The Further Study Of Pancreatic Carcinoma And The Correlation With Histopathologic Findings In Specimens And Resectability Of Pancreatic Carcinoma With High Field MRI

Part One Pancreatic carcinoma and MR imaging: diagnostic value of 3D VIBE with iPAT sequenceObjective To compare the difference of each MRI sequences for signal intensity(SI) and contrast-to-noise ratio(CNR), and to analyze the value of 3D VIBE with iPAT sequence in diagnosis of pancreatic carcinoma.Materials and Methods Forty-nine patients with pancreatic carcinoma proved by histo-pathology or clinical follow up were underwent MR examination. T|WI 2D FLASH, T₁WI 3D VIBE with and without iPAT , T₁WI 3D VIBE sequences with iPAT in pancreatic parenchymal phase(PPP) , hepatic phase(HP) scanning and T₁WI 2D FLASH in delayed phase(DP) scanning were acquired. The crude SI, the difference of SI among normal pancreas, pancreatic carcinoma and no tumor pancreas, the normal pancreas-to-tumor contrast-to-noise ratio were defined.Results (1) The crude SI of the normal pancreas, pancreatic carcinoma and no tumor pancreas were differed statistically significantly within the sequence except the normal pancreas and no tumor pancreas during contrast enhancement scanning. (2) The analysis of variance showed the SI of normal pancreas during PPP and HP, pancreatic carcinoma and no tumor pancreas during plain T₁WI 3D VIBE with and without iPAT, no tumor pancreas during PPP and HP did not differ statistically significantly from each other, but the others showed a significant difference between the sequences. (3) The SI difference between normal pancreas and pancreatic carcinoma showed a significant difference between the sequences except plain VIBE and VIBE (iPAT). (4) TheSI change of normal pancreas subtract pancreatic carcinoma before and aftercontrast with VIBE (iPAT)sequence displayed inversive V shape. The PPP wereon a higher level than the rest. (5) During the sequence, PPP demonstratedthe best CNR of normal pancreas to pancreatic carcinoma, delayed phase thelowest, others did not show a significant difference.Conclusion VIBE sequence with iPAT before and after contrast can obtain wellSI and CNR in pancreatic carcinoma MR examination, PPP demonstrated the bestCNR of normal pancreas to pancreatic carcinoma.Part Two Diagnostic evaluation of 2D and 3D MRI in pancreatic carcinoma and the optimal critical point of resectability of pancreatic carcinoma invasion to peripancreatic vesselsOneDiagnostic evaluation of 2D and 3D MRI in pancreatic carcinomaObjective To evaluate the value of various sequences and modified dynamic contrast-enhanced MRI method in diagnosis of pancreatic carcinoma.Materials and Methods Forty-nine patients with pancreatic carcinoma proved by histo-pathology or clinical follow up were underwent MR examination. Plain MR sequences include: T₁WI 2D FLASH,T₁WI 3D VIBE,TSE T₂WI and HASTE with FS. Modified dynamic contrast-enhanced scanning procedures were as following: The coronal scans with 3D FLASH and axial scans using T₁WI 3D VIBE sequences were, in turn, used to upper abdominal area and pancreatic areas, the coronal imaging were used to obtain 3D peripancreatic vessels during arterial phase and portal vein phase scanning respectively; the axial imaging were used to detect the tumor well in pancreatic parenchymal phase and hepatic phase scanning respectively, final scanning with T₁WI 2D FLASH axial sequence was covered the whole upper abdomen.Results (1) Of 49 cases, the lesions showed hypo-intensity in 45 and iso-intensity in 4 on 3D VIBE. The lesions demonstrated hypo-intensity in 46 and iso-intensity in 3 on 2D FLASH. On TSE T₂+FS, lesions contour are not clearly well. Three of them displayed iso-intensity, others were iso- and hyper-intensity. (2) During pancreatic parenchymal phase, 48 lesions showed hypo-intensity and ring enhancement could be seen in 39 cases. Twenty-four lesions displayed progressive inequable or marginal tubercle and inner compartment enhancement during hepatic phase and delayed phase. Six showed iso- and hyper-intensity in delayed phase. (3) According to thirty-seven lesions underwent surgical operation, MRI and MRA made a good correlation with surgical findings for peripancreatic vessels invasive or non-invasive by tumors except three superior mesenteric arteries and four superior mesenteric veins were misdiagnosed.Conclusion As "one-stop-shop" MRI examination, modified MRI sequences including dynamic coronal and axial contrast-enhanced scanning is feasible and applicable and has great value in diagnosing, staging and evaluating resectability of pancreatic carcinoma. TwoPancreatic carcinoma invasion to peripancreatic vessels: The optimal critical point of evaluation the resectability with MRI Objective Compared with surgical exploration, to discuss the sensitivity, specificity and accuracy of MRI in evaluating the invasion of peripancreatic vessels and respectability of pancreatic carcinoma, and further to establish optimal critical point.Materials and methods Forty-one cases confirmed pancreatic carcinoma by surgery-pathology were collected, to which MRI plain and contrast enhanced scans, include 37 of them with coronal MRA scans were performed. The invasionof peripancreatic vessels were preoperatively evaluated based on MRI signs. According to the invasion degree, such grading method as 1, 2a, 2b, 3a, 3b and 4 grades, the number of vessels invasion and non-invasion were calculated, and compared with the findings of surgery. The sensitivity and specificity of each grade for vessels invasion were studied and draw receiver operator characteristic curve, the sensitivity, specificity, positive predictive value, negative predictive value and accuracy of unresectability for 2 degree and 2a degree were also calculated respectively. The invaded arteries and veins of 2 degree and resectability were analyzedResults Twenty-two of forty-one patients were resected, among which twenty were curative resection. Compared with the findings of surgery, seven vessels including three artery and four veins were misdiagnosed by MRI. Using 1, 2a, 2b, 3a and 3b grade as standard of respectability, The sensitivity of unresectability was 78.3%,84.8%,67.4%,56.5% and 47.8% respectively. Receiver operator characteristic curve demonstrated 2a degree was the optimal critical point. The sensitivity, specificity, positive predictive value, negative predictive value and accuracy of 2a and 2 degree in identifying predictive unresectability were 84.8%, 98.5%, 92.9%, 96.6%, 95.9% and 67.4%, 99.5 %, 97.9%, 93.0%, 89.4%. Three of six artery involved as 2 degree by MRI were resected, while only six of eighteen veins were resected, among which include five of eleven as 2a and one of seven as 2b.Conclusion Compared with 2 degree, 2a degree could improve the sensitivity, negative predictive value and accuracy of unresectability. In clinical practice, 2a degree is the optimal critical point as the respectability standard. Part threeThe head of pancreatic carcinoma: A comparative study of in vivo and ex vivo MRI manifestation and histopathologic findings in specimensObjective: To prospectively analyze the MRI signal intensity characteristics of in vivo and ex vivo of pancreatic head carcinoma and to assess the correlations with histopathologic specimens.Materials and methods Sixteen patients with pancreatic head carcinoma were examined with 1. 5T MRI. MR sequences include T₁WI 2D FLASH, TSE T₂WI with section thickness of 5 mm and no intersection gap, routine plain and contrast enhanced sequences. All resected fresh specimens were reexamined with MRI plain scans in an hour of resection, the sequence similar to that of in vivo. After examination, the specimens were cut into 5mm thick, corresponding to the same plane as that Of MRI scans. The specimens were embedded in paraffin using routine hematoxylin and eosin staining with serial section of gross pathologic specimen excised. MRI findings were thus compared with pathology. Results (1) In vivo MRI findings: Fourteen lesions demonstrated hypo-intensity on T₁WI +FS, 2 were iso-intensity. On TSE T₂WI, all lesions displayed hetero-intensity except 3 iso-intensity and had poor boundary with peripancreas parenchyma. During pancreatic parenchymal phase, 15 lesions showed hypo-intensity and ring enhancement could be seen in 11 cases .Nine lesions displayed progressive inequable or marginal tubercle and inner compartment enhancement during hepatic phase and delayed phase. Four showed iso- and hyper-intensity in delayed phase. (2) Ex vivo MRI findings: Signal intensity of lesions and pancreas parenchyma is superior to that of in vivo. All lesions demonstrated hypo-intensity on T₁WI +FS and hetero-intensity On TSE T₂WI. Compared with in vivo, the range and boundary of hyper-intensity on T₂WI were more clear, among which 4 cases could distinguish relic pancreas parenchyma in lesions and 5 cases could see tumors infitrating peripancreas parenchyma. (3) MRI and Radiologic-pathologic correlation: The different proportion of the tumor tissue may has slight signal intensity difference on T₁WI and T₂WI. Tumor cell, colloidal fiber and inflammatory infiltration could demonstrate hypo-intensity on T₁WI . iso-intensity on T₁WI correspondedto more tumor cell and inflammatory infiltration and less colloidal fiber. On the contrary, hypo- and iso-intensity indicated more colloidal fiber, slight hyper- intensity indicated relative gently desmoplastic reaction, hyper- intensity areas were mucous degeneration or little ductal dilatations on T₂WI. Ring enhancement areas of perilesions during pancreatic parenchymal phase were caused by multiple factor, progressive inequable or marginal tubercle and inner compartment enhancement during hepatic phase and delayed phase were partly the results of desmoplastic reaction.Conclusion (1) Being different proportion of pathological tissue in pancreatic head carcinoma, the signal intensity of lesions is difference on plain and contrast enhanced MRI sequence. (2) MRI could reveal the pathologic character of pancreatic head carcinoma. Signal intensity of ex vivo is superior to that in vivo, could provide more information to react different pathological component.