Background: The prognosis of patients with localized gastroesophageal carcinoma (LGC) can be defined after chemoradiation by the standardized uptake value (SUV) of positron emission tomography (PET). High SUV (HSUV) after chemoradiation portends a poor prognosis. The authors retrospectively examined the role of surgery in patients with HSUV after chemoradiation.
Methods: The authors analyzed the postchemoradiation PET scans of 204 LGC patients. One hundred twenty-nine patients had HSUV. Two postchemoradiation variables were evaluated: SUV and surgery and their association with overall survival (OS) and event-free survival (EFS). The log-rank test, multivariate Cox proportional hazards model, and Kaplan-Meier survival plots were used to assess the association between OS or EFS and the dichotomized SUV (using the median SUV as the cutoff) and surgery.
Results: The median SUV was 4.6. The OS of the 52 patients who had an SUV above the median and did not undergo surgery (HSUV-NS) (median OS, 1.22 years; 95% confidence interval [95% CI], 1.02-2.16 years) was much shorter than that of the 77 patients with an SUV above the median who underwent surgery (HSUV-S) (median OS, 2.7 years; 95% CI, 2.43 years to not reached [P<.0001]). Similarly, the EFS for patients with HSUV-NS was significantly shorter than that for patients with HSUV-S (P=.001). In the multivariate analyses, patients who underwent surgery (irrespective of SUV) had a lower risk of death (P=.0001) and disease progression (P=.002).
Conclusions: The data from the current study suggest that surgery may prolong OS and EFS in patients with a poor prognosis after chemoradiation as defined by PET. However, these data need confirmation.
Copyright © 2010 American Cancer Society.