RPS 501a - New insights in pancreatic imaging

Author Block: M. M. L. Engels¹, S. Hoogenboom², A. Chuprin¹, J. van Hooft³, J. Legout¹, M. Wallace¹, C. Bolan¹; ¹Jacksonville, FL/US, ²Amsterdam/NL, ³Leiden/NL
Purpose or Learning Objective: To analyse the radiological errors associated with pre-diagnostic pancreatic ductal adenocarcinoma (PDAC).
Methods or Background: This is a single-centre, retrospective, case-control study of 87 pre-diagnostic cross-sectional imaging studies (CT (N=60) and MRI (N=27)) of patients diagnosed with PDAC <3 years later, and 338 matched control studies (CT (N=235), MRI (N=103)). Two board-certified radiologists, blinded to case/control status, independently reassessed each exam. One radiologist was then unblinded and utilised the 2016 revised RADPEER criteria to grade radiological errors and further classify them according to the classification proposed by Kim and Mansfield (AJR, 2014).
Results or Findings: Pancreatic masses were retrospectively suspected in around 57.0% of CT and MRI cases, and 1.6% of controls. Respectively, in comparison to available original CT and MRI reports, the reviewers concurred with the original report in 43.8% and 26.8% (RADPEER 1), with understandable discrepancies in 29.2% and 33.3% (RADPEER 2) and unacceptable discrepancies in 27.0% and 38.1% (RADPEER 3). The ‘under-reading’ error was by far the most common, occurring in 57.2% of CT and 45.5% of MRI cases. Errors of ‘satisfaction of search’ occurred in 31.9% of MRI cases and 10.7% of CT cases, and ‘faulty reasoning’ in 17.9% of CT cases and 4.5% of MRI cases.
Conclusion: Unidentified, early findings of PDAC that should not have been overlooked occur up to 27.0% of CT and 38.1% of MRI cases. Given the detrimental effects of a delayed diagnosis in this deadly disease, radiologists must be aware of these findings and be trained to prevent the most commonly occurring errors.
Limitations: Retrospective design and selection bias.
Ethics committee approval: Mayo Clinic IRB approval #18-002403.
Funding for this study: Funding received by Champions for Hope / Funk-Zitiello foundation.

Author Block: G. H. Pöhler, F. Klimes, L. Behrendt, H. Winther, F. Wacker, K. I. Ringe; Hannover/DE
Purpose or Learning Objective: The aim of this study was to evaluate the influence of different clinical preconditions on hepatic tissue shrinkage following computed tomography (CT)-guided microwave ablation (MWA) of malignant liver tumours.
Methods or Background: 25 patients (f=14, median age 61 years) with liver tumours (primary n=15; metastases n=10; mean size 18 mm) referred for CT-guided MWA, were included. Pre- and postinterventional CT images were post-processed using rigid/non-rigid registration and Liver/tumour segmented. Pre- and post-interventional 3D tissue volume analysis was performed voxel-wise using Jacobian Determinant (JD) mapping of the tumour (zone 1), a 5 cm perimeter around the tumour (zone 2) and the whole liver (zone 3) and was compared (tumour subcapsular vs. non-subcapsular; cirrhosis, yes vs. no; primary vs. secondary tumour; history of chemotherapy, yes vs. no. Shapiro-Wilk test, Wilcoxon rank-sum test, p<0.05 deemed significant). Zone 1-3 shrinkage was correlated with local tumour recurrence (Spearman’s correlation coefficient (ρ)).
Results or Findings: The median JD tissue volume change in zone 1 was +6%, zone 2 -0.8 % and zone 3 +0.9 %. In non-subcapsular tumours, shrinkage in zone 2 was significantly pronounced as compared to subcapsular (-3.8% (-8.5 - 1.1%) vs. +3.4% (-0.07 8.1%), p=0.004). In cirrhotic patients, shrinkage in zone 1 was significantly less compared to non-cirrhotic patients (+20.1% (8.6-67.5%) vs. -0.7% (-8.1-3.2%), p=0.0027). The extent of tissue shrinkage in zones 2 and 3 correlated inversely significantly with local tumour recurrence (ρ=-0.51, p=0.017, ρ=-0.69, p<0.001).
Conclusion: Tissue shrinkage after hepatic MWA is more pronounced in patients with a non-subcapsular tumour location and in patients with non-cirrhotic liver parenchyma. Limited tumour and tissue shrinkage may be associated with a higher risk of local tumour recurrence after hepatic MWA.
Limitations: Manageable patients' numbers in subgroups.
Ethics committee approval: IRB-approved study
Funding for this study: Funding was received by the Deutsche Forschungsgesellschaft

Author Block: F. Mambrin, M. Bariani, G. Zamboni, M. C. Ambrosetti, E. Bannone, G. Marchegiani, G. Mansueto; Verona/IT
Purpose or Learning Objective: The ISGPS has recently defined that a diagnosis of post-pancreatectomy acute pancreatitis (PPAP) requires: sustained postoperative hyperamylasemia, clinically relevant features and radiological alterations consistent with PPAP. Our purpose is to analyse the CT features that correlate with early postoperative serum pancreatic amylase (spAMY) trends after partial pancreatic resections.
Methods or Background: Patients who underwent major pancreatic resections between 2016 and 2021, and had a contrast-enhanced postoperative MDCT between POD3 and POD15 were included. Patients were divided into 3 groups based on their spAMY patterns: spAMY1: values always within/below the reference range or with a single increase in spAMY>upper limit of normal at any POD; spAMY2: sustained increase on POD0+1; spAMY3: sustained increase in spAMY including POD1+2. Two readers in consensus, blinded to spAMY trends, analysed the exams and logged arterial enhancement homogeneity, main pancreatic duct (MPD) diameter, peripancreatic oedema, and fluid collections.
Results or Findings: 473 patients were included (288 males, 185 females, mean age 63 years). Inhomogeneous arterial enhancement was more common in spAMY3 patients (49%) than in spAMY2 or spAMY1 patients (both 40%, p<0.001). MPD was larger in spAMY1 (mean calibre 2.40±1.68mm; p<0.001). Peripancreatic fat stranding (p=0.005) and fluid collections, especially peripancreatic (p<0.001), were significantly more common in spAMY3.
Conclusion: A clear definition of the CT features of PPAP is fundamental for the radiological description of this entity, which will lead to its accurate and early diagnosis and thus to better treatment.
Limitations: SpAMY evaluation was conducted within a single centre, with homogeneous surgical approaches and postoperative management.
Ethics committee approval: Approved by the Ethics Committee of the Provinces of Verona and Rovigo; approval number: 1101CESC.
Funding for this study: No financial help was obtained for this study.

Author Block: G. Chiti, F. Flammia, G. Grazzini, P. Tortoli, S. Bettarini, V. Miele; Florence/IT
Purpose or Learning Objective: This single-centre retrospective study aims to assess whether contrast-enhanced computed tomography (CECT) radiomics analysis is predictive of gastroenteropancreatic neuroendocrine neoplasms (GEP-NENs) grade based on the 2019 World Health Organization (WHO) classification and to establish a tumour grade (G) prediction model.
Methods or Background: Preoperative CECT images of 72 patients with GEP-NENs were retrospectively reviewed and divided into two groups (G1-G2 in class 0, G3-NEC in class1). A total of 107 radiomics features were extracted from each neoplasm ROI in CECT arterial phases acquisitions with a 3DSlicer. The Mann-Whitney test and the LASSO regression method were performed in R for feature selection and reduction to build the radiomic-based predictive model. The model was developed for a training cohort (75% of the total) and validated on the independent validation cohort (25%). ROC curves and AUC values were generated on training and validation cohorts.
Results or Findings: 40 features were found to be significant in class distinction. From the LASSO regression, 3 features were identified as suitable for group classification and used to construct the tumour grade radiomic-based prediction model: MajorAxisLength, Mean and 90Percentile, also individually statistically significant for the group differentiation (p-values respectively 0.001, <0.001 and <0.001). The prediction model resulted in AUC values of 0.84 (95% CI: 0.72-0.97) and 0.82 (95% CI: 0.62-1) for the training and validation cohorts, respectively.
Conclusion: CT-radiomics analysis may aid in differentiating the histological grade for GEP-NENs.
Limitations: Due to the low incidence of the disease, the study population is numerically limited to perform 4 groups (G1, G2, G3, NEC separately) classification statistics.
Ethics committee approval: Approved by the Ethics Committee of our Institution (register number 13261_oss, approved on 02/02/2021).
Funding for this study: No funding was received for this study.

Author Block: R. M. Pozzi-Mucelli¹, C. F. Moro¹, M. Del Chiaro², R. Valente³, L. K. Blomqvist¹, N. Papanikolaou⁴, J-M. Löhr¹, N. Kartalis¹; ¹Stockholm/SE, ²Aurora, CO/US, ³Umeå/SE, ⁴Lisbon/PT
Purpose or Learning Objective: Current guidelines base the management of IPMN on well-established resection criteria (RC), including cyst size. However, malignancy may occur in small cysts. Since branch-duct (BD) IPMN aren’t perfect spheres, the volumetric and morphologic analysis might better correlate with mucin production and grade of dysplasia. Nonetheless, their role in malignancy (high-grade dysplasia/invasive cancer) prediction has been poorly investigated. Previous studies evaluating RC also included patients with IPMN-associated/concomitant pancreatic cancer (PC), which may affect RC’s yield. This study aimed to assess the role of volume, morphology, and other well-established RC in malignancy prediction in patients with BD- and mixed-type IPMN after excluding IPMN-associated/concomitant PC.
Methods or Background: Retrospective study of 106 patients (2008-2019) with histopathological diagnosis of BD- and mixed-type IPMN (without associated/concomitant PC) and preoperative MRI available. Standard imaging and clinical features were collected and cyst volume and elongation value [EV=1-(width/length)] calculated on T2-w images. Logistic regression analysis was performed and predicted probabilities (PP) were calculated. Statistical significance set at two-tails, p<0.05.
Results or Findings: Neither volume [Odds ratio (OR)=1.01, 95%CI:0.99-1.02, p=0.12) nor EV (OR=0.38, 95%CI:0.02-5.93, p=0.49) were associated to malignancy. Contrast-enhancing mural nodules (MN), main pancreatic duct (MPD)>=5mm and elevated CA19-9 serum levels (>37𝜇mol/L) were associated to malignancy [MN OR:4.32, 95%CI:1.18-15.76, p=0.02; MPD>=5mm OR:4.2, 95%CI:1.34-13.1, p=0.01; CA19-9 OR:6.72, 95%CI:1.89-23.89, p=0.003]. The PP for malignancy in a hypothetical male >=70-years-old increased from 0.08 (95%CI:0.02-0.27) with no risk-factors, to 0.92 (95%CI:0.63-0.99) with all three risk-factors.
Conclusion: Volume and EV cannot predict malignancy in BD- and/or mixed-type IPMN. MN, MPD>=5mm and elevated CA19-9 are associated with higher malignancy risk even after the exclusion of IPMN-associated/concomitant PC.
Limitations: Small sample, only including operated patients with IPMN-diagnosis.
Ethics committee approval: Institutional Review Board approval was obtained (EPN 2015/154431/4).
Funding for this study: This study has received funding from Stockholm Regional Council.