Comparative prognosis of long-term follow-up over 10 years and dropout from follow-up after resection of lung cancer

Background: Most recurrences of non-small cell lung cancer (NSCLC) after lung resection occur within 5 years, which is why 5-year overall survival rates are used to give prognoses for lung cancer. Elderly individuals also often show comorbidities and may die from other diseases. Few studies have examined the long-term prognosis of elderly patients with NSCLC, and no reports have investigated drop-out from follow-up after resection of NSCLC, including in elderly patients. This retrospective cohort study analyzed and surveyed long-term prognosis and drop-out from follow-up, including in elderly patients, after resection of lung cancer. Methods: We identified 349 consecutive patients after lung resection between January 2009 and March 2011. Twenty-two cases were excluded because of small cell lung cancer, past metachronous multiple lung cancers, recurrences of lung cancer, surgical biopsy, and other reasons. We investigated recurrences and causes of death in all patients and defined cases for which follow-up could not be conducted even by telephone or documentation from patients or the public office as cases of drop-out from follow-up. Results: Of the 327 cases analyzed, 81 cases dropped out from follow-up and 246 cases completed >10 years of follow-up. Multivariable analysis demonstrated age >= 75 years [odds ratio 1.83; 95% confidence interval (CI): 1.01-3.32] and female sex (odds ratio 1.87, 95% CI: 1.06-3.3) as independent risk factors for drop-out from follow-up. Recurrence was detected >5 years after surgery in 5 cases (2.0%, 5/246 cases). Five- and 10-year overall survival rates were 67.2% and 52.9% for patients <75 years of age, and 42.1% and 21.1% for patients of age >= 75 years (P<0.001). Overall survival was significantly better for those of age <75 years than for those of age >= 75 years. Disease-specific survival did not differ significantly between groups (5-year disease-specific survival rate: 71.9% vs. 73.4%; 10-year disease-specific survival: 66.0% vs. 61.2%, P=0.80). Using Cox proportional hazard regression, age >= 75 years (hazard ratio 2.221; 95% CI: 1.507-3.274; P<0.001) and stage >= 2 (hazard ratio 2.628; 95% CI: 1.868-3.698; P<0.001) were significantly associated with overall survival. Conclusions: In patients with NSCLC after lung resection, age >= 75 years and female sex were risk factors for dropping out from follow-up over 10 years. In addition, patients >= 75 years of age have a high possibility of dying from other diseases and sufficient consideration of and informed consent for surgical indications are necessary.