ABSTRACT
BACKGROUND: Concurrent chemoradiotherapy is an established component of the nonoperative management of locally advanced head and neck squamous cell carcinoma (HNSCC), but the standard dose of 100 mg/m2 cisplatin every 3 weeks is associated with clinically significant toxicity. Interest in a more tolerable regimen has led to the widespread use of weekly lower dose cisplatin, but few randomized trials have compared these approaches. METHODS: We examined outcomes of patients with stage III-IVb HNSCC treated with definitive intent chemoradiotherapy using either high-dose cisplatin (HDC) or low-dose cisplatin (LDC), using population-based Veterans Affairs data. In an intent-to-treat analysis, patients were assigned to the HDC vs LDC group according to the dose of their first cycle. Variables potentially influencing treatment decisions including cancer site, stage, smoking/alcohol use, and comorbidities were used to generate propensity scores (PS) for the use of HDC. We compared overall survival (OS) by treatment group using Cox regression, adjusting for PS. We then determined the risk of toxicities using PS-adjusted logistic regression. RESULTS: A total of 2901 patients were included in the analysis; 2200 received HDC (mean initial dose 100 mg/m2). The mean initial dose of LDC was 40 mg/m2. After PS adjustment, HDC was not associated with improved OS over LDC (hazard ratio = 0.94, 95% confidence interval = 0.80 to 1.04). Adjusting for PS, HDC was associated with an increased risk of acute kidney injury, neutropenia, dehydration/electrolyte disturbance, and hearing loss. CONCLUSION: In this large, population-based study of US military veterans, LDC was associated with similar survival to HDC in the nonoperative definitive management of locally advanced HNSCC of the oral cavity, oropharynx, and hypopharynx/larynx. HDC was associated with statistically significantly more toxicity than LDC. Adoption of LDC may reduce toxicity burden while maintaining OS.
Subject(s)
Antineoplastic Agents/administration & dosage , Chemoradiotherapy/methods , Cisplatin/administration & dosage , Head and Neck Neoplasms/therapy , Radiation-Sensitizing Agents/administration & dosage , Squamous Cell Carcinoma of Head and Neck/therapy , Acute Kidney Injury/chemically induced , Aged , Antineoplastic Agents/adverse effects , Chemoradiotherapy/adverse effects , Confidence Intervals , Drug Administration Schedule , Female , Head and Neck Neoplasms/mortality , Head and Neck Neoplasms/pathology , Humans , Intention to Treat Analysis , Male , Middle Aged , Propensity Score , Radiation-Sensitizing Agents/adverse effects , Squamous Cell Carcinoma of Head and Neck/mortality , Squamous Cell Carcinoma of Head and Neck/pathology , Treatment Outcome , VeteransABSTRACT
BACKGROUND: The addition of cisplatin or cetuximab to radiation therapy (RT) improves outcomes in comparison with RT alone in the nonoperative management of head and neck squamous cell carcinoma (HNSCC), but limited data exist for comparing these approaches. Using Veterans Health Affairs data, this study compared the outcomes of patients treated with RT plus cisplatin or cetuximab. METHODS: Patients with stage III to IVb HNSCC who had been treated nonsurgically with RT and cisplatin or cetuximab from 2000 to 2016 within the Veterans Health Affairs system were identified. Patients were analyzed by the drug used in the first treatment cycle (intent to treat). Overall survival (OS) was compared by treatment group with Cox regression models, and propensity score (PS) methods were used to account for a treatment allocation bias. The risk of toxicities was determined, with logistic regression models fit into propensity-matched cohorts. RESULTS: A total of 4520 patients were included in the analysis with a median follow-up of 3 years: 83% received cisplatin. Cisplatin patients were younger (P < .001) and had fewer comorbidities (P < .001). In an unmatched analysis, cetuximab was associated with inferior OS (P < .001). After PS matching, cetuximab treatment remained statistically significantly associated with inferior OS (1.7 vs 4.1 years; hazard ratio, 1.61; 95% confidence interval, 1.44-1.79; P < .001). These differences remained significant across all primary HNSCC subsites and in comparison with low- and high-dose cisplatin. CONCLUSIONS: Cetuximab with RT yields inferior OS in comparison with cisplatin for the nonoperative management of stage III to IVb HNSCC. According to this study, cisplatin may be the most appropriate partner for RT in this setting.
