By Florian Siebzehnrubl
Brain cancers are fortunately rare tumours in adults. Nevertheless, the most common type of brain cancer, glioblastoma, is incurable and almost always lethal. Because of the lack of effective therapies, the average survival with glioblastoma is only 15 months. US senator, John McCain suffered from this form of tumour, as did former British MP Tessa Jowell, who was outspoken about the need for more experimental treatments.
Finding new and more effective therapies for glioblastoma is very difficult for many reasons. Glioblastoma cells can move far away from the main tumour into the healthy brain, which makes complete surgical removal of the cancer impossible. These cancers are typically also very resistant to current drugs and radiation therapy. As a result, glioblastomas usually regrow after treatment, and these “recurrent” tumours tend to resist all efforts to treat them and ultimately cause the patient’s demise.
The so-called “blood-brain barrier” prevents many anti-cancer drugs that work well in other organs from entering the brain in the first place. This severely limits the number of drugs that can be applied. Glioblastomas are also very dissimilar between patients, and no genes are known that are mutated in all glioblastomas. Even within a tumour in a single patient, there are large differences in the tumour cells themselves. It is these differences between tumour cells that we wanted to understand better.
Feeding on fats and sugars
In a recent study, we compared glioblastoma cells and how fast they divided. And we saw that there were remarkable differences in the division rate of these tumour cells.
This led us to classify glioblastoma cells into faster-dividing and slower-dividing cancer cells and we investigated these further. Intuitively, we expected faster-dividing cells to be more aggressive, as they cause a tumour to grow more rapidly. But surprisingly, we found that glioblastoma cells that are dividing slower were more likely to resist chemotherapy.
Likewise, slower cells were more likely to move away from the main tumour, making them harder to remove surgically. Genetic signatures of these slower cells resemble signatures of “recurrent” glioblastoma, which indicates that slower cells may in fact be the ones that cause tumours to regrow.
Interestingly, the genetic signatures of “recurrent” glioblastomas and slower dividing glioblastoma cells were similar and revealed an increased dependence on fats for energy consumption of these tumours.
Since the 1920s it has been hypothesised that cancers predominantly rely on a less effective way than normal cells of using sugars for their energy support. This is known as the Warburg effect.
Our research shows this to be true for faster-dividing cells, but not for slower cells, which mainly use fats as their source of energy. In experimental models, we could show that withholding sugars affects only the faster-dividing cells. Conversely, in experiments where we blocked the metabolism of slower-dividing cells, their faster counterparts continued to grow unchecked. When we combined treatments to stop both faster and slower dividing cells in experimental models, we found that this was most effective to stop the tumours from growing.
In our study, we also identified a certain transport protein for fatty acids that is specific to slower-dividing glioblastoma cells. An important question arising from our work is whether this transport protein carries specific types of fatty acids into these tumour cells, and if this can be stopped for future therapies of glioblastoma. We used an experimental drug to block this transport protein, which was effective in our models in combination with treatments targeting faster dividing cells.
It is too early to tell whether this drug could one day be effective in the clinic, but based on our findings, it seems a question that is worth exploring.
Florian Siebzehnrubl is Research Fellow, European Cancer Stem Cell Research Institute at the Cardiff University.
Stay updated with all the insights.
Navigate news, 1 email day.
Subscribe to Qrius