Study On The Model Of Pathological Staging Prediction For Elderly Patients With Cancerous Acute Abdomen Based On The 1H-NMR Metabonomics

Objective: using 1h-NMR metabolomics technology and GO technology,study sex in different periods of patients with acute abdominal pain and pathological features of thyroid carcinoma in elderly,found that older cancer patients with an acute metabolic network disturbance mechanism,establish a preoperative can cancerous accurate staging method of acute abdominal pain of the age.Methods: The use of cohort study to design,included in the Sichuan People's Hospital emergency surgical treatment of colorectal cancer caused by acute abdomen patients,the following is the inclusion criteria:(1)age 60 years and above;(2)Emergency to hospitalized with acute abdomen,the clinical diagnosis of colorectal cancer patients(3)patients and their families agree to participate in this study,and signed to confirm;the following is the exclusion criteria:(1)diabetes,andother endocrine problems(2)the merger of other tumors;(3)is in pregnancy or lactation of women;(4)mental state can not effectively cooperate,There is a certain mental problem,the lack of good self-control and efficient expression of patients;(5)actually added to the relevant clinical analysis..According to TNM staging method in different periods of patients can be divided into four groups I,II,III and IV.Respectively in the group of blood on the day after a(about 2 ml)blood pumping,high speed centrifugal,extract its plasma for metabolomic analysis(1h-nmr).Using Bruker Avvance DRx600 Fourier transform nuclear magnetic resonance instrument to detect serum metabolites,and get the original one dimensional NMR spectra.For metabolomics data,first of all to Fourier transform,and then use the MsetReNova software will automatically integral spectra is decomposed into 199 area;And the actual data from within the platform to carry out by means of MATLAB software supporting operation to check the partial least squares(Orthogonal Signal Correction Partial further Squares-Discriminant Analysis OPLS-DA),set up data matrix,the data is centralized and normalized processing after the Orthogonal Signal Correction(OSC).And based on gene ontology(gene ontology,GO)method to establish the metabolic type--GO pathway back analysis technique to establish metabolic enzyme network model,and then through the principal component analysis and partial least squares and support vector data analysis.All the data using Matlab software to deal with.Calculated data using Wilcoxon rank and inspection(non-normal distribution analysis)or analysis of variance(normal distribution)to determine the correlation between different factorsResults: 62 patients with different stages of selection of the age,weight,bloody no difference,according to the plasma nuclear magnetic resonance(NMR)map of different periods of elderly cancer patients with acute abdominal pain,the plasma resonance spectrum fingerprint characteristic to distinguish the different tumor stage.In patients with different TNM staging,the metabolites of the plasma metabolic pool obvious differences,III,IV than in the period of I,II phase of the plasma of patients with a variety of amino acids.Exist within the blood of a large number of amino acids,has extremely significantly reduced,which covers the leucine,alanine,and many other aspects of the elements,the difference on the level of statistical significance(P< 0.05),while in the blood within the relevant materials such as lactic acid,choline and creatine is also extremely significantly lower(P< 0.05),but,within the associated plasma of patients with stage I kept a very low state of lipoprotein cholesterol,and materials such as acetone has certain promotion,at the same time,there exist difference of statistical value(P< 0.05).After calculation,finally screening leucine,valine,beta hydroxy butyric acid,glycine,eight key metabolic marker is 112 pairs of enzymes and genes.By use of the main metabolic markers GO technology were analyzed,and the fast preoperative clinical staging prediction based on this model,the prediction model of root mean square error is 5.2,R2= 0.47(P = 0.036),the fruit shows good prediction effect.Conclusion: based on 1h-NMR metabonomics technology and gene ontology established elderly cancer patients with the acute pathological staging prediction model is established,effectively in clinical staging in patients with preoperative elderly cancer sex acute abdomen.