RPS 406 - Moving forward in clinical molecular oncologic imaging

Author Block: J. H. R. Cairns¹, S. J. Arain¹, A. Challapalli², A. Bahl², T. Ali³, J. Sage⁴, R. Frood¹, G. Petrides³, A. Scarsbrook¹; ¹Leeds/UK, ²Bristol/UK, ³Newcastle/UK, ⁴Oxford/UK
Purpose: Prostate-specific membrane antigen (PSMA) PET/CT is widely used in prostate cancer management. Indeterminate bone lesions (IBLs) remain a diagnostic challenge when using [18F]PSMA-1007. The study compared the bone uptake metastatic probability (BUMP) score - a composite model combining imaging and clinical parameters with PSMA Reporting and Data System (PSMA-RADS) Version 2.0 for predicting skeletal metastases in IBLs.
Methods or Background: [18F]PSMA-1007 PET/CT scans from three tertiary cancer centres were retrospectively analysed. Bone lesions with focal PSMA uptake were evaluated. For each lesion SUVmax, CT-derived mean Hounsfield unit and androgen deprivation therapy status were recorded to calculate the BUMP score. Lesional PSMA-RADS scores were determined by experienced radiologists. Ground truth was established using follow-up imaging, prostate specific antigen (PSA) kinetics and multidisciplinary consensus. Receiver operating characteristic (ROC) analysis was used to assess diagnostic performance of both methods.
Results or Findings: In total 136 patients (median age 70.5) with 427 bone lesions were analysed; 58.8% (n=251) of lesions were metastatic. Median SUVmax for malignant lesions was 7.55 (Interquartile range (IQR) 10.3) versus 3.53 (IQR 1.51) for benign lesions. BUMP achieved an area under the ROC curve (AUROC) of 0.81 (95% confidence interval (CI): 0.78 - 0.85) with an optimal threshold at 0.125 yielding an F1 score of 0.77. PSMA-RADS 2 outperformed BUMP with higher sensitivity and specificity (98% and 96%, versus 74% and 75% respectively). In equivocal PSMA-RADS 3 lesions (n=44), BUMP had reduced discriminative power (AUROC 0.61, 95%: CI 0.40-0.79).
Conclusion: BUMP performed well in predicting metastases but underperformed compared to expert PSMA-RADS assessment. In equivocal lesions BUMP offered limited value, highlighting importance of expert interpretation.
Limitations: Retrospective design, limited subgroup power and reliance on clinical rather than histopathological ground truth.
Funding for this study: No specific funding was obtained for this study.
Has your study been approved by an ethics committee? Not applicable
Ethics committee - additional information:

Author Block: W. Jutidamrongphan¹, J. Hu¹, J. Wang¹, Y. Chen¹, Y. Rakvongthai², R. Seifert¹, K. Shi¹, A. Rominger¹, A. Afshar-Oromieh¹; ¹Bern/CH, ²Bangkok/TH
Purpose: Appropriate patient selection is crucial for effective [177Lu]Lu–prostate–specific membrane antigen (PSMA) radioligand therapy (RLT) in metastatic castration–resistant prostate cancer (mCRPC). While quantitative tumor burden assessment from PSMA PET/CT is prognostically valuable, its accuracy depends on segmentation methodology. This study compares fixed SUV 4 thresholding (SUV_4) with personalized SUVmean–based segmentation (SUV_MEAN) for predicting 12–month disease progression.
Methods or Background: This single–center study retrospectively analyzed 51 mCRPC patients treated with [177Lu]Lu–PSMA–617. Pre–treatment PSMA PET/CT was performed with [68Ga]Ga–PSMA–11 (n=6) or [18F]F–PSMA–1007 (n=45). Whole–body molecular tumor volume (MTV), total lesion uptake (TLU), and total lesion quotient (TLQ) were derived using both segmentation methods. Progression–free survival (PFS) was assessed with Cox regression; 12–month progression was evaluated using LASSO–penalized multivariable logistic regression.
Results or Findings: Mean age was 70.6 years (range 54–89); median PSA 264 ng/mL (range 1.6–6946); median treatment cycles 4 (range 1–11). At 12 months, 27 patients (53%) progressed. Both high MTV_4 and MTV_MEAN were associated with shorter PFS (HR 2.50 [95% CI 1.30–4.80], p=0.006), as were high TLU_4 and TLU_MEAN (HR 2.02 [1.06–3.85], p=0.032). For 12–month progression prediction, the SUV_MEAN–based model (MTV_MEAN, treatment cycles, baseline PSA) achieved superior performance with AUC 0.923 (95% CI 0.849–0.996) and cross–validated AUC 0.842 (0.691–0.993), with MTV_MEAN as the only significant PET–derived predictor. No SUV_4–derived metrics achieved significance in the final model.
Conclusion: Personalized SUV_MEAN segmentation provides superior prognostic performance compared to fixed SUV_4 thresholding for predicting early progression after [177Lu]Lu–PSMA–617 RLT. By incorporating patient–specific tumor uptake patterns, SUV_MEAN generates more clinically representative and reproducible tumor burden metrics. Validation in larger multicenter cohorts could improve patient selection and provide a standardized foundation for AI–driven prognostic tools.
Limitations: Not applicable
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: The ethics committee approval can be found under the number 2023–01877

Author Block: Y. Dauytova, Z. J. Zholdybay, Z. Zhakenova, Z. M. Amankulov, A. Dyo, L. Aitzhanova, M. S. Telkhazhaeyeva; Almaty/KZ
Purpose: To assess the added prognostic value of MTV and Dmax beyond the Deauville score in predicting progression-free survival (PFS) in lymphoma patients
Methods or Background: 18F-FDG PET/CT is indispensable in the management of lymphoma. While the Deauville 5-point scale remains the cornerstone for response assessment, it often fails to capture the full biological heterogeneity of disease. Novel quantitative biomarkers such as metabolic tumor volume (MTV) and the maximum distance between the two most metabolically active lesions (Dmax) may provide complementary insights into tumor burden and dissemination.
We retrospectively evaluated 120 patients with newly diagnosed Hodgkin and non-Hodgkin lymphoma who underwent baseline and interim ^18F-FDG PET/CT. MTV was derived from semi-automated segmentation, and Dmax was calculated as the Euclidean distance between the most distant FDG-avid lesions. Patients were stratified into high vs. low MTV and Dmax groups. Kaplan–Meier and multivariate Cox analyses were performed with PFS as the primary endpoint.
Results or Findings: Patients with high baseline Dmax (>30 cm) and high MTV had significantly inferior outcomes. The 2-year PFS was 82% in the low-Dmax group versus 48% in the high-Dmax group (p < 0.01), and 80% versus 50% for low vs. high MTV (p < 0.01). On multivariate analysis, both Dmax and MTV remained independent predictors of PFS beyond interim PET response. The combined model (Deauville + MTV + Dmax) achieved superior risk stratification, identifying a subgroup with poor prognosis despite negative interim PET findings.
Conclusion: MTV and Dmax extend the prognostic power of FDG PET/CT beyond the Deauville scale. Their integration into routine reporting can refine risk stratification and support personalized treatment strategies in lymphoma.
Limitations: N/a
Funding for this study: No funding was received for this work
Has your study been approved by an ethics committee? Not applicable
Ethics committee - additional information:

Author Block: F. Herr, C. A. Dascalescu, R. Ebner, M. L. Schnitzer, J. Ricke, M. Heimer, M. Brendel, R. Werner, C. C. Cyran; Munich/DE
Purpose: To evaluate the prognostic value of integrated clinical and laboratory biomarkers for longitudinal overall survival (OS) prediction in patients with gastroenteropancreatic neuroendocrine tumors (GEP-NETs) undergoing [177Lu]Lu-DOTA-TATE peptide receptor radionuclide therapy (PRRT).
Methods or Background: In this retrospective single-center study, 178 patients with histologically confirmed, well-differentiated (Grade 1–2) GEP-NETs treated with PRRT between 2012 and 2023 were included. Clinical and laboratory parameters (BMI, CRP, albumin, leukocytes, erythrocytes, hemoglobin, platelets, NSE, CgA) were collected at baseline, follow-up 1 (FU1; after 2 cycles), and follow-up 2 (FU2; after 4 cycles). Optimal cut-offs were determined using maximally selected rank statistics. Univariate and multivariate Cox regression analyses were performed; model performance was assessed by Harrell’s concordance index (C-index) and Akaike Information Criterion (AIC). Kaplan–Meier analysis with log-rank testing was applied.
Results or Findings: Median OS across the cohort was 70.1 months after a median follow-up of 60.8 months. At FU1, albumin ≤ 4.1 g/dL and CgA > 299 ng/mL were independent predictors of shorter OS. At FU2, albumin ≤ 4.5 g/dL, NSE > 29.1 ng/mL, and CgA > 179 ng/mL retained independent prognostic value. Models combining these biomarkers showed superior prognostic accuracy (C-index up to 0.77) compared with single-parameter approaches. Kaplan–Meier analyses confirmed significant separation between high- and low-risk groups.
Conclusion: Longitudinal assessment of albumin, NSE, and CgA provides independent and complementary prognostic information in PRRT-treated GEP-NET patients. Integration of these biomarkers into multivariable models improves OS prediction beyond baseline or single-marker assessment, supporting their role in personalized follow-up and treatment planning.
Limitations: Retrospective, single-center design; external validation in multicenter cohorts is required.
Funding for this study: No
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: This retrospective single-center study was approved by the local ethics committee (approval number 19–027). Due to the retrospective nature, informed consent was waived.

Author Block: D. Kifjak, M. J. Hochmair, A. Korajac, S. Pochepnia, R-I. Milos, A. Hoda, K. Sinn, L. Beer, H. Prosch; Vienna/AT
Purpose: To evaluate the clinical impact of pre-operative 18F-FDG-PET/CT in non-small cell lung cancer (NSCLC) patients who received neoadjuvant immunochemotherapy.
Methods or Background: This retrospective analysis included 68 potentially operable NSCLC patients (39 female, 29 male) who were treated with neoadjuvant immunochemotherapy and underwent subsequent pre-operative 18F-FDG-PET/CT for surgical planning. Histological subtypes included 42 adenocarcinomas and 26 squamous cell carcinomas. Pre-operative 18F-FDG-PET/CT scans after neoadjuvant immunochemotherapy were assessed for new metastatic disease and their influence on treatment decisions. Patient treatment decisions, including surgical resection and reasons for non-operative management, were reviewed.
Results or Findings: Of the 68 patients, 62 underwent surgical resection. Six patients did not undergo surgery. Among these, one patient was found to have new osseous metastases in the pelvis on pre-operative 18F-FDG-PET/CT, that were only detected on the metabolic PET component, which excluded surgery and therefore altered the treatment decision. Five patients did not undergo surgery for reasons unrelated to 18F-FDG-PET/CT findings: one declined surgery, one was excluded by multidisciplinary tumor board decision due to a large primary tumor, one was referred for radiotherapy because of N3 disease, and in two patients surgical resection was technically not feasible. Thus, pre-operative 18F-FDG-PET/CT changed the treatment strategy in only one patient (1.5%).
Conclusion: While the current National Comprehensice Cancer Network (NCCN) guidelines recommend contrast-enhanced CT with optional 18F-FDG-PET/CT for restaging prior to surgery, our results suggest that pre-operative 18F-FDG-PET/CT may not be warranted, raising an ethical dilemma regarding addional costs, radiation exposure and patient burden given its minimal impact on treatment decision-making.
Limitations: This was a retrospective analysis of a relatively small cohort size.
Funding for this study: The financial support by the Austrian Federal Ministry for Digital and Economic Affairs, the National Foundation for Research, Technology and Development and the Christian Doppler Research Association is gratefully acknowledged.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: Approved by the institutional ethics committee (EC No: 1521/2015).

Author Block: G. Kaya¹, O. Kodaz², A. E. Yıldız², S. Akın², P. Ozgen Kıratlı²; ¹Yozgat/TR, ²Ankara/TR
Purpose: Recent ESMO-EANM guidelines recognize ¹⁸F-FDG-PET/CT and whole-body diffusion-weighted MRI (WB-MRI) as key imaging modalities in multiple myeloma (MM), while ¹⁸F-Fluorocholine-(FCH)-PET/CT has emerged as a potentially more sensitive alternative. However, no prior study has directly compared all three modalities in the same cohort. This study aimed to evaluate the diagnostic yield of FCH PET/CT, FDG PET/CT, and WB-MRI in newly diagnosed MM patients and correlate imaging findings with laboratory and histopathological markers.
Methods or Background: This retrospective approved-study included 7 adults with newly diagnosed MM (median age: 63). All underwent WB-MRI, FDG-PET/CT, and FCH-PET/CT within four weeks. PET scans were evaluated using IMPeTUs criteria; WB-MRI was scored semi-quantitatively. Lesion- and patient-based analyses were performed. SUVmax and ADC were recorded for three target lesions per patient.
Results or Findings: FCH-PET/CT and WB-MRI detected disease in all patients(100%), outperforming FDG-PET/CT (positive in 6/7). FCH identified significantly more lesions than FDG or WB-MRI (p = 0.005), especially in diffuse marrow involvement. FDG showed better performance in cases with low CD138 expression. Imaging patterns varied markedly, even in similar disease phenotypes. SUVmax and ADC values showed weak correlation (r < 0.4). FDG Deauville scores correlated strongly with urinary kappa light chains, while FCH scores correlated with focal WB-MRI lesion counts and lower ADC values. Interestingly, higher β₂-microglobulin levels were inversely associated with FCH scores. WB-MRI findings correlated with urinary lambda light chains and protein levels in patients with focal lesions.
These findings underscore FC- PET/CT’s translational potential as a sensitive and widely accessible tool in MM staging.
Conclusion: FCH-PET/CT provided the most extensive disease detection, particularly in diffuse marrow infiltration. The complementary nature of FCH, FDG, and WB-MRI imaging suggests value in combined multimodality approaches.
Limitations: Retrospective nature, small cohort.
Funding for this study: This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: Hacettepe Üniversitesi Sağlık Bilimleri Araştırma Etik Kurulu; Hacettepe University Institutional Review Board for Health Sciences Research; SBA25/604

Author Block: M. A. Drews, A. Milosevic, S. Zensen, M. Opitz, L. Umutlu, W. Fendler, K. Herrmann, U. Dirksen, B. M. Schaarschmidt; Essen/DE
Purpose: Ewing sarcoma (EwS) is the second most prevalent malignant bone tumour among paediatric and juvenile patients. Due to its highly malignant character, an accurate diagnosis is crucial to determine most suitable treatment and to identify patients with unfavourable courses as early as possible. In this context, hybrid 18F-FDG-PET/MRI combines high-contrast morphological information with functional diffusion imaging and metabolic data which could deliver valuable prognostic information for treatment response and survival prediction. Therefore, we evaluated correlations between ADC and SUV as established parameters for tumour cellularity and metabolic activity and Salzer-Kuntschik scale(SKS) for treatment response as well as overall survival(OS) in this rare entity.
Methods or Background: In a monocentric retrospective study, 39 patients with histopathologically proven EwS who underwent 18F-FDG PET/MRI before therapy start were evaluated(14female, median age 16yrs [IQR:11-23]). A polygonal region of interest was drawn encompassing the entire primary tumour mass on fused PET/MR images to determine SUVmax and SUVmean and copied onto the ADC maps to quantify ADCmean and ADCmin. Dependencies between these parameters and SKS were analysed using Spearman's correlation coefficients and Hazard ratio was used for OS.
Results or Findings: A strong inverse correlation between SUVmax and ADCmean(ρ=-0.709,p<0.001) and SUVmean and ADCmean(ρ=-0.689,p<0.001) was observed. Furthermore, moderate correlations between SKS and SUVmax(ρ=0.502,p=0.009), SUVmean(ρ=0.389,p=0.049) and ADCmean(ρ=-0.425,p=0.03) were seen as well as correlations between OS and SUVmax(HR:5.324,95%-CI 1.065-26.615,p=0.042) and SUVmean(HR: 5.558,95%-CI 1.114-27.733,p=0.037).
Conclusion: Our study demonstrates a strong correlation between metabolic activity(SUV) and tumour cellularity(ADC) in EwS. SUVmax and SUVmean can further be used as prognostic markers for neoadjuvant treatment response and OS, while ADCmean also correlates inversely with treatment response but does not deliver complementary information.
Limitations: Limitations lie in the retrospective design and small patient number, but which is high for this rare entity.
Funding for this study: Marcel Drews received funding by the Junior Clinical Scientist Program at the University of Essen Medical School to carry out this project.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: Approval by the local ethics committee was obtained and the need for informed consent was waived due to the retrospective nature of this study (application number: 24-11729-BO; approval date: 23.02.2024).