Three-dimensional amide proton transfer (APT) imaging appliable to navigation surgery can present comparable metabolic activity of glioblastoma to 11C-Methionine PET

Background

Amide proton transfer (APT) imaging has been proposed as a technique to assess tumor metabolic activity. We have previously ¹¹C-methionine positron emission tomography (¹¹C-Met-PET) can evaluate the metabolic activity of peritumoral area including infiltrating tumor cells in glioblastoma (GBM). To resolve disadvantages of ¹¹C-Met-PET, in the present study, we aimed to evaluate whether three-dimensional fast spin echo-based APT (3D FSE-APT) imaging is usable for not only presenting the metabolic activity of brain tumors, but also detecting areas where infiltrating tumor cells including glioma stem cells (GSCs) could exist, by applying an image-guided navigation system incorporating 3D FSE-APT to glioblastoma surgery.

Methods

Twenty-six consecutive patients with GBMs were enrolled in this study. Among these 26 cases, 10 patients underwent ¹¹C-Met-PET examination. All 26 patients underwent two-dimensional single shot fast spine echo-based APT acquisition with a chemical exchange saturation transfer sequence (2D SSFSE-APT). The most recent 14 cases underwent 3D FSE-APT to examine whether 3D APT imaging was applicable to the navigation system. We investigated the clinical applicability of 3D FSE-APT by comparison with 2D SSFSE-APT and evaluated the utility of 3D FSE-APT as a metabolic imaging guide in the intraoperative navigation system. We also analyzed whether 3D FSE-APT can depict the extent of infiltrating tumor cells including GSCs in the peritumoral area in GBM.

Results

The most recent 14 cases underwent 3D FSE-APT. The 3D FSE-APT was visually almost equivalent to 2D SSFSE-APT and mean APT intensity (APT_mean) in GBM obtained by 3D FSE-APT was almost equal to that from 2D SSFSE-APT. Mean APT_mean on 2D SSFSE-APT at the site showing a tumor-to-contralateral normal brain tissue ratio (TNR) of 1.4 on ¹¹C-Met-PET was 1.52 ± 0.16%. In contrast, mean APT_mean on 3D FSE-APT at the same site was 1.30 ± 0.06%. The optimal cut-off value for APT_mean on 3D FSE-APT was evaluated as 1.28%, offering 100% sensitivity and 100% specificity. Incorporating 3D FSE-APT into the navigation system allowed tumor resection including infiltrating tumor cells under image-guided navigation. Mean Ki-67 staining index in the area with a mean APT_mean of 1.28% was 11.8% (range, 5.0–20.0%).

Conclusions

The area of tumor invasion could be evaluated by 3D FSE-APT in a similar way to ¹¹C-Met-PET, and the cut-off value for deciding the borderline between the area including infiltrating tumor cells and that with almost no tumor cells was 12.8%. In addition, 3D FSE-APT could be applied to navigation systems and may have great potential as an imaging modality replacing ¹¹C-Met-PET in GBM surgery.