BACKGROUND: The natural history and the best modality of follow-up of atypical lung carcinoids (AC) remain ill defined. The aim of this study was to analyze recurrence-free survival (RFS) after complete resection (R0) of stage I-III pulmonary AC. Secondary objectives were prognostic parameters, the location of recurrences, and the modality of follow-up. METHODS: A retrospective review of 540 charts of AC patients treated between 1998 and 2008 at 10 French and Italian centers with experience in lung neuroendocrine tumor management was undertaken. The exclusion criteria were MEN1-related tumor, history of another cancer, referral after tumor relapse, and being lost to follow-up. A central pathological review was performed in each country. RESULTS: Sixty-two patients were included. After a median follow-up time of 91 months (mean 85, range 6-165), 35% of the patients experienced recurrence: 16% were regional recurrences and 19% were distant metastases. Median RFS was not reached. The 1-, 3-, and 5-year RFS rate was 90, 79, and 68%, respectively. In univariate analysis, lymph node involvement (p = 0.0001), stage (p = 0.0001), mitotic count (p = 0.004), and type of surgery (p = 0.043) were significantly associated with RFS. In multivariate analysis, lymph node involvement was significantly associated with RFS (HR 95% CI: 0.000-0.151; p = 0.004). During follow-up, somatostatin receptor scintigraphy, fibroscopy, and abdominal examination results were available for 22, 12, and 25 patients, respectively. The median time interval for imaging follow-up was 10 months. CONCLUSIONS: After complete resection of AC, recurrences were observed mostly within the first 5 years of follow-up, within bronchi, mediastinal nodes, the liver, and bones. In R0 patients, lymph node involvement could help to stratify follow-up intervals. Suboptimal imaging is evidenced.
BACKGROUND: The natural history and the best modality of follow-up of atypical lung carcinoids (AC) remain ill defined. The aim of this study was to analyze recurrence-free survival (RFS) after complete resection (R0) of stage I-III pulmonary AC. Secondary objectives were prognostic parameters, the location of recurrences, and the modality of follow-up. METHODS: A retrospective review of 540 charts of AC patients treated between 1998 and 2008 at 10 French and Italian centers with experience in lung neuroendocrine tumor management was undertaken. The exclusion criteria were MEN1-related tumor, history of another cancer, referral after tumor relapse, and being lost to follow-up. A central pathological review was performed in each country. RESULTS: Sixty-two patients were included. After a median follow-up time of 91 months (mean 85, range 6-165), 35% of the patients experienced recurrence: 16% were regional recurrences and 19% were distant metastases. Median RFS was not reached. The 1-, 3-, and 5-year RFS rate was 90, 79, and 68%, respectively. In univariate analysis, lymph node involvement (p = 0.0001), stage (p = 0.0001), mitotic count (p = 0.004), and type of surgery (p = 0.043) were significantly associated with RFS. In multivariate analysis, lymph node involvement was significantly associated with RFS (HR 95% CI: 0.000-0.151; p = 0.004). During follow-up, somatostatin receptor scintigraphy, fibroscopy, and abdominal examination results were available for 22, 12, and 25 patients, respectively. The median time interval for imaging follow-up was 10 months. CONCLUSIONS: After complete resection of AC, recurrences were observed mostly within the first 5 years of follow-up, within bronchi, mediastinal nodes, the liver, and bones. In R0 patients, lymph node involvement could help to stratify follow-up intervals. Suboptimal imaging is evidenced.