Pii: S1010-7940(00)00384-5

Objective: This retrospective study evaluates probability of survival and mode of recurrence in patients with a microscopically positive bronchial resection margin following resection for primary bronchogenic carcinoma, as well as in ̄uence of radiotherapy on survival. Methods: From January 1986 to July 1997, 40 patients had a microscopically positive bronchial resection margin following a macroscopically complete resection (17 lobectomies, three bilobectomies, four sleeve-lobectomies, and 16 pneumonectomies). Tissue diagnosis was squamous cell carcinoma in 32 patients, adenocarcinoma in four, adenosquamous carcinoma in two and neuroendocrine carcinoma in two. Lymph node status was N0 in 14 patients, N1 in 10, and N2 in 16. The bronchial margin contained carcinoma in situ in 20 patients, invasive mucosal carcinoma in ®ve, and peribronchial in®ltration in 15. All patients except the three most recent underwent adjuvant radiation therapy. Results: At the conclusion of the study (January 31st, 1999), 30 patients had died: two with post-operative complications, 17 with progressive disease, ten without relation to cancer, and one under unde®ned circumstances. Six of 10 unrelated deaths were interpreted as respiratory complications of radiotherapy. Recurrent disease appeared in 24 patients (60%). Nineteen had progression of initial disease (47.5%): metastatic spread in 12 (30%), isolated local recurrence in four (10%), and combined local recurrence and metastases in three (7.5%). Five patients developed metachronous cancer, with bronchial location in four (10%) and laryngeal in one (2.5%). 5-year survival (Kaplan±Meier) in 20 patients with carcinoma in situ was 38:7 ^ 13:7% (median 31 months), but rose to 55:0 ^ 16:6% when excluding seven deaths not related to cancer (®ve of whom were secondary to radiotherapy) (x ˆ 3:080; P ˆ 0:0792). Survival in 13 patients classi®ed N0 was 51:3 ^ 16:3% (median 61 months), and 71:1 ^ 18:0% following exclusion of unrelated deaths (x ˆ 3:939; P ˆ 0:0472). Adverse prognosis of peribronchial in®ltration was correlated to a positive N status (13 N2 and 2 N1), 5-year survival being 20:0 ^ 10:3% (median: 18 months). Conclusions: Prognosis of peribronchial in®ltration is similar to N2 disease. In situ carcinoma does not in ̄uence survival per se. Local control of disease is probably in part due to radiotherapy. However, the high prevalence of unrelated late deaths suggests an adverse impact of radiotherapy on survival. q 2000 Elsevier Science B.V. All rights reserved.