Dexamethasone administration during definitive radiation and temozolomide renders a poor prognosis in a retrospective analysis of newly diagnosed glioblastoma patients

Lisa B. E. Shields, Brent J. Shelton, Andrew J. Shearer, Li Chen, David A. Sun, Sarah Parsons, T. David Bourne, Renato LaRocca, Aaron C. Spalding

Dexamethasone (DXM) is commonly used in the management of cerebral edema in patients diagnosed with glioblastoma multiforme (GBM). Bevacizumab (BEV) is FDA-approved for the progression or recurrence of GBM but has not been shown to improve survival when given for newly diagnosed patients concurrently with radiation (RT) and temozolomide (TMZ). Both DXM and BEV reduce cerebral edema, however, DXM has been shown to induce cytokine cascades which could interfere with cytotoxic therapy. We investigated whether DXM would reduce survival of GBM patients in the setting of concurrent TMZ and BEV administration. We reviewed the treatment of all 73 patients with GBM who received definitive therapy at our institution from 2005 to 2013 with RT (60 Gy) delivered with concurrent daily TMZ (75 mg/m(2)). Of these, 34 patients also were treated with concurrent BEV (10 mg/kg every two weeks). Patients received adjuvant therapy (TMZ or TMZ/Bev) until either progression, discontinuation due to toxicity, or 12 months after radiation completion. All patients who had GBM progression with TMZ were offered BEV for salvage therapy, with 19 (56 %) receiving BEV. With a median follow-up of 15.6 months, 67 (91.8 %) patients were deceased. The OS for the entire cohort was 15.9 months, while the PFS was 7.7 months. The extent of resection was a prognostic indicator for OS (p = .0044). The median survival following gross tumor resection (GTR) was 22.5 months, subtotal resection (STR) was 14.9 months, and biopsy was 12.1 months. The addition of BEV to TMZ with RT was borderline significantly associated with increased PFS (9.4 vs. 5.1 months, p = 0.0574) although was not significantly associated with OS (18.1 vs. 15.3 months respectively, p = 0.3064). In patients receiving TMZ, DXM use concurrent with RT was a poor prognostic indicator of both OS (12.7 vs. 22.6 months, p = 0.003) and PFS (3.6 vs. 8.4 months, p <0.0001). DXM did not reduce OS in patients who received TMZ and BEV concurrently with RT (22.9 vs 22.8 months, p = 0.4818). On multivariable analysis, DXM use predicted an unfavorable OS hazard ratio (HR) = 1.72, p = 0.045). Our results with TMZ, BEV, and RT are similar to previous studies in terms of PFS and OS. DXM use during RT with concurrent TMZ correlated with reduced OS and PFS unless BEV was administered.