Accuracy of FDG-PET and spiral CT for the early prediction of non-response to preoperative chemoradiotherapy in patients with esophageal cancer.

Background:
Surgical resection is the preferred treatment of potentially curable esophageal cancer. To improve long term patient outcome, many institutes apply neoadjuvant chemo-(radio-)therapy. In a large proportion of patients no response to chemoradiotherapy is achieved. These patients suffer from toxic and ineffective neoadjuvant treatment, while appropriate surgical therapy is delayed. For this reason a diagnostic test that allows for accurate prediction of tumor response early during chemoradiotherapy is of crucial importance. CT-scan and endoscopic ultrasonography have limited accuracy in predicting histopathologic tumor response. Data suggest that metabolic changes in tumor tissue as measured by FDG-PET predict response better.

Objective:
To compare FDG-PET and CT-scan for the early prediction of non-response to preoperative chemoradiotherapy in patients with potentially curable esophageal cancer.
Design: Prognostic accuracy study, embedded in a randomized multicenter Dutch trial comparing neoadjuvant chemoradiotherapy for 5 weeks followed by surgery versus surgery alone for esophageal cancer. This prognostic accuracy study is performed only in the neoadjuvant arm of the randomized trial (CROSS).

Intervention:
In 8 centers, 150 consecutive patients will be included in this prognostic accuracy study over a 3 year period. FDG-PET and CT-scan will be performed independently before and 2 weeks after the start of the chemoradiotherapy. All patients complete the 5 weeks regimen of neoadjuvant chemoradiotherapy, regardless the test results.

Reference standard:
Histology in the surgical resection specimen. Responders are defined as patients with < 10% viable residual tumor cells (Mandard-score).
Data analysis: Difference in accuracy (area under ROC curve) and negative predictive value between FDG-PET and CT-scan.

Economic evaluation:
The economic evaluation has been designed as a cost-effectiveness study, comparing survival and costs associated with the use of FDG-PET (or CT-scan) to predict tumor response with survival and costs of neoadjuvant chemoradiotherapy without prediction of response (reference strategy). Patient outcome and costs after false positive and false negative results will be based on the data of the randomized clinical trial in which this accuracy study is embedded.

To compare FDG-PET and CT-scan for the early prediction of non-response to preoperative chemoradiotherapy in patients with potentially curable esophageal cancer.

150 consecutive patients will be included in this prognostic accuracy study over a 3 year period. FDG-PET and CT-scan will be performed independently before and 2 weeks after the start of the chemoradiotherapy.
All patients complete the 5 weeks regimen of neoadjuvant chemoradiotherapy, regardless the test results.