P11.20.A Correlation between thyroid function and regorafenib activity in recurrent IDH wild-type(IDHwt) glioblastoma(GBM) patients: a large monocentric study

Abstract

Abstract Background Glioblastoma is the most frequent CNS malignant tumor, with high aggressiveness and poor prognosis. Regorafenib has recently demonstrated promising activity in terms of survival in recurrent glioblastoma patients. The impact of thyroid function on regorafenib activity has alreasy described in in patients with other types of tumors. This study aimed to investigate the relationship between baseline thyroid variables (TSH, fT3, fT4, fT3/fT4 ratio) and survival in IDHwt GBM patients who were treated with regorafenib. Material and Methods We retrospectively evaluated all consecutive recurrent IDHwt GBM patients who were treated with regorafenib at the Veneto Institute of Oncology in Padua (Italy) for which baseline thyroid function assessment was available prior to starting regorafenib. Major inclusion criteria were: histological diagnosis of IDHwt glioblastoma, regorafenib as second-line treatment, basel thyroid values available.The relationships between baseline thyroid variables (TSH, fT3, fT4, fT3/fT4 ratio) and outcomes (PFS, OS) were investigated with Cox regression models, where thyroid variables were modeled with first order polynomial or restricted cubic splines. Results We enrolled 108 recurrent IDHwt glioblastoma patients treated with regorafenib at our center From November 2015 to January 2022: 70% were male and median age was 50 years old. All patients received post-surgical treatment with concomitant chemotherapy and subsequent temozolomide according to Stupp protocol as first-line therapy. 32 patients (30%) underwent a second surgery upon recurrence. From starting regorafenib the median follow-up was 7.3 months (IQR 4.0-12.7). MedianPFS was 2.2 months (95% CI 2.0 to 3.4), and PFS rate was 43-13-4% at 3-6-12ms, respectively. MedianOS was 10.4 months (95% CI 7.5 to 14.5), and OS rate was 92-70-46% at 3-6-12 months, respectively. Disease Control Rate (DCR) was 42.7%. Univariate analysis suggested that the relationship between PFS and baseline fT4 may be modelled with first order polynomial (linear term p=0.06, non-linear term p=0.61) and also suggested a non-linear relationship between PFS and baseline fT3/fT4 (linear term p=0.06, non-linear term p=0.04).When adjusting for major clinical confounding factors (age, ECOG PS, second surgery, MGMT), multivariate analysis identified that baseline fT4 (as a continuous variable) is an independent risk factor for PFS (HR 1.09, 95% CI 1.02 to 1.17; p=0.02). We did not find any statistically significant associations between all baseline thyroid variables with OS and response. Conclusion Our study demonstrated fT4 value to be a predictive biomarker of PFS in recurrent glioblastoma patients treated with regorafenib. No correlation was showed between baseline thyroid function and survival.