RPS 2001 - Upper GI imaging

Author Block: J. Liu¹, B. Fan¹, Y. Liao², X. Li¹, W. Deng²; ¹Nanchang/CN, ²Guangzhou/CN
Purpose: Extracellular volume fraction (ECV) derived from spectral CT has been proposed as a potential imaging biomarker reflecting tumor stromal characteristics. This study aim to investigate the diagnostic value of spectral CT-derived ECV for preoperative gastric cancer staging, histological differentiation and peritumoral invasion.
Methods or Background: Eighty-six patients with surgically resected and pathologically confirmed gastric cancer were prospectively enrolled. Pathological results served as the reference standard, and patients were grouped according to: T stage (T1–2, n=30 vs T3–4, n=56), N stage (negative, n=31 vs positive, n=55), perineural and vascular invasion (negative, n=35 vs positive, n=51), and histologic differentiation (low differentiation, n=49 vs moderate–high differentiation, n=37). Delayed-phase images were acquired 180 seconds after contrast injection. Regions of interest (ROIs) were manually delineated and ECV values calculated based on iodine concentration. Receiver operating characteristic (ROC) analysis was performed to evaluate the diagnostic efficacy of ECV in different pathological groups.
Results or Findings: ECV showed limited diagnostic performance for discriminating histologic differentiation (AUC=0.602, 95%CI:0.461-0.740 ), but demonstrated good diagnostic value for predicting T stage (AUC=0.852, 95%CI: 0.747-0.946), lymph node metastasis (AUC=0.827, 95%CI: 0.727-0.927) and perineural/vascular invasion (AUC=0.821, 95%CI: 0.726-0.916). These findings suggest that ECV predominantly reflects tumor stromal features, with advantages in assessing tumor- and peritumoral infiltration in gastric cancer.
Conclusion: Delayed-phase ECV derived from spectral CT may serve as a non-invasive imaging biomarker for gastric cancer aggressiveness, particularly for tumor infiltration, lymph node metastasis, and perineural/vascular invasion, while its value for assessing histologic differentiation is limited.
Limitations: This single-center retrospective study with a limited sample size did not include subgroup analyses, which may affect the stability and generalizability of the results.
Funding for this study: None
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: Jiangxi Provincial People's Hospital Ethics Committee

Author Block: I. Cecchi, D. Cozzi, G. Angeluzzi, E. Buttitta, D. Rossini, D. Lavacchi, L. Antonuzzo, V. Miele; Firenze/IT
Purpose: The aim of this study is to investigate if CT assessment of baseline sarcopenia could be considered as a prognostic factor for overall survival (OS) and relapse-free survival (RFS) in a population of 133 patients with locally advanced gastric cancer recruited from Careggi Hospital, Florence.
Methods or Background: We included 133 patients with a diagnosis of locally advanced gastric cancer who underwent the same neoadjuvant chemotherapy followed by surgery. Skeletal muscle cross-sectional area at the third lumbar vertebra (L3) was measured using 3D Slicer software on non-contrast-enhanced CT scan performed before neoadjuvant therapy. Different cut-off points for men and women were used to classify individuals as sarcopenic or non-sarcopenic.
Results or Findings: Sarcopenia was identified in 26.3% of patients. Median OS was numerically different between sarcopenic and non-sarcopenic patients (59.0 vs 67.3 months); in the same way, RFS was numerically lower in sarcopenic subjects compared with non-sarcopenic ones (38.0 vs 51.0 months). Nevertheless, no statistically significant differences in survival outcomes were observed between the two groups.
Conclusion: Baseline sarcopenia, as assessed by CT, was not associated with a significant difference in OS or RFS among patients with locally advanced gastric cancer undergoing neoadjuvant chemotherapy and surgery.
Limitations: The limitations of the study are the small and heterogeneous population. Moreover, sarcopenia was examined only at baseline. Future studies should expand the patient cohort, assessing sarcopenia at the end of neoadjuvant therapy to evaluate longitudinal changes and considering a multivariate analysis including socio-demographic and clinical variables.
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:

Author Block: D. Liao¹, E. B. Mark¹, M. B. Poulsen¹, L. Davidsen¹, V. Rosca², V. Drug², C. Brock¹, J. B. Frøkjær¹, A. M. Drewes¹; ¹Aalborg/DK, ²Iasi/RO
Purpose: Traditional techniques for evaluating gastric physiology, such as scintigraphy and gastric manometry, face noteworthy limitations, especially in capturing the dynamic behaviour of the stomach wall during phases of filling and emptying. To overcome these challenges, we introduce VIGA (VIrtual GAster), a novel integrated analytical platform that leverages Magnetic Resonance Imaging (MRI) to quantify stomach geometry and motility.
Methods or Background: VIGA and its standardized analytical pipeline were developed and applied to static and dynamic MRI datasets from 12 healthy participants following the consumption of a 600 ml liquid test meal (300 kcal). This enabled comprehensive assessment of stomach morphology and wall dynamics. Morphological analysis involved measuring the volume and surface geometry of the fundus, corpus, and antrum. Motility was characterized through a detailed motility map, from which contraction frequency, amplitude, and propagation speed were extracted.
Results or Findings: The mean volume of the entire stomach was 785.7 ± 220 ml after the test meal, where the gastric liquid volume was 464.1 ± 59 ml. The mean contraction frequency was 2.83 ± 0.06 cycles per minute. The analysis identified distinct regional differences in contraction patterns, with the distal stomach showing significantly greater contraction occlusion (15.8±4.0%) compared to the proximal region (8.9±4.0%, p<0.001). Furthermore, contraction propagation speed tended to be higher in the distal stomach (4.87±1.7 mm/s) than in the proximal (4.01±0.9 mm/s, p=0.07). Surface geometry also varied across compartments, with the fundus exhibiting the highest inverse mean curvature, an indicator of the ratio between wall tension and intragastric pressure (p=0.003).
Conclusion: VIGA offers a robust, non-invasive imaging-based framework for evaluating gastric function, with promising implications for enhancing clinical assessment of gastric disorders.
Limitations: The subjects were investigated in supine position, negating the effects of gravity in the gastric content.
Funding for this study: This study was supported by the North Denmark Region’s Health Innovation Fund.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: The study was approved by the regional ethics committee (N-20220059).

Author Block: G. Rinaldi, M. C. Ambrosetti, M. Bariani, E. Strazimiri, G. Zamboni, S. Giacopuzzi, J. Weindelmayer, M. Sacco, G. Puppini; Verona/IT
Purpose: Although in the case of caustic ingestion treatment is usually based on endoscopic evaluation (Zargar classification), CT could provide an accurate assessment and potentially reduce overtreatment. Our aim was to assess reproducibility and effectiveness of a CT-based classification of caustic injuries in a specialized center, for predicting need for emergency intervention and long-term complications.
Methods or Background: five radiologists and six surgeons blinded to patient outcomes independently reviewed CT of 36 patients admitted to the Emergency Department between 2014-2025 for caustic ingestion and who underwent endoscopy, applying the 3-grade score by Chirica: grade I (regular), grade IIA (thickened, hypointense and vascularized mucosa, wall/perivisceral fat tissue edema), grade IIB (necrotic mucosa with slight external wall enhancement), grade III (transmural necrosis/lack of wall enhancement) for both esophagus and stomach. Kendall’s Kappa inter-operator concordance and CT sensitivity and specificity for each physician were calculated for emergency treatment (grade III) and stenosis risk (grades IIa-IIb), comparing these results to the Zargar classification.
Results or Findings: for emergency treatment (Zargar III), mean sensitivity was 37.1% for esophagus evaluation, 30.2% for stomach, and 34.2% for both, with the best performance reached by the radiologist for esophageal evaluation (SE38.4%). For stenosis prediction (Zargar IIA-IIB), the mean sensitivity was 69% for esophagus evaluation, 59.7% for stomach, and 68% for both, with the best performance reached by both radiologist and surgeon for stomach evaluation (SE100%).
Interobserver concordance was higher for radiologists than for surgeons (k=0.867-0.895 vs k=0.198-0.573).
Conclusion: CT classification, compared to endoscopic evaluation, is not sensitive enough for predicting the need for emergency surgery. Sensitivity is higher for risk of stenosis. While CT offers a promising alternative by potentially reducing overtreatment and predicting long-term complications, the interobserver variability highlights the need for standardization.
Limitations: Retrospective study
Funding for this study: None
Has your study been approved by an ethics committee? Not applicable
Ethics committee - additional information:

Author Block: G. Veneziano, D. Palumbo, S. Battaglia, C. Sarzo, U. Elmore, R. Rosati, F. De Cobelli, F. Puccetti, A. COSSU; Milano/IT
Purpose: Anastomotic leakage (AL) remains the most feared complication after Ivor Lewis esophagectomy, with significant morbidity/mortality and difficult, non-standardized, management. CT-based topographical characterization of AL may provide crucial insights into its pathogenesis, clinical course, and prognosis, potentially guiding risk stratification and therapeutic decisions.
Methods or Background: This is a single-center retrospective analysis of 85 consecutive patients who underwent Ivor Lewis esophagectomy (01/2015-12/2024) and performed contrast enhanced CT scan because of AL development. ALs were classified into four topographical subtypes based on CT findings: left-anterior, right-anterior, left-posterior, right-posterior. Radiological features (perianastomotic collections, air within collections, pleural effusion, pneumothorax, pneumomediastinum) and clinical outcomes were analyzed using bivariate and multivariate regression models; the course of right gastroepiploic arteries was also noted.
Results or Findings: The most common subtype of AL was found to be the left-anterior one (40%), followed by right-anterior (20%), left-posterior (14%), and right-posterior (11%). Most ALs manifested within eight postoperative day (72%). Left-anterior leaks demonstrated strongest association with perianastomotic collections containing air (p<0.001) and significantly correlated with respiratory and circulatory complications. Multivariate analysis revealed left-anterior leaks as independent predictors of severe outcomes (Clavien-Dindo 4b-5; p<0.001), with highest mortality rates predominantly from respiratory failure. Notably, the left-anterior position corresponds to the region opposite to the right gastroepiploic artery vascular territory, potentially representing an area of relative hypoperfusion.
Conclusion: CT-based topographical classification of AL after Ivor Lewis esophagectomy provides valuable prognostic stratification. The predominance and severity of left-anterior leaks may reflect underlying vascular vulnerability mechanisms that warrant further investigation. These findings underscore the need for targeted preventive strategies, including enhanced intraoperative vascular assessment and potentially modified surgical techniques focusing on this high-risk anatomical region
Limitations: Small sample of patients in one high-volume center and need for standardization of indications.
Funding for this study: None
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: San Raffaele Scientific Institute Ethics Committee

Author Block: V. Patanè, P. Senneca, M. C. Brunese, G. Sarnelli, R. Grassi, S. Cappabianca, A. Reginelli; Naples/IT
Purpose: Esophagogastric junction outflow obstruction (EGJOO) according to Chicago Classification v4.0 requires manometric, clinical, and supportive evidence of impaired bolus transit. Timed barium esophagogram (TBE) is a widely available imaging test providing objective evidence of functional relevance. This study evaluated the diagnostic performance and reproducibility of standardized TBE in suspected EGJOO.
Methods or Background: In this prospective single-center study, 124 consecutive patients underwent standardized TBE for esophageal emptying assessment. Examinations were performed and interpreted blinded to symptoms and high-resolution manometry (HRM). Barium column height and width were measured at 1, 2, and 5 minutes after ingestion of low-density barium sulfate. HRM, performed according to CCv4.0, served as the reference. Interobserver agreement between a senior radiologist and a junior resident was assessed.
Results or Findings: TBE was positive in 20 patients (16%), all with functionally relevant EGJOO (FR-EGJOO) on HRM. The mean barium column height at 1 minute was 122.9 ± 65.4 mm, decreasing to 89.0 ± 54.6 mm at 2 minutes. Residual barium at 5 minutes occurred in eight patients, all with the largest 1-minute dimensions. Compared with HRM, TBE showed specificity and positive predictive value of 100%, sensitivity of 31%, and negative predictive value of 58%. Interobserver agreement was excellent.
Conclusion: Standardized TBE is a highly specific and reproducible tool for confirming the functional relevance of EGJOO diagnosed by CCv4.0. Although sensitivity is modest, a positive TBE strongly correlates with manometric findings and helps differentiate EGJOO from achalasia. Integration into the diagnostic pathway supports accurate patient stratification and may avoid unnecessary treatment.
Limitations: Being a single-center study with relatively few positive cases may limit generalizability. Moreover, suboptimal sensitivity could underestimate EGJOO in milder presentations. Larger multicenter cohorts are required to validate findings and refine cutoff values.
Funding for this study: This research received no external funding
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013). The study’s protocol was approved by the local ethics committee at the University Hospital of Campania “L. Vanvitelli” and AORN “Ospedale dei Colli”.

Author Block: G. Balestrucci¹, M. Brunese¹, V. Patanè², N. Giordano², M. Alessandrella², S. Cappabianca², A. Reginelli²; ¹Campobasso/IT, ²Naples/IT
Purpose: To summarize the current role of imaging modalities in gastric cancer staging and highlight the emerging potential of magnetic resonance imaging (MRI) with advanced functional sequences and artificial intelligence (AI) applications.
Methods or Background: A systematic literature search was conducted in PubMed using predefined search strings. We included systematic reviews and meta-analyses published from January 2020 to December 2024 focusing on imaging techniques for gastric cancer staging. About 30 studies met the inclusion criteria, covering CT, PET/CT, endoscopic ultrasound, MRI, and AI-based approaches.
Results or Findings: Computed tomography (CT) remains the reference technique for baseline staging, although it has limited sensitivity in assessing tumor wall invasion (T-staging) and in detecting peritoneal metastases. PET/CT improves distant staging accuracy; innovative tracers such as Ga-FAPI have demonstrated significantly higher sensitivity than FDG-PET, particularly for peritoneal and nodal metastases. EUS remains the most accurate method for local T-staging and nodal sampling, but its performance is strongly operator-dependent.
Magnetic resonance imaging is emerging as a promising complementary tool, with DWI enabling improved differentiation between malignant and non-malignant tissues and perfusion MRI providing insights into tumor vascularity and microenvironment. Several studies reported superior accuracy of MRI compared to CT for local staging and prediction of treatment response. AI and radiomics approaches further expand diagnostic potential, showing promise in predicting tumor aggressiveness and therapy response.
Conclusion: MRI, especially when combined with functional sequences and AI, represents an emerging complementary modality for gastric cancer staging. It has the potential to improve diagnostic accuracy, refine patient stratification, and support personalized surgical and oncological management.
Limitations: Evidence remains limited, with heterogeneous imaging protocols and few prospective multicenter studies. Larger standardized investigations are warranted to validate the clinical role of MRI and AI in gastric cancer staging.
Funding for this study: None
Has your study been approved by an ethics committee? Not applicable
Ethics committee - additional information: