Gefitinib for oesophageal cancer progressing after chemotherapy (COG): a phase 3, multicentre, double-blind, placebo-controlled randomised trial

摘要: Summary Background Evidence is scarce for the effectiveness of therapies oesophageal cancer progressing after chemotherapy, and no randomised trials have been reported. We aimed to compare gefitinib with placebo in previously treated advanced cancer. Methods For this phase 3, parallel, randomised, placebo-controlled trial, eligible patients were adults or type I/II Siewert junctional tumours, histologically confirmed squamous-cell carcinoma adenocarcinoma, who had progressed WHO performance status 0–2, measurable evaluable disease on CT scan. Participants recruited from 48 UK centres randomly assigned (1:1) (500 mg) matching by simple randomisation stratification factors. Patients, clinicians, trial office staff masked treatment allocation. Treatment continued until progression, unacceptable toxicity, patient choice. The primary outcome was overall survival, analysed intention treat. This registered, number ISRCTN29580179. Findings Between March 30, 2009, Nov 18, 2011, 450 groups (one withdrew consent; 224 allocated 225 included analyses). Overall survival did not differ between (median 3·73 months, 95% CI 3·23–4·50, vs 3·67 2·97–4·37, placebo; hazard ratio [HR] 0·90, 0·74–1·09, p=0·29). Among prespecified patient-reported outcomes (110 121 completed both baseline 4 week questionnaires analyses), odynophagia significantly better group (adjusted mean difference −8·61, −14·49 −2·73; n=227; p=0·004), whereas other improved compared placebo: global quality life (2·69, −2·33 7·72, n=231, p=0·293), dysphagia (−3·18, −8·36 2·00, p=0·228), eating (−4·11, −9·96 1·75, n=229, p=0·168). Median progression-free marginally longer than it (1·57 1·23–1·90 1·17 1·07–1·37 group; HR 0·80, 0·66–0·96, p=0·020). most common toxicities diarrhoea (36 [16%] six [3%] placebo) skin toxicity (46 [21%] two [1%]), mostly grade 2. commonest 3–4 fatigue (24 [11%] 13 [6%] patients) (13 [1%]). Serious adverse events reported 109 (49%) 101 (45%) placebo. 54 (24%) achieved control at 8 weeks, as 35 (16%) (p=0·023). Interpretation use a second-line unselected does improve but has palliative benefits subgroup these difficult-to-treat short expectancy. Future research should focus identification predictive biomarkers identify benefiting patients. Funding Cancer Research UK.