| Objectives:Systematic review of MUC1 on the invasion of pancreatic ductal papillary mucinous adenoma of significance, in order to reveal the MUC1 expression in the infiltrative IPMNs help the choice of surgical procedure and prognosis.Methods:By the two researchers independently in the MEDLINE database, EMBASE databases, the Chinese Biomedical Literature Database (CBMdisc), China Journal Full Database (CNKI) and the Cochrane Controlled Trials Register database retrieval center. Retrieval immunohistochemistry to detect MUC1 in pancreatic ductal papillary mucinous adenoma test the'expression of literature, and references included . in the study secondary search. Literature published in time for 1989 to 2008-10-31. Development of literature and documents into the standard exclusion criteria. Extracted from the literature of MUC1 in invasive ductal papillary mucinous adenoma and non-invasive pancreatic ductal papillary mucinous adenoma of the expression of tables from the data to calculate the odds ratio (OR) or can be obtained directly from the literature value OR 95% confidence interval. According to the literature of the design type, sample size, selection bias, test the implementation process, if there is pathologic evidence, interpretation of test results are blind to. Literature by the two researchers blinded quality assessment, quality assessment scale development literature. Extraction into the research literature tables from the data, its entry into ReviewManager4.2 software, quantitative data synthesis, the first test for heterogeneity, if the heterogeneity test P>0.05, and when the 12≤50% select the fixed effect model; if heterogeneity test P<0.05, and I2>50% chose random effects model. Second, calculate the average effect of the merger OR values and 95% confidence interval (CI). The literature by removing poor quality before and after the conclusion of the differences and choose a different statistical model, the effect of the merger value of OR point estimates and 95% confidence interval (CI) of differences of sensitivity analysis. Meanwhile funnel plot is symmetric with stata 8.0 software linear regression method (Egger's test) and the Begg funnel plot to show whether the publication bias.Results: 1. The initial literature search was 326, by screening, supplementary and secondary search,15 articles met the inclusion criteria, including a literature because of its sensitivity=1-specificity, the theory is completely worthless, statistical software automatically deletes not included in statistical analysis, the final total of 14 articles included in the study, three of them Chinese, English,11, the cumulative infiltration groups (iIPMN) cases,234/cases of non-invasive groups (no-iIPMN) cases of 345 patients.2. According to the literature quality assessment scale (Table 1) evaluation criteria, the standard view from the inside, all the studies have established the gold standard (pathology of evidence), all of the gold standard diagnostic interpretation of the results without reference to MUC1, the MUC1 results interpretation can not suggest that in the absence of reference to clinical data or pathological diagnosis of the case to make. Only four studies clearly are two people read the piece, I research were not mentioned. All the studies were reviewed retrospectively. Which the study sample size 6 too few (<40). Information from the outside look, most of the studies did not report pancreatic cancer stage and grade. All studies failed to show a continuous choice case. All studies were described in detail immunohistochemical methods, and detailed reporting threshold to define, but the study between the use of reagents and equipment, which may vary between the threshold of setting the same value or method failure. Quality score inside and outside the aggregate score 5-7 points in between.3.14 literature by the heterogeneity test, Chi2=17.64, df=13, p=0.17, p>0.05, 12=26.3%,I2<50%, so the fixed effects model. The results showed that combined, effect of point estimates for the value of OR 4.88,95% confidence interval CI, 3.11-7.65,Z=6.91, p<0.00001, statistically significant, indicating MUC1 in invasive ductal papillary mucinous adenoma of the pancreas express significance than the non-invasive pancreatic ductal papillary mucinous adenoma obvious.4. Sensitivity analysis shows that this system of evaluation research literature combined with high-quality results compared to good stability. In addition, the funnel plot and rank correlation test and Egger Bgeg weighted regression models show that' this study less affected by publication bias. Conclusion:The results of the current research Meta-analysis, MUC1 expression in IPMNs can be used as an indicator of invasion, thus contributing to the pancreatic ductal papillary mucinous adenoma of the surgical options and prognosis.
|
PDF Full Text Request