Comparison of four lymph node staging systems for predicting prognosis for stage IV rectum cancer

Background: With recommendation of surgical management in primary site, both the positive and negative lymph nodes (LNs) retrieved have been emphasized to predict prognosis in stage IV rectum cancer. Therefore, we attempt to compare the prognostic performance of American Joint Committee on Cancer (AJCC) N-stage relative to lymph node ratio (LNR), log odds of metastatic lymph nodes (LODDS), and N-score in stage IV rectal cancer. Methods: Total 5,090 patients taken surgical resection of primary site in rectum cancer with distant metastasis were extracted from Surveillance, Epidemiology, and End Results Program (SEER) database. Harrell's C statistic (C-index) and Akaike's Information Criterion (AIC) were used to evaluate the discriminative ability of the different LN staging systems. Results: Of the 3,243 patients without radiotherapy, 82.46% (n=2,675) had been found with lymph nodes metastasis with median number of 16 lymph nodes collected (IQR: 11-22). Modeled as categorical cutoff variables for further clinical usage, when number of LNs was between 12 and 25 (C-index: 0.5997, AIC: 1,698.015), 8(th) AJCC N-stage outperformed other three schemas with increasing C-index and less AIC value. Assessed as continuous values, the LODDS shown as the best schemas with greatest discriminatory power (C-index: 0.5971, AIC: 3,680.017), generally. On the other hand, in the cohorts of other 1274 patients taken radiation, the median number of lymph nodes retrieved was 13 (IQR: 9-18). LODDS still remained remarkable performance as continuous (C-index: 0.5912; AIC: 1,058.765) and categorical variables (C-index: 0.5700; AIC: 1,061.703), while N-staging outperformed with less than 25 lymph nodes retrieved (LNs <12 C-index: 0.5678, AIC: 481.94; 12< LNs <25 C-index: 0.5933, AIC: 390.395). Conclusions: When assessed as categorical variables, N-stage performed superiorly with adequate lymph nodes examined, whether the patients have got radiotherapy prior to surgery or not. LODDS showed, when assessed as a continuous variable, good discriminative ability and goodness of fit in predicting survival for stage IV rectum cancer patients regardless of radiation therapy status.