Evaluation of a treatment strategy for optimising preoperative chemoradiotherapy in stage III non-small-cell lung cancer.

OBJECTIVE: Concurrent chemoradiotherapy is standard of care in stage III non-small-cell lung cancer, although surgery may be beneficial in selected patients in whom induction therapy has achieved 'down-staging' of mediastinal nodal disease. Previous studies incorporated treatment 'splits' for re-evaluation, and such gaps lead to poorer survival in patients undergoing chemoradiotherapy. We describe the outcome of a treatment strategy to limit the duration of treatment splits. METHODS: A prospective database (2003-2007) of stage III non-small-cell lung cancer patients treated with concurrent chemoradiotherapy outwith clinical trials at our centre was reviewed. Preoperative chemoradiotherapy consisted of one induction course of cisplatin-gemcitabine, followed by two courses of cisplatin-etoposide with once-daily thoracic radiotherapy using four-dimensional involved-field treatment planning. After a dose of 46-50 Gy, potentially resectable patients without disease progression underwent immediate planned mediastinal re-staging and patients with persistent N2 disease or who were unfit for surgery continued to full-dose radiotherapy. Effort was made to shorten the treatment split by substituting mediastinoscopy for endoscopic procedures (transbronchial and -oesophageal). RESULTS: A total of 34 patients had potentially resectable disease at the start of treatment. Toxicity of chemoradiotherapy was predominantly leucocytopaenia grade III/IV in 38% of courses and grade III oesophagitis in five patients (15%), but was manageable and reversible. After re-staging, 24 patients (71%) proceeded to surgery. A radical resection was achieved in 23 patients; nine had a complete pathological response. Re-staging was accurate with only one false-negative mediastinoscopy. One patient died 10 days after surgery. Median time from end of induction treatment to re-staging or surgery was 12 (range: 0-51 days) and 35 days (range: 18-63 days), respectively. Median survival for resected patients was not reached. Six patients had persisting N2 disease, of which two continued radiotherapy after a split of 3 and 4 days. CONCLUSIONS: Image-guided, involved-field preoperative chemoradiotherapy can be performed with acceptable toxicity, and the present strategy achieves the goal of limiting splits in treatment delivery that may adversely affect survival in patients who do not undergo down-staging with induction therapy.