Purpose
This study investigated the prognostic role of tumor 18F-FDG uptake on pretreatment scans as an independent indicator and whether its addition improves risk prediction from Response Evaluation Criteria in Solid Tumors Version 1.1 (RECIST 1.1).
Methods
We measured the SUVmax of the most 18F-FDG–avid tumor lesions on pretreatment scans from 222 patients (age, 60.5 ± 9.5 years; males, 55.2%) with advanced nonsquamous non–small-cell lung cancer who were enrolled in a prospective phase II clinical trial. We then examined the prognostic value of SUVmax compared with other clinical factors, including chemotherapy response according to RECIST 1.1 criteria.
Results
A multivariable Cox proportional hazards model revealed that an SUVmax greater than 16.3 was an independent predictor of poor progression-free survival (hazards ratio, 3.50; 95% confidence interval, 1.89–6.51; P < 0.000) and overall survival (hazards ratio, 6.87; 95% confidence interval, 2.51–18.76; P < 0.000), whereas RECIST 1.1 did not show a significant association with any survival outcome. Furthermore, improvement was achieved by adding SUVmax to RECIST 1.1, which increased the net reclassification index (27.4%; P = 0.046) and integrated discrimination improvement (integrated discrimination improvement, 10.6%; P = 0.026). Similarly, adding RECIST 1.1 to SUVmax also improved net reclassification index (68.9%, P = 0.006) and integrated discrimination improvement (25.4%, P = 0.006) for prognosis prediction.
Conclusions
High tumor 18F-FDG uptake on a pretreatment scan is an independent prognostic indicator that can significantly improve risk stratification when added to RECIST 1.1 for patients with advanced nonsquamous non–small-cell lung cancer.