Summary of Key Points
This article outlines the evolution of animal models in cancer research over the past century: from the early use of chickens and rabbits to reveal the truth that viruses/chemical substances can cause cancer, to nude mice allowing human tumors to thrive in a mouse host, to genetically engineered mice that can simulate the entire process of cancer development, and finally to the use of spontaneous tumors in pets such as cats and dogs to more accurately mimic human cancer. These models have gradually transformed cancer from an unpredictable disaster into a disease that can be studied, replicated, and manipulated, providing crucial tools for cancer research.
1. Chickens and Rabbits: The First Step in Cancer Research
In 1911, a chicken with a tumor, donated by a farmer's wife, led scientist Rous to conduct a groundbreaking experiment: he ground the tumor and filtered it to remove all known bacteria, then injected it into a healthy chicken. A few weeks later, the healthy chicken developed the same tumor! This demonstrated that cancer could be caused by viruses smaller than bacteria (later identified as the Rous sarcoma virus). However, the prevailing medical community believed that cancer was due to cellular metabolic disorders, and Rous's discovery was ignored for 55 years before he was awarded a Nobel Prize at the age of 87.
Almost simultaneously, Japanese scientist Katsushiro Yamaguchi observed that workers exposed to coal tar were prone to skin cancer. Upon returning to Japan, he repeatedly applied coal tar to the ears of rabbits, and five months later, malignant tumors appeared—this was the first time cancer was induced in animals using a chemical substance. Yamaguchi nominated for the Nobel Prize four times but was not awarded; in 1926, the prize went to a researcher who studied the carcinogenic effects of parasites (later disproven). Nevertheless, his work is considered a cornerstone of cancer research.
These experiments shattered the notion that cancer was a random disaster, showing that it has clear causes and can be replicated in the laboratory.
2. Nude Mouse Models: Enabling Human Tumors to Survive in Mice
In 1962, a scaly, sickly mouse was discovered in a Scottish laboratory; it turned out to have a congenital absence of the thymus, preventing the maturation of T cells (the immune system's fighters), resulting in little rejection of foreign tissues. In 1969, Danish scientists transplanted human colon cancer tissue into this nude mouse, and the tumor survived! This marked the beginning of the "patient-derived xenograft (PDX)" model era.
However, nude mice still possess B cells and NK cells, so not all tumors could be successfully induced. Scientists continued to improve these models: SCID mice lack T/B cells, NOD-SCID mice are even more immunocompromised, and NSG mice have nearly complete immune deficiencies, allowing human immune cells to grow within them for testing of immunotherapy drugs (such as PD-1 inhibitors).
PDX models retain the original characteristics of patient tumors but are costly and time-consuming, so they are often used in conjunction with cell line models for rapid drug screening.
3. Genetically Engineered Mice: "Customizing" the Cancer Development Process
Nude mice allow tumors to be introduced, but human cancer develops from normal cells. Genetically engineered mice have solved this issue by enabling tumors to develop in vivo through gene editing:
- Early Attempts: In 1984, the OncoMouse model was created by placing a cancer gene under a breast promoter, causing tumors to form spontaneously in the mouse's mammary glands—proving that activating a single cancer gene can lead to cancer.
- Tumor Suppressor Gene Models: In the 1990s, mice with the p53 gene knocked out (a guardian against cancer) developed multiple tumors quickly, showing that the absence of tumor suppressor genes can also cause cancer.
- Precise Control: The Cre-LoxP technology allows for targeted gene activation in specific organs at specific times. For example, the KPC model activates KRAS mutations and deletes p53, resulting in pancreatic cancers highly similar to those in humans.
- CRISPR Acceleration: While it used to take years to create such models, CRISPR can now edit multiple genes within weeks, allowing researchers to observe the effects of different mutation combinations.
These models enable the observation of the entire process of cancer development, including resistance and immune evasion.
4. Pets (Cats and Dogs): The Most Realistic Cancer Models
Laboratory mice develop tumors artificially, which differ from those in humans, who are exposed to complex environments (pollution, stress). Scientists turned to cats and dogs, as they share the same diet and living conditions, leading to similar cancer types:
- Canine Osteosarcoma: The osteosarcomas in large dogs (such as Golden Retrievers and Rottweilers) have molecular similarities to those in human teenagers. Immunotherapy studies on dogs have been translated to clinical trials in humans.
- Feline Breast Cancer: Feline breast cancer is often hormone receptor-negative, similar to the most difficult-to-treat triple-negative breast cancer in humans. In 2026, the journal Science reported that feline breast cancers with the FBXW7 mutation have the same prognosis as human cases.
The advantages of cats and dogs include spontaneous tumors, genetic diversity, an immune microenvironment closer to that of humans, and a short lifespan (which allows for early evaluation of treatment effects). This approach also supports the concept of "one health" by benefiting both pets and humans.
5. The Future: Will Animal Models Be Replaced?
EU data shows a decline in the number of experimental animals used in cancer research between 2015 and 2020 due to the emergence of new technologies:
- Organoids: Small amounts of patient tumor tissue are cultured into three-dimensional cell clusters that can be used to test drug responses within 2–4 weeks, allowing for the simultaneous screening of hundreds of drug combinations in 384-well plates.
- Organ Chips: Tiny chips the size of a fingernail simulate human organs; for example, lung chips can mimic lung expansion during breathing to test the effectiveness of lung cancer drugs.
- AI Digital Twins: Virtual models containing genomic and proteomic data are created for each patient, combined with organoids to quickly identify optimal treatment options.
These technologies will reduce the need for animal experiments, but animal models will not be completely replaced. Complex immune responses and tumor microenvironments still require live animals for research. The future may see a combination of animal models and new technologies to advance cancer research.
Conclusion: From chickens and rabbits to cats and dogs, from nude mice to genetically engineered mice, animal models have helped us move from ignorance to a better understanding of cancer. In the future, new technologies will work alongside animal models to continue tackling this challenging disease.