[18F]DASA-23 PET/MRI evaluation in newly-diagnosed and recurrent high-grade glioma

High-grade gliomas pose significant diagnostic imaging challenges in their initial diagnosis, evaluation of treatment-related change versus recurrent/progressive disease, and prognostication. 1-((2-fluoro-6-[18F]fluorophenyl)sulfonyl)-4-((4-methoxyphenyl)sulfonyl)piperazine ([18F]DASA-23) which binds pyruvate kinase M2 (PKM2), an important enzyme in glycolysis, was developed to probe the metabolic activity of high-grade gliomas. This pilot study sought to evaluate [18F]DASA-23 PET’s diagnostic performance in high-grade gliomas, how its uptake over time is associated with survival, and its relationship with conventional MR findings. This prospective pilot study enrolled healthy controls and patients with high-grade gliomas who received [18F]DASA-23 PET/MRI’s. In the latter group, scans were performed at the time of initial diagnosis or recurrence, with a sub-group also receiving follow-up scans during their therapy course. Three nuclear medicine physicians independently adjudicated the presence or absence of focal intracranial [18F]DASA-23 uptake, from which the sensitivity, specificity, accuracy, positive predictive value, and negative predictive value were determined. For patients with tumors, tumor-to-background SUVmax was calculated and correlated with progression-free survival (PFS). Correlation analysis between MRI perfusion parameters and [18F]DASA-23 TBRmax was also performed. As an exploratory analysis, the amount of [18F]DASA-23 uptake overlap with sites of both MRI T1 post-contrast enhancement and non-enhancing T2-FLAIR hyperintense signal about the resection cavity pre-therapy was determined. Four healthy controls and nine patients with high-grade gliomas were included in this study. The baseline [18F]DASA-23 PET demonstrated a sensitivity of 0.78, specificity of 1.00, and accuracy of 0.85 in the diagnosis of high-grade glioma. The follow-up [18F]DASA-23 PET exams in 5 patients at a mean of 46 days after initiation of therapy showed tumor to background ratio SUVmax (TBRmax) was significantly inversely correlated with PFS (r = -0.91, p = 0.03). MR dynamic contrast susceptibility (DSC) derived normalized cerebral blood volume (nCBV) and arterial spin labeling (ASL) derived normalized cerebral blood flow max (nCBFmax) were not significantly associated with [18F]DASA-23 uptake. Within the [18F]DASA-23 uptake regions of interest, 20.2% of the volume was occupied by T1 post-contrast enhancement, and 28.3% of the volume was non-enhancing T2 FLAIR hyperintensity, with no significant difference in the volume of enhancement and non-enhancing T2 FLAIR hyperintensity. This pilot study demonstrated favorable [18F]DASA-23 PET diagnostic performance in the diagnosis of high-grade gliomas, and its uptake after initiation of therapy may with further research be useful as a marker to evaluate treatment efficacy and patient prognosis. Additionally, the absence of a significant relationship between [18F]DASA-23 uptake and MR perfusion markers, and [18F]DASA-23 PET’s limited overlap with enhancement and non-enhancing T2 FLAIR hyperintensity raises the possibility that future [18F]DASA-23 PET studies with larger patient samples could uncover clinically relevant information not fully captured with MR alone.