Background: Neoadjuvant chemotherapy (NACT) targets lymph node metastasis (LN), as well as the primary tumor (PT) in esophageal squamous cell carcinomas (ESCC). (18)F-fluorodeoxyglucose-positron emission tomography (FDG-PET) reflects viable tumor volume and may be more useful for evaluating NACT responses than conventional radiography. Moreover, FDG-PET may elucidate the clinical significance of NACT responses for LN, which is not always identical to those for PT.
Patients and methods: We retrospectively investigated prognostic factors in 38 node-positive ESCC patients who had undergone NACT (5-fluorouracil, adriamycin, and cisplatin) and surgical resection. The NACT response was evaluated separately by both PET and computed tomography (CT) for each PT and LN.
Results: Although NACT effect for PT and LN was similar by PET evaluation (SUVmax reduction; average 70.58% vs 71.57%), they did not show significant correlation, revealing discordance for 13 (34.2%) patients when SUVmax reduction of more than 70% was classified as a PET responder. An opposite relationship existed in that the pre-NACT SUVmax of PT was significantly lower in PET responders than in PET non-responders (9.92 +/- 4.3 vs 12.96 +/- 3.8, P = .032), while that of LN tended to be higher in responders than in non-responders (5.70 +/- 3.2 vs 3.77 +/- 0.9, P = .072). Multivariate analysis identified the number of PET-positive LN (P = .018, HR = 5.464) and PET response for PT (P = .015, HR = 4.620) and for LN (P = .028, HR = 3.854) as independent prognostic predictors. The NACT response for PT or LN on CT evaluation was not a significant prognostic predictor.
Conclusion: PET is superior to CT for evaluating the NACT response from the viewpoint of survival analysis. The NACT response should be evaluated for both LN and PT because of their different behaviors during chemotherapy. Further studies of larger sample number should be conducted in the future.