PET-Guided Resection in High-Grade Gliomas

Background and Scientific Rationale: Glioblastoma (GBM), classified as a CNS WHO grade 4 tumor under the 2021 WHO Classification of Central Nervous System Tumors, is the most common and aggressive primary brain malignancy. Despite the current standard of care-maximal surgical resection followed by radiotherapy and chemotherapy-its prognosis remains poor, with a mean overall survival (OS) of approximately 15 months. [11]Surgery plays a critical role in disease control, as the extent of resection (EOR) is one of the few modifiable prognostic factors. Traditionally, surgical resection is guided by magnetic resonance imaging (MRI), primarily through gadolinium contrast-enhanced T1-weighted (CE-T1) and Fluid-attenuated inversion recovery (FLAIR) T2-weighted sequences. Although there is no universally accepted definition of gross total resection (GTR), a near-complete removal (90-100%) of the contrast-enhancing tumor volume is associated with better outcomes. [5, 6]However, glioma cells frequently infiltrate areas beyond the contrast-enhanced margins, often extending into the FLAIR hyperintense region. Resection of these regions has been shown to confer a survival benefit, albeit at the potential cost of functional deficits, as FLAIR hyperintensity is non-specific and may also represent peritumoral edema. [1]This challenge has led to the exploration of additional imaging modalities, such as positron emission tomography (PET), magnetic resonance spectroscopy, and hybrid imaging techniques, to improve glioma delineation.[4, 8] PET imaging, particularly using amino acid tracers such as 18F-DOPA, has emerged as a valuable tool for identifying the true biological extent of gliomas. Unlike 18F-fluorodeoxyglucose (18F-FDG), which lacks specificity in brain imaging, radiolabeled amino acids offer superior tumor-to-background contrast due to their selective uptake via the L-type amino acid transporters (LAT). [7]Molecular imaging with 18F-DOPA PET allows for the delineation of the so-called biological tumor volume (BTV), highlighting metabolically active and potentially more aggressive tumor regions. Importantly, the BTV frequently extends beyond the contrast-enhancing margins and, in some cases, may identify tumor areas not detected by FLAIR.[3, 10] Evidence supporting the oncological relevance of PET-guided resection is growing. Several studies have demonstrated that PET-defined tumor volumes correspond to regions of viable neoplastic tissue, and surgical resection of these hypercaptant areas has been associated with improved OS. [2]Among the various amino acid radiotracers available, 18F-DOPA has shown particular promise not only in assessing tumor grade and aggressiveness but also in monitoring treatment response and distinguishing true progression from treatment-related changes. [9] Preliminary data from the study initial cohort of 24 patients further support this approach. In a subset of 5 patients, histopathological analyses of different tissue samples confirmed that PET hypercaptant areas outside the contrast-enhancing tumor region harbored a high mitotic index and a significant neoplastic component. Notably, these areas exhibited a higher proliferation rate compared to regions identified solely through FLAIR imaging. These findings, presented at national and international neurosurgical conferences, reinforce the hypothesis that PET-guided resection may enhance oncological outcomes and warrant further investigation through a dedicated multicenter study.

This study seeks to validate the clinical utility of PET-guided glioma resection by comparing outcomes in patients undergoing PET-RM integrated surgery versus standard MRI-guided resection. The authors hypothesize that PET-guided resection will yield superior survival benefits, as it enables more selective yet extensive tumor removal, providing a more precise target for supramaximal resection refining the balance between oncological efficacy and functional preservation.

Objectives: The authors hypothesize that 18F-DOPA PET can provide a more reliable and achievable surgical target than FLAIR hyperintensity in supramaximal resection, as it may differentiate infiltrative tumor tissue from peritumoral edema. Since PET-defined tumor volumes are larger than contrast-enhancing regions but smaller than FLAIR hyperintensity, PET-guided resection could optimize the balance between maximal tumor removal and functional preservation, enhancing the maximal safe resection approach.

Furthermore, the authors hypothesize-supported by the study preliminary data-that PET hypercaptant areas beyond the contrast-enhancing tumor harbor viable tumor cells and should be considered a biologically relevant resection target.

The main objectives of the present study are:

1. To evaluate the impact of 18F-DOPA PET-guided resection on overall survival (OS) and progression-free survival (PFS) compared to conventional MRI-guided resection.
2. To validate the biological significance of PET hypercaptant areas through histopathological correlation, confirming the presence of neoplastic cells in PET-positive regions beyond contrast enhancement.
3. To analyze the correlation between PET parameters and molecular markers to improve glioma characterization and surgical planning

This study is a multicenter, prospective, non-randomized trial conducted at IRCCS Ospedale Policlinico San Martino (Genoa) and AOU Città della Salute e della Scienza (Turin). Patients with newly diagnosed, supratentorial high-grade gliomas will be enrolled and undergo preoperative 18F-DOPA PET and MRI. The aim is to compare outcomes between patients undergoing PET-guided resection and those receiving conventional MRI-guided resection.

Methods and study design The study population includes adult patients diagnosed with high-grade glioma, located supratentorially and eligible for surgical resection. Inclusion criteria consist of age ≥18 years, have a high-grade glioma (WHO grade III/IV) diagnosed on MRI as assessed by the neurosurgeon, and provide written informed consent. Exclusion criteria include tumors located in the cerebellum, brainstem, or midline. Patients with medical conditions precluding MRI, such as those with a pacemaker, are not eligible. Other exclusion criteria include the inability to provide written informed consent, secondary high-grade gliomas resulting from malignant transformation of a low-grade glioma, and a second primary malignancy within the past five years. Eligible patients are scheduled for a preoperative 18F-DOPA PET scan within one week prior to surgery. PET and MRI images are co-registered using neuronavigation systems to optimize surgical planning. Neurosurgical resections are performed with a maximal safe resection approach, prioritizing function preservation. In the neurosurgical procedure, tumor samples are collected from five distinct areas: contrast-enhancing, PET-positive, FLAIR-positive, PET-only, and MRI-negative regions. These tissue samples undergo histopathological and molecular characterization to validate PET-defined tumor infiltration. Five different areas are considered: a) CE area (inside both PET and FLAIR volumes); b) n-CE areas (PET+ and FLAIR+); c) only-PET+; d) only-FLAIR+; e) not evident MRI disease. An experienced pathologist evaluates these samples blindly, analyzing: 1) neoplastic component (%), 2) necrosis (%), 3) mitosis (10/high power field), 4) vascular proliferation (absent, focal, or diffuse), 5) infiltrative component (%), and 6) normal/reactive component (%). After surgery, patients will be categorized into two groups based on the extent of PET-guided resection: those in whom the entire PET-positive tumor volume has been removed and those with residual PET-positive tissue. Postoperative assessment includes MRI at 48 hours to evaluate the extent of resection (EOR) according to RANO criteria and 18F-DOPA PET at 30 ±7 days to determine residual metabolic tumor activity.Longitudinal follow-up is conducted, tracking progression-free survival (PFS) and overall survival (OS) through neuro-oncology teams at both centers.

Imaging Protocol MRI data will be acquired using a 3 Tesla Siemens Prisma scanner with an advanced imaging protocol, including T1, T2, FLAIR, DTI, DSC, and DCE sequences. For each patient, relative cerebral blood volume (CBV) and diffusion-weighted imaging (DWI)-derived metrics will be evaluated across clinical outcome groups. The dataset will include standard deviations and medians of these parameters, categorized by tumor diagnosis and molecular subtypes. Log-rank tests will evaluate the association between rCBV, ADC metrics, and PFS. The extent of resection (EOR) will be quantified using volumetric analysis according to RANO criteria. PET imaging will be conducted on a Siemens Biograph 16 PET-CT scanner, with 18F-DOPA injected 15 minutes before a static 90-minute scan. The collected data will be analyzed using semi-quantitative and dynamic approaches. PET images will be co-registered with MRI using intraoperative neuronavigation systems (Brainlab AG or Medtronic S8 StealthStation). A low-dose CT scan will be used for attenuation correction.

Muldisciplinary collaboration The study relies on close collaboration between specialists. Nuclear medicine experts will oversee PET acquisition and interpretation, ensuring accurate metabolic tumor delineation. Neuroradiologists will analyze MRI data, quantify tumor volumes, and perform PET/MRI co-registration. Neurosurgeons will integrate PET findings into their surgical strategy, aiming to maximize resection while preserving neurological function. Neuropathologists will examine tumor samples to confirm the correlation between PET findings and histological tumor infiltration. Neuro-oncologists will manage patient follow-up, recording survival outcomes and treatment responses, contributing to the overall evaluation of the impact of PET-guided surgery on glioma management.

Data Analysis Statistical analysis involves Kaplan-Meier survival curves to estimate progression-free survival (PFS) and overall survival (OS). Additionally, multivariate regression models will identify independent prognostic factors, considering variables such as age, tumor location, molecular subtypes, and EOR. Sample size calculation is based on the prevalence of high-grade gliomas in the participating centers, with a target enrollment of at least 100 patients, evenly distributed between the two groups. Histopathological analysis will include assessments of neoplastic density, necrosis, mitotic count, and vascular proliferation. Pearson's chi-square test will be used for categorical variables, while a Wilcoxon-Mann-Whitney test or Fisher's Exact test will be applied for numerical variables. A paired sample Student's t-test will compare numerical variables between samples.

Work Plan and Timeline The study spans four years and is divided into sequential phases. Phase 1 (Months 0-6) focused on study preparation, including ethical approvals, imaging and surgical protocol development, staff training, and PET/MRI integration into neuronavigation workflows.

Phase 2 (Months 7-30) is ongoing, with active patient recruitment at both centers. Eligible patients undergo preoperative PET and MRI, followed by surgery based on their assigned strategy. Postoperative imaging and clinical follow-up monitor disease progression and survival outcomes.

Phase 3 (Months 31-42) will involve data analysis and evaluation. Survival outcomes, PET parameters, molecular markers, and histopathological findings will be analyzed, with subgroup assessments to refine prognostic factors.

Phase 4 (Months 43-48) will focus on disseminating results through presentations at major conferences and publications in peer-reviewed journals.