RPS 2016 - The evolving landscape of imaging for neuroendocrine tumours and sarcomas

Author Block: M. Yuan; Zhengzhou/CN
Purpose: To explore the value of spectral CT in characterizing gastric stromal tumors and extra-gastric stromal tumors by combining clinical and conventional imaging features and spectral CT quantitative parameters.
Methods or Background: The clinical data of 55 patients with gastrointestinal stromal tumor (GIST) scanned by spectral CT from March 2022 to November 2023 were retrospectively collected from the First Affiliated Hospital of Zhengzhou University, which were classified into the gastric stromal tumor (GST) group (n = 42) and the extra-gastric stromal tumor (E-GST) group (n = 13) according to the location of the tumor. The clinical, conventional imaging features and spectral CT quantitative parameters of the two groups were compared and analyzed, and the differences between the two groups were compared by univariate analysis.
Results or Findings: Compared to the E-GST group, only CD34 showed a significant difference in clinicopathological indexes (P < 0.05), while no significant differences were found in other demographics and biomarkers. Regarding conventional imaging features, significant differences were observed in lesion diameters, shape, ulceration, enhancement homogeneity, and growth pattern (P < 0.05), but not in peritumoral fat, necrosis, calcification, or lymph node status. Among spectral parameters, IC, NIC, CT values at 40–140 keV, λHU, and Zeff in both arterial and venous phases differed significantly (P < 0.05), whereas ED values did not.
Conclusion: GST and E-GST have many similarities and differences in pathology and clinical presentation. Spectral CT is of high clinical value for quantitative and qualitative assessment of imaging features of both.
Limitations: this is a small single-center study with limited E-GST cases
Funding for this study: None
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: None

Author Block: J. Sun¹, Z. Zhou², T. Xie², W. Liu², Y. Chen², Y. Wang²; ¹FuZhou/CN, ²Shanghai/CN
Purpose: To preoperatively predict the Ki-67 index of pancreatic neuroendocrine tumors (pNETs) using quantitative and qualitative variables obtained from dual-layer spectral detector CT (DLCT).
Methods or Background: This retrospective study assessed 196 patients with pathologically confirmed pNETs who underwent DLCT between November 2020 and December 2024. Patients were randomly divided into a training set (n=138) and a testing set (n=58) . Quantitative DLCT parameters included iodine concentration (IC), VMI40keV, VMI70keV, and effective atomic number (Zeff) during arterial phase (AP) and venous phase (VP). Morphological CT features included tumor location, shape, margin, heterogeneity, calcification, necrosis, parenchymal atrophy, and main pancreatic duct dilatation. Patients were stratified into low- and high-proliferation groups using a 5% Ki-67 cutoff. Univariate analysis identified significant predictors, followed by forward stepwise binary logistic regression for further screening and prediction model construction. A multiparametric model was visualized as a nomogram, and its performance was evaluated using ROC analysis.
Results or Findings: A total of 196 patients (124 low- and 72 high-proliferation; median age 54.3 years, IQR 44.2–62.3) were included. Multivariate analysis revealed that heterogeneity (OR = 3.96, 95% CI: 1.77-9.08, p = 0.001) and Zeff_PP (OR = 0.24, 95% CI: 0.12-0.47, p < 0.001) were independent predictors. The nomogram combining Zeff_PP and heterogeneity achieved the highest AUCs of 0.808 (95% CI: 0.731-0.886) and 0.809 (95% CI: 0.686-0.932) for the training and testing sets, significantly exceeding the morphological CT model (p=0.0009 and p=0.0014, respectively).
Conclusion: DLCT-derived parameters, particularly when combined with morphological CT features in a nomogram, provide valuable noninvasive preoperative prediction of Ki-67 stratification in pNETs, which may aid in surgical planning and treatment stratification.
Limitations: Single-center retrospective design and limited sample size
Funding for this study: This work was supported by the following projects: National Nature Science Foundation of China (82471981); Shanghai Science and Technology Innovation Action Plan and Hongkong, Macao and Taiwan Science and Technology Cooperation Project (22490760800); Artifical Intelligence Medical Hospital Cooperation Project of Shanghai Xuhui District Science and Technology Commission (23XHYD-13); Shanghai Anticancer Association Soar Project (SACA-AX202309); Shanghai Anticancer Association Eyes Project (SACA-CY23B05); Clinical Research Special Project of Shanghai Municipal Health Commission (202340123).
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: This retrospective study was approved by the Institutional Review Board.

Author Block: J. Sun¹, Z-R. Zhou², T. Xie², W. Liu², Y. Chen², Y. Wang²; ¹FuZhou/CN, ²Shanghai/CN
Purpose: To predict the histological grade of pancreatic neuroendocrine tumors (pNETs) preoperatively using quantitative and qualitative variables obtained from dual-layer spectral detector CT (DLCT).
Methods or Background: This retrospective study assessed 177 patients with pathologically confirmed pNETs who underwent DLCT between November 2017 and December 2024. For quantitative analysis, both conventional and spectral parameters, such as effective atomic number (Zeff) and iodine concentration (IC), were measured in each phase. The iodine enhancement fraction (IEF_PP/VP) was calculated as the ratio of IC in pancreatic parenchymal phase to venous phase. Normalized IC (nIC) was derived from the lesion-to-aortic IC ratio at the same slice level. Qualitative variables of necrosis and enhancement uniformity were evaluated on virtual mono-energetic images (VMIs) at 40 keV. Univariate analysis identified significant predictors, followed by forward stepwise binary logistic regression for further screening and prediction model construction. A multiparametric logistic regression model integrating spectral parameters was constructed. Diagnostic performance was evaluated using receiver operating characteristic (ROC) analysis.
Results or Findings: Among 177 pNETs, 62 (35%), were grade G1 and 115 (65%) were grade G2 and G3. Independent predictors included necrosis (OR = 3.546, p < 0.001), enhancement uniformity (OR = 0.103, p < 0.001), Zeff_VP (OR = 0.141, p = 0.001), and IEF_PP/VP (OR = 0.208, p < 0.001). IEF_PP/VP (AUC = 0.708, p = 0.016) exhibited better predictive performance than Zeff_VP (AUC = 0.663), enhanced uniformity (AUC = 0.680), and necrosis (AUC = 0.523). The combination model integrating necrosis, enhancement uniformity, Zeff_VP, and IEF_PP/VP achieved superior diagnostic efficacy, with an AUC of 0.822 (p < 0.001), sensitivity of 88.7%, and specificity of 66.1%.
Conclusion: DLCT-derived spectral parameters are valuable for noninvasive preoperative prediction of pNET histological grade.
Limitations: Single-center retrospective design and limited sample size
Funding for this study: This work was supported by the following projects: National Nature Science Foundation of China (82471981); Shanghai Science and Technology Innovation Action Plan and Hongkong, Macao and Taiwan Science and Technology Cooperation Project (22490760800); Artifical Intelligence Medical Hospital Cooperation Project of Shanghai Xuhui District Science and Technology Commission (23XHYD-13); Shanghai Anticancer Association Soar Project (SACA-AX202309); Shanghai Anticancer Association Eyes Project (SACA-CY23B05); Clinical Research Special Project of Shanghai Municipal Health Commission (202340123)
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: This retrospective study was approved by the Institutional Review Board

Author Block: P. Summer¹, P. Sorgo¹, P. M. Sbeghen², A. Y. Yasin¹, S. Pötter-Lang¹, N. Bastati-Huber¹, A. Ba-Ssalamah¹, U. I. Attenberger¹; ¹Vienna/AT, ²Verona/IT
Purpose: Intrapancreatic ectopic spleen (IPES) is a benign entity that often mimics pancreatic neuroendocrine tumors (NETs) on imaging. Conventional CT, ECCM-MRI, Ga-68 DOTA-NOC PET/CT, and Tc-99m HDRBC SPECT have important diagnostic limitations. Superparamagnetic iron oxide (SPIO)–enhanced MRI may allow reliable noninvasive differentiation. We evaluated the diagnostic performance of multimodal imaging with emphasis on SPIO-MRI.
Methods or Background: This retrospective study included 23 patients (10 NET, 13 IPES; mean age 59 ± 14 years) who underwent ECCM-MRI and SPIO-MRI. Subsets also had CECT (n=22), Ga-68 DOTA-NOC PET/CT (n=13), and Tc-99m HDRBC SPECT (n=6). Two blinded abdominal radiologists assessed qualitative and quantitative imaging features and diagnostic confidence. ROC analyses were performed to evaluate modality performance.
Results or Findings: NETs were significantly larger than IPES (1.9 vs. 1.3 cm, p<0.05). ECCM-MRI and CECT showed overlapping enhancement patterns with low accuracy (~50%). Ga-68 DOTA-NOC PET/CT demonstrated higher uptake in NETs compared with IPES (SUVmax 82 vs. 36, p=0.06), but with poor discriminatory value. Tc-99m HDRBC SPECT was inconclusive due to limited spatial resolution. In contrast, all IPES lesions showed marked T2/T2* signal loss after SPIO administration, whereas NETs remained unchanged. Both readers achieved 100% accuracy, with excellent interobserver agreement (κ=0.95, ICC=0.91). ROC analysis confirmed SPIO-MRI as the only modality with near-perfect diagnostic performance (AUC 0.79–1.00, p<0.001).
Conclusion: SPIO-enhanced MRI enables confident, noninvasive differentiation between IPES and NET, clearly outperforming conventional imaging and nuclear medicine techniques. It should be implemented as the preferred next step in suspected IPES to avoid unnecessary interventions.
Limitations: This retrospective, single-center study included a small cohort with uneven availability of comparator imaging modalities, limiting generalizability. The reported 100% accuracy of SPIO-MRI requires prospective validation in larger, multicenter cohorts.
Funding for this study: No funding.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: Protocol Number: 1441/2016

Author Block: L. Carone, M. G. Di Niso, S. Kalantari, M. Zacchino, F. Sirchia, L. Preda; Pavia/IT
Purpose: NF1 is a rare inherited genetic syndrome characterized by wide phenotypic variability and a predisposition to tumor development. In recent years, WB-MRI has emerged as a valuable imaging tool for the comprehensive evaluation of multisystem alterations associated with NF1 and for the early detection of incidental findings of potential clinical relevance. This prospective, monocentric observational study aimed to assess the role of WB-MRI in the characterization of clinically relevant radiological findings (ONCO-RADS) in adult NF1 patients, promoting a tailored approach.
Methods or Background: Since July 2024, 20 adult patients with NF1 (mean age 36 years, range 18–68) referred to the Medical Genetics Unit of Policlinico San Matteo, Pavia, underwent WB-MRI from head to toe. Each case was discussed by a multidisciplinary team to evaluate disease-related alterations and define appropriate management.
Results or Findings: Plexiform neurofibromas (PN), with variable size and location, were identified in 11/20 patients (55%), of which 6 were newly diagnosed.
Other extracranial findings included cutaneous and subcutaneous neurofibromas (100%), skeletal alterations (60%), thyroid nodules (20%), splenomegaly (15%), diffuse neurofibromas (15%), and bullous dystrophy (5%).
In one patient, an early-stage malignant lung tumor was diagnosed, allowing timely treatment. In another patient with a symptomatic inoperable PN, experimental treatment with Selumetinib was proposed.
Conclusion: This case series confirms the established role of WB-MRI as a safe tool for surveillance and multidisciplinary management of NF1-related manifestations, enabling the detection of previously unknown and clinically relevant lesions.
Limitations: Small cohort study; limited observation period; only preliminary restults are avalaible..
Funding for this study: No
Has your study been approved by an ethics committee? Not applicable
Ethics committee - additional information:

Author Block: L. Alonzo¹, R. Cannella¹, R. Laudicella², V. Benfante¹, P. Purpura¹, G. Micci¹, M. Galia¹, G. Brancatelli¹, P. Alongi¹; ¹Palermo/IT, ²Messina/IT
Purpose: The diagnosis of Neuroendocrine Tumors (NETs) can be challenging because of small lesion size, indolent growth, and variable metabolic activity, particularly in cases of unknown primary origin (CUP-NETs). Early and accurate identification of the primary lesion is essential for therapeutic decisions, including surgery, peptide receptor radionuclide therapy (PRRT), or somatostatin analogues.
This systematic review investigated the diagnostic accuracy and clinical impact of molecular imaging, with particular emphasis on PET/CT with radiolabeled somatostatin analogs, in patients with suspected NETs and CUP-NETs.
Methods or Background: A comprehensive search of PubMed and Scopus was performed following PRISMA guidelines. Twenty-one clinical studies published between 2010 and 2024 were included, encompassing 1,857 patients (420 CUP-NETs and 452 suspected NETs). Extracted data included study design, radiotracers, detection rates, and changes in patient management. Methodological quality was assessed using CASP tools.
Results or Findings: PET/CT with [68Ga]Ga-DOTA-peptides consistently outperformed conventional [111In]Pentetreotide SPECT. Reported detection rates for primary tumors in CUP-NETs ranged from 38% to 83% with PET/CT, compared with less than 10% with SPECT. In suspected NETs, PET/CT achieved sensitivity up to 95% and specificity above 85%, leading to therapeutic changes in up to one-third of cases. Additional diagnostic yield was provided by dual-tracer protocols combining [68Ga]Ga-DOTATATE with [18F]FDG and by newer tracers such as [18F]DOPA and [18F]-OC, particularly for aggressive variants and rare subtypes (e.g., paragangliomas, insulinomas). Detection of unknown primaries frequently enabled curative surgery or more tailored systemic treatment.
Conclusion: [68Ga]Ga-DOTA-peptide PET/CT is a key tool in the diagnostic evaluation of suspected and unknown primary NETs. Beyond its superior sensitivity compared with conventional imaging, it significantly influences clinical decision-making and should be integrated into standard diagnostic algorithms.
Limitations: Several retrospective, single-center, and with small sample sizes articles, which may limit generalizability.
Funding for this study: None
Has your study been approved by an ethics committee? Not applicable
Ethics committee - additional information:

Author Block: H. Zhang, Y. Zhou; Shanghai/CN
Purpose: To develop and validate a nomogram combining ADC histogram parameters with conventional MR features for preoperative differentiation between endometrial stromal sarcoma (ESS) and cellular leiomyoma (CL).
Methods or Background: This retrospective, dual-center, diagnostic accuracy study received institutional review board approval. The reference standard was postoperative histopathology. A total of 155 women were included (ESS: n=57, CL: n=98), split into a derivation cohort (n=111 from Hospital A) and an external validation cohort (n=44 from Hospital B). All underwent preoperative contrast-enhanced pelvic MRI. A nomogram was built using stepwise logistic regression on MRI features and ADC histogram parameters and was evaluated with receiver-operating characteristic curve analysis.
Results or Findings: A total of 155 women (mean age, 44 years ± 9 [standard deviation]) were analyzed. A nomogram incorporating irregular margin, cystic change, and mean ADC value was developed. The nomogram achieved an area under the curve (AUC) of 0.828 (95% CI: 0.742, 0.911) in the derivation cohort and 0.873 (95% CI: 0.768, 0.977) in the external validation cohort. At an optimal threshold of 0.477, the nomogram showed a sensitivity of 78.5% and a specificity of 87.9% in the derivation set, and 87.5% and 75.0% in the validation set, respectively.
Conclusion: The nomogram integrating ADC histogram parameters and MR features provides an accurate tool for the preoperative differentiation of ESS from CL.
Limitations: It is a retrospective stufy.
Funding for this study: This study has received funding by the Science and Technology Commission of Shanghai Municipality (Project No. 25ZR1401036).
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: review board of the obstetrics and gynecological hospital, Medical College, Fudan University.

Author Block: F. Schön, P. Günther, A. Weidlich, J. Kirchberg, M-L. Kromrey, R-T. Hoffmann, H. Fritzsche, J-P. Kühn; Dresden/DE
Purpose: This study investigated the potential of quantitative magnetic resonance imaging (MRI) to differentiate benign from malignant soft tissue tumors.
Methods or Background: Between 10/2022 and 01/2025, patients with soft tissue tumors were prospectively enrolled for quantitative MRI prior to histopathological confirmation. All 3.0T MRI examinations included chemical shift-encoded MRI, T1-/T2-mapping, and diffusion-weighted imaging (intravoxel incoherent motion [IVIM]). All MRI datasets were post-processed to generate R2*-, PDFF- (proton density fat fraction), T1-/T2-maps, and Diffusion (IVIM D), Pseudo-Diffusion (IVIM D*), Perfusion-Fraction (IVIM f). Histopathological diagnosis was defined as standard of reference. Mean values of two readers of each map were compared using Mann-Whitney U tests. A logistic regression model was established to predict malignancy.
Results or Findings: Fifty-nine patients (30 women; mean age 59.1+/-15.5 years) were examined, including 23 with malignant and 36 with benign tumors. Malignant tumors revealed significantly lower values for R2* (13.5+/-25.9 vs. 42.7+/-29.2; p<0.001), PDFF (29.3+/-38.1 vs. 60.2+/-42.4; p=0.003), IVIM D (1221.5+/-716.6 vs. 683.4+/-810.7; p=0.007), and IVIM D* (26.0+/-48.2 vs. 107.0+/-170.6; p=0.004) compared to benign tumors. In contrast, T1 relaxation times were significantly higher (1843.0+/-1065.4 vs. 1113.6+/-948.3; p=0.031). No significant differences were observed for IVIM f and T2 relaxation times (p>0.05). Univariate logistic regression analyses identified R2*, PDFF, IVIM D, and T1 as predictors for malignancy (p<0.05). Following backward elimination, R2* remained as independent variable in the multivariate model (p=0.002; odds ratio=0.96), yielding an AUC of 0.800 with a Youden-Index of 0.539.
Conclusion: By enabling non-invasive differentiation of soft tissue tumors, quantitative MRI - particularly through the independent predictive value of R2* - may serve as a valuable tool to determine lesion dignity in clinical practice.
Limitations: Limited sample size and single-center setting.
Funding for this study: This work was supported by the Clinical Scientist Program MeDDrive of the University Medicine Dresden.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: Approved by the local ethics committee of the Technical University Dresden (BO-EK-342072022_2)

Author Block: A. K. Nawaz, S. Ahmed, K. Siddique; Lahore/PK
Purpose: - To illustrate diverse diverse CT appearances of pulmonary metastasis in osteosarcoma.
- To correlate imaging appearances with their frequency in clinical cohort.
- To highlight the role of CT in staging, surveillance and treatment response assessment.
Methods or Background: Lungs are the most common site of metastasis in osteosarcoma. While soft tissue and calcified nodules are well known, atypical patterns such as cavitary, pleural or endobronchial disease are often underappreciated. Pulmonary metastasis in osteosarcoma are not just nodules - they represent a spectrum every radiologist must recognize.
For this purpose, we retrospectively reviewed chest CTs of 46 patients with histologically confirmed osteosarcoma and pulmonary metastasis betrween 2023-2025. Imaging appearances were categorized into six groups: soft tissue nodules, calcified nodules, cavitary lesions, pleural involvement, endobronchial disease and mass like lesions.
Results or Findings: Soft tissue nodules were frequent finding, observed in 15 patients (32.6%). Calcified metastases were present in 14 (30.4%), pleural involvement in 11 (23.9%) and mass like lesions or consolidations in 5 (10.8%). Cavitary lesions and endobronchial disease were less common, identified in 2 patients each (4.3%).
Conclusion: Pulmonary metastases in osteosarcoma demonstrate a wide CT spectrum. Beyond the classic nodule, recognizing calcified, cavitary, pleural and airway patterns is essential for accurate staging, guiding treatment decisions and monitoring disease progression.
Limitations: None
Funding for this study: None
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: This retrospective anonymized review was conducted in accordance with our institutional IRB guidelines. No additional patient consent was required.

Author Block: A. K. Nawaz, M. Nawab, A. Usman, K. Siddique; Lahore, Pakistan/PK
Purpose: Ewing sarcoma is typically a bone tumor, but extra skeletal forms, though uncommon, can occur in diverse anatomical sites. Early recognition of these rare presentations is essential for accurate diagnosis and management. The objective of this study was to analyze the anatomical distribution and imaging characteristics of 12 histologically proven extra skeletal Ewing sarcoma cases from 2020–2025.
Methods or Background: In this retrospective study, we analyzed 208 cases of histologically proven cases of Ewing Sarcoma from January 20202 to January 2025. From these 208 cases, 12 extra skeletal cases were identified after excluding lesions in the upper limb, lower limb, mandible, and thorax. Demographic, anatomical, and imaging data were reviewed.
Results or Findings: Extra skeletal disease represented 5.8% of the cohort. Sites included: head & neck (5, 41.7%), pelvis (3, 25%), abdomen/retroperitoneum (2, 16.7%), thyroid (1, 8.3%), and uterus (1, 8.3%). Patients ranged from 14–32 years; male-to-female ratio was 1.3:1. Imaging revealed well-defined soft tissue masses without bone involvement.
Conclusion: Although rare, extra skeletal Ewing sarcoma can involve head & neck, pelvis, abdomen, thyroid, and uterus. Recognizing these histologically proven atypical sites is vital for timely diagnosis and guiding therapy. This five-year review highlights their frequency, distribution, and imaging features, emphasizing the need for awareness among radiologists and oncologists.
Limitations: None
Funding for this study: None
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: This retrospective anonymized review was conducted in accordance with our institutional IRB guidelines. No additional patient consent was required.

Author Block: B. J. Kim, M. J. Kim, J. Y. Kim; Daegu/KR
Purpose: To date, there are no studies examining radiologic findings based on histologic patterns of primary hepatic angiosarcoma. This study aims to elucidate radiologic findings of primary hepatic angiosarcoma according to distinct histologic patterns.
Methods or Background: From January 2010 to October 2024, a cohort of 17 patients was included in this study. All patients were pathologically diagnosed with primary hepatic angiosarcoma and underwent CT with or without MRI. Histologic patterns were categorized into mass-forming and non-mass forming groups. The mass-forming type was subdivided into vasoformative and non-vasoformative (epithelioid, spindled) patterns, while the non-mass forming type was divided into sinusoidal and peliotic patterns. Two radiologists independently reviewed the CT and MRI images, categorizing the lesions into diffusely infiltrating and mass-forming types. Additionally, the presence of hypervascular portions and targetoid patterns was analyzed. The correlation between histologic patterns and radiologic findings was assessed using Fisher’s exact test.
Results or Findings: The mass-forming pattern was observed in 13 patients (76.5%), while the non-mass-forming pattern was identified in 4 patients (23.5%). There was a significant correlation (P<.05) between the radiologically classified patterns of diffusely infiltrating and mass-forming lesions and their pathological counterparts. However, the pathological classification of vasoformative and non-vasoformative lesions did not show a significant correlation with the presence of a hypervascular portion on imaging.
Conclusion: The pathological classification of hepatic angiosarcoma into mass-forming and non-mass-forming patterns closely corresponds to the radiological classification of mass-forming and diffusely infiltrating lesions, indicating that pathological findings are well reflected in imaging characteristics.
Limitations: Due to its retrospective design and relatively small number of patients, there may be inherent selection and sampling biases, and in cases with mixed histologic patterns, both histologic classification and radiologic interpretation were based on the predominant component, potentially limiting accuracy.
Funding for this study: The histologic pattern of primary hepatic angiosarcoma may be reflected in imaging as diffusely infiltrating or mass-forming appearances.
Has your study been approved by an ethics committee? Not applicable
Ethics committee - additional information:

Author Block: S. Kaur, S. S. Victory, M. S. Swarup, G. Sindhwani, C. Paruthi, R. Kanaujiya, R. G. Ghasi, A. Malik; Delhi/IN
Purpose: To present the clinical spectrum and imaging features of rhabdomyosarcoma in children resulting in early diagnosis and better patient outcome.
Methods or Background: The retrospective analysis was done at a tertiary care institute providing comprehensive radiological services from August 2023 to April 2025. Cases of rhabdomyosarcoma were assessed and findings compiled.
Results or Findings: A wide spectrum of rhabdomyosarcoma presentations was observed across diverse anatomical locations.
• Head and Neck: Orbital lesions presented with rapidly progressive proptosis and soft tissue masses; non-parameningeal sites revealed well-defined masses with local infiltration, whereas parameningeal tumors demonstrated aggressive features with skull base and intracranial extension.
• Genitourinary System: Bladder involvement was identified with intraluminal polypoid masses causing obstructive symptoms. Testicular rhabdomyosarcoma in males presented as painless testicular enlargement, while vaginal involvement in females appeared as polypoid soft tissue masses often misdiagnosed clinically.
• Perianal Region: Rare cases manifested as infiltrative perianal soft tissue masses, mimicking other benign and malignant conditions.
Conclusion: Rhabdomyosarcoma is a biologically diverse pediatric sarcoma with varied anatomical presentations. Cross-sectional imaging remains pivotal in accurate diagnosis, staging, and treatment planning, thereby guiding multidisciplinary management and improving overall outcomes.
Limitations: No limitations were identified for this 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:

Author Block: M. Ben salah Faria, G. Raia, F. Peier, L. Widmer, H. Thoeny; Fribourg/CH
Purpose: Our systematic review and meta-analysis aim to evaluate the diagnostic performance of Node Reporting and Data System 1.0 (Node-RADS) for identifying lymph node (LN) metastases, using histopathology as a reference standard, across a range of abdominal and pelvic malignancies.
Methods or Background: This review was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. PubMed, ScienceDirect and Cochrane Library databases were systematically searched until August 2025. We included studies analysing patients diagnosed with an abdominal or pelvic malignancy, who underwent contrast-enhanced CT or MRI for loco-regional work-up, with LN assessment by Node-RADS, associated with LN histopathological analysis. Study quality was assessed by modified Quality Assessment of Diagnostic Accuracy Studies (QUADAS-2) tool. Diagnostic performance was estimated by the random-effects model. Forest plots of pooled sensitivity and specificity were generated for two different Node-RADS thresholds.
Results or Findings: Eighteen retrospective studies were included, comprising 10 using CT and 8 using MRI. These studies covered a variety of malignancies: gynecologic (endometrial, ovarian, cervical), urologic (prostate, bladder, renal), and gastrointestinal cancers (colon, rectal, gastric, cholangiocarcinoma, periampullary). When applying a Node-RADS threshold of ≥3 to define malignant LN, the pooled sensitivity and specificity were 0.80 (95%CI, 0.69-0.87) and 0.79 (95%CI, 0.68-0.87), respectively. For a threshold of ≥4, sensitivity decreased to 0.60 (95%CI, 0.50-0.69), while specificity improved to 0.94 (95%CI, 0.90-0.96). The hierarchical summary receiver operating characteristic (HSROC) analysis showed area under the curves (AUCs) of 0.85 (95%CI, 0.79-0.89) and 0.87 (95%CI, 0.80-0.92) for thresholds ≥3 and ≥4, respectively.
Conclusion: Node-RADS score demonstrated reliable diagnostic performance in assessing lymph node malignancy across abdomino-pelvic malignancies, allowing a standardized approach that enhances reproducibility and supports clinical decision-making in oncologic staging.
Limitations: Heterogeneity of the included studies, which are of retrospective design.
Funding for this study: None
Has your study been approved by an ethics committee? Not applicable
Ethics committee - additional information:

Author Block: W. Zhao, D. Han, B. He; Kunming/CN
Purpose: Pleural mesothelioma (PM) is a rare malignancy with a poor prognosis. This study aimed to develop and validate a habitat subregion radiomics model for predicting the prognosis and overall survival (OS) of PM under immunotherapy.
Methods or Background: The PM patients with immunotherapy from two medical centers were enrolled as the training cohort (119 cases), while patients from a third medical center served as the external validation cohort (48 cases). Based on unenhanced CT images, tumor clustering was performed to extract highly correlated radiomic features from habitat subregions. A habitat subregion radiomic model was constructed using the Cox proportional hazards model to predict OS at 6, 12, 18, and 24 months, respectively. Furthermore, a clinical-habitat radiomic combined model and nomogram were developed, incorporating both clinical factors and radiomic features. Finally, the clinical-habitat radiomics model was validated using the external validation cohort.
Results or Findings: The clinical-habitat subregion radiomics combined model outperformed both the pure habitat radiomics model and the clinical model in predicting prognostic outcomes. The clinical-habitat subregion radiomics combined model was visualized as a nomogram, with OS at 6, 12, 18 and 24 months yielding AUC values of 0.807, 0.826, 0.846, and 0.839 in the training set, and 0.769, 0.809, 0.769, and 0.874 in the validation set, respectively.
Conclusion: The clinical-habitat radiomics model showed strong potential for predicting OS in patients with PM undergoing immunotherapy and provided insights into imaging and biological markers potentially implicated in immunotherapy-related tumor progression.
Limitations: First, as a retrospective study, it is prone to selection bias. Second, some patients in this study did not undergo enhanced CT imaging, so radiomic features were extracted only from non-enhanced CT images.
Funding for this study: National Natural
Science Foundation of China(82360344).
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: Approval NO. 2020-L-27,Habitat Subregion Radiomics for Predicting Overall Survival in Immunotherapy-Treated Pleural Mesothelioma: A Multicenter Study