RESUMO
OBJECTIVE: Administration of postoperative chemotherapy to patients with completely resected stage I NSCLC is still a matter of debate. The aim of the present study was to evaluate the value of different baseline 18F-FDG PET parameters in identifying surgical stage I NSCLC patients who are at high risk of recurrence, and thus are indicated for further postoperative treatment. METHODS: This is a retrospective study, which included 49 patients (28 males, 21 females) with the median age of 69 years (range 28-84), who had pathologically proven stage I NSCLC. All patients underwent 18F-FDG PET/CT at baseline followed by complete surgical resection of the tumor (R0). Baseline SUVmax, MTV and TLG were measured. Patients' follow-up records were retrospectively reviewed, and DFS (disease-free survival) was assessed. For each parameter, the most accurate cut-off value for the prediction of recurrence was calculated using the ROC curve analysis and the Youden index. DFS was evaluated for patients above and below the calculated cut-off value using the Kaplan-Meier method and the difference in survival between the two groups was estimated using the log-rank test. RESULTS: Median observation time of the patients after surgery was 28.7 months (range 3.5-58.8 months). 9 patients developed recurrence. The calculated cut-off values for SUVmax, MTV and TLG were 6, 6.6 and 33.6, respectively. Using these cut-offs, the observed sensitivity for SUVmax, MTV and TLG for prediction of recurrence was 100%, 89% and 89%, respectively, while the observed specificity was 43%, 73% and 65%, respectively. The difference in survival between patients below and above the cut-off value was statistically significant in all three studied parameters. The highest AUC was observed for MTV (AUC = 0.825, p = 0.003), followed by TLG (AUC = 0.789, p = 0.007), and lastly SUVmax (AUC = 0.719, p = 0.041). ROC curve analysis showed that volumetric parameters had better predictive performance than SUVmax as regards recurrence. CONCLUSION: PET-derived parameters at baseline were predictive of recurrence in stage I surgical NSCLC patients. Moreover, the metabolic volume of the tumor was the most significant parameter for this purpose among the studied indices.