A phase II trial of perioperative concurrent chemoradiotherapy (CCRT) and gefitinib (G) in locally advanced esophagus (E) and gastroesophageal junction (GEJ) cancer
4547 Background: Since 1999, Cleveland Clinic protocols for locally advanced E or GEJ cancer have used peri-operative CCRT. Despite high locoregional control (LRC) rates, distant metastatic control (DMC) and overall survival (OS) are suboptimal. This phase II study added G to our CCRT regimen in an effort to improve these outcomes. Methods: Eligibility required a diagnosis of E or GEJ squamous cell or adenocarcinoma (ACA), with an esophageal ultrasound stage of at least T3, N1, or M1a. Four-day (d) continuous IV infusions of cisplatin (20mg/m2/d), and fluorouracil (1,000 mg/m2/d) began on d1 of pre-operative radiation (30Gy at 1.5Gy bid). Surgery followed in 4–6 weeks (wks), followed by identical CCRT given 6–10 wks postoperatively. G (250 mg/d) was also started on d1 and given for 4 wks. It was restarted with the postoperative CCRT and given for 2 years. Targeted accrual was 80 patients (pts) with primary endpoints of OS and DMC. Results were retrospectively compared to 93 pts treated with an identical regimen without G. Results: Between 4/03 and 7/06, 80 pts were enrolled. Pt and tumor characteristics were similar to the non-G cohort. Median age was 58 (range 33–75) years; 73 pts (91%) were male and 75 (94%) Caucasian. Tumor began in distal E in 28 pts (35%) and GEJ in 46 (57%), and was an ACA in 75 (94%). Clinical staging was T3N0 in 15 pts (19%), T2–3N1 in 41 (51%) and T4 or M1a in 24 (30%). Induction toxicity was similar in G v. non-G pts except for grade 3–4 esophagitis in 3% v. 14% (p=0.007), neutropenia 1 rash in 42 pts (53%), and grade >1 diarrhea in 44 (55%). Resection was accomplished in 73 pts (91%) and post-operative CCRT completed in 60 (75%). With a median follow up of 33 (range 11–54) months, 3-year Kaplan-Meier estimates (G v. non-G pts) include: OS (40% v. 28%, p=0.07), DMC (46% v. 33%, P=0.20), and LRC (82% v. 77%, p=0.28). Only 50% of pts are projected to complete 2 years of maintenance G. Conclusions: Although toxicity was not increased by adding G to peri-operative CCRT, maintenance G was difficult. The suggestion of a survival advantage in this retrospective comparison merits additional investigation. No significant financial relationships to disclose.