A prospective Phase II clinical trial of 5-aminolevulinic acid to assess the correlation of intraoperative fluorescence intensity and degree of histologic cellularity during resection of high-grade gliomas

OBJECTIVE There is evidence that 5-aminolevulinic acid (ALA) facilitates greater extent of resection and improves 6-month progression-free survival in patients with high-grade gliomas. But there remains a paucity of studies that have examined whether the intensity of ALA fluorescence correlates with tumor cellularity. Therefore, a Phase II clinical trial was undertaken to examine the correlation of intensity of ALA fluorescence with the degree of tumor cellularity. METHODS A single-center, prospective, single-arm, open-label Phase II clinical trial of ALA fluorescence-guided resection of high-grade gliomas (Grade III and IV) was held over a 43-month period (August 2010 to February 2014). ALA was administered at a dose of 20 mg/kg body weight. Intraoperative biopsies from resection cavities were collected. The biopsies were graded on a 4-point scale (0 to 3) based on ALA fluorescence intensity by the surgeon and independently based on tumor cellularity by a neuropathologist. The primary outcome of interest was the correlation of ALA fluorescence intensity to tumor cellularity. The secondary outcome of interest was ALA adverse events. Sensitivities, specificities, positive predictive values (PPVs), negative predictive values (NPVs), and Spearman correlation coefficients were calculated. RESULTS A total of 211 biopsies from 59 patients were included. Mean age was 53.3 years and 59.5% were male. The majority of biopsies were glioblastoma (GBM) (79.7%). Slightly more than half (52.5%) of all tumors were recurrent. ALA intensity of 3 correlated with presence of tumor 97.4% (PPV) of the time. However, absence of ALA fluorescence (intensity 0) correlated with the absence of tumor only 37.7% (NPV) of the time. For all tumor types, GBM, Grade III gliomas, and recurrent tumors, ALA intensity 3 correlated strongly with cellularity Grade 3; Spearman correlation coefficients (r) were 0.65, 0.66, 0.65, and 0.62, respectively. The specificity and PPV of ALA intensity 3 correlating with cellularity Grade 3 ranged from 95% to 100% and 86% to 100%, respectively. In biopsies without tumor (cellularity Grade 0), 35.4% still demonstrated ALA fluorescence. Of those biopsies, 90.9% contained abnormal brain tissue, characterized by reactive astrocytes, scattered atypical cells, or inflammation, and 8.1% had normal brain. In nonfluorescent (ALA intensity 0) biopsies, 62.3% had tumor cells present. The ALA-associated complication rate among the study cohort was 3.4%. CONCLUSIONS The PPV of utilizing the most robust ALA fluorescence intensity (lava-like orange) as a predictor of tumor presence is high. However, the NPV of utilizing the absence of fluorescence as an indicator of no tumor is poor. ALA intensity is a strong predictor for degree of tumor cellularity for the most fluorescent areas but less so for lower ALA intensities. Even in the absence of tumor cells, reactive changes may lead to ALA fluorescence.