Precision diagnostics to predict brain cancer

​It seems that a combination of several advanced examinations could accurately predict sites in the brain at the greatest risk of developing tumours in brain cancer patients. 

Is it possible to treat aggressive cancer in the brain more effectively by predicting precisely where the cancer tumour will return? This is what Michael Lundemann, MSc Eng. and postdoc at the Department of Clinical Physiology, Nuclear Medicine and PET is hoping for. 

With a brand new diagnostic model that combines several different types of examinations, he has helped create the conditions for more targeted treatment.

"Brain cancer is one of the most deadly types of cancer.  In recent years, there’s been progress in surgical treatment as well as medical treatment, but not to the same degree within radiotherapy. Therefore, I want to examine whether we can develop better diagnostics and, in the longer term, improve treatment," said Michael Lundemann. 

Today, most patients with aggressive brain cancer receive radiotherapy after their tumour has been surgically removed. The radiation is targeted on the basis of a traditional MRI scan, so the patient receives radiation in an area of two cm around the tumour. This is because the tumour will usually return in the same place. The target area for the radiation is limited to avoid unnecessary side effects in healthy parts of the brain. Overall, the treatment is effective, but only moderately. 

More clues to follow

Michael Lundemann has examined whether a combination of several advanced examinations can be used to target radiotherapy even more precisely. 

"We've tried to combine the anatomical MRI scan with a PET scan, which provides knowledge about the brain's function. We've examined patients with a tracer aimed at brain tumours and with a substance that is normally used for cancer. We've also examined how water molecules move in individual patients – they move differently in tumour tissue than in healthy tissue. This gives us a picture of where the cancer is going – you could say that we can predict the route of the cancer cells. We've also measured the blood vessels to see whether they are permeable. We know that tumours stimulate the brain to form rather strange new blood vessels," said Michael Lundemann. 

The different examinations were combined to predict precisely where the cancer would recur in the patient – and we then compared our results with the actual recurrence in patients. 

"We could see that some of the parameters on which we measured had great significance, and the combination gave us a surprisingly good prediction. When comparing our map of expected recurrences with the actual recurrences, we were spot on in 77% of cases," said Michael Lundemann. The results are based on nine patients and were published as a "Proof of concept" showing that the model is suitable for continuing work. 

Follow-up in progress

The researchers are currently working on being able to continue their efforts in clinical practice for some of the patients.
"We'll adjust to a slightly smaller set-up, without the PET scan and the cancer tracer, because this gave the least significant information. We'll continue the other examinations to test and adjust the model on even more patients. In the long term, we hope that our results can be used to give a higher dose of radiation in each patients' high-risk areas, and a lower dose in low-risk areas. But we need to be absolutely certain that the predictions are solid before we can go on to test any changes in the treatment," said Michael Lundemann. 

Around 300 Danes are diagnosed with aggressive brain cancer each year. About 50% are treated at Rigshospitalet, where about 100 patients receive radiotherapy as part of their treatment.

Responsible editor