Comparison of two novel staging systems with the TNM system in predicting stage III colon cancer survival

Real-world survival of colon cancer after radical surgery: A single-institutional retrospective analysis

The survival rate for colon cancer after radical surgery has been the focus of extensive debate. To assess the postoperative survival and prognostic factors for overall survival (OS), we collected clinicopathological information for 2,655 patients. The survival time and potential risk factors for OS were analyzed by using Kaplan–Meier curves, Cox proportional hazards models, best subset regression (BSR), and least absolute shrinkage and selection operator (LASSO). The 5-year survival rates of stage I–IV colon cancer were 96.6%, 88.7%, 69.9%, and 34.3%, respectively. Adjuvant chemotherapy improved the survival rate (90.4% vs. 82.4%, with versus without adjuvant chemotherapy, respectively) in stage II patients with high-risk factors. Elevated preoperative carcinoembryonic antigen (CEA) and carbohydrate antigen 19-9 (CA19-9) were significantly associated with worse OS compared with patients without these elevations. Less than 12 versus more than 12 harvested lymph nodes (LNs) affected prognosis (84.6% vs. 89.7%, respectively). Regarding the lymph node ratio (LNR), the 5-year OS rate was 89.2%, 71.5%, 55.8%, and 34.5% in patients with LNR values of 0, 0.3, 0.3–0.7, and >0.7, respectively. We constructed a nomogram comprising the independent factors associated with survival to better predict prognosis. On the basis of these findings, we propose that stage II colon cancer patients without high-risk factors and with both elevated preoperative CEA and CA199 should receive adjuvant therapy. Furthermore, the LNR could complement TNM staging in patients with <12 harvested LNs. Our nomogram might be useful as a new prognosis prediction system for colon cancer patients.

Deep-learning model for predicting the survival of rectal adenocarcinoma patients based on a surveillance, epidemiology, and end results analysis

Background

We collected information on patients with rectal adenocarcinoma in the United States from the Surveillance, Epidemiology, and EndResults (SEER) database. We used this information to establish a model that combined deep learning with a multilayer neural network (the DeepSurv model) for predicting the survival rate of patients with rectal adenocarcinoma.

Methods

We collected patients with rectal adenocarcinoma in the United States and older than 20 yearswho had been added to the SEER database from 2004 to 2015. We divided these patients into training and test cohortsat a ratio of 7:3. The training cohort was used to develop a seven-layer neural network based on the analysis method established by Katzman and colleagues to construct a DeepSurv prediction model. We then used the C-index and calibration plots to evaluate the prediction performance of the DeepSurv model.

Results

The 49,275 patients with rectal adenocarcinoma included in the study were randomly divided into the training cohort (70%, n = 34,492) and the test cohort (30%, n = 14,783). There were no statistically significant differences in clinical characteristics between the two cohorts (p > 0.05). We applied Cox proportional-hazards regression to the data in the training cohort, which showed that age, sex, marital status, tumor grade, surgery status, and chemotherapy status were significant factors influencing survival (p < 0.05). Using the training cohort to construct the DeepSurv model resulted in a C-index of the model of 0.824, while using the test cohort to verify the DeepSurv model yielded a C-index of 0.821. Thesevalues show that the prediction effect of the DeepSurv model for the test-cohort patients was highly consistent with the prediction resultsfor the training-cohort patients.

Conclusion

The DeepSurv prediction model of the seven-layer neural network that we have established can accurately predict the survival rateand time of rectal adenocarcinoma patients.

Comparison of Different Lymph Node Staging Systems in Patients With Resectable Colorectal Cancer

Background and Objectives: Currently, the United States Joint Commission on Cancer (AJCC) N staging, lymph node positive rate (LNR), and log odds of positive lymph nodes (LODDS) are the main lymph node (LN) staging systems. However, the type of LN staging system that is more accurate in terms of prognostic performance remains controversial. We compared the prognostic accuracy of the three staging systems in patients with CRC and determine the best choice for clinical applications. Methods: From the Surveillance, Epidemiology, and End Results (SEER) database, 56,747 patients were identified who were diagnosed with CRC between 2004 and 2013. Akaike's Information Criterion (AIC) and Harrell's Consistency Index (c-index) were used to assess the relative discriminative abilities of different LN staging systems. Results: In 56,747 patients, when using classification cut-off values for evaluation, the LNR of Rosenberg et al. showed significantly better predictive power, especially when the number of dissected lymph nodes (NDLN) were insufficient. When analyzed as a continuous variable, the LODDS staging system performed the best and was not affected by the NDLN. Conclusions: We suggest that the LNR of Rosenberg et al. should be introduced into the AJCC system as a supplement when the NDLN is insufficient until the optimal LODDS cut-off values are calculated.

The Prognostic Impact of the Metastatic Lymph Nodes Ratio in Colorectal Cancer

Background: This study was designed to validate the prognostic significance of the ratio of positive to examined lymph nodes (LNR) in patients with colorectal cancer. Methods: 218,314 patients from the SEER database and 1,811 patients from the three independent multicenter were included in this study. The patients were divided into 5 groups on a basis of previous published LNR: LNR0, patients with no metastatic lymph nodes; LNR1, patients with the LNR between 0.1 and 0.17; LNR2, patients with the LNR between 0.18 and 0.41; LNR3, patients with the LNR between 0.42 and 0.69; LNR4, patients with the LNR >0.7. The 5-year OS and CSS rate were estimated using Kaplan-Meier method and the survival difference was tested using log-rank test. Multivariate Cox analysis was used to further assess the influence of the LNR on patients' outcome. Results: The 5-year OS rate of patients within LNR0 to LNR4 group was 71.2, 55.8, 39.3, 22.6, and 14.6%, respectively (p < 0.001) in the SEER database. While the 5-year OS rate of those with LNR0 to LNR4 was 75.2, 66.1, 48.0, 34.0, and 17.7%, respectively (p < 0.001) in the international multicenter cohort. In the multivariate analysis, LNR was demonstrated to be a strong prognostic factor in patients with < 12 and ≥12 metastatic lymph nodes. Furthermore, the LNR had a similar impact on the patients' prognosis in colon cancer and rectal cancer. Conclusion: The LNR allowed better prognostic stratification than the positive node (pN) in patients with colorectal cancer and the cut-off values were well validated.