659P - The log odds of positive lymph nodes is an independent prognostic factor for gastric cancer recurrence and mortality after curative resection

Date 29 September 2014
Event ESMO 2014
Session Poster Display session
Topics Gastric Cancer
Pathology/Molecular Biology
Basic Scientific Principles
Presenter Fatma Paksoy Türköz
Citation Annals of Oncology (2014) 25 (suppl_4): iv210-iv253. 10.1093/annonc/mdu334
Authors F. Paksoy Türköz1, M. Solak2, S. Buyukbas3, B. Oksuzoglu3, S. Yalcin4
  • 1Medical Oncology, Ataturk Chest Disease and Chest Surgery Training and Research Hospital, 06280 - ANKARA/TR
  • 2Medical Oncology, Hacettepe University Oncology Institutte, Ankara/TR
  • 3Department Of Medical Oncology, Ankara Dr.A.Y.Oncology Research and Education Hospital, 06200 - ANKARA/TR
  • 4Medical Oncology, Hacettepe University Cancer Institute, 06100 - ANKARA/TR



To evaluate the prognostic value of the log odds of positive lymph nodes (LODDS) system and lymph node ratio (LNR) compared with pathological nodal (pN; 7th edition UICC/ AJCC) stage in gastric cancer (GC) patients after curative resection.


A total of 513 GC patients who underwent radical gastrectomy (R0) from Jan 2005 to Dec 2013 were analyzed. LNR was defined as the ratio of metastatic lymph nodes (LN) to total LN. LODDS was calculated as log[(pnod + 0.5)/ (tnod-pnod + 0.5)], where the "pnod" is the number of positive LN and "tnod" is the total number of LN retrieved. The primary endpoint of the study was overall survival. The patients were classified into four LNR (LNR0: 0; LNR1: > 0 to 0.2; LNR2: > 0.2 to 0.5 and LNR3: > 0.50 to 1.00) and four LODDS (LODDS1: -2 to ≤ -1, LODDS2: -1 to ≤ 0, LODDS3: 0 to ≤ +1 and LODDS4: +1 to ≤ +2) groups by the best cut-off approach. Kaplan-Meier analysis with log-rank test and Cox regression model were performed.


The median numbers of involved and dissected LN were 5 (0-50), and and 21 (2-91), respectively. The median follow-up period was 36.7 months. In multivariate analysis, LODDS system was independent prognostic factor for GC mortality (HR:2.02, 95%CI 1.61-2.54; p< 0.0001) and recurrence (HR:2.02 95%CI 1.47-2.77; p< 0.0001). LODDS system significantly discriminated the prognostic groups in patients with 6 to ≤ 10 LN retrieved (plog-rank = 0.022) and 11 to < 15 LN retrieved (plog-rank = 0.014) whereas pN or LNR classifications did not. In patients with ≤ 5 LN retrieved, the prediction of prognosis was insignificant for all 3 staging systems (plog-rank> 0.05). LODDS system also significantly discriminated the prognostic groups both in the patients with D1 and D2 lymphadenectomy, as well as in patients who received adjuvant chemotherapy and radiotherapy.


LODDS system is an independent prognostic factor for GC patients. It has more potential to minimize stage migration in patients with insufficient number of LN retrieved compared to the pN and LNR classifications. Although limited practical application due to calculation procedure, LODDS system should be recommended in the countries where both D1 and D2 lymphadenectomy are performed. Moreover, it would impact the selection of patients who may benefit most from the adjuvant treatments.


All authors have declared no conflicts of interest.