OS6.5 The prognostic usability of Ga68-DOTA-TATE PET-CT in irradiated meningioma

Abstract

Introduction: Most meningiomas feature overexpression of somatostatin receptors. The utilization of their specific radionuclide conjugate, the 68Galium-labeled tetraazacyclododecanetetraacetic acid-octreotate (Ga68-DOTA-TATE) enables for accurate and contrast-rich visualization of these tumors in PET-CT scans. This greatly aids the treatment modalities in meningiomas: surgery and irradiation - primarily by helping determine the tumor extent. On the other hand, a quantitative analysis of various PET radiotracer uptake parameters can be used to investigate the tumor's biology and prognosis. This has been widely studied for 18FDG, but not for Ga68-DOTA-TATE. In the present study we analyze an array of PET-CT uptake parameters of irradiated meningiomas in an attempt to determine their usability to predict the risk of treatment failure. Material And Methods: We analyzed 54 meningiomas of 51 patients irradiated in our institution between 2013 and 2015. All patients had a Ga68-DOTA-TATE PET-CT scan prior to the treatment. The following quantifiers were calculated for each tumor: maximum standardized uptake value (SUVmax); three different SUV thresholds were used to measure mean SUV, standard deviation (SD) and metabolic tumor volume (MTV): gluteus muscle-based(m), liver-based(l) and gradient-based(g). Composite parameters were also calculated for each threshold individually: the total lesion activity (TLA=SUVmean, MTV) and coefficient of variation (COV=SD/SUVmean). All quantifiers as well as a number of clinical variables were investigated in uni- and multivariate models for their ability to predict the risk of tumor progression at a minimum of 12 months of followup as well as their influence on progression-free survival (PFS). Results: The multivariate Cox model identified SUVmax [hazard ratio(HR)=1,08, 95% confidence interval(CI):1,03-1,15; p=0,004] and MTVg (HR=1,03, 95%CI: 1,007-1,057; p=0,01) as prognostic for the risk of tumor progression. Using receiver operating characteristics model we identified the threshold values that stratified the patients into groups of high and low risk of progression. These were tested in a Kaplan-Meier univariate model which found MTVg, TLGg and MTVm significant (log-rank test results of p=0,00006, p=0,011 and p=0,0002, respectively). The strongest of these predictors, MTVg identified patient subgroups of progression risk 66,6% vs 8,9%. Conclusions: SUVmax and MTV of gradient- and muscle-based threshold are significant and objective predictors of meningioma progression following irradiation. An early diagnosis of treatment failure can allow for salvage treatment options such as surgery of re-irradiation. Therefore, the quantifiers that our study identified as unfavorable prognostic factors can be used along with the clinical variables to select a group of patients who may benefit from a more intensive surveillance.