Prognostic Role of Log Odds of Lymph Nodes After Resection of Pancreatic Head Cancer

• Published in: Journal of gastrointestinal surgery Vol. 20; no. 10; pp. 1707 - 1715
• Main Authors: Riediger, Hartwig, Kulemann, Birte, Wittel, Uwe, Adam, Ulrich, Sick, Olivia, Neeff, Hannes, Höppner, Jens, Hopt, Ulrich T., Makowiec, Frank
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.10.2016; Springer Nature B.V
• Subjects: Adult; Aged; Cancer therapies; Female; Gastroenterology; Humans; Lymph Nodes - pathology; Lymphatic system; Male; Medical prognosis; Medicine; Medicine & Public Health; Middle Aged; Multivariate Analysis; Neoplasm Staging; Original Article; Pancreas - pathology; Pancreas - surgery; Pancreatic cancer; Pancreatic Neoplasms - mortality; Pancreatic Neoplasms - pathology; Pancreatic Neoplasms - surgery; Prognosis; Surgeons; Surgery; Survival Analysis; Survival Rate; Veins & arteries; Berlin Germany; Germany; LogODDS lymph nodes; Resection; Survival; Pancreatic cancer; Prognostic factors; Lymph node ratio
• ISSN: 1091-255X; 1873-4626
• DOI: 10.1007/s11605-016-3200-5

Introduction Nodal status is a strong prognostic factor after resection of pancreatic cancer. The lymph node ratio (LNR) has been shown to be superior to the pN status in several studies. The role of log odds of the ratio between positive and negative nodes (LODDS) as a suggested new indicator of prognosis, however, has been hardly evaluated in pancreatic cancer. Methods Prognostic factors for overall survival after resection for cancer of the pancreatic head were evaluated in 409 patients from two institutions (prospectively maintained databases). The lymph node status, LNR, and LODDS were separately analyzed and independently compared in multivariate survival analysis. Results The median numbers of examined and positive lymph nodes were 16 and 2, respectively. Actuarial 3- and 5-year survival rates were 29 and 16 %. All three classifications of nodal disease significantly predicted survival in the entire group ( n  = 409), in patients with free resection margins ( n  = 297), and in patients with <12 examined nodes. In multivariate analysis, however, both LNR and LODDS were equally superior to the nodal status. In node-negative patients ( n  = 110), LODDS could not identify subgroups with different prognosis. Conclusion Both LNR and LODDS are superior to the classical nodal status in predicting prognosis in resected pancreatic cancer. However, LODDS has not shown any advantage over LNR in our series, neither in the entire patient group nor in the subgroups with free margins, negative nodes or a low number of examined nodes. Therefore, the use of LODDS to predict the outcome after resection of pancreatic head cancer cannot be recommended.