did under the skin, indicating that the location of the tumor and the specific environment it creates play a major role. The enzyme that helps release the drug is known to be more active in low-oxygen conditions, which are common in fast-growing tumors, but this activity can vary widely from one tumor to another. The researchers propose that looking for this enzyme might be a better way to predict who will benefit from the treatment than looking for the HER2 switch. This shifts the focus from the target on the cell surface to the chemical environment surrounding the tumor, opening a new path for understanding how these powerful medicines can be used against cancers that were previously thought to be out of reach.