The ideal method for non-invasive monitoring of biological or metabolic activity in the brainstem has not yet been established. Existing methods for evaluating cerebral tissue (e.g. PET) have limitations when applied to the brainstem. New noninvasive imaging methods are currently being studied. We are using dynamic susceptibility contrast MR perfusion, dynamic enhanced MRI, proton nuclear magnetic spectroscopic imaging, and FDG PET to further investigate the imaging characteristics of pontine gliomas. There are four clinical trials which currently fall under this project. Nearly all pediatric patients with CNS tumors undergo extensive imaging on at least one of these studies. Over the past several years, we have demonstrated the difficulty in using standard MRI for definition of response in clinical trials and have proposed a new definition of clinical benefit for these patients based on these results. Response to treatment and therefore efficacy of new agents is typically measured using two-dimensional tumor measurements following WHO criteria, or more recently, using RECIST criteria. RECIST criteria have not been validated for the measurement of pediatric CNS tumors and two-dimensional measurements continue to be the standard in consortium trials. We compared the 1D, 2D and 3D measurements of diffuse intrinsic pontine gliomas by four separate blinded readers. The results show that there is significant variability (0-197%) in all measurements and change in tumor size is therefore a poor indicator of response for clinical trials. We have also electronically surveyed pediatric oncologists and neuro-oncologists across the country to determine the standard practice of tumor measurement. Our results show that there is no standard measurement method being used in the measurement of these lesions. We have published a report recommending the use of FLAIR images to measure DIPG and comparing these with the best response scans on study. Images should be compared side by side. Central review was also recommended for clinical trials involving DIPG. We are continuing to evaluate the utility of additional MR sequences, including perfusion and spectroscopic imaging in an attempt to define the role of specific imaging techniques (e.g. spectroscopy, PET) in these patients. Recently, we have demonstrated that changes in maximum CHO:NAA ratios over time using proton spectroscopic imaging were prognostic of outcome. In addition, we demonstrated that increased perfusion at any time point was a poor prognostic factor. These results have now been published..
