Utility and limitations of ¹⁸F-fluciclovine PET for biopsy targeting in a nonenhancing diffuse glioma with a gliomatosis cerebri–like growth pattern: illustrative case

Abstract

BACKGROUND

Contrast-enhanced MRI reflects blood-brain barrier (BBB) disruption rather than tumor biological activity, whereas ¹⁸F-fluciclovine (¹⁸F-FACBC) PET reflects amino acid metabolism independent of BBB integrity. The authors compared regional ¹⁸F-FACBC uptake with histopathological and molecular findings in a diffuse astrocytoma, IDH-wildtype, not elsewhere classified, with a gliomatosis cerebri (GC)–like infiltrative growth pattern.

OBSERVATIONS

A woman in her 50s presented with headache. MRI revealed a nonenhancing, diffusely infiltrating right parietal white matter lesion. ¹⁸F-FACBC PET showed focal uptake that guided stereotactic biopsy. Six weeks later, progression caused left uncal herniation requiring emergency decompression, during which PET-negative left temporal tissue was resected. The PET-positive region showed a high Ki-67 index, high-level PDGFRA amplification, and CDKN2A/B homozygous deletion, with a methylation profile matching pediatric-type diffuse high-grade glioma, RTK1 subclass. The PET-negative region was predominantly gliotic, with only low-level PDGFRA amplification. Thus, uptake marked a more proliferative, molecularly altered component. However, both the fatal mass effect and out-of-field recurrence following chemoradiotherapy occurred in regions without corresponding uptake.

LESSONS

¹⁸F-FACBC PET identified an aggressive, high-risk tumor component, supporting its use for biopsy targeting in nonenhancing gliomas. However, because life-threatening progression occurred without corresponding uptake, ¹⁸F-FACBC uptake reflects proliferative burden rather than the full infiltrative extent.

https://thejns.org/doi/10.3171/CASE26761

MRI, owing to its superior soft tissue contrast, remains indispensable for the assessment of brain tumors.¹ However, gadolinium enhancement primarily reflects blood-brain barrier (BBB) disruption and does not necessarily correspond to tumor cell density, proliferative activity, or the full extent of infiltrative disease.

The amino acid tracer ¹⁸F-fluciclovine (¹⁸F-FACBC) is transported into tumor cells mainly via ASCT2 (alanine-serine-cysteine ­transporter 2) and LAT1 (l-type amino acid transporter 1).^2,3 Since tracer uptake is not dependent on BBB disruption, ¹⁸F-FACBC PET can depict nonenhancing lesions and infiltrative tumor margins that may be inconspicuous on contrast-enhanced MRI.⁴ Previous studies have shown that fluciclovine uptake correlates with the histological grade, invasiveness, and Ki-67 labeling index of gliomas, and this tracer has been proposed as an adjunct for biopsy targeting and tumor delineation.^5–7

Gliomatosis cerebri (GC) is no longer recognized as a distinct tumor entity in the current WHO classification, but rather as a diffuse infiltrative growth pattern (GC-like growth pattern) observed in adult-type diffuse gliomas.⁸ Because GC-like diffuse gliomas extensively infiltrate the cerebral white matter, accurate biopsy targeting and assessment of the disease extent are particularly challenging. The clinical utility of ¹⁸F-FACBC PET in this infiltrative growth pattern has not yet been well characterized.

In the 2021 WHO Classification of Tumours of the Central Nervous System, the suffix not elsewhere classified (NEC) is applied when the required diagnostic workup has been successfully completed; however, the integrated histological, immunohistochemical, and molecular findings do not fit a recognized WHO tumor type. Within this framework, an IDH-wildtype diffuse astrocytic glioma that lacks the molecular features required for glioblastoma, IDH-wildtype, including EGFR amplification, combined whole-chromosome 7 gain/10 loss (+7/−10), and TERT promoter mutation, may remain classified as diffuse astrocytoma, IDH-wildtype, NEC.^9,10 The clinical behavior of this category is heterogeneous, and optimal treatment strategies remain undefined.¹¹ Moreover, whether focal ¹⁸F-FACBC uptake corresponds to molecularly high-risk tumor components—and whether it reflects the extent of infiltrative disease responsible for clinical progression—within nonenhancing NEC tumors remains unclear.

Here, we present an illustrative case of diffuse astrocytoma, IDH-wildtype, NEC, with a GC-like infiltrative growth pattern, in which regional ¹⁸F-FACBC uptake was compared with histopathological, genomic, and DNA methylation findings, including a PET-negative region from which the fatal mass effect subsequently arose.

Illustrative Case

A woman in her 50s developed a persistent orbital-to-frontal headache following COVID-19 infection. Two months later, an MRI examination performed at a local hospital revealed a diffusely infiltrating lesion centered in the right parietal white matter. Under a working diagnosis of postinfectious inflammatory/autoimmune CNS disorder, she was admitted to the neurology department and received intravenous steroid pulse therapy for 3 days. However, follow-up MRI showed no reduction in lesion size, raising suspicion of a brain tumor, and the patient was referred to our department.

At our institution, MRI revealed a lesion centered in the right parietal white matter extending into the white matter of both cerebral hemispheres, including the splenium of the corpus callosum, in a pattern suggestive of GC-like infiltrative growth. The lesion was hypointense on T1-weighted imaging and hyperintense on fluid-attenuated inversion recovery (FLAIR), showed diffusion restriction on diffusion-weighted imaging, and demonstrated no increase in cerebral blood volume or enhancement on gadolinium-enhanced T1-weighted imaging (Fig. 1A–E). MR spectroscopy demonstrated an elevated choline-to–n-acetylaspartate ratio, indicating a neoplastic process. As glioma was suspected, ¹⁸F-FACBC PET was performed to define the biopsy target and revealed focal uptake in the right parietal lobe (maximum standardized uptake value [SUVmax] 0.96, contralateral normal-appearing white matter SUVmax 0.3, tumor-to-background ratio 3.2) (Fig. 1F). No other areas of increased ¹⁸F-FACBC uptake were identified in the brain (Fig. 1G).

FIG. 1.

Preoperative axial MR and ¹⁸F-FACBC PET images. A and B: T1-weighted (A) and FLAIR (B) images showing a diffusely infiltrating, nonenhancing lesion centered in the right parietal white matter, extending across the midline through the splenium of the corpus callosum, which was hypointense on the T1-weighted image and hyperintense on the FLAIR image. C: Diffusion-weighted image demonstrating diffusion restriction in the right parietal region. D:Cerebral blood volume map showing no corresponding increase in perfusion. E: Gadolinium-enhanced T1-weighted image showing no enhancement. F: ¹⁸F-FACBC PET image demonstrating focal tracer uptake in the right parietal lobe (SUVmax 0.96, tumor-to-background ratio 3.2), which was selected as the stereotactic biopsy target (arrow). G:No other areas showing increased ¹⁸F-FACBC uptake were identified in the brain. H: Postbiopsy noncontrast CT image confirming accurate sampling of the targeted PET-positive region (arrow).

A stereotactic biopsy of the PET-positive region was performed as planned (Fig. 1H) and showed atypical glial cells and spongiform changes, accompanied by histiocytic infiltration and scattered small lymphocytes, although microvascular proliferation and necrosis were absent (Fig. 2A). Immunohistochemistry was negative for IDH1 R132H and showed a Ki-67 labeling index of 60% (hotspot) (Fig. 2B). Although these findings raised the possibility of high-grade glioma, a definitive classification was deferred pending genomic and DNA methylation analyses.

FIG. 2.

Histopathological findings of the PET-positive (A and B) and PET-negative (C) regions. A: Hematoxylin and eosin (H&E) staining showing an infiltrating glial neoplasm composed of atypical cells with enlarged round nuclei and plump, astrocyte-like cells with spongiform change, accompanied by scattered small lymphocytes and histiocytes. Microvascular proliferation and necrosis were absent. B:Ki-67 immunostaining showing a high labeling index (60%). C: H&E staining revealing gliosis with only mildly increased cellularity, composed of small cells with scant cytoplasm. Scale bars = 100 µm.

Approximately 6 weeks after the biopsy, the patient developed headaches and vomiting due to increased intracranial pressure. She was comatose (Glasgow Coma Scale score of 6) with a fixed, dilated left pupil. Imaging showed left-predominant interval disease progression with marked cerebral edema and left uncal herniation (Fig. 3A–D). Emergency decompressive craniotomy with internal decompression was performed, during which the PET-negative tissue from the herniating left temporal lobe was resected and examined. Postoperative MRI and CT confirmed adequate decompression (Fig. 3E–H).

FIG. 3.

Perioperative axial CT and MR images obtained at the time of uncal herniation and after emergency decompression. A–D: Preoperative noncontrast CT (A and B) and FLAIR (C and D) images showing marked, left-predominant cerebral edema with effacement of the basal cisterns, compression of the lateral ventricles, and midline shift. The arrow indicates left uncal herniation. E–H:Postoperative noncontrast CT (E and F) and FLAIR (G and H) images obtained after emergency decompressive craniotomy with internal decompression of the left temporal lobe, confirming adequate decompression with reexpansion of the previously compressed lateral ventricles and reduced mass effect. The left temporal resection cavity is visible (E and G). The PET-negative specimen was obtained from the resected left temporal lobe.

Genomic analysis of an earlier stereotactic biopsy specimen using a bimodal custom DNA/RNA panel¹² showed that the PET-positive site was IDH-wildtype and lacked EGFR amplification, combined whole-chromosome 7 gain/10 loss (+7/−10), and a TERT promoter mutation. As the tumor did not meet the molecular criteria for glioblastoma, IDH-wildtype, and the integrated histological, immunohistochemical, and molecular findings did not correspond to any other established WHO entity, it was diagnosed as diffuse astrocytoma, IDH-wildtype, NEC, although it harbored molecular alterations associated with high-grade behavior, including high-level PDGFRA amplification (copy number 45.0), CDKN2A/B homozygous deletion, and an unmethylated MGMT promoter. In contrast, the PET-negative left temporal region was consistent with gliosis (Fig. 2C). Next-generation sequencing revealed low-level PDGFRA amplification (copy number 4.3) as the only detected alteration. Therefore, both regions shared PDGFRA amplification, but the higher amplification level and additional high-risk alterations were confined to the PET-positive region. On DNA methylation analysis, the PET-positive right parietal lesion clustered most closely with pediatric-type diffuse high-grade glioma, RTK1 subclass (RTK1C), with a high calibrated score of 0.951 despite the patient’s age (Fig. 4). In contrast, the PET-negative region showed a methylation profile consistent with a control brain tissue class, with a high calibrated score of 0.843, which was consistent with nonneoplastic cortical tissue.

FIG. 4.

DNA-methylation–based classification. Uniform manifold approximation and projection of the tumor methylation profile relative to the reference CNS tumor methylation classes (color key on right). The present tumor (“Our case,” black dot) maps most closely to pediatric-type diffuse high-grade glioma, RTK1C (pedHGG_RTK1C; calibrated score 0.951), despite the patient’s adult age.

Given the high-grade molecular features, treatment followed a glioblastoma protocol, with radiotherapy (40.5 Gy in 15 fractions) and concomitant temozolomide (75 mg/m²/day) plus bevacizumab (10 mg/kg) beginning approximately 2 weeks after the emergency decompression. As the disease was diffuse and bihemispheric, the radiation field encompassed the supratentorial brain while sparing the posterior fossa. Given the large treatment volume, hypofractionated radiotherapy at a reduced total dose was administered. Bevacizumab was added to control severe cerebral edema and mass effects that persisted despite the emergency decompression. After treatment, her Karnofsky Performance Status score was 60. She was transferred to another facility where maintenance temozolomide and bevacizumab were continued every 4 weeks. Approximately 6 months after the start of chemoradiotherapy, MRI demonstrated a recurrent lesion in the cerebellum outside the initial radiation field, which showed no contrast enhancement, mirroring the nonenhancing appearance at presentation (Fig. 5). The cerebellar lesion progressively enlarged, causing obstructive hydrocephalus and herniation, and the patient died approximately 12 months after diagnostic biopsy.

FIG. 5.

Cerebellar recurrence approximately 6 months after chemoradiotherapy. A: Axial FLAIR image obtained before biopsy at the time of initial diagnosis, showing an unremarkable posterior fossa. B: Axial FLAIR image obtained at recurrence, demonstrating a new cerebellar lesion (arrow). C: Gadolinium-enhanced T1-weighted image showing no enhancement, mirroring the nonenhancing appearance of the supratentorial lesion at presentation. Cerebellar recurrence occurred outside the initial radiation field.

Informed Consent

The necessary informed consent was obtained in this study.

Discussion

Observations

We report a case of diffuse astrocytoma, IDH-wildtype, NEC, presenting as a nonenhancing, bihemispheric white matter lesion with a GC-like infiltrative pattern, in which ¹⁸F-FACBC PET was used to select the biopsy target and establish the diagnosis. As disease progression caused left uncal herniation, decompressive surgery yielded tissue from a PET-negative left temporal region, enabling paired histopathological, molecular, and DNA methylation comparisons. The key observation was twofold: focal ¹⁸F-FACBC uptake corresponded to a histologically overt tumor component carrying high-risk molecular alterations, whereas the fatal mass effect and subsequent recurrence both arose from regions without corresponding uptake. Thus, ¹⁸F-FACBC uptake marked the proliferative, molecularly aggressive component, but did not capture the diffuse infiltrative extent that ultimately governed the clinical course.

In the PET-positive region, the uptake corresponded to an infiltrating glial neoplasm composed of atypical cells with spongiform change (without microvascular proliferation or necrosis), a markedly high Ki-67 labeling index (60%), high-level PDGFRA amplification, a CDKN2A/B homozygous deletion, and an unmethylated MGMT promoter. The Ki-67 index has been reported as a prognostic factor in diffuse astrocytoma, IDH-wildtype, NEC, and is associated with alterations such as PDGFRA/KIT amplification and ATM mutation;¹¹ PDGFRA amplification is itself an adverse prognostic factor supporting a biologically aggressive phenotype.^13–15 As ¹⁸F-FACBC uptake reflects amino acid transport and metabolic activity,⁶ the focal signal plausibly corresponds to this proliferative, highly cellular component. Although NEC is reported to have a somewhat better prognosis than conventional glioblastoma, its behavior is heterogeneous and partly driven by molecular risk;¹¹ the combination of these alterations with high proliferative activity likely underlie early recurrence and poor outcome, underscoring that NEC should not be regarded as a uniformly low-risk category.

By contrast, the fatal mass effect and left uncal herniation originated in the PET-negative left temporal lobe, which was gliosis predominant and classified as “control brain tissue” by DNA methylation. Because this classification requires adequate neoplastic cell content, and low-level PDGFRA amplification was nonetheless retained, this region reflected low tumor cell purity rather than a true absence of tumor. This dissociation between local histology and clinical behavior is consistent with the GC-like growth pattern, now regarded as a diffuse infiltrative pattern of adult-type gliomas rather than a distinct entity⁸ characterized by widespread infiltration beyond MRI-visible lesions and marked spatial heterogeneity.¹⁶ Together, these findings indicate that ¹⁸F-FACBC uptake primarily reflects tumor cellularity and proliferative activity and may not predict the extent of diffuse infiltrative progression or the development of a life-threatening mass effect.

Several amino acid PET tracers, including ¹¹C-methionine, ¹⁸F-fluoroethyl-l-tyrosine (¹⁸F-FET), and ¹⁸F-fluoro-l-DOPA, have established roles in glioma imaging.¹⁷ Among these, ¹⁸F-FET has been more extensively validated for glioma imaging, whereas ¹⁸F-FACBC remains newer and less extensively studied.^{17,18 18}F-FACBC offers very low physiological uptake in normal brain, potentially yielding favorable tumor-to-brain contrast;^17,18 in our case, a tumor-to-background ratio of 3.2 identified the biologically active component despite a modest SUVmax of 0.96. However, ¹⁸F-FACBC lacks the larger evidence base and established interpretation thresholds available for ¹⁸F-FET. Conversely, even ¹⁸F-FET PET–defined margins may not encompass microscopic infiltrative tumor, and quantitative metrics are not directly interchangeable between tracers because of differences in kinetics and imaging protocols. Direct head-to-head studies are therefore needed to define the relative strengths and limitations of ¹⁸F-FACBC and ¹⁸F-FET for biopsy targeting and diffuse tumor delineation.

On DNA methylation analysis, the tumor was confidently classified as pediatric-type diffuse high-grade glioma, RTK1C (calibrated score 0.951), despite the patient’s adult age. Because PDGFRA alterations are characteristic of this group,^19,20 this classification provides a biological context for the aggressive, PDGFRA-driven phenotype, although its clinical significance in adult NEC tumors remains uncertain. This case illustrates that NEC may encompass diagnostically challenging tumors that fall within the recognized pediatric-type RTK-driven high-grade glioma methylation classes. Notably, both the initial lesion and cerebellar recurrence remained nonenhancing, indicating that a biologically aggressive disease may remain radiologically occult on contrast-enhanced MRI throughout its clinical course.

This report has several limitations. It describes a single case, and the PET-positive and PET-negative specimens were obtained from anatomically distinct sites approximately 6 weeks apart; therefore, the observed differences reflect regional heterogeneity and interval disease progression in addition to uptake-related variation, and no causal relationship between tracer uptake and specific molecular alterations can be inferred. Nonetheless, this paired comparison offers insights into intratumoral heterogeneity in NEC, and prospective studies integrating molecular subtyping with amino acid PET are warranted.

Lessons

In this diffuse astrocytoma, an IDH-wildtype, NEC, with a GC-like infiltrative growth pattern, focal ¹⁸F-FACBC uptake marked a tumor component with higher proliferative activity and high-risk molecular alterations, including high-level PDGFRA amplification and CDKN2A/B homozygous deletion. This supports the use of amino acid PET for targeted biopsy of biologically aggressive components that are inconspicuous on contrast-enhanced MRI, particularly in nonenhancing gliomas with ambiguous molecular classifications. However, the life-threatening mass effect and subsequent recurrence both arose from PET-negative regions, indicating that the uptake reflects the proliferative burden rather than the full infiltrative extent. Therefore, amino acid PET should complement, rather than replace, the assessment of diffuse infiltrative burden.

Disclosures

The authors report no conflict of interest concerning the materials or methods used in this study or the findings specified in this paper.

Author Contributions

Conception and design: Higa, Matsueda, Makino, Yonezawa. Acquisition of data: Matsueda, Makino, Akahane, Yonezawa, Tanimoto. Analysis and interpretation of data: Matsueda, Akahane, Kirishima, Yonezawa. Drafting the article: Matsueda. Critically revising the article: Hanaya. Reviewed submitted version of manuscript: Matsueda, Kirishima, Yonezawa, Hanaya. Approved the final version of the manuscript on behalf of all authors: Higa. Administrative/technical/material support: Yonezawa. Study supervision: Higa, Hanaya.

Correspondence

Nayuta Higa: Kagoshima University Graduate School of Medical and Dental Sciences, Kagoshima, Japan. nayuhiga@m.kufm.kagoshima-u.ac.jp.