Prognostic significance of positive circumferential resection margin post neoadjuvant chemotherapy in patients with esophageal or gastro-esophageal junction adenocarcinoma

2019 
Abstract Background The aim of the present study was to assess the prognosis of patients with esophageal or gastroesophageal junction (E/GEJ) adenocarcinoma extending beyond the muscularis propria layer (≥ypT3) and positive circumferential resection margin (CRM), post neoadjuvant chemotherapy. Methods 177 patients were retrospectively studied. The majority (94.9%) received ECX (epirubicin, cisplatin, capecitabine), and all had clear proximal/distal resection margins. CRM was defined as positive (CRM+) when it was directly infiltrated (infiltrated CRM) or when tumor cells were detected within 1 mm from CRM (close CRM) and as negative (CRM-) when tumor cells were found in a distance > 1 mm from CRM. Results CRM+ was found in 83 patients (46.9%). Of them, infiltrated CRM was recorded in 36 (20.3%) and close CRM in 47 (26.6%). Adjuvant chemotherapy was administered to 132 patients (74.6%). Lymphovascular invasion and primary site in the lower esophagus were independently associated with higher risk of CRM+. Patients with infiltrated CRM, compared to those with close CRM and those CRM-, had the shortest median time-to-relapse (11.4 vs. 15.6 vs. 22.1 months, respectively, p  = 0.005) and overall survival (18.7 vs. 23.1 vs. 38.8 months, respectively, p  = 0.001). However, CRM status was not an independent predictor of poor outcome. Symptomatic isolated locoregional recurrences were rare in both CRM+ and CRM-patients (4/56 [7.1%] vs. 5/52 [9.6%], p  = 0.736), as well as in infiltrated vs. non-infiltrated CRM (CRM- and close CRM) (0/26 [0%] vs. 9/82 [11.0%], p  = 0.110). Conclusion Although CRM status is associated with poor outcome, it does not represent an independent prognostic factor. The status of CRM did not significantly influence the pattern of cancer relapse.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    30
    References
    7
    Citations
    NaN
    KQI
    []