Purpose: A retrospective review of surgically treated lung cancer patients imaged preoperatively by F-18 fluorodeoxyglucose-positron emission tomography ([(18)F]FDG-PET) to determine if the primary tumor standardized uptake value (SUV) predicts survival.
Patients and methods: Non-small-cell lung cancer or carcinoid pT1-4, N0-2, M0 patients treated by R0 surgical resection alone were imaged with computed tomography scan and PET within 90 days before surgery. Prognostic variables were assessed by log-rank test; survival was assessed by the method of Kaplan and Meier.
Results: One hundred consecutive patients (48 men, 52 women) were retrospectively reviewed. Median follow-up for surviving patients was 28 months (range, 16 to 81 months). Median maximal SUV (SUV(MAX)) was 9. The 2-year survival for patients with SUV(MAX) more than 9 was 68% and for those with SUV(MAX) less than 9, it was 96% (P <.01, log-rank test). In a multivariate analysis including pathologic tumor size, involved nodes, histology, and SUV(MAX), only tumor size (T) more than 3 cm and SUV(MAX) more than 9 and their interaction were significant predictors of survival (P =.01, 0.02, and < 0.01, respectively). The 3-year survivals for patients with both T less than 3 cm and SUV(MAX) less than 9 was 97%; for those with T less than 3 cm and SUV(MAX) more than 9, it was 94%; for those with T more than 3 cm and SUV(MAX) less than 9, it was 93%; and for those with T more than 3 cm and SUV(MAX) more than 9, it was 47% (P <.01).
Conclusion: In surgically managed lung cancer patients, SUV is a predictor of overall survival after resection. The addition of SUV(MAX) to pathologic tumor size identifies a subgroup of patients at highest risk for death as a result of recurrent disease after resection.