Multivariate Prognostic Analysis And MicroRNA Profiles In Rectal Adenocarcinoma

Purpose:Rectal cancer has a poor prognosis and high incidence. Rectal and colon cancer were often analyzed as one disease, even their treatment and prognosis were different. Controversies still remain regarding the prognostic factors as well as the role of adjuvant therapy in rectal cancer. So we choose the appropriate cases to evaluate the effect of adjuvant chemotherapy. In the mean time, using univariate and multivariate analyses to identify the prognostic value of the clinic pathological factors of rectal cancer, we expect to find some undocumented prognostic factors and create the individual risk assessment model for predicting the prognosis of the patients with rectal cancer.MicroRNAs (miRNAs) are short ribonucleic acid (RNA) molecules, on average only 22 nucleotides long. miRNAs are post-transcriptional regulators that bind to complementary sequences on target messenger RNA transcripts (mRNAs), usually resulting in translational repression and gene silencing. Several miRNAs have been found to have links with some types of cancer including colorectal adenocarcinoma. However, the miRNA expression profiles in rectal cancer are not examined till now in our country. The second objective of this study is to compare and contrast miRNAs expression between rectal tumors and normal tissues, and to identify some of the genes possibly associated with development of rectal cancer.Material and method:Between January 2001 and December 2005, a total of 159 patients with rectal adenocarcinoma, who received adjuvant chemotherapy after surgery in the general hospital of PLA, were included in the studies. The data concerning surgical treatment, pathological factors, regimens of chemotherapy and outcome were collected. The univariate and multivariate analyses were applied to evaluate the prognostic factors.1-year,3-year and 5-year survival rates were calculated by Kaplan-Meier curve method, univariate analysis was done through Log-rank and multiple factors comparison through Cox regression analysis. All tests were performed at the 0.05 level of significances. Statistical analysis was completed using the PASW Statistics 18.0 statistic software.In part two of this research, we studied 2 rectal adenocarcinomas,2 rectal adenomas, and 2 rectal normal tissue samples to examine miRNA expression profiles through miRNA microarray. We also predicted the possible targets of miRNA by bioinformatics.Result:159 patients with rectal cancer were registered in this study. The ratio of male to female is 1.04:1. The median follow-up time of the patients was 50.0 months. The distribution of the age was 53.24±11.14ys. The survival rate for 1, 3,5years was 87.2%,63.9%,55.0% respectively. The 5-year survival rates of patients in pathologic TNM (pTNM) stage I, II, III and IV were 100.0%,80.4%,41.2%,15.3%, respectively (P<0.01). Of all 7 regimens,5-FU was most widely used. The regimens did not show any significant association with survival rate (p=0.589). The survival rate for 1,3,5years was 92.6%,77.9%,65.3% when continuous intravenous infusion of 5-FU was used in the regimen, and 85.2%,57.0%,51.1% respectively in the group with traditional infusion of 5-FU. The survival rate for 5years was 56.3% when the regimens contain cisplatin,53.3% with oxaliplatin, and 51.9% without them. From the univariate analysis, preoperative course, weight loss, blood carcino-embryonic antigen level, macroscopic appearance, histopathological classification, depth of invasion, lymph node metastasis, remote metastasis, lymphovascular invasion, operation modes, perioperative blood transfusion, were found to be significantly associated with the cumulative survival rate. Of the variables selected by the univariate analysis, only seven (remote metastasis, histopathological classification, weight loss, lymph node metastasis, lymphovascular invasion, perioperative blood transfusion and macroscopic appearance) were recognized as significant prognostic factors by the multivariate analysis.By using miRNAs microarray,38 miRNAs were found to be differentially expressed between rectal adenomas and normal tissue, with 15 miRNAs upregulated while 23 downregulated.37 miRNAs were upregulated and 18 downregulated in rectal adenocarcinomas when comparing with normal issues.42 miRNAs, including 34 upregulated and 8 downregulated, were found to be differentially expressed between rectal adenocarcinomas and rectal adenomas.5 miRNAs (miR-18a*, miR-224, miR-503, miR-424*, miR-31) were gradually upregulated in the procession of the initiation and development of oncogenesis for rectal cancer. Bioinformatics indicates that SMAD2 may be the targets of miR-18a*, and that BACH2 may be the target of miR-31.Conclusion:1. No regimen is definitely prominent for overall survival rate in the 7 regimens in treating rectal cancer; while continuous intravenous infusion of 5-fluorouracil or regimens with cisplatin/oxaliplatin may be more effective.2. Remote metastasis, histopathological classification, weight loss, lymph node metastasis, lymphovascular invasion, perioperative blood transfusion and macroscopic appearance have been recognized as significant prognostic factors. Lymph node ratios may aid in predicting prognosis when fewer lymph nodes are harvested after rectal cancer surgery.3. We founded the equation for risk factors for prognosis originally: [h(t,xi)]=ho(t)exp(1.713x1+0.261x2+1.202x3+0.647x4+1.153x5+0.612x6+0.378x7)4. There is a specific miRNA profile in rectal adenocarcinoma.5. miRNA pathways play an role in genesis and development in rectal adnocarcinoma. 6. SMAD2 may be the targets of miR-18a*; and BACH2 may be the target of miR-31.