Responsabile: Dr. M. Catarci
PROTOCOLLO ROMA 3
(Responsabile: Dr. Catarci)
PROGNOSTIC FACTORS AFTER R0 SURGERY FOR GASTRIC CANCER
In order to overcome the problem of stage migration induced by extended lymph node dissection [1, 2], during the last years a new independent prognostic factor was largely investigated and validated: the lymph node ratio (LNR), defined as the absolute ratio between metastatic and examined lymph nodes, appeared to be a strong independent indicator of prognosis [3-7]. There still is some controversy regarding its prognostic role in case of inadequate nodal staging (< 15 examined lymph nodes), with some studies supporting it [8, 9] and others denying it [3, 10]. The IRGGC defined four intervals for LNR: 0; 0.01-0.09; 0.10-0.25; > 0.25. These intervals corresponded to a definite different prognosis even within the same pN status, independently from the extent of nodal dissection .
It was already demonstrated that LNR is independently influenced by the number of metastatic lymph nodes but not by the number of examined lymph nodes [5, 11, 12], being therefore suitable as a prognostic factor independent from the stage-migration phenomenon.
As a consequence, a Tumor-Ratio-Metastasis (TRM) instead of a Tumor-Node-Metastasis (TNM) staging system was suggested .
In the recent new 7th Edition of the TNM staging system , the UICC elected to retain the pN classification based on the number of metastatic nodes, modifying pN grouping with a new pN1 group with 1 or 2 metastatic lymph nodes, and to introduce a different stage grouping. It seems that this new pN stage-grouping is more accurate as a prognostic predictor than its precursor . It will be very interesting, therefore, to investigate the prognostic yeld of this new TNM stage-grouping faced to LNR.
The main issue in this analysis is statistical in nature, as the two variables (pN grouping and IRGGC-LNR categories) are highly co-linear and automated Cox’s proportional hazards model  building in multivariate analysis can generate controversial results .
A retrospective analysis of prospective database from centers of the IRGGC -- 11 years of accrual (1998-2008 included). Minimum follow-up 35 months with censor at November 30, 2011.
Inclusion criteria: all the R0 resections for “pure” gastric cancer (Siewert II-III excluded), preferentially treated by D1, D1alfa-beta, D2 and D3 lymph node dissection according to the JGCA  rules. All cases submitted to be verified for the number of examined lymph nodes according to the IRGGC standards .
Endpoints: Disease-free survival and disease recurrence rates calculated at 5 and at 10 years from the operation by the Kaplan Meier method . Univariate analysis performed applying the Log-rank test . Multiple variate analysis applying Cox’s proportional hazards model  with proportion of explained variation (PEV) and bootstrap analysis [22, 23]. The proportion of explained variation (PEV) analysis [24, 25] will be adopted in our study to compare the relative importance of LNR or pN classifications over the other prognostic factors statistically using a bootstrap technique (Bootstrap N=1500). In addition, D-measures for prognostic separation [26,27] will be computed for both LNR and pN. The bootstrap 95% CI for the difference (Bootstrap N=1500) will be computed to further compare the prognostic power between LNR classification and pN classification.