5-Aminolevulinic Acid in Low-Grade Gliomas

5 5-Aminolevulinic Acid in Low-Grade Gliomas

Georg Widhalm, Mitchel S. Berger, and Johannes Wölfer

Abstract
5-aminolevulinic acid (5-ALA)-induced tumor fluorescence represents a powerful tool to optimize the resection of high-grade gliomas (HGG). Use of 5-ALA for the fluorescence-guided surgery (FGS) of low-grade glioma (LGG) tumors in patients is actively being studied. Currently, 5-ALA-induced fluorescence helps identify anaplastic tumor areas in diffuse infiltrating gliomas without MRI contrast enhancement that would have gone unnoticed otherwise, potentially subjecting patients to insufficient therapies. Further, specific imaging and metabolic parameters that enhance the propensity of suspected low-grade tumors to show significant and surgically useful fluorescence have been identified. However, the task of intraoperatively identifying the margins of a diffuse, straight LGG cannot yet be reliably addressed by the technique in its current setup, even though the differences of 5-ALA metabolism between LGG and HGG seem to be only quantitative rather than qualitative. Introducing methods such as fluorescence spectroscopy or confocal microscopy might considerably expand the field of FGS in order to optimize intraoperative visualization and thus maximize resection of LGG.

Keywords: 5-ALA, low-grade gliomas, anaplastic focus, sampling error, fluorescence spectroscopy, confocal microscopy

5.1 Introduction

Selective intraoperative visualization of malignant gliomas has become feasible with the introduction of 5-aminolevulinic acid (5-ALA)-induced fluorescence into the neurosurgical field in the last years. Fluorescence-guided resection using 5-ALA has been established as standard for surgery of high-grade gliomas (HGG) at many neurosurgical centers worldwide. Low-grade gliomas (LGG) were generally thought to be inaccessible to this technique, but recent data do provide support in estimating its usefulness in a subgroup of these tumors that do not show the classical hallmarks of malignant gliomas. This chapter will try to span the distance between the evidence for the use of 5-ALA in LGGs with the current visualization techniques and the possibly extended range of indications with future developments.

5.2 Low-Grade Gliomas

5.2.1 Background

Diffusely infiltrating gliomas (DIGs) represent the most common primary brain tumors in adults.1 Based on specific histopathological criteria defined by the World Health Organization (WHO), the usually slow-growing LGG (WHO grade II) are distinguished from the aggressive HGG (WHO grades III and IV).¹ Annually, 2,700 to 4,600 cases of LGG are diagnosed in the United States, and such tumors frequently present in the second to fourth decade.^2,3 The first clinical symptom of LGG is the occurrence of new epileptic seizures in up to 80% of cases.² Further frequent symptoms include changes of mental status, clinical signs of increased intracranial pressure, and focal neurological deficits.⁴

5.2.2 Preoperative Imaging

MRI represents the imaging technique of choice to further investigate suspected LGG.⁵ To detect a potential disruption of the blood–brain barrier (BBB), additional administration of contrast medium during the MR investigation is of major importance.⁵ Contrary to HGG, significant contrast enhancement (CE) on MRI is usually absent in LGG.⁶ To further characterize suspected LGG, advanced MRI methods such as MR spectroscopy (MRS), perfusion MRI, diffusion-weighted images (DWI), and diffusion tensor imaging (DTI) were established in the last decade.^5,7 Additionally, positron emission tomography (PET) using amino acid tracers such as ¹⁸F-fluoroethyl-L-tyrosine (FET) and ¹¹C-methionine (MET) is a powerful technique to characterize suspected LGG.^8,9,10

5.2.3 Treatment

Generally, the primary treatment in patients with suspected LGG is surgical resection whenever possible. The formerly often advocated “watch and wait” strategy has been replaced by an early and aggressive surgical treatment approach. This has been supported by a Norwegian retrospective study comparing early surgery to observation after biopsy of suspected LGG tumors.¹¹ Early surgical management was associated with better overall survival.¹¹ Nowadays, there is clear evidence that more extensive resections of LGG result in improved patient outcomes.^12,13,14 Additionally, maximal resection of LGG tumors prolongs the time span to malignant transformation.¹⁵ Consequently, the current aim of surgery in suspected LGG is maximum safe tumor removal with preservation of neurological function. The Response Assessment in Neuro-Oncology (RANO) criteria define a complete resection of LGG as total removal of the abnormality on preoperative MRI T2-weighted/fluid-attenuated inversion recovery (FLAIR) sequences.^6,16 To avoid new postoperative neurological deficits and to achieve maximal safe resection, brain mapping and intraoperative stimulation have nowadays become indispensable techniques in LGG surgery.¹⁷ Additionally, intraoperative navigation with DTI data provides a powerful tool to localize and avoid injury to relevant white matter tracts in LGG surgery.⁷ Ultrasound may be used as an easily available adjunct, even though echodensity of LGG is variable and often incongruent to T2 and/or FLAIR by MRI.¹⁸

After surgical resection, a postoperative “watch and wait” strategy with regular imaging follow-up, but without initial adjuvant treatment, is frequently conducted in LGG patients.¹⁹ However, new molecular markers such as isocitrate dehydrogenase 1 (IDH1)/IDH2 mutation and 1p19q co-deletion status are important prognostic factors that also impact treatment decisions with such tumors.^1,20

5.2.4 Shortcomings of LGG Surgery and Current Solutions

Suspected LGG pose a special challenge for the neurosurgeon both in the preoperative planning phase and during resection itself. Typically, surgery of gliomas with nonsignificant CE on MRI is associated with specific drawbacks. First of all, incomplete resection of LGG is reported in the literature in 54 to 88% of cases resulting in worse patient prognosis.12,15,21,22 One of the major causes for incomplete resection is insufficient visualization of the margin of LGG during resection since these tumors demonstrate only slight differences in macroscopic appearance and texture consistency compared to normal brain. Moreover, histopathological undergrading of gliomas due to the so-called sampling error is not uncommon.^8,23 Gliomas often show histopathological heterogeneity, and thus circumscribed intratumoral areas of malignant transformation, so-called anaplastic foci, may arise in initial pure LGG.²³ Missing such an anaplastic focus can lead to histopathological undergrading and inadequate postoperative patient management.

To improve LGG resection, the use of neuronavigation, navigation-guided tissue sampling from a PET or MRI/MRS “hotspot” to avoid undergrading, and intraoperative MRI have been proposed.^10,24,25,26 However, navigation systems tend to lose their initial accuracy during the course of resection due to brain shift, while intraoperative MRI is time-consuming and expensive, and therefore not widely available. Consequently, different surgical tools are required to overcome the current limitations of LGG surgery.

5.3 5-Aminolevulinic Acid in Low-Grade Glioma

Intraoperative visualization of malignant brain tumor tissue using 5-ALA-induced fluorescence is a well-validated tool to optimize resection. 5-ALA fluorescence-guided surgery (FGS) is relatively inexpensive, widely available, and remains unaffected by intraoperative brain shift. Although this method has been primarily applied in HGG with significant CE on MRI,^27,28 there has been growing interest to analyze the value of 5-ALA-induced fluorescence in suspected LGG with nonsignificant CE as well.

5.3.1 Background

Initially, the breakdown of the BBB was felt to be a prerequisite of visible 5-ALA-induced fluorescence in brain tumors, and thus only HGG with significant CE were thought to be amenable to this technique.²⁸ Currently, additional factors such as the metabolism of 5-ALA to protoporphyrin IX (PpIX) in the heme biosynthesis pathway are known to play a crucial role for the presence of 5-ALA-induced fluorescence.²⁹ According to current knowledge, the principles of PpIX accumulation in HGG cells seem to apply to LGG as well. Most of them pertain to increased uptake, production, and/or decreased utilization of PpIX during heme biosynthesis.^30,31,32,33 The initial 5-ALA uptake into tumor cells is only partially understood, but altogether the differences between LGG and HGG seem to be based more on quantitative than on qualitative properties. In this sense, one of the key factors of 5-ALA metabolization in glioma cells seems to be the expression of the enzyme ferrochelatase.³⁴ This enzyme degrades fluorescing PpIX into nonfluorescing heme by the incorporation of iron (Fe).³⁴ Recently, a downregulation of ferrochelatase mRNA expression was found in glioblastomas as compared to low-grade astrocytomas.³⁴ This explains why a better fluorescence effect can be expected in HGG rather than LGG.

5.3.2 Initial Observations

In the first observations of two studies including 10 LGG patients, no visible 5-ALA tumor fluorescence could be found in any of these cases.^35,36 Similarly, Stummer et al were not able to detect visible 5-ALA-induced fluorescence in a patient suffering from a secondary HGG in the nonenhancing LGG portion.²⁸ Interestingly, however, the authors found focal fluorescence in the contrast-enhancing intratumoral area with malignant transformation.²⁸ In 2007, Ishihara et al performed an ex vivo analysis of glioma specimens of different WHO grades derived from fluorescence resections using 5-ALA.³⁷ In this analysis, no visible fluorescence was detected in any of the specimens of the two analyzed WHO grade II gliomas.³⁷ In contrast, specimens with fluorescence visible as well as absent were observed in the two analyzed WHO grade III gliomas.³⁷

5.3.3 Clinical Studies

Based on these initial observations, the first clinical study to better characterize 5-ALA tumor fluorescence in suspected LGG was performed by Widhalm et al in 2010.³⁸ Altogether, 17 patients with suspected DIGs, but without significant CE on preoperative MRI, were included.³⁸ 5-ALA-induced fluorescence was observed in a subgroup of these patients.³⁸ A circumscribed intratumoral area of fluorescence was found in eight of nine gliomas classified as WHO grade III after histopathologic analysis.³⁸ In contrast, no visible fluorescence was detected in any intratumoral region of all eight histologically confirmed WHO grade II gliomas.³⁸ A subsequent study from this group included 59 patients with suspected LGG.³⁹ They found that 23 of 26 WHO grade III gliomas demonstrated focal intratumoral fluorescence, whereas 29 of 33 WHO grade II gliomas did not reveal any visible fluorescence at all after 5-ALA administration.³⁹ In this study, high sensitivity (89%), specificity (88%), and positive (85%) and negative (91%) predictive values of visible fluorescence for high-grade histology were found.³⁹ In 2011, Ewelt et al reported similar findings in an independent patient cohort.⁴⁰ In their study, the authors found visible fluorescence in 12 of 17 WHO grade III and IV gliomas, but no visible fluorescence in 12 of 13 WHO grade II gliomas.⁴⁰ In the largest series to date (2017), Jaber et al found visible fluorescence in 59 of 76 WHO grade III gliomas.⁴¹ In contrast, visible fluorescence was absent in 69 of 82 WHO grade II gliomas.⁴¹ In summary, it can be stated that the vast majority of WHO grade II gliomas do not show visible fluorescence, whereas most WHO grade III/IV gliomas demonstrate visible (focal) 5-ALA-induced fluorescence.

5.3.4 Imaging Parameters

Amount of Contrast Enhancement on Preoperative MRI

The breakdown of the BBB and thus the presence of visible CE on MRI seems to be one important indicator of visible 5-ALA-induced fluorescence.28 In this sense, a recent study found a positive correlation of 5-ALA-induced fluorescence with the amount of CE on preoperative MRI in radiologically suspected LGG.³⁹ In this study, visible fluorescence was observed especially in gliomas with patchy/faint CE (53%) and focal CE (88%), whereas such visible fluorescence was frequently absent in gliomas with no CE (87%).³⁹ Similarly, in a further study, visible 5-ALA-induced fluorescence was found in 78% of gliomas with CE on MRI, but only in 16% of gliomas with no CE.⁴¹

Metabolic Activity According to PET

The metabolic activity assessed by PET seems to serve as a further powerful indicator of visible 5-ALA-induced fluorescence in suspected LGG. A significantly higher PET uptake of either MET or FET was found in gliomas with visible (focal) fluorescence as compared to nonfluorescing gliomas.^38,39,40 Recently, an FET-PET uptake ratio of more than 1.9 was identified as a strong predictor for visible fluorescence in gliomas.⁴¹ Interestingly, regions of focal fluorescence in suspected LGG topographically correlated with the intratumoral area of maximum MET-PET tracer uptake as well.^38,39

5.3.5 Histopathology and Molecular Markers

Histopathology

In radiologically suspected LGG, a correlation between visible 5-ALA-induced fluorescence and specific histopathological findings has been found. The proliferation rate assessed by the MIB-1 labeling index (LI) was found to be significantly higher in fluorescing gliomas than in nonfluorescing gliomas.^38,39,41 Moreover, the proliferation rate was found to be significantly higher in tumor regions with visible fluorescence compared to nonfluorescing areas within the same glioma³⁸ ( Fig. 5.1). Additionally, visible 5-ALA-induced fluorescence correlated with a MIB-1 LI over 10% in suspected LGG.³⁸ However, in a recent publication, it was shown that visible 5-ALA-induced fluorescence correlates not only with the proliferation rate in gliomas with nonsignificant CE, but also with specific histopathological criteria of anaplasia defined by the WHO.^1,39 According to the data of this study, cell density, nuclear pleomorphism, and mitotic rate were found to be significantly higher in intratumoral regions with focal fluorescence as compared to nonfluorescing areas.³⁹ These clinical studies highlight the ability of visible 5-ALA-induced fluorescence to intraoperatively identify focal intratumoral areas of malignant transformation without being affected by brain shift (Video 5.1).