Insights from the Breast Cancer Screening Trials: How Screening Affects the Natural History of Breast Cancer and Implications for Evaluating Service Screening Programs

Considerations for using potential surrogate endpoints in cancer screening trials

The requirement of large-scale expensive cancer screening trials spanning decades creates considerable barriers to the development, commercialisation, and implementation of novel screening tests. One way to address these problems is to use surrogate endpoints for the ultimate endpoint of interest, cancer mortality, at an earlier timepoint. This Review aims to highlight the issues underlying the choice and use of surrogate endpoints for cancer screening trials, to propose criteria for when and how we might use such endpoints, and to suggest possible candidates. We present the current landscape and challenges, and discuss lessons and shortcomings from the therapeutic trial setting. It is hugely challenging to validate a surrogate endpoint, even with carefully designed clinical studies. Nevertheless, we consider whether there are candidates that might satisfy the requirements defined by research and regulatory bodies.

Revolutionizing Breast Cancer Detection With Artificial Intelligence (AI) in Radiology and Radiation Oncology: A Systematic Review

The number one cause of cancer in women worldwide is breast cancer. Over the last three decades, the use of traditional screen-film mammography has increased, but in recent years, digital mammography and 3D tomosynthesis have become standard procedures for breast cancer screening. With the advancement of technology, the interpretation of images using automated algorithms has become a subject of interest. Initially, computer-aided detection (CAD) was introduced; however, it did not show any long-term benefit in clinical practice. With recent advances in artificial intelligence (AI) methods, these technologies are showing promising potential for more accurate and efficient automated breast cancer detection and treatment. While AI promises widespread integration in breast cancer detection and treatment, challenges such as data quality, regulatory, ethical implications, and algorithm validation are crucial. Addressing these is essential for fully realizing AI's potential in enhancing early diagnosis and improving patient outcomes in breast cancer management. In this review article, we aim to provide an overview of the latest developments and applications of AI in breast cancer screening and treatment. While the existing literature primarily consists of retrospective studies, ongoing and future prospective research is poised to offer deeper insights. Artificial intelligence is on the verge of widespread integration into breast cancer detection and treatment, holding the potential to enhance early diagnosis and improve patient outcomes.

Transfer learning for the generalization of artificial intelligence in breast cancer detection: a case-control study

BACKGROUND

Some researchers have questioned whether artificial intelligence (AI) systems maintain their performance when used for women from populations not considered during the development of the system.

PURPOSE

To evaluate the impact of transfer learning as a way of improving the generalization of AI systems in the detection of breast cancer.

MATERIAL AND METHODS

This retrospective case-control Finnish study involved 191 women diagnosed with breast cancer and 191 matched healthy controls. We selected a state-of-the-art AI system for breast cancer detection trained using a large US dataset. The selected baseline system was evaluated in two experimental settings. First, we examined our private Finnish sample as an independent test set that had not been considered in the development of the system (unseen population). Second, the baseline system was retrained to attempt to improve its performance in the unseen population by means of transfer learning. To analyze performance, we used areas under the receiver operating characteristic curve (AUCs) with DeLong's test.

RESULTS

Two versions of the baseline system were considered: ImageOnly and Heatmaps. The ImageOnly and Heatmaps versions yielded mean AUC values of 0.82±0.008 and 0.88±0.003 in the US dataset and 0.56 (95% CI=0.50-0.62) and 0.72 (95% CI=0.67-0.77) when evaluated in the unseen population, respectively. The retrained systems achieved AUC values of 0.61 (95% CI=0.55-0.66) and 0.69 (95% CI=0.64-0.75), respectively. There was no statistical difference between the baseline system and the retrained system.

CONCLUSION

Transfer learning with a small study sample did not yield a significant improvement in the generalization of the system.

Mammography-based deep learning model for coronary artery calcification

AIMS

Mammography, commonly used for breast cancer screening in women, can also predict cardiovascular disease. We developed mammography-based deep learning models for predicting coronary artery calcium (CAC) scores, an established predictor of coronary events.

METHODS AND RESULTS

We evaluated a subset of Korean adults who underwent image mammography and CAC computed tomography and randomly selected approximately 80% of the participants as the training dataset, used to develop a convolutional neural network (CNN) to predict detectable CAC. The sensitivity, specificity, area under the receiver operating characteristic curve (AUROC), and overall accuracy of the model's performance were evaluated. The training and validation datasets included 5235 and 1208 women, respectively [mean age, 52.6 (±10.2) years], including non-zero cases (46.8%). The CNN-based deep learning prediction model based on the Resnet18 model showed the best performance. The model was further improved using contrastive learning strategies based on positive and negative samples: sensitivity, 0.764 (95% CI, 0.667-0.830); specificity, 0.652 (95% CI, 0.614-0.710); AUROC, 0.761 (95% CI, 0.742-0.780); and accuracy, 70.8% (95% CI, 68.8-72.4). Moreover, including age and menopausal status in the model further improved its performance (AUROC, 0.776; 95% CI, 0.762-0.790). The Framingham risk score yielded an AUROC of 0.736 (95% CI, 0.712-0.761).

CONCLUSION

Mammography-based deep learning models showed promising results for predicting CAC, performing comparably to conventional risk models. This indicates mammography's potential for dual-risk assessment in breast cancer and cardiovascular disease. Further research is necessary to validate these findings in diverse populations, with a particular focus on representation from national breast screening programmes.

Cancer screening with multicancer detection tests: A translational science review

Multicancer detection (MCD) tests use a single, easily obtainable biospecimen, such as blood, to screen for more than one cancer concurrently. MCD tests can potentially be used to improve early cancer detection, including cancers that currently lack effective screening methods. However, these tests have unknown and unquantified benefits and harms. MCD tests differ from conventional cancer screening tests in that the organ responsible for a positive test is unknown, and a broad diagnostic workup may be necessary to confirm the location and type of underlying cancer. Among two prospective studies involving greater than 16,000 individuals, MCD tests identified those who had some cancers without currently recommended screening tests, including pancreas, ovary, liver, uterus, small intestine, oropharyngeal, bone, thyroid, and hematologic malignancies, at early stages. Reported MCD test sensitivities range from 27% to 95% but differ by organ and are lower for early stage cancers, for which treatment toxicity would be lowest and the potential for cure might be highest. False reassurance from a negative MCD result may reduce screening adherence, risking a loss in proven public health benefits from standard-of-care screening. Prospective clinical trials are needed to address uncertainties about MCD accuracy to detect different cancers in asymptomatic individuals, whether these tests can detect cancer sufficiently early for effective treatment and mortality reduction, the degree to which these tests may contribute to cancer overdiagnosis and overtreatment, whether MCD tests work equally well across all populations, and the appropriate diagnostic evaluation and follow-up for patients with a positive test.

[Structured screening for sporadic breast cancer]

BACKGROUND

The aim of secondary prevention of breast cancer is to detect the disease at the earliest curable stage and thus to reduce breast cancer-specific mortality. To this end, the nationwide population-based mammography screening program (MSP) was set up in Germany in 2005 in addition to an interdisciplinary prevention project for high-risk groups.

OBJECTIVE

Overview of the current state of the MSP, the upcoming age expansion, and potential further developments.

MATERIAL AND METHODS

Narrative review article with topic-guided literature and data search.

RESULTS

Approximately 50% of the 70,500 new cases of breast cancer that occur each year are related to the age group of the MSP. 10 years after introduction of the MSP, the incidence of advanced breast cancer stages and breast cancer-related mortality of the screening target group have steadily decreased by about one quarter, while no relevant trends were seen in the neighboring age groups at the population level.

CONCLUSION

The MSP has effectively contributed to a reduction of breast cancer mortality. With the expansion of the age groups to 45-75 years, more women have access to structured, quality assured screening. With the use of advanced stratifications and diagnostics as well as artificial intelligence, the MSP could be further optimized.

Autoantibodies, antigen-autoantibody complexes and antigens complement CA125 for early detection of ovarian cancer

BACKGROUND

Multiple antigens, autoantibodies (AAb), and antigen-autoantibody (Ag-AAb) complexes were compared for their ability to complement CA125 for early detection of ovarian cancer.

METHODS

Twenty six biomarkers were measured in a single panel of sera from women with early stage (I-II) ovarian cancers (n = 64), late stage (III-IV) ovarian cancers (186), benign pelvic masses (200) and from healthy controls (502), and then split randomly (50:50) into a training set to identify the most promising classifier and a validation set to compare its performance to CA125 alone.

RESULTS

Eight biomarkers detected ≥ 8% of early stage cases at 98% specificity. A four-biomarker panel including CA125, HE4, HE4 Ag-AAb and osteopontin detected 75% of early stage cancers in the validation set from among healthy controls compared to 62% with CA125 alone (p = 0.003) at 98% specificity. The same panel increased sensitivity for distinguishing early-stage ovarian cancers from benign pelvic masses by 25% (p = 0.0004) at 95% specificity. From 21 autoantibody candidates, 3 AAb (anti-p53, anti-CTAG1 and annt-Il-8) detected 22% of early stage ovarian cancers, potentially lengthening lead time prior to diagnosis.

CONCLUSION

A four biomarker panel achieved greater sensitivity at the same specificity for early detection of ovarian cancer than CA125 alone.

Clinical performance and utility: A microsimulation model to inform the design of screening trials for a multi-cancer early detection test

OBJECTIVES

Designing cancer screening trials for multi-cancer early detection (MCED) tests presents a significant methodology challenge, as natural histories of cell-free DNA-shedding cancers are not yet known. A microsimulation model was developed to project the performance and utility of an MCED test in cancer screening trials.

METHODS

Individual natural history of preclinical progression through cancer stages for 23 cancer classes was simulated by a stage-transition model under a broad range of cancer latency parameters. Cancer incidences and stage distributions at clinical presentation in simulated trials were set to match the data from Surveillance, Epidemiology, and End Results program. One or multiple rounds of annual screening using a targeted methylation-based MCED test (GalleriⓇ) was conducted to detect preclinical cancers. Mortality benefit of early detection was simulated by a stage-shift model.

RESULTS

In simulated trials, accounting for healthy volunteer effect and varying test sensitivity, positive predictive value in the prevalence screening round reached 48% to 61% in 6 natural history scenarios. After 3 rounds of annual screening, the cumulative proportions of stage I/II cancers increased by approximately 9% to 14%, the incidence of stage IV cancers was reduced by 37% to 46%, the reduction of stages III and IV cancer incidences was 9% to 24%, and the reduction of mortality reached 13% to 16%. Greater reductions of late-stage cancers and cancer mortality were achieved by five rounds of MCED screening.

CONCLUSIONS

Simulation results guide trial design and suggest that adding this MCED test to routine screening in the United States may shift cancer detection to earlier stages, and potentially save lives.

Temporal patterns of cancer burden in Asia, 1990-2019: a systematic examination for the Global Burden of Disease 2019 study

Background

Cancers represent a challenging public health threat in Asia. This study examines the temporal patterns of incidence, mortality, disability and risk factors of 29 cancers in Asia in the last three decades.

Methods

The age, sex and year-wise estimates of incidence, mortality, and disability-adjusted life years (DALYs) of 29 cancers for 49 Asian countries from 1990 through 2019 were generated as a part of the Global Burden of Disease, Injuries and Risk Factors 2019 study. Besides incidence, mortality and DALYs, we also examined the cancer burden measured in terms of DALYs and deaths attributable to risk factors, which had evidence of causation with different cancers. The development status of countries was measured using the socio-demographic index. Decomposition analysis was performed to gauge the change in cancer incidence between 1990 and 2019 due to population growth, aging and age-specific incidence rates.

Findings

All cancers combined claimed an estimated 5.6 million [95% uncertainty interval, 5.1-6.0 million] lives in Asia with 9.4 million [8.6-10.2 million] incident cases and 144.7 million [132.7-156.5 million] DALYs in 2019. The age-standardized incidence rate (ASIR) of all cancers combined in Asia was 197.6/100,000 [181.0-214.4] in 2019, varying from 99.2/100,000 [76.1-126.0] in Bangladesh to 330.5/100,000 [298.5-365.8] in Cyprus. The age-standardized mortality rate (ASMR) was 120.6/100,000 [110.1-130.7] in 2019, varying 4-folds across countries from 71.0/100,000 [59.9-83.5] in Kuwait to 284.2/100,000 [229.2-352.3] in Mongolia. The age-standardized DALYs rate was 2970.5/100,000 [2722.6-3206.5] in 2019, varying from 1578.0/100,000 [1341.2-1847.0] in Kuwait to 6574.4/100,000 [5141.7-8333.0] in Mongolia. Between 1990 and 2019, deaths due to 17 of the 29 cancers either doubled or more, and 20 of the 29 cancers underwent an increase of 150% or more in terms of new cases. Tracheal, bronchus, and lung cancer (both sexes), breast cancer (among females), colon and rectum cancer (both sexes), stomach cancer (both sexes) and prostate cancer (among males) were among top-5 cancers in most Asian countries in terms of ASIR and ASMR in 2019 and cancers of liver, stomach, hodgkin lymphoma and esophageal cancer posted the most significant decreases in age-standardized rates between 1990 and 2019. Among the modifiable risk factors, smoking, alcohol use, ambient particulate matter (PM) pollution and unsafe sex remained the dominant risk factors between 1990 and 2019. Cancer DALYs due to ambient PM pollution, high body mass index and fasting plasma glucose has increased most notably between 1990 and 2019.

Interpretation

With growing incidence, cancer has become more significant public health threat in Asia, demanding urgent policy attention and guidance. Its heightened risk calls for increased cancer awareness, preventive measures, affordable early-stage detection, and cost-effective therapeutics in Asia. The current study can serve as a useful resource for policymakers and researchers in Asia for devising interventions for cancer management and control.

Funding

The GBD study is funded by the Bill and Melinda Gates Foundation.

Artificial Intelligence for Breast Cancer Detection on Mammography: Factors Related to Cancer Detection

RATIONALE AND OBJECTIVES

Little is known about the factors affecting the Artificial Intelligence (AI) software performance on mammography for breast cancer detection. This study was to identify factors associated with abnormality scores assigned by the AI software.

MATERIALS AND METHODS

A retrospective database search was conducted to identify consecutive asymptomatic women who underwent breast cancer surgery between April 2016 and December 2019. A commercially available AI software (Lunit INSIGHT, MMG, Ver. 1.1.4.0) was used for preoperative mammography to assign individual abnormality scores to the lesions and score of 10 or higher was considered as positive detection by AI software. Radiologists without knowledge of the AI results retrospectively assessed the mammographic density and classified mammographic findings into positive and negative finding. General linear model (GLM) analysis was used to identify the clinical, pathological, and mammographic findings related to the abnormality scores, obtaining coefficient β values that represent the mean difference per unit or comparison with the reference value. Additionally, the reasons for non-detection by the AI software were investigated.

RESULTS

Among the 1001 index cancers (830 invasive cancers and 171 ductal carcinoma in situs) in 1001 patients, 717 (72%) were correctly detected by AI, while the remaining 284 (28%) were not detected. Multivariable GLM analysis showed that abnormal mammography findings (β = 77.0 for mass, β = 73.1 for calcification only, β = 49.4 for architectural distortion, and β = 47.6 for asymmetry compared to negative; all Ps < 0.001), invasive tumor size (β = 4.3 per 1 cm, P < 0.001), and human epidermal growth receptor type 2 (HER2) positivity (β = 9.2 compared to hormone receptor positive, HER2 negative, P = 0.004) were associated with higher mean abnormality score. AI failed to detect small asymmetries in extremely dense breasts, subcentimeter-sized or isodense lesions, and faint amorphous calcifications.

CONCLUSION

Cancers with positive abnormal mammographic findings on retrospective review, large invasive size, HER2 positivity had high AI abnormality scores. Understanding the patterns of AI software performance is crucial for effectively integrating AI into clinical practice.

Normal Risk Ovarian Screening Study: 21-Year Update

PURPOSE

The Normal Risk Ovarian Screening Study (NROSS) tested a two-stage screening strategy in postmenopausal women at conventional hereditary risk where significantly rising cancer antigen (CA)-125 prompted transvaginal sonography (TVS) and abnormal TVS prompted surgery to detect ovarian cancer.

METHODS

A total of 7,856 healthy postmenopausal women were screened annually for a total of 50,596 woman-years in a single-arm study (ClinicalTrials.gov identifier: NCT00539162). Serum CA125 was analyzed with the Risk of Ovarian Cancer Algorithm (ROCA) each year. If risk was unchanged and <1:2,000, women returned in a year. If risk increased above 1:500, TVS was undertaken immediately, and if risk was intermediate, CA125 was repeated in 3 months with a further increase in risk above 1:500 prompting referral for TVS. An average of 2% of participants were referred to TVS annually.

RESULTS

Thirty-four patients were referred for operations detecting 15 ovarian cancers and two borderline tumors with 12 in early stage (I-II). In addition, seven endometrial cancers were detected with six in stage I. As four ovarian cancers and two borderline tumors were diagnosed with a normal ROCA, the sensitivity for detecting ovarian and borderline cancer was 74% (17 of 23), and 70% of ROCA-detected cases (12 of 17) were in stage I-II. NROSS screening reduced late-stage (III-IV) disease by 34% compared with UKCTOCS controls and by 30% compared with US SEER values. The positive predictive value (PPV) was 50% (17 of 34) for detecting ovarian cancer and 74% (25 of 34) for any cancer, far exceeding the minimum acceptable study end point of 10% PPV.

CONCLUSION

While the NROSS trial was not powered to detect reduced mortality, the high specificity, PPV, and marked stage shift support further development of this strategy.

Tumour stage, treatment, and survival of women with high-grade serous tubo-ovarian cancer in UKCTOCS: an exploratory analysis of a randomised controlled trial

BACKGROUND

In UKCTOCS, there was a decrease in the diagnosis of advanced stage tubo-ovarian cancer but no reduction in deaths in the multimodal screening group compared with the no screening group. Therefore, we did exploratory analyses of patients with high-grade serous ovarian cancer to understand the reason for the discrepancy.

METHODS

UKCTOCS was a 13-centre randomised controlled trial of screening postmenopausal women from the general population, aged 50-74 years, with intact ovaries. The trial management system randomly allocated (2:1:1) eligible participants (recruited from April 17, 2001, to Sept 29, 2005) in blocks of 32 using computer generated random numbers to no screening or annual screening (multimodal screening or ultrasound screening) until Dec 31, 2011. Follow-up was through national registries until June 30, 2020. An outcome review committee, masked to randomisation group, adjudicated on ovarian cancer diagnosis, histotype, stage, and cause of death. In this study, analyses were intention-to-screen comparisons of women with high-grade serous cancer at censorship (Dec 31, 2014) in multimodal screening versus no screening, using descriptive statistics for stage and treatment endpoints, and the Versatile test for survival from randomisation. This trial is registered with the ISRCTN Registry, 22488978, and ClinicalTrials.gov, NCT00058032.

FINDINGS

202 562 eligible women were recruited (50 625 multimodal screening; 50 623 ultrasound screening; 101 314 no screening). 259 (0·5%) of 50 625 participants in the multimodal screening group and 520 (0·5%) of 101 314 in the no screening group were diagnosed with high-grade serous cancer. In the multimodal screening group compared with the no screening group, fewer were diagnosed with advanced stage disease (195 [75%] of 259 vs 446 [86%] of 520; p=0·0003), more had primary surgery (158 [61%] vs 219 [42%]; p<0·0001), more had zero residual disease following debulking surgery (119 [46%] vs 157 [30%]; p<0·0001), and more received treatment including both surgery and chemotherapy (192 [74%] vs 331 [64%]; p=0·0032). There was no difference in the first-line combination chemotherapy rate (142 [55%] vs 293 [56%]; p=0·69). Median follow-up from randomisation of 779 women with high-grade serous cancer in the multimodal and no screening groups was 9·51 years (IQR 6·04-13·00). At censorship (June 30, 2020), survival from randomisation was longer in women with high-grade serous cancer in the multimodal screening group than in the no screening group with absolute difference in survival of 6·9% (95% CI 0·4-13·0; p=0·042) at 18 years (21% [95% CI 15·6-26·2] vs 14% [95% CI 10·5-17·4]).

INTERPRETATION

To our knowledge, this is the first evidence that screening can detect high-grade serous cancer earlier and lead to improved short-term treatment outcomes compared with no screening. The potential survival benefit for women with high-grade serous cancer was small, most likely due to only modest gains in early detection and treatment improvement, and tumour biology. The cumulative results of the trial suggest that surrogate endpoints for disease-specific mortality should not currently be used in screening trials for ovarian cancer.

FUNDING

National Institute for Health Research, Medical Research Council, Cancer Research UK, The Eve Appeal.

Breast Cancer Screening for Women at Higher-Than-Average Risk: Updated Recommendations From the ACR

Early detection decreases breast cancer death. The ACR recommends annual screening beginning at age 40 for women of average risk and earlier and/or more intensive screening for women at higher-than-average risk. For most women at higher-than-average risk, the supplemental screening method of choice is breast MRI. Women with genetics-based increased risk, those with a calculated lifetime risk of 20% or more, and those exposed to chest radiation at young ages are recommended to undergo MRI surveillance starting at ages 25 to 30 and annual mammography (with a variable starting age between 25 and 40, depending on the type of risk). Mutation carriers can delay mammographic screening until age 40 if annual screening breast MRI is performed as recommended. Women diagnosed with breast cancer before age 50 or with personal histories of breast cancer and dense breasts should undergo annual supplemental breast MRI. Others with personal histories, and those with atypia at biopsy, should strongly consider MRI screening, especially if other risk factors are present. For women with dense breasts who desire supplemental screening, breast MRI is recommended. For those who qualify for but cannot undergo breast MRI, contrast-enhanced mammography or ultrasound could be considered. All women should undergo risk assessment by age 25, especially Black women and women of Ashkenazi Jewish heritage, so that those at higher-than-average risk can be identified and appropriate screening initiated.

Dying To Find Out: The Cost of Time at the Dawn of the Multicancer Early Detection Era

Cancer is a significant burden worldwide that adversely impacts life expectancy, quality of life, health care costs, and workforce productivity. Although currently recommended screening tests for individual cancers reduce mortality, they detect only a minority of all cancers and sacrifice specificity for high sensitivity, resulting in a high cumulative rate of false positives. Blood-based multicancer early detection tests (MCED) based on next-generation sequencing (NGS) and other technologies hold promise for broadening the number of cancer types detected in screened populations and hope for reducing cancer mortality. The promise of this new technology to improve cancer detection rates and make screening more efficient at the population level demands the development of novel trial designs that accelerate clinical adoption. Carefully designed clinical trials are needed to address these issues.

Screening Mammography Utilization in the United States

Breast cancer is the second leading cause of cancer mortality in adult women in the United States. Screening mammography reduces breast cancer mortality between 22% and 48%; however, screening mammography remains underutilized. Screening mammography utilization data are available from insurance claims, electronic medical records, and patient self-report via surveys, and each data source has unique benefits and challenges. Numerous barriers exist that adversely affect the use of screening mammography in the United States. This article will review screening mammography utilization in the United States, explore factors that impact utilization, and briefly discuss strategies to improve utilization.

Evaluation of an AI Model to Assess Future Breast Cancer Risk

Background Accurate breast cancer risk assessment after a negative screening result could enable better strategies for early detection. Purpose To evaluate a deep learning algorithm for risk assessment based on digital mammograms. Materials and Methods A retrospective observational matched case-control study was designed using the OPTIMAM Mammography Image Database from the National Health Service Breast Screening Programme in the United Kingdom from February 2010 to September 2019. Patients with breast cancer (cases) were diagnosed following a mammographic screening or between two triannual screening rounds. Controls were matched based on mammography device, screening site, and age. The artificial intelligence (AI) model only used mammograms at screening before diagnosis. The primary objective was to assess model performance, with a secondary objective to assess heterogeneity and calibration slope. The area under the receiver operating characteristic curve (AUC) was estimated for 3-year risk. Heterogeneity according to cancer subtype was assessed using a likelihood ratio interaction test. Statistical significance was set at P < .05. Results Analysis included patients with screen-detected (median age, 60 years [IQR, 55-65 years]; 2044 female, including 1528 with invasive cancer and 503 with ductal carcinoma in situ [DCIS]) or interval (median age, 59 years [IQR, 53-65 years]; 696 female, including 636 with invasive cancer and 54 with DCIS) breast cancer and 1:1 matched controls, each with a complete set of mammograms at the screening preceding diagnosis. The AI model had an overall AUC of 0.68 (95% CI: 0.66, 0.70), with no evidence of a significant difference between interval and screen-detected (AUC, 0.69 vs 0.67; P = .085) cancer. The calibration slope was 1.13 (95% CI: 1.01, 1.26). There was similar performance for the detection of invasive cancer versus DCIS (AUC, 0.68 vs 0.66; P = .057). The model had higher performance for advanced cancer risk (AUC, 0.72 ≥stage II vs 0.66 Collapse

Key Words Collapse

MESH Headings

• Humans
• Female
• Middle Aged
• Breast Neoplasms/diagnostic imaging
• Carcinoma, Intraductal, Noninfiltrating
• Artificial Intelligence
• Case-Control Studies
• Retrospective Studies
• State Medicine

Collapse

Grants

• 28689 Cancer Research UK
• C49757/A28689 Cancer Research UK

Collapse

Affiliation(s) Collapse

Standalone AI for Breast Cancer Detection at Screening Digital Mammography and Digital Breast Tomosynthesis: A Systematic Review and Meta-Analysis

Background There is considerable interest in the potential use of artificial intelligence (AI) systems in mammographic screening. However, it is essential to critically evaluate the performance of AI before it can become a modality used for independent mammographic interpretation. Purpose To evaluate the reported standalone performances of AI for interpretation of digital mammography and digital breast tomosynthesis (DBT). Materials and Methods A systematic search was conducted in PubMed, Google Scholar, Embase (Ovid), and Web of Science databases for studies published from January 2017 to June 2022. Sensitivity, specificity, and area under the receiver operating characteristic curve (AUC) values were reviewed. Study quality was assessed using the Quality Assessment of Diagnostic Accuracy Studies 2 and Comparative (QUADAS-2 and QUADAS-C, respectively). A random effects meta-analysis and meta-regression analysis were performed for overall studies and for different study types (reader studies vs historic cohort studies) and imaging techniques (digital mammography vs DBT). Results In total, 16 studies that include 1 108 328 examinations in 497 091 women were analyzed (six reader studies, seven historic cohort studies on digital mammography, and four studies on DBT). Pooled AUCs were significantly higher for standalone AI than radiologists in the six reader studies on digital mammography (0.87 vs 0.81, P = .002), but not for historic cohort studies (0.89 vs 0.96, P = .152). Four studies on DBT showed significantly higher AUCs in AI compared with radiologists (0.90 vs 0.79, P < .001). Higher sensitivity and lower specificity were seen for standalone AI compared with radiologists. Conclusion Standalone AI for screening digital mammography performed as well as or better than radiologists. Compared with digital mammography, there is an insufficient number of studies to assess the performance of AI systems in the interpretation of DBT screening examinations. © RSNA, 2023 Supplemental material is available for this article. See also the editorial by Scaranelo in this issue.

Women 75 Years Old or Older: To Screen or Not to Screen?

Breast cancer is the most common cancer in women, with the incidence rising substantially with age. Older women are a vulnerable population at increased risk of developing and dying from breast cancer. However, women aged 75 years and older were excluded from all randomized controlled screening trials, so the best available data regarding screening benefits and risks in this age group are from observational studies and modeling predictions. Benefits of screening in older women are the same as those in younger women: early detection of smaller lower-stage cancers, resulting in less invasive treatment and lower morbidity and mortality. Mammography performs significantly better in older women with higher sensitivity, specificity, cancer detection rate, and positive predictive values, accompanied by lower recall rates and false positives. The overdiagnosis rate is low, with benefits outweighing risks until age 90 years. Although there are conflicting national and international guidelines about whether to continue screening mammography in women beyond age 74 years, clinicians can use shared decision making to help women make decisions about screening and fully engage them in the screening process. For women aged 75 years and older in good health, continuing annual screening mammography will save the most lives. An informed discussion of the benefits and risks of screening mammography in older women needs to include each woman's individual values, overall health status, and comorbidities. This article will review the benefits, risks, and controversies surrounding screening mammography in women 75 years old and older and compare the current recommendations for screening this population from national and international professional organizations. ^©RSNA, 2023 Quiz questions for this article are available through the Online Learning Center.

Performance analysis of screening mammography in Asian women under 40 years

BACKGROUND

Screening mammography performance among young women remains uncertain in East Asia, where the proportion of young breast cancer patients is higher than that in Western countries. Thus, we analyzed the performance of screening mammography in women under 40 years in comparison with older age groups.

METHODS

This retrospective study comprised 95,431 Asian women with 197,525 screening mammograms. The reference standard was determined by linkage to the national cancer registry data and the 12-month follow-up outcomes after the index mammogram. The performance metrics included sensitivity, specificity, cancer detection rate (CDR), positive predictive value (PPV), recall rate, and areas under the receiver operating characteristic curve (AUCs), with comparisons across age groups (30 s, 40 s, and ≥ 50 s).

RESULTS

For young women aged < 40 years, sensitivity and AUC (95% confidence interval [CI]) of screening mammography were 60.4% (50.4-69.7) and 0.73 (0.68-0.77), respectively, with no significant difference compared to women in their 40 s (sensitivity: 64.0% [95% CI: 57.8-69.8], P = 0.52; AUC: 0.75 [95% CI: 0.73-0.78], P = 0.35). The CDR (95% CI) was 0.8 (0.6-1.1) per 1,000 mammograms for young women, poorer than 1.8 (1.6-2.1) per 1,000 mammograms for women in their 40 s (P < 0.001). The PPV and recall rate (95% CI) for young women were 0.6% (0.4-0.7) and 14.9% (14.6-15.1), poorer than 1.4% (1.2-1.6) and 13.3% (13.1-13.5) for women in their 40 s (P < 0.001), respectively.

CONCLUSION

The accuracy of screening mammography for young women in their 30 s was not significantly different from that for women in their 40 s, but the cancer detection and recall rates were poorer.

Mammographic Breast Density: Current Assessment Methods, Clinical Implications, and Future Directions

Mammographic breast density is widely accepted as an independent risk factor for the development of breast cancer. In addition, because dense breast tissue may mask breast malignancies, breast density is inversely related to the sensitivity of screening mammography. Given the risks associated with breast density, as well as ongoing efforts to stratify individual risk and personalize breast cancer screening and prevention, numerous studies have sought to better understand the factors that impact breast density, and to develop and implement reproducible, quantitative methods to assess mammographic density. Breast density assessments have been incorporated into risk assessment models to improve risk stratification. Recently, novel techniques for analyzing mammographic parenchymal complexity, or texture, have been explored as potential means of refining mammographic tissue-based risk assessment beyond breast density.

Effects of electronic personal health information technology on American women's cancer screening behaviors mediated through cancer worry: Differences and similarities between 2017 and 2020

Backgrounds

Thanks to their accessibility and low cost, electronic personal health information (ePHI) technologies have been widely used to facilitate patient-physician communication and promote health prevention behaviors (e.g. cancer screening). Despite that empirical evidence has supported the association between ePHI technology use and cancer screening behaviors, the underlying mechanism through which ePHI technology use influences cancer screening behaviors remains a topic of discussion.

Objective

This study investigates the relationship between ePHI technology uses and cancer screening behaviors of American women and examines the mediating role of cancer worry.

Methods

Data for this study were from the Health Information National Trends Survey (HINTS) collected in 2017 (HINTS 5 Cycle 1) and 2020 (HINTS 5 Cycle 4). The final sample included 1914 female respondents in HINTS 5 Cycle 1 and 2204 in HINTS 5 Cycle 4. Mann-Whitney U test, two-sample t-test, and mediation analysis were performed. We also referred to the regression coefficients generated by min-max normalization as percentage coefficients (bp) for the comparison.

Results

This study reports increased usage of ePHI technologies (from 1.41 in 2017 to 2.19 to 2020), increased cancer worry (from 2.60 in 2017 to 2.84 in 2020), and a stable level of cancer screening behaviors (from 1.44 in 2017 to 1.34 in 2020) among American women. Cancer worry was found to mediate the ePHI effect on cancer screening behaviors (bp = 0.005, 95% confidence interval [0.001, 0.010]) in a positive complementary mediation in 2020.

Conclusions

The research findings support a positive association between ePHI technology use and cancer screening behaviors, and cancer worry has been identified as a salient mediator. An understanding of the mechanism that prompts US women's cancer screening practices provides practical implications for health campaign practitioners.

The Landmark Series-Addressing Disparities in Breast Cancer Screening: New Recommendations for Black Women

Randomized, clinical trials have established the efficacy of screening mammography in improving survival from breast cancer for women through detection of early, asymptomatic disease. However, disparities in survival rates between black women and women from other racial and ethnic groups following breast cancer diagnosis persist. Various professional groups have different, somewhat conflicting, guidelines with regards to recommended age for commencing screening as well as recommended frequency of screening exams, but the trials upon which these recommendations are based were not specifically designed to examine benefit among black women. Furthermore, these recommendations do not appear to incorporate the unique epidemiological circumstances of breast cancer among black women, including higher rates of diagnosis before age 40 years and greater likelihood of advanced stage at diagnosis, into their formulation. In this review, we examined the epidemiologic and socioeconomic factors that are associated with breast cancer among black women and assess the implications of these factors for screening in this population. Specifically, we recommend that by no later than age 25 years, all black women should undergo baseline assessment for future risk of breast cancer utilizing a model that incorporates race (e.g., Breast Cancer Risk Assessment Tool [BCRAT], formerly the Gail model) and that this assessment should be conducted by a breast specialist or a healthcare provider (e.g., primary care physician or gynecologist) who is trained to assess breast cancer risk and is aware of the increased risks of early (i.e., premenopausal) and biologically aggressive (e.g., late-stage, triple-negative) breast cancer among black women.

Depiction of breast cancers on digital mammograms by artificial intelligence-based computer-assisted diagnosis according to cancer characteristics

OBJECTIVE

To evaluate how breast cancers are depicted by artificial intelligence-based computer-assisted diagnosis (AI-CAD) according to clinical, radiological, and pathological factors.

MATERIALS AND METHODS

From January 2017 to December 2017, 896 patients diagnosed with 930 breast cancers were enrolled in this retrospective study. Commercial AI-CAD was applied to digital mammograms and abnormality scores were obtained. We evaluated the abnormality score according to clinical, radiological, and pathological characteristics. False-negative results were defined by abnormality scores less than 10.

RESULTS

The median abnormality score of 930 breasts was 87.4 (range 0-99). The false-negative rate of AI-CAD was 19.4% (180/930). Cancers with an abnormality score of more than 90 showed a high proportion of palpable lesions, BI-RADS 4c and 5 lesions, cancers presenting as mass with or without microcalcifications and invasive cancers compared with low-scored cancers (all p < 0.001). False-negative cancers were more likely to develop in asymptomatic patients and extremely dense breasts and to be diagnosed as occult breast cancers and DCIS compared to detected cancers.

CONCLUSION

Breast cancers depicted with high abnormality scores by AI-CAD are associated with higher BI-RADS category, invasive pathology, and higher cancer stage.

KEY POINTS

• High-scored cancers by AI-CAD included a high proportion of BI-RADS 4c and 5 lesions, masses with or without microcalcifications, and cancers with invasive pathology. • Among invasive cancers, cancers with higher T and N stage and HER2-enriched subtype were depicted with higher abnormality scores by AI-CAD. • Cancers missed by AI-CAD tended to be in asymptomatic patients and extremely dense breasts and to be diagnosed as occult breast cancers by radiologists.

Diagnostic performance of the Kaiser score in the evaluation of breast lesions on contrast-enhanced mammography

OBJECTIVES

We aimed to investigate whether the Kaiser score (KS) could improve the diagnostic performance of breast imaging reporting and data system (BI-RADS) in evaluating breast enhancing lesions on contrast-enhanced mammography (CEM).

METHODS

Three hundred fifty-nine patients with 375 lesions (231 malignant and 144 benign) were included in this retrospective study from April 2019 to December 2021.Two readers with different levels of experience in breast imaging were asked to give a BI-RADS assessment category according to the CEM BI-RADS and final score based on the KS. The diagnostic performance of all lesions, mass and non-mass enhancement (NME) were assessed by receiver operating characteristic (ROC) analysis, and the areas under the ROC curve (AUCs) were measured. The weighted kappa coefficients were calculated to investigate the interreader agreement.

RESULTS

The AUCs of the KS for all lesions were 0.915 (95 %CI: 0.884-0.947) and 0.876 (95 %CI: 0.838-0.914) for two readers. When mass and NME were evaluated separately, the AUCs of the KS for mass were higher than those for NME (p < 0.001). The AUCs of BI-RADS for all lesion diagnoses ranged between 0.821 (95 %CI: 0.778-0.864) and 0.842(95 %CI: 0.801-0.883) for two readers. The AUCs of the KS were higher than those of BI-RADS (p < 0.001, p = 0.016). There were no significant differences in the sensitivity between the KS (97.4 %) and BI-RADS (99.6 %) for all lesions (p = 0.130). The specificity of the KS was significantly higher than that of BI-RADS (p < 0.001). Compared with BI-RADS, the application of the KS could have potentially obviated 41.7 % to 47.9 % unnecessary biopsies in 144 benign lesions. Interreader agreement between the two readers of the KS was almost perfect (k = 0.883 [95 % CI: 0.842-0.924]).

CONCLUSION

The use of the KS provided a high diagnostic performance in distinguishing malignant and benign breast lesions on CEM and outperformed BI-RADS. The application of the KS can downgrade up to 47.9% of unnecessary biopsies of benign breast lesions.

Value of contrast-enhanced mammography combined with the Kaiser score for clinical decision-making regarding tomosynthesis BI-RADS 4A lesions

OBJECTIVES

To investigate the diagnostic performance of contrast-enhanced mammography (CEM) combined with the Kaiser score (KS) in digital breast tomosynthesis (DBT) BI-RADS 4A lesions to potentially reduce unnecessary breast biopsies.

METHODS

This retrospective study evaluated 106 patients with 109 DBT BI-RADS 4A lesions from June 2019 to June 2021. For the absence of enhancement on CEM, the lesions were downgraded to BI-RADS 3. For lesions with enhancement, the readers were asked to classify all enhancing lesions referring to the KS for breast MRI. Receiver operating characteristic (ROC) curve analysis was used to evaluate the diagnostic performance. Two readers rated all cases and interreader agreement was assessed by Cohen's kappa coefficients.

RESULTS

There were ninety-five benign lesions and 14 malignant lesions. CEM combined with KS's accuracy, represented by the area under the curve (AUC), ranged between 0.880 and 0.906. The use of the KS improved the performance, with a significant difference relative to a single BI-RADS reading or US (p < 0.001). CEM with KS had higher specificity than CEM with BI-RADS or US (p < 0.001), without difference in sensitivity (p > 0.05). CEM combined with KS could have potentially obviated 72 (75.8%) to 78 (82.1%) unnecessary benign biopsies in 95 benign lesions previously DBT classified as BI-RADS 4A. The interreader agreement was substantial (kappa: 0.727) for KS.

CONCLUSIONS

CEM combined with KS may be used in DBT BI-RADS 4A lesions to substantially reduce unnecessary benign biopsies.

KEY POINTS

• CEM combined with the Kaiser scoring system shows high diagnostic performance for DBT BI-RADS 4A lesions. • The application of CEM combined with the Kaiser scoring system may avoid 75.8% to 82.1% of unnecessary benign breast biopsies. • CEM combined with the KS aids clinical decision-making in DBT BI-RADS 4A lesions.

10-year opportunistic mammographic screening scenario in Brazil and its impact on breast cancer early detection: a nationwide population-based study

Background

Methods

Results

Conclusions

Misinformation and Facts about Breast Cancer Screening

Quality medical practice is based on science and evidence. For over a half-century, the efficacy of breast cancer screening has been challenged, particularly for women aged 40–49. As each false claim has been raised, it has been addressed and refuted based on science and evidence. Nevertheless, misinformation continues to be promoted, resulting in confusion for women and their physicians. Early detection has been proven to save lives for women aged 40–74 in randomized controlled trials of mammography screening. Observational studies, failure analyses, and incidence of death studies have provided evidence that there is a major benefit when screening is introduced to the general population. In large part due to screening, there has been an over 40% decline in deaths from breast cancer since 1990. Nevertheless, misinformation about screening continues to be promoted, adding to the confusion. Despite claims to the contrary, a careful reading of the guidelines issued by major groups such as the U.S. Preventive Services Task Force and the American College of Physicians shows that they all agree that most lives are saved by screening starting at the age of 40. There is no scientific support for using the age of 50 as a threshold for screening. All women should be provided with the facts and not false information about breast cancer screening so that they can make “informed decisions” for themselves about whether to participate.

The Benefits of Early Detection: Evidence From Modern International Mammography Service Screening Programs

Research from randomized controlled trials initiated up to 60 years ago consistently confirms that regular screening with mammography significantly reduces breast cancer mortality. Despite this success, there is ongoing debate regarding the efficacy of screening, which is confounded by technologic advances and concerns about cost, overdiagnosis, overtreatment, and equitable care of diverse patient populations. More recent screening research, designed to quell the debates, derives data from variable study designs, each with unique strengths and weaknesses. This article reviews observational population-based screening research that has followed the early initial long-term randomized controlled trials that are no longer practical or ethical to perform. The advantages and disadvantages of observational data and study design are outlined, including the three subtypes of population-based observational studies: cohort/case-control, trend, and incidence-based mortality/staging. The most recent research, typically performed in countries that administer screening mammography to women through centralized health service programs and directly track patient-specific outcomes and detection data, is summarized. These data are essential to understand and inform construction of effective new databases that facilitate continuous assessment of optimal screening techniques in the current era of rapidly developing medical technology, combined with a focus on health care that is both personal and equitable.

The Effects of Prior Mammography Screening on the Performance of Breast Cancer Detection in Taiwan

The aim of this study was to investigate the influence of previous mammography screening on the performance of breast cancer detection. The screened women were divided into first-visit and follow-up groups for breast cancer screening. The positive predictive value (PPV), cancer detection rate (CDR), and recall rate were used to evaluate and analyze the overall screening performance among the two groups. Among them, 10,040 screenings (67.2%) were first visits and 4895 screenings (32.8%) were follow-up visits. The proportion of positive screening results for first-visit participants was higher than that for their follow-up counterparts (9.3% vs. 4.0%). A total of 98 participants (74 first-visit and 24 follow-up visit) were confirmed to have breast cancer. The PPV for positive mammography for women who underwent biopsy confirmation was 28.7% overall, reaching 35.8% for the follow-up visit group and 27.0% for the first-visit group. The CDR was 6.6 per 1000 overall, reaching 7.4 per 1000 for first-visit group and 4.9 per 1000 for the follow-up group. The overall recall rate was 7.9%, reaching 9.7% for the first-visit group and 4.2% for the follow-up group. The PPV is improved and the recall rate is decreased if prior mammography images are available for comparison when conducting mammography screening for breast cancer. By this study, we concluded that prior mammography plays an important role for breast cancer screening, while follow-up mammography may increase the diagnostic rate when compared to the prior mammography. We suggest that the public health authority can encourage subjects to undergo screenings in the same health institute where they regularly visit.

Multicancer Early Detection Technologies: A Review Informed by Past Cancer Screening Studies

More than 75% of cancer-related deaths occur from cancers for which we do not screen. New screening liquid biopsies may help fill these clinical gaps, although evidence of benefit still needs to be assessed. Which lessons can we learn from previous efforts to guide those of the future? Screening trials for ovarian, prostate, pancreatic, and esophageal cancers are revisited to assess the evidence, which has been limited by small effect sizes, short duration of early-stage disease relative to screening frequency, study design, and confounding factors. Randomized controlled trials (RCT) to show mortality reduction have required millions of screening-years, two-decade durations, and been susceptible to external confounding. Future RCTs with late-stage incidence as a surrogate endpoint could substantially reduce these challenges, and clinical studies demonstrating safety and effectiveness of screening in high-risk populations may enable extrapolation to broader average-risk populations. Multicancer early detection tests provide an opportunity to advance these practical study designs. Conditional approvals based on RCTs with surrogate endpoints, contingent upon real world evidence generation and continuation of trials to definitive endpoints, may lower practical barriers to innovation in cancer screening and enable greater progress.

How Did CNBSS Influence Guidelines for So Long and What Can That Teach Us?

The biased randomization and other quality concerns about the Canadian National Breast Screening Studies (CNBSS) were documented and criticized for decades, even by several individuals very close to the research. CNBSS were the outlier studies among several RCTs of the era and yet were given equal weighting and occasionally higher importance than the remainder of the canon of mammography RCTs. These studies have had an ongoing influence on subsequent evidence review, guideline formation, and, ultimately, patient access to screening. This article explores possible reasons for the ongoing inclusion of CNBSS in the body of mammography screening evidence, discusses the lack of expertise in critical healthcare guideline processes, and, ultimately, suggests several actions and reforms.

How a faecal immunochemical test screening programme changes annual colorectal cancer incidence rates: an Italian intention-to-screen study

Background

This study aimed to evaluate the effectiveness of a biennial faecal immunochemical test (FIT) screening programme in reducing annual colorectal cancer (CRC) incidence in its dynamic target population.

Methods

The target population included over 1,000,000 persons aged 50–69 living in a region of northern Italy. The average annual response rate to invitation was 51.4%. Each observed annual age-standardised (Europe) rate per 100,000 persons between 2005, the year of introduction of the programme, and 2016 was compared with each expected annual rate as estimated with age-period-cohort (men) and age-period (women) models.

Results

For both sexes, the rates observed in 1997–2004 and those expected in 2005–2016 were stable. Observed rates increased in 2005, peaked in 2006 (the first full year of screening), dropped significantly below the expected level in 2009, and continued to decrease until 2013 (the eighth full year), after which no further significant changes occurred. In the pooled years 2013–2016, the observed incidence rate per 100,000 persons was 102.2 [95% CI: 97.4, 107.1] for men, 75.6 [95% CI: 71.6, 79.7] for women and 88.4 [95% CI: 85.3, 91.5] for both sexes combined, with an observed:expected incidence rate ratio of 0.68 [95% CI: 0.65, 0.71], 0.79 [95% CI: 0.76, 0.82] and 0.72 [95% CI: 0.66, 0.81], respectively.

Discussion

The study provided multiple consistent proofs of a causal relationship between the introduction of screening and a stable 28% decrease in annual CRC incidence after eight years.

Breast cancer screening in low and middle-income countries

Breast cancer incidence rates are rising in low and middle-income countries (LMIC), and these populations have reduced access to advanced multidisciplinary treatment. Screening and early detection are therefore critical in these regions but must be affordable and sustainable. Mammography screening programs are well established in more affluent countries, but alternative strategies to reduce the breast cancer burden of LMIC (such as clinical breast examination, general breast health awareness, and addressing modifiable lifestyle factors including obesity) are likely to be more realistic at the present time.

Intratumoral analysis of digital breast tomosynthesis for predicting the Ki-67 level in breast cancer: A multi-center radiomics study

PURPOSE

To non-invasively evaluate the Ki-67 level in digital breast tomosynthesis (DBT) images of breast cancer (BC) patients based on subregional radiomics.

METHODS

A total of 266 patients who underwent DBT scans were consecutively enrolled at two centers, between September 2017 and September 2021. The whole tumor region was partitioned into various intratumoral subregions, based on individual- and population-level clustering. Handcrafted radiomics and deep learning-based features were extracted from the subregions and from the whole tumor region, and were selected by least absolute shrinkage and selection operator (LASSO) regression, yielding radiomics signatures (RSs). The area under the receiver operating characteristic curve (AUC), accuracy, sensitivity, and specificity were calculated to assess the developed RSs.

RESULTS

Each breast tumor region was partitioned into an inner subregion (S1) and a marginal subregion (S2). The RSs derived from S1 always generated higher AUCs compared with those from S2 or from the whole tumor region (W), for the external validation cohort (AUCs, S1 vs. W, handcrafted RSs: 0.583 [95% CI, 0.429-0.727] vs. 0.559 [95% CI, 0.405-0.705], p-value: 0.920; deep RSs: 0.670 [95% CI, 0.516-0.802] vs. 0.551 [95% CI, 0.397-0.698], p-value: 0.776). The fusion RSs, combining handcrafted and deep learning-based features derived from S1, yielded the highest AUCs of 0.820 (95% CI, 0.714-0.900) and 0.792 (95% CI, 0.647-0.897) for the internal and external validation cohorts, respectively.

CONCLUSIONS

The subregional radiomics approach can accurately predict the Ki-67 level based on DBT data; thus, it may be used as a potential non-invasive tool for preoperative treatment planning in BC.

Utility of Digital Breast Tomosynthesis with Two-Dimensional Synthesized Mammography Images: A Pictorial Essay

Background  Mammography has been established as the key modality in the detection and diagnosis of breast cancers. Digital breast tomosynthesis (DBT) has emerged as a mammographic technique which allows improved visualization of abnormalities by reducing the effect of overlapping breast tissue.

Purpose  This article is a pictorial essay which highlights the advantages of DBT with two-dimensional (2D) synthesized mammography (2DSM) images, its clinical applications, and its role in breast imaging.

Materials and Methods  Selenia Dimensions HD mammography machine performs DBT which acquires a series of low-dose digital mammographic images of the compressed breast followed by full-field digital mammography. Software using specialized algorithms helps to create a 2DSM image reconstructed from the DBT data set. The images are interpreted on a dedicated work station on high-resolution monitors by the radiologist. American College of Radiology Breast Imaging-Reporting and Data System (BI-RADS) lexicon is used for reporting. High-resolution breast ultrasound which includes evaluation of the axilla is done for all cases.

Conclusion  DBT improves detection and better characterization of lesions which thereby increases confidence of interpretation of mammograms and assigning BI-RADS categories for further management.

Effective Surveillance of High-Risk Women

BACKGROUND

This study addresses the effectiveness of risk models and screening breast magnetic resonance imaging (MRI) in women who have atypical hyperplasia (AH), lobular carcinoma in situ (LCIS), or a family history of breast cancer, but not a genetic mutation.

PATIENTS AND METHODS

A retrospective review of 444 women who had 458 breast screening MRIs at a community teaching hospital over a 12-month period between March 25, 2014 and March 31, 2015 was performed. The patients underwent high risk screening with breast MRIs alternating with mammograms every 6 months. After excluding patients with prior breast or ovarian cancer, genetic mutations, and chest wall radiation, 200 remaining patients constituted the study cohort. Over the following 5 years, the patients were screened with MRIs alternating with mammograms every 6 months. A total of 961 total MRI screenings were performed over the entire 5-year period of the study.

RESULTS

A total of 200 women fit the study criteria. Of these 103 had a prior history of AH or LCIS. Over the 5-year period, 60 women dropped out of the screening regimen, 6 patients were diagnosed with breast cancer on screening MRIs, and 2 additional patients were diagnosed with breast cancer on screening mammograms. Surprisingly, the highest Tyrer-Cuzick (T-C) scores did not correlate with increased development of breast cancers in our population.

CONCLUSIONS

This study shows that there is wide variation in the results of risk assessment models. Risk models may overestimate breast cancer risk, suggesting that re-evaluation of current risk assessment and screening protocols is warranted.

Toward robust mammography-based models for breast cancer risk

Improved breast cancer risk models enable targeted screening strategies that achieve earlier detection and less screening harm than existing guidelines. To bring deep learning risk models to clinical practice, we need to further refine their accuracy, validate them across diverse populations, and demonstrate their potential to improve clinical workflows. We developed Mirai, a mammography-based deep learning model designed to predict risk at multiple timepoints, leverage potentially missing risk factor information, and produce predictions that are consistent across mammography machines. Mirai was trained on a large dataset from Massachusetts General Hospital (MGH) in the United States and tested on held-out test sets from MGH, Karolinska University Hospital in Sweden, and Chang Gung Memorial Hospital (CGMH) in Taiwan, obtaining C-indices of 0.76 (95% confidence interval, 0.74 to 0.80), 0.81 (0.79 to 0.82), and 0.79 (0.79 to 0.83), respectively. Mirai obtained significantly higher 5-year ROC AUCs than the Tyrer-Cuzick model ( P < 0.001) and prior deep learning models Hybrid DL ( P < 0.001) and Image-Only DL ( P < 0.001), trained on the same dataset. Mirai more accurately identified high-risk patients than prior methods across all datasets. On the MGH test set, 41.5% (34.4 to 48.5) of patients who would develop cancer within 5 years were identified as high risk, compared with 36.1% (29.1 to 42.9) by Hybrid DL ( P = 0.02) and 22.9% (15.9 to 29.6) by the Tyrer-Cuzick model ( P < 0.001).

Current trends in incidence, characteristics, and surgical management of metastatic breast cancer to the spine: A National Inpatient Sample analysis from 2005 to 2014

Management of metastatic breast cancer to the spine (MBCS) incorporates a multimodal approach. Improvement in screening and nonsurgical therapies may alter the trends in surgical management of MBCS. The objective of this study is to assess trends in surgical management of MBCS and short-term outcomes based on the National Inpatient Sample (NIS) database. The NIS database was queried for patients with MBCS who underwent surgery from 2005 to 2014. The weighted frequencies of spinal decompression alone, spinal stabilization +/- decompression, and vertebral augmentation were calculated. Multivariate analysis was performed to analyze the effect of patient characteristics on outcomes stratified by procedure. The most common procedure performed was vertebral augmentation (11,114, 53.4%), followed by stabilization +/- decompression (6,906, 33.2%) and then decompression alone (3,312, 13.4%). The total population-adjusted rate of surgical management for MBCS remained stable, while the rate of spinal stabilization increased (P < 0.001) and vertebral augmentation decreased (p < 0.003). The risk of complication increased with spinal stabilization and decompression compared to vertebral augmentation procedures in those with fewer comorbidities. This relative increase in risk abated in patients with higher numbers of pre-operative comorbidities. Any single complication was associated with increases in length of stay, cost, and mortality. The rate of in-hospital interventions remained stable over the study period. Stratified by procedure, the rate of stabilizations increased with a concomitant decrease in vertebral augmentations, which suggests that patients who require hospitalization for MBCS are becoming more likely to represent advanced cases that are not amenable to palliative vertebral augmentation procedures.

Breast Cancer Screening Recommendations Inclusive of All Women at Average Risk: Update from the ACR and Society of Breast Imaging

Breast cancer remains the most common nonskin cancer, the second leading cause of cancer deaths, and the leading cause of premature death in US women. Mammography screening has been proven effective in reducing breast cancer deaths in women age 40 years and older. A mortality reduction of 40% is possible with regular screening. Treatment advances cannot overcome the disadvantage of being diagnosed with an advanced-stage tumor. The ACR and Society of Breast Imaging recommend annual mammography screening beginning at age 40, which provides the greatest mortality reduction, diagnosis at earlier stage, better surgical options, and more effective chemotherapy. Annual screening results in more screening-detected tumors, tumors of smaller sizes, and fewer interval cancers than longer screening intervals. Screened women in their 40s are more likely to have early-stage disease, negative lymph nodes, and smaller tumors than unscreened women. Delaying screening until age 45 or 50 will result in an unnecessary loss of life to breast cancer and adversely affects minority women in particular. Screening should continue past age 74 years, without an upper age limit unless severe comorbidities limit life expectancy. Benefits of screening should be considered along with the possibilities of recall for additional imaging and benign biopsy and the less tangible risks of anxiety and overdiagnosis. Although recall and biopsy recommendations are higher with more frequent screening, so are life-years gained and breast cancer deaths averted. Women who wish to maximize benefit will choose annual screening starting at age 40 years and will not stop screening prematurely.

Ovarian cancer population screening and mortality after long-term follow-up in the UK Collaborative Trial of Ovarian Cancer Screening (UKCTOCS): a randomised controlled trial

BACKGROUND

Ovarian cancer continues to have a poor prognosis with the majority of women diagnosed with advanced disease. Therefore, we undertook the UK Collaborative Trial of Ovarian Cancer Screening (UKCTOCS) to determine if population screening can reduce deaths due to the disease. We report on ovarian cancer mortality after long-term follow-up in UKCTOCS.

METHODS

In this randomised controlled trial, postmenopausal women aged 50-74 years were recruited from 13 centres in National Health Service trusts in England, Wales, and Northern Ireland. Exclusion criteria were bilateral oophorectomy, previous ovarian or active non-ovarian malignancy, or increased familial ovarian cancer risk. The trial management system confirmed eligibility and randomly allocated participants in blocks of 32 using computer generated random numbers to annual multimodal screening (MMS), annual transvaginal ultrasound screening (USS), or no screening, in a 1:1:2 ratio. Follow-up was through national registries. The primary outcome was death due to ovarian or tubal cancer (WHO 2014 criteria) by June 30, 2020. Analyses were by intention to screen, comparing MMS and USS separately with no screening using the versatile test. Investigators and participants were aware of screening type, whereas the outcomes review committee were masked to randomisation group. This study is registered with ISRCTN, 22488978, and ClinicalTrials.gov, NCT00058032.

FINDINGS

Between April 17, 2001, and Sept 29, 2005, of 1 243 282 women invited, 202 638 were recruited and randomly assigned, and 202 562 were included in the analysis: 50 625 (25·0%) in the MMS group, 50 623 (25·0%) in the USS group, and 101 314 (50·0%) in the no screening group. At a median follow-up of 16·3 years (IQR 15·1-17·3), 2055 women were diagnosed with tubal or ovarian cancer: 522 (1·0%) of 50 625 in the MMS group, 517 (1·0%) of 50 623 in the USS group, and 1016 (1·0%) of 101 314 in the no screening group. Compared with no screening, there was a 47·2% (95% CI 19·7 to 81·1) increase in stage I and 24·5% (-41·8 to -2·0) decrease in stage IV disease incidence in the MMS group. Overall the incidence of stage I or II disease was 39·2% (95% CI 16·1 to 66·9) higher in the MMS group than in the no screening group, whereas the incidence of stage III or IV disease was 10·2% (-21·3 to 2·4) lower. 1206 women died of the disease: 296 (0·6%) of 50 625 in the MMS group, 291 (0·6%) of 50 623 in the USS group, and 619 (0·6%) of 101 314 in the no screening group. No significant reduction in ovarian and tubal cancer deaths was observed in the MMS (p=0·58) or USS (p=0·36) groups compared with the no screening group.

INTERPRETATION

The reduction in stage III or IV disease incidence in the MMS group was not sufficient to translate into lives saved, illustrating the importance of specifying cancer mortality as the primary outcome in screening trials. Given that screening did not significantly reduce ovarian and tubal cancer deaths, general population screening cannot be recommended.

FUNDING

National Institute for Health Research, Cancer Research UK, and The Eve Appeal.

Examination of incidence, mortality and disability-adjusted life years and risk factors of breast cancer in 49 Asian countries, 1990-2019: estimates from Global Burden of Disease Study 2019

BACKGROUND

This study presents an up-to-date, comprehensive and comparative examination of breast cancer's temporal patterns in females in Asia in last three decades.

METHODS

The estimates of incidence, mortality, disability-adjusted-life-years and risk factors of breast cancer in females in 49 Asian countries were retrieved from Global Burden of Disease 2019 study.

RESULTS

In Asia, female breast cancer incidence grew from 245 045[226 259-265 260] in 1990 to 914 878[815 789-1025 502] in 2019 with age-standardized incidence rate rising from 21.2/100 000[19.6-22.9] to 35.9/100 000[32.0-40.2] between 1990 and 2019. The death counts more than doubled from 136 665[126 094-148 380] to 337 822[301 454-375 251]. The age-standardized mortality rate rose marginally between 1990 and 2019 (1990: 12.1[11.0-13.1]; 2019: 13.4[12.0-14.9]). In 2019, age-standardized incidence rate varied from 17.2/100 000[13.95-21.4] in Mongolia to 122.5[92.1-160.7] in Lebanon and the age-standardized mortality rate varied 4-fold from 8.0/100 000 [7.2-8.8] in South Korea to 51.9[39.0-69.8] in Pakistan. High body mass index (5.6%), high fasting plasma glucose (5.6%) and secondhand smoke (3.5%) were the main contributory risk factors to all-age disability-adjusted-life-years due to breast cancer in Asia.

CONCLUSION

With growing incidence, escalating dietary and behavioural risk factors and lower survival rates due to late-disease presentation in low- and medium-income countries of Asia, breast cancer has become a significant public health threat. Its rising burden calls for increasing breast cancer awareness, preventive measures, early-stage detection and cost-effective therapeutics in Asia.

Digital Breast Tomosynthesis: an Overview

Breast cancer is emerging as the most common malignancy in Indian women. Mammography is one of the few screening modalities available to the modern world that has proved itself of much use by aiding early detection and treatment of non-palpable, node-negative breast cancers. However, due to its two-dimensional nature, many cases of malignancies are still missed, to be detected at a later date or by an alternate modality. In 2011, FDA approved the supplemental use of digital breast tomosynthesis (DBT) in screening and diagnostic set ups. The acquisition of multiple low-dose projection images of the compressed parenchyma provided a 'third' dimension to the mammogram whereby the breast tissue could be seen layer by layer on the workstation. It improves cancer detection rate, and reduces recall rate and false-positive findings by improving lesion characterization. The current review discusses the principle of DBT with a comprehensive study of the literature.

Supplementary Information

The online version contains supplementary material available at 10.1007/s13193-021-01310-y.

Application of artificial intelligence-based computer-assisted diagnosis on synthetic mammograms from breast tomosynthesis: comparison with digital mammograms

OBJECTIVE

To compare the diagnostic agreement and performances of synthetic and conventional mammograms when artificial intelligence-based computer-assisted diagnosis (AI-CAD) is applied.

MATERIAL AND METHOD

From January 2017 to April 2017, 192 patients (mean age 53.7 ± 11.7 years) diagnosed with 203 breast cancers were enrolled in this retrospective study. All patients underwent digital breast tomosynthesis (DBT) with digital mammograms (DM) simultaneously. Commercial AI-CAD was applied to the reconstructed synthetic mammograms (SM) from DBT and DM respectively and abnormality scores were calculated. We compared the median abnormality scores between DM and SM with the Wilcoxon signed-rank test and used the Bland-Altman analysis to evaluate agreements between the two mammograms and to investigate clinicopathological factors which might affect agreement. Diagnostic performances were compared using an area under the receiver operating characteristic curve (AUC).

RESULT

The abnormality scores showed a mean difference (bias) of - 3.26 (95% limits of agreement: - 32.69, 26.18) between the two mammograms by the Bland-Altman analysis. The concordance correlation coefficient was 0.934 (95% CI: 0.92, 0.946), suggesting high reproducibility. SM showed higher abnormality scores in cancer with distortion and occult findings, T1 and N0 cancer, and luminal type cancer than DM (all p ≤ 0.001). Diagnostic performance did not differ between the mammograms (AUC 0.945 for conventional mammograms, 0.938 for synthetic mammograms, p = 0.499).

CONCLUSION

AI-CAD can also work well on synthetic mammograms, showing good agreement and comparable diagnostic performance compared to its application to DM.

KEY POINTS

• AI-CAD which was developed based on imaging findings of digital mammograms can also be applied to synthetic mammograms. • AI-CAD showed good agreement and similar diagnostic performance when applied to both synthetic and digital mammograms. • With AI-CAD, synthetic mammograms showed relatively higher abnormality scores in cancer with distortion and occult findings, T1 and N0 cancer, and luminal type cancer than digital mammograms.