SS 701a - Benign and malignant pancreatic diseases: advances in imaging
SS 701a - Benign and malignant pancreatic diseases: advances in imagingThursday, March 2, 14:00 - 15:30 Room: B Session Type: Scientific Session Topics: Oncologic Imaging, Abdominal Viscera Moderators: N. Alberti (Bordeaux/FR), D. Weishaupt (Zurich/CH) Add session to my schedule In your schedule (remove)
Pancreatic lipid deposition is determined by environmental rather than genetic factors: a classical twin study
Purpose: Intrapancreatic fat can be assessed by non-contrast-enhanced computed tomography (CT) and is associated with beta-cell dysfunction. The role of genetic and environmental factors in pancreatic lipid accumulation is unclear. Therefore, we sought to evaluate the contribution of genetics and environmental factors on pancreatic lipid content within a cohort of adult twins.
Methods and Materials: We investigated 77 twin pairs (47 monozygotic [MZ] and 30 dizygotic [DZ] same gender pairs) with a 256-slice CT-scanner. Using non-enhanced CT images we measured the average value of pancreatic attenuation (Hounsfield units [HU]) in three regions of interest. Blood samples were collected before the CT scan. Intra-pair correlations were calculated and structural equation model was used for evaluating additive genetic (A), dominant genetic (D) and unique environmental (E) components.
Results: Main clinical and laboratory findings of the twins were: age 56.1±9.4 years, BMI 27.7±5.3kg/m2, fasting glucose 97.5±25.7mg/dL, HbA1c 5.5±1.0% (mean±SD). Average pancreatic attenuation was 47.2±11.3 HU in MZ and 47.6±11.8 HU in DZ twins. The intra-pair correlation between HU values were stronger in MZ as compared to DZ twins (rMZ=0.498, p<0.001; rDZ=0.080, p=0.674). Using the structural equation model, a predominant environmental influence (E: 59%) and a moderate additive genetic dependence (A: 41%) was found. Dominant genetic influence was not identified (D: 0%).
Conclusion: We found a moderate genetic and a much stronger environmental dependence of pancreatic lipid accumulation in our twin cohort indicating that environmental factors and lifestyle characteristics are predominantly involved in the development of fat accumulation in the pancreas.
Differences in pancreatic proton-density fat fraction by MRI in subjects with prediabetes, diabetes, and controls from the general population
Purpose: To determine differences in pancreatic adipose tissue between subjects with prediabetes, diabetes, and normal controls as measured by magnetic resonance imaging (MRI) and its association with other adipose tissue compartments.
Methods and Materials: Subjects without history of cardiovascular disease with established diabetes or prediabetes as well as normal controls were included from a population-based cohort study and underwent whole-body 3T MRI. The imaging protocol included a 3D multi-echo Dixon sequence for derivation of proton-density fat fraction of the pancreas (PDFFpanc) and liver (PDFFhepatic) as well as a two-point Dixon to determine subcutaneous and visceral adipose tissue (SAT and VAT, respectively). Univariate and multivariate analyses were employed to determine associations.
Results: A total of 385 subjects were included (median age: 57 years, 58.2% males) and the prevalence of diabetes and prediabetes was moderate (13.8% and 24.7%, respectively). The median PDFFpanc was 5.2% [IQR 3.3; 9.4], and significantly higher in subjects with prediabetes and diabetes as compared to controls (6.2% [3.5; 12] vs. 8.6% [4.3; 17.5] vs. 4.9% [3.1; 7.4], respectively; pTrend<0.001). After adjusting for age, gender and BMI, the association was attenuated. In addition, while in univariate analysis BMI, PDFFhepatic, SAT and VAT were associated with PDFFpanc (all p<0.05), only VAT predicted PDFFpanc independently (Beta: 0.02; p<0.001).
Conclusion: While there is a significant difference in PDFFpanc between subjects with prediabetes, diabetes and controls, this association is largely confounded by the amount of VAT. Thus, VAT may play a prominent role in the pathophysiology of the development of hyperglycaemic states.
Optimising quality and diagnostic performance of the MDCT by using low tube voltage in patients with suspected pancreatic adenocarcinoma
Purpose: Attenuation characteristics of the pancreatic adenocarcinoma(PA) and contrast material uptake vary depending on histologic grade, pancreatic parenchyma condition and scanning parameters. The purpose of this study was to evaluate the impact of low-tube voltage multiphase contrast-enhanced MDCT on tumor conspicuity and overall diagnostic performance.
Methods and Materials: 42 consecutive patients with clinically suspected PA and indications for MDCT were enrolled in the study and randomized into two groups based on the tube voltage (group 1 - 120 kV and group 2 - 100 kV). We compared mean tumor and parenchyma attenuation, noise, effective dose as well as qualitative score between groups. Relative tumor enhancement ratio (RTE) was calculated for each contrast enhancement (CE) phase.
Results: The mean BMI was 25.5±3.56kg/m2 and 26.3±3.14kg/m2, respectively (p=0.552). 34 PA and 8 chronic pancreatitis were identified by CT. Parenchyma attenuation was significantly higher during arterial and venous phase in 100kV group - 106±17 HU and 95±18 HU vs 91±14 and 89±9 HU in 120kV group, respectively (p<0,0001). Arterial RTE in 100kV group was -0.73±0.13 (-0,88±0,05 in 120kV group, p<0.001), venous RTE was -0.33±0.23 (-0,45±0,25 in 120kV group, p>0.05), delayed RTE was 0.59±0.27 (0,21±0,53 in 120kV group, p>0.05). Qualitative analysis showed better lesion conspicuity in 100kV group, especially on arterial and delayed phases. Diagnostic accuracy of the MDCT was higher in 100kV group (sensitivity 97,3 vs 91,3%, specificity 84,1 vs 80%).
Conclusion: Low-tube-voltage scanning protocol may be used for improving diagnostic quality and performance of the multiphase contrast-enhanced MDCT during pancreatic adenocarcinoma diagnostic workflow.
Timing-specific contrast media protocol enhances image quality at reduced contrast volume and radiation dose during computed tomography of the pancreas
Purpose: To investigate the opacification of the pancreatic vasculature and parenchyma during CT utilizing a patient-specific contrast formula.
Methods and Materials: This hybrid prospective and retrospective study was institutional review board approved. In 220 consecutive patients, pancreatic CT was performed with one of two protocols: protocol A, 100mL of contrast material injected via timed bolus triggering technique; or protocol B, employing a patient-specific contrast media protocol specifically timed at the gastroduodenal artery. Measured attenuation of pancreatic parenchymal, arterial and venous vasculature. Effective dose was calculated. Data were compared with the independent two-sample t-test. Receiver operating characteristic (ROC), visual grading characteristic (VGC) and Cohens’ kappa analyses were performed.
Results: Pancreatic measurements in each of the segments during the arterial and venous phase were significantly higher in Protocol B (mean±standard deviation, art: 96.59HU±27.37; venous: 91.28HU±20.88) compared to A (art:77.86HU±21.14; venous:73.99HU± 14.75)(p<0.0001). Arterial opacification was significantly higher in protocol B compared to A(p<0.036). In the venous circulation, the inferior vena cava, superior mesenteric, portal and splenic veins demonstrated reduction in vascular opacification protocol B compared to A(p<0.001). The contrast media volume in protocol B(57.60±12.25 mL) was significantly lower than in protocol A(100±1 mL)(p<0.001). Effective dose was significantly reduced in protocol B(2.75±0.63mSv) compared to A(4.015±0.89mSv) (p<0.001). ROC and VGC analysis demonstrated significantly higher area under the curve for protocol B(p<0.0001)(p<0.034) respectively, with inter-reader agreement increasing from good to excellent in pancreatic lesion detection.
Conclusion: Timing-specific contrast media protocol enhances image quality at reduced contrast volume and radiation dose during CT of the pancreas.
CT features of non-hypervascular endocrine tumours of the pancreas: a comparison with pancreatic adenocarcinoma
Purpose: To review the CT features of non-hypervascular endocrine tumours of the pancreas (nonH-NETs) and compare them with those of pancreatic adenocarcinomas.
Methods and Materials: We reviewed the multiphasic MDCTs performed on 38 nonH-NETs. These were compared with 38 pancreatic adenocarcinomas matched by size. All patients underwent multiphasic MDCT after administration of a weight-based amount of high-concentration contrast agent. One reader evaluated tumour size, margins, homogeneity, enhancement pattern (subjectively increasing enhancement from arterial to venous phase vs other patterns) and presence of vascular invasion. All diagnoses were confirmed at pathology. Data were analysed with Fisher’s test.
Results: Mean tumour diameter was 46 mm for nonH-NETs and 45.5 mm for adenocarcinomas (p=ns). 41/57 nonH-NETs had well-defined margins, in contrast to 15/57 adenocarcinomas (p<0,0001). Calcifications were seen in 5/57 nonH-NETs and none of the adenocarcinomas (specificity 100%). No difference in tumour homogeneity was observed between the two groups (p=ns). 33/57 nonH-NETs were hypodense in the arterial phase and 22/57 were isodense, 54/57 adenocarcinomas were hypodense and 3/57 isodense (p<0,0001). 15/57 nonH-NETs and 11/57 adenocarcinomas showed increasing enhancement from the arterial to the venous phase (p=ns). The MPD upstream to the tumour had significantly larger calibre in adenocarcinomas (5.8 vs 2.2 mm; p<0,0001). No difference was observed for arterial or venous invasion between the two groups (p=ns).
Conclusion: Non-hypervascular endocrine tumours have well-defined margins, cause less MPD dilatation, and are more commonly isodense than adenocarcinoma. The presence of tumoural calcifications is specific for nonH-NETs.
Insulinoma detection with cross-sectional imaging: comparison of biphasic-enhanced CT, volume perfusion CT and 3T multiparametric MR
Purpose: To evaluate the diagnostic performance of biphasic contrast-enhanced CT (CECT), volume perfusion CT(VPCT) and multiparametric MR (mp-MR) at 3T, in patients with clinically suspected insulinomas.
Methods and Materials: Sixty-four patients with clinically suspected insulinomas were prospectively enrolled. All patients underwent biphasic CECT, VPCT and 3T mp-MR with identical protocols. Two radiologists independently determined the presence/absence of tumour using a 5-scale confidence level. Conspicuity of the lesion and clarity of tumour-to-pancreatic duct distance were graded. CECT, VPCT and mp-MR were evaluated in random order, with time interval of at least 7 days. Receiver operating characteristic (ROC) analysis was performed to compare the radiologists' diagnostic confidence.
Results: Forty-seven patients were tumour positive, with a total of 51 tumours. The Az values for tumour detection were as follows: 0.715 (CECT), 0.903 (VPCT) and 0.955 (mp-MR) for reader 1, and 0.738 (CECT), 0.895 (VPCT) and 0.956 (mp-MR) for reader 2. VPCT and mp-MR were significantly more accurate than CECT for insulinoma detection (p=0.02 and 0.01 for reader 1, and p=0.03 and 0.01 for reader 2). Lesion conspicuity was better on mp-MR and VPCT compared to CECT (both p<0.01). Tumour-to-pancreatic duct distance was better appreciated on MR, compared to CECT and VPCT (both p<0.01). The weighted k values indicate good to excellent agreement between observers for determining tumour presence/absence (k=0.64 for CECT, 0.80 for VPCT and 0.84 for mpMR).
Conclusion: VPCT and mp-MR were significantly more accurate than CECT for insulinoma detection, and mp-MR demonstrate better tumour conspicuity and clearer tumour-to-duct distance.
Correlation between appearance of the retroportal fat plane at preoperative CT and pathology findings in resected adenocarcinoma of the pancreatic head
Purpose: To correlate the CT and pathology findings in resected adenocarcinomas of the pancreatic head (PDAC).
Methods and Materials: We included 48 patients with resected PDAC of the pancreatic head (24M, 24F, mean age 65 years). All patients underwent multiphasic preoperative MDCT less than 30 days before surgery. All cases were re-evaluated at pathology for the state of the retroportal lamina, lymphatic and perineural invasion. CT images were reviewed in consensus by two radiologists for assessment of the fat plane between the pancreatic head and the mesenteric vessels: this was graded in two categories (clear, effaced/infiltrated). Fisher's test was used to assess the correlation between CT and pathology findings.
Results: A clear fat plane between the pancreatic head and the mesenteric vessels was significantly associated with a negative retroportal lamina at pathology (p=0.0037). No different results were observed between effaced and clearly infiltrated fat. No association was observed between the appearance of the fat planes at CT and the presence of lymphatic or perineural invasion (p=ns).
Conclusion: A clear fat plane between the pancreatic head and the mesenteric vessels is significantly associated with negative retroportal lamina at pathology. CT is not accurate in predicting lymphatic or perineural invasion.
Body composition parameters, pancreatic volume and texture as radiological predictors of pancreatic fistula after Whipple procedure
Purpose: Patients after Whipple procedure (pancreatoduodenectomy) experience postoperative complications in up to 50-60% of cases. Many surgical and radiological reports describe pancreatic fistula (PF) as the most frequent complication (10-24%). Aim was to find out the radiological predictors of PF in patients after Whipple procedure.
Methods and Materials: We retrospectively analysed 32 patients who underwent Whipple procedure and had a preoperative CT scan as staging in our centre. CT images were processed to obtain measures of pancreatic texture (density in nonenhanced scan), whole pancreatic volume, volume of pancreatic remnant and body composition parameters including whole body fat-free mass index (FFMI), whole body fat mass index (FMI), L3 skeletal muscle index (SMI).
Results: PF occurred in 15 patients who were included in the first group, the other 17 patients did not have PF and were included to the second group. There were significant differences (p≤0.05) between the two groups in the whole pancreatic volume (32.1+/-3.4cm3 vs 21.4+/-3.0cm3), texture of the pancreas (38.5+/-3.6HU vs 21.7+/-3.0HU) and volume of pancreatic remnant (18.4+/-2.1cm3 vs 12.1+/-1.4cm3). The significant difference of SMI (44.3 ± 8.3cm2/m2 vs 40.5 ± 8.8cm2/m2), FFMI (15.4 ± 2.4kg/m2 vs 14.3 ± 2.5kg/m2), FMI (8.8±2.5kg/m2 vs 9.5±2.5kg/m2) was not determined.
Conclusion: Pancreatic texture, preoperative whole pancreatic volume and volume of pancreatic remnant are independent predictors of pancreatic fistula. Body composition parameters (FFMI, FMI, SMI) do not affect the development of pancreatic fistula.
Purpose: To define feasibility of the 640-slice dynamic volume computed tomography in diagnostics of vascular complications and assessment of pancreas graft’s (PG) perfusion.
Methods and Materials: 23 patients after pancreas transplantation were investigated using Toshiba 640-slice computer tomography. The contrast was injected through the central venous 16G-catheter. Data acquisitions were done without breath holding. Arterial blood flow parameters (AF, ml/min/100ml) were measured at the level of graft head, body, tail areas, and the time of maximum PG’s tissue peak (TTP, sec.) contrasting was determined. The obtained results were analysed and compared with those of others authors dealing with perfusion of native pancreas.
Results: 12 cases showed adequate PG’s blood flow at both arteries with normal organ function. The TTP-19[17;21], AF-117[97;123], 113[98;125], 104[89;126]. 9 cases revealed blood flow absence in a PG’s superior mesenteric artery (PG blood flow was going separately through splenic artery’s system). There were neither clinical nor laboratory signs PG dysfunction. AF-137[117;160], 126[120;132], 125[105;137], TTP-18[17;19] and statistically didn't differ from first patients group and a native pancreas perfusion. Also there were in 2 patients expressed depression of PG’s blood supply in both arteries which resulted in depressed perfusion: AF - 44[20;66], 52[18;85], 57[21;92], TTP - 11[9;13]. Clinical and laboratory signs of PG’s dysfunction were also found in this patients.
Conclusion: 640-slice dynamic volume perfused computed tomography allows to visualise whole-PG’s vascular structure and anatomy. It also helps objectively estimate the adequacy of PG’s blood supply and clinical importance of vascular complications.
Diagnostic role of new generation multidetector-CT scanner in detection and characterisation of incidental pancreatic cystic lesions: comparison with MRCP
Purpose: To compare the diagnostic accuracy of MDCT to MRI with MRCP sequences, in detecting pancreatic cystic lesion and those features useful for their characterization.
Methods and Materials: We evaluated 50 patients with no history of pancreatitis, neoplasm and prior surgical procedures in pancreas, who underwent both MDCT and MRI with MRCP (time interval lower 90 days),for pancreatic cysts characterization. The reviewer recorded number of lesions detected in each patients and the main morphologic features of defined target lesion: location (head,uncinate process,neck,body,tail), number, size, communication with main pancreatic duct (MPD), maximum MPD diameter, presence of septa, mural enhancing nodules. Data were analysed to compare diagnostic performance of MDCT and MRI.
Results: A greatest number of pancreatic cysts was identified on MRI than CT examinations (227 vs 128); however ICC value of 0.76 suggested a good agreement.Very good agreement (ICC 0.98) was found regarding the diameter of target lesions(21.4 mm on CT vs 21.8 mm on MRI),location of target lesion (κ=0.90), and detection of MPD dilatation (κ=1.00) and septa (κ=0.86). Fair agreement about the evaluation of MPD communication (κ=0.33) and calcifications (κ=0.22) was found.
Conclusion: MDCT can be consider almost equivalent to MRI with MRCP in the evaluation of pancreatic cystic lesion, offering detailed morphologic features helpful for their characterisation and those correlated with a high risk of malignancy.
Purpose: - Compare the diagnostic accuracy of 256 multislice CT and endoscopic ultrasound (EUS) in characterization and assessment of resectability of pancreatic masses. - Comparison of 256 slice MDCT and EUS findings with histopathological/operative findings.
Methods and Materials: Total of 36 patients with pancreatic masses were included. Clinical and laboratory findings were recorded followed by dual phase examination using pancreatic protocol. EUS was performed using transducer of 5-13 MHz. FNAC was done wherever feasible. Parameters regarding tumor size, location, imaging morphology and vessel involvement were recorded. Findings were compared with histopathological /operative diagnosis/clinical follow up.
Results: -Out of 36 patients, MDCT and EUS established diagnosis consistent with tissue diagnosis in 30 (83%) and 22(61%) patients respectively. However, best results were obtained with combined use of MDCT and EUS ie 34(95%) patients. The number of patients categorized as inconclusive by MDCT were lower compared to EUS (2/36 cases on MDCT vs 10/36 cases on EUS) -MDCT showed specificity and PPV of 100% in assessment of resectability of pancreatic adenocarcinoma compared to EUS which had specificity and PPV of 75% and 92.3% respectively. The sensitivity and NPV was 100% for both modalities.
Conclusion: -MDCT is first line imaging modality in detection, characterization of pancreatic masses and assessment of resectability in malignant solid neoplasms. The specificity and PPV of MDCT was higher than EUS in determining resectability of pancreatic adenocarcinoma. -EUS is beneficial in detection of masses <2 cm in size causing only mild pancreatic contour deformity on CT, and for guiding FNAC.