Use of preoperative chemotherapy with or without postoperative radiotherapy in technically resectable stage IIIA non-small-cell lung cancer. Provincial Lung Cancer Disease Site Group.

GUIDELINE QUESTION: Should preoperative (neoadjuvant) cisplatin-based chemotherapy with or without postoperative radiotherapy be offered to patients with technically resectable stage IIIA non-small-cell lung cancer (NSCLC) to improve survival? (Resectability should be determined preoperatively by a thoracic surgeon.) OBJECTIVE: To make recommendations about the use of preoperative cisplatin-based chemotherapy with or without postoperative radiotherapy in technically resectable stage IIIA NSCLC. OUTCOMES: Survival is the primary outcome of interest. PERSPECTIVES (VALUES): Evidence was collected and reviewed by 4 members of the Lung Cancer Disease Site Group (LCDSG) of the Cancer Care Ontario Practice Guidelines Initiative. The evidence was then presented to the full LCDSG and discussed extensively at 5 of its meetings. The LCDSG comprises medical and radiation oncologists, pathologists, surgeons, epidemiologists, a psychologist and a medical sociologist. A community representative was present at one meeting during which the recommendation was discussed. QUALITY OF EVIDENCE: Four small randomized controlled trials (RCTs) were available for review; 2 were completed and were reported in full in the literature, 1 was published in abstract form, and 1 was closed and was reported as an interim analysis. Although the RCTs used appropriate clinical trials methodology, including planned interim analyses and early stopping rules, retrospective review revealed inconsistencies between the treatment arms for subsets of stage IIIA disease and for prognostic factors. These factors and the small samples in each study limit the interpretation of the results. BENEFITS: The data from 2 of the 4 trials were not combined because the data were not mature in one case and not extractable in the other. The 2 fully published, completed trials reported a survival benefit for patients treated with preoperative chemotherapy with or without postoperative radiotherapy compared with those not given preoperative chemotherapy. One trial reported a median survival of 26 months in the treatment group versus 8 months in the control group (p < 0.001). A second trial reported an estimated median survival of 64 months versus 11 months (p < 0.008) and a 3-year survival rate of 56% versus 15% respectively. A pooled analysis of the 2-year survival data from the 2 completed RCTs yielded an odds ratio for death of 0.18 (95% confidence interval 0.06 to 0.51) in favour of preoperative chemotherapy. HARMS: There was no difference in the postoperative mortality between the trials reviewed. Toxic effects associated with the chemotherapy were limited primarily to neutropenic fever, nausea and vomiting.