A phase III trial of personalized chemotherapy based on serum tumor marker decline in poor-prognosis germ-cell tumors: Results of GETUG 13.

Abstract

LBA4500 Background: Poor-prognosis GCT (IGCCCG, J Clin Oncol 1997) remains a challenge with no improvement in the 50% survival demonstrated in phase III trials for 25 years. Day 21 serum tumor marker decline rate identified a subgroup of patients (pts) with a better outcome (J Clin Oncol 2004, 22: 3868-76). The hypothesis we tested in this study is that treatment allocation based on early tumor marker decline will improve the progression-free survival (PFS). Methods: Pts with IGCCCG poor-prognosis GCT were treated with a first cycle of BEP. AFP and hCG were assessed at day 18–21: 1) Pts with a favorable decline continued BEP for a total of 4 courses (Fav-BEP); 2) Pts with an unfavorable decline were randomized to receive either BEP (Unfav-BEP) or a dose-dense regimen (Unfav-dose-dense), consisting of paclitaxel-BEP plus day-10 oxaliplatin x 2 cycles, followed by 2 cycles of cisplatin, ifosfamide, and continuous infusion bleomycin (depending on lung function) + G-CSF. The primary endpoint was PFS (hypothesis: 20% difference, type 1 error: 5%, power 80%, 196 randomized pts needed). Results: 263 pts were enrolled and 254 were evaluable at day 21 (6 early deaths, 3 withdrawals): 51 pts (20%) had favorable tumor marker decline and 203 had unfavorable decline (randomized: 105 Unfav-dose-dense arm, 98 Unfav-BEP). The prognostic value of early tumor marker decline (Fav-BEP vs Unfav-BEP) was confirmed: 70% vs 48% for 3-year PFS (p=0.01), and 84% vs 65% for overall survival (OS) (p=0.02). The 3-year PFS was 59% in the Unfav-dose-dense arm vs 48% in the Unfav-BEP arm (p=0.05; HR: 0.66 [0.44-1.00]). 3-year OS was 73% and 65%, respectively. More ≥ grade 2 neurotoxicity (21% vs 4%) and more hematotoxicity occurred in the dose-dense arm, with no excess febrile neutropenia (17% each arm) or toxic deaths (1 each arm). Salvage high-dose chemotherapy + stem-cell transplant were required in 6% in the Unfav-dose-dense arm and 16% in the Unfav-BEP arm (p=0.01). Conclusions: An algorithm of individualized treatment intensification determined by the rate of early tumor marker decline reduces the risk of progression or death in men with poor-prognosis GCT. Clinical trial information: NCT00104676.