Increased Brain Tumor Resection Using Fluorescence Image Guidance in a Preclinical Model

Division of Pathology, St. Michael’s Hospital, Toronto, Ontario, Canada M5B 1W8Background and Objectives: Fluorescence image-guided brain tumor resection is thought to assist neuro-surgeons by visualizing those tumor margins that mergeimperceptibly into normal brain tissue and, hence, aredifficult to identify. We compared resection completenessand residual tumor, determined by histopathology, afterwhite light resection (WLR) using an operating microscopeversus additional fluorescence guided resection (FGR).StudyDesign/MaterialsandMethods:Weemployedanintracranial VX2 tumor in a preclinical rabbit model and afluorescence imaging/spectroscopy system, exciting anddetecting the fluorescence of protoporphyrin IX (PpIX)induced endogenously by administering 5-aminolevulinicacid (ALA) at 4 hours before surgery.Results: Using FGR in addition to WLR significantly in-creased resection completeness by a factor 1.4 from 68 38to 98 3.5%, and decreased the amount of residual tumorpost-resection by a factor 16 from 32 38 to 2.0 3.5% ofthe initial tumor volume.Conclusions: Additional FGR increased completeness ofresection and enabled more consistent resections betweencases. Lasers Surg. Med. 35:181–190, 2004. 2004 Wiley-Liss, Inc.Key words: malignant glioma; VX2; 5-aminolevulinicacid; protoporphyrin IX; fluorescence imaging and spectro-scopyINTRODUCTIONThe treatment of patients with high-grade gliomasremainsamajorchallenge.Theprognosisforthesepatientsis poor, with a median survival time after diagnosis andtreatment of less than 1 year [1,2]. It has been suggestedthat the prognosis is linked to the completeness of tumorremoval [3–6]. A recent [7] study of 416 patients withglioblastoma multiforme indicated that resection of 89% ofthe tumor volume is necessary to improve survival, whileresection of 98% or more resulted in a significant survivaladvantage of 4.2 months compared with a resection of lessthan 98%. However, such a high degree of tumor resectionis often limited by the surgeon’s ability to distinguishresidual tumortissue from surrounding braintissue underconventional white light microscope illumination [8].Hence,methodsenablingbetterintraoperativediscrimina-tion of viable tumor borders should be valuable.Fluorescence imaging and spectroscopy using 5-amino-levulinic acid (ALA)-induced protoporphyrin IX (PpIX) is apotential technique to enhance contrast of viable tumorborders.Althoughadministrationoffluorescentmarkerstoenhance contrast of malignant gliomas is not new [9–13],marking tumors with ALA is conceptually different fromearlier investigations, since ALA is not itself fluorescentbut is metabolized into fluorescent PpIX by a number ofmalignant tumors in situ through enzymes of the heme-biosynthesis pathway [14]. ALA-PpIX has been usedwidely,bothforfluorescencediagnosticsandphotodynamictherapy [15]. It may avoid problems that arise when afluorescent marker is administered directly, such as leak-age from the tumor into surrounding normal brain tissue[16]. We have shown previously [17–20] that ALA-inducedPpIXlevelsinnormalbraintissue,especiallywhitematter,are very low. Clinical and preclinical studies suggest thatthe resulting selectivity of ALA-PpIX in certain braintumors [14,17–25] is a result of various factors. The lowpermeability of ALA at the blood–brain barrier (BBB)[17,26] reduces uptake in normal brain, whereas thecompromised BBB in certain brain tumors is thought topermit selective ALA transport. Different activities ofenzymes in the heme pathway have also been observedbetween tumor and normal tissues, which subsequentlyenable selective production of PpIX [27,28].