TY  - CONF
AU  - Lohmann, Philipp
AU  - Lerche, Christoph
AU  - Bauer, Elena
AU  - Steger, Jan
AU  - Stoffels, Gabriele
AU  - Blau, Tobias
AU  - Dunkl, Veronika
AU  - Filss, Christian P
AU  - Stegmayr, Carina
AU  - Neumaier, Bernd
AU  - Shah, Nadim J
AU  - Fink, Gereon
AU  - Langen, Karl-Josef
AU  - Galldiks, Norbert
TI  - NMNIMG-32. DIFFERENTIATION OF PSEUDOPROGRESSION FROM TUMOR PROGRESSION IN GLIOBLASTOMA PATIENTS BASED ON FET PET RADIOMICS
SN  - 1523-5866
M1  - FZJ-2022-02694
PY  - 2017
AB  - AbstractBACKGROUNDDifferentiation of pseudoprogression (PsP) from tumor progression (TP) in glioblastoma patients can be difficult with standard MRI. Textural feature analysis as part of the concept of radiomics offers a quantitative method to describe tumor heterogeneity. We investigated the potential of textural features of O-(2-[18F]fluoroethyl)-L-tyrosine (FET) PET to discriminate between PsP and TP in glioblastoma patients.METHODSTwenty-six newly diagnosed glioblastoma patients with MRI findings suspicious for TP within 12 weeks after completion of chemoradiation with temozolomide underwent an additional dynamic FET PET scan. Volumes-of-interest were defined on summed images (20-40 min post-injection) by a 3-dimensional auto-contouring process using a tumor-to-brain ratio (TBR) of 1.6 or more. TBRs and time-activity curves (TACs) of FET uptake were determined. Dynamic FET PET parameters time-to-peak (TTP) and slope (slope of the linear regression line 20-50 min post-injection) were evaluated. Additionally, 39 textural parameters were calculated using the software LifeX. The diagnostic accuracy of TBRs, TTP, slope, and textural parameters to discriminate between PsP and TP was evaluated using ROC analyses using the results of histopathology or of the clinico-radiological course as reference. In order to further increase the diagnostic accuracy, parameters were combined using linear logistic regression for classification of PsP and TP.RESULTSFifteen patients had TP and 11 patients had PsP. The parameters TBRmean, TBRmax and TTP yielded a diagnostic accuracy to discriminate between PsP and TP of 70%, 74%, 59%, respectively. The dynamic FET PET parameter slope yielded the highest diagnostic accuracy of 81%. The two best textural features showed an accuracy of 74%. Combining TBR with textural features lead to an improved accuracy of 78%.CONCLUSIONSTextural features might yield additional valuable information for this highly relevant clinical problem without the need for acquiring a more time-consuming dynamic PET acquisition and should be further evaluated prospectively in larger cohorts.
T2  - Joint Conference of 22nd Annual Scientific Meeting and Education Day of the Society-for-Neuro-Oncology / Conference of the Society-for-CNS-Interstitial-Delivery-of-the-Therapeutics (SCIDOT) on Therapeutic Delivery to the CNS
CY  - 16 Nov 2017 - 19 Nov 2017, San Francisco (USA)
Y2  - 16 Nov 2017 - 19 Nov 2017
M2  - San Francisco, USA
LB  - PUB:(DE-HGF)1
DO  - DOI:10.1093/neuonc/nox168.607
UR  - https://juser.fz-juelich.de/record/908570
ER  -