Objective: To assess the feasibility of a policy of conservative resection for carcinoid tumours, to validate this by a study of recurrence rates and survival, and to assess those factors that might frustrate such a policy. Methods: Over 20 years, 95 patients with a final histological diagnosis of carcinoid tumour were assessed for surgery. During this time we had a policy of conservative resection for such tumours, with preservation of functional lung parenchyma wherever possible. Results: The mean age at presentation was 51 years (range 14-81). Symptoms were present in 62 patients (65.3%). Fifty-eight rumours (61.1%) were central in position. Surgical resection was performed in 92 patients. Three patients (3.3%) underwent pneumonectomy: in two because the situation was complicated by destruction or severe damage to the distal lung parenchyma, and in one the initial biopsy was interpreted as lung cancer. Less than 50% of patients were referred with the correct histological diagnosis. In 18% preoperative biopsies were interpreted as non-small cell lung cancer (NSCLC). At thoracotomy similar confusion remained with 26% of frozen section reports suggesting NSCLC. In those patients coming to thoracotomy, lymph node involvement was present in 15 patients (16.3%) (N1 in 13 patients, N2 in two patients) being found in 11 of 81 (13.6%) patients whose tumours showed typical histological features, and four of the 14 patients whose rumours (28.5%) displayed atypical features. During followup from 6 months to 12 years (mean 3.9 years) four of the 92 operated patients were found to have local recurrence and underwent further surgery and three others developed distant metastases (3.2%), two of whom have died. Conclusions: Bronchial carcinoid is now considered to be a low-grade, but malignant tumour. Despite this we have found over the last 20 years that a policy of conservative resection is feasible and safe whenever the true histology is known and the distal lung parenchyma is functional. This is not affected by the presence of nodal involvement or atypical features and the long-term results of conservative resection are not affected by the presence of nodal disease. These factors should not influence the extent of surgical resection. (C) 2000 Elsevier Science B.V. All rights reserved.
