5-Aminolevulinic Acid and High-Grade Gliomas

3 5-Aminolevulinic Acid and High-Grade Gliomas

Ricardo Díez Valle and Walter Stummer

Abstract
5-aminolevulinic acid (5-ALA) is a compound used for fluorescence-guided tumor resection in neurosurgery. It was the first drug to be developed specifically for this purpose and is presently the only agent approved by the European Medicines Agency and the Food and Drug Administration for intraoperative tissue imaging. This chapter reviews the experimental and clinical background of this drug for its use in malignant glioma surgery.

Keywords: malignant glioma, fluorescence, ALA, porphyrins, fluorescence-guided resection, glioblastoma

3.1 Background

3.1.1 High-Grade Gliomas

High-grade gliomas (HGGs) are diffuse tumors that extensively infiltrate the brain and seldom have a histologically distinct border. In most cases, these tumors consist of a core of solidly proliferating tumor cells surrounded by brain containing invading cells, but little information is available about the amount, characteristics of invasive cells, and the distance of invasion. Serial biopsy studies performed in the past suggested that enhancing areas on CT and MRI corresponded mostly to solid tumor, while hypointense or surrounding T2 abnormal areas contained variable amounts of invasive cells.1 After these early studies, little relevant information has been acquired on this topic, and, importantly, there is no standard method to quantify and report the degree of invasion in specific areas of the brain.

3.1.2 Role of Resection in High-Grade Gliomas

Given the diffuse nature of malignant glial tumors, surgery is never curative, and the benefit of surgery has been a matter of discussion. There is no level 1 evidence from appropriate randomized trials on this subject. One randomized trial found resection superior to biopsy.² However, the conclusions from this study were limited by the inclusion of only elderly patients, small patient numbers, a combination of WHO grade III and IV cases, and a lack of standardized adjuvant therapies. Other available studies are retrospective in nature. Nevertheless, a plethora of data suggest that resection of HGG has a favorable impact on overall survival (OS).³ When the degree of resection is measured by postoperative MRI, a strong positive correlation can be observed between extent of resection (EOR) and longer OS. This correlation was initially shown with the resection of glioblastoma (GBM) and postoperative MRI,⁴ and confirmed in successive, larger series,^5,6 so that the contrast-enhancing tumor has become the target of resection.

For GBM, the study with the best evidence⁷ is a cohort study that provides level 2b evidence that survival depends on complete resection of enhancing tumor (CRET) in GBM. A larger retrospective series, with detailed volumetric analysis over 721 patients, found a continuous, nonlinear relationship between OS and extent of resection, supporting a maximal safe resection approach to GBM.⁸ For other types of HGG, series are smaller, and there are less high-quality data, mostly supporting the same philosophy. Given the available data, it is highly improbable that a randomized trial will be performed for resection versus nonresection of a GBM, since available data prevent equipoise.

3.1.3 Results of Resection

Due to the diffuse, infiltrative nature of HGGs, neurosurgeons have understood that resection should be restricted to tissue areas that are visualized by MRI contrast enhancement. Thus, CRET has been the surgical paradigm in most neurosurgical centers. However, multiple reports indicate that this aim is seldom actually achieved. Rates of around 30% or less have been reported in leading institutions of the world, as reviewed by Sanai and Berger.⁹ There are several reasons why the resection of contrast-enhancing tumor is not always complete. As surgery is not curative, it is important to perform a safe resection, strictly avoiding neurological deficits. Intentional sparing of assumed eloquent regions invaded by tumor is one reason for incomplete resections. However, safety does not seem to be the main reason. At least two studies have compared the visual impression of the surgeon at the end of surgery to the actual resection as verified by postcontrast MRI. The first study from 1994 found that surgeons expected complete resections in 54% of the cases based on their visual impression of the resection cavity. MRI disproved them, showing that the intraoperative impression was true for only 18% (one-third of cases).⁴ A similar study published in 2012, more than 20 years later, found that resection was complete in only 30% of cases in which the surgeon had expected resections to be complete.¹⁰ The authors of this particular study also identified 17 cases in which complete resections were considered safe by expert panel review. Only 24.5% of these patients were subject to a complete resection. Such results show very clearly that the contrast-enhancing limit of the tumor is not reliably identified by the surgeon during standard surgery based on the visual impression and haptic information only.

3.2 Preclinical Data of 5-Aminolevulinic Acid Fluorescence

3.2.1 Physiopathology

5-aminolevulinic acid (5-ALA) is the first committed precursor of heme synthesis, situated in the metabolic pathway just beyond the controlling step. External, excess supply of the compound to cells increases porphyrin synthesis, ending in protoheme. 5-ALA is not a fluorophore, but some of the dependent metabolites in the heme synthesis are. Prior work found that administration of 5-ALA could produce a chemical model of porphyria, as some of its metabolites are phototoxic. This observation later spawned the idea of using 5-ALA for photodynamic therapy (PDT). This therapeutic concept exploits the accumulation of photosensitizing molecules at a high enough concentration to eliminate tumor cells with light exposure and creation of toxic reactive oxygen species (ROS). A number of in vitro and animal experiments completed in the past showed that accumulation of fluorescent porphyrin compounds, especially protoporphyrin IX (PpIX), was indeed much higher in certain tumors (skin, bladder, and gastrointestinal tumors) than in normal tissue. PDT has a number of challenges, as it requires the delivery of light to cells in sufficient amounts, subsequent formation of enough toxic metabolites (ROS) to kill cells, and a wide distribution in the tumor to be effective.

Visualization of fluorescence as guidance for resection emerged as a more straightforward application of 5-ALA and its PpIX metabolite. PpIX is readily excited by light in the blue/violet range of the spectrum around 400 nm, and it is fluorescent with two emission peaks, at 635 and at 704 nm, which is observed by the human eye as red light. The differential concentration between malignant tumor cells and normal cells is enough for differential fluorescence visualization of tumors.

As 5-ALA is readily absorbed within the intestine and rapidly passes into the bloodstream (within 1 hour), it is administered orally and needs to be given prior to general anesthesia. Once 5-ALA reaches the plasma, viable glioma cells start to accumulate PpIX minutes later, in an active metabolic process through the heme biosynthesis pathway. Most of the synthesized PpIX remains intracellular. 5-ALA is a small, polar molecule that does not readily cross the intact blood–brain barrier (BBB).11 However, due to its size (131.131 g/mol), only minor perturbations of the BBB function are needed for 5-ALA to reach tumor cells, even in areas that do not accumulate MR contrast (gadolinium). The tumor cell targeting of 5-ALA, the number of viable tumor cells, and their intrinsic metabolic activity will impact the generation of PpIX and fluorescence emission. Necrotic tumor regions do not produce visible 5-ALA-induced fluorescence, as PpIX accumulation requires intact tumor cell metabolism. 5-ALA fluorescence will not be present in vascular compartment, or tissues that do not actively synthetize porphyrins in great quantity. However, the administration of oral 5-ALA does produce an increase in porphyrin synthesis in organs such as the skin and liver. 5-ALA administration can lead to transient skin phototoxicity or transiently (but not critically) elevated liver enzymes.

3.2.2 Experimental Data In Vitro and In Vivo

The first experimental results utilizing 5-ALA in glioma cells were published in 1998.¹² 5-ALA was shown to induce fluorescence in incubated C6 rat glioma cells, and in orthotopic syngeneic tumors derived from the intracranial implantation of C6 cells in rats. After intraperitoneal administration of 5-ALA, at a dose of 100 mg/kg, tumors were visualized as intensely fluorescent, while faint fluorescence could be seen in some areas of normal tissue, pia, choroid plexus, and surrounding white matter tracts. In the same year, another group published similar results utilizing C6 and also 9 L glioma models with a higher dose of 5-ALA (200 mg/kg). Tumor fluorescence was visualized between 2 and 8 hours after 5-ALA administration and disappeared by 22 hours. After using this dose, the group noted fluorescence in the ependyma, and also slight fluorescence in the normal brain.¹³

In subsequent experiments with multiple different cell lines, differences in 5-ALA-induced fluorescence intensities were found. However, tumor fluorescence was still sufficient for differentiating tumor cells from normal neuronal cell lines and cultured astrocytes at an incubation time of 120 minutes.¹⁴ No malignant glioma cell line has been reported to be negative for 5-ALA-induced fluorescence. PpIX fluorescence is also present in GBM cancer stem cells in vitro in concentrations that have been demonstrated to be sufficient for cell killing by PDT.¹⁵

The amount of PpIX accumulation in tumor cells and visible fluorescence can be modified, at least in vitro, by a number of substances that affect porphyrin metabolism. Iron chelation with deferoxamine increases fluorescence in the glioma xenograft U251 model,¹⁶ and also in cancer stem cells.¹⁷ On the other hand, antiepileptic drugs, such as phenytoin and valproic acid, can diminish accumulation of PpIX. The most commonly used antiepileptic in brain tumor patients, levetiracetam, shows no effect.¹⁸ In the same experimental series, the addition of dexamethasone to the other drugs appeared to further diminish the amount of PpIX produced, while increasing the retention inside the cells. Dexamethasone and phenytoin together have been shown to lower intracellular accumulation of PpIX.¹⁸

3.3 Clinical Application

3.3.1 Early Clinical Experience

The first clinical use of fluorescence-guided surgery (FGS) with 5-ALA for gliomas was reported in 1998.¹⁹ In a cohort of nine HGG patients, utilizing an oral dose of 10 mg/kg, the authors used a violet–blue (375- to 440-nm) xenon excitation light and a 455-nm long-pass filter adapted to an operating microscope. A specifically developed 455-nm long-pass filter allowed red light emitted by the PpIX to reach the observer along with a small fraction of the excitation light, the latter necessary for allowing background illumination with red porphyrin fluorescence observable on a blue rather than black background ( Fig. 3.1). This allowed blue field surgery to be possible even under the fluorescence mode of the microscope.

In this first trial, 89 tissue biopsies of specimens with or without 5-ALA-induced fluorescence were analyzed. A calculated sensitivity of 85% and a specificity of 100% for the detection of malignant tissue were found in this study. Chapter 2 in this book deals with the problem of evaluating intraoperative diagnostic methods, and how the use of traditional descriptors of diagnostic accuracy, such as sensitivity and specificity in this setting, is difficult to determine amidst a number of biases and confounders. Nevertheless, the first results confirmed the potential utility of 5-ALA for detection of tumor in HGGs, as all the fluorescent samples contained tumor, and most of the nonfluorescent samples did not. For seven of nine patients, visible 5-ALA-induced fluorescence led to additional resection of tumor. The authors also closely assessed the effect of photobleaching (porphyrin degeneration due to standard operating white light illumination) as a potential problem. Under operating light conditions, fluorescence decayed to 36% in 25 minutes for violet–blue light using intensities comparable to those obtained during surgery, and in 87 minutes for white light. This study also contained the first description of a seemingly nonenhancing glioma with an anaplastic focus, which was detected by fluorescence after 5-ALA administration.

Fig. 3.1 Border of resection of a glioblastoma showing an area with normal aspect under white light (a) where fluorescence with 5-ALA shows residual tumor (b).

Fig. 3.2 Image showing the different quality of intense, red fluorescence versus pale, pink fluorescence.

Two years later, the same group published the results of surgery utilizing 5-ALA fluorescence guidance in 52 HGG patients.20 In this phase II trial, 5-ALA was administered at an oral dose of 20 mg/kg of body weight, which has now become the standard dose administered to patients. The same combination of excitation light and long-pass filter coupled to the surgical microscope was used. Fluorescence was useful for guiding the resection in all cases, and two discernible fluorescence qualities were observed: a strong solid red fluorescence in viable solid tumor, surrounded by a transition zone with less intense, pink fluorescence in the region of the infiltrating tumor margin ( Fig. 3.2). Necrotic tumor tissue was not fluorescent but could be differentiated under normal light. Selective biopsies with different fluorescence qualities were taken and analyzed for the presence of tumor. Solid fluorescence was usually characterized by solid, coalescent tumor, with PpIX fluorescence in the cytoplasm of the cells, while vague or weaker pink fluorescence usually featured infiltrating tumor of medium to high density.

This trial compared the results of the FGS to the postoperative contrast MRI, the standard for resection control in GBM, and the effect of residual fluorescence on OS. CRET was achieved in 33 patients (63%). Resection was mostly limited due to functional considerations. The resection cavity had collapsed and fluorescence was not visualized in only one patient with a cystic lesion. Complete resection of the contrast-enhancing tumor was achieved in 16 of 17 patients. Complete resection of fluorescent tissue was achieved in 9 of 12 cases with resection of all areas with intense fluorescence. Only in 8 of 23 patients was intense tissue fluorescence left behind. There was a significant correlation between any kind of residual fluorescence and residual enhancing tumor on MRI, and an even stronger correlation between residual solid fluorescence and residual enhancing tumor on the MRI. The Kaplan–Meier analysis revealed survival was longer in patients with CRET than in those without CRET (no enhancement, 103 ± 11 weeks; residual enhancement, 54 ± 5 weeks), and similarly, resection of fluorescent tissue was superior to leaving fluorescing tissue behind (no fluorescence, 101 ± 15 weeks; vague fluorescence, 79 ± 6 weeks; solid fluorescence, 51 ± 3 weeks). The difference was significant only for residual solid, intense fluorescence. In summary, this work confirmed the association between removal of 5-ALA-induced tumor fluorescence and more extensive tumor resections with improved outcomes in OS. Increasing the rate of CRET was better achieved with 5-ALA FGS in comparison to historical studies of HGG tumor resection at that time.