Rationale: Subventricular zone (SVZ) involvement has been proposed as an adverse prognostic factor in glioblastomas (GBM). The true extent of ventricular involvement at surgery is often difficult to establish and is poorly studied. Tumour fluorescence provides us with an exciting opportunity to interrogate tumour extent intraoperatively.
Methods: We conducted a retrospective analysis of all cases of GBMs operated using aminolevulinic acid-induced fluorescence and analyzed radiological SVZ involvement alongwith the incidence of ventricular entry at surgery, ependymal fluorescence and histological correlation of the ependymal involvement.
Results: Of 30 GBMs, radiological SVZ involvement was seen in 26 of which ventricles were opened at surgery in 19. Diffuse ependymal fluorescence was seen in 10 of the 19 cases (51%) and histology revealed tumour infiltration in only one of the five cases where ependymal tissue was sampled. Focal ependymal fluorescence seen in two of the 19 cases was always pathological. Diffuse ependymal fluorescence did not always correlate with gross appearance of the ventricular lining at surgery. Nor did it correlate with SVZ involvement.
Conclusions: Pathological significance of diffusely fluorescing ependymal lining seen during surgery is questionable and need not represent tumour extension. Ependymal fluorescence may sometimes not be visualized even when the tumour appears to involve the SVZ. These results highlight the potential limitations of fluorescence especially in the bordering infiltrating zone where its predictive value is diminished.
Keywords: Fluorescence-guided resection; aminolevulinic acid; ependymal involvement; glioblastoma; subventricular zone; surgical resection.