Assessment of treatment response in patients with glioblastoma using O-(2-18F-fluoroethyl)-L-tyrosine PET in comparison to MRI

Abstract

The assessment of treatment response in glioblastoma is difficult with MRI because reactive blood-brain barrier alterations with contrast enhancement can mimic tumor progression. In this study, we investigated the predictive value of PET using O-(2-(18)F-fluoroethyl)-l-tyrosine ((18)F-FET PET) during treatment.

Methods: In a prospective study, 25 patients with glioblastoma were investigated by MRI and (18)F-FET PET after surgery (MRI-/FET-1), early (7-10 d) after completion of radiochemotherapy with temozolomide (RCX) (MRI-/FET-2), and 6-8 wk later (MRI-/FET-3). Maximum and mean tumor-to-brain ratios (TBR(max) and TBR(mean), respectively) were determined by region-of-interest analyses. Furthermore, gadolinium contrast-enhancement volumes on MRI (Gd-volume) and tumor volumes in (18)F-FET PET images with a tumor-to-brain ratio greater than 1.6 (T(vol 1.6)) were calculated using threshold-based volume-of-interest analyses. The patients were grouped into responders and nonresponders according to the changes of these parameters at different cutoffs, and the influence on progression-free survival and overall survival was tested using univariate and multivariate survival analyses and by receiver-operating-characteristic analyses.

Results: Early after completion of RCX, a decrease of both TBR(max) and TBR(mean) was a highly significant and independent statistical predictor for progression-free survival and overall survival. Receiver-operating-characteristic analysis showed that a decrease of the TBR(max) between FET-1 and FET-2 of more than 20% predicted favorable survival [corrected], with a sensitivity of 83% and a specificity of 67% (area under the curve, 0.75). Six to eight weeks later, the predictive value of TBR(max) and TBR(mean) was less significant, but an association between a decrease of T(vol 1.6) and PFS was noted. In contrast, Gd-volume changes had no significant predictive value for survival.

Conclusion: In contrast to Gd-volumes on MRI, changes in (18)F-FET PET may be a valuable parameter to assess treatment response in glioblastoma and to predict survival time.