Background and objectives: New potential prognostic indicators aside from the TNM classification have been proposed. The aim of this study was to analyze the prognostic relevance of tumor length as well as number of involved regional lymph nodes (LNM) in patients with esophageal carcinoma.
Methods: Two hundred thirteen patients with esophageal carcinoma (116 squamous cell- and 97 adenocarcinoma) were included in this study. Treatment of choice was subtotal en bloc esophagectomy including "2-field" lymphadenectomy. The median number of examined lymph nodes (LNs) was 28. Eighty patients (38%) received preoperative radio-chemotherapy according to a standardized protocol. Histopathology consisted of tumor stage, residual tumor, grading, and number of examined and involved LN. Univariate and multivariate prognostic values were calculated.
Results: Length of tumor correlated with pT/ypT-category (P<0.01). Univariate but not multivariate analysis showed better survival for tumors<or=3 cm (P<0.05). Patients with 1-5 LNM had significantly better prognoses than those with more than 5 LNM (Hazard ratio 2.7, 95% CI=1.7-4.2) (P<0.01). Patients without LNM and more than 15 examined LN showed significantly better prognosis than those with fewer examined LN (Hazard ratio=0.3, 95% CI=0.1-0.6) (P<0.01).
Conclusions: A revision of the TNM classification for esophageal carcinoma should subdivide the pN1-category according to the number of LNM.
Copyright (c) 2006 Wiley-Liss, Inc.