[18F] FLT PET predicts cellular proliferation in pediatric brain tumors

Objectives: Brain tumors are the most common solid tumor in children and are responsible for substantial morbidity and mortality in the pediatric population. [18F] FDG PET/CT has been of limited utility in the evaluation of brain tumors, due to the high level of FDG uptake in the normal brain. Other PET radiopharmaceuticals, such as [18F] FLT, a putative measure of cellular proliferation, have demonstrated promise for the evaluation of brain tumors in adults, but have not been studied systematically in children. The aim of this study is to evaluate the use of [18F] FLT to assess cellular proliferation in pediatric brain tumors.

Methods: [18F] FLT PET was performed in 22 pediatric patients (age 6-21 y) in which MRI suggested a newly diagnosed or recurrent brain tumor. Participants were enrolled at Boston Children’s Hospital and five institutions of the Pediatric Brain Tumor Consortium. Study protocols were approved by an IRB at each institution. Informed written consent was obtained. At each institution, [18F] FLT PET or PET/CT was performed on a camera included in a multi-institution validation program. No pre-study preparation was required of participants, and brain PET/CT was performed 30-45 minutes after intravenous administration of 5.2 MBq/kg [F18] FLT. No or minimal uptake was observed in normal brain tissue. PET images were analyzed on a stand-alone workstation (Hermes Medical Systems, Montreal) in a single image analysis core. Image analysis used ROI and VOI analysis, with boundaries determined on the PET scan or on co-registered FLAIR and post-constrast T1 MR images. Ki-67 labeling index as a measure of cellular proliferation was determined with MIB-1 immunostaining of paraffin sections of surgical specimens, which was performed or confirmed in a single central review laboratory. Statistical analysis was performed using Pearson’s correlation.

Results: Eight participants had a new diagnosis of brain tumor and 9 participants had recurrent tumor. In 5 participants, subsequent surgical histology did not confirm a malignant primary brain tumor (instead demonstrating lymphoma, vasculitis, or inflammation), and these participants were not included in further analysis. Tumor uptake of FLT, assessed by SUVmax ranged from 0.11 (oligoastrocytoma, with uptake similar to background brain) to 4.00 (pilocytic astrocytoma). The Ki-67 labeling index (fraction of cells staining with MIB antibody) ranged between 3% and 66%, in keeping with a wide range of tumor types and grades. The most common tumor type (N=4) was pilocytic astrocytoma. In the 17 participants with new or recurrent brain tumor, tumor uptake of FLT (SUVmax) correlated significantly with Ki-67 labeling index (Pearson r = 0.74, p < 0.001). In the 9 participants with possible tumor recurrence, SUVmax ranged from 1.54 to 4.00, and all were demonstrated to have tumor recurrence.

Conclusions: In pediatric brain tumors, uptake of [18F] FLT on PET does predict the degree of cellular proliferation. This information may be of prognostic value and may guide the initial surgical approach, particularly, when only partial resection or biopsy is planned or possible. These findings also suggest that [18F] FLT PET/CT may have utility in evaluating patients in whom anatomic imaging has suggested possible recurrence of a previously treated brain tumor.