A pair of research articles published May 9 in Nature Genetics shed new light on the cellular complexity of glioblastoma, the most aggressive type of brain cancer. An international team of scientists, including researchers at the Yale Cancer Centre, analysed tumour samples from 59 glioblastoma patients to better understand how diverse cell types within a tumour change over time and in response to standard therapy. Their findings identify previously unrecognised patterns of cancer cell activity and may help guide future treatment strategies for this disease.

The work described in the two articles was supervised by five senior researchers, including Roel Verhaak, PhD, Harvey and Kate Cushing Professor of Neurosurgery at Yale School of Medicine. His research has long focused on the identification and characterisation of gene expression subtypes in glioblastoma. The current articles build on Verhaak’s prior research by leveraging the latest genomic technologies.

“Using high-resolution technologies that enable us to measure gene expression at the single-cell level, we are now able to specifically pinpoint characteristics of glioblastoma cells and the mechanisms behind disease progression. This new work applies these technologies at scale, revealing the heterogeneity and evolution of this aggressive disease,” he says.

The multilayered transcriptional architecture of glioblastoma ecosystems

The first article presents a detailed analysis of 121 primary and recurrent glioblastoma samples from the 59 patients to discover cancer cell types not previously identified in earlier, smaller studies. The large data set included about 430,000 cells and led to the identification of three novel glioblastoma cell “states,” in addition to confirming previously identified ones, that may contribute to a glioblastoma’s ability to adapt and evade therapies.

Glioblastoma is different from patient to patient, and its cellular composition is varied even within the same tumour. Despite this variability, the researchers found some common cellular programs across patients that are influenced by specific gene mutations and a tumour’s surrounding cells. These common patterns define three overarching “ecosystems” that reflect distinct cellular communities.

“By dissecting glioblastoma at the single-cell resolution, we’re beginning to understand how individual cancer cells function collectively as an ecosystem. Mapping this cellular landscape gives us critical insights into how glioblastomas develop, evolve, and ultimately resist treatment,” says Kevin Johnson, PhD, research scientist in the Department of Neurosurgery at Yale School of Medicine and co-first author of the studies.

Deciphering the longitudinal trajectories of glioblastoma ecosystems by integrative single-cell genomics

The second article examined how the glioblastoma cellular ecosystems change between a patient’s initial diagnosis and disease recurrence. These findings both added to, and reinforced the existing complex picture of varied glioblastoma cells and how they evolve and develop resistance to treatment.

Most recurrent glioblastomas retain the cellular makeup associated with the primary tumour, but some do not. For example, tumours with higher levels of the gene MGMT, which is related to chemotherapy resistance, can transition to a more aggressive form when they recur. Another subgroup, with genetic patterns linked to therapeutic radiation resistance, displayed a low-oxygen (hypoxic) profile in the recurring tumour that may assist glioblastoma cells in surviving standard radiation therapy.

Together, the insights from these articles mark a major step toward decoding glioblastoma’s notorious complexity and treatment resistance mechanisms. The articles noted that further genetic studies are still needed to reveal additional recurrence patterns that could inform treatment decisions that improve patient outcomes.