Construction Of A Novel Colorectal Cancer Staging System Based On The Log Odds Of Positive Lymph Nodes And Its Clinical Application

PurposeThe TNM staging system is currently the most widely used staging system,and is widely used to predict the prognosis of colorectal cancer(CRC)patients,as well as the basis for guiding postoperative adjuvant chemotherapy.It has been found that there is a"survival paradox" in TNM staging that is seriously inconsistent with clinical reality,and most studies have shown that this is related to the unreasonable traditional N-staging,which is susceptible to "staging migration".The main objective of this study is to compare the traditional N-staging,Lymph node ratio(LNR)and log odds of positive lymph nodes(LODDS)staging,to find a more stable and reliable lymph node staging,and to construct a new lymph node staging based on the new lymph node staging with T-staging and M-staging,A new staging system for colorectal cancer was constructed based on the new lymph node staging,T-staging and M-staging,to resolve the survival paradox in the staging system and to weaken the influence of staging migration on tumour staging,so as to achieve accurate prognostic assessment of CRC patients.MethodsA training set of 71,927 CRC patients with a definite pathological diagnosis and treated with radical surgery between 2010 and 2015 was selected from the SEER database.The K-M(Kaplan-Meier)method was used to verify the existence of the"survival paradox" in TNM staging.To elucidate how LODDS can improve the accuracy of lymph node status assessment,scatter plots were used to depict the relationship between LODDS values and the distribution of LNR values and number of positive lymph nodes.A Classification and Regression Tree(CART)was used to establish the cut-off values of LODDS and LNR and convert them into categorical variables to form the LODDS staging and LNR staging.COX univariate and multivariate analyses were applied to examine the independent predictive ability of LODDS staging,LNR staging and conventional N staging on the prognosis of CRC patients.To compare the ability of LODDS staging with LNR staging and conventional N staging in identifying the prognostic risk,resistance to staging migration and statistical predictive power of CRC patients without lymph node metastases,where the difference in prognosis of CRC patients without lymph node metastases and resistance to staging migration were identified using the Log-Rank test and the statistical power was measured using the Akaike The Akaike information criterion(AIC)and the Harrell’s concordance index(C-index)were used for statistical power.A better lymph node staging was selected to replace the traditional N-staging and to form a new tumour staging system with T-staging and M-staging.The C-index,Area under Curve of Receiver Operating Characteristic(AUC)and AIC were used to compare the prediction accuracy and goodness of fit of the new tumour staging with TNM staging in the training set.929 patients who were first diagnosed and underwent surgery between 2012 and 2016 in Subei People’s Hospital affiliated to Yangzhou University were used as the external validation set to verify the goodness of fit and predictive stability of the new staging system.ResultIn this study,a clear survival paradox was observed in the training set under TNM staging,where the 5-year survival rate of patients with stage ⅢA was better than that of stage ⅡB and ⅡC,and even better than that of stage ⅡA;the 5-year survival rate of stageⅢB was also better than that of stage ⅡB and ⅡC,and this phenomenon also existed in the population with at least 12 lymph nodes cleared,again demonstrating the shortcomings of the traditional TNM staging system.From the scatter plot depicting the distribution of LODDS values,LNR values,and the number of lymph node metastases,it can be concluded that LODDS values are closely related to the number of lymph node metastases,and the distribution of LNR values is close to that of LODDS values,but when LNR values are 0 or 1,the distribution of LODDS values is more different,indicating that there is a clear prognosis of LODDS compared to LNR and positive lymph node counts in differentiating CRC patients advantage.The LODDS and LNR were transformed into five categorical variables based on the cut-off values obtained from the classification and regression tree model,and are referred to as LODDS staging and LNR staging.LODDS staging is superior to LNR staging and conventional N staging in terms of differentiating patients without lymph node metastases,combating"staging migration" and predictive efficacy.A new TLM staging system was constructed using LODDS staging instead of traditional N staging to benchmark against TNM staging.The study shows that the "survival paradox" is no longer present in the new TLM staging system and that LODDS staging is more resistant to "staging migration",enhancing the predictive performance of patients with inadequate lymph node dissection.Compared with the conventional TNM staging model,the TLM staging model has a lower AIC value,higher C-index and better AUC in the training set,indicating that the accuracy and goodness of fit of the TLM are better than the conventional TNM staging,and this advantage is more prominent in patients with inadequate lymph node clearance.External validation has shown that the new TLM staging system has a better fit and is no less accurate than TNM staging,and that the system has good stability.Conclusion1.LODDS staging is a valid tool for assessing the status of lymph nodes,identifying the prognostic risk of patients without lymph node metastases and providing a more comprehensive assessment of the prognosis of CRC patients;the three lymph node stages(traditional N stage,LNR stage and LODDS stage)are affected by the phenomenon of "staging migration" to the extent that N stage>LNR stage>LODDS stage,indicating that LODDS stage has a significant advantage when lymph node clearance is inadequate.LNR staging>LODDS staging,indicating that LODDS staging has a significant advantage when lymph node dissection is insufficient.2.The new TLM staging system not only resolves the survival paradox,but also has better predictive performance than the 8th version of the TNM staging system for CRC patients,especially for those with inadequate lymph node clearance and negative lymph nodes.3.External validation shows that the TLM staging system has good stability...