RPS 1001 - Characterisation of upper GI-tract malignancies and beyond

Author Block: G. Fontanella¹, S. Borrelli², B. Brogna³; ¹Benevento/IT, ²Mirabella Eclano/IT, ³Avellino/IT
Purpose: The aim of the study was to establish a quali-quantative fluoroscopic severity assessment for achalasia, comparable to the equivalent clinical Eckhard scoring system.
Methods or Background: From September 2020 to August 2022, 69 patients already diagnosed with achalasia and scored with ESS, were recruited and evaluated with our fluoroscopy barium protocol. The AP sequence was used to divide the esophagus into nine segments, according to Brombart's classic description, plus the gastro-esophageal junction. Three scoring items were chosen, after a profiling study of achalasia, to depict the features, some mutually exclusive, of the three clinical subtypes: lumen dilation, stasis, spasm. Each esophageal segment was scored for the three items (1 point, item present; 0 points, no item), the IVA score was calculated by summing points up until a maximum of 20 points for each subtype was reached. IVA scores were then normalised on a 0-12 scale to be compared to ESS.
Results or Findings: IVA and ESS scores were not found to be statistically diverging in 60/69 patients (86.95%, p=0.05). IVA scores were diverging and superior to ESS in 6/69 patients (8.69%); in this group of patients, the ESS' 'chest pain'/'weight loss' items were found to be biasing factors. IVA scores were inferior to ESS in just 3/69 patients (4.34%). In all the patients with a diverging IVA score (9/9), ESS scores were found to be lower than 6/12.
Conclusion: IVA score was found to be consistent and compatibile with ESS scores, especially in patients with moderate to severe achalasia. The apparent superiority of imaging scores in a small proportion of patients might instead be used as a revealing tool to call out patients in which the ESS does not reflect the disease's severity, due to internal biases.
Limitations: This was a monocentric study and there was a 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: This study was approved by the FBF BN ethical committee.

Author Block: M. W. Ling, Y. Zhou, J. Gao; Zhengzhou/CN
Purpose: Early tumour shrinkage (ETS) is a promising parameter for assessing treatment responses. Our study hypothesised that an ETS with an optimal cut-off value was an imaging biomarker for advanced esophageal squamous cell carcinoma (ESCC) treated with first-line immunotherapy.
Methods or Background: We retrospectively enrolled 129 patients with unresectable locally advanced ESCC treated with first-line immunotherapy between 2019 and 2021. ETS was defined as the relative change in the sum of the target lesions' longest diameters at the first evaluation compared with that at baseline. Multivariate analyses were conducted to identify the significant prognostic variables in progression-free survival (PFS) and overall survival (OS).
Results or Findings: The median value of ETS was 29.5%. An ETS with a 10% cut-off value was statistically significantly associated with PFS in the univariate analysis (hazard ratio [HR]: 2.26; 95% confidence interval [95% CI]: 1.21-4.24; p = 0.009). Besides, in the univariate analysis, the longest diameter, maximum short diameter, central necrosis on enhanced computed tomography, enhanced pattern, and ETS values were statistically significant predictive factors for OS. In the multivariate analysis, ETS with a 10% cut-off value was an independent predictive factor for OS (HR: 3.14; 95% CI: 1.45-6.83; p = 0.004).
Conclusion: ETS is associated with survival outcomes in patients with advanced ESCC treated with immunotherapy. Early tumour size shrinkage of at least 10% can be regarded as a promising biomarker predictor for PFS and OS.
Limitations: First, there might have been some bias. Second, the ETS was not sufficient to fully reflect the dynamic process of tumour response to treatment, including the combination of tumour response and treatment time points. The tumour size reduction pattern, morphological response based on imaging, and other factors might be related to the patient's prognosis.
Funding for this study: Funding was provided by the Youth Project of the Henan Natural Science Foundation (no. 212300410271) and the Youth Project in Medical Science and Technology of Henan Province (no. SBGJ202003016).
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: Ethics committee approval was provided by the first affiliated hospital of Zhengzhou University.

Author Block: C. Xie¹, Z. Ning¹, Y. Gao², J. Chen¹, Q. Zhang¹, Z. Liu¹, Y. Hu¹; ¹Guangzhou/CN, ²Jinan/CN
Purpose: Neoadjuvant immunotherapy has emerged as a promising therapeutic approach for locally advanced ESCC. This study aims to systematically evaluate suspicion concerning lymph node involvement by synthesising well established imaging observations based on routine workup and reassessment CT images.
Methods or Background: In our retrospective study, we enrolled a total of 100 patients diagnosed with locally advanced ESCC. CT images were meticulously evaluated by experienced radiologists, and enlarged LN were further analysed for size measurement (long-axis diameters, short-axis diameters, and corresponding ratios) and morphological appearances (shape, enhancement pattern, the completeness of the extracapsular border, the presence of fatty hilum, necrosis, fusion, and conglomeration) for prediction of lymph node metastasis (LNM). Subsequently, we compared clinicopathological characteristics between enlarged LN groups presenting different radiological features for exploration of their potential biological significance.
Results or Findings: Neoadjuvant immunotherapy yielded a radiologically enlarged LN in 27/100 (27%) of patients. We observed a significant increase in the size of the long-axis diameter (63% vs 39%, p = 0.041) in pathologically negative enlarged LN than malignant LN. A cut-off value of >40% change in lymph node long-axis size was established as a statistically significant discriminator of LNM (AUC = 0.747). Reactive enlargements of LNs are more likely to occur in patients with favourable prognostic and predicative biomarkers (PD-L1 positivity, lower expression of ECD, EGFR, and CD44V6) .
Conclusion: Our study has contributed to our understanding of the correlation between CT-based morphological features and enlargements of LNs, potentially addressing a gap in current knowledge. Our findings indicate the observed radiological appearance of cancer progression in lymph nodes may actually be attributed to a special response pattern following neoadjuvant immunotherapy. This step is crucial for ensuring the development of an appropriate clinical treatment plan.
Limitations: The sample size was limited.
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: No information provided by the submitter.

Author Block: G. Cho; Seoul/KR
Purpose: The purpose of this study was to compare the size of metastatic lymph nodes on preoperative CT in poorly cohesive advanced gastric cancer with other types of advanced gastric cancers.
Methods or Background: Literature concerning poorly cohesive gastric cancer is scarce and lymph node metastasis is a well known prognostic factor in gastric cancer patients. So, herein, we evaluated poorly cohesive AGC with emphasis on the difference in size of metastatic LNs, compared with other types of AGC.
We retrospectively included AGC patients, who underwent gastric cancer surgery at Chung-Ang University Hospital from February 2018 to May 2023. Two abdominal radiologists independently reviewed abdomen CT scans and evaluated the largest size of visible LNs on each gastric LN stations. Measurable LNs (Defined as >3 mm SD) were matched with full pathology report on electronic database records and the metastasis status was determined. We evaluated the size difference of metastatic LNs using independent t-test.
Results or Findings: A total of 140 patients (median age, 67 years [IQR, 58-77 years]; 92 men) were evaluated. Poorly cohesive cancer was present in 27 patients (19.7%). Total number of the measurable LNs was 425. 216 out of 425 LNs were matched as malignant based on pathology report. The size of metastatic LNs in poorly cohesive AGC was significantly smaller than metastatic LNs in other types of AGC (p<0.001, geometric mean size: 5.774 vs 7.613).
Conclusion: The size of metastatic LNs in poorly cohesive AGC is significantly smaller than metastatic LNs in other types of AGC. Therefore, lowering the size threshold to 5 mm could improve preoperative CT evaluation of metastatic LNs in poorly cohesive AGC.
Limitations: The study is a retrospective single-centre study. We evaluated only the largest metastatic LNs per station rather than each and every visible lymph node.
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: This study was approved by the Chung Ang University Hostpital International Review Board.

Author Block: Y. Liangliang, J. Li, J. Qu; Zhengzhou/CN
Purpose: To explore the feasibility of quantitative DCE parameters derived from DCE-MRI combined with ADC values to predict the molecular typing of gastric cancer (GC).
Methods or Background: 43 patients were enrolled in this retrospective study. Mean values, 10th, 25th, 50th, 75th, 90th percentile values of quantitative DCE parameters (Ktrans, Kep, Ve) and ADC values were manually extracted. The specimens were performed with five biomarkers, including EBER in situ hybridisation, MLH1, PMS2, E-Cadherin and P53. According to the different expression results, they were divided into five molecular types. The aberrant E-cadherin group and aberrant P53 group were combined into a high-grade malignant group, and the other three groups were combined into a low-grade malignant group. The quantitative DCE parameters or ADC values between two malignant groups were compared.
Results or Findings: There were significant differences in Ktrans mean, Ktrans 25%, Ktrans 50%, Ktrans 75%, Ktrans 90%, Kep mean, Kep 10%, Kep 25%, Kep 50%, Kep 75%, Kep 90%, Ve 10%, Ve 25% and ADC between two malignant groups of GC (p = 0.006, 0.044, 0.007, 0.007, 0.009, 0.004, 0.032, 0.024, 0.004, 0.005, 0.016, 0.021, 0.028, 0.018, respectively). Ktrans 90% and ADC were independent risk factors for predicting the two malignant groups. The AUC values, sensitivity, specificity, positive predictive value, and negative predictive value of Ktrans 90%, ADC, Ktrans 90% + ADC were 0.720, 0.688, 0.798; 44.4%, 88.9%, 88.9%; 88.0%, 52.0%, 60.0%; 72.7%, 57.1%, 61.5%; 68.7%, 86.7%, 88.2%, respectively.
Conclusion: Quantitative DCE parameters combined with ADC values can assess different malignant groups based on molecular types of GC, which may provide new directions for the evaluation of GC.
Limitations: Firstly, the sample size is too small. Secondly, this study did not classify molecularly type of GC according to TCGA.
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: This study was approved by the Ethics Review Committee of Henan cancer hospital.

Author Block: Y. Liangliang, J. Li, J. Qu; Zhengzhou/CN
Purpose: The aim of this study was to explore the differences in ADC values based on the primary tumour and perigastric lymph nodes for distinguishing N stages of gastric cancer (GC).
Methods or Background: 160 GC patients from April 2019 to April 2022 were enrolled in this retrospective study. ADC values and relative ADC values (ADCT, rADCT, ADCLN, rADCLN) based on primary tumours and perigastric first station lymph nodes were measured separately. The ANOVA or Kruskal Wallis test was used to compare differences in ADCT, rADCT, ADCLN, rADCLN values between different N stages. ROC curves were used to determine the optimal parameters and diagnostic efficacy for predicting N0 + 1 and N2 + 3 stages.
Results or Findings: There were significant differences in ADCT, rADCT, ADCLN, rADCLN values to distinguish different N stages of GC (p values of <0.001, 0.023, <0.001, <0.001, respectively). The AUC values, sensitivity, specificity, positive predictive value, and negative predictive value of ADCT, rADCT, ADCLN, and rADCLN values in predicting N0 + 1 and N2 + 3 stages of GC were 0.714, 0.632, 0.739, 0.743; 87.7%, 46.2%, 52.3%, 83.1%; 49.5%, 72.6%, 80.0%, 57.9%; 54.3%, 53.6%, 64.2%, 57.4%; 85.5%, 66.3%, 71.0%, 83.3%, respectively.
Conclusion: Both ADC values and rADC values can be used to distinguish different N stages of GC, and rADC values based on perigastric lymph nodes have the highest diagnostic efficacy in predicting N0 + 1 and N2 + 3 stages of GC.
Limitations: Firstly, this study is a single-centre retrospective study, which may bring some bias to the results. Secondly, as this study mainly evaluates whether there are differences between ADCT and ADCLN values in distinguishing N stages of GC, we did not evaluate T stage and clinical stage, which will be our future research direction.
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: This study was approved by the Ethics Review Committee of Henan cancer hospital.

Author Block: I. Capitoni, G. Bagnacci, L. Ferradini, N. Di Meglio, A. Perrella, L. Volterrani, M. A. Mazzei; Siena/IT
Purpose: Recent evidence has shown that in patients with diffuse-type gastric carcinoma (GC), a high percentage of signet ring cells (SRC) represents a positive prognostic factor (reduced 5-year mortality risk by four times).
The aim of this study was therefore to predict the percentage of SRC in GC through radiomics and morphological criteria applied to staging CT scans.
Methods or Background: 44 patients were selected based on the following inclusion criteria: (1) re-evaluation of the percentage of SRC through histopathological examination and (2) good quality of preoperative staging CT scans.
The 10% cut-off of SRC was considered to classify patients into PC-NOS (pure poorly cohesive) and PC-SRC (with signet ring cell component). The CT images were assessed by two readers with different levels of experience, who evaluated 12 dichotomous criteria and post-contrast enhancement: the readers also segmented the tumours, and a radiomic analysis was performed using specific software (pyradiomics).
Results or Findings: Among the two groups (SRC<10% and SRC>10%), significant differences were found in the distribution of six dichotomous variables and in the post-contrast behavior (ΔHU late-portal 21.5±24.9 VS -26.1±27.2, p=0.001). From the radiomic analysis, 81 out of 106 variables were found to be reproducible between the two readers (ICC>0.75), and the developed model, including three variables, had an area under the curve of 0.726.
Conclusion: Patients with different percentages of ring-shaped cells can be accurately identified on CT scans. A larger patient cohort is desirable in the future, considering the prognostic impact of this information, which can only be extracted from a comprehensive analysis of the primary neoplasm.
Limitations: This was a monocentric study.
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 obtained from all individual participants included in the study.

Author Block: F. Mohsin; Karachi/PK
Purpose: It is important from a prognostic point of view to distinguish primary GI lymphoma from secondary extra-nodal involvement by disseminated nodal disease. Although it has a vast variety of imaging features and exact diagnosis relies on histopathologic analysis, certain imaging appearances like a bulky mass with diffuse infiltration, fat planes preservation, no signs of obstruction, multiple site involvement, associated bulky lymphadenopathy can strongly suggest the diagnosis.
Methods or Background: Primary gastrointestinal (GI) lymphoma is an uncommon disease but is the most frequently occurring extra-nodal lymphoma (10–30%) and is almost exclusively of non-Hodgkin type. The stomach, small bowel, large bowel, and oesophagus are involved in decreasing order of frequency. Risk factors for the development of gastrointestinal lymphoma include H/pylori, immunosuppression, Celiac disease, IBD and HIV.
Results or Findings: In the stomach, they typically demonstrate marked thickening of the wall with homogeneous enhancement and submucosal spread.
In the small bowel, distal ileum is classically the most common site because of the greater amount of lymphoid tissue in this portion of the bowel. Typical presentation is a thick walled infiltrating mass with aneurysmal dilatation without obstruction which occurs due to tumour invasion into the muscularis propria causing destruction of its intramural autonomic nerve plexus.
In the large bowel, they usually appear as bulky polypoid masses on CT, larger than the ones that can be encountered in colorectal adenocarcinomas and may extend beyond the bowel wall, thus presenting as enormous peritoneal masses, that can also be cavitated. Colonic lymphoma usually involves a longer segment, moreover, usually located near the ileocaecal valve and grows into the terminal ileum, not invading or obstructing neighbouring viscera.
Conclusion: It is important to decipher the mystery of primary GI lymphomas on the ground of radiology for better management.
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? Not applicable
Ethics committee - additional information: Not applicable

Author Block: G. Bagnacci¹, A. Perrella¹, N. Di Meglio¹, L. Funicelli², F. Pittiani³, A. Veltri⁴, P. Morgagni⁵, G. Mura⁶, M. A. Mazzei¹; ¹Siena/IT, ²Milan/IT, ³Brescia/IT, ⁴Orbassano/IT, ⁵Forli/IT, ⁶Arezzo/IT
Purpose: The introduction of neoadjuvant therapy (NAT) for gastric cancer (GC) has led to the need for radiologists to assess response to treatment using CT. The shortcomings of the RECIST 1.1 criteria in assesing NAT response for gastric cancer are well known. Currently, prognostic predictions rely heavily on TNM staging information obtained from staging CT.
Methods or Background: A prospective Italian registry ("METAGASTRO"), focusing on patients with metastatic gastric cancer (GC), is currently ongoing. Our subanalysis included patients who had undergone at least one cycle of neoadjuvant chemotherapy and had both staging and restaging CT scans reviewed by experienced radiologists. We selected 124 patients from six centres affiliated with the GIRCG (Italian Research Group for Gastric Cancer). For each patient, detailed data were collected on approximately 90 variables related to T, N, and M parameters in both staging and restaging.
Results or Findings: As expected, patients without peritoneal involvement or with fewer than two hepatic metastases or isolated pathological paraortic lymph nodes demonstrated significantly better survival (p=0.001). RECIST 1.1 offered no prognostic insights (p = 0.233). New criteria, incorporating different lymph node thresholds and Peritoneal Cancer Index (PCI), resulted in effective stratification (p <0.001). The combination of new response criteria with oligometastatic status provided optimal prognostic stratification (p <0.001), with the most favourable group exhibiting a median survival of 41.3 months. Interestingly, non-oligometastatic patients displaying a partial response had a prognosis similar to stable oligometastatic patients.
Conclusion: The combination of initial staging and new criteria for response provided satisfactory prognostic stratification for patients affected by metastatic GC.
The prognostic value of CT scans can be improved significantly with the expertise of skilled radiologists.
Limitations: The small sample size, limited interreader agreement as well as slight variation in time interval between staging and restaging CT were identified as limitations.
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 received institutional review board approval and written informed consent was obtained from all participants.
The prospective study was approved by the Ethical committee "Comitato Etico Regionale per la Sperimentazione
Clinica della Regione Toscana, Sezione AREA VASTA SUD EST", protocol number 13082_2018, 21/05/2018.

Author Block: M. Mitrovic, J. Kovac, L. Lazic, A. Jankovic, D. Šaponjski, S. Milosevic, K. Ebrahimi, D. Mašulović, A. Djuric-Stefanovic; Belgrade/RS
Purpose: The aim of the study is to determine the morphological characteristics of the tumour obtained by the analysis of the conventional computed tomography examination and texture analysis parameters, which may be useful as imaging biomarkers for the preoperative prediction of high-risk gastrointestinal stromal tumours.
Methods or Background: This was a prospective cohort study that was carried out in the period from 2020 to 2023. The study included 79 patients who underwent CT examination and texture analysis, surgical resection of a lesion that was suspicious for GIST, as well as pathohistological and immunohistochemical analysis.
Results or Findings: Textural analysis pointed out Min Norm (p=0.032) as a histogram parameter of the first order statistically significant in the prediction of HR GIST, while Min Norm (p=0.007), Skewness (p=0.035) and Kurtosis (p=0.003) showed significance in predicting high grades of this tumour. Univariate regression analysis identified tumour diameter, margin appearance, growth pattern, lesion shape, structure, mucosal continuity, presence of enlarged feeding or draining vessel (EFDV) and Max Norm as significant predictive factors for HR GIST. Multivariate regression analysis extracted interrupted mucosa (p <0.001) and presence of EFDV (p <0.001) as independent predictive CT features for HR GIST with an AUC of 0.878 (CI: 0.797-0.959), sensitivity of 94%, specificity of 77% and accuracy of 88% in predicting HR GIST.
Conclusion: The morphological characteristics of the tumour detected by conventional CT examination still have the greatest value in the preoperative stratification of the metastatic risk of gastric GIST. The incorporation of texture analysis into the basic imaging protocol may further improve the preoperative assessment of risk stratification.
Limitations: Our study did not include a follow-up of the involved patients.
Funding for this study: This research received no external funding.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: Our research was permitted by an Ethical Committee of the School of Medicine, University of Belgrade, code 1322/II-6, and written informed consent was obtained from all patients.

Author Block: C. Sallemi, F. Bodini, F. Rosella; Brescia/IT
Purpose: Duodenal stump fistula (DSF) is one of the most serious complications following gastrectomy, with a high risk of morbidity and mortality and a long period of hospitalisation.
When conservative management fails, percutaneous transhepatic biliary drainage is useful to reduce duodenal pressure and fistula output by aspirating bile. However, it needs a long healing time and often needs to be followed by other treatments, such as percutaneous or endoscopic injection of glue.
The aim of this study was to explore the feasibility and efficacy of a novel technique of percutaneous DFS embolisation with ethylene vinyl alcohol copolymer (EVOH) combined with an occlusion balloon in the duodenal stump for the treatment of DSF after gastrectomy for malignant or benign disease.
Methods or Background: From 2018 to 2023, 11 consecutive patients underwent PTB-A embolisation with EVOH for the treatment of DSF. Clinical and technical success, morbidity and mortality were analysed. Fistula recurrence was also evaluated.
Technical success was defined as the absence of contrast media extravasation from the duodenal stump at fistolography after the embolisation. Clinical success was defined as no leakage from the percutaneous tract of the fistula 48 hours after the embolisation.
Recurrence was defined as a fistula that recurred after clinically complete healing.
Results or Findings: Technical success was achieved in all cases. Clinical success was obtained in 8/10 patients in a single treatment. 2/10 patients were re-treated to achieve complete fistula healing. No procedure-related complications and mortality were recorded. No relapse of fistula occurred during follow-up.
Conclusion: PTB-A embolisation with EVOH of DSF after gastrectomy is a feasible and safe procedure and seems to be effective to achieve complete healing of the fistula.
Limitations: This was a retrospective study.
Funding for this study: No funding was received for this study
Has your study been approved by an ethics committee? Not applicable
Ethics committee - additional information: IRB approval was not waived; written informed consent for PTB-A embolisation was obtained from patients who signed a specific institutional procedure-related consent valid for retrospective observational studies.