BACKGROUND Patients with oesophageal/gastro-oesophageal junction adenocarcinoma (EAC) not showing early metabolic response (EMR) to chemotherapy have poorer survival and histological response rates <5%. We investigated whether tailoring neoadjuvant therapy can improve outcomes in these patients. PATIENTS AND METHODS Patients with resectable EAC were enrolled and randomised into two single-arm, multicentre phase II trials. After induction cisplatin and 5-fluorouracil (CF), all were assessed by day 15 positron emission tomography (PET). Patients with an EMR [maximum standardised uptake values (SUVmax) ≥35% reduction from baseline to day 15 PET] received a second CF cycle then oesophagectomy. Non-responders were randomised 1 : 1 to two cycles of CF and docetaxel (DCF, n = 31) or DCF + 45 Gy radiotherapy (DCFRT, n = 35) then oesophagectomy. The primary end point was major histological response (<10% residual tumour) in the oesophagectomy specimen; secondary end points were overall survival (OS), progression-free survival (PFS), and locoregional recurrence (LR). RESULTS Of 124 patients recruited, major histological response was achieved in 3/45 (7%) with EMR, 6/30 (20%) DCF, and 22/35 (63%) DCFRT patients. Grade 3/4 toxicities occurred in 12/45 (27%) EMR (CF), 13/31 (42%) DCF, and 25/35 (71%) DCFRT patients. No treatment-related deaths occurred. LR by 3 years was seen in 5/45 (11%) EMR, 10/31 (32%) DCF, and 4/35 (11%) DCFRT patients. PFS [95% confidence interval (CI)] at 36 months was 47% (31% to 61%) for EMR, 29% (15% to 45%) for DCF, and 46% (29% to 61%) for DCFRT patients. OS (95% CI) at 60 months was 53% (37% to 67%) for EMR, 31% (16% to 48%) for DCF, and 46% (29% to 61%) for DCFRT patients. CONCLUSIONS EMR is associated with favourable OS, PFS, and low LR. For non-responders, the addition of docetaxel augmented histological response rates, but OS, PFS, and LR remained inferior compared with responders. DCFRT improved histological response and PFS/LR outcomes, matching the EMR group. Early PET/CT has the potential to tailor therapy for patients not showing an early response to chemotherapy. TRIAL REGISTRATION ACTRN12609000665235.

中文翻译：

---

对可切除食管腺癌 (AGITG DOCTOR) 的顺铂和氟尿嘧啶早期反应不佳后，术前顺铂、氟尿嘧啶和多西他赛联合或不联合放疗：来自多中心、随机对照 II 期试验的结果。

背景 食管/胃食管交界处腺癌 (EAC) 患者对化疗未表现出早期代谢反应 (EMR)，生存率较差，组织学反应率 <5%。我们调查了量身定制的新辅助治疗是否可以改善这些患者的预后。患者和方法 可切除 EAC 患者被纳入并随机分配到两个单臂、多中心 II 期试验中。诱导顺铂和 5-氟尿嘧啶 (CF) 后，均通过第 15 天正电子发射断层扫描 (PET) 进行评估。EMR [从基线到第 15 天 PET 的最大标准化摄取值 (SUVmax) 减少 ≥35%] 的患者接受了第二个 CF 周期，然后进行了食管切除术。无反应者以 1:1 的比例随机分配至两个周期的 CF 和多西他赛（DCF，n = 31）或 DCF + 45 Gy 放疗（DCFRT，n = 35），然后进行食管切除术。主要终点是食管切除标本的主要组织学反应（<10% 残留肿瘤）；次要终点是总生存期（OS）、无进展生存期（PFS）和局部复发（LR）。结果 在招募的 124 名患者中，3/45 (7%) EMR、6/30 (20%) DCF 和 22/35 (63%) DCFRT 患者实现了主要组织学反应。12/45 (27%) EMR (CF)、13/31 (42%) DCF 和 25/35 (71%) DCFRT 患者发生 3/4 级毒性。没有发生与治疗相关的死亡。在 5/45 (11%) EMR、10/31 (32%) DCF 和 4/35 (11%) DCFRT 患者中观察到 3 年的 LR。EMR 在 36 个月时的 PFS [95% 置信区间 (CI)] 为 47%（31% 至 61%），DCF 为 29%（15% 至 45%），DCFRT 为 46%（29% 至 61%）患者。60 个月时，EMR 的 OS（95% CI）为 53%（37% 至 67%），DCF 为 31%（16% 至 48%），DCFRT 患者为 46%（29% 至 61%）。结论 EMR 与良好的 OS、PFS 和低 LR 相关。对于无反应者，加入多西他赛可提高组织学反应率，但与反应者相比，OS、PFS 和 LR 仍然较差。DCFRT 改善了组织学反应和 PFS/LR 结果，与 EMR 组匹配。早期 PET/CT 有可能为对化疗没有早期反应的患者量身定制治疗方案。试用注册 ACTRN12609000665235。早期 PET/CT 有可能为对化疗没有早期反应的患者量身定制治疗方案。试用注册 ACTRN12609000665235。早期 PET/CT 有可能为对化疗没有早期反应的患者量身定制治疗方案。试用注册 ACTRN12609000665235。