Two Preclinical Studies Test Liver and Colorectal Cancer Strategies

Two independent preclinical studies have produced new ways to investigate liver and colorectal cancer, but neither offers a treatment ready for patients. Rockefeller University researchers followed virus-driven liver disease in mice for 18 months. A Cedars-Sinai team tested an experimental compound called C3 in cancer cells and several mouse tumor models.

The projects address different problems. The Rockefeller work creates a living model of how chronic viral hepatitis can progress through inflammation and fibrosis to hepatocellular carcinoma, or HCC. The Cedars-Sinai study targets a protein network that supports growth in liver and colorectal cancer cells.

Both studies move laboratory research forward because they give scientists something concrete to test. Their results remain tied to animal and cell experiments, however, and do not establish a diagnostic test, a safe human dose or a clinical benefit.

A Virus Model Followed Liver Disease for 18 Months

Human hepatitis C virus does not naturally infect ordinary mouse liver cells, which has made the full course from infection to liver cancer difficult to reproduce in a small animal. The Rockefeller team used a mouse-adapted Norway rat hepacivirus, or NrHV, a relative of hepatitis C that infects liver tissue.

Researchers temporarily depleted CD4 T cells before infection so that the virus could become established. CD4 T-cell levels recovered within about a month, and the animals otherwise retained functioning immune systems during the long follow-up. The team used two cohorts of C57BL/6J mice, monitored infection and liver injury over time, and examined liver tissue at intervals up to 18 months.

More than 60% of infected mice still carried the virus at 18 months. Their livers developed inflammation, progressive collagen accumulation and molecular changes that overlapped with patterns seen in human hepatitis C. The resemblance was substantial but not complete: most infected mice did not develop full cirrhosis, and the model depends on an engineered virus and temporary immune manipulation.

At 18 months, 67% of infected animals had histologically confirmed HCC, compared with 4% of mock-infected controls. The rate also differed by sex. Across the two cohorts, 15 of 17 infected males and seven of 16 infected females developed HCC. Tumors were usually well differentiated and often showed immune-cell infiltration, features that resemble one class of human HCC.

The model gives investigators a way to study how persistent infection, immune activity, fibrosis and cancer interact over time. It may also support future tests of antifibrotic, antiviral and anticancer strategies. It does not show that a biomarker has been validated or that a therapy works in people.

C3 Disrupted a Cancer-Related Protein Scaffold

The Cedars-Sinai study focused on GIT1 and MAT2B, proteins found at unusually high levels in liver and colorectal cancers. Together they form part of a scaffold that supports the RAS-RAF-MEK-ERK signaling pathway, which helps regulate cell growth. The researchers asked whether a small molecule could interfere with that scaffold.

Because a complete crystal structure for GIT1 was unavailable, the team built a computer model of an accessible domain and screened chemical candidates. Nine compounds were tested experimentally, and C3 was selected for more detailed work. Binding experiments indicated that C3 interacted with GIT1 rather than the related GIT2 protein.

In liver and colorectal cancer cell lines, C3 reduced GIT1-MAT2B interaction, lowered downstream MEK and ERK activity, limited colony formation and migration, and arrested cells during mitosis. Its growth-suppressing effect weakened when researchers first reduced GIT1 expression, supporting the proposed target mechanism.

The compound was then evaluated in three mouse settings. These included subcutaneous mouse colorectal tumors, human colorectal cancer cells implanted in the livers of immunodeficient mice, and a colorectal liver-metastasis model in mice with intact immune systems. C3 slowed tumor signals in these experiments and produced no overt toxicity under the tested schedules.

Those safety observations are preliminary. The tumor studies generally involved five or six animals per group, treatment lasted days or weeks, and the investigators did not use statistical methods to determine sample size in advance. “No overt toxicity” in a short mouse experiment cannot establish long-term safety in humans.

The Two Results Should Not Be Blended

The Rockefeller model concerns primary liver cancer that follows chronic hepacivirus infection. The C3 experiments concern a molecular target shared by liver and colorectal cancer cells, while the in vivo treatment work centered on colorectal tumors and colorectal cancer spreading to the liver. A colorectal metastasis in the liver is not the same disease as HCC arising from liver cells.

That distinction matters when assessing what has been demonstrated. The NrHV study did not test C3, and the C3 study did not use the 18-month virus model. Neither project showed that the compound prevents virus-driven liver cancer. Presenting the two papers as one treatment pathway would create a connection the experiments did not test.

The studies also answer different scientific questions. A disease model is valuable when it reproduces enough of human pathology to reveal mechanisms or compare interventions. A candidate compound is valuable when its target, activity, exposure and toxicity can be reproduced across increasingly demanding experiments. Neither type of result is a patient outcome.

Promising Tools Still Need Harder Tests

The next steps for the NrHV model are comparative. Researchers need to show which findings persist across laboratories, how the temporary CD4 depletion affects later disease, and whether therapies that succeed in the model predict anything useful about human HCC. Differences in cirrhosis, telomere biology and viral species remain important limits.

C3 faces a different path. Its chemistry must be refined, its binding structure and possible unintended effects need fuller study, and longer toxicology and dosing work would be required before a human trial. The mouse experiments used routes and schedules designed for proof of concept, not a clinical regimen.

The strongest conclusion is therefore about research capacity, not treatment success. One team created a model in which chronic hepacivirus infection can progress to spontaneous liver cancer. Another identified a compound that interrupted cancer-related signaling in cells and reduced tumor growth in small mouse studies. These are credible preclinical advances. Their value to patients will depend on the evidence produced next, not on calling the first result a breakthrough.