Background

Management of patients with locoregionally advanced head and neck squamous cell carcinoma (HNSCC) when cisplatin is contraindicated is controversial. We aimed to assess whether radiotherapy with concurrent and adjuvant durvalumab would improve outcomes compared with radiotherapy with cetuximab.

Methods

NRG-HN004 was designed as an open-label, multicentre, parallel-group, randomised, phase 2/3 trial with safety lead-in conducted at 89 academic and community medical centres in North America. Eligible patients were aged 18 years or older with American Joint Committee on Cancer 8th edition stage III–IVB p16-negative HNSCC or unfavourable stage I–III p16-positive oropharyngeal or unknown primary carcinoma, who had a contraindication to cisplatin (Eastern Cooperative Oncology Group [ECOG] performance status 2, renal or hearing impairment, peripheral neuropathy, aged at least 70 years with moderate or severe comorbidity, or aged younger than 70 years with severe comorbidity). Patients were randomly assigned (2:1) by permuted block randomisation (multiples of 6) to intravenous durvalumab 1500 mg starting 2 weeks before radiotherapy then every 4 weeks starting week 2 of radiotherapy (seven cycles) or intravenous cetuximab 400 mg/m² 1 week before radiotherapy then 250 mg/m² weekly beginning week 1 of radiotherapy (eight cycles), with intensity-modulated radiotherapy (70 Gy in 35 fractions over 7 weeks). Stratification factors were tumour and nodal stage, ECOG performance status and comorbidity, and primary site and p16 status. The phase 2 primary endpoint was progression-free survival in the intention-to-treat population. There was one prespecified interim futility analysis at 50% of progression-free survival information. If the observed hazard ratio was 1·0 or more, favouring cetuximab, early stopping would be considered. Extended follow-up analysis was post hoc. This trial is registered with ClinicalTrials.gov, NCT03258554, and is closed to enrolment.

Findings

Following a ten-patient safety lead-in, the phase 2 trial enrolled 190 patients from March 12, 2019, to July 30, 2021, 186 of whom were randomly assigned (123 to durvalumab and 63 to cetuximab). Median age was 72 years (IQR 64–77), 30 (16%) patients were women and 156 (84%) were men. Phase 2 accrual was suspended in July 30, 2021, following an interim futility analysis, and permanently closed in Sept 1, 2022. The phase 3 part of the trial was not conducted. At a median follow-up of 2·3 years (IQR 1·9–3·1) for the extended follow-up (data cutoff July 31, 2023; post-hoc analysis), 2-year progression-free survival was 50·6% (95% CI 41·5–59·8) in the durvalumab group versus 63·7% (51·3–76·1) in the cetuximab group (hazard ratio 1·33 [95% CI 0·84–2·12]; p=0·89). Adverse events were similar in both groups. The most common grade 3–4 adverse events were dysphagia (26 [22%] of 119 patients in the durvalumab group vs 18 [30%] of 61 patients in the cetuximab group), lymphopenia (33 [28%] vs 20 [33%]), and oral mucositis (13 [11%] vs 11 [18%]). Four (3%) patients in the durvalumab group and one (2%) in the cetuximab group died from treatment-related adverse events (death not otherwise specified, laryngeal oedema, lung infection, and respiratory failure in the durvalumab group and sudden death not otherwise specified in the cetuximab group).

Interpretation

Our findings suggest that durvalumab did not improve outcomes compared with cetuximab in patients with HNSCC with contraindications to cisplatin. Further trials are needed to define the standard of care for this population.

Funding

US National Cancer Institute and AstraZeneca.

中文翻译：

---

西妥昔单抗或 durvalumab 放疗治疗有顺铂禁忌症的患者局部晚期头颈癌 （NRG-HN004）：一项开放标签、多中心、平行组、随机、2/3 期试验

 背景

对于局部晚期头颈部鳞状细胞癌 （HNSCC） 患者禁忌使用顺铂的管理存在争议。我们旨在评估与西妥昔单抗放疗相比，同时和辅助 durvalumab 放疗是否会改善结局。

 方法

NRG-HN004 被设计为一项开放标签、多中心、平行组、随机、2/3 期试验，在北美的 89 个学术和社区医疗中心进行安全导入。符合条件的患者年龄在 18 岁或以上，患有美国癌症联合委员会第 8 版 III-IVB 期 p16 阴性 HNSCC 或预后不良的 I-III 期 p16 阳性口咽癌或不明原发癌，有顺铂禁忌症（东部肿瘤合作组 [ECOG] 体能状态 2，肾脏或听力障碍，周围神经病变，年龄至少 70 岁，有中度或重度合并症， 或年龄小于 70 岁且患有严重合并症）。通过排列区组随机化（6 的倍数）将患者随机分配 （2：1） 在放疗前 2 周开始静脉注射 durvalumab 1500 mg，然后从放疗第 2 周开始每 4 周一次（7 个周期）或静脉注射西妥昔单抗 400 mg/m²，放疗前 1 周，然后每周 250 mg/m²，从放疗第 1 周开始（八个周期）， 调强放疗（70 Gy，在 7 周内分 35 次）。分层因素为肿瘤和淋巴结分期、 ECOG 体能状态和合并症以及原发部位和 p16 状态。2 期主要终点是意向治疗人群的无进展生存期。在 50% 的无进展生存期信息中，有一项预先指定的中期无效分析。如果观察到的风险比为 1·0 或更高，有利于西妥昔单抗，则应考虑提前停止。扩展随访分析是事后进行的。此试用版已在 ClinicalTrials.gov NCT03258554注册，并且已停止注册。

 发现

在 10 名患者的安全性导入后，2 期试验于 2019 年 3 月 12 日至 2021 年 7 月 30 日招募了 190 名患者，其中 186 名患者被随机分配（123 名接受度伐利尤单抗治疗，63 名接受西妥昔单抗治疗）。中位年龄为 72 岁 （IQR 64-77），30 名 （16%） 患者为女性，156 名 （84%） 为男性。在进行临时无效分析后，第 2 阶段的应计于 2021 年 7 月 30 日暂停，并于 2022 年 9 月 1 日永久关闭。试验的 3 期部分未进行。在延长随访的中位随访 2·3 年 （IQR 1·9–3·1） 时（数据截止日期为 2023 年 7 月 31 日;事后分析），度伐利尤单抗组的 2 年无进展生存率为 50·6% （95% CI 41·5–59·8），而西妥昔单抗组为 63·7% （51·3–76·1） （风险比 1·33 [95% CI 0·84–2·12];p=0·89）。两组的不良事件相似。最常见的 3-4 级不良事件是吞咽困难（度伐利尤单抗组 119 名患者中有 26 名 [22%] 对西妥昔单抗组 61 名患者中有 18 名 [30%]）、淋巴细胞减少症（33 名 [28%] 对 20 名 [33%]）和口腔粘膜炎（13 名 [11%] 对 11 名 [18%]）。度伐利尤单抗组有 4 例 （3%） 患者和西妥昔单抗组有 1 例 （2%） 患者死于治疗相关不良事件（度伐利尤单抗组未另行说明的死亡、喉水肿、肺部感染和呼吸衰竭，西妥昔单抗组未另行说明的猝死）。

 解释

我们的研究结果表明，与西妥昔单抗相比，度伐利尤单抗并未改善有顺铂禁忌症的 HNSCC 患者的结局。需要进一步的试验来确定该人群的护理标准。

 资金

美国国家癌症研究所和阿斯利康。