TwitterFollow us on Twitter

FacebookFollow us on Facebook

Online Support

FacebookLGG Support Group

Join our mailing list


Longitudinal study of MRI and Spectroscopy techniques in the evaluation of the natural history of Low-Grade Gliomas

Principal Investigator: Dr. Jeremy Rees, Senior Lecturer in Medical Neuro-oncology, Institute of Neurology, National Hospital for Neurology & Neurosurgery, Queen Square, London


This project sets out to advance the management of untreated adult supratentorial low-grade gliomas (LGGs), regarded by many as the most controversial are in neuro-oncology. The main thrust of the proposed work is to study the natural history of these tumours non-invasively by employing the most advanced MR imaging techniques available in order to gain a better understanding of the factors contributing to tumour progression. This is an area where there has been a dearth of research, and to the best of my knowledge has not been systematically explored by any other group in the world. By the end of the study, we should have a clearer understanding of the role of different imaging techniques in predicting tumour behaviour, which in turn will allow a more individualized approach to patient management. In addition, refinement of these techniques will allow better monitoring of tumour responses to new treatments.

Extract from introduction, written in 2001

Conventional management of low-grade gliomas (LGGs) is conservative because there is no evidence that early treatment with either surgery or radiotherapy confers any survival advantage or prevents subsequent malignant transformation. Most people adopt a ‘wait and see’ policy and only intervene if the clinical or radiological features suggest progression. This may not be sensitive to the earliest changes of malignant transformation, the process by which these tumours progress from a low-grade into a high-grade lesion.

Low-grade gliomas of the cerebral hemispheres are common in both children and adults and comprise more than 50% of primary brain tumours. Patients may live for many years without symptoms apart from epilepsy. However, these tumours often infiltrate adjacent brain structures diffusely and have an inherent tendency to transform into malignant gliomas at an unpredictable time in their natural history. This process is associated with neurological symptoms including raised intracranial pressure, worsening epilepsy and progressive focal deficit.

By Dr.Jeremy Rees

As at August 2005, 72 patients have been recruited, 60 of whom were eligible for long term study. Over the last five years, 36 patients have transformed i.e. the tumour has become malignant and have required further surgery, radiotherapy or chemotherapy. Because of the active surveillance protocol, we are picking up transformation before the majority of patients develop neurological problems, which we feel is beneficial for all the patients in the study. The long-term benefit of such a proactive approach is unknown but 14 out of 18 transformers who had been studied for at least one year before transformation are still alive and well. Unfortunately, as expected with this type of tumour, 8 transformers have died, as has one non-transformer (who died suddenly of an unrelated problem).

The tumour volumes of 27 patients who have had at least three scans have been carefully and reproducibly measured using an in-house programme and a paper has now been resubmitted to a prestigious neurology journal reporting the results of this work. The findings surprised us – even so called ‘stable’ tumours were growing steadily at around 13%/year while tumours that eventually transformed grew even faster at 26%/year and then at 56%/year just before transformation occurred. This work has refined our understanding of the kinetics of low-grade gliomas and may be the first step towards tailoring treatment to individual patients whose tumours are growing above a critical rate. Because of the highly significant differences in growth rates between the non-transformers and the transformers, we have carried out complex statistical modeling to derive probabilities of a tumour transforming within the next 6, 12 or 18 months given a certain growth rate and volume.

Another paper has now been published by the American Journal of Neuroradiology showing how the various techniques that we are using can demonstrate differential sensitivity to chemotherapy within the same tumour depending on whether that part of the tumour was high grade or low grade. This work will be extended to develop imaging techniques that can predict early on whether or not a tumour is going to respond to chemotherapy.

We have developed the use of histogram analysis which allows different parameters from the whole tumour to be plotted in graphical format and which permits comparison between patients and between studies from the same patient across time. This powerful methodology has been applied to our diffusion-weighted imaging, perfusion imaging and permeability studies. Four papers, in various stages of preparation, will be produced and submitted by the end of the year. In summary, we have found the following:

  1. Perfusion imaging and analysis of relative Cerebral Blood Volume (rCBV) maps predicts progression by as much as twelve months and will probably turn out to be the most useful method of monitoring low-grade gliomas.
  2. Diffusion weighted imaging may be useful in distinguishing astrocytomas from oligodendrogliomas.
  3. Diffusion-weighted imaging reflects progression rather than predicts progression.
  4. Analysis of gadolinium enhancement across the whole tumour is sensitive to very small changes in leakage of the blood-brain-barrier and may provide another method of predicting early transformation in low-grade gliomas.
  5. Magnetic resonance spectroscopy cannot distinguish between transformers and non-transformers either at baseline or after prolonged follow-up. This sequence has therefore been dropped from the protocol.

By Dr.Jeremy Rees

We have identified three promising markers of early transformation using the different scanning techniques - volume change, gadolinium enhancement and perfusion.

Volume change
We have measured tumour volumes on 35 tumours, 19 transformers and 16 non-transformers. Growth rates in transformers were twice as fast as non-transformers even up to six months prior to transformation. More
excitingly, growth rate in the first six months of measurement (i.e the first two scans) was an independent predictor of subsequent transformation even after controlling for other known prognostic factors

Gadolinium enhancement
This is a measure of tumour blood vessel leakiness and is one of the earliest visible markers of malignant transformation. We have developed techniques to measure subtle enhancement (not visible to the naked eye) and have quantified the amount of overall enhancement within a whole tumour histogram, and shown clear differences between transformers and
non-transformers even at the very first scan.

This is a way of looking at blood flow within a tumour. As expected, transformers showed increasing perfusion using relative cerebral blood volume as a surrogate marker even twelve months prior to transformation, while non-transformers didn't change.

All these data have been analysed on relatively small numbers of patients so far and we are hoping to extend and confirm these preliminary but significant observations with more patient follow-up. Without the support of the SDBTT Astro Brain Tumour Fund, this project could never have been done.

The Future

We still have an enormous amount of data that requires analysis and new projects are being finalized. In particular we are carrying out a pilot study of a brand new technique, developed by the Physics Group in the NMR Unit, called quantitative Magnetisation Transfer. We are aiming to study 15 patients with three different subtypes of low-grade glioma (astrocytomas, oligodendroglioma and oligoastrocytoma) to see whether this can help distinguish between the three and thereby avoid the need for a biopsy.

We are also in discussion with a pharmaceutical company which is testing a new drug which inhibits angiogenesis (production of new blood vessels in the tumour) and hope to recruit patients whose scans are showing early signs of transformation, as evidenced either by acceleration of growth or increase in perfusion, for testing with this new drug.

By Dr.Jeremy Rees

To download a pdf of the presentation given by Dr.Rees at the Low-Grade Glioma Information Day held in London in November 2008, incorporating his final research findings and references to other current research related to low-grade gliomas, please click here

Dissemination of results at scientific meetings
British Neuro-oncology Society, Portsmouth, June 2007
International Society for Magnetic Resonance Imaging in Medicine, Berlin, May 2007.
International Symposium for Biomedical Imaging, 2007
British Society of Neuroradiology, Kingston-upon-Hull, Oct 2006.
European Society of Neuroradiology, Geneva, Sept 2006.
European Federation of Neurological Sciences, Glasgow, Sept 2006.
European Neurological Society, Lausanne, May 2006.
World Federation of Neuro-oncology, Edinburgh, May 2005.
International Society of Magnetic Resonance in Medicine, Miami, May 2005.
Association of British Neurologists, Belfast April, 2005.
British Neurosurgical Research Group, London, March, 2005.
European Federation of neurological Sciences 2004.
Association of British Neurologist Spring Meeting 2004.
British Neuro-oncology Group 2004

Published Papers
DJ Tozer, HR Jager, N danchaivijitr, CE Benton, PS Tofts, JH Rees, AD Waldman. Apparent Diffusion Coefficient Histograms may predict Low Grade Glioma Subtype.NMR in Biomedicine 2007; 20: 49-57

PS Tofts, CE Benton, R Weil, DJ Tozer, DR Altmann, HR Jager, AD Waldman and JH Rees. Quantitative analysis of whole tumour Gd enhancement histograms predicts malignant transformation in low-grade gliomas. J Magn Res Imaging 2007; 25: 208-214

N Danchaivijitr, AD Waldman, DJ Tozer, CE Benton, G Brasil Caseiras, PS Tofts, JH Rees and HR Jager. Longitudinal Perfusion-Weighted MR Imaging in Patients with Low Grade Gliomas: Do Changes in rCBV Measurements Predict Malignant Transformation? Radiology (awaiting acceptance)