Portal de Periódicos da CAPES

NI-56 * FET-PET ADDS PREDICTIVE VALUE TO LOCALIZATION OF TREATMENT FAILURE IN HIGH GRADE GLIOMAS

2014; Oxford University Press; Volume: 16; Issue: suppl 5 Linguagem: Inglês

10.1093/neuonc/nou264.54

1523-5866

Michael Lundemann, Junia C. Costa, Ian Law, Silke Engelholm,

Medical Imaging Techniques and Applications

PURPOSE: To evaluate if post-surgical magnetic resonance imaging (MRI) in combination with positron emission tomography (PET) is a better predictor for localization of treatment failure than MRI alone in patients with high grade gliomas. METHODS: Twenty-one patients with primary glioblastoma WHO grade IV were included in this retrospective study. Post-surgical, contrast-enhanced MRI was used to define the gross tumor volume (GTV-MR). In addition, 18F-fluoro-ethyltyrosine (FET) PET was performed, and the FET-positive volume was semi-automatically determined by including tissue with an uptake exceeding 1.6 times the uptake in normal appearing gray matter, and subsequently manually edited to exclude non-tumor. We defined GTV-MR in combination with FET as GTV-MRPET. Clinical target volume (CTV) was defined by adding a two centimeter margin to GTV-MR and modified to include the FET-volume. Patients were treated with surgery and subsequent chemo-radiation (RCX) to 60 Gy in 30 fractions to the CTV plus concurrent and adjuvant temozolomide (TMZ). Treatment failure volume (TFV) was defined as the earliest radiological evidence of residual or recurrent disease on post-therapeutic MRI. The maximum distances from GTV-MR and GTV-MRPET to encompass TFV, and the sensitivity as well as specificity between TV and TFV, were calculated and pairwise compared (Wilcoxon). RESULTS: The maximum distance to encompass TFV was significantly higher (p < 0.001) for GTV-MR (12.7 mm [3.20-73.6]) compared to GTV-MRPET (11.4 mm [3.15-73.6]). GTV-MRPET was significantly more sensitive in predicting the location of TFV (p < 0.001), however the specificity was significantly lower (p < 0.001). CONCLUSION: GTV-MRPET required smaller margin to encompass the TFV, was more sensitive but less specific than GTV-MR.