Background: Surgical interventions for cervical esophageal cancer (CEC), particularly larynx-preserving procedures, have not yet been standardized in terms of short- and long-term outcome.
Methods: We retrospectively analyzed 100 consecutive surgeries for CEC in our department. We compared clinicopathological parameters and long-term outcomes between larynx-preserving esophagectomies (LP group) and nonpreserving procedures (NP group). We also evaluated preoperative predictive parameters for larynx-preservation.
Results: Compared with the NP group, the LP group had significantly lower cT (P < 0.001) and cStage (P = 0.001) and shorter tumor length (P = 0.0108). Multivariate logistic regression analysis identified early cT stage, early cStage, and response to preoperative treatment as significant predictive parameters of larynx preservation. Larynx-preserving procedures could be performed for 90.5 % of T1-2 tumors regardless of preoperative treatment response. In contrast in T3-4 tumors, most nonresponders (92.3 %) were ineligible for larynx-preservation (P = 0.0012), whereas 54.3 % of responders could achieve larynx preservation. The average shortening of upward extension in T3-4 tumors after preoperative treatment was 20.0 mm in the LP group vs. 10.2 mm in the NP group (P = 0.051). The two groups were similar in terms of overall morbidity (including pneumonia), mortality, and postoperative hospital days. Importantly, larynx preservation for CEC neither worsened patient prognosis nor increased locoregional recurrence compared to the NP group.
Conclusions: Larynx-preserving esophagectomy for CEC is feasible and oncologically acceptable. The cT, cStage, and clinical response to preoperative treatment are important preoperative predictors of a patient's suitability for larynx-preservation.