| Part Ⅰ Evaluation of the Potential Value of Serum Exosomal miRNA in the Differential Diagnosis of Benign and Malignant Pancreatic Diseases Purpose:Pancreatic ductal adenocarcinoma(PDAC)is a deadly disease with a low rate of surgical resection and low sensitivity to chemoradiotherapy.It is difficult to diagnose PD AC at an early stage,resulting in an unfavaroble prognosis.Therefore,it is urgent to find clinical tumor biomarkers which have high pooled sensitivity and specificity in PD AC diagnosis.Mucinous cystic neoplasms(MCN)and intraductal papillary mucinous neoplasms(IPMN)are potentially-malignant pancreatic neoplams,early diagnosis and treatment is essential for patients’prognosis.Exosomal miRNA could be a potential biomarker for PDAC,which has a good stability and is easy to be detected.This study investigated the potential value of exosomal miRNA in the differential diagnosis between benign and malignant pancreatic diseases.Materials and methods:A total of 166 serum samples were included into our study,which consisted of 71 pathologically proved PD AC samples,36 pathologically proved pancreatic potentially-malignant neoplasm(PPN)samples(17 mucinous cystic neoplasms,MCN;19 intraductal papillary mucinous neoplasms),and 59 control samples(34 healthy controls and 25 pathologically proved chronic mass-forming pancreatitis,CMFP).The 166 serum samples were divided into training set(13 PD AC,11 PPN,and 12 Ctrl),validation set 1(29 PD AC,25 PPN,and 22 Ctrl),and validation set 2(29 PDAC and 25 Ctrl).In the training set,exoEasy Maxi Kit was adopted for exosome extraction,and TEM,NTA and WB were adopted to identify the exosome.After total RNA extraction,RNA library construction,and quality control,the Illumina high throughput sequencing was used in PDAC,PPN,and Ctrl groups.Differentially expressed miRNA was analyzed with volcano maps,heatmaps,and venn diagrams.Targeted gene analysis,including GO biological function and module analysis,KEGG pathway enrichment,and miRNA-mRNA-KEGG pathway network diagram,were carried out.Finally,TaqMan probes were used to validate the differentially expressed miRNA in the training and validation sets.Results:1.We successfully extracted exosomes with exoEasy Maxi Kit.The TEM results showed that the exosomes were cup-like vesicle with the double lipid layer.The NTA results showed most of exosomes were distributed with a peak at size 110 nm.Moreover,they expressed ALIX,TSG101,CD63,but did not express pancreatic cell line specific protein Calnexin.2.During miRNA differential expression analysis,differentially expressed miRNA was obtained between different groups.The expression of hsa-let-7f-5p(FC:3.98,p<0.001),hsa-let-7g-5p(FC:2.81,p=0.008),hsa-miR-148a-3p(FC:2.58,p<0.001)and hsa-miR-192-5p(FC:3.52,p=0.036)were significantly up-regulated in PDAC group compared with control group;The expression of hsa-let-7f-5p(FC:2.28,p<0.001),hsa-let-7g-5p(FC:1.73,p=0.036),hsa-miR-148a-3p(FC:4.66,p=0.018)and hsa-miR-192-5p(FC:2.28,p=0.013)were significantly up-regulated in PDAC group compared with PPN group,while hsa-miR-150-5p(FC:0.40,p<0.001)was down-regulated;The expression of hsa-let-7f-5p(FC:1.74,p=0.018),hsa-let-7g-5p(FC:1.62,p=0.034),hsa-miR-148a-3p(FC:1.91,p=0.043),hsa-miR-151a-3p(FC:1.85,p=0.004),hsa-miR-192-5p(FC:1.55,p=0.046)and hsa-miR-451a(FC:1.90,p=0.022)were significantly up-regulated in PPN group compared with control group.In addition,hsa-miR-486-5p and hsa-miR-423-5p were chosen as internal references.3.For 1st validation set,hsa-let-7f-5p(FC:2.3492,p=0.001),hsa-let-7g-5p(FC:2.0278,p=0.0349),and hsa-miR-192-5p(FC:2.882,p=0.0068)was up-regulated in PDAC group compared with PPN group;As compared to control group,hsa-let-7f-5p(FC:8.8555,p<0.001),hsa-let-7g-5p(FC:13.1170,p=0.0349),and hsa-miR-192-5p(FC:8.9165,p<0.001)was up-regulated in PDAC group.As compared to control group,hsa-let-7f-5p(FC:3.7697,p=0.0039),hsa-let-7g-5p(FC:6.4687,p=0.0167),hsa-miR-192-5p(FC:3.0872,p=0.0433),and hsa-miR-451a(FC:2.0902,p=0.0226)was up-regulated in PPN group.For 2nd validation set,hsa-let-7f-5p(FC:1.8508,p=0.0209),hsa-let-7g-5p(FC:2.1696,p=0.0039),and hsa-miR-192-5p(FC:2.2291,p=0.002)was up-regulated in PDAC group compared with control group.4.For differential miRNA expression analysis between PDACEs and PDACLS groups,hsa-let-7f-5p(FC:3.2235,p<0.001),hsa-let-7g-5p(FC:2.7705,p=0.0048),hsa-miR-148a-3p(FC:2.8998,p<0.001)and hsa-miR-192-5p(FC:3.7371,p=0.0032)were up-regulated in the PDACLS group as compared to PDACES group in the 1st validation set.For 2nd validation set,hsa-let-7f-5p(FC:2.5846,p=0.0065),hsa-let-7g-5p(FC:2.1820,p=0.0234),hsa-miR-148a-3p(FC:2.0217,p=0.0363)and hsa-miR-192-5p(FC:2.0569,p=0.0239)were up-regulated in the PDACLS group as compared to PDACES group.Conclusions:This study used the Illumina high throughput sequencing,total RNA extraction,the construction and quality control of RNA library,and miRNA expression analyses to screen the differentially expressed miRNA(hsa-let-7f-5p,hsa-let-7g-5p,hsa-miR-148a-3p,hsa-miR-192-5p)in PDAC,PPN,and control groups.Subsequently,TaqMan probes were used to validate the differentially expressed miRNA between different groups.Part Ⅱ Evaluation of the Potential Value of CT-based Radiomics in the Staging of Pancreatic CancerPurpose:To investigate the potential value of contrast-enhanced CT(CECT)-based radiomics in the staging of pancreatic ductal adenocarcinoma(PDAC).Materials and methods:Seventy-one pathologically proved PD AC patients with a mean age of 61.55±8.53 years were included into our study.There were 39 males and 32 females.Among 71 PD AC patients,40 were at an early stage and the others were at a late stage.All patients underwent CE-CT examinations.Following imaging features were evaluated:tumor location,morphology,boundary,texture,calcification,main pancreatic duct or bile duct dilatation,lymph node metastasis,vascular invasion,and distal metastasis.ITK-SNAP software was used for tumor segmentation at the arterial and portal phases.Analysis Kit software(version V3.0.0.R,GE Healthcare)was adopted to extract radiomics features,and then student t or Mann-Whitney u test and maximum relevance and minimum redundancy(MRMR)method was used for features selection.After that,nine predictive features were retained and used for model construction with random forest method.Ten-times LGOCV method was used to test the stabiliy and reproducibility of the radiomics model.Results:1.All 71 PDAC cases had a single lesion with a mean diameter of 3.60 ± 1.11cm.Tumor location was divided into head&neck(42 cases)and body&tail(29 cases).Sixty-four cases had an ill-defied margin.Most of tumors(62,87.3%)were predominantly solid.Calcification was less common in tumors and only 2(2.82%)had calcifications.Forty-five cases(63.4%)had main pancreatic duct dilatation and 25 cases had bile duct dilatation.Following CT imaging findings were assessed in PD AC cases:37 cases(52.1%)presented with vascular invasion,40 cases(56.3%)presented with lymph node metastasis,and 24 cases(33.8%)presented with distal metastasis.2.A total of 396 radiomics features were extracted from the arterial and portal phases,respectively.They were divided into six categories:42 histogram features,9 morphological features,144 GLCM features,180 RLM features,11 GLSZM features,and 10 Haralick features.Nine features(a_Correlation_angle135_offset4,p_Correlation_angle90_offset7,a_Correlation_angle45_offset7,a_GLCMEntropy_AllDirection_offset4_SD,a_Compactness2,a_GLCMEntropy_AllDirection_offset7_SD,p_Compactness2,a_InverseDifferenceMoment_AllDirection_offset4_SD and p_HighGreyLevelRunEmphasis_AllDirection_offset)were retained after statistical analysis.Random forest method was used to construct the radiomics model with the nine retained features,which yielded an AUC of 0.99.Then 10-times LGOVCV method was used to test the stability and reproducibility of the radiomics model,which showed an average AUC of 0.75,indicating a relatively reliable and reproducible radiomics model.Conclusions:CE-CT based radiomics is a useful tool in the staging of PD AC,which has good stability and reproducibility.Part Ⅲ Correlation Analysis between Serum Exosomal miRNA and CT Imaging Findings of Pancreatic CancerPurpose:Pancreatic ductal adenocarcinoma(PDAC)has an unfavarable biological behavior,which has a great impact on the management strategy and prognosis evaluation.Poor-differentiated PD AC usually has a high frequency of lymph node metastasis,vascular invasion,or distal metastasis.In this study,we performed staging and biological behavior analysis of PDAC based on CT imaging findings and pathological results,and we aimed to investigate the correlation between serum exosomal miRNA and CT imaging findings of pancreatic cancer.Materials and methods:Seventy-one pathologically proved PDAC patients with a mean age of 61.55±8.53 years old were included into our study.There were 39 males and 32 females.Biological behaviors of PDAC were evaluated based the combination of CT imaging findings and pathological results.The expression of the following miRNA,including hsa-let-7f-5p,hsa-let-7g-5p,hsa-miR-15b-5p,hsa-miR-148a-3p,hsa-miR-150-5p,hsa-miR-151 a-3p,hsa-miR-192-5p,hsa-miR-370-3p and hsa-miR-451a,were analyzed between PDAC with or without lymph nodes metastasis,with or without vascular invasion,and with or without metastasis.Hsa-miR-486-5p and hsa-miR-423-5p were chosen as internal references.Results:1.For differential miRNA expression analysis between PDAC with or without lymph node metastasis,hsa-let-7f-5p(FC:2.4434,p=0.0033),hsa-let-7g-5p(FC:2.7489,p=0.008),hsa-miR-148a-3p(FC:2.5898,p<0.0001)and hsa-miR-192-5p(FC:3.4084,p=0.0059)were up-regulated in the former group compared with the latter group in the 1st validation set.The 2nd validation set showed similar results:hsa-let-7f-5p(FC:1.8956,p=0.0474),hsa-let-7g-5p(1.9943,p=0.0321),hsa-miR-148a-3p(2.4314,p=0.0041)and hsa-miR-192-5p(2.2885,p=0.008);2.For differential miRNA expression analysis between PDAC with or without vascular invasion,hsa-let-7f-5p(FC:2.4434,p=0.0033),hsa-let-7g-5p(FC:3.4847,p=0.0007),hsa-miR-148a-3p(FC:2.3631,p=0.0003)and hsa-miR-192-5p(FC:4.1086,p=0.002)were up-regulated in the former group compared with the latter group in the 1st validation set.The 2nd validation set showed similar results:hsa-let-7f-5p(FC:2.5846,p=0.0065),hsa-let-7g-5p(2.1820,p=0.0234),hsa-miR-148a-3p(2.0217,p=0.0363)and hsa-miR-192-5p(2.0569,p=0.0239);3.For differential miRNA expression analysis between PD AC with or without metastasis,hsa-let-7f-5p(FC:2.3580,p=0.0023),hsa-let-7g-5p(FC:3.0590,p=0.0021),hsa-miR-148a-3p(FC:2.0043,p=0.0029)and hsa-miR-192-5p(FC:3.7171,p=0.0046)were up-regulated in the former group compared with the latter group in the 1st validation set.The 2nd validation set showed similar results:hsa-let-7f-5p(FC:3.2456,p=0.0269),hsa-let-7g-5p(3.1268,p=0.0178),hsa-miR-148a-3p(2.1939,p=0.1577)and hsa-miR-192-5p(2.5922,p=0.0285).Conclusions:Serum exosomal miRNA,including hsa-let-7f-5p,hsa-let-7g-5p,hsa-miR-148a-3p and hsa-miR-192-5p,were valuable diagnostic biomarkers in biological behavior assessment,which could be used to differentiate between PD AC with or lymph node metastasis,with or without vascular invasion,and PDAC with or without metastasis. |
PDF Full Text Request