Prognostic value of maximal SUV on preoperative FDG PET in patients with esophageal squamous cell cancer

Objectives We investigated whether maximal SUV can be used to predict disease free survival in patients with esophageal cancer.

Methods This study enrolled 28 consecutive patients (m:f = 26:2; age = 62.7±9.2 yr) with esophageal squamous cell cancer diagnosed between April 2003 and December 2007. All of patients had operation after preoperative FDG PET study for stage work up. FDG finding (SUV of primary lesion, number of detected LNs), pathologic results (tumor size, grade, pathologic stage), localization of tumor, treatment status (Op vs. Op + CTx or RTx) were included as variables. Disease free-survival of each variable was estimated by the Kaplan-Meier method. We evaluated multivariable analysis using Cox's regression method to discriminate of independent prognostic variable.

Results From univariate analysis, maximal SUV =5.5, T stage (T1-T2, T3-4) had significant prognostic predictors of disease free survival (p<0.05). Tumor grade and localization had weak significant prognostic predictor value of disease free survival (p<0.01). After multivariate analysis, only maximal SUV = 5.5 was independent predictor value of poor outcome (p<0.002).

Conclusions With maximal SUV of primary lesion, we can predict disease free survival in patients with esophageal squamous cell cancer who were fit to undergo surgery.