https://orcid.org/0000-0002-4528-8971
Abstract
MRI alone has limited accuracy for delineating tumor margins and poorly predicts the aggressiveness of gliomas, especially when tumors do not enhance. This study evaluated simultaneous 3,4-dihydroxy-6-[18F]fluoro-L-phenylalanine (FDOPA)-PET/MRI to define tumor volumes compared to MRI alone more accurately, assessed its role in patient management, and correlated PET findings with histopathology.
Ten patients with known or suspected gliomas underwent standard of care surgical resection and/or stereotactic biopsy. FDOPA-PET/MRI was performed prior to surgery, allowing for precise co-registration of PET, MR, and biopsies. The biopsy sites were modeled as 5-mm spheres, and the local FDOPA uptake at each site was determined. Correlations were performed between measures of tumor histopathology, and static and dynamic PET values: standardized uptake values (SUVs), tumor to brain ratios, metabolic tumor volumes, and tracer kinetics at volumes of interest (VOIs) and biopsy sites.
Tumor FDOPA-PET uptake was visualized in 8 patients. In 2 patients, tracer uptake was similar to normal brain reference with no histological findings of malignancy. Eight biopsy sites confirmed for glioma had FDOPA uptake without T1 contrast enhancement. The PET parameters were highly correlated only with the cell proliferation marker, Ki-67 (SUVmax: r = 0.985, P = .002). In this study, no statistically significant difference between high-grade and low-grade tumors was demonstrated. The dynamic PET analysis of VOIs and biopsy sites showed decreasing time-activity curves patterns. FDOPA-PET imaging directly influenced patient management.
Simultaneous FDOPA-PET/MRI allowed for more accurate visualization and delineation of gliomas, enabling more appropriate patient management and simplified validation of PET findings with histopathology.
In the United States, for 2019, there were an estimated 86 970 new cases of brain and other CNS tumors diagnosed.1 Gliomas represent approximately 30% of all brain and CNS tumors and 80% of all malignant brain tumors and are associated with a high mortality rate.2 These tumors represent a significant health problem and constitute a major imaging challenge for volume determination, delineation of margins, and therapeutic evaluation.
Neuroimaging plays a critical role in the initial clinical diagnosis, surgical treatment planning, and posttreatment follow-up, with MRI currently the modality of choice given its superior resolution and high soft-tissue contrast. A key feature of many high-grade gliomas (HGGs) is the compromise of the blood-brain barrier (BBB), which can be detected on contrast-enhancement MRI, although many low-grade gliomas (LGGs) and some HGGs, mainly grade III, show no or little contrast enhancement in regions.3,4 Lack of contrast enhancement on MRI makes delineation of the entire tumor volume and surgical margins much more challenging in gliomas. The ability of MRI to differentiate between treatment-induced changes and residual or recurrent tumor is also limited because the imaging features have substantial overlap.5–7
PET is a noninvasive imaging technique used to visualize and quantify various biochemical and physiological processes in living beings. The PET radiotracer most widely used for oncologic imaging, 18F-FDG (2-deoxy-2-[fluorine-18]fluoro-D-glucose), has poor tumor-to-background contrast in gliomas, limiting its utility for brain tumor imaging. Furthermore, FDG-PET provides limited additional value to MRI for differentiation between malignant glioma recurrence and radionecrosis.8,9
Several amino acid PET tracers have established utility for imaging gliomas: 11C–methyl-L-methionine (MET), O-(2-[18F]fluoroethyl)-L-tyrosine (FET), and 3,4-dihydroxy-6-[18F]fluoro-L-phenylalanine (FDOPA).10–13 These radiolabeled amino acids target the system L substrates, which are upregulated in gliomas and do not depend on the compromise of the BBB.14,15 Unlike contrast-enhanced MRI, radiolabeled amino acid substrates transported by system L can visualize both contrast-enhancing and nonenhancing brain tumors.16,17 These biological properties make this class of PET tracer well suited for brain tumor imaging, particularly for defining gross tumor volume and tumor margins. Improved estimation of tumor volume and delineation of tumor margins has important implications for biopsy, resection, and radiation treatment planning.
Studies of amino acid PET tracers with histological confirmation of imaging findings have provide validation of PET imaging. These studies with MET, FET, and FDOPA have demonstrated that glioma boundaries are underestimated by contrast-enhanced MRI, and amino acid PET more accurately delineates both contrast-enhancing and nonenhancing brain tumors.18–20 Better delineation of tumor margins is of considerable importance given that gross total resection of gliomas is a primary goal of surgery and is associated with increased survival.21,22 The use of FDOPA-PET/MRI for presurgical planning has not been systematically investigated.
The objectives of this pilot study were: 1) to evaluate the role of FDOPA-PET/MRI for more accurately defining tumor volumes and surgical margins compared to MRI alone using histopathology as the reference standard; 2) to correlate FDOPA-PET/MRI imaging findings with tumor World Health Organization (WHO) grade and histology, and established molecular markers in gliomas: isocitrate dehydrogenase 1 (IDH1) R132H; p53; Ki-67; alpha-thalassemia/mental retardation, X-linked (ATRX); and chromosomes 1p and 19q; and 3) to determine the impact of FDOPA-PET/MRI on patient management. This study used simultaneous PET/MRI, which facilitates the implementation of multimodality, multiparametric techniques for producing and integrating morphologic, functional, and molecular imaging. This modality allows for a complete neuro-oncologic imaging assessment in a single session, reducing the patient’s imaging time and providing precise registration of the PET and MRI data. A key outcome for this study was to determine whether changes in patient management occurred with the addition of FDOPA-PET to the MR imaging, and what the changes were, as documented in the patient’s clinical record.
Methods
The present prospective study was conducted after approved by the Siteman Cancer Center Protocol Review Monitoring Committee and the Washington University School of Medicine institutional review board. Written informed consent was obtained from each of our patients prior to enrollment.
Study Population and Design
We studied 10 patients with known or suspected intracranial gliomas between May 2015 and July 2017 (5 men, 5 women; mean age ± SD, 43.8 ± 14.0 years; range, 25-66 years), who planned to undergo standard of care surgical resection and/or stereotactic biopsy under intraoperative MRI guidance. FDOPA-PET/MRI was performed prior to tissue sampling, allowing for correlation of imaging findings to histopathological results. Figure 1 presents the study design and Supplemental Table S1 patient demographic information, tumor location, and histology. For this study, substantial nonenhancing gliomas on MRI were defined as having contrast-enhancing volumes of less than 50% of the total estimated tumor volume. This inclusion criterion was intended to enrich the study population in patients with gliomas likely to be incompletely evaluated with MRI for tumor volume and margins.
![Study design for the FDOPA-PET/MRI preoperative evaluation of gliomas with substantial nonenhancing regions. FDOPA, 3,4-dihydroxy-6-[18F]fluoro-L-phenylalanine; HGG, high-grade glioma; LGG, low-grade glioma; VOI, volume of interest.](https://oup.silverchair-cdn.com/oup/backfile/Content_public/Journal/nop/7/6/10.1093_nop_npaa044/1/m_npaa044f0001.jpeg?Expires=1749042196&Signature=SNM73KDIH5ZByfwaodxlZhURiVEVj2vTnOSixCufb1pVIEYTD3b9MsmoHuyCyXvBIqIAiFuLaIE~couDsIsP8mUytZLFAGULMw0MaprIdp-f2UW9z9ORUkN0O2ebwHTKrMHbkGycePf86EleHze0gdJq6iT7082iFceLA4aHKz2P7ssmnyjwjaxw5CbxwefViz7FgG~k6HPcNSsfEbFerLP3kIWk2bNnUY5gq5C6X-w~OpmIFQhEc7KcWA6gutQzQw3Z3z~l~v6Z~4naT2MdersL3aU8qzlgKco8WrO5~UCLE0SV~Z0MeuYoPQpEL1pPJx8UJCcieJH8vqhPdkZEOA__&Key-Pair-Id=APKAIE5G5CRDK6RD3PGA)
Study design for the FDOPA-PET/MRI preoperative evaluation of gliomas with substantial nonenhancing regions. FDOPA, 3,4-dihydroxy-6-[18F]fluoro-L-phenylalanine; HGG, high-grade glioma; LGG, low-grade glioma; VOI, volume of interest.
Dynamic 18F-FDOPA PET/MRI Imaging
FDOPA was produced by the Washington University Cyclotron Facility using an electrophilic fluorination method with an automated system25 based on a previously reported synthetic pathway.26
The patients underwent FDOPA-PET/MRI imaging of the brain on a simultaneous 3-Tesla PET/MRI system, Siemens Biograph mMR (Siemens Health Care). After an intravenous injection of 5 mCi (185 MBq) ± 10% of 18F-FDOPA, dynamic PET data were acquired over the entire acquisition, corrected, and reconstructed using the 3-dimensional ordered-subset expectation maximization method at a framing rate of 20 × 3 seconds, 12 × 10 seconds, 6 × 20 seconds, 10 × 60 seconds, and at least 6 × 5 minutes. The attenuation correction was performed using the Dixon MRI sequence as recommended by the manufacturer.
Multisequence, multiplanar, contrast-enhanced MR images were acquired simultaneously during PET data acquisition using the standard-of-care protocol for brain tumor imaging at Washington University. The MRI included precontrast and postcontrast 3-dimensional volumetric acquisitions, T2-weighted, and fluid-attenuated inversion recovery (FLAIR) sequences to assess tumor margins with MRI, and a Dixon sequence for PET data attenuation correction. Intravenous contrast consisted of Multihance (Bracco Imaging) gadolinium-based contrast at a dosage of 0.1 mmol/kg.
Histopathology
The biopsy tissue samples underwent standard histopathological evaluation for gliomas at Washington University, and the information was used in the care of the patient. The tissue analysis included IDH1 R132H, ATRX, tp53, and Ki-67 labeling index, and evaluation for 1p/19q codeletions. Targeted next-generation sequencing was also performed for selected genetic mutations as part of clinical diagnostic workup based on the standard of care Genomics and Pathology Services. Gliomas were classified according to the WHO guidelines27 by a board-certified neuropathologist (S.D.).
Image Analysis
All FDOPA-PET/MRI studies were interpreted by a board-certified nuclear medicine radiologist (M.R.P.) and neuroradiologists (T.L.S.B./M.M.M.-T.). The patient’s clinical information was available to the reviewing radiologists.
The FDOPA-PET/MR images were reviewed and semiquantitative analysis performed using MIM Encore (MIM Software, Inc), a commercially available software package. Volumes of interest (VOIs) were drawn manually over the suspected brain tumor using the region of relatively increased FDOPA uptake (P+). For tumors that did not show visibly increased FDOPA uptake, MR T1 contrast enhancement (T+) and T2/FLAIR hyperintensity (F+) were used. The reference VOIs in normal brain were drawn over the contralateral hemisphere at the level of the suspected tumor. The reference VOI was used to generate standardized uptake value (SUV) thresholds, from 1.0 to 2.0 in 0.1 increments, and applied to the tumor VOI for each patient, defining a tumor volume at each SUV threshold value. The PET data summed at 10 to 15 minutes, the time of peak tumor uptake, and at 35 to 45 minutes, the time of decreasing uptake, were used in this analysis. Areas of FDOPA uptake in the tumor VOI above the 1.5-fold threshold were considered suspected tumor, with areas of discrepancy between tumor defined by MRI and FDOPA-PET noted. The 1.5-fold threshold was selected based on reported FET-PET values and our own experience with FDOPA-PET/MRI in pediatric brain tumor patients.28,29 PET analysis parameters included maximum and mean SUV (SUVmax, SUVmean), tumor to normal brain ratios (TBRs), and metabolic tumor volume (MTV) at the 1.5-fold threshold. TBRs were calculated by dividing the SUVmax of the tumor by the SUVmean of the reference region for TBRmax, the mean SUV of the tumor by the mean SUV of the reference region for TBRmean, and the maximum SUV of the tumor by the mean SUV at the 1.5-fold threshold for TBRmean1.5.
Decay-corrected dynamic PET time-activity curves (TACs) were generated to determine the time course of tracer uptake in the tumor, normal brain reference, and at the 1.5-fold threshold for the VOIs and biopsy sites. Time to peak (TTP) values were calculated, defined as the time in minutes to peak tracer uptake (SUVmax) in the tumor. The tumor retention index of the tumor VOIs and biopsy sites were obtained by dividing the SUVmean at 35 to 45 minutes by the SUVmean at 10 to 15 minutes and expressed as a percentage change. The TACs were obtained by projecting the VOIs and biopsy sites on the dynamic FDOPA-PET data over the entire acquisition period. TAC values within the first time frames (approximately 3 minutes) were excluded from the analyses and figures because of noise.
Assessment of the Role of FDOPA-PET/MRI on Clinical Patient Management
For each patient, in the planning session for the standard of care surgical resection and/or stereotactic biopsy, the MRI alone was first reviewed, and then the FDOPA-PET imaging added. The neurosurgeon would indicate whether the addition of FDOPA-PET changed the surgical plan, and any changes to the patient’s clinical management were recorded. When feasible and at the discretion of the neurosurgeon performing the resection, areas of suspected tumor identified on the FDOPA-PET but not the MRI alone would undergo tissue sampling. The comparison of FDOPA-PET/MR imaging findings with histopathological evaluation of biopsy sites provided a method to assess patient management in FDOPA-PET/MRI.
Correlation of Tissue Sampling and FDOPA-PET/MRI
Nine of 10 patients underwent standard of care surgical resection or stereotactic biopsy of their presumed glioma after their FDOPA-PET/MRI study. Tumor sampling was performed intraoperatively using a Stealth Station Neuronavigation System (Medtronic Sofamor Danek USA, Inc), and a subset of tissue sampling sites spatial coordinates were captured for correlation with the FDOPA-PET/MRI. The intraoperative MRI biopsies’ spatial coordinates were coregistered with the PET/MRI T1 contrast-enhanced and FDOPA-PET images and modeled as 5-mm spheres to minimize errors in sampling size and spatial coordinates. Correlations were performed between static and dynamic PET measures of FDOPA uptake with tumor grade, and the standard histopathological evaluation for gliomas, for the tissue specimens with spatial coordinates.
Statistical Analysis
The FDOPA-PET parameters and histopathology data from the tumor VOIs and biopsy sites were correlated using 3-parameter Weibull distribution curves modified with a vertical scale multiplier and fitted to the data. The maxima from these fits were used herein as maximum responses for the PET parameters of tumor SUVmean, SUVmax, and SUVmean1.5. Pearson correlations between the PET parameters and numeric histopathology data were performed, and P values less than .05 were considered statistically significant. Discriminant analysis was used to test for associations between nonnumeric histopathology variables and PET parameters. Additionally, the FDOPA-PET data were categorized into LGGs and HGGs and analyzed similarly. All statistical analysis was performed using SAS 9.4 software (SAS Institute Inc).
Results
All 10 patients successfully underwent simultaneous FDOPA-PET/MRI studies prior to the scheduled standard of care surgical resection and/or stereotactic biopsy. No adverse events related to the FDOPA-PET studies occurred, and the PET/MRI protocol was well tolerated by all patients. The initial diagnosis was suspected recurrent tumor in 4 patients, and newly diagnosed cerebral glioma in 6 patients. The FDOPA-PET/MR imaging was performed within 2 weeks of the surgical procedure in 8 patients. In patient 05, there was a 6-month interval between PET imaging and surgical excision given the absence of symptoms and the radiological stability of posttreatment changes. Conventional MR imaging follow-ups only were performed on patient 09, with suspected recurrence, and given the lack of FDOPA uptake to suggest malignancy, no biopsy was performed. In this study, 5 patients underwent resection and biopsy, 4 underwent biopsy only, and 1 patient underwent only MR imaging follow-up.
FDOPA PET/MRI Imaging Analysis
In 7 patients, the tumors were readily visualized with FDOPA-PET, demonstrating increased tracer uptake above the normal brain reference. In 2 patients, tracer uptake was similar to normal brain reference, and in 1 patient tracer uptake below the normal brain reference was demonstrated. Visual assessment of the MRI demonstrated no T1 contrast enhancement in 6 patients, and partial enhancement in 4 patients. The FDOPA-PET tumor regions were substantially greater in size than those defined by contrast enhancement in all patients. The T2/FLAIR hyperintensity extending beyond the regions delineated by FDOPA in 4 patients was approximately equal to the FDOPA in 2 patients, and in 1 patient, FDOPA extended beyond the region defined by T2/FLAIR hyperintensity (Table 1).
FDOPA PET/MRI Volumes of Interest and Biopsy Samples: Imaging Assessment and Histopathology
| Patient ID | Sampling | Histopathology | WHO grade | Imaging evaluation | Time to peak, MM:SS | Tumor SUVmean | Tumor SUVmax | TBRmax | TBRmean | TBRmean1.5 | Tumor retention index, % | Reference SUVmean | Ki-67 | MTV, cm3 | FDOPA extends beyond T1 | FDOPA extends beyond T2/FLAIR |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 01 | VOI | Anaplastic oligodendroglioma | III | P+, T±, F+ | 8:39 | 4.5 | 14.8 | 10.7 | 3.2 | 2.6 | 83% | 1.4 | 90% | 129.0 | Yes | Yes |
| 01 | Biopsy | Anaplastic oligodendroglioma | III | P+, T+, F+ | 6:02 | 4.1 | 4.5 | 3.2 | 2.1 | 2.9 | 59% | 70% | ||||
| 01 | Biopsy | Anaplastic oligodendroglioma | III | P+, T–, F+ | 6:42 | 8.5 | 10.4 | 7.5 | 5.0 | 6.1 | 65% | 2% | ||||
| 01 | Biopsy | Anaplastic oligodendroglioma | III | P+, T–, F+ | 11:22 | 12.3 | 13.1 | 9.3 | 6.2 | 8.8 | 58% | 5% | ||||
| 01 | Biopsy | Anaplastic oligodendroglioma | III | P+, T+, F+ | 9:22 | 7.0 | 7.6 | 5.4 | 3.6 | 5.0 | 77% | 2% | ||||
| 02 | VOI | Suspected autoimmune disease | N/A | P, T±, F+ | NA | NA | NA | NA | NA | NA | NA | 1.1 | N/A | N/A | N/A | N/A |
| 02 | Biopsy | Suspected autoimmune disease | N/A | P, T–, F+ | NA | NA | NA | NA | NA | NA | NA | |||||
| 03 | VOI | Oligodendroglioma | II | P+, T–, F+ | 14:28 | 2.7 | 5.0 | 3.7 | 2.0 | 2.0 | 71% | 1.4 | 5% | 36.4 | T- | NO |
| 03 | Biopsy | Oligodendroglioma | II | P+, T–, F+ | 7:28 | 3.2 | 3.5 | 2.5 | 1.7 | 2.3 | 80% | |||||
| 04 | VOI | Oligodendroglioma | II | P+, T–, F+ | 14:29 | 1.6 | 2.9 | 2.4 | 1.4 | 1.5 | 83% | 1.2 | 5% | 1.7 | T- | EQUAL |
| 04 | Biopsy | Oligodendroglioma | II | P+, T–, F+ | 10:29 | 2.1 | 2.3 | 1.9 | 1.3 | 1.7 | 64% | |||||
| 04 | Biopsy | Oligodendroglioma | II | P+, T–, F+ | 14:29 | 1.9 | 2.3 | 1.9 | 1.3 | 1.6 | 76% | |||||
| 05 | VOI | Anaplastic astrocytoma | III | P+, T–, F+ | 11:29 | 3.2 | 5.2 | 4.5 | 2.7 | 1.6 | 69% | 1.2 | 11% | 5.0 | T- | NO |
| 06 | VOI | Diffuse astrocytoma | II | P–, T–, F+ | 11:29 | 0.5 | 1.5 | NA | NA | NA | NA | 0.8 | 4% | N/A | N/A | N/A |
| 07 | VOI | Oligodendroglioma | II | P+, T–, F+ | 13:29 | 1.6 | 2.8 | 2.3 | 1.3 | 1.4 | 72% | 1.2 | 2% | 1.2 | T- | Equal |
| 07 | Biopsy | Oligodendroglioma | II | P+, T–, F+ | 8:29 | 1.6 | 1.9 | 1.6 | 1.0 | 1.3 | 70% | |||||
| 08 | VOI | Oligodendroglioma w/ focal anaplasia | III | P+, T±, F+ | 10:29 | 2.8 | 5.8 | 4.8 | 2.3 | 2.1 | 58% | 1.2 | 7% | 33.9 | Yes | No |
| 08 | Biopsy | Oligodendroglioma w/ focal anaplasia | III | P+, T–, F+ | 10:29 | 1.8 | 2.0 | 1.7 | 1.1 | 1.5 | NP | |||||
| 08 | Biopsy | Oligodendroglioma w/ focal anaplasia | III | P+, T–, F+ | 7:29 | 5.0 | 5.5 | 4.6 | 3.0 | 4.2 | NP | |||||
| 09 | VOI | No biopsy/surgery | N/A | P, T–, F– | NA | NA | NA | NA | NA | NA | NA | 1.3 | N/A | N/A | N/A | N/A |
| 10 | VOI | Glioblastoma | IV | P+, T±, F+ | 13:29 | 1.9 | 3.1 | 2.4 | 1.5 | 1.6 | 62% | 1.3 | N/P | 2.7 | Yes | No |
| 10 | Biopsy | Glioblastoma | IV | P+, T+, F+ | 10:29 | 2.8 | 3.1 | 2.4 | 1.6 | 2.2 | 52% | |||||
| 10 | Biopsy | No tumor | N/A | P, T–, F+ | NA | NA | NA | NA | NA | NA | NA |
| Patient ID | Sampling | Histopathology | WHO grade | Imaging evaluation | Time to peak, MM:SS | Tumor SUVmean | Tumor SUVmax | TBRmax | TBRmean | TBRmean1.5 | Tumor retention index, % | Reference SUVmean | Ki-67 | MTV, cm3 | FDOPA extends beyond T1 | FDOPA extends beyond T2/FLAIR |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 01 | VOI | Anaplastic oligodendroglioma | III | P+, T±, F+ | 8:39 | 4.5 | 14.8 | 10.7 | 3.2 | 2.6 | 83% | 1.4 | 90% | 129.0 | Yes | Yes |
| 01 | Biopsy | Anaplastic oligodendroglioma | III | P+, T+, F+ | 6:02 | 4.1 | 4.5 | 3.2 | 2.1 | 2.9 | 59% | 70% | ||||
| 01 | Biopsy | Anaplastic oligodendroglioma | III | P+, T–, F+ | 6:42 | 8.5 | 10.4 | 7.5 | 5.0 | 6.1 | 65% | 2% | ||||
| 01 | Biopsy | Anaplastic oligodendroglioma | III | P+, T–, F+ | 11:22 | 12.3 | 13.1 | 9.3 | 6.2 | 8.8 | 58% | 5% | ||||
| 01 | Biopsy | Anaplastic oligodendroglioma | III | P+, T+, F+ | 9:22 | 7.0 | 7.6 | 5.4 | 3.6 | 5.0 | 77% | 2% | ||||
| 02 | VOI | Suspected autoimmune disease | N/A | P, T±, F+ | NA | NA | NA | NA | NA | NA | NA | 1.1 | N/A | N/A | N/A | N/A |
| 02 | Biopsy | Suspected autoimmune disease | N/A | P, T–, F+ | NA | NA | NA | NA | NA | NA | NA | |||||
| 03 | VOI | Oligodendroglioma | II | P+, T–, F+ | 14:28 | 2.7 | 5.0 | 3.7 | 2.0 | 2.0 | 71% | 1.4 | 5% | 36.4 | T- | NO |
| 03 | Biopsy | Oligodendroglioma | II | P+, T–, F+ | 7:28 | 3.2 | 3.5 | 2.5 | 1.7 | 2.3 | 80% | |||||
| 04 | VOI | Oligodendroglioma | II | P+, T–, F+ | 14:29 | 1.6 | 2.9 | 2.4 | 1.4 | 1.5 | 83% | 1.2 | 5% | 1.7 | T- | EQUAL |
| 04 | Biopsy | Oligodendroglioma | II | P+, T–, F+ | 10:29 | 2.1 | 2.3 | 1.9 | 1.3 | 1.7 | 64% | |||||
| 04 | Biopsy | Oligodendroglioma | II | P+, T–, F+ | 14:29 | 1.9 | 2.3 | 1.9 | 1.3 | 1.6 | 76% | |||||
| 05 | VOI | Anaplastic astrocytoma | III | P+, T–, F+ | 11:29 | 3.2 | 5.2 | 4.5 | 2.7 | 1.6 | 69% | 1.2 | 11% | 5.0 | T- | NO |
| 06 | VOI | Diffuse astrocytoma | II | P–, T–, F+ | 11:29 | 0.5 | 1.5 | NA | NA | NA | NA | 0.8 | 4% | N/A | N/A | N/A |
| 07 | VOI | Oligodendroglioma | II | P+, T–, F+ | 13:29 | 1.6 | 2.8 | 2.3 | 1.3 | 1.4 | 72% | 1.2 | 2% | 1.2 | T- | Equal |
| 07 | Biopsy | Oligodendroglioma | II | P+, T–, F+ | 8:29 | 1.6 | 1.9 | 1.6 | 1.0 | 1.3 | 70% | |||||
| 08 | VOI | Oligodendroglioma w/ focal anaplasia | III | P+, T±, F+ | 10:29 | 2.8 | 5.8 | 4.8 | 2.3 | 2.1 | 58% | 1.2 | 7% | 33.9 | Yes | No |
| 08 | Biopsy | Oligodendroglioma w/ focal anaplasia | III | P+, T–, F+ | 10:29 | 1.8 | 2.0 | 1.7 | 1.1 | 1.5 | NP | |||||
| 08 | Biopsy | Oligodendroglioma w/ focal anaplasia | III | P+, T–, F+ | 7:29 | 5.0 | 5.5 | 4.6 | 3.0 | 4.2 | NP | |||||
| 09 | VOI | No biopsy/surgery | N/A | P, T–, F– | NA | NA | NA | NA | NA | NA | NA | 1.3 | N/A | N/A | N/A | N/A |
| 10 | VOI | Glioblastoma | IV | P+, T±, F+ | 13:29 | 1.9 | 3.1 | 2.4 | 1.5 | 1.6 | 62% | 1.3 | N/P | 2.7 | Yes | No |
| 10 | Biopsy | Glioblastoma | IV | P+, T+, F+ | 10:29 | 2.8 | 3.1 | 2.4 | 1.6 | 2.2 | 52% | |||||
| 10 | Biopsy | No tumor | N/A | P, T–, F+ | NA | NA | NA | NA | NA | NA | NA |
Abbreviations: F+, MR T2/FLAIR hyperintensity; F–, No MR T2/FLAIR hyperintensity; FDOPA, 3,4-dihydroxy-6-[18F]fluoro-L-phenylalanine; FLAIR, fluid-attenuated inversion recovery; Ki-67, cell proliferation index; MM:SS, minutes:seconds; MTV, metabolic tumor volume; NA, not applicable; NP, not performed; P, FDOPA-PET uptake similar to reference; P+, FDOPA-PET uptake above reference; P–, FDOPA PET uptake below reference; T+, MR T1 contrast enhancement; T–, no MR T1 contrast enhancement; SUV, standardized uptake value; SUVmax, maximum standardized uptake value; SUVmean, mean standardized uptake value; T±, partial MR T1 contrast enhancement; TBRmax, maximum standardized uptake value of tumor volume divided by mean standardized uptake value of reference volume; TBRmean, mean standardized uptake value of tumor volume divided by mean standardized uptake value of reference volume; TBRmean1.5, maximum standardized uptake value of tumor volume divided by mean standardized uptake value of 1.5-fold threshold; Tumor Retention Index, mean standardized uptake value at 35 to 45 minutes divided by mean standardized uptake value at 10 to 15 minutes, expressed as a percentage; w/, with; WHO, World Health Organization; VOI, volume of interest.
FDOPA PET/MRI Volumes of Interest and Biopsy Samples: Imaging Assessment and Histopathology
| Patient ID | Sampling | Histopathology | WHO grade | Imaging evaluation | Time to peak, MM:SS | Tumor SUVmean | Tumor SUVmax | TBRmax | TBRmean | TBRmean1.5 | Tumor retention index, % | Reference SUVmean | Ki-67 | MTV, cm3 | FDOPA extends beyond T1 | FDOPA extends beyond T2/FLAIR |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 01 | VOI | Anaplastic oligodendroglioma | III | P+, T±, F+ | 8:39 | 4.5 | 14.8 | 10.7 | 3.2 | 2.6 | 83% | 1.4 | 90% | 129.0 | Yes | Yes |
| 01 | Biopsy | Anaplastic oligodendroglioma | III | P+, T+, F+ | 6:02 | 4.1 | 4.5 | 3.2 | 2.1 | 2.9 | 59% | 70% | ||||
| 01 | Biopsy | Anaplastic oligodendroglioma | III | P+, T–, F+ | 6:42 | 8.5 | 10.4 | 7.5 | 5.0 | 6.1 | 65% | 2% | ||||
| 01 | Biopsy | Anaplastic oligodendroglioma | III | P+, T–, F+ | 11:22 | 12.3 | 13.1 | 9.3 | 6.2 | 8.8 | 58% | 5% | ||||
| 01 | Biopsy | Anaplastic oligodendroglioma | III | P+, T+, F+ | 9:22 | 7.0 | 7.6 | 5.4 | 3.6 | 5.0 | 77% | 2% | ||||
| 02 | VOI | Suspected autoimmune disease | N/A | P, T±, F+ | NA | NA | NA | NA | NA | NA | NA | 1.1 | N/A | N/A | N/A | N/A |
| 02 | Biopsy | Suspected autoimmune disease | N/A | P, T–, F+ | NA | NA | NA | NA | NA | NA | NA | |||||
| 03 | VOI | Oligodendroglioma | II | P+, T–, F+ | 14:28 | 2.7 | 5.0 | 3.7 | 2.0 | 2.0 | 71% | 1.4 | 5% | 36.4 | T- | NO |
| 03 | Biopsy | Oligodendroglioma | II | P+, T–, F+ | 7:28 | 3.2 | 3.5 | 2.5 | 1.7 | 2.3 | 80% | |||||
| 04 | VOI | Oligodendroglioma | II | P+, T–, F+ | 14:29 | 1.6 | 2.9 | 2.4 | 1.4 | 1.5 | 83% | 1.2 | 5% | 1.7 | T- | EQUAL |
| 04 | Biopsy | Oligodendroglioma | II | P+, T–, F+ | 10:29 | 2.1 | 2.3 | 1.9 | 1.3 | 1.7 | 64% | |||||
| 04 | Biopsy | Oligodendroglioma | II | P+, T–, F+ | 14:29 | 1.9 | 2.3 | 1.9 | 1.3 | 1.6 | 76% | |||||
| 05 | VOI | Anaplastic astrocytoma | III | P+, T–, F+ | 11:29 | 3.2 | 5.2 | 4.5 | 2.7 | 1.6 | 69% | 1.2 | 11% | 5.0 | T- | NO |
| 06 | VOI | Diffuse astrocytoma | II | P–, T–, F+ | 11:29 | 0.5 | 1.5 | NA | NA | NA | NA | 0.8 | 4% | N/A | N/A | N/A |
| 07 | VOI | Oligodendroglioma | II | P+, T–, F+ | 13:29 | 1.6 | 2.8 | 2.3 | 1.3 | 1.4 | 72% | 1.2 | 2% | 1.2 | T- | Equal |
| 07 | Biopsy | Oligodendroglioma | II | P+, T–, F+ | 8:29 | 1.6 | 1.9 | 1.6 | 1.0 | 1.3 | 70% | |||||
| 08 | VOI | Oligodendroglioma w/ focal anaplasia | III | P+, T±, F+ | 10:29 | 2.8 | 5.8 | 4.8 | 2.3 | 2.1 | 58% | 1.2 | 7% | 33.9 | Yes | No |
| 08 | Biopsy | Oligodendroglioma w/ focal anaplasia | III | P+, T–, F+ | 10:29 | 1.8 | 2.0 | 1.7 | 1.1 | 1.5 | NP | |||||
| 08 | Biopsy | Oligodendroglioma w/ focal anaplasia | III | P+, T–, F+ | 7:29 | 5.0 | 5.5 | 4.6 | 3.0 | 4.2 | NP | |||||
| 09 | VOI | No biopsy/surgery | N/A | P, T–, F– | NA | NA | NA | NA | NA | NA | NA | 1.3 | N/A | N/A | N/A | N/A |
| 10 | VOI | Glioblastoma | IV | P+, T±, F+ | 13:29 | 1.9 | 3.1 | 2.4 | 1.5 | 1.6 | 62% | 1.3 | N/P | 2.7 | Yes | No |
| 10 | Biopsy | Glioblastoma | IV | P+, T+, F+ | 10:29 | 2.8 | 3.1 | 2.4 | 1.6 | 2.2 | 52% | |||||
| 10 | Biopsy | No tumor | N/A | P, T–, F+ | NA | NA | NA | NA | NA | NA | NA |
| Patient ID | Sampling | Histopathology | WHO grade | Imaging evaluation | Time to peak, MM:SS | Tumor SUVmean | Tumor SUVmax | TBRmax | TBRmean | TBRmean1.5 | Tumor retention index, % | Reference SUVmean | Ki-67 | MTV, cm3 | FDOPA extends beyond T1 | FDOPA extends beyond T2/FLAIR |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 01 | VOI | Anaplastic oligodendroglioma | III | P+, T±, F+ | 8:39 | 4.5 | 14.8 | 10.7 | 3.2 | 2.6 | 83% | 1.4 | 90% | 129.0 | Yes | Yes |
| 01 | Biopsy | Anaplastic oligodendroglioma | III | P+, T+, F+ | 6:02 | 4.1 | 4.5 | 3.2 | 2.1 | 2.9 | 59% | 70% | ||||
| 01 | Biopsy | Anaplastic oligodendroglioma | III | P+, T–, F+ | 6:42 | 8.5 | 10.4 | 7.5 | 5.0 | 6.1 | 65% | 2% | ||||
| 01 | Biopsy | Anaplastic oligodendroglioma | III | P+, T–, F+ | 11:22 | 12.3 | 13.1 | 9.3 | 6.2 | 8.8 | 58% | 5% | ||||
| 01 | Biopsy | Anaplastic oligodendroglioma | III | P+, T+, F+ | 9:22 | 7.0 | 7.6 | 5.4 | 3.6 | 5.0 | 77% | 2% | ||||
| 02 | VOI | Suspected autoimmune disease | N/A | P, T±, F+ | NA | NA | NA | NA | NA | NA | NA | 1.1 | N/A | N/A | N/A | N/A |
| 02 | Biopsy | Suspected autoimmune disease | N/A | P, T–, F+ | NA | NA | NA | NA | NA | NA | NA | |||||
| 03 | VOI | Oligodendroglioma | II | P+, T–, F+ | 14:28 | 2.7 | 5.0 | 3.7 | 2.0 | 2.0 | 71% | 1.4 | 5% | 36.4 | T- | NO |
| 03 | Biopsy | Oligodendroglioma | II | P+, T–, F+ | 7:28 | 3.2 | 3.5 | 2.5 | 1.7 | 2.3 | 80% | |||||
| 04 | VOI | Oligodendroglioma | II | P+, T–, F+ | 14:29 | 1.6 | 2.9 | 2.4 | 1.4 | 1.5 | 83% | 1.2 | 5% | 1.7 | T- | EQUAL |
| 04 | Biopsy | Oligodendroglioma | II | P+, T–, F+ | 10:29 | 2.1 | 2.3 | 1.9 | 1.3 | 1.7 | 64% | |||||
| 04 | Biopsy | Oligodendroglioma | II | P+, T–, F+ | 14:29 | 1.9 | 2.3 | 1.9 | 1.3 | 1.6 | 76% | |||||
| 05 | VOI | Anaplastic astrocytoma | III | P+, T–, F+ | 11:29 | 3.2 | 5.2 | 4.5 | 2.7 | 1.6 | 69% | 1.2 | 11% | 5.0 | T- | NO |
| 06 | VOI | Diffuse astrocytoma | II | P–, T–, F+ | 11:29 | 0.5 | 1.5 | NA | NA | NA | NA | 0.8 | 4% | N/A | N/A | N/A |
| 07 | VOI | Oligodendroglioma | II | P+, T–, F+ | 13:29 | 1.6 | 2.8 | 2.3 | 1.3 | 1.4 | 72% | 1.2 | 2% | 1.2 | T- | Equal |
| 07 | Biopsy | Oligodendroglioma | II | P+, T–, F+ | 8:29 | 1.6 | 1.9 | 1.6 | 1.0 | 1.3 | 70% | |||||
| 08 | VOI | Oligodendroglioma w/ focal anaplasia | III | P+, T±, F+ | 10:29 | 2.8 | 5.8 | 4.8 | 2.3 | 2.1 | 58% | 1.2 | 7% | 33.9 | Yes | No |
| 08 | Biopsy | Oligodendroglioma w/ focal anaplasia | III | P+, T–, F+ | 10:29 | 1.8 | 2.0 | 1.7 | 1.1 | 1.5 | NP | |||||
| 08 | Biopsy | Oligodendroglioma w/ focal anaplasia | III | P+, T–, F+ | 7:29 | 5.0 | 5.5 | 4.6 | 3.0 | 4.2 | NP | |||||
| 09 | VOI | No biopsy/surgery | N/A | P, T–, F– | NA | NA | NA | NA | NA | NA | NA | 1.3 | N/A | N/A | N/A | N/A |
| 10 | VOI | Glioblastoma | IV | P+, T±, F+ | 13:29 | 1.9 | 3.1 | 2.4 | 1.5 | 1.6 | 62% | 1.3 | N/P | 2.7 | Yes | No |
| 10 | Biopsy | Glioblastoma | IV | P+, T+, F+ | 10:29 | 2.8 | 3.1 | 2.4 | 1.6 | 2.2 | 52% | |||||
| 10 | Biopsy | No tumor | N/A | P, T–, F+ | NA | NA | NA | NA | NA | NA | NA |
Abbreviations: F+, MR T2/FLAIR hyperintensity; F–, No MR T2/FLAIR hyperintensity; FDOPA, 3,4-dihydroxy-6-[18F]fluoro-L-phenylalanine; FLAIR, fluid-attenuated inversion recovery; Ki-67, cell proliferation index; MM:SS, minutes:seconds; MTV, metabolic tumor volume; NA, not applicable; NP, not performed; P, FDOPA-PET uptake similar to reference; P+, FDOPA-PET uptake above reference; P–, FDOPA PET uptake below reference; T+, MR T1 contrast enhancement; T–, no MR T1 contrast enhancement; SUV, standardized uptake value; SUVmax, maximum standardized uptake value; SUVmean, mean standardized uptake value; T±, partial MR T1 contrast enhancement; TBRmax, maximum standardized uptake value of tumor volume divided by mean standardized uptake value of reference volume; TBRmean, mean standardized uptake value of tumor volume divided by mean standardized uptake value of reference volume; TBRmean1.5, maximum standardized uptake value of tumor volume divided by mean standardized uptake value of 1.5-fold threshold; Tumor Retention Index, mean standardized uptake value at 35 to 45 minutes divided by mean standardized uptake value at 10 to 15 minutes, expressed as a percentage; w/, with; WHO, World Health Organization; VOI, volume of interest.
The normal brain reference VOIs yielded very similar mean SUVs within individual patients, with a reference SUVmean of 1.3 ± 0.1 (range, 1.2-1.4). These reference regions were used to generate thresholds for defining MTVs, and the 1.5-fold threshold was selected for further analysis based on the literature, given that the patient sample size in this study is too small to determine the optimal threshold for defining MTVs. The SUVmean of tumor VOIs based on the 1.5-fold threshold (SUVmean1.5) was 2.9 ± 1.3 (range, 1.9-5.6).
Imaging assessment and FDOPA parameters using the static PET images at 10 to 15 minutes and 30 to 45 minutes on the 7 patients with tracer uptake above reference are presented in Table 1. For the tumor VOIs with FDOPA uptake, mean ± SD for SUVmax was 5.6 ± 4.2 (range, 2.8-14.8), and SUVmean was 2.6 ± 1.0 (range, 1.6-4.5). The TBRs were TBRmax = 4.4 ± 3.0 (range, 2.3-10.7), TBRmean = 2.0 ± 0.7 (range: 1.3-3.2), and TBRmean1.5 = 1.8 ± 0.4 (range, 1.4-2.6). The MTV defined by the 1.5-fold threshold based on the normal brain uptake was 30.0 ± 46.3 cm3 (range, 1.2-129.0 cm3).
The dynamic PET analysis of VOIs and biopsy sites with imaging correlation demonstrated decreasing TAC patterns in all FDOPA-avid tumors, with SUV peaking early within approximately the first 10- to 15-minute interval followed by a constant descent thereafter. No increasing TAC pattern was identified in our cohort. The tumor VOIs TTP mean ± SD was 12:03 ± 2:02 minutes (range, 8:39-14:29 minutes), with the earliest VOI TTP, 8:39 minutes, corresponding to the highest Ki-67 value of 90% (patient 01), with longer TTP and lower Ki-67 values in the other patients. The 4 biopsies from patient 01, within different regions of the tumor, demonstrated a similar decreasing TAC pattern with TTPs at 6:02, 6:42, 11:22, and 9:22 minutes, and Ki-67 values of 70%, 2%, 5%, and 2%, respectively (Figure 2).
A 43-year-old woman with anaplastic oligodendroglioma grade III (patient 01). (A) FDOPA-PET/MRI: T1-weighted MRI with contrast (ceT1), Fluid-attenuated inversion recovery (FLAIR), FDOPA + ceT1, FDOPA + FLAIR at the biopsy sites (a-d, magenta dots), and total metabolic tumor volume (e, magenta outline) and normal brain reference (e, yellow outline). (B) Dynamic time-activity curves (TACs) of FDOPA-PET/MRI of tumor SUVmean and SUVmax, and normal brain reference SUVmean at biopsy sites (a–d) and total metabolic tumor volume (e). The TAC of SUVmax at the biopsy site (a) and total metabolic tumor volume (e) show an early peak following by a rapid decreasing kinetic pattern with a high cellular marker for proliferation (Ki-67) of 70% and 90%, respectively. The biopsy sites (b–d) show similar early SUV peak with slower decreasing patterns and lower Ki-67 values ranging from 2–5%.
Histopathology and Imaging Correlations
Nine of 10 patients underwent stereotactic biopsy of their presumed gliomas after their FDOPA-PET/MRI study, with histopathology confirmation of glioma in 8 patients. No biopsy was performed on patient 09, who showed no definitive clinical or radiological findings of recurrence with MR imaging follow-ups indicating stable postsurgical changes. Patient 02 had histopathology findings compatible with an autoimmune disease. Decreased tracer uptake, compared to normal brain background, was demonstrated in patient 06, with histopathology findings of diffuse astrocytoma grade II. The static 10- to 15-minute PET images demonstrated tumor SUV values (SUVmax = 1.0 and SUVmean = 0.5) lower than the normal brain reference (SUVmax = 1.4 and SUVmean = 0.8), with TACs over the entire acquisition period showing consistently lower tumor SUVs than reference (Figure 3).
A 30-year-old man with diffuse astrocytoma grade II (patient 06). (A) T1-weighted MRI with contrast (B) Fluid-attenuated inversion recovery (FLAIR) MRI (C) FDOPA-PET/MRI, (D-E) FDOPA-PET/MRI dynamic time-activity curves (TACs). Non-enhancing, FLAIR/T2 hyperintensity mass with no FDOPA uptake (A–C). The FDOPA tumor TACs show decreased tracer uptake (magenta trace) below the normal brain reference (cyan trace).
A total of 30 stereotactic biopsy samples on 9 patients were collected and underwent standard histopathological evaluation for gliomas, with the results summarized in Supplemental Table S2.
The spatial coordinates were recorded by the Stealth neuronavigational system on a subset (n = 13) of biopsy samples on 7 patients and coregistered with the FDOPA-PET/MRI. The visual assessment and PET analysis parameters SUVmax, SUVmean, TBRmax, TBRmean, and TBRmean1.5 for each biopsy with coordinates are given in Table 1 and correlated with tumor grade and histology. The biopsy results coregistered with the FDOPA-PET/MRI were used to evaluate concordant and discordant PET and MR regions. Eight biopsy sites had FDOPA uptake without T1 contrast enhancement, with histopathology of glioma in all 8 locations. Two patients had FDOPA uptake similar to reference with no histologic findings of malignancy, and these were patient 09 with no recurrent glioblastoma multiform, and patient 02 with findings suggestive of autoimmune disease. No biopsy site demonstrated positive T1 contrast enhancement and negative FDOPA uptake.
In the 7 patients with FDOPA uptake, 3 had low-grade tumors, and 4 had high-grade tumors. These were divided into LGGs and HGGs, and the tumor VOIs and biopsy site values for PET parameters, TTP, and tumor retention index in each group were analyzed. There was considerable overlap between the 2 groups, with no statistically significant difference between high-grade and low-grade tumors. In this study, TACs for high-grade and low-grade tumors were also similar in terms of TTP and tumor retention index.
Impact of FDOPA-PET on Patient Management
In several patients, FDOPA-PET imaging affected patient management. In patient 04, the original targets were diagnosed as benign upon histological examination, and 4 additional biopsies in the region of FDOPA uptake were obtained, consistent with oligodendroglioma grade II.
The conventional brain MRIs of patient 05, with a history of anaplastic astrocytoma grade III status postresection 3 years previously, were reported as stable posttreatment changes. The FDOPA-PET/MRI demonstrated uptake in the inferior rectus gyrus and superior aspect of the surgical cavity. Given the absence of symptoms and the apparent radiological stability, follow-up with MRI and FDG-PET/CT were performed with inconclusive findings. MR imaging 5 months after positive FDOPA uptake revealed interval new enhancement and extension of FLAIR abnormalities, confirmed on biopsy to be recurrent anaplastic astrocytoma grade III (Figure 4).
A 39-year-old man with recurrent anaplastic astrocytoma grade III (patient 05). (A) T1-weighted MRI with contrast (ceT1) (B) Fluid-attenuated inversion recovery (FLAIR) MRI (C) FDOPA-PET/MRI (D–E) FDG-PET, and (F) hematoxylin-eosin staining (60× objective). Nonenhancing (A), FLAIR/T2 hyperintensity in the left frontal lobe (B–C, white arrows). FDG-PET shows minimal tracer uptake (D–E, long black arrows). Tumor histology showed increased cellularity, atypia, and scattered mitoses (F, long black arrow).
Patient 08 presented with a new diagnosis of T2/FLAIR hyperintensity, a minimally enhancing, right frontal–based infiltrative mass extending through the corona radiata into the basal ganglia and thalamus. Preoperative tractography (Figure 5) demonstrated the lesion emanating from the corticospinal tract, making extensive resection not appropriate given the risk of neurological compromise. The FDOPA-PET images revealed tracer uptake deeper and smaller than the region of T2/FLAIR abnormality, with uptake projecting into the inferior right frontal lobe distant to the cortical fiber tracts. The biopsies targeted the area of increased FDOPA uptake to obtain the tumor’s histopathology, oligodendroglioma with focal anaplasia grade III, for treatment planning.
A 47-year-old man with oligodendroglioma with focal anaplasia grade III (patient 08). (A–B) FDOPA-PET/MRI tractography (C) FDOPA-PET/MRI fluid-attenuated inversion recovery (FLAIR) (D) hematoxylin-eosin staining (60× objective). Tractography (A–B, white arrows) show disruption of the corticospinal tracts by the FDOPA avid tumor (yellow arrows). FDOPA-PET demonstrates the tumor is deeper and smaller than the FLAIR abnormality (C, thin black arrow). The biopsy targeted the region of increased FDOPA uptake with a lower risk of neurologic compromise (C, thick black arrow). Histology displayed variably increased cellularity, atypia, and mitoses (D, long black arrow).
Patient 10 had presented with episodes of numbness on the left side, and a slight increase in the size of a minimally enhancing, right temporal lesion with extensive T2/FLAIR hyperintensity. The FDOPA-PET showed a small region of uptake within the T2/FLAIR hyperintensity. Two biopsies were performed, one in the region of FDOPA uptake and no increased relative cerebral blood volume consistent for glioblastoma grade IV. The second biopsy in an area of T2/FLAIR hyperintensity and negative FDOPA uptake did not demonstrate malignancy.
Statistical Analysis
The Pearson correlations between the PET parameters and histopathology parameters were highly and significantly correlated only with SUV parameters and the Ki-67 index. No statistically significant correlation that differentiated between LGGs and HGGs and WHO grade based on histopathology and PET parameters from the VOIs and biopsy sites was found. The correlations between the Ki-67 labeling index and SUV parameters were SUVmax: r = 0.985, P = .002; SUVmean: r = 0.946, P = .015; and SUVmean1.5: r = 0.954, P = .012 for a mixture of recurrent and new diagnosed tumors.
Discussion
To our knowledge, this is the first study to use simultaneous FDOPA-PET/MRI to integrate imaging findings with the precise registration of biopsy locations and histopathology, WHO grade, and established molecular markers, in gliomas and evaluate its impact on patient management.
Despite promising updates to the Response Assessment in Neuro-Oncology (RANO) criteria for HGGs30 and LGGs31 to assess tumor response on MRI, it continues to remain a challenge. The problems associated with the accurate evaluation of contrast enhancement secondary to BBB alterations due to treatment and possible tumor extent within nonspecific regions of T2/FLAIR hyperintensity still frequently require serial imaging follow-up or biopsy to provide the final diagnosis. Furthermore, despite the use of advanced imaging techniques such as MR spectroscopy, and diffusion-weighted and perfusion-weighted imaging to assist image interpretation, none have provided a reliable diagnosis of treatment-related effects vs true progression. Thus, the most recent recommendations from the RANO working group and European Association for Neurooncology emphasize the value of integrating PET imaging (FDG, FET, MET, FDOPA) into the clinical workflow to evaluate gliomas.32
Multiple human studies have explored the use of FDOPA-PET for imaging of newly diagnosed and previously treated brain tumors, including comparisons with MRI, FDG-PET, and MET-PET.16,33,34 FDOPA has been shown to be more accurate than FDG for imaging low-grade tumors, evaluating recurrent tumors, and distinguishing tumor recurrence from radiation necrosis.35–37 FDOPA has very similar brain tumor imaging properties when compared with MET and FET, as expected based on their shared mechanism of uptake by system L amino acid transport.24,28,38
Our study focused on the evaluation of totally or substantially nonenhancing gliomas using FDOPA-PET/MRI, given that MRI assessment is particularly challenging. The static and dynamic PET findings were correlated with histopathology for validation. We observed MTV as defined by FDOPA uptake extend beyond the region of T1 contrast enhancement in all our patients. In 4 patients, T2/FLAIR hyperintensity extended beyond the volume of FDOPA uptake, limiting MR characterization. Our data demonstrated that FDOPA-PET/MRI provides more accurate tumor delineation than MRI, and enhanced patient management by improving diagnostic accuracy, in agreement with other FDOPA-PET studies with histopathology validation.20,33
Analysis of dynamic amino acid PET TACs has been used to quantify and capture multiple aspects of tumor biology.34,39,40 Published studies of FET TACs have shown 3 distinct patterns, possible correlation with IDH1/2 mutational and 1p/19q codeletion status and differentiation between LGG and HGG.41,42 In contrast, we observed that FDOPA had only one pattern of TAC, an early TTP around 10 to 20 minutes after injection, followed by a decrease of FDOPA uptake with a variable slope. Previously reported studies have shown FDOPA uptake correlated significantly with Ki-67,43 especially in patients with newly diagnosed brain tumors.33 The few initial studies of dynamic FDOPA-PET in gliomas have inconsistent findings based on the use of 2-compartment models.34,39,44 A more recent study with FDOPA TACs demonstrated prediction of the molecular classification, the presence or absence of IDH mutations, in newly diagnosed gliomas.45 Our study results are not in complete agreement, which may be because of the highly specific study population. In our study, dynamic FDOPA-PET was independent of WHO grade and did not differentiate between LGGs and HGGs with statistical significance.
We noted reduced FDOPA uptake below normal background with histopathology findings of diffuse astrocytoma grade II in one patient. In the literature, amino acid reduced uptake in tumors was noted with FDOPA43 and FET,48 with the exact mechanism responsible not completely understood. A large study of negative FET-PET scans in glioma patients observed that those with photopenic defects had a greater risk of HGGs.49 This result suggests that not only increased but also decreased FDOPA uptake should be considered an abnormal finding with further assessment using dynamic TACs.
Numerous studies cited in the RANO report on PET imaging in gliomas32 have demonstrated that molecular and metabolic information provided by amino acid PET imaging complements conventional MR and benefits the clinical management of glioma patients. Changes in clinical management with FDOPA-PET/MRI occurred in several patients, and these included additional biopsies in the region of FDOPA uptake, changes to the surgical plan based on the involvement of eloquent areas, and confirmation of nonspecific MR findings. Multiparametric PET/MRI has the potential to radically alter clinical management by eliminating much of the complexity associated with coregistration of PET, conventional and advanced MRI, and biopsy data.
The main limitation of this prospective pilot study is the small number of patients preventing robust statistical correlation of PET parameters and histopathology data. Although a strong correlation between tumor SUV variables and cell proliferation index (Ki-67) was observed, further investigation will be required. Also, not all patients have their stereotactic biopsy spatial coordinates recorded on the Stealth neuronavigational system; thus, only a subset of biopsy sites was available. Despite these limitations, this pilot study demonstrated that FDOPA-PET/MRI provides better diagnostic information than either modality alone for noncontrast-enhancing gliomas, with PET complementing conventional and advanced MRI to improve diagnostic accuracy, tumor delineation, and patient management.
Conclusions
In our pilot study, simultaneous FDOPA-PET/MRI allowed for more accurate tumor visualization and delineation in enhancing and nonenhancing gliomas, enabling more appropriate patient management in a significant fraction of patients. Visual and quantitative assessment of static and dynamic FDOPA findings with histopathology at tumor VOIs and biopsy sites allowed for better characterization of tumor extent and biology, which may lead to improved diagnosis and treatment. The multimodality and multiparametric functionality of PET/MRI positively affected clinical management in our study and has the potential to reduce the complexity associated with the complementary roles of PET and MRI.
Funding
This work was supported by the Barnes-Jewish Hospital Cancer Frontier Fund.
Conflict of interest statement
Dr McConathy has consulted and spoken at PET/MRI user meetings for GE Healthcare and Siemens Healthcare. Dr Benzinger has consulted for Eli Lilly and received research funding from Avid Radiopharmaceuticals.
