| Part Ⅰ: Analysis of risk factors for colorectal cancer complicated with intestinal obstructionObjective: Intestinal obstruction complicated by colorectal cancer is an unfavorable prognostic factor for patients.In addition to tumor staging,it is unknown whether there are other influencing factors.We used propensity score matching(PSM)analysis to further explore the difference of colorectal cancer specific inflammatory indicators and tumor tissue intestinal microflora in colorectal cancer patients with intestinal obstruction and non-obstruction and the impact of related risk factors on the prognosis of patients after balancing tumor stage and other factors.Methods: A retrospective analysis of 1470 colorectal cancer patients in the database was performed,including 198 patients with colorectal cancer complicated with intestinal obstruction and 1272 patients with non-obstruction.First,the PSM grouping was performed according to the patient’s baseline data and tumor stage.Indicators related to inflammation were then selected,including: preoperative neutrophil-to-lymphocyte ratio(NLR),platelet-to-lymphocyte ratio(PLR),lymphocyte-monocyte ratio(LMR),prognostic nutritional index,systemic immune inflammatory index(SII),aspartate aminotransferase to lymphocyte ratio index(ALRI)score,model of end-stage liver disease(MELD)score,etc.were statistically analyzed,and factors related to patient obstruction and prognosis were screened out and integrated to construct CRC-specific Inflammatory factor score(CRC-specific inflammatory score,CPIS).Univariate and multivariate logistic regression was used to analyze the association between inflammatory indicators and the occurrence of intestinal obstruction in patients.Then,according to the results of 16 S rRNA sequencing,the diversity and composition of tumor tissue microbiota were compared between patients with intestinal obstruction(N=29)and non-obstruction(N=29).Kaplan-meier method was used to analyze the correlation between the different microbiota and the overall survival rate of patients.Results:(1): After PSM,328 patients were divided into two groups: colorectal complicated with intestinal obstruction and non-obstruction group.The results of COX multivariate regression analysis showed that age,number of positive lymph nodes and blood transfusion were independent prognostic factors for OS in patients with colorectal cancer complicated with intestinal obstruction.Age,perioperative blood transfusion,and primary anastomosis were independent prognostic factors for DFS in patients with colorectal cancer complicated with intestinal obstruction.The overall survival of patients in the non-obstructive group was significantly better than that in the obstructed group,but there was no significant difference in tumor recurrence rate(P=0.551)between the two groups.(2): Multivariate Logistic regression results showed that LMR(β=-1.576,OR=0.207,P<0.001),ALRI(β=0.018,OR=1.018,P=0.015)and prognostic nutritional index(β=-0.067,OR= 0.936,P=0.009)was a specific inflammation-related independent risk factor for intestinal obstruction in patients.Based on the β coefficient,a predictive model CPIS was generated: CPIS=-1.576× LMR-0.067 × prognostic nutrition index +0.018×ALRI.It was also an independent predictor of intestinal obstruction in patients.(3): The microbial community in the tumor tissue of CRC patients with intestinal obstruction was significantly more abundant than that of patients without intestinal obstruction,and there was a statistically significant difference in the alpha diversity index between the two groups of patients(P<0.01).β diversity showed that the microbes in the tumor tissues of the two groups were independently divided into two groups,and the difference was statistically significant(P<0.05).The K-M survival curve results showed that the highly expressed Firmicutes,the lowly expressed Proteobacteria and the high expressed Bacteroidetes in the alpha diversity were significantly associated with poor long-term OS in CRC patients.Conclusions:(1)In addition to clinical characteristics such as tumor stage,other prognostic factors are also important,including age,whether perioperative blood transfusion is performed,whether primary anastomosis is performed,tumor markers and other factors not included.(2)The poor prognostic factors of intestinal obstruction in colorectal cancer may be related to the inflammatory response,but the specific mechanism needs to be further studied.(3)Intestinal obstruction significantly affects the distribution of bacteria in tumor tissue of patients with colorectal cancer,and the expression levels of different bacteria are related to the long-term prognosis of patients.Part Ⅱ:Construction of a predictive model for the prognosis of patients with colorectal cancer complicated with intestinal obstructionObjective: The prognosis of patients with colorectal cancer complicated with intestinal obstruction(CCRC)is worse than that of non-obstructive patients.However,there has been little screening of prognostic factors and construction of individualized prediction models for OCRC patients.We aimed to screen the clinical,pathological,and laboratory factors that affect the prognosis of OCRC,and use these indicators to construct a nomogram model to predict the prognosis of OCRC patients and provide a reference for individualized treatment of patients.Methods: We retrospectively analyzed the data of 198 patients diagnosed with OCRC in the MPE team database of Wuhan Union Medical College Hospital.The patients were randomly divided into the training set and the test set,and the parameters of the two cohorts were compared.Lasso regression and SVM-RFE were used to screen the prognostic indicators of DFS and OS respectively,and the overlapped indicators screened by the two methods were included in the final model analysis.Then,the ability of relevant inflammatory indicators to predict patients’ survival was analyzed and evaluated by drawing Receiver operating characteristic(ROC)curves.The results were analyzed and a rosette map integrating clinicopathological parameters was developed.Results: Based on the screening results of Lasso regression and SVM-RFE,the final selection included: neuroinvasion,TNM stage,MCHC,CEA,LDH and CA199 six valuable factors to establish a predictive model for DFS.Time-dependent(TD)-receiver operating characteristic(ROC)analysis was used to assess the validity of nomograms in predicting DFS at 1,3,and 5 years.The predictive power of the survival nomogram as measured by the area under the ROC curve(AUC)in the training set was 0.856,0.847,and 0.824 when assessing recurrence-free survival at 1,3,and 5 years;0.753 and 0.645.For OS,seven valuable factors in Lasso regression and SVM-RFE including: neuroinvasion,TNM stage,MCHC,MPV,LDH,TBA and DBIL were selected to build the predictive model.TD-ROC analysis was used to estimate the efficiency of nomograms for predicting 1-,3-,and 5-year OS.The predictive power of the survival nomogram as measured by AUC was 0.841,0.857,and 0.862 in the training set when assessing 1-,3-,and 5-year survival,and 0.641,0.845,and 0.883 in the validation set,respectively.In OCRC patients,the calibration curves of the two rograms for predicting patient OS and DFS probabilities showed good agreement between prediction and observation in the test set and validation set,with no deviation from the reference line and high confidence.Conclusions: The factors affecting OS and DFS in OCRC patients included not only tumor-related TNM stage and Perineural invasion,but also other non-tumor factors such as MCHC,MPV,LDH,CEA,CA199,TBA and DBIL.The survival rolograph constructed by us based on Lasso regression and SVM-RFE screening for important features can intuitively display these results and have a better ability to distinguish the prognosis of OCRC patients. |
