Study participants and inclusion and exclusion criteria
Twenty-three patients attending the Department of Neurosurgery, 10th Military Research Hospital, Bydgoszcz, Poland for stereotactic biopsy due to suspected primary brain tumors of glial origin having undergone contrast-enhanced MRI were enrolled in this prospective study.
The inclusion criteria were: age > 18 years; suspected CNS glioma based on MRI; and in general good health permissive of surgery. The exclusion criteria were: age < 18 years; KPS < 60 points; pregnancy; and disqualification from surgery for medical reasons.
The reasons for performing CNS diagnostics in these patients included: headaches, epileptic seizures, cerebellar syndrome, speech disorders, and paresis. In a few patients, the tumor was detected accidentally due to diagnostics performed for a head injury. The basic characteristics of the participants are shown in Table 1.
Hybrid PET-MRI and pre-biopsy examinations
Patients meeting the inclusion criteria were referred for hybrid 18F-FET-PET-MRI examination in one procedure with simultaneous data collection at the Department of Nuclear Medicine, Franciszek Lukaszczyk Oncology Center, Bydgoszcz, Poland. PET was performed as described previously 25. PET-MRI image analysis was performed by a nuclear medicine specialist and neurosurgeon to confirm increased FET uptake during early acquisition 5-15 minutes post injection. The final biopsy site was selected by two experienced neurosurgeons.
The median time to biopsy was seven days (IQR 3-13 days). Before biopsy, each patient had: 18F-FET PET-MRI; routine MRI of the head for neuronavigation (T1 and T1 sequences with gadolinium contrast, T2 and T2 FLAIR sequences); routine chest X-ray for surgery; electrocardiography; routine pre-operative blood tests; and head CT (stereo thin layers) with the stereotactic frame attached. The stay at the Department of Neurosurgery lasted four days.
PET-guided stereotactic biopsy
A routine biopsy procedure was used except for selecting targets based on fused PET-MRI images rather than MRI alone, with more samples taken than usual for histopathological examination. In some cases, it was necessary to plan several trajectories with an extra trepanation hole. The procedure was performed under local anesthesia in analgosedation.
The frame was first fixed to the patient's head prior to stereo CT examination and patient transfer to the operating theater. Using BrainLab software (Munich, Germany) in the operating theater, CT, MRI, and PET images were fused and the biopsy target parameters determined. The following biopsy sites were determined in each case: (i) site of contrast enhancement in T1 MRI sequences after contrast administration and simultaneous increased FET uptake in PET images 5-15 min post injection; (ii) site of contrast enhancement in MRI but no increase in FET uptake; (iii) site of increased FET uptake but without contrast enhancement on MRI; and (iv) peripheral areas hyperintense in T2 FLAIR without increased FET uptake and contrast enhancement.
It was not always necessary to plan four trajectories to collect the material, since it was sometimes possible to collect different samples on one trajectory. For example, if on a given trajectory the area of isolated FET uptake was more superficial and the MRI contrast enhancement was deeper, material corresponding to two different targets could be taken from one trajectory but at different depths.
Each target was defined as a 0 point measuring a few millimeters in area, i.e., the target ranged, for example, from -6 to 0, where material was taken every 1 mm from -6 depth to 0 along a planned trajectory. Every first sample was tested intraoperatively, and every second sample was taken for histopathological and molecular examination. Approximately six serial biopsies were taken in areas of contrast enhancement, increased FET uptake, and peripheral areas. Therefore, in total, about 18 samples were obtained from each patient, i.e., ~400 samples in total for the entire cohort. The highest-grade malignancy detected at each serial biopsy site is presented in the results.
Hyperintensity was seen as white compared with areas without enhancement in MRI T1 images with contrast. FET uptake was visualized at the time of biopsy target determination as blue, yellow, red, and white areas, with white representing the highest tracer uptake (Figure 1).
PET examination was performed 5-15 minutes after administration of 18F-FET and again after about 60 minutes (dual timepoint examination). Biopsies were planned based on the PET images taken 5-15 minutes after administration of the radiopharmaceutical.
After completion of planning, the patient's head was attached to the operating table through the previously installed frame. The x, y, and z coordinates of the entry point were calculated by the stereotaxis program, thus determining where the head should be shaved, the skin incision, and the site for the trepanation hole. After washing and preparation, the site was anesthetized with 1% lignocaine solution and a small skin incision was then made to expose the skull bones. With a manual trepan, and for tumors in the posterior cranial fossa, a small trepanation was made with a high-speed drill. Then, the visualized dura mater was coagulated, incised, and, in accordance with the calculated parameters, a cannulated biopsy needle was placed in the brain. Forceps were inserted through the cannula and several small samples taken for histopathological examination at 1 mm intervals. A neuropathologist was present in the operating room throughout the procedure, who initially assessed the material. The remaining tumor fragments were placed into sample tubes with formalin and then processed and assessed by a pathologist.
Histopathological and molecular evaluation
A histopathological diagnosis was made intraoperatively with complete histopathological assessment performed postoperatively. Gliomas were graded according to the 2016 WHO classification.
For intraoperative evaluation, tissues were smeared and crushed onto glass slides before staining with methylene blue and microscopic examination. Postoperatively, formalin-fixed, paraffin-embedded tissue samples were stained with hematoxylin and eosin for microscopic examination, with immunocytochemical and immunohistochemical methods used for diagnosis where applicable.
For molecular evaluation, biopsy samples were dewaxed. DNA was extracted using the magnetic method (Maxewell®️16 FFPE Tissue LEV DNA Purification Kit, Promega, Madison, WI). The Plexor®️HY System (Promega) was used to determine the DNA concentration. Molecular testing became standard practice during the course of this study, so only some patients had genetic test results. The molecular methods used in our institution are described in 31.
Statistical analysis
All analyses were conducted in PQStat (PQStat Software, Poznan, Poland). A p-value of <0.05 was considered statistically significant. Fisher’s exact test or the Fisher-Freeman-Halton test were calculated to investigate relationships between categorical variables. Cohen's kappa concordance coefficient (with linear weights) was calculated to determine the agreement of both assessments.
Study approval
The institution’s bioethics committee approved the study (KB 647/2015), and informed consent was obtained from all participants.