Comprehensive Clinical Staging for Resectable Lung Cancer: Clinicopathological Correlations and the Role of Brain MRI

INTRODUCTION: In our model of comprehensive clinical staging (CCS) for lung cancer, patients with a computerized tomography scan of the chest and upper abdomen not showing distant metastases will then routinely undergo whole body positron emission tomography/computerized tomography and magnetic resonance imaging (MRI) of the brain before any therapeutic decision. Our aim was to determine the accuracy of CCS and the value of brain MRI in this population. METHODS: A retrospective analysis of a prospectively entered database was performed for all patients who underwent lung cancer resection from January 2012 to June 2014. Demographics, clinical and pathological stage (seventh edition of the American Joint Committee on Cancer/Union for International Cancer Control tumor, node, and metastasis staging manual), and costs of staging were collected. Correlation between clinical and pathological stage was determined. RESULTS: Of 315 patients with primary lung cancer, 55.6% were female and the mean age was 70 ± 9.6 years. When correlation was analyzed without consideration for substages A and B, 49.8% of patients (158 of 315) were staged accurately, 39.7% (125 of 315) were overstaged, and 10.5% (32 of 315) were understaged. Only 4.7% of patients (15 of 315) underwent surgery without appropriate neoadjuvant treatment. Preoperative brain MRI detected asymptomatic metastases in four of 315 patients (1.3%). At a median postoperative follow-up of 19 months (range 6-43), symptomatic brain metastases developed in seven additional patients. The total cost of CCS in Canadian dollars was $367,292 over the study period, with $117,272 (31.9%) going toward brain MRI. CONCLUSION: CCS is effective for patients with resectable lung cancer, with less than 5% of patients being denied appropriate systemic treatment before surgery. Brain MRI is a low-yield and high-cost intervention in this population, and its routine use should be questioned.