RPS 2102 - Current innovations in contrast enhanced mammography

Author Block: I. Allajbeu¹, R. Manavaki¹, T. Wegman², J. A. Saenz³, T. Van Nijnatten², F. J. Gilbert¹; ¹Cambridge/UK, ²Maastricht/NL, ³Barcelona/ES
Purpose: Quantitative analysis of enhancement on contrast-enhanced mammography (CEM) has shown promise in distinguishing benign from malignant breast lesions. We evaluated diagnostic models combining low-energy (LE) mammographic features with quantitative enhancement metrics.
Methods or Background: Data from 251 CEM examinations (2018-2021) with identifiable, histologically confirmed lesions were retrospectively analysed. Lesion characteristics, including lesion diameter (LD), type (mass, calcification, distortion, asymmetry), background parenchymal enhancement (BPE), breast density (BD), and conspicuity, were assessed on LE and recombined images by three radiologists using BI-RADS criteria. Enhancement metrics were computed from early (CEearly) and late (CElate) views, with percent residual signal difference (%RSD) used to classify enhancement patterns as progressive, plateau, or wash-out. Model construction utilised nested cross-validation (CV) with stratified sampling. In each of ten outer folds, data were stratified into 90% training and 10% testing subsets. Feature selection was optimised within the training data using five-fold inner CV. Minimum redundancy–maximum relevance identified non-redundant, informative predictors for logistic regression. Performance was assessed on the outer test sets and averaged across folds using area-under-the-curve (AUC), accuracy, sensitivity, and specificity.
Results or Findings: Of 251 lesions, 155 (61.7%) were malignant and 96 (38.3%) benign. Eleven features (age, LD, type, BI-RADS score, BD, BPE, conspicuity, CEearly, CElate, RSD, enhancement type) were considered for model construction. The most predictive subset included BI-RADS score, age, lesion conspicuity, BD, CELate, and LD. Models with these features outperformed those using BI-RADS score alone, with higher AUC (0.88 vs 0.84), accuracy (83% vs 78%), sensitivity (89% vs 72%), and comparable specificity (85%).
Conclusion: Integrating standard BI-RADS descriptors with quantitative enhancement metrics can improve lesion discrimination on CEM versus BI-RADS alone. This combined approach may enhance diagnostic confidence and reduce unnecessary biopsies in clinical practice
Limitations: Single-centre study
Funding for this study: Cambridge Biomedical Research Centre (BRC)
CUH NHS Foundation Trust
Has your study been approved by an ethics committee? Not applicable
Ethics committee - additional information:

Author Block: S. Mohammadzadeh, A. Abdi, I. Kiani, F. Zeinalkhani; Tehran/IR
Purpose: Contrast-enhanced mammography (CEM) is an emerging imaging technique for breast cancer detection. The Kaiser score (KS), a structured diagnostic tool developed initially for breast MRI with high consistency and accuracy. We aimed to determine whether the KS could enhance the diagnostic performance of the CEM in evaluating suspicious breast lesions and to assess the effect of combining the KS with calcification evaluation (KS+).
Methods or Background: This retrospective, single-center study involved 143 breast lesions from 130 patients. CEM exams were evaluated by two radiologists, assigning both Breast Imaging-Reporting and Data System (BI-RADS) and KS classifications. Diagnostic performance metrics, including sensitivity, specificity, and area under the ROC curve (AUC), were calculated for BI-RADS, KS, and KS+. Moreover, subgroup analyses based on mass/non-mass lesions and background parenchymal enhancement (BPE) were performed.
Results or Findings: Overall, the KS showed an AUC of 0.892, with 83.3% sensitivity and 89.7% specificity. The BI-RADS classification produced an AUC of 0.867, with a higher sensitivity of 94.7% but a lower specificity of 58.6%. The combined KS+ model reached an AUC of 0.894. Diagnostic performance was significantly better for mass lesions than for non-mass lesions across all systems. For non-mass lesions, the KS maintained a specificity of 83.3%, whereas the BI-RADS specificity decreased to 33.3%. All methods experienced reduced diagnostic accuracy in cases with moderate BPE.
Conclusion: The KS acts as a helpful complement to BI-RADS when interpreting CEM. By offering a structured and numerical assessment, the KS significantly improves diagnostic specificity, particularly for challenging non-mass lesions, without substantially compromising sensitivity. This enhancement can help lower false-positive results and reduce unnecessary biopsies.
Limitations: 1) Retrospective design and single-center setting may limit the generalizability of our findings. 2) Inter-reader agreement was not evaluated
Funding for this study: None
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: The institutional review board of the Imam Khomeini Hospital Complex approved our study.

Author Block: A. Stiglbauer-Tscholakoff¹, G. Chiffi², P. D. Stelzer¹, T. H. Helbich¹, P. Clauser¹, P. A. Baltzer¹; ¹Wien/AT, ²Rome/IT
Purpose: To compare detection rate, lesion visibility and conspicuity in contrast enhanced mammography using two different doses.
Methods or Background: A total of 256 lesions were included: in 122 lesions Iomeron®400 was administered with a dose of 1 ml/kg bodyweight at 2 ml/s flow (group 1) and in 134 lesions with 1.5 ml/kg at 2 ml/s (group 2). Two breast radiologists independently evaluated lesion visibility (overall, 1-5) and conspicuity (in comparison to background enhancement, 0-3). Lesions were classified as benign/malignant (biopsy-proven). Distribution of malignant/benign lesions (chi-square test), diagnostic performance, intra-reader comparisons, and inter-reader differences (Student’s t-test) were assessed.
Results or Findings: There was no significant difference in the distribution of malignant/benign lesions in group1/group2 (P = 0.4627). Almost all malignant lesions were detected by both readers (group 1: 70/71, group 2: 83/84). Mean lesion conspicuity in group 1 was 1.7 (SD 0.9) for Reader 1 and 1.5 (SD 1.1) for Reader 2, in group 2 2.2 (SD 0.8, R1) and 2.1 (SD 1.0, R2). Mean lesion visibility in group 1 was 3.4 (SD 1.5, R1) and 2.9 (SD 1.3, R2), in group 2 3.5 (SD 1.4, R1) and 3.7 (SD 1.3, R2). Intra-reader analyses showed lower conspicuty for R1 and R2 in group 1 (R1&R2, t-test P < 0.0001), equal visibility (R1, t-test P = 0.5816) and lower visibility (R2, t-test P < 0.0001) in this group.
Conclusion: Lesion visibility in CEM was higher using 1,5 ml. Lesion conspicuity was reader dependent. The detection rate of malignant lesion was high with both doses, suggesting that a higher dose of contrast agent may improve lesion conspicuity in comparison to background enhancement but a lower dose might be sufficient for lesion detection.
Limitations: Single-centre design; selection bias; no clinical outcomes correlation.
Funding for this study: Bracco Imaging
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: Data were collected within a prospective study comparing the diagnostic value of CEM to CE-MRI in a problem-solving setting (Ethics Review Board number 2282/2019).

Author Block: L. Corradini, C. Depretto, C. De Berardinis, G. Della Pepa, E. D'Ascoli, G. Irmici, A. Bonanomi, F. Sardanelli, G. P. Scaperrotta; Milan/IT
Purpose: To evaluate the diagnostic performance of contrast-enhanced mammography (CEM) in preoperative locoregional staging of breast cancer, with a specific focus on the contribution of 7-minute acquisitions, through a retrospective analysis of consecutive cases at a reference oncology center.
Methods or Background: We included patients who underwent CEM between January 2018 and January 2024 at the National Cancer Institute (Milan). Two independent readers, each with over five years of experience in CEM, performed blinded assessments without knowledge of clinical history or biopsy results. A BI-RADS score (0–5) was assigned to each breast, first based soley on 2-minute acquisitions and subsequently incorporating information from 7-minute images. The reference standard included histopathological findings (diagnostic or definitive), follow-up, or comparison with prior imaging. Diagnostic performance was compared using McNemar’s test.
Results or Findings: A total of 286 breasts in 144 patients (including two cases of prior unilateral mastectomy) were evaluated. The additional review of 7-minute acquisitions led to a significant modification of the BI-RADS score and/or other assessments in 12/286 breasts (4.2%), with clinical impact in 2/144 patients (1.39%, 95% CI 0.17−4.93%). The diagnostic performance of 2-minute acquisitions alone was not significantly different from that obtained when including 7-minute acquisitions: sensitivity 94.0% vs. 94.6% (p = 1.00); specificity 97.7% vs. 96.2% (p = 0.48).
Conclusion: Delayed 7-minute CEM acquisitions do not provide clinically relevant additional information for preoperative breast cancer assessment. A protocol limited to 2-minute acquisitions reduces exposure to ionizing radiation and simplifies workflow without compromising the diagnostic value of the examination.
Limitations: Retrospective, single-center study with a limited number of patients.
Funding for this study: This study received no funding.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: The study was approved by Comitato Etico Territoriale Lombardia 4

Author Block: G. Gennaro¹, E. Baldan¹, P. Belli², D. Bernardi³, E. Bezzon¹, A. Coran¹, V. Iotti⁴, I. Polico¹, F. Caumo¹; ¹Padua/IT, ²Rome/IT, ³Rozzano, Metropolitan City of Milan/IT, ⁴Reggio Emilia/IT
Purpose: To compare the diagnostic performance of contrast-enhanced mammography (CEM) versus breast MRI in women at increased risk for breast cancer.
Methods or Background: In this paired-case study, 462 women underwent both CEM and MRI. Four breast radiologists interpreted CEM, while a separate group of four radiologists evaluated MRI. Multi-reader, multi-case (MRMC) ROC analysis was used to calculate AUC for each modality. Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and overall accuracy were also estimated. Low-energy CEM (LE-CEM) readings were included as a baseline reference. Statistical significance was set at p<0.05 for all comparisons.
Results or Findings: LE-CEM to CEM comparisons confirmed prior evidence, with mean AUC improving from 0.832 (95% CI: 0.809–0.855) to 0.913 (95% CI: 0.897–0.930, p<0.001). For CEM versus MRI, mean AUCs were 0.936 (95% CI: 0.915–0.958) for CEM and 0.933 (95% CI: 0.916–0.949) for MRI (p=0.665). Sensitivity, specificity, PPV, NPV, and accuracy were also similar between CEM and MRI (sensitivity 0.83 vs 0.86, specificity 0.91 vs 0.87, PPV 0.77 vs 0.70, NPV 0.94 vs 0.95, accuracy 0.89 vs 0.86; all p>0.30).
Conclusion: CEM demonstrates diagnostic performance equivalent to breast MRI and significantly outperforms mammography (LE-CEM). These results support CEM as a high-quality, more accessible alternative to MRI for risk-adapted screening.
Limitations: Retrospective analysis with readers not fully crossed between modalities.
Funding for this study: Veneto Region, grant number RSF-2017-00000562
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: Institutional Ethics Committee Veneto Institute of Oncology (IOV)— IRCCS, Padua, Italy. Protocol code CE IOV #2017/92

Author Block: V. Lamagna, C. Bellini, G. Bicchierai, D. De Benedetto, F. Di Naro, S. Vidali, F. Pugliese, V. Miele, J. Nori; Florence/IT
Purpose: To compare the performance of Contrast-Enhanced Mammography (CEM) vs. ultrasound (US) in women with personal history of breast cancer (PHBC) who are recalled for suspicious mammographic findings.
Methods or Background: This retrospective single-center study included women with PHBC recalled for suspicious findings on digital mammography between November 2016 and June 2023. All underwent both CEM and targeted US. Two experienced breast radiologists independently reviewed the images in a blinded manner. For each modality, findings were categorized as positive (BIRADS 4 o 5) or negative (BIRADS 1,2 or 3) based on BI-RADS classification. Histopathology served as the reference standard for positive cases, while a negative 24-month follow-up was used for negative cases. Sensitivity, specificity, positive and negative predictive values, and overall accuracy were calculated for each modality.
Results or Findings: Among 295 women included, 54 (18.3%) had malignant lesions. CEM demonstrated significantly higher sensitivity than ultrasound for lesion detection (98.2% vs 35.2%, p<0.001) and higher specificity (91.7% vs 72.2%, p<0.01). CEM accurately identified 53 of 54 malignant lesions confirmed on histopathology with a PPV of 72.6% and a NPV of 99.5%, including several that were occult or equivocal on ultrasound that showed a PPV of 22.1% and a NPV of 83.3%. The only missed case was a non-enhancing ductal carcinoma in situ. Overall diagnostic accuracy was higher for CEM than for ultrasound (92.9% vs 65.4%, p<0.001).
Conclusion: CEM compared to US is a reliable and efficient first-line imaging technique in in women with a personal history of breast cancer, potentially improving workflow efficiency in breast cancer follow-up imaging.
Limitations: This study was limited by its retrospective single-center design.
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: E. Meltem¹, R. Wijgers², U. Lalji², C. R. Jeukens², J. E. Wildberger², E. C. Nijssen², T. Van Nijnatten²; ¹Istanbul/TR, ²Maastricht/NL
Purpose: To evaluate the true extent of iodine-based contrast media (ICM)-related risk of kidney function decline in a breast cancer screening population, recalled and scheduled for contrast-enhanced mammography (CEM).
Methods or Background: This retrospective study included women recalled from the Dutch National Breast Cancer Screening Program to Maastricht University Medical Centre+ (MUMC+) between January 2013 and January 2022. Recalled patients routinely underwent CEM, while those with contraindications received full-field digital mammography (with or without digital breast tomosynthesis). The most recent estimated glomerular filtration rate (eGFR) before and after imaging was obtained from electronic medical records. Renal safety was assessed by changes in eGFR; a ≥25% decline within 30 days after CEM was considered clinically relevant.
Results or Findings: A total of 2707 women were recalled, of whom 2437 (90%) underwent CEM. Paired pre- and post-CEM eGFR values were available in 1184 patients, with pre-CEM values obtained a median of 177 days before (IQR 420-50) and post-CEM values 141 days after imaging (IQR 28-423). The median eGFR difference was 0.5 (IQR -5.4 to 6.3). A ≥25% decline in eGFR occurred in 31 patients (2.6%); only one (0.08%) dropped below an eGFR of 30. Most declines reflected shifts from baseline eGFR >60 to the 45–60 range (n=74, 8%), while eGFR remained stable in those with baseline values <45 (n=13). In the subgroup with post-CEM eGFR within 30 days (n=306), ≥25% decline was observed in six patients(2.0%); none dropped below eGFR 45, and changes were again mainly shifts from >60 to 45–60 (n=15, 6%).
Conclusion: Clinically relevant renal function decline after CEM was rare and never severe, supporting the notion that kidney safety concerns should not restrain the implementation of CEM in clinical practice.
Limitations: Variability in the timing of eGFR measurements.
Funding for this study: Institutional grant support from Bayer
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: Institutional ethical approval/2023

Author Block: F. Pugliese¹, F. Di Naro¹, S. E. Baldi Giorgi¹, C. Maiello², V. Lamagna¹, D. De Benedetto¹, G. Bicchierai¹, C. Bellini¹, J. Nori¹; ¹Florence/IT, ²Casoria (Na)/IT
Purpose: To evaluate the role of lesion conspicuity (LC) at 12 months on contrast-enhanced mammography (CEM) follow-up as a predictor of cryoablation outcome in breast cancer.
Methods or Background: We conducted a retrospective analysis of 69 patients (mean age: 86 years; range: 60–94) with unifocal invasive breast cancer who underwent percutaneous cryoablation. CEM imaging was performed at baseline and within 12 months post-ablation. LC was categorized based on the degree of enhancement relative to background as: low, moderate, or high. Tumor response was assessed at 12 months using standard imaging modalities (ultrasound and CEM) according to RECIST 1.1 criteria. Statistical analyses were conducted to evaluate the correlation between changes in post-treatment conspicuity and later tumor response.
Results or Findings: At 12 months post-ablation, CEM showed a complete absence of enhancement in 68.1% of lesions. Among the 22/69 (32%) lesions showing enhancement, 6 had low LC, 11 moderate, and 5 high. The agreement between CEM and biopsy results was strong, with 100% sensitivity and positive predictive value for CEM. A substantial concordance (Kappa=0.80) was observed between CEM and RECIST criteria. Among lesions that were minimally conspicuous or non-enhancing early post-treatment, 80.7% had a complete response at 12 months, compared to only 1.8% of lesions that retained enhancement. In tumors ≤2 cm, early post-ablation CEM showing reduced or absent conspicuity was significantly associated with complete response (Kappa=0.74), with a 78.8% complete response rate in this subgroup.
Conclusion: A reduction in lesion conspicuity on early post-ablation CEM strongly correlates with a favorable tumor response in breast cancer patients treated with cryoablation. This imaging feature may serve as an early, noninvasive biomarker for treatment evaluation and could help guide clinical decision-making.
Limitations: retrospective study
small sample size
did't compare CEM-MRI
Funding for this study: None
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: Meyer Hospital Florence

Author Block: L. M. Neeter¹, M. B. I. Lobbes², K. M. Duvivier³, F. Raat⁴, M. Smidt¹, J. E. Wildberger¹, B. Essers¹; ¹Maastricht/NL, ²Sittard-Geleen/NL, ³Amsterdam/NL, ⁴Roermond/NL
Purpose: The randomised controlled clinical RACER trial studied the diagnostic work-up with contrast-enhanced mammography (CEM) compared to conventional imaging as primary tool in women recalled from breast cancer screening. This current trial-based economic evaluation was performed from a hospital perspective.
Methods or Background: Cost prices were retrieved from the financial departments of the Maastricht University Medical Center. Health-related quality of life was measured five times over 18 months using the EQ-5D-5L questionnaire. Cost-utility analysis outcome was expressed as costs per quality-adjusted life year (QALY). Multiple imputation was used for missing data, and non-parametric bootstrap analysis was performed to examine uncertainty in the difference in costs and incremental costs per QALY. Post hoc subgroup analysis was performed per BI-RADS recall score.
Results or Findings: Work-up with CEM showed lower total costs of €-117 (95% CI €-254; €+22) compared to conventional imaging. Average imaging costs were significantly lower with CEM (mean difference €-130; 95% CI €-153; €-105). Mean QALY for the CEM group was 1.2034 versus 1.2137 for the control group. Bootstrap analysis showed that 95% of all simulated ICERs were in the quadrants that indicate cost-savings, although 25% of the ICERs showed a small gain in QALY and 70% a small QALY loss. The probability of CEM being cost-effective is 53% at a threshold of €10,000 to accept a QALY loss. In the subgroup of BI-RADS 0 recalls, specifically, this probability is 85%.
Conclusion: CEM as primary tool in the diagnostic work-up is a cost-effective diagnostic strategy, especially in BI-RADS 0 (low suspicion for breast cancer) recalls.
Limitations: The proportion of missing data related to the EQ-5D-5L questionnaire increased over time up to 42% and 34% in the control and CEM group, respectively.
Funding for this study: ZonMw Efficiency Studies (Grant: 843001801); GE Healthcare
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: Institutional Review Board approval: METC171082/NL62788.068.17

Author Block: J. Theunissen¹, Y. Vissers¹, B. Van Grinsven², A. Valentijn-Morsing¹, M. B. I. Lobbes¹, L. Bouwman²; ¹Sittard-Geleen/NL, ²Maastricht/NL
Purpose: Contrast-enhanced mammography (CEM) has witnessed an increasing implementation in clinical practice and is currently even considered for large-scale screening programs. However, successful adoption of screening CEM is dependent on patient acceptance. This study investigated women’s experiences with CEM, focusing on pain and willingness for future CEM examinations.
Methods or Background: In this prospective survey at a specialized breast center, women undergoing CEM completed a 10-minute questionnaire. A questionnaire collected data on demographics, medical history, expectations, and patient-reported outcomes. Procedural data such as prior imaging, biopsy during the same visit, and compression force were collected from clinical records.
Results or Findings: Currently, 478 women (median age 58 years) completed the questionnaire, with 382 (79.9%) also participating in the national screening program. Median Visual Analogue Scale pain scores were 1 (IQR 0 – 2) for intravenous cannulation and 3 (IQR 1-6) for breast compression. Overall, 475 (99.4%) participants reported willingness to undergo CEM again if indicated, and 458 (95.8%) would undergo CEM screening if it improved cancer detection rates. Twenty-three participants (4.8%) reported pain during compression as a potential obstacle, and 15 participants (3.1%) reported the same for intravenous cannulation.
Amongst 25 women who had undergone a prior MRI, 16 (64%) considered CEM more patient-friendly, 7 (28%) rated both modalities equally, and 2 (8%) preferred MRI. Multivariable analysis of factors associated with pain experience and willingness for CEM screening is planned once all data are collected.
Conclusion: CEM is well tolerated, with limited discomfort reported during both intravenous cannulation and compression. Nearly all participants expressed willingness to undergo CEM again, both clinically and for screening.
Limitations: The single-center questionnaire may limit generalizability. Outcomes relied on self-reported data which introduces possible response bias.
Funding for this study: Zuyderland MC
Has your study been approved by an ethics committee? Not applicable
Ethics committee - additional information:

Author Block: G. Vatteroni¹, R. Levi¹, M. Filippini², M. Costa³, R. M. Trimboli¹, D. Bernardi¹; ¹Milan/IT, ²Nuvolera/IT, ³Catania/IT
Purpose: To evaluate patient acceptance and experience of contrast-enhanced mammography (CEM) compared to standard digital mammography (DM) in women operated for breast cancer (BC) undergoing follow-up.
Methods or Background: This study included women operated for BC who underwent CEM instead of DM for BC surveillance as part of a prospective study (COMBO-TRIAL). A 16-item anonymous questionnaire was administered after the exam, evaluating perceived discomfort, pain, exam duration, reassurance, information received, anxiety, contrast injection, satisfaction and comparative experience with prior DM. Responses used Likert-type scales (1-5). Descriptive and comparative analyses (T-Test or Mann-Whitney U test, Pearson’s R, Spearman’s R correlation) were conducted (p<0.05 significant).
Results or Findings: Among 969 patients, 936 (median-age 58.2 years, IQR 51–66.25) completed the questionnaire. CEM was reported to cause similar discomfort compared to DM by 67.4% of women, while 14.1% experienced greater discomfort, strongly associated with higher pre-exam anxiety(p<0.001) and lower reassurance(p<0.001). Likewise, 77.5% reported similar pain, whereas 7.4% perceived CEM as more painful, also linked to increased anxiety(p<0.001) and reduced reassurance (p<0.001). Examination positioning was considered comparable by 76.6% of patients, and with a similar duration (64.2%).
Overall, 92.6% felt reassured and 95% well informed. Elevated anxiety was reported by 21.2%, with only 32.6% reporting high comfort during contrast injection. Pre-exam anxiety was inversely correlated with injection comfort (r =–0.265, p<0.001). Importantly, 90.6% accepted the risk of contrast-related events if CEM enabled earlier BC detection. Patient satisfaction was high (96.1%), and nearly all (99.6%) would recommend CEM, highlighting its perceived diagnostic value and role in early detection.
Conclusion: CEM showed excellent acceptance, tolerability and satisfaction compared with DM in BC follow-up. Distress was limited to a small subgroup with elevated pre-exam anxiety, highlighting the radiologist’s key role in communication and patient reassurance.
Limitations: n/a
Funding for this study: n/a
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: n/a

Author Block: M. B. I. Lobbes¹, J. Theunissen¹, R. Körver-Steeman¹, C. Frotscher¹, A. Valentijn-Morsing¹, B. De Vries¹, W. L. Mok², Y. Vissers¹, L. Bouwman¹; ¹Sittard-Geleen/NL, ²Roermond/NL
Purpose: Contrast-enhanced mammography (CEM) can reveal lesions not visible on low-energy images, targeted ultrasound, or tomosynthesis. When such lesions are detected, contrast-enhanced stereotactic biopsy (CESB) is required for tissue diagnosis. However, the optimal number of tissue samples needed for a reliable diagnosis remains unclear.
Methods or Background: We included patients with lesions visible only on recombined CEM-images, warranting CESB. Using a 9-gauge vacuum-assisted biopsy (VAB) needle, 18 consecutive tissue samples were obtained per protocol. A pathologist evaluated each sample in the order collected to determine the cumulative diagnostic yield per sample. Additional parameters assessed included patient age, procedure duration (excluding IV catheter placement), adverse events, patient-reported discomfort (10-point scale), and the positive predictive value (PPV) for malignancy.
Results or Findings: Sixty women (mean age 57.8 years, SD 10.6) underwent CESB. Mean procedure time was 13.5 minutes. One procedure was discontinued due to syncope. Median discomfort score was 3. The PPV for malignancy was 31.7%. The median number of samples needed to establish a diagnosis was five (range: 1–18). Diagnostic yield increased with each additional sample, reaching 98.3% by the 12th biopsy. In one case, all 18 samples were required, as no lesion was identified; follow-up is ongoing.
Conclusion: CESB is a well-tolerated and safe technique with minimal discomfort and low complication rates. Our findings suggest that obtaining 12 samples with a 9-gauge VAB needle provides a high diagnostic yield and is sufficient in most cases to reach a reliable diagnosis.
Limitations: Relatively small sample size, but study is currently recruiting.
Funding for this study: Hologic Inc.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: Ethical committee approval acquired.

Author Block: R. Alcantara Souza, J. Saenz, M. Pitarch, P. Miranda Martínez, E. Vall Foraster, E. Vila-Trias Jover, N. Arenas, X. Castells; Barcelona/ES
Purpose: To evaluate the false-positive rate and histopathological spectrum of recombined (RC)-only lesions requiring contrast-enhanced mammography (CEM)-guided biopsy.
Methods or Background: Retrospective analysis of consecutive CEM-guided biopsies performed at a tertiary centre between 2019–2025. Demographics, CEM descriptors (mass, non-mass, enhancing asymmetry), and histopathology were extracted from a REDCap database. Pathology was categorised according to the NHSBSP system: B1 normal, B2 benign, B3 lesions of uncertain malignant potential, and B5 malignant. False positives were defined as B1–B3. RC-only lesions were enhancing findings without a low-energy correlate and represented the main indication for CEM-guided biopsy at our centre.
Results or Findings: A total of 174 RC-only lesions in 170 patients (median age 59 years, range 22–88) underwent CEM-guided biopsy with pathology. B1 was found in 3 (1.7%), B2 in 73 (42.0%), B3 in 27 (15.5%), and B5 in 71 (40.8%). The false-positive rate (B1–B3) was 59.2%. Malignancy rates were 45.4% in masses, 36.5% in non-mass enhancements, and 33.3% in enhancing asymmetries (OR mass vs non-mass 1.45; 95% CI 0.78–2.69). Benign outcomes included fibroadenomas, UDH, adenosis, and fat necrosis.
Among B3 lesions, 18 underwent intervention (14 lumpectomies, 3 mastectomies, 1 vacuum-assisted excision) and 10 upgraded (55.6%). By subtype: ADH 5/6 (83%), papillary 3/5 (60%) of 9 papillary overall, LIN 1/3 (33%), FEA 0/2, phyllodes 0/1, radial scar 0/0, and other atypia 1/1 (100%).
Conclusion: False-positive outcomes are frequent among RC-only CEM findings requiring biopsy, spanning benign and B3 lesions. Although many yield non-malignant results, the high upgrade risk of some subtypes supports intervention. Findings are consistent with published CEM series and comparable to MRI-guided biopsy. Despite this false-positive burden, CEM-guided biopsy remains essential for characterisation of RC-only lesions not accessible with conventional guidance.
Limitations: Retrospective design, single-centre setting, and small B3 subgroups.
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: Protocol number 2019/8890

Author Block: A. F. Murano¹, D. De Benedetto², M. Sacco¹, L. R. La Rocca¹, V. Lamagna², F. Di Naro², M. G. Vigilante¹, J. Nori², V. Romeo¹; ¹Naples/IT, ²Firenze/IT
Purpose: Contrast-enhanced mammography (CEM)-guided biopsy allows the histological characterization of “enhancing only” CEM or MRI findings. Its recent introduction made it a valuable alternative the less available and more expensive MRI guided biopsy. However, “contrast-enhanced only” findings may be non-specific, raising the risk of unnecessary biopsies. The study aimed to identify clinical and/or imaging parameters predictive of malignancy in patients undergoing CEM-guided biopsy in a multicentric setting.
Methods or Background: We retrospectively included women with suspicious “contrast-enhanced only” lesions visible only on CEM and sampled with CEM-guided biopsy in two different institutions. Clinical data (age, concurrent/previous breast cancer (BC), family history of BC, high-risk factors) and CEM features (size, enhancement type, distribution, pattern, conspicuity, background parenchymal enhancement) were collected. Univariate and logistic regression analyses were performed to identify clinical and/or CEM findings that could independently predict malignancy after CEM-guided biopsy.
Results or Findings: A total of 126 women with 126 suspicious CEM “enhancing only” findings, of which 98 benign and 28 malignant, were included. Among the clinical and imaging parameters, patient age, the presence of concurrent breast cancer, and early contrast enhancement were statistically significant at both univariate (p value: < 0.001, 0.004, and 0.015, respectively) and logistic regression analyses, with a combined AUC of 0.798 (95% CI: 0.718–0.864) and accuracy of 81.75% (95% CI: 73.9% - 81%) in predicting malignancy.
Conclusion: Patient age, the presence of a concurrent breast cancer, and early enhancement may help selecting patient who could benefit from CEM-guided biopsy. Larger studies are needed to confirm these preliminary results, improve the performance of the model, and develop clinical decision tools.
Limitations: Small sample size
Funding for this study: None
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: The study was approved by the ethics committee Campania 3