Randomized phase II trial of gemcitabine/cisplatin (GC) with or without cetuximab (CET) in patients (pts) with advanced urothelial carcinoma (UC).

Abstract

4506 Background: In UCpts EGFR over-expression correlates with high grade, stage, progression, short survival. Preclinical data show enhanced anti-tumor effect with CET + chemotherapy (CT). In several solid tumors CET added to CT improved survival. CET is hypothesized to improve UC response to CT. Methods: Pts with ECOG PS 0-2, adequate organ function, unresectable, locally recurrent or metastatic UC were stratified by disease state and randomized (1:2) to C 70mg/m2 d1, G 1000mg/m2 d1, d8, d15 (Arm A) or same CT + CET 500mg/m2 d1, d15 (Arm B) q28 days. Due to high thromboembolic events rate in Arm B, G dose was reduced to 800 mg/m2. 6 GC cycles were planned. Arm A pts progressing after 2 GC cycles had CET added. Arm B pts were to continue on CET monotherapy after 4-6 CT cycles. Imaging was q8 weeks. Primary endpoint: RECIST response rate (RR); secondary: safety, PFS, survival. Since E-cadherin impacts UC response to CET in vitro, serum level (sE-cad) was assessed at baseline, after 2 cycles, end of CT and at progression. With 1-stage design, assuming RR 50% (Arm A) and at least 65% (Arm B), planned sample size was 27/54 pts. RR improvement of 5% has 81% probability if true difference is 15%. Results: 88 pts were eligible and randomized, Arm A/B: median age 66/61 years; bladder primary 71.4%/77.2%; metastatic disease 92.9%/89.5%, median CT cycles number 6/5. 28 Arm B pts continued CET (median 3 cycles, range 1-29); 1 Arm A pt had CET added. Most common G3/4 adverse events (AEs) (Arm A, B): neutropenia (34.5%, 33.9%), anemia (10.3%, 5.1%), thrombocytopenia (27.6%, 22%), thromboembolism (6.9%, 18.6%), hyponatremia (3.5%, 10.2%); Arm B only: rash 28.8%, fatigue 11.9%, hypomagnesemia 11.9% and 2 deaths. 84 pts were response evaluable (see table below). sE-cad did not correlate with outcome. sE-cad increased in Arm A, but decreased in Arm B over time. Conclusions: GC/CET was feasible with no outcome improvement but higher AEs. sE-cad did not correlate with outcome. Decreased sE-cad in Arm B suggests potential EGFR-mediated E-cad proteolysis. [Table: see text]