RESUMO
BACKGROUND: The meaning of the ventricular wall fluorescence during 5-aminolevulinic (5-ALA)-guided surgery in patients with glioblastoma (GBM) is still unknown. The authors studied the association between ventricle fluorescence, clinical outcome and survival, and described the histopathological findings of selective biopsies from the ventricular wall. METHODS: One hundred and forty patients diagnosed of GBM underwent fluorescence-guided surgery (FGS); 65 of them were naive GBM and ventricle fluorescence during surgery was annotated prospectively. Selective biopsies were collected from the ventricular wall when possible. Clinical and radiological data were registered, including age, Karnofsky Performance Scale (KPS) score, presence of hydrocephalus, overall survival (OS), tumour volume and location (periventricular vs non-periventricular) and leptomeningeal dissemination. RESULTS: During FGS the ventricle wall was opened just when the tumour was periventricular in the preoperative MRI (45 out of 65). In 28 of them (60 %) the fluorescence extended far away from the site of opening, while in 17 it ended just in the few millimetres around the tumour. All four patients who developed hydrocephalus had periventricular tumours and the ventricle wall had been opened during surgery. Statistically significant differences were seen in OS according to periventricular location (15 m vs 33 m, P = 0.008 log rank). However, there was not significant relationship between ventricle fluorescence and hydrocephalus (P = 0.75), nor survival (14 m vs 15.5 m, P = 0.64). CONCLUSIONS: Preoperative MRI predicts if the ventricle will be opened using the 5-ALA fluorescence, according to tumour location. It does not predict, however if the ventricle wall is going to be fluorescent or not. The fluorescence of the ventricle wall is not a predictor for complications or survival. Periventricular tumour location is an independent bad prognostic factor in GBM.