Objective: To compare the prognostic performance of American Joint Committee on Cancer/International Union Against Cancer seventh N stage relative to lymph node ratio (LNR), log odds of metastatic lymph nodes (LODDS), and N score in gastric adenocarcinoma.
Background: Metastatic disease to the regional LN basin is a strong predictor of worse long-term outcome following curative intent resection of gastric adenocarcinoma.
Methods: A total of 804 patients who underwent surgical resection of gastric adenocarcinoma were identified from a multi-institutional database. The relative discriminative abilities of the different LN staging/scoring systems were assessed using the Akaike's Information Criterion (AIC) and the Harrell's concordance index (c statistic).
Results: Of the 804 patients, 333 (41.4%) had no lymph node metastasis, whereas 471 (58.6%) had lymph node metastasis. Patients with ≥N1 disease had an increased risk of death (hazards ratio = 2.09, 95% confidence interval: 1.68-2.61; P < 0.001]. When assessed using categorical cutoff values, LNR had a somewhat better prognostic performance (C index: 0.630; AIC: 4321.9) than the American Joint Committee on Cancer seventh edition (C index: 0.615; AIC: 4341.9), LODDS (C index: 0.615; AIC: 4323.4), or N score (C index: 0.620; AIC: 4324.6). When LN status was modeled as a continuous variable, the LODDS staging system (C index: 0.636; AIC: 4304.0) outperformed other staging/scoring systems including the N score (C index: 0.632; AIC: 4308.4) and LNR (C index: 0.631; AIC: 4225.8). Among patients with LNR scores of 0 or 1, there was a residual heterogeneity of outcomes that was better stratified and characterized by the LODDS.
Conclusions: When assessed as a categorical variable, LNR was the most powerful manner to stratify patients on the basis of LN status. LODDS was a better predicator of survival when LN status was modeled as a continuous variable, especially among those patients with either very low or high LNR.