Research Presentation Session: Oncologic Imaging

RPS 1816 - Liver, biliary tract and pancreas

March 2, 09:30 - 11:00 CET

7 min
An image-based comparison of histological subtypes in intraductal papillary mucinous neoplasm of the pancreas
Jessica Ritter, Munich / Germany
Author Block: J. Ritter, I. Pergolini, I. E. Demir, F. Lohöfer, R. Braren; Munich/DE
Purpose: Intraductal papillary mucinous neoplasms (IPMN) are precursor lesions of pancreatic ductal adenocarcinoma (PDAC). IPMN can be distinguished by histomorphological and immunohistochemical characteristics as intestinal, oncocytic, pancreatobiliary and gastric subtypes. The Fukuoka consensus guideline is a classification system for risk assessment in IPMN. The aim of the study was to compare these subtypes according to the Fukuoka criteria.
Methods or Background: MRI scans of 104 patients undergoing pancreatic surgery for IPMN were re-examined for the Fukuoka criteria. Statistics on the distribution of Fukuoka criteria and the incidence of malignancy were performed.
Results or Findings: In IPMN with pancreatobiliary subtype (n=16), 75% had carcinoma compared to 44% with intestinal subtype (n=41) and only 25% with gastric subtype (n=47). Further analysis using the Fukuoka criteria compared all malignant versus benign IPMNs, regardless of the histological subtype. Here significantly more worrisome features (P-value 0.017) and high-risk stigmata (P-value 0.0077) were detected in the presence of malignancy. For the individual histological subtypes, no significant difference in the number of worrisome features and high-risk stigmata was found between the benign and malignant cases. A cross-subgroup comparison within the malignant IPMN showed that patients with PDAC based on pancreatobiliary or intestinal subtype had significantly more worrisome features than the gastric cohort (2.08/2.22 vs 1.25; P-value 0.040/0.026). Regarding the high-risk stigmata all three subtypes presented similarly.
Conclusion: According to our findings pancreatobiliary and intestinal subtype are more susceptible to malignant progression than gastric subtype. At the same time, our results show that the Fukuoka criteria are only of limited use for preoperative subtyping in IPMN.
Limitations: A limitation of this study is the relatively small number of patients of different subtypes. At least similarly sized patient groups could be formed in the case of malignancy.
Funding for this study: No funding was received for this study.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: The study was approved by the ethics committee of the university hospital rechts der Isar, technical university of Munich (number 87/18S)
7 min
Dual-layer detector spectral CT: a non-invasive preoperative tool for predicting histopathological differentiation in pancreatic ductal adenocarcinoma
Wei Liu, Shanghai / China
Author Block: W. Liu, Z. Zhou, T. Xie, Y. Wang; Shanghai/CN
Purpose: The study aimed to predict histopathological differentiation grades in patients with pancreatic ductal adenocarcinoma (PDAC) before surgery, with quantitative and qualitative variables obtained from dual-layer spectral detector CT (DLCT).
Methods or Background: A total of 128 patients with histopathologically confirmed PDAC and preoperative DLCT were retrospectively enrolled and categorised into the low-grade (LG) (well and moderately differentiated, n=82) and high-grade (HG) (poorly differentiated, n=46) subgroups. The regions of interest were placed on tumours, and both conventional and spectral variables for PDAC were determined, including CT attenuation, contrast enhancement fraction (CEF), iodine concentration (IC), effective atomic number (Zeff), iodine enhancement fraction (IEF), and the slope of the spectral curve (K-slope). Necrosis was visually assessed on both conventional CT images (necrosis_con) and virtual mono-energetic images (VMIs) at 40 keV (necrosis_40keV). Receiver operating characteristic (ROC) curves were used to evaluate the efficiencies of variables in predicting tumour grade.
Results or Findings: Necrosis_con (odds ratio [OR]=2.84, 95% confidence interval [CI]: 1.13–7.13; p<0.001), necrosis_40keV (OR=5.82, 95% CI: 1.98–17.11; p=0.001) and IEF in the arterial (AP) and venous (VP) phases (OR=1.12, 95% CI:1.07–1.17; p<0.001) were significantly different between the LG and HG PDAC groups by univariable and multivariable analyses. IEF_AP/VP (AUC=0.754, p=0.016) and the combination model (AUC=0.812, p<0.001) had better predictive performances than necrosis_con (AUC=0.580). The combination model yielded the highest sensitivity (72%) and accuracy (79%), while IEF_AP/VP exhibited the highest specificity (89%).
Conclusion: Variables derived from DLCT have the potential to preoperatively evaluate PDAC tumour grade. Furthermore, spectral variables and their combination exhibited superior predictive performances than conventional CT variables.
Limitations: The small sample size, and at single institution, was an identified limitation.
Funding for this study: Funding was received from: Shanghai Science and Technology Innovation Action Plan, Hong Kong, Macao and Taiwan Science and Technology Cooperation Project (22490760800) and Shanghai Minhang District Medical Characteristic Specialty Development Project (2020MWFC05).
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: The study was approved by the Ethics Committee of Fudan University Shanghai Cancer Center.
7 min
Preoperative scoring model for identifying proliferative hepatocellular carcinoma on multiphase liver MRI and its implication for surgical resection
Junhan Pan, Hangzhou / China
Author Block: J. Pan, Y. Zhu, F. Chen; Hangzhou/CN
Purpose: The study aimed to develop and validate a scoring model for preoperative identification of proliferative hepatocellular carcinomas (HCCs) using multiphase liver MRI, and to compare the early recurrence rates between patients with model-predicted proliferative and nonproliferative HCCs after surgical resection.
Methods or Background: Between September 2019 and August 2021, 476 consecutive patients with surgically proven HCC who underwent preoperative multiphase liver MRI were retrospectively enrolled. Logistic regression analyses were conducted to determine predictors associated with proliferative HCC. We established a scoring model based on the identified predictors (training cohort n=332) and verified it in a time-independent validation cohort (n=144). Early recurrence rates were evaluated by the Kaplan–Meier method with log-rank test.
Results or Findings: Serum alpha-fetoprotein level >100 ng/mL, irregular tumour margin, rim arterial phase hyperenhancement, marked diffusion restriction, and tumour-to-liver ratio at arterial phase ≤1.4 were independent predictors of proliferative HCCs. The optimal threshold of the scoring model for diagnosing proliferative HCCs was >18 points, with an area under the curve of 0.83 and 0.80 in the training and validation cohort, respectively. In both two cohorts, patients with model-predicted proliferative HCCs exhibited significantly higher early recurrence rates compared to those with model-predicted nonproliferative HCCs (all P<0.05). Moreover, patients with BCLC stage B-C and predicted nonproliferative HCCs exhibited similar early recurrence rates to those with BCLC stage 0-A and predicted proliferative HCCs in both cohorts (all P>0.05).
Conclusion: The developed scoring model, incorporating four MRI features and serum alpha-fetoprotein level, showed promising potential for predicting proliferative HCCs and identifying suitable surgical candidates for patients with HCC.
Limitations: This was a single-centre retrospective study with inherent selection bias.
Funding for this study: This study received funding from the National Natural Science Foundation of China (12090020 and 12090025) and Zhejiang Provincial Natural Science Foundation Committee-Zhejiang Society for Mathematical Medicine Joint Fund Major Project (LSD19H180003).
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: This study was approved by the Institutional Review Board of the First Affiliated Hospital, College of Medicine, Zhejiang University.
7 min
Intraindividual comparison of CT and MRI for predicting vessels encapsulating tumour clusters in hepatocellular carcinoma: which imaging modality is better?
Junhan Pan, Hangzhou / China
Author Block: J. Pan, Y. Zhu, F. Chen; Hangzhou/CN
Purpose: The study aimed to establish and validate scoring models for identifying vessels encapsulating tumour clusters (VETC) pattern in hepatocellular carcinoma (HCC) using CT and MRI, and to compare the predictive performance across the modalities.
Methods or Background: Between June 2019 and August 2020, 324 consecutive patients with surgically confirmed HCC who underwent preoperative multiphase CT and MRI were retrospectively enrolled. Logistic regression analyses were conducted to determine predictors associated with the VETC positive HCC. We established two separate scoring models based on the identified predictors (training cohort n=227) and verified it in an independent validation cohort (n=97). Generalised estimating equations were used to compare the diagnostic performance of the two models.
Results or Findings: Tumour size >5.0 cm, intra-tumoural artery, three or four types of enhancement pattern, tumour-to-liver ratio at arterial phase >1.57, and serum AFP > 400 ng/ml were identified as independent predictors for VETC positive HCC on CT. Conversely, independent predictors on MRI included incomplete "capsule", necrosis or serve ischaemia, marked diffusion restriction, tumour-to-liver ratio at ADC map ≤1.03, and serum AFP >400 ng/mL. The CT-based and MRI-based scoring models were constructed using these predictors, with cut-off values of >17 points and >21 points, respectively. Compared to the MRI-based model, CT-based model exhibited lower sensitivity but higher specificity. However, the CT-based and MRI models demonstrated similar AUC values in both training (0.83 versus 0.85, P=0.468) and validation cohorts (0.81 versus 0.80, P=0.966).
Conclusion: Preoperative CT and MRI may show comparable predictive performances for identifying the VETC pattern in HCC. The application of CT or MRI should be determined by the preferred trade-off between sensitivity and specificity in clinical practice.
Limitations: This was a single-centre retrospective study with inherent selection bias.
Funding for this study: This study received funding from the National Natural Science Foundation of China (12090020 and 12090025) and Zhejiang Provincial Natural Science Foundation Committee-Zhejiang Society for Mathematical Medicine Joint Fund Major Project (LSD19H180003).
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: This study was approved by the Institutional Review Board of the First Affiliated Hospital, College of Medicine, Zhejiang University.
7 min
CT evaluation of hepatic steatosis as a predictive factor for the development of surgical complications in patients with pancreatic adenocarcinoma resected after neoadjuvant chemotherapy
Elisa Boffa, Padova PD / Italy
Author Block: E. Boffa, G. Zamboni, M. Antolini, A. Spezia, G. Malleo, G. Mansueto; Verona/IT
Purpose: The study aimed to assess the impact of neoadjuvant chemotherapy (NAT) on the onset of moderate-to-severe liver steatosis and postoperative complications in patients with pancreatic adenocarcinoma (PDAC).
Methods or Background: This retrospective study included patients with PDAC who received NAT, underwent CT before and after NAT and underwent major pancreatic resections. All patients were included in a prospectively maintained registry; 71 patients were also enrolled in an IRB-approved prospective study and underwent intraoperative liver biopsy. Two readers in consensus drew multiple ROIs on liver and spleen parenchyma on the venous phase: the difference between the two mean densities was calculated and used to quantify steatosis. Imaging assessment of steatosis was compared with biopsy results in 71 patients. The population was divided into two groups based on the steatosis degree (group 1: <30%; group 2: >30%). Post-surgical complications and Clavien-Dindo Index (CDI) were compared between groups applying the chi-squared test.
Results or Findings: A total of 234 patients were included (113 males, 121 females; mean age 62 years). Liver steatosis at CT was significantly correlated with biopsy results (p=0.0002). After NAT, 82 patients (35%) developed steatosis or worsened their degree of steatosis. 109 patients (47%) developed post-surgical complications (POPF, liver failure, bile leak, PPAP). Moderate-severe steatosis before NAT was significantly correlated with a CDI ≥3 (p=0.0453) and with 90-days postoperative mortality (p=0.0262). A positive trend of correlation was observed between moderate-severe steatosis after NAT and CDI ≥3 and PPAP.
Conclusion: In patients affected by PDAC undergoing neoadjuvant treatment, liver steatosis evaluated on routine CT is associated with a higher risk of developing complications after major pancreatic resections.
Limitations: No limitations were identified.
Funding for this study: No funding was received for this study.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: Informed consent was provided by all patients (PAD-R registry, n1101CESC). For this retrospective study, IRB approval was not required. IRB approval was obtained for the prospective LIMBO study.
7 min
Apparent diffusion coefficient and tissue stiffness are associated with different tumour microenvironment features in hepatocellular carcinoma
Jie Chen, Chengdu / China
Author Block: J. Chen, T. Chen, Z. Zhang, W. Zhenru, B. Song; Chengdu/CN
Purpose: The study aimed to investigate associations between tissue diffusion, stiffness, and cell density, nuclear volume fraction, tumour-stroma ratio (TSR), tumour stemness, and CD8+ T cell infiltration in resected hepatocellular carcinoma (HCC).
Methods or Background: We included 72 HCC patients with preoperative magnetic resonance (MR) diffusion-weighted imaging and MR elastography examination. Free hands region of interests was placed on the central four slices of the apparent diffusion coefficient (ADC) map and the stiffness map, and the mean value of the tumour and peri-tumour parenchyma was estimated. Cell density, nuclear volume fraction, TSR, CK19 expression, and CD8+ T cell infiltration was estimated in the whole slide histopathological image of the resected tumour sample. Spearman’s rank correlation coefficients and intraclass correlation coefficients were calculated.
Results or Findings: Inter-reader agreement was excellent regarding ADC and stiffness measurement. Tumour ADC correlated with cell density (r=-0.39, p=0.001) and nuclear volume fraction (r=-0.39, p=0.001). Tumour stiffness (TS) correlated with nuclear volume fraction (r=-0.28, p=0.02) and TSR (r=-0.33, p=0.005). Peri-tumour ADC correlated with CK19 expression (r=0.40, p=0.001) in HCC. Peri-tumour stiffness correlated with CD8+ T cell infiltration (r=0.27, p=0.02) in HCC.
Conclusion: In HCC, tumour ADC reflects tumour cellularity and nuclear volume fraction, tumour stiffness reflects nuclear volume fraction and TSR. Peri-tumour ADC is associated with CK19 expression in tumour, and the peri-tumour stiffness is associated with CD8+ T cell infiltration in tumour.
Limitations: The limitations of the study are its retrospective nature and small sample size.
Funding for this study: Funding was provided by the Science and Technology Support Program of Sichuan Province (grant number: 2022YFS0072) and the 1.3.5 Project for Disciplines of Excellence, West China Hospital, Sichuan University (grant number: ZYJC21012).
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: The study was approved by the Ethics Committee on Biomedical Research, West China Hospital of Sichuan University
7 min
Pancreatic cancer in photon-counting CT: low keV virtual monoenergetic images improve tumour conspicuity
Matthias Michael Wöltjen, Minden / Germany
Author Block: M. M. Wöltjen, J. R. Kroeger; Minden/DE
Purpose: The aim of the study was to identify differences in the tumour conspicuity of pancreatic
adenocarcinomas in different monoenergetic or polyenergetic (T3D) series and
contrast phases in photon-counting CT (PCCT).
Methods or Background: A total of 34 patients were retrospectively enrolled in this study. Quantitative image analysis was performed with ROI measurements in different monoenergetic levels ranging from 40 up to 70 keV (5-point steps) and polyenergetic series. Tumour-parenchyma attenuation differences and CNR were calculated. A qualitative image analysis was accomplished by four radiologists. Differences between groups were evaluated for statistical significance using the Friedman test and in case of significant differences pair-wise post-hoc testing with Bonferroni correction was applied.
Results or Findings: Tumour-parenchyma attenuation difference was significantly different between the different image reconstructions for both arterial- and portal-venous-phase-images (p<0.001). Tumour-parenchyma attenuation difference was significantly higher on arterial-phase images at mono 40 keV compared to T3D images (p<0.001) and mono 55 keV images or higher (p<0.001). For portal-venous-phase images tumour-parenchyma attenuation difference was significantly higher on mono 40 keV images compared to T3D images (p<0.001) and mono 50 keV images (p=0.03) or higher (p<0.001). The same trend was seen for CNR. Tumour conspicuity was rated best on mono 40 keV images with 4.3±0.9 for arterial-phase images and 4.3±1.1 for portal-venous-phase images. In contrast, overall image quality was rated best on T3D images with 4.8±0.5 for arterial-phase images and 4.7±0.6 for portal-venous-phase images.
Conclusion: Low keV virtual monoenergetic images significantly improve the tumour conspicuity of pancreatic adenocarcinomas in PCCT based on quantitative and qualitative results. On the other hand, readers prefer polyenergetic images for overall image quality.
Limitations: Limitations were the monocentric study design and the limited number of patients.
Funding for this study: No funding was received for this study.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: The study was conducted according to the guidelines of the Declaration of Helsinki and approved by the institutional review board (AZ: 2021-805).
7 min
Is computed tomography-based body composition a reliable predictor of chemotherapy-related toxicity in pancreatic cancer patients?
Vito Chianca, Naples / Italy
Author Block: V. Chianca1, M. Cefalì2, C. Reiner3, S. Stocker3, M. Kissling4, F. Del Grande1, S. De Dosso2, S. Rizzo1; 1Lugano/CH, 2Bellinzona/CH, 3Zurich/CH, 4Basle/CH
Purpose: Malnutrition, loss of weight and of skeletal muscle mass are frequent in pancreatic cancer patients, a majority of whom will undergo chemotherapy over the course of their disease. Available data suggest a negative prognostic role of these changes in body composition on disease outcomes; however, it is unclear whether tolerance to chemotherapeutic treatment is similarly and/or negatively affected. We aimed to explore this association by retrospectively assessing changes in body composition and chemotherapy-related toxicity in a cohort of advanced pancreatic cancer patients.
Methods or Background: 131 patients (mean age 69.7 ± 9.0 years, 45% women and 55% men), most of whom (81.5%) had metastatic disease at diagnosis, were enrolled in this study.
Body composition was evaluated through clinical parameters and through radiological assessment of muscle mass, skeletal muscle area, skeletal muscle index and skeletal muscle density; an assessment of fat distribution by subcutaneous adipose tissue and visceral adipose tissue was performed. We performed descriptive statistics, pre/post chemotherapy comparisons and univariate and multivariate analyses to assess the relation between changes in body composition and toxicity.
Results or Findings: Toxicity risk increased with an increase of skeletal muscle index (OR: 1.03) and body mass index (OR: 1.07), whereas it decreased with an increase in skeletal muscle density (OR: 0.96). Multivariate analyses confirmed a reduction in the risk of toxicity only with an increase in skeletal muscle density (OR: 0.96).
Conclusion: This study suggests that the retrospective analysis of changes in body composition is unlikely to be useful in predicting toxicity to gemcitabine-nab-paclitaxel.
Limitations: The small sample size was an identified limitation.
Funding for this study: No funding was received for this study.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: The Ethics Committee approved this retrospective study with a waiver for informed consent (2020-01085).
7 min
Intrahepatic mass-forming cholangiocarcinoma: is there additional prognostic value in using Gd-EOB enhanced MRI?
Sebastian Halskov, Berlin / Germany
Author Block: S. Halskov, F. Krenzien, L. K. Segger, D. Geisel, B. Hamm, J. Ihlow, W. Schöning, T. A. Auer, U. Fehrenbach; Berlin/DE
Purpose: The study aimed to investigate the prognostic value of enhancement patterns of intrahepatic mass-forming cholangiocarcinomas (IMCCs) during the hepatobiliary phase (HBP) in gadoxetic acid (Gd-EOB) enhanced MRI.
Methods or Background: We retrospectively identified 66 consecutive patients with histopathologically proven IMCCs (reference standard: resection) and preoperative Gd-EOB-enhanced MRI. Gd-EOB retention area was subjectively rated based on areas of intermediate signal intensity. Lesions were classified as either hypointense (0-25% retention area) or significantly-retaining (>25% retention area). Clinical, radiological and prognostic features were compared between these groups. The primary endpoint was recurrence-free-survival (RFS) after primary surgical resection.
Results or Findings: In total, 73% (48/66) of lesions were rated as hypointense and 29% (19/66) as significantly-retaining. While the hypointense subgroup more frequently featured local and distant intrahepatic metastases (p=0.039 and p=0.022) and an infiltrative growth pattern (p=0.005), RFS and clinical features did not differ significantly with Gd-EOB retention area or quantitatively measured HBP enhancement ratios. Lymph node metastasis was an independent predictor of poor RFS (p=0.001).
Conclusion: Gd-EOB-enhanced MRI reveals two subtypes of IMCC in the HBP: hypointense and signal-retaining. The hypointense subtype is associated with more frequent intrahepatic metastases and an infiltrative growth pattern, indicating potential tumour aggressiveness. However, this did not result in a significant difference in survival after primary resection of IMCC.
Limitations: This was a retrospective study of a surgical cohort, which could lead to selection bias: the prognostic value of imaging features in unresectable HCC remains unclear. Histopathological analysis to identify small-duct and large-duct IMCCs could have been of added value for identifying prognostic imaging subtypes. Quantitative analysis of HBP enhancement evaluated the whole tumour without taking intra-lesion heterogeneity in enhancement into account.
Funding for this study: No funding was provided for this study.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: Our institutional review board approved this retrospective study (internal registration number: EA2/016/14) and waived informed consent due to the retrospective nature. The study protocol conforms to the ethical guidelines of the 2002 Declaration of Helsinki.
7 min
Efficacy and evaluation of response in hepatocellular carcinoma (HCC) treated with stereotactic body radiotherapy (SBRT)
Francesca Castagnoli, London / United Kingdom
Author Block: F. Castagnoli1, A. Villanacci2, M. Bertuletti2, B. Frittoli2, L. Grazioli2; 1London/UK, 2Brescia/IT
Purpose: The study aimed to evaluate changes in ADC values, size, wash-in, wash-out, T2 signal intensity, necrosis and capsule of HCCs treated with SBRT.
Methods or Background: A total of 62 patients with HCC (65 lesions), treated with SBRT, who underwent baseline gadoxetate-enhanced MRI within four weeks before treatment and follow-up MRI within three months after SBRT were prospectively enrolled; 44 patients underwent a further MRI follow-up 12 months after SBRT. Three radiologists evaluated tumour size (∂), ADC, wash-in, wash-out, capsule, necrosis and T2 signal of all lesions. Local control (LC) was evaluated after a median follow-up (FU) of 2.1 years. Delta(Δ)∂ and ADC were correlated using Spearman correlation coefficient. Δ∂ and ΔADC were compared (Mann-Whitney test) between patients with and without LC. A logistic regression model was trained to estimate the probability of response of patients using categorical imaging variables.
Results or Findings: The median HCC ADC (10⁻³mm²/s) and dimension were 0.94 and 22.5 mm at baseline; 1.326 and 16.4 mm at 3-month follow-up; 1.352 and 10.6 mm at 12-month follow-up. At 12-months, the ADC values increased by 41.1% (p<0.001), whereas tumour size decreased by 52.7% (p<0.001).
At 3-month follow-up there was no correlation between the increase in ADC values and decrease in lesion dimension (p=0.41).
At 12-month follow-up, there was correlation between the increase in ADC values and decrease in lesion dimension (p=0.04).
At 3- and 12-month follow-up, there was significant difference in Δ(∂) and presence of wash-out between patients with and without LC (p=0.05, p=0.03; p=0.04, p=0.02). No significant correlation were found for Δ(ADC), wash-in, T2w signal, necrosis and capsule between patients with and without LC.
Conclusion: SBRT showed excellent LC for HCC, which was associated with the degree of size change after treatment and absence of wash-out at follow-up.
Limitations: Identified limitations were the small sample size and the limited cases of local recurrence (n=7).
Funding for this study: No funding was received for this study.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: The current study was carried out according to the Declaration of Helsinki (1964) and was approved by the Ethics Committee of the ASST Spedali Civili, Brescia.

This session will not be streamed, nor will it be available on-demand!