P16.17 * F-18-FET PET IMAGING FOR DIAGNOSIS OF LATE PSEUDOPROGRESSION IN PATIENTS WITH HIGH-GRADE GLIOMAS

Abstract

BACKGROUND: Pseudoprogression is characterized by increasing contrast-enhancement in the tumor area mimicking tumor progression, which, however, remains stable or subsides in control MRI 4-8 weeks later. Pseudoprogression is thought to be a sign of response to therapy rather than rapid tumor growth. It mainly occurs within three months after radiochemotherapy but it is increasingly observed that pseudoprogression may also develop substantially later and may be particularly difficult to distinguish from true tumor progression which is much more prevalent at this stage. As conventional MRI lacks specificity for pseudoprogression, there is still an unmet need for additional imaging modalities in this setting. PATIENTS ANDMETHODS:Twenty-two clinically stable patients diagnosed with highgrade glioma were examined by F-18-fluoroethyl-tyrosine PET (FET PET) at the time of either new contrast enhancement or ≥25% increase of an existing contrast-enhanced lesion on conventionalMRI ≥12 weeks after last radiochemotherapy. For static analysis, maximum and mean tumor/brain ratios (TBRmax, TBRmean) of FET uptake were determined. In addition, dynamic FET PET parameters were analysed: time-to peak (TTP) and time-activity curve patterns. Diagnosis of late pseudoprogression or progression was established according toRANOcriteria relying on follow-up MRI obtained at least 6-8 weeks from the indexMRIandserved as reference forPETdata evaluation. RESULTS: Based upon follow-up MRI 11 patients were diagnosed with late pseudoprogression, in 72.7% (8 out of 11) of whom MGMT methylation was tested positive. Both TBRmax and TBRmean were significantly lower compared with progressive tumor (TBRmax, 2.12 ± 0.22 vs. 3.63 ± 0.18; TBRmean, 1.47 ± 0.17 vs. 2.19 ± 0.05; both p < 0.001). Receiver-operating-characteristic (ROC) analysis yielded an optimal cut-off of 3.05 for TBRmax (sensitivity, 100%; specificity, 91.7%; AUC, 0.96; 95%CI, 0.88 - 1.0; p,0.001) and 1.95 forTBRmean(specificity, 91.7%; sensitivity, 77.8%; AUC, 0.94; 95% CI, 0.47 - 1.0; p = 0.001). ROC analysis of TTP revealed at an optimal cutoff of 19 min a sensitivity of 100% along with a specificity of90%for predicting pseudoprogression (AUC, 0.92;95%CI, 0.76 - 1.0; p = 0.01). However, there was no significant association between the type of curve pattern and the diagnosis of late pseudoprogression (x2 (1) = 3.19, p = 0.117). CONCLUSION: F-18-FET PET may provide valuable information for the assessment of late pseudoprogression. Systematic prospective analyses are warranted.