Journal of Neuro-Oncology 41: 291–298, 1999. © 1999 Kluwer Academic Publishers. Printed in the Netherlands. Clinical Study Variation of post-treatment H-MRSI choline signal intensity in pediatric gliomas Jorge A. Lazareff 1 , Rakesh K. Gupta 2 and Jeffry Alger 2 1 Department of Surgery, Division of Neurosurgery, 2 Department of Radiological Sciences, University of California Los Angeles School of Medicine, Los Angeles, CA, USA Key words: low grade astrocytoma, proton spectroscopy, pediatric, choline Summary Pediatric brain gliomas are not always amenable for complete surgical excision, therefore adjuvant treatment for a large tumor mass is often required. As tumor volume shrinkage may not be a reliable method for assessing response to treatment, information about the tumor growth potential is desirable for an adequate follow-up of the patients. Choline (Cho) signal intensity, determined by proton magnetic resonance spectroscopy imaging (H-MRSI), has proved to be a reliable indicator of the metabolic activity and of tumor progression in various intracranial tumors. In this study we have sought to determine if H-MRSI can be of use in monitoring the response of pediatric gliomas to different forms of therapy. We performed pretreatment and post-treatment H-MRSI in 10 children with biopsed or partially excised brain gliomas. The follow-up period ranged between 6 and 40 months. A total of 38 H-MRSI were performed. All the patients had chemotherapy or radiotherapy. As an indicator of tumor activity we utilized the ratio between tumor/brain Cho signal intensity. Treatment response was evaluated as a function of tumor volume and clinical outcome. In 6 patients whose tumor volume decreased or remained stable we observed that the Cho ratio decreased (p< 0.01) after treatment and remained low during longitudinal follow-up. In the 4 patients whose tumors progressed the Cho ratio increased after treatment. These observations suggest that serial H-MRSI can provide valuable information regarding the response to therapy in pediatric gliomas and therefore be of use in the follow-up of these neoplasms of childhood. Introduction The management of pediatric brain gliomas represents a particular challenge for the neuro oncology team. While it is true that astrocytomas of the pediatric variety are less aggressive compared to their adult counterparts of the same histopathological grade [1], incompletely excised pediatric gliomas may exhibit a highly variable evolution subsequent to partial resection. Some tumors spontaneously regress while others rapidly progress and thereby compromise neighboring vital structures [2]. Further, the available treatments (radiation and/or chemotherapy) are not without short and long term complications [3] explaining the reluctance to indi- cate additional cytoreductive regimes that may have permanent implications on the quality of survival of these young patients. The efficacy of brain tumor treatments is tradition- ally assessed in terms of variations in tumor mass by serial MRI. In pediatric gliomas, however, significant tumor shrinkage may not always correlate with modifi- cations of tumor behavior [4]. Factors other than tumor volume require consideration in the evaluation of ther- apeutic response. Ex vivo tests, such as quantification of BdUr labeling, provide valuable information about the cellular proliferation of the tumor tissue at the time of the surgical biopsy [5], but have limited value for longitudinal analysis of response to treatment. This problem can be circumvented by in vivo methods such as proton magnetic resonance spectroscopic imaging