Cisplatin and fluorouracil with or without panitumumab in patients with recurrent or metastatic squamous-cell carcinoma of the head and neck (SPECTRUM): An open-label phase 3 randomised trial

摘要

Background: Previous trials have shown that anti-EGFR monoclonal antibodies can improve clinical outcomes of patients with recurrent or metastatic squamous-cell carcinoma of the head and neck (SCCHN). We assessed the efficacy and safety of panitumumab combined with cisplatin and fluorouracil as first-line treatment for these patients. Methods: This open-label phase 3 randomised trial was done at 126 sites in 26 countries. Eligible patients were aged at least 18 years; had histologically or cytologically confirmed SCCHN; had distant metastatic or locoregionally recurrent disease, or both, that was deemed to be incurable by surgery or radiotherapy; had an Eastern Cooperative Oncology Group performance status of 1 or less; and had adequate haematological, renal, hepatic, and cardiac function. Patients were randomly assigned according to a computer-generated randomisation sequence (1:1; stratified by previous treatment, primary tumour site, and performance status) to one of two groups. Patients in both groups received up to six 3-week cycles of intravenous cisplatin (100 mg/m2 on day 1 of each cycle) and fluorouracil (1000 mg/m2 on days 1-4 of each cycle); those in the experimental group also received intravenous panitumumab (9 mg/kg on day 1 of each cycle). Patients in the experimental group could choose to continue maintenance panitumumab every 3 weeks. The primary endpoint was overall survival and was analysed by intention to treat. In a prospectively defined retrospective analysis, we assessed tumour human papillomavirus (HPV) status as a potential predictive biomarker of outcomes with a validated p16-INK4A (henceforth, p16) immunohistochemical assay. Patients and investigators were aware of group assignment; study statisticians were masked until primary analysis; and the central laboratory assessing p16 status was masked to identification of patients and treatment. This trial is registered with ClinicalTrials.gov, number NCT00460265. Findings: Between May 15, 2007, and March 10, 2009, we randomly assigned 657 patients: 327 to the panitumumab group and 330 to the control group. Median overall survival was 11·1 months (95% CI 9·8-12·2) in the panitumumab group and 9·0 months (8·1-11·2) in the control group (hazard ratio [HR] 0·873, 95% CI 0·729-1·046; p=0·1403). Median progression-free survival was 5·8 months (95% CI 5·6-6·6) in the panitumumab group and 4·6 months (4·1-5·4) in the control group (HR 0·780, 95% CI 0·659-0·922; p=0·0036). Several grade 3 or 4 adverse events were more frequent in the panitumumab group than in the control group: skin or eye toxicity (62 [19%] of 325 included in safety analyses vs six [2%] of 325), diarrhoea (15 [5%] vs four [1%]), hypomagnesaemia (40 [12%] vs 12 [4%]), hypokalaemia (33 [10%] vs 23 [7%]), and dehydration (16 [5%] vs seven [2%]). Treatment-related deaths occurred in 14 patients (4%) in the panitumumab group and eight (2%) in the control group. Five (2%) of the fatal adverse events in the panitumumab group were attributed to the experimental agent. We had appropriate samples to assess p16 status for 443 (67%) patients, of whom 99 (22%) were p16 positive. Median overall survival in patients with p16-negative tumours was longer in the panitumumab group than in the control group (11·7 months [95% CI 9·7-13·7] vs 8·6 months [6·9-11·1]; HR 0·73 [95% CI 0·58-0·93]; p=0·0115), but this difference was not shown for p16-positive patients (11·0 months [7·3-12·9] vs 12·6 months [7·7-17·4]; 1·00 [0·62-1·61]; p=0·998). In the control group, p16-positive patients had numerically, but not statistically, longer overall survival than did p16-negative patients (HR 0·70 [95% CI 0·47-1·04]). Interpretation: Although the addition of panitumumab to chemotherapy did not improve overall survival in an unselected population of patients with recurrent or metastatic SCCHN, it improved progression-free survival and had an acceptable toxicity profile. p16 status could be a prognostic and predictive marker in