Randomized, Prospective Double-Blinded Study Comparing 3 Different Doses of 5-Aminolevulinic Acid for Fluorescence-Guided Resections of Malignant Gliomas

Abstract

Background: Five-aminolevulinic acid (5-ALA) is used for fluorescence-guided resections of malignant glioma at a dose of 20 mg/kg; yet, it is unknown whether lower doses may also provide efficacy.

Objective: To perform a double-blinded randomized study comparing 3 different doses of 5-ALA.

Methods: Twenty-one patients with suspected malignant glioma were randomly assigned to 0.2, 2, or 20 mg/kg 5-ALA. Investigators were unaware of dose. Intraoperatively, regions of interest were first defined in tumor core, margin, and adjacent white matter under white light. Under violet-blue illumination, the surgeon's impression of fluorescence was recorded per region, followed by spectrometry and biopsy. Plasma was collected after administration and analyzed for 5-ALA and protoporphyrin IX (PPIX) content.

Results: The positive predictive value of fluorescence was 100%. Visual and spectrometric fluorescence assessment showed 20 mg/kg to elicit the strongest fluorescence in tumor core and margins, which correlated with cell density. Spectrometric and visual fluorescence correlated significantly. A 10-fold increase in 5-ALA dose (2-20 mg/kg) resulted in a 4-fold increase of fluorescence contrast between marginal tumor and adjacent brain. t max for 5-ALA was 0.94 h for 20 mg/kg (0.2 kg: 0.50 h, 2 mg/kg: 0.61 h). Integrated PPIX plasma levels were 255.8 and 779.9 mcg*h/l (2 vs 20 mg/kg). Peak plasma concentrations were observed at 1.89 ± 0.71 and 7.83 ± 0.68 h (2 vs 20 mg/kg; average ± Standard Error of Mean [SEM]).

Conclusion: The highest visible and measurable fluorescence was yielded by 20 mg/kg. No fluorescence was elicited at 0.2 mg/kg. Increasing 5-ALA doses did not result in proportional increases in tissue fluorescence or PPIX accumulation in plasma, indicating that doses higher than 20 mg/kg will not elicit useful increases in fluorescence.

Keywords: 5-ALA; Aminolevulinic acid; Fluorescence-guided resections; Histology; Malignant glioma; Randomized study; Spectrography.

FIGURE 1.

Procedure for sampling and spectrometry. In a first step, after removal of necrotic tumor, regions of interest were identified under white light consisting of non-necrotic solid tumor (core), marginal tumor, and normal adjacent tissue under normal white light illumination. Second, the macroscopic fluorescence quality was assessed by the surgeon in each region (“strong,” “weak,” “none”), immediately followed by spectrometry in the respective region and biopsies. In “normal” adjacent tissue, biopsies were only collected if this tissue was intended for resection.

FIGURE 2.

Surgeons’ subjective perception of global fluorescence quality and extent in identifiable tumor (under white light), stratified by dose. “Extent” is defined as the area of fluorescing tissue relative to the area of abnormal tissue as identified under white light illumination.

FIGURE 3.

Spectrometrical assessment of fluorescence as measured in the region identified as tumor “core” and “margin” under white light, stratified by 5-ALA dose. Values for each fluorescence quality in single patient averaged.

FIGURE 4.

Mean fluorescence intensities as measured spectrometrically vs fluorescence quality (strong, weak, none) in tumor core. Box plots signify minimum, first quartile, median, third quartile, and maximum. Jonckheere–Terpstra test: P < .0001; Wilcoxon–Mann–Whitney test: none vs weak fluorescence: P = .0001; weak vs strong: P = .002; none vs strong: P < .0001.

FIGURE 5.

Mean tumor cell density of biopsies vs mean fluorescence intensities as measured spectrometrically, stratified by 5-ALA dose (all localizations including normal brain, all fluorescence qualities, values for single fluorescence qualities in single patient averaged). A, 0.2 mg/kg b.w. B 2 mg/kg b.w. C 20 mg/kg b.w. We only found a relationship between tumor cell density and fluorescence for the 20 mg/kg group. The mean fluorescence intensities per treatment group were 0.767 ± 0.0564 a.u. (0.2 mg/kg); 0.245 ± 0.340 a.u. (2 mg/kg); and 1.71 ± 1.65 a.u. (20 mg/kg; P < .0001 by analysis of variance).

FIGURE 6.

Geometric mean concentrations of 5-ALA [μg/L] vs time [h] for 3 different doses. Inset: dose proportionality of area under curve (AUC) of 5-ALA p.o. in the dose range of 0.2 to 20 mg/kg b.w.

FIGURE 7.

A, Geometric mean concentrations of protoporphyrin-IX (PPIX) [μg/L] vs time [h] for 2 and 20 mg/kg averaged at each time point. No PPIX was detectable for the lowest dose of 0.2 mg/kg. B, First, peak, and final values for tissue PPIX averaged for each patients with time point of measurement (averages ± SEM) for 2 and 20 mg/kg b.w. On the average, plasma PPIX peaked at 7.8 after 20 mg/kg b.w. and at 4.7 h after 2 mg/kg b.w.