BACKGROUND: In patients with non-small cell lung cancer (NSCLC), previous studies have shown a prognostic benefit of maximum standardized uptake (SUV(max)) values on positron emission tomography (PET). Because tumor size is also prognostic and is associated with SUV(max), we sought to better characterize their relationship. We hypothesize that the ratio of SUV(max) to tumor size is a clinically useful measurement. METHODS: A retrospective review was performed for patients (tumors ≥ 1 cm) undergoing resection of NSCLC. Patients were placed into quartiles (SUV(max) and SUV(max) to tumor size ratio) and compared for clinical and pathologic factors. Predictors of SUV(max) and SUV(max) to tumor size ratio on survival were evaluated. RESULTS: Among 530 patients, increasing tumor size (odds ratio [OR], 2.04; confidence interval [CI], 1.68-2.47; p < 0.001) was an independent predictor of higher SUV(max). Patients in quartiles by the ratio of SUV(max) to tumor size demonstrated no significant difference in median tumor size. Those patients with the highest ratios (QR4, 3.21-27.5) more frequently had poorly differentiated tumors (51%; p < 0.001), were likely to have lymph node metastases (30%; p < 0.001), and had poor 3-year disease-free survival (DFS) (58%; p = 0.013). On multivariate analysis, as a continuous variable SUV(max) to tumor size ratio was a stronger independent predictor of survival than SUV(max) alone (hazard ratio [HR], 1.06; CI, 1.00-1.13 versus HR, 1.02; CI, 0.99-1.06). Using cutpoint analysis, a high SUV(max) to tumor size ratio was also a stronger predictor of survival than was high SUV(max) alone, particularly for tumors 1-3 cm (HR, 1.53; CI, 0.93-2.53 versus HR, 1.15; CI, 0.69-1.93). CONCLUSIONS: The ratio of SUV(max) to tumor size may be a more important indicator of prognosis than SUV(max) alone in patients with NSCLC. In particular, the use of the ratio may be appropriate for identifying patients with small tumors who are at high risk for lymph node metastases and poor survival.
BACKGROUND: In patients with non-small cell lung cancer (NSCLC), previous studies have shown a prognostic benefit of maximum standardized uptake (SUV(max)) values on positron emission tomography (PET). Because tumor size is also prognostic and is associated with SUV(max), we sought to better characterize their relationship. We hypothesize that the ratio of SUV(max) to tumor size is a clinically useful measurement. METHODS: A retrospective review was performed for patients (tumors ≥ 1 cm) undergoing resection of NSCLC. Patients were placed into quartiles (SUV(max) and SUV(max) to tumor size ratio) and compared for clinical and pathologic factors. Predictors of SUV(max) and SUV(max) to tumor size ratio on survival were evaluated. RESULTS: Among 530 patients, increasing tumor size (odds ratio [OR], 2.04; confidence interval [CI], 1.68-2.47; p < 0.001) was an independent predictor of higher SUV(max). Patients in quartiles by the ratio of SUV(max) to tumor size demonstrated no significant difference in median tumor size. Those patients with the highest ratios (QR4, 3.21-27.5) more frequently had poorly differentiated tumors (51%; p < 0.001), were likely to have lymph node metastases (30%; p < 0.001), and had poor 3-year disease-free survival (DFS) (58%; p = 0.013). On multivariate analysis, as a continuous variable SUV(max) to tumor size ratio was a stronger independent predictor of survival than SUV(max) alone (hazard ratio [HR], 1.06; CI, 1.00-1.13 versus HR, 1.02; CI, 0.99-1.06). Using cutpoint analysis, a high SUV(max) to tumor size ratio was also a stronger predictor of survival than was high SUV(max) alone, particularly for tumors 1-3 cm (HR, 1.53; CI, 0.93-2.53 versus HR, 1.15; CI, 0.69-1.93). CONCLUSIONS: The ratio of SUV(max) to tumor size may be a more important indicator of prognosis than SUV(max) alone in patients with NSCLC. In particular, the use of the ratio may be appropriate for identifying patients with small tumors who are at high risk for lymph node metastases and poor survival.