RPS 1107 - Female genitourinary imaging

Author Block: S. Bottazzi¹, L. Russo¹, D. Cusumano², R. Autorino¹, L. Boldrini¹, A. Amerighi¹, A. Rame¹, E. Sala¹, B. Gui¹; ¹Rome/IT, ²Olbia/IT
Purpose: Early Regression Index (ERI) is an image-based biomarker which combines gross tumour volume (GTV) measured at staging (Vpre) and at mid-therapy (Vmid) to quantify early tumour shrinkage. It showed promising results in predicting pathological complete response (pCR) in rectal and cervical cancer treated with neoadjuvant chemoradiotherapy. This study aims to evaluate the feasibility of ERI in predicting local control (LC) after cisplatinum-based chemotherapy and external beam radiotherapy (CRT) followed by brachytherapy in locally advanced cervical cancer (LACC), starting from T2-WI and apparent diffusion coefficient (ADC) maps.
Methods or Background: 91 patients with LACC (FIGO IB3-IVA) underwent MRI for staging and after CRT. GTV was delineated on the axial oblique T2-WI and ADC map and ERI was calculated on both sequences. Response to CRT and brachytherapy was evaluated six months after the end of treatment using MRI and PET/CT exams. LC was considered in case of the absence of residual disease.
The ERI performance was quantified by calculating the area (AUC) under the Receiver Operating Characteristic (ROC) curve and measuring sensitivity and specificity at the best threshold value.
Results or Findings: The performance of ERI-T2 (AUC=0.84; 95% CI 0.76-0.95) was superior to that reported by ERI-ADC (AUC=0.72; 95% CI 0.63-0.81). At the best cut-off threshold, ERI-T2 showed excellent specificity (100.0%) with limited sensitivity (67.4%), while ERI-ADC showed high specificity (87.2%) and low sensitivity (59.1%). At subgroup analysis, both ERI-T2 and ERI-ADC showed lower accuracy in adenocarcinoma (72.3% and 70.4%) compared to squamous subgroup (87.5% and 81.3%).
Conclusion: ERI is a promising biomarker in LACC treated with concurrent CRT. Combining ERI-T2 and ERI-ADC, it is possible to identify poor responders after CRT thus modulating the brachytherapy boost consequently.
Limitations: Retrospective single-center 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: Not applicable.

Author Block: R. V. Ninkova, A. Calabrese, F. Curti, S. Riccardi, M. Gennarini, V. Miceli, A. Cupertino, C. Catalano, L. Manganaro; Rome/IT
Purpose: The aim of this study is to explore the diagnostic accuracy, specificity, sensitivity, positive and negative predictive value (PPV and NPV) of Node-RADS in assessing the risk of metastatic lymph node (LN) involvement of cervical cancer (CC) patients using magnetic resonance imaging (MRI). We also evaluated the applicability and feasibility of the score among three different readers with different years of experience.
Methods or Background: From December 2014 to June 2023, 140 patients were treated at our department. 68 patients underwent staging MRI, neoadjuvant chemotherapy and radical surgery, and were included in our study. MRI images were retrospectively reviewed and Node-RADS scores were assigned for the main bilateral pelvic LN stations. LNs were evaluated in a scoring category of 1 to 5 to assess risk of metastatic involvement: 1 (very low), 2 (low), 3 (equivocal), 4 (high), 5 (very high).
The adequacy of the score with respect to the actual neoplastic involvement was then evaluated by comparing it with the results of histological examination.
Results or Findings: The resulting sensitivities, specificities, PPVs, NPVs, and accuracies were 100%, 45%, 56%, 100%, 68% for >1, 92.8%, 72.5%, 70.3%, 93.5%, 80.9% for >2, 71.4%, 100%, 100%, 83.3%, 88.2% for >3, 42.9%, 100%, 100%, 71.4%, 76.4% for >4. The inter-observer agreement between the Node-RADS scores assigned by the senior reader compared with the scores assigned by junior reader 1 and the scores assigned by the senior reader compared with junior reader 2 was 0.888 and 0.738, respectively.
Conclusion: The Node-RADS score is a valid system for correct and standardised evaluation of LN stations, increasing diagnostic accuracy particularly for patients with CC. It represents an easily applicable standardisation system that can be useful in clinical practice, improving the proper management of patients with CC.
Limitations: The main limitation is the patient sample and the retrospective nature of the 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: This study was conducted in accordance with the Declaration of Helsinki.

Author Block: J. Ren, Y-L. He, L. Mao, Y. Li, Z. Yuan, Z. Y. Jin, X. Li, H-D. Xue; Beijing/CN
Purpose: The study aimed to assess the value of CT-derived radiomics in preoperatively identifying ovarian clear cell carcinoma (CCC) subtype in epithelial ovarian cancer (EOC) using multicenter datasets.
Methods or Background: A total of 457 patients with EOCs were retrospectively recruited from three medical centres: Centre A (n=349), divided into training (n=280) and internal testing (n=69) sets based on surgery date, while Centre B (n=74) and Centre C (n=34) were external testing sets. Patients were grouped into CCC or non-CCC subtype based on postoperative pathology. We documented six clinical characteristics and assessed ten radiological characteristics of each lesion by two radiologists. Three models to identify CCC subtype in EOC were built: a radiomic signature using selected radiomic features, a traditional model with selected clinical and radiological characteristics, and an integrated model combining radiomic signature, clinical characteristics, and radiological characteristics. These models were constructed using JMIM feature selection and logistic regression in 10-fold cross-validation. Diagnostic performance was assessed in internal and external test sets.
Results or Findings: The mean age of the 457 patients was 53.7 ± 10.4 years. CCC accounted for 21.0% (96/457) of all EOC cases. The integrated model, which utilized the radiomic signature plus one clinical characteristic and three radiological characteristics, demonstrated better performance over the traditional model and the radiomic signature. The integrated model achieved AUCs of 0.890 in the internal testing set, 0.885 in external testing set 1, and 0.781 in external testing set 2, surpassing the traditional model with AUCs of 0.840, 0.884, and 0.735, and the radiomic signature with AUCs of 0.811, 0.836, and 0.778.
Conclusion: CT-derived radiomics analysis proved to be valuable in the identification of the CCC subtype in EOC, suggesting its potential to enhance subtype-specific therapeutic approaches.
Limitations: No limitations were identified.
Funding for this study: Funding was provided by Natural Science Foundation of China [grant No. 81901829]; and National High Level Hospital Clinical Research Funding [grant No. 2022-PUMCH-A-004 & 2022-PUMCH-A-109].
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: This retrospective study was approved by the local institutional review board, which waived the requirement for written informed consent.

Author Block: S. Bozer, D. Kuru Öz, G. A. Erden; Ankara/TR
Purpose: The ovarian-adnexal reporting and data system-magnetic resonance imaging (O-RADS MRI) guideline determines the malignancy risk of adnexal lesions and provides a common language between radiologists and clinicians. The aim of this study was to determine the diagnostic performance of the O-RADS MRI scoring system.
Methods or Background: Patients who underwent pelvic MRI with any clinical prediagnosis between January 2011-April 2023 were screened. Exclusion criteria were as fallows, history of malignancy or gynaecological surgery for non-malignant reasons, O-RADS 1 category lesions and contraindication for contrast administration. MR images were evaluated by two radiologists who were unaware of clinical and laboratory findings. Lesions with different scores were reviewed by consensus. Histopathologic results were accepted as the reference standard. Lesions that did not have a histopathologic diagnosis but regressed on follow-up imaging or remained stable for at least six months were considered benign. Lesions in the O-RADS 4 category and above were considered malignant. For statistical analysis, borderline tumours were included in the malignant category.
Results or Findings: A total of 209 adnexal lesions, of which 184 were benign, three were borderline, and 22 were malignant, were included in the study. The sensitivity, specificity, positive predictive value, negative predictive value and accuracy of the O-RADS MRI score were 100%, 97.2%, 83.3%, 100%, and 97.6%. Interobserver agreement was very high (k=0.95).
Conclusion: The sensitivity, specificity and overall accuracy of the O-RADS MRI score in this study were similar to those reported for the original score. Our findings support the use of the O-RADS MRI score in the evaluation of adnexal masses with high interobserver reliability without loss of diagnostic accuracy.
Limitations: The limitations of the study are small sample size, relatively low number of malignant lesions and retrospective nature.
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 committee.

Author Block: E. A. Kirillova¹, I. Mashchenko¹, P. Kozlova¹, L. Kashapova², T. Mamiseishvili², E. Shelepova¹, I. Zazerskaya¹, G. Trufanov¹; ¹St. Petersburg/RU, ²Moscow/RU
Purpose: Placentomegaly may be associated with severe foetal and maternal complications. The aim of this study was to provide useful diagnostic information on the differential diagnosis of placentomegaly and to describe the specific features of thrombohaematoma-associated cases.
Methods or Background: Placental thrombohaematomas (PTHs) represent rare cases of placentomegaly that are associated with poor pregnancy outcomes and can be classified as subamniotic, subchorionic, retroplacental or marginal haemorrhages.
Results or Findings: We encountered six cases of PTHs in pregnant women aged 29.50 ± 5.96 years. The diagnosis was unclear on a routine ultrasound exam and required additional confirmation using MRI of the placenta performed at 23.33 ± 2.58 weeks.

Unfavourable pregnancy outcome (foetal demise; n=3) included (1) acute massive retroplacental haematomas with a fluid-fluid level as a result of a sedimentation effect, (2) in two cases, all four types of PTHs were observed, (3) the mean gestational age at the time of delivery was 22 weeks.

Unfavourable pregnancy outcome (live births; n=3) involved (1) all three patients having subamniotic PTHs (massive [n=2] and local [n=1] PTHs) and placental villous oedema, (2) two cases of preterm birth at 27 GW (in both cases of massive PTHs), and one case of full-term birth at 37 GW, (3) the mean gestational age at the time of delivery was 31 weeks.

Some newborns demonstrated congenital anomalies (n=2), neurologic deficits and gastrointestinal complications (n=3 each). Respiratory support was required in three cases.
Conclusion: MRI of the placenta is a valuable tool for the differential diagnosis of placentomegaly. Early diagnosis is important for risk stratification and prenatal outcome prediction. Special caution should be exercised in case of thrombohaematoma-associated placentomegaly as this finding may be associated with a higher risk of preterm birth and poor neonatal outcomes.
Limitations: No limitations were identified.
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: Not applicable.

Author Block: C. Panico, G. Avesani, V. Simeon, M. Mangialardi, P. Sciaccotta, B. Gui, E. Sala; Rome/IT
Purpose: Appropriate management of patients with adnexal masses needs an accurate characterisation. Up to 20% of adnexal masses remains uncharacterised on US and MRI is needed for full characterisation. The O–RADS MRI score relies on intravenous gadolinium contrast. The purpose of this study was to show interim data of a multicentre prospective external evaluation of the non-contrast MRI adnex Score (NCMS) in characterising the adnexal masses.
Methods or Background: 73 patients from six Italian centres were evaluated. NCMS was used to predict the malignancy of the ovarian lesions. The score is based on five categories (four and five indicate malignancy). One senior and one junior radiologist per centre read the images, unaware of the patients' clinical information except for age. Presence of malignant tumours was verified during histopathology after surgery or during the radiological follow-up after 12 months.
Diagnostic accuracy in terms of the percentage of correctly classified diagnoses was measured. Agreement between senior and junior radiologist has been evaluated performing kappa statistics.
Results or Findings: 26 patients were classified as malignant. Malignant lesions were correctly classified in 78.08% of cases (Conf. Int. 68.38% - 86.92%). The agreement between senior and junior radiologist was 93.15% (under an expected agreement of 53.74%), with a kappa statistic of 0.79 (Conf. Int. 0.59-0.985).
Conclusion: The results on the accuracy and reproducibility of the diagnosis are in line with the potential use of the NCMS in clinical practice when contrast medium cannot be used for logistic or patient factors.
Limitations: The limitation of this study is the number of observations being low (only 20% of the number of observations needed according to the sample size calculation); the expansion of the number of cases in this multicentric study will overcome this problem.
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 proposal to the Ethical Committee of Fondazione Policlinico Gemelli (coordinating center) was presented on June 21st, 2021 and approved. Each centre had their participation approved by their own ethical committee.

Author Block: S. Park, S. E. Rha, H. Kim, Y. R. Shin; Seoul/KR
Purpose: The purpose of this study was to evaluate the diagnostic value of known qualitative magnetic resonance imaging (MRI) features and diffusion-weighted imaging (DWI)-based quantitative parameters in the differentiation of uterine low-grade endometrial stromal sarcoma (LGESS) from benign atypical leiomyoma (ALM).
Methods or Background: This retrospective study includes 113 women (ALM= 93, LGESS= 20) imaged with pelvic MRI with DWI prior to surgery. Two readers independently evaluated each lesion for conventional qualitative features as well as quantitative parameters including the mean ADC value and the relative contrast ratio (RCR) on high b-value DWI. The diagnostic performance of both qualitative and quantitative parameters in discriminating LGESS and ALM was assessed.
Results or Findings: The qualitative MRI features that significantly differed between LGESS and ALM were ill-defined or infiltrative margin (p=0.018), irregular or nodular shape (p<0.001), intratumoral low-SI band on T2WI (p<0.001), intramyometrial nodular extension (p< 0.001), nodule-in-nodule appearance (p= 0.017), and cystic changes (p= 0.010). The mean ADC value of LGESS (0.99 ± 0.29 x 10−3mm2/s) was significantly lower than that of ALM (1.25 ± 0.31 x 10−3mm2/s) (p< 0.001) and the mean RCR of LGESS was 8.3 ± 5.5, which was significantly higher than that of ALM, 4.2 ± 2.7 (p< 0.001).
Conclusion: Although there are substantial overlaps in conventional imaging features between LGESS and ALM, applying quantitative parameters based on DWI and ADC map in addition to qualitative imaging features may help distinguish LGESS from ALM.
Limitations: The study is retrospective. There are heterogeneous MRI acquisition parameters, but no major differences were detected in the protocols among different equipment. In addition, the study includes a relatively small number of patients with low-grade ESS. Nonetheless, it's important to highlight that low-grade ESS is an exceedingly rare tumor, and our research encompasses a more substantial number of cases compared to previous reports.
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 catholic university of Korea, IRB

Author Block: C. Li, M. Sun, R. Xu, J. Guan; Guangzhou/CN
Purpose: The purpose of this study was to recognise and analyse the cause of “cysts margination” sign in ovarian thecoma-fibroma groups (OTFG). And to investigate the value of this specific sign in the differential diagnosis between OTFG and other ovarian cystic and solid tumours.
Methods or Background: A total of 112 patients with surgically proven OTFG were screened consecutively from January 2016 to December 2022. Preoperative CT/MR examinations were reviewed by one radiologist with 20 years of experience in genital/reproductive imaging. The “cysts margination” sign was clearly recognised in 25 patients (group 1), while absent in the remaining 87 patients (group 2). Clinical data were recorded and compared between the two groups. Preoperative CT/MR imaging of 25 OTFG patients with “cysts margination” sign and 50 matched patients with surgically proven other ovarian cystic and solid tumours were retrospectively reviewed by two independent radiologists (five and nine years of experience) before and after learning the sign. The diagnostic accuracy of OTFG and other ovarian tumours was calculated.
Results or Findings: 22.3% (25/112) of patients with OTFG presented with “cysts margination” sign. Histopathology confirmed these peripheral cysts to be ovarian cysts. The age (56±15 vs 43±16, p= 0.02), menopausal state (p< 0.001), and presence of endometrium thickening (p< 0.001) were significantly different between the two groups. With the addition of the sign into radiological evaluation, the diagnostic accuracy of OTFG by two independent radiologists improved from 10% and 35% to 90% and 90%, respectively.
Conclusion: “Cysts margination” sign can be seen in around 20% of OTFG patients, mostly in peri- or post-menopausal state, and may be associated with the elevated oestrogen production of the tumour. Familiarity with this specific sign can significantly increase the diagnostic efficacy of OTFG in menopausal women, thereby avoiding unnecessary further examination or treatment.
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: The study was approved by the First Affiliated Hospital, Sun Yat-Sen University

Author Block: D. Kuru Öz, M. Araz, S. N. Yılmazer Zorlu, B. Varlı, C. Soydal, S. Taşkın, E. Ozkan, G. A. Erden, O. Kücük; Ankara/TR
Purpose: The aim of this study is to evaluate the diagnostic performance of F18-FDG PET/MRI in endometrial cancer (EC) staging, determine the individual contributions of PET and MRI, compare PET/CT and PET/MRI findings, quantitatively assess the primary tumor (PT) and metastatic lymph nodes (LN), and investigate the differences in quantitative measurements among histological subtypes.
Methods or Background: Between February 2020 and March 2022, a total of 44 patients who were diagnosed with EC via biopsy were included in the study. Following whole-body PET/CT, pelvic PET/MRI was performed. For quantitative analysis, the apparent diffusion coefficient (ADC) values for PT and LN were measured with MRI, while the maximum standard uptake value(SUVmax), metabolic tumor volume(MTV), and total lesion glycolysis(TLG) values were measured with PET/CT and PET/MRI. For statistical analysis Spearman correlation and Mann-Whitney U tests were performed.
Results or Findings: Out of the 44 patients, 39 (88.6%) had concordance between PET/MRI staging and pathological staging, while 5 (11.3%) were discordant.
There were nine patients (20.4%) with pelvic and/or para-aortic LN metastasis. In three patients (6.8%) with suspected LN involvement on MRI, PET accurately predicted LN involvement with pathological activity.
In LN evaluation, the sensitivity, specificity, positive predictive value, negative predictive value, and accuracy for PET/CT were 66.6%, 93.9%, 66.6%, 91.1%, and 88%, respectively. For PET/MRI, the corresponding values were 88.8%, 88.2%, 66.6%, 96.7%, and 88.3%, respectively.
For both PT and pathological LN, no strong correlation was found between ADC and SUVmax values (r> -0.009, p> 0.095,respectively).
Conclusion: PET/MRI, can be used in preoperative staging for the accurate assessment of both the primary tumor and nodal involvement. It offers high negative predictive value and has the potential to prevent unnecessary lymph node dissections while providing a high diagnostic performance as a hybrid imaging method.
Limitations: The limitation of this study is the small sample size.
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 Institutional Ethics Committee approved this retrospective study protocol and waived informed consent

Author Block: M. Tzoumpa¹, T. Siempis², G. Alexiou¹, M. Argyropoulou¹, A. C. Tsili¹; ¹Ioannina/GR, ²Belfast/UK
Purpose: The purpose of this study was to compare the diagnostic performance of multidetector CT (MDCT), MRI including diffusion-weighted imaging and FDG PET/CT in the detection of peritoneal metastases (PMs) in ovarian cancer (OC) patients.
Methods or Background: A comprehensive search was performed for articles published from January 2000 to February 2023, in PubMed database. Inclusion criteria were the following: diagnosis/suspicion of PMs in patients with ovarian/fallopian/primary peritoneal cancer; initial staging or suspicion of recurrence (primary outcome); MDCT, MRI and/or FDG PET/CT performed for the detection of PMs; population of at least 10 patients; at least three data sets available for each imaging modality; surgical results, histopathologic analysis and/or radiologic follow-up, used as reference standard; per-patient and per-region data; and data for calculating sensitivity and specificity reported.
Results or Findings: Thirty-three studies were assessed, including 487 women with OC and PMs. On a per-patient basis, MRI (P = 0.03) and FDG PET/CT (P < 0.01) had highest sensitivity compared to MDCT. MRI and PET/CT had comparable sensitivities (P = 0.84). On a per-lesion analysis, no differences in sensitivity estimates were noted between MDCT and MRI (P = 0.25), MDCT and FDG PET/CT (P = 0.68), MRI and FDG PET/CT (P = 0.35).
Conclusion: Based on our results, FDG PET/CT and MRI had a higher diagnostic performance in the detection of PMs compared to MDCT, on a per-patient analysis. No differences were found on a per-lesion basis.
FDG PET/CT and MRI can be considered equivalent alternatives for the detection of PMs in OC. MDCT can be used as an alternative.
Limitations: The limitations of this study are due to selection bias, and the heterogeneity of the studies.
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: No information provided by the submitter.

Author Block: G. Avesani, L. D'Erme, A. Iacono, C. Panico, S. Bottazzi, F. Capomacchia, C. Marchetti, B. Gui, E. Sala; Rome/IT
Purpose: The aim of this study was to evaluate the radiological distribution of disease at the first recurrence in patients with advanced ovarian cancer treated with different maintenance therapies (PARP-inhibitors versus Bevacizumab).
Methods or Background: We retrospectively collected data on patients with recurrent advanced ovarian cancer from January 2017 to December 2022, who had III-IV FIGO stage at the diagnosis, a complete debulking (primary or interval surgery) followed by adjuvant chemotherapy and maintenance therapy. We included those with an available contrast-enhanced CT at the recurrence time (clinical, serological or radiological).
We evaluated whether the presence of recurrence was visible (visible/non-visible solid tissue), the type (macronodular/micronodular) and the number of localisations (oligometastatic (<4 sites)/multi-metastatic) and if the relapse was only in lymph nodes.
Results or Findings: We evaluated 80 patients; 40 had Bevacizumab and 40 had PARP-i.
The incidence of recurrence without any clearly visible soft tissue was significantly higher during Bevacizumab (20% versus 10%; p= 0.04). Micronodular diffuse relapse was more frequent in patients treated with Bevacizumab (58% versus 30%; p= 0.03). Oligometastatic recurrence was more frequent in PARP-i (45% versus 24%; p= 0.03). No difference was found in only nodal relapse between the two groups (18% versus 15%, p> 0.05).
Conclusion: The recurrence pattern in ovarian cancer differs based on the administered maintenance therapy. Secondary cytoreduction, which can be performed in oligometastatic disease, is more frequent during PARP-i maintenance. Radiologists should know this to make the correct diagnosis, especially avoiding delayed diagnosis of recurrence in patients treated with Bevacizumab.
Limitations: This is a retrospective study with a small sample size.
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 with the approval code: 6031