Background Extremely dense breast tissue is associated with increased breast cancer risk and reduced sensitivity of digital mammography (DM). This study evaluated the cost-effectiveness of abbreviated breast MRI (AB-MRI) and contrast-enhanced mammography (CEM) versus DM for screening women with extremely dense breasts. Methods This micro-simulation study applied the validated SimRisc Breast model. Screening strategies included biennial supplemental full-protocol MRI (FP-MRI), supplemental AB-MRI, supplemental CEM, AB-MRI alone and CEM alone, with biennial DM (ages 50-74) as reference. Costs and life-years gained (LYG) were discounted at 3%. Average cost-effectiveness ratios (ACERs) were calculated versus the reference, and incremental cost-effectiveness ratios (ICERs) for pairwise comparisons. A willingness-to-pay threshold of €20,000 per LYG was applied. Results Compared with the reference, all alternative screening strategies increased cancer detection and LYG, while reducing interval cancers and breast cancer deaths. CEM alone and supplemental CEM slightly increased in radiation-induced tumors. In cost-effectiveness analysis, supplemental AB-MRI and supplemental FP-MRI were not favorable (ACERs of €24,700–€34,400 per LYG), whereas CEM alone, AB-MRI alone, and supplemental CEM were cost-effective (ACERs €13,400–€19,600 per LYG). Incremental comparisons indicated that all supplemental strategies were dominated, with CEM alone optimal (ICER €13,400 per LYG vs. reference) and AB-MRI alone at €49,400 per LYG relative to CEM alone. Conclusion Both AB-MRI and CEM could reduce breast cancer mortality through earlier detection, with low additional risks and reasonable incremental costs. Among the modeled strategies, CEM was estimated to be the most cost-effective, although further prospective clinical studies are needed to confirm these findings.

Background : Extremely dense breast tissue is associated with an increased risk of breast cancer and reduced sensitivity of digital mammography (DM). This study evaluated the cost-effectiveness of abbreviated breast MRI (AB-MRI) and contrast-enhanced mammography (CEM) compared with DM for screening women with extremely dense breasts.Methods : A validated micro-simulation model (Simrisc Breast) was applied. Screening strategies included biennial supplemental full-protocol MRI (FP-MRI), supplemental AB-MRI, supplemental CEM, AB-MRI alone and CEM alone, with biennial DM (age 50-74) as reference. Costs and life-years gained (LYG) were discounted at 3%. Average cost-effectiveness ratios (ACERs) were calculated relative to the reference, and incremental cost-effectiveness ratios (ICERs) were estimated for pairwise comparisons. A willingness-to-pay threshold of €20,000 per LYG was applied.Results: Compared with the reference, all alternative screening strategies increased cancer detection and LYG, while reducing interval cancers and breast cancer deaths. CEM alone and supplemental CEM showed a slight increase in radiation-induced tumors. In cost-effectiveness analysis, supplemental AB-MRI and supplemental FP-MRI were not favorable (ACERs of €24,700–€34,400 per LYG), whereas CEM alone, AB-MRI alone, and supplemental CEM were cost-effective (ACERs €13,400–€19,600 per LYG). Incremental comparisons indicated that all supplemental strategies were dominated, with CEM alone optimal (ICER €13,400 per LYG vs. reference) and AB-MRI alone at €49,400 per LYG relative to CEM alone.Conclusion: Both AB-MRI and CEM are cost-effective alternatives to DM and could reduce breast cancer deaths through earlier detection, with low additional risks and reasonable incremental costs. CEM is the preferred strategy due to superior cost-effectiveness.

Overdiagnosis estimates of ductal carcinoma in situ (DCIS) vary from 20-91%, which complicates screening communication and optimization. The aim was to quantify the influence of follow-up time and screening setting on overdiagnosis of DCIS. The fully validated micro-simulation Markov model for DCIS (SimDCIS) was used to estimate DCIS overdiagnosis in different screening settings with varying follow-up. DCIS overdiagnosis was defined as the number of diagnosed DCIS (screen-, clinically detected, or progressed to invasive breast cancer) in screening that would not have been diagnosed without screening. Outcomes were presented as overdiagnosed proportion and rate. The base cohort was screened biennially from age 50-74 with 76% compliance and 25 years (y) follow-up and compared to the non-screened cohort. Follow-up was varied from 2-25y, screening start 40-74y, screening interval 1-5y and compliance 50-100%. DCIS overdiagnosis was estimated at 20% of all diagnosed DCIS and 38.1 overdiagnosed DCIS per 100,000 women screened biennially from age 50-74 at 76% compliance and 25y follow-up. The proportion of overdiagnosed DCIS increased with shorter follow-up (27% at 2y to 20% at 25y), older screening start age (1% at 40y to 15% at 74y), decreased screening interval (23% at 1y to 12% at 5y), and increased compliance (16% at half to 20% at full participation). In conclusion, reliable DCIS overdiagnosis estimates require attention to screening setting and ≥20 years follow-up. Older women (74y) showed up to seven times more overdiagnosis at initial screening than younger women (50y). Improved estimates can provide guidance in screening communication and optimization.

BACKGROUND:Breast cancer (BC) screening can detect BC early to reduce mortality. This study evaluated the impact of BC screening on the incidence of late-stage BC in a population-based setting in the Netherlands. METHODS:All Dutch women aged 50-74 years diagnosed with invasive BC or ductal carcinoma in situ (DCIS) between 2007 and 2016 (n = 108,253) were included from the Netherlands Cancer Registry. BC was classified as screen-related if diagnosed within 24 months the last screening attendance, and as screen-detected if diagnosed within 12 months after positive screening result. Late-stage BC was defined in two ways: advanced BC, including TNM stages T3, T4, or N2, N3, or M1, and metastatic BC, defined as M1 disease. Multivariable logistic regression adjusted for age and socioeconomical status was used to assess associations between screen-related and screen-detected BC and late-stage BC. Analyses were done overall and stratified by HR/HER2-defined subtypes. FINDINGS:BC incidence increased between 2007 and 2013 and decreased slightly thereafter. Advanced BC incidence decreased between 2007 and 2016, while metastatic BC rates remained stable. Non-screen-related BCs were significantly more likely to be present as late disease compared with screen-related (advanced BC: aOR = 3.24, 95%CI = 3.12-3.37; metastatic BC: aOR = 6.40, 95%CI = 5.98-6.85). Similarly, non-screen-detected BCs had substantial higher odds of being late than screen-detected BCs (advanced BC: aOR = 5.54, 95%CI = 5.31-5.78; metastatic BC: aOR = 12.66, 95% CI = 11.41-14.05) than screen-detected BCs. These associations were observed across all HR/HER2-defined subtypes. INTERPRETATION:Population-based screening is strongly associated with earlier-stage breast cancer at diagnosis, consistently across all immunohistochemistry subtypes.

Background and Objectives: Emphysema and coronary artery calcium (CAC) share common lifestyle-related risk factors, yet their association in Chinese populations remains understudied. This study investigated how lifestyle factors influence the association between emphysema and CAC score in an urban Chinese general population. Methods: The study included 1000 participants from the Chinese Nelcin-B3 urban general population study originating in 2017 who underwent low-dose CT (LDCT) screening and comprehensive CT assessment. Emphysema was visually assessed by subtype and severity. CAC was measured using the Agatston method and categorized as 0, 1-100, and >100. Questionnaire-based lifestyle factors (smoking, BMI, diet, physical activity, alcohol consumption and environmental exposures) were categorized based on number of unfavorable behaviors. Multivariable multinomial logistic regression adjusted for age, sex, education and cardiovascular risk factors examined the associations between emphysema and CAC, with interactions and stratified analyses for lifestyle effects. Results: Emphysema was present in 62.3% of the participants, with centrilobular being the most common subtype (61.5%). Paraseptal emphysema was associated with both CAC 1-100 (OR: 2.07 [1.03-4.15]) and CAC > 100 (OR: 2.94 [1.26-6.84]). Severe emphysema was linked to CAC > 100 (OR: 3.50 [1.38-8.84]). These associations were stronger in the intermediate unhealthy lifestyle group for paraseptal (OR: 5.41 [1.70-17.22] and moderate and severe emphysema (OR: 9.64 [1.64-56.55]; OR: 3.73 [1.07-13.06]), respectively, but not significantly different. Conclusions: While paraseptal and severe emphysema are associated with higher CAC scores, there is no modifying effect of lifestyle factors. These findings suggest that cardiovascular risk assessment could be of importance in individuals with emphysema. Further longitudinal studies are needed to clarify the clinical implications.

To evaluate and compare low-dose CT (LDCT)-defined pulmonary nodule features between individuals who never smoked and who smoke(d) in a Chinese general population. This study included 2033 participants from the Nelcin-B3 cohort who underwent baseline LDCT. Trained radiologists reviewed each CT scan and assessed nodule CT features, including nodule density, size, location, shape, edge, attachment type, calcification and perifissural nodules (PFNs). Multilevel logistic regression (adjusted for age and sex) was performed to evaluate the relationship between nodule CT features and smoking status, accounting for nodule clustering within participants. Overall, 36.7

BACKGROUND:We studied changes in mammographic density (MD) among premenopausal women with a pathogenic germline variant (PGV) in the BRCA1 or BRCA2 gene, comparing those who did and did not undergo risk-reducing salpingo-oophorectomy (RRSO) in the interval between mammograms, accounting for changes in exogenous oral contraceptive or hormone replacement therapy (HRT) use. METHODS:From five studies of the International BRCA1/2 Carrier Cohort Study consortium, we included 691 participants who had two or more screening mammograms available, were less than 47 years at the time of RRSO (N = 208), or premenopausal at all mammograms without RRSO (N = 483). MD metrics [percent density (PD), dense area (DA), and non-DA] were quantified using STRATUS. Multivariable linear mixed models assessed changes in MD metrics between groups, adjusting for confounders. RESULTS:The mean PD at first mammogram was 26.8% ± 15.3 (RRSO) and 31.3% ± 18.1 (no RRSO). In a median 1.1 years between mammograms, PD decreased on average by 0.9% [95% confidence interval (CI), -1.6 to -0.2] among women who did not undergo RRSO in the interval between mammograms compared with 5.9% (95% CI, -7.4 to -4.5) among women who underwent RRSO in the interval (adjusted difference, -5.9%; 95% CI, -9.5 to -2.2; P = 0.002). Results were driven primarily by MD changes among BRCA2 PGV carriers. The use of HRT after RRSO attenuated the decline in PD. CONCLUSIONS:On average, PD and DA decrease following RRSO in premenopausal carriers, particularly among BRCA2 PGV carriers. HRT formulation affects MD changes. IMPACT:A decrease in MD may inform the potential protective effect of RRSO against breast cancer.

To assess the impact of reconstruction parameters on AI’s performance in detecting and classifying risk-dominant nodules in a baseline low-dose CT (LDCT) screening among a Chinese general population. Baseline LDCT scans from 300 consecutive participants in the Netherlands and China Big-3 (NELCIN-B3) trial were included. AI analyzed each scan reconstructed with four settings: 1 mm/0.7 mm thickness/interval with medium-soft and hard kernels (D45f/1 mm, B80f/1 mm) and 2 mm/1 mm with soft and medium-soft kernels (B30f/2 mm, D45f/2 mm). Reading results from consensus read by two radiologists served as reference standard. At scan level, inter-reader agreement between AI and reference standard, sensitivity, and specificity in determining the presence of a risk-dominant nodule were evaluated. For reference-standard risk-dominant nodules, nodule detection rate, and agreement in nodule type classification between AI and reference standard were assessed. AI-D45f/1 mm demonstrated a significantly higher sensitivity than AI-B80f/1 mm in determining the presence of a risk-dominant nodule per scan (77.5

OBJECTIVES:To investigate the performance of two segmentation algorithms for nodule volumetric classification at participant/scan level in the NELCIN-B3 cohort (Netherlands and China Big-3), a lung cancer screening program (LCS) using low-dose CT (LDCT). METHODS:Baseline scans with qualified LDCT images from consecutive NELCIN-B3 participants were included from June 2017 to July 2018. Performance of two software algorithms were independently evaluated by two radiologists: software A (Syngo.via VB30A) by reader 1 and software B (AVIEW v1.1.39.14) by reader 2. According to the NELSON2.0 protocol, nodules with a solid component ≥ 100 mm3 were classified as indeterminate-positive, while all other nodules were classified as negative. Disagreements in classification were resolved by consensus with three senior radiologists. These results served as a reference standard for identifying positive misclassifications (PM) and negative misclassifications (NM). RESULTS:In total, 300 participants were evaluated comprising 159 women (53.0 %) and 193 (64.3 %) never smokers, with a mean ± standard deviation age of 61.2 ± 7.1 years. There were disagreements in 17 cases: in 11 (11/300, 3.7 %), this was due to differences in nodule selection and nodule type classification between readers; and in 6 (6/300, 2.0 %), this was due to variations in nodule volume metrics between algorithms. Inter-software agreement was almost perfect (κ = 0.88 [95 %CI: 0.83-0.93]). In the consensus read, reader 1/software A generated 12 misclassifications (11 PM, 1 NM), giving a negative predictive value of 99.6 % (95 % CI: 98.9 %-100.0 %). Reader 2/software B generated 5 misclassifications (2 PM, 3 NM), giving a negative predictive value of 98.9 % (95 % CI: 97.7 %-100.0 %). CONCLUSION:Two software algorithms (Syngo.via VB30A and AVIEW v1.1.39.14) showed comparable performance for lung nodule volumetric classification at participant/scan level. Further research is needed to confirm the results in other LDCT LCS programs.

To assess the co-occurrence of incidental CT lung findings (emphysema, bronchiectasis, and airway wall thickening) as well as associated risk factors in low-dose CT (LDCT) lung cancer screening in a Chinese urban population. Data from 978 participants aged 40–74 years from the Chinese NELCIN-B3 urban population study who underwent LDCT screening were selected. CT scans were reviewed for incidental lung findings: emphysema, bronchiectasis and airway wall thickness. Emphysema was defined in three ways (≥ trace, ≥ mild, or ≥ moderate) depending on severity. Participants were described and stratified by presence or absence of incidental lung findings. Logistic regression analyses were performed to examine the relationship between participant characteristics and CT findings. Mean age was 61.3 years ± 6.8 and 533 (54.6

To develop a lobe-based bronchial scoring system in a general Chinese urban population undergoing low-dose CT (LDCT) screening and examining the association between the scores and the presence and absence of respiratory symptoms. A total of 989 Chinese participants aged 40–74 from the NELCIN-B3 study underwent LDCT screening. The scoring system assessed bronchiectasis by summing up CT findings in each of the five lung lobes, including severity and extent of bronchial dilatation and airway wall thickness, as well as the presence of mucoid impaction. The modified Reiff score was used as comparison. Multivariable logistic regression analyses were performed to examine the relationship between bronchial scores and respiratory symptoms. The study included 44.8

Supplementary Figure from Airflow Limitation Increases Lung Cancer Risk in Smokers: The Lifelines Cohort Study

Objective:Given the improved sensitivity of magnetic resonance imaging (MRI) for detecting ductal carcinoma in situ (DCIS), the omission of routine mammography (MG) or digital breast tomosynthesis (DBT) in high-risk breast cancer screening is under consideration. We aim to conduct a systematic review and meta-analysis to compare the screening sensitivity of MRI, MG and DBT for detecting DCIS in high-risk females. Methods:PubMed, Embase, and Web of Science were searched for studies reporting the sensitivity of detecting DCIS in high-risk females up to July 02, 2025. Study quality was assessed with quality assessment of diagnostic accuracy studies-2 (QUADAS-2). Pooled sensitivity was estimated using a random-effects model, overall and stratified by age (<40 and ≥40 years old) and BRCA status (BRCA1 and BRCA2). Meta-regression was used to compare modalities. Results:Seventeen studies (18,348 participants, 211 with DCIS) were included. MRI showed significantly higher pooled sensitivity [85%, 95% confidence interval (95% CI): 74%-94%] than MG (36%, 95% CI: 23%-50%; P<0.001). No DBT data were available. Combined MRI and MG yielded the highest sensitivity (99%, 95% CI: 97%-100%), but offered no significant gain over MRI alone in females <40 years old (P=0.091) and in BRCA1 mutation carriers (P=0.143). Conclusions:MRI is more sensitive than MG for DCIS detection in high-risk females. In females <40 years old and BRCA1 mutation carriers, adding MG to MRI provides no additional diagnostic value. Considering the potential trade-offs, the routine use of MG in these subgroups should be carefully reconsidered.

PURPOSE:To develop a novel simulation model for ductal carcinoma in situ (DCIS), fully validate it, and provide new estimates for DCIS in the setting of population-based biennial screening. METHODS:A micro-simulation Markov model for DCIS (SimDCIS) was developed. Input parameters were independently derived from the literature and transition parameters were age- and grade-dependent. The model was applied to the Dutch biennial screening program. SimDCIS was internally, cross, and externally validated by comparison of the model output to data from the Netherlands Cancer Registry, a modelling study on the United Kingdom Frequency Trial, and the United Kingdom screening program, respectively. Univariate and probabilistic sensitivity analyses were performed to estimate uncertainty. DCIS regression, progression to invasive breast cancer (IBC), clinical detection, and screen-detection were estimated in Dutch screening setting. RESULTS:SimDCIS matched observed data in internal, external, and cross-validation. The model was most sensitive to DCIS onset probability, and the maximum variation in screen-detection rate was 11%. In Dutch screening setting, DCIS regression, progression to IBC, clinical detection, and screen-detection were estimated at 8% (0-14%), 19% (16-24%), 8% (0-13%), and 61% (56-65%), respectively. Grade distribution was 20% grade 1, 38% grade 2, and 42% grade 3. CONCLUSION:SimDCIS provides strong accuracy across validation methods and is particularly sensitive to DCIS onset probability. Most DCIS will be found through screening, of which less than 50% of DCIS will be grade 3, less than 1 in 10 will regress, and 1 out of 5 DCIS will progress to IBC in biennial screening setting.

AIM:To estimate ductal carcinoma in situ (DCIS) overdiagnosis overall and by grade in population-based screening and to determine the variation in overdiagnosis estimates by definition. METHODS:Using a fully validated micro-simulation Markov model for DCIS (SimDCIS), the number, rate, and proportion of DCIS overdiagnoses were estimated overall and by grade. Overdiagnoses comprised excess DCIS cases in the screened versus the unscreened population; overdiagnosis rate equaled the number of DCIS overdiagnoses per 100,000 screened women; and DCIS overdiagnosis proportion equaled overdiagnosed DCIS divided by total diagnosed DCIS in the screened population. Base estimates for overdiagnosed DCIS were from a population perspective (ages 50-100 years) and included screen-detected, clinically detected, or progressed DCIS (i.e., invasive breast cancer with DCIS precursor). Overdiagnosis was also estimated for alternative definitions and perspectives. Univariate and probabilistic sensitivity analyses were performed to estimate uncertainty. RESULTS:Base definitions yielded an overdiagnosis rate of 38.1 (range, 25.7-58.7) per 100,000 screened women and a proportion of 20 % (range 13 %-30 %). Stratification by grade showed 24 %, 20 %, and 18 % proportion overdiagnosis for grades 1, 2, and 3, respectively. Varying the definition led to overdiagnosis estimates from 18 % to 94 %; these overdiagnosis estimates increased by 36 %-49 % when excluding invasive breast cancer and by 54 %-71 % when including only screen-detected DCIS. Individual perspective estimates were 12 % higher than population perspective estimates. CONCLUSION:In biennial screening, approximately 1 in 5 DCIS is overdiagnosed, but with minimal variation between grades. A consensus definition and perspective for overdiagnosis would reduce the observed variation in DCIS overdiagnosis estimates.

BACKGROUND:Accurate breast density evaluation allows for more precise risk estimation but suffers from high inter-observer variability. PURPOSE:To evaluate the feasibility of reducing inter-observer variability of breast density assessment through artificial intelligence (AI) assisted interpretation. STUDY TYPE:Retrospective. POPULATION:Six hundred and twenty-one patients without breast prosthesis or reconstructions were randomly divided into training (N = 377), validation (N = 98), and independent test (N = 146) datasets. FIELD STRENGTH/SEQUENCE:1.5 T and 3.0 T; T1-weighted spectral attenuated inversion recovery. ASSESSMENT:Five radiologists independently assessed each scan in the independent test set to establish the inter-observer variability baseline and to reach a reference standard. Deep learning and three radiomics models were developed for three classification tasks: (i) four Breast Imaging-Reporting and Data System (BI-RADS) breast composition categories (A-D), (ii) dense (categories C, D) vs. non-dense (categories A, B), and (iii) extremely dense (category D) vs. moderately dense (categories A-C). The models were tested against the reference standard on the independent test set. AI-assisted interpretation was performed by majority voting between the models and each radiologist's assessment. STATISTICAL TESTS:Inter-observer variability was assessed using linear-weighted kappa (κ) statistics. Kappa statistics, accuracy, and area under the receiver operating characteristic curve (AUC) were used to assess models against reference standard. RESULTS:In the independent test set, five readers showed an overall substantial agreement on tasks (i) and (ii), but moderate agreement for task (iii). The best-performing model showed substantial agreement with reference standard for tasks (i) and (ii), but moderate agreement for task (iii). With the assistance of the AI models, almost perfect inter-observer variability was obtained for tasks (i) (mean κ = 0.86), (ii) (mean κ = 0.94), and (iii) (mean κ = 0.94). DATA CONCLUSION:Deep learning and radiomics models have the potential to help reduce inter-observer variability of breast density assessment. LEVEL OF EVIDENCE:3 TECHNICAL EFFICACY: Stage 1.