Liver cancer stem cells shown to use immune system as shield to spark disease recurrence | News Center

Liver cancer cells that rub shoulders with a class of immune cells called macrophages promote treatment-resistant recurrence of the cancer, according to a study led by Stanford Medicine researchers. Two drugs that restrain that cozy relationship eliminated the liver cancer cells in two mouse models of recurrent liver cancer and slowed or prevented tumor recurrence in the animals.

The study used techniques developed at Stanford Medicine to precisely map the location of more than 1 million single cells in more than 100 human liver tumor samples after initial chemotherapy treatment. Although this form of treatment, called chemoembolization, is often successful — leading to no evidence of cancer — the tumors frequently return within a few months.

“We wanted to know where these residual tumor cells are hiding,” said Renumathy Dhanasekaran, MD, assistant professor of gastroenterology and hepatology. “You can’t see them on a scan, and there are no biomarkers in the blood we can track. Our study shows that the location of the cells in the tumor — their biological ZIP code — matters. We found that tumor cells enlist nearby macrophages to shield them from attack by other immune cells.”

Sheltering in place, the tumor cells function as cancer stem cells, lying dormant and undetectable for weeks or months before waking to promote the growth of a new tumor.

Dhanasekaran, who is a member of the Stanford Cancer Institute, and professor of medicine and of pathology Dean Felsher, MD, PhD, are the senior authors of the study, which was published Sept. 20 in Nature Cancer.  Postdoctoral scholars Lea Lemaitre, PhD; former medical student Nia Adeniji, MD; and former undergraduate student Akanksha Suresh are the lead authors of the paper.

Sifting through a million cells

The researchers had to overcome several obstacles to conduct the study. Studying residual disease in a solid tumor such as liver cancer is difficult because the tissue remaining after treatment is mostly made up of dead cells, so many samples are needed. And identifying not just the precise location of every cell relative to one another — a cellular milieu known as the tumor microenvironment — but also the genes and proteins each cell is producing has been made possible only by recent computational and machine-learning advances. 

The researchers drew on biopsies of 108 previously treated human liver tumors conducted at Stanford Medicine, comprising more than 1 million cancer cells, and used a spatial mapping technique developed at Stanford Medicine called co-detection by indexing (CODEX) to identify 11 major cell types in the tumors. They picked three cell types — tumor cells, macrophages and the immune cells known as T cells — for further study.