NIMG-80. IMPACT OF KETOGENIC DIET ON MOLECULAR METABOLIC IMAGING WITH [18F]FDG AND [18F]DASA-23 POSITRON EMISSION TOMOGRAPHY IN A PATIENT WITH GLIOBLASTOMA

Abstract The standard-of-care for neuroimaging in patients with glioblastoma is contrast-enhanced MRI, which does not provide tumor metabolism information. Although [18F]FDG PET is used to image tumor metabolism in many solid organ cancers, it has limited benefit in atients with brain tumors due to the high level of glucose uptake in normal brain cells. [18F]DASA-23 is a novel PET radiotracer that assesses the levels of pyruvate kinase M2 (PKM2), a protein highly expressed in cancer cells, which undergo aberrant glycolysis. Despite [18F]DASA-23 having a higher tumor-to-background ratio (TBR) of standardized uptake values (SUVs) in the brain compared to [18F]FDG, both may be susceptible to altered carbohydrate metabolism in the setting of a ketogenic diet (KD). We report the case of a patient with glioblastoma on a KD, whose [18F]FDG and [18F]DASA-23 PET brain scans were limited in identifying disease progression (PD). Background was defined as the white matter contralateral to the tumor, at the level of the centrum semiovale, for TBR of SUVmax. A 39-year-old man was diagnosed with right temporal glioblastoma after previous diagnoses of grade 2 gliofibrillary oligodendroglioma (age 20) and grade 3 anaplastic oligodendroglioma (age 22). He began a KD after the glioblastoma diagnosis. 21 months later, TBR on [18F]DASA-23 PET was 1.38. One month later, surveillance brain MRI showed PD. Ten days later, TBR on [18F]FDG PET was 1.03. The patient died 15 months later. Despite MRI-based PD, the TBR of each radiotracer was not markedly elevated, although the [18F]DASA-23 TBR 1-month pre-PD was 34% greater than the [18F]FDG PET TBR 10-days post-PD. This may suggest a limitation of metabolic imaging by PET in the setting of a KD, and provide an opportunity to develop novel PET radiotracers that are not susceptible to the patient’s diet.