Researchers at the University of California San Francisco (UCSF) have discovered how cancer cells evade targeted therapy using an escape mechanism called phenotype switching, potentially explaining the failure of precision therapies for glioblastoma to date.
Glioblastoma is a type of incurable brain cancer with a median survival rate of under two years. The tumors are highly resistant to treatment and difficult to operate on due to their location in the brain. Standard of care for the disease usually involves a maximal safe resection followed by radiotherapy and chemotherapy. However, success rates are low and tumors tend to return.
Reporting in Nature Cancer, scientists at UCSF have identified a possible reason for the recurrence. Using 86 primary-recurrent, patient-matched, paired glioblastoma samples from a UCSF biobank containing tumor tissue and adjacent non-malignant tissue, the team analyzed communications between the two types of cells.
Upon comparison, the researchers found that non-malignant nerve cells in the tumor adjacent tissue called glia expressed pro-growth signals causing the cancer cells to regrow. These cell extrinsic signals in the tumor microenvironment led to mesenchymal transition, tumor recurrence and finally therapy resistance.
The scientists identified this as a stress response following standard glioblastoma therapy called phenotype switching due to the change in cellular phenotype from terminally differentiated back to a mesenchymal, cancerous and radiation-resistant state. According to the researchers, the pathway responsible for this change was stimulated by the activator protein (AP1).
“We asked if there is another mechanism that explains therapeutic resistance,” said Aaron Diaz, PhD, associate professor of neurological surgery at the UCSF Weill Institute for Neurosciences and senior author of the study. “Our study concludes that, rather than a genetic evolution, there is a phenotypic plasticity or transition which allows these cells to evade therapy.”
For each tumor biopsy the scientists performed single-nucleus RNA sequencing measuring the transcriptome-wide gene expression in individual cells. Using this technique, they were able to compare expression of signal receptors to their known ligands between the different tissues and identify the drivers of tumor recurrence in glioblastoma. Finally spatial transcriptomics was conducted to validate the results.
“This is the first single-cell longitudinal study of this scale in glioblastoma. It’s also a study that could be done only at UCSF, because it represents decades of careful biobanking of surgical specimens. Since all the specimens came from UCSF, we know that the treatment histories are homogeneous, in that each patient received only standard-of-care therapy. It’s both this cohort’s scale and treatment uniformity that enable us to see past patient specific effects to the underlying biology of the disease,” Diaz concluded.