Amide proton transfer-weighted magnetic resonance image-guided stereotactic biopsy in patients with newly diagnosed gliomas

Abstract

Purpose: Pathological assessment using World Health Organization (WHO) criteria is the gold standard for diagnosis of gliomas. However, the accuracy of diagnosis is limited by tissue sampling, particularly for infiltrating, heterogeneous tumours. We assessed the accuracy of amide proton transfer-weighted (APTw) magnetic resonance imaging (MRI)-guided tissue sampling to identify regions of high-grade glioma via radiographic-histopathologic correlation in patients with newly suspected glioma.

Patients and methods: Twenty-four patients with previously undiagnosed gliomas underwent a volumetric APTw MRI prior to their first neurosurgical procedure. A total of 70 specimens were collected via APTw image-directed stereotactic biopsy. Cellularity, necrosis, proliferation and glioma WHO grade were analysed for all specimens and correlated with corresponding APTw signal intensities.

Results: Thirty-three specimens displayed grade-II pathology, 14 grade-III, 15 grade-IV, and eight specimens revealed only peritumoural oedema. Multiple glioma grades were found within a single lesion in six patients. APTw signal intensities of the biopsied sites and the maximum APTw values across all biopsied sites in each patient were significantly higher for high-grade versus low-grade specimens. APTw signal intensities were significantly positively correlated with cellularity (R = 0.757) and proliferation (R = 0.538). Multiple linear regression analysis showed that tumour cellularity and proliferation index were the best predictors of APTw signal intensities.

Conclusion: APTw imaging identified tumour areas of higher cellularity and proliferation, allowing identification of high-grade regions within heterogeneous gliomas. APTw imaging can be readily translated for more widespread use and can assist diagnostic neurosurgical procedures by increasing the accuracy of tumour sampling in patients with infiltrating gliomas.

Keywords: Amide proton transfer-weighted imaging; Glioma; Histopathologic validation; Image-guided biopsy; Imaging biomarker.

Figure 1

Patient selection flowchart and reasons for exclusion from analysis.

Figure 2

Registration pipelines between the APTw (slice thickness, 4.4 mm) and conventional (slice thickness, 2.2 mm for T₂w and FLAIR or 1.1 mm for T₁w and Gd-T₁w) images used for APTw-directed stereotactic biopsies of gliomas on the BrainLab VectorVision neuro-navigation system. All MR images were re-sliced to a thickness of 4.4 mm. The acquired APTw image series (dashed line box) was corrected for possible motion artifacts caused during the scanning. For the APTw-directed tumor biopsy (blue arrows), the conventional MR images were registered to the APTw images. The ROIs were determined primarily on APTw images, and then transferred to the registered conventional MR images and uploaded into the BrainLab Neuro-navigation System. The exact biopsied sites were indicated by cursors placed by the surgeons with the intraoperative screenshots. For the APTw-histopathology correlation analysis (red arrows), the APTw images were registered to the conventional MR images. The biopsied sites on the screenshots were automatically linked to regions on the registered and resliced APTw images.

Figure 3

Conventional MR, APTw MR, and microscopic images for a patient with a histopathologically confirmed WHO grade-III anaplastic astrocytoma in the right frontal lobe (Patient 19). (A) MR images. The Gd-T₁w image revealed no post-contrast enhancement. APTw MRI showed hyperintensity in part of the lesion. (B) Stereotactic sites on the clinically obtained Gd-T₁w images. Tissues were obtained in two areas (1 and 2), showing distinct APTw enhancement. (C) Histology images. Tissue “1” with APTw hyperintensity (APTw = 3.52%) was diagnosed as grade-III. The tissue showed nuclear pleomorphism and elevated mitotic activity, demonstrating higher cellularity (1098/FOV) and significant proliferation (24.4%). Tissue “2” visually with APTw iso-intensity (APTw = 1.85%) was diagnosed as grade-II. The cellularity of the neoplastic astrocytes was moderately increased (819/FOV), mitosis was absent, and proliferative activity was comparatively low (1.5%). If the first area had been missed during biopsy, the clinical diagnosis would be inaccurate.

Figure 4

Conventional MR, APTw MR, and microscopic images for a patient with a histopathologically confirmed glioblastoma in the right parietal lobe (Patient 13). (A) Conventional MR images demonstrate a peripherally Gd-enhancing mass with central coagulative necrosis (showing T₂w hyperintensity and substantially suppressed water signals on FLAIR image), as well as mild adjacent vasogenic edema. The APTw image shows hyperintensity in the Gd-enhancing area, compared to CNAWM. (B) Stereotactic site on the clinically obtained Gd-T₁w images. One specimen was obtained from the rim of the tumor (with Gd enhancement and clear APTw hyperintensity of 4.28%). (C) Microscopic examination revealed extremely high cellularity (2468/FOV), prominent proliferation, and mitotic activity (Ki-67 of 22.3%), as well as the vascular proliferation typical of glioblastoma.

Figure 5

APTw imaging intensities of individual specimens and correlation analysis with histopathology indices. (A, B) Quantitative comparison of APTw imaging intensities of (A) all biopsied sites and (B) APTw^max values for each patient. (C, D) The correlation analysis results (C) between APTw intensities and Cell_count (linear regression: y = 1.23 + 1.11E-3*x), as well as (D) between APTw intensities and Ki-67 index (linear regression: y = 2.12 + 0.03*x).