Radiomics strategies for risk assessment of tumour failure in head-and-neck cancer

Soft-tissue Sarcomas (STS) are a heterogeneous group of malignant neoplasms with a relatively high mortality rate from distant metastases. Early prediction or quantitative evaluation of distant metastases risk for patients with STS is an important step which can provide better-personalized treatments and thereby improve survival rates. Positron emission tomography-computed tomography (PET-CT) image is regarded as the imaging modality of choice for the evaluation, staging and assessment of STS. Radiomics, which refers to the extraction and analysis of the quantitative of high-dimensional mineable data from medical images, is foreseen as an important prognostic tool for cancer risk assessment. However, conventional radiomics methods that depend heavily on hand-crafted features (e.g. shape and texture) and prior knowledge (e.g. tuning of many parameters) therefore cannot fully represent the semantic information of the image. In addition, convolutional neural networks (CNN) based radiomics methods present capabilities to improve, but currently, they are mainly designed for single modality e.g., CT or a particular body region e.g., lung structure. In this work, we propose a deep multi-modality collaborative learning to iteratively derive optimal ensembled deep and conventional features from PET-CT images. In addition, we introduce an end-to-end volumetric deep learning architecture to learn complementary PET-CT features optimised for image radiomics. Our experimental results using public PET-CT dataset of STS patients demonstrate that our method has better performance when compared with the state-of-the-art methods.

Simple SummaryOesophago-gastric cancer is one of the commonest cancers worldwide, yet it can be particularly difficult to diagnose given that initial symptoms are often non-specific and routine screening is not available. Cancer risk-assessment tools, which calculate cancer risk based on symptoms and other risk factors present in the primary care record, can aid decisions on referrals for cancer investigations, facilitating earlier diagnosis. Diagnosing common cancers earlier could help improve survival rates. Using UK primary care electronic health record data, we compared five different machine learning techniques for probabilistic classification of cancer patients against a current widely used UK primary care cancer risk-assessment tool. The machine learning algorithms outperformed the current risk-assessment tool, with a higher overall accuracy and an ability to reasonably identify 11–25% more cancer patients. We conclude that machine-learning-based risk-assessment tools could help better identify suitable patients for further investigation and support earlier diagnosis.Oesophago-gastric cancer is difficult to diagnose in the early stages given its typical non-specific initial manifestation. We hypothesise that machine learning can improve upon the diagnostic performance of current primary care risk-assessment tools by using advanced analytical techniques to exploit the wealth of evidence available in the electronic health record. We used a primary care electronic health record dataset derived from the UK General Practice Research Database (7471 cases; 32,877 controls) and developed five probabilistic machine learning classifiers: Support Vector Machine, Random Forest, Logistic Regression, Naïve Bayes, and Extreme Gradient Boosted Decision Trees. Features included basic demographics, symptoms, and lab test results. The Logistic Regression, Support Vector Machine, and Extreme Gradient Boosted Decision Tree models achieved the highest performance in terms of accuracy and AUROC (0.89 accuracy, 0.87 AUROC), outperforming a current UK oesophago-gastric cancer risk-assessment tool (ogRAT). Machine learning also identified more cancer patients than the ogRAT: 11.0% more with little to no effect on false positives, or up to 25.0% more with a slight increase in false positives (for Logistic Regression, results threshold-dependent). Feature contribution estimates and individual prediction explanations indicated clinical relevance. We conclude that machine learning could improve primary care cancer risk-assessment tools, potentially helping clinicians to identify additional cancer cases earlier. This could, in turn, improve survival outcomes.

A prognostic scoring system for locoregional control in nasopharyngeal carcinoma following conformal radiotherapy

There is no validated, discriminating, and easy-to-apply tool for estimating risk of colorectal neoplasia. We studied whether the National Cancer Institute's (NCI's) Colorectal Cancer (CRC) Risk Assessment Tool, which estimates future CRC risk, could estimate current risk for advanced colorectal neoplasia among average-risk persons. This cross-sectional study involved individuals age 50 to 80 years undergoing first-time screening colonoscopy. We measured medical and family history, lifestyle information, and physical measures and calculated each person's future CRC risk using the NCI tool's logistic regression equation. We related quintiles of future CRC risk to the current risk of advanced neoplasia (sessile serrated polyp or tubular adenoma ≥ 1 cm, a polyp with villous histology or high-grade dysplasia, or CRC). All statistical tests were two-sided. For 4457 (98.5%) with complete data (mean age = 57.2 years, SD = 6.6 years, 51.7% women), advanced neoplasia prevalence was 8.26%. Based on quintiles of five-year estimated absolute CRC risk, current risks of advanced neoplasia were 2.1% (95% confidence interval [CI] = 1.3% to 3.3%), 4.8% (95% CI = 3.5% to 6.4%), 6.4% (95% CI = 4.9% to 8.2%), 10.0% (95% CI = 8.1% to 12.1%), and 17.6% (95% CI = 15.5% to 20.6%; P < .001). For quintiles of estimated 10-year CRC risk, corresponding current risks for advanced neoplasia were 2.2% (95% CI = 1.4% to 3.5%), 4.8% (95% CI = 3.5% to 6.4%), 6.5% (95% CI = 5.0% to 8.3%), 9.3% (95% CI = 7.5% to 11.4%), and 18.4% (95% CI = 15.9% to 21.1%; P < .001). Among persons with an estimated five-year CRC risk above the median, current risk for advanced neoplasia was 12.8%, compared with 3.7% among those below the median (relative risk = 3.4, 95 CI = 2.7 to 4.4). The NCI's Risk Assessment Tool, which estimates future CRC risk, may be used to estimate current risk for advanced neoplasia, making it potentially useful for tailoring and improving CRC screening efficiency among average-risk persons.

Uncertain Temporal Logics for Risk Analysis.- Individual Differences in Risk Perception and Risk-Taking Preferences.- Technological Risk Perception: Similarities and Dissimilarities in French and American Samples.- Risk Perception and Energy Policy: A Swedish Case Study.- Dispersion of Dense Air Toxics.- Carpet/4-Phenylcyclohexene Toxicity: The EPA Headquarters Case.- Analysis of Community Risk Resulting from Rupture of a Sour Gas Pipeline.- Risk Assessment for Children Playing on Lawns Treated with a Pesticide.- Uncertainty Management in Engineering Risk Assessment.- An Approach to the Analysis of Accident Precursors.- Risk Perception and the Politics of Citizen Participation: The Case of Radioactive Waste Management.- Screening Lifetime Risks to Help Prevent Radon Exposure: A Methodology for Public Health Policy Makers.- Dose-Response Relationship Between Arsenic Inhalation Exposure and Risk of Lung Cancer.- Development of Uncertainty Factors for Nonhuman Receptors.- Addressing Uncertainties in Environmental Site Audits.- Farmworkers and Pesticide Exposure: Perceived Risk and Self-Protective Behavior.- Age-Dependent Risk Quantification Using Standard Maintenance Records.- Use of Risk Analysis on Remedial Alternatives.- Victims, Agents, and Outrage.- An Introduction to the Texas Risk Communication Process.- Risk Communication and the Cognitive Representation of Uncertainty.- Common Methodological Flaws in Risk Assessment.- Use of Risk Assessment Methodologies for Sewage Sludge Disposal Regulations.- Presentation of Risk Assessments of Carcinogens.- Exposure to Environmental Contaminants Through Breast Milk.- Cancer Risk from the Application of Newspaper to Farmland.- Alar in Fruit: Limited Regulatory Action in the Face of Uncertain Risks.- Extending Biologically-Based Cancer Risk Modeling to Apply to Benzene-Induced Leukemogenesis.- Reducing the Risk in Buying Risk Analysis Software.- Probabilistic Causality and Its Applications to Risk Analysis.- A Comprehensive Risk Analysis Example.- Public Perception and Response to the Parkfield California Earthquake Prediction.- Preliminary Analysis of Off-Normal Emissions from a Hazardous Waste Incinerator.- Estimating Emissions from Municipal Solid Waste Incinerators.- The Role of Structure-Activity Relationships in Risk Assessment.- A Framework for Computer Security Risk Management.- The Rules of the Game: What Recent Rulings Say About Courts' and Regulators' Differing Approaches to Establishing Causation for Chronic Health Risks.- Determinants of Severity in Acutely Hazardous Chemical Releases.- The 1988 Ashland Oil Spill: Lessons Learned Regarding Emergency Response Planning for Hazmat Releases.- Potential Uses and Abuses of Toxic Release Inventory Data.- Robust Estimation of Lung Retention in Animals Exposed to a-Emitting Radionuclides.- Dispositional Differences of 2, 3, 7, S-Tetrachlorodibenzo-p-dioxin in Rats and Humans: Implications in Cancer Mechanisms and Risk Assessment.- The Risk Analysis of Extreme Events in Queuing Theory.- On the Inclusion of Organizational and Managerial Influences in Probabilistic Safety Assessments of Nuclear Power Plants.- Risk Communication and Regulatory Culture Clash.- Bridging the Gap Between Risk Assessment by Professionals and Acceptance by Lay Decision Makers.- Acceptability of Risk: Vinyl Chloride and RCRA Subtitle D, Potential Decision Framework.- Consistent Estimates of Urethane Carcinogenic Potency Using Pharmacokinetic and Time-Dependent Models.- Carcinogenic Impurities in Food and Color Additives-An Analysis of Presumptive Risk Levels.- Decision Theory, Failure Tolerance Analysis, and Quantitative Risk Analysis -Synergistic Application to the Space Station Freedom WP-2 Program.- Sources and Consequences of Hypothetical Bias in Economic Analyses of Risk Behavior.- Establishing Communication with Brookhurst: A Case Study in Community Involvement.- Accepting New Technology: Community Relations for Mobile Incineration in Illinois.- Evaluation of Analytical Methods in Water for the Chemicals Listed Under the California Safe Drinking Water and Toxic Enforcement Act of 1986 (Proposition 65).- Use of Health Risk Estimates in U.S. EPA Activities.- The Use of Quantitative Risk Assessment in the Continuing Risk Management of a Chlorine Handling Facility.- Ozone Risk Assessment Implementation: Transferring Risk Analysis Technology to Ozone Nonattainment Planning.- Acute Ozone Exposure-Response Relationships for Use in Health Risk Assessment.- Risk-Benefit Balancing in Risk Management: Measures of Benefits and Detriments.- Risk Assessment of Soil-Related Foundation Failure.- Estimating Mine Equipment Injury Rates.- Ecological Models for Risk Assessment/Risk Management.- Science Quality and Risk Assessment: The Leaking Landfill.- Risk Assessment-Risk Management: The Need for a Synthesis.- Application of an Urban-Scale Population Exposure Model (NEM/SAI) to Ozone Exposure Assessment.- New Approaches and Tools for Quantitative Cancer Risk Assessment.- Statistical Properties of a Model-Free Approach to Low-Dose Extrapolation.- Author Index.

Background: Contralateral axillary lymph node metastasis (CAM) in breast cancer is currently classified as a stage IV disease but its prognosis is still controversial. Purpose: To determine outcomes in overall survival (OS) and disease-free survival (DFS) in patients with and without locoregional tumor recurrence who present with contralateral axillary lymph node metastasis (CAM). Methods: Patients with pathologically confirmed invasive breast cancer with metachronous CAM who received treatment between 1988 and 2017 were retrospectively reviewed. Patients with other distant metastases at the time of CAM diagnosis were excluded. The outcome of CAM in cases of IBTR and regional recurrence (RR) were compared to CAM not accompanied by locoregional tumor recurrence. Results: Thirty-eight patients with metachronous CAM were included in the study. Metachronous CAM occurred 55 months (interquartile range, 17-77 months) after surgical treatment of the primary tumor and median follow-up was 95 months (interquartile range, 49-117 months) from the initial operation date and 40 months (interquartile range, 15-54 months) from the diagnosis of CAM. At the time of initial CAM diagnosis, 11 patients had IBTR, 12 patients had RR, and 15 patients had no locoregional recurrence. The estimated 5-year OS was 49.1% and 5-year DFS was 45.3%. Although statistically insignificant due to small sample size, when stratified by loco regional recurrence, the prognosis of CAM patients with IBTR appeared to be better than those without locoregional recurrence (5-year OS: 88.9% vs. 41.4%, HR 5.88, p = 0.09) whereas the prognosis of CAM patients with RR was worse than those without locoregional recurrence (5-year OS: 35.4% vs. 41.4%, HR 0.44, p = 0.20). Axillary lymph node dissection (ALND) improved median OS (83 vs. 36 months, p = 0.069) in all patients. When stratified, improvement in median OS was 13 vs 27 months (p = 0.094) in patients with RR, and 36 vs. 65 months (p = 0.061) in patients without locoregional recurrence. For patients accompanied by IBTR, ALND was performed in 8 out of 11 and only one patient died during the follow-up period. Conclusion: Our study indicates that the patients with CAM have superior survival outcome when compared to other stage IV patients, especially when CAM was accompanied by other loco regional recurrences. These data suggest that the CAM patients may benefit from active loco regional treatment. Citation Format: Ji-Jung Jung, Hyeong-Gon Moon, Wonshik Han, Han-Byoel Lee, Hong-Kyu Kim, Jung Whan Chun, Eunhye Kang, Changjin Lim, Jang-il Kim, Hyunsu Yeoh. Contralateral Axillary Lymph Node Metastasis after Ipsilateral Breast Tumor Recurrence: Is it distant metastasis or locoregional progression? [abstract]. In: Proceedings of the 2022 San Antonio Breast Cancer Symposium; 2022 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2023;83(5 Suppl):Abstract nr P5-01-04.

Often breast cancer can be treated by breast-conserving surgery (BCS), after which 10 % locoregional recurrences (LRR) occur within 10 years. After BCS mastectomy is recommended at first LRR, although another BCS could be possible. Changes in clinical parameters and in tumor biology from primary breast cancer to first and multiple LRR are described and correlated with further LRR and overall survival (OS). 380 patients with ≥1 ≤3 LRR (1997-2007) were evaluated retrospectively and followed until 5/2009. Patients' age, tumor size, nodal involvement, distant metastases, histological subtype, hormone receptor (HR) and Her-2/neu status were assessed. LRR therapy options were evaluated. 247 patients had one LRR (94 two and 39 three). Mean OS was 10.1 years. Number of LRR was not correlated with OS. Positive HR status was significantly correlated with longer OS. Patients, who changed from primarily ER negative to positive at first LRR had a significantly longer OS compared to those, who remained or changed to ER negative (p < 0.01). Tumor size and grading correlated inversely with OS (both: p < 0.001). BCS at first LRR correlated with a significantly better OS than mastectomy (p < 0.001). LRR cases with chemotherapy had a shorter OS. Irradiation and/or endocrine therapy after LRR were not correlated with OS. Patients with positive HR status had the best survival data. HR should always be determined. In positive cases, endocrine therapy is recommended. As clinical data are good, BCS at first LRR can be suggested for more patients.

Objective: Recent data in select pre-clinical models suggest that radiation can activate normal stroma to promote tumor metastases and aggressiveness. We hypothesized that if these were occurring clinically, there would be a lower survival after locoregional recurrence (LRR) in patients after post-mastectomy radiation therapy (PMRT) compared to mastectomy (Mx) alone. This study used two independent datasets to compare survival after LRR in women treated with versus without PMRT.Methods: Data from 229 of 1,505 patients who experienced LRR after treatment on sequential non-randomized institutional prospective trials at the MD Anderson Cancer Center (MDA) and 66 of 318 patients enrolled in the British Columbia (BC) PMRT randomized trial who experienced LRR were analyzed. All patients underwent Mx and level I/II axillary dissection. In both data sets analysis was based on treatment received. Patients from MDA received doxorubicin based chemotherapy +/- PMRT, with 45 LRR after PMRT and 184 LRR after Mx alone). Patients treated on the BC trial received CMF chemotherapy +/- PMRT, with LRR in 14/160 after PMRT versus 52/158 after Mx alone. Survival was calculated from time of LRR to death using Kaplan-Meier and log rank statistics.Results:MDA Data: Median follow up of living patients was 192 months. Analyzing data from all patients with LRR regardless of distant metastasis (DM), patients with LRR after PMRT were younger (47 vs. 51 y, p = 0.033) and had shorter time to first LRR (40mo vs. 51 mo, p = 0.018). 5-yr/10-yr OS were 31%/16% without PMRT and 20%/7% after PMRT (p = 0.008). However, PMRT-treated patients had increased risk factors for DM (advanced T and N stage) and more PMRT-treated patients developed DM prior to LRR (58% vs. 36% p = 0.009). Analyzing only patients without DM there was no difference in OS between groups (p = 0.67), and a separate analysis of all patients who developed metastatic disease (N = 385 no PMRT, 233 after PMRT) revealed no difference in 5 or 10-yr OS after DR (15%/4% without PMRT vs. 13%/6% after PMRT, p = 0.5).BC Data: Median follow up of living patients was 235 months. The distributions of age, T stage, N stage, grade, LVI, ER status, excised nodes and nodal ratio were similar between patients with LRR after Mx alone vs. Mx plus PMRT. (all p &amp;gt; 0.05). The mean time to first LRR was 39 mo in patients treated with Mx alone and 57 mo in patients treated with PMRT, p= 0.27). The rate of DM was similar in patients with LRR after Mx with vs without PMRT (93% vs. 96%, p=0.60). Distant relapse free survival after LRR was similar in Mx alone vs. PMRT-treated patients (log rank p=0.75). Overall survival was also similar in the two groups (5-yr/10-yr OS 21%/8% without PMRT vs. 23%/12% with PMRT, log rank p=0.93).Conclusions: Decades of randomized data have demonstrated that PMRT reduces LRR and improves overall survival. In the non-randomized dataset, removing the competing risk of DM which is higher in patients selected for PMRT by studying patients with isolated LRR, we find no difference in survival after LRR in the PMRT setting. Analysis of the randomized PMRT trial dataset confirmed the finding of similar survival among women with LRR irrespective of PMRT use. Citation Information: Cancer Res 2009;69(24 Suppl):Abstract nr 4101.

Distribution of Locoregional Breast Cancer Recurrence in Relation to Postoperative Radiation Fields and Biological Subtypes

BackgroundThe role of post-mastectomy radiotherapy (PMRT) in the treatment of patients with T1–2N1 breast cancer is controversial. This study’s purpose was to evaluate the risk of recurrence of T1–2N1 breast cancer and the efficacy of PMRT in low-, medium- and high-risk groups of patients.MethodsPost-mastectomy patients with T1–2N1 breast cancer were restaged according to the American Joint Committee on Cancer Staging Manual, 8th edition (AJCC 8th ed.) staging system. Recurrence scores were generated using prognostic factors identified for loco-regional recurrence and distant metastasis in patients without PMRT, and three risk groups were identified. Rates of loco-regional recurrence and distant metastasis were calculated with a competing risk model and compared using Gray’s test. Disease-free survival and overall survival were calculated using the Kaplan-Meier method and compared using the log-rank test. The Cox proportional hazards regression model was used for the multivariate analysis.ResultsData from 1986 patients (1521without PMRT; 465 with PMRT) were analyzed. Patients without PMRT were stratified into low-, intermediate- and high-risk groups by age, tumor location, AJCC 8th ed. stage, number of positive nodes and lympho-vascular invasion. The 5-year loco-regional recurrence rate and distant metastasis rates for the three risk groups were significant at 2.5, 5.4 and 16.2% (p < 0.001) respectively, and 4.9, 8.4 and 18.6% (p < 0.001) respectively. In the high-risk group, loco-regional recurrence (p < 0.001), and distant metastasis (p = 0.044) were significantly reduced, and disease free survival (p = 0.004), and overall survival (p = 0.029) were significantly improved after PMRT. In the low- and intermediate-risk groups, PMRT had no significant effect on loco-regional recurrence (p = 0.268), distant metastasis (p = 0.252), disease free survival (p = 0.608) or overall survival (p = 0.986).ConclusionOur results showed no benefits of PMRT in the low-risk group, and thus, omitting PMRT radiotherapy in this population could be considered.

To determine the impact of delayed regional metastases, distant metastases, and second primary tumors on the therapeutic outcomes in squamous cell carcinomas of the larynx and hypopharynx. Chart review and statistical analysis. A retrospective tumor registry analysis was made of patients with squamous cell carcinomas of the larynx and hypopharynx who were treated with curative intent in the Department of Otolaryngology-Head and Neck Surgery and the Radiation Oncology Center of the Washington University School of Medicine (St. Louis, MO) between January 1971 and December 1991 and developed delayed regional metastases (2 y after treatment), distant metastases, and second primary malignancies. In 2550 patients, the mean age (59.8 y), sex (8.5 male patients and 1 female patient), and tumor differentiation did not affect the incidence of delayed distant, regional, or second primary malignancies. The overall incidence of delayed regional metastases was 12.4% (317/2550 patients); distant metastases, 8.5% (217/2550); and second primary tumors, 8.9% (228/2550), with a 5-year disease-specific survival of 41%, 6.4%, and 35%, respectively. Second primary malignancies were not statistically related to the origin of the primary tumor, tumor staging, or delayed regional and distant metastases (P =.98). Delayed regional metastases and distant metastases were related to advanced primary disease (T4 stage), lymph node metastases (node positive [N+]), tumor location (hypopharynx), and locoregional tumor recurrence (P < or =.028). Advanced regional metastases at initial diagnosis (N2 and N3 disease) increased the incidence of delayed and distant metastases threefold (P =.017). These two metastatic parameters were significantly greater in hypopharyngeal tumors than in laryngeal tumors (P =.037). The incidences of delayed regional metastases by anatomical location of the primary tumor were as follows: glottic, 4.4%; supraglottic, 16%; subglottic, 11.5%; aryepiglottic fold, 21.9%; pyriform sinus, 31.1%; and posterior hypopharyngeal wall, 18.5%. The incidences of distant metastases were as follows: glottic, 4%; supraglottic, 3.7%; subglottic, 14%; aryepiglottic fold, 16%; pyriform fossa, 17.2%; and posterior hypopharyngeal wall, 17.6%. Seventeen hypopharyngeal tumors (2%) presented with M1 disease. Delayed regional metastases to the ipsilateral treated neck had a significantly worse survival prognosis than delayed metastases to the contralateral nontreated neck (P =.001). Conclusions are as follows: 1) The incidence of second primary tumors is independent from the primary tumor staging and distant and delayed regional metastases. The highest incidence occurred in patient groups with the highest disease-free survival rates (P =.0378). 2) Highest incidence of delayed and distant metastases occurred in hypopharyngeal tumors and was three times greater than in laryngeal cancers (P =.028). 3) Salvage therapeutic rates were poor for delayed metastases to the ipsilateral treated nodes and distant metastases as compared with contralateral neck metastases and second primary tumors (P =.001). 4) Delayed and distant lymph node metastases were significantly higher in advanced primary disease (T4 stage), locoregional recurrences, and regional disease (N2 and N3) (P =.028) in both the larynx and hypopharynx. 5) The higher incidence of delayed and distant metastatic disease was related to more advanced initial tumor presentation in hypopharyngeal cancer as compared with laryngeal cancer (P =.039). 6) Incidence of distant metastases was greatest between 1.5 and 6 years after initial treatment with a mean incidence being less than or equal to 3.2 years.

Dramatic advances in our understanding of the genetic basis for cancer have led to the development of new technologies and tools for genetic cancer risk assessment. Yet, cancer is a complex disorder, and risk assessment, counseling, and management strategies need to consider several important domains: state of cancer genetics knowledge, state of mind (previous cancer experience within the family), state of technology, and state of the art in terms of management. There are several barriers to the efficient identification and counseling of patients and families at high risk for cancer because of inherited susceptibility mutations. Chief among these concerns is the lack of access to competent counseling and education services that are equipped to handle the complex and rapidly evolving medical, technological, and ethical issues. Cancer risk assessment is developing into a distinct discipline in which established empiric risk models are recast along with rapidly evolving genetic technologies for estimation of individual cancer risk. Cancer genetics consultants are an important resource for primary care physicians, gynecologists, surgeons, and oncologists. However, no formal qualification criteria exist for either physicians or allied health care professionals who subspecialize in this new field. This article covers the unique domains of cancer genetics in health care and surveys models for delivery of cancer genetics services and tools for risk assessment. Coupled with innovative cancer diagnostic and preventive services and research, we have the potential to make great strides in cancer prevention and control. Cancer 1999;86:2483–92. © 1999 American Cancer Society.

Commentary on Graubard et al., [p. 2430][1] In 2010, it is projected that there will be 207,090 diagnoses and 39,840 deaths from breast cancer among women in the United States. Approximately 10,300 will be diagnosed before the age of 40 ([1][2]), the age when the American Cancer Society ([2][3])

Dramatic advances in our understanding of the genetic basis for cancer have led to the development of new technologies and tools for genetic cancer risk assessment. Yet, cancer is a complex disorder, and risk assessment, counseling, and management strategies need to consider several important domains: state of cancer genetics knowledge, state of mind (previous cancer experience within the family), state of technology, and state of the art in terms of management. There are several barriers to the efficient identification and counseling of patients and families at high risk for cancer because of inherited susceptibility mutations. Chief among these concerns is the lack of access to competent counseling and education services that are equipped to handle the complex and rapidly evolving medical, technological, and ethical issues. Cancer risk assessment is developing into a distinct discipline in which established empiric risk models are recast along with rapidly evolving genetic technologies for estimation of individual cancer risk. Cancer genetics consultants are an important resource for primary care physicians, gynecologists, surgeons, and oncologists. However, no formal qualification criteria exist for either physicians or allied health care professionals who subspecialize in this new field. This article covers the unique domains of cancer genetics in health care and surveys models for delivery of cancer genetics services and tools for risk assessment. Coupled with innovative cancer diagnostic and preventive services and research, we have the potential to make great strides in cancer prevention and control.

We aimed to develop a user-centered, web-based, decision support tool for breast cancer risk assessment and personalized risk management. Using a novel model choice algorithm, iPrevent® selects one of two validated breast cancer risk estimation models (IBIS or BOADICEA), based on risk factor data entered by the user. Resulting risk estimates are presented in simple language and graphic formats for easy comprehension. iPrevent® then presents risk-adapted, evidence-based, guideline-endorsed management options. Development was an iterative process with regular feedback from multidisciplinary experts and consumers. To verify iPrevent®, risk factor data for 127 cases derived from the Australian Breast Cancer Family Study were entered into iPrevent®, IBIS (v7.02), and BOADICEA (v3.0). Consistency of the model chosen by iPrevent® (i.e., IBIS or BOADICEA) with the programmed iPrevent® model choice algorithm was assessed. Estimated breast cancer risks from iPrevent® were compared with those attained directly from the chosen risk assessment model (IBIS or BOADICEA). Risk management interventions displayed by iPrevent® were assessed for appropriateness. Risk estimation model choice was 100 % consistent with the programmed iPrevent® logic. Discrepant 10-year and residual lifetime risk estimates of >1 % were found for 1 and 4 cases, respectively, none was clinically significant (maximal variation 1.4 %). Risk management interventions suggested by iPrevent® were 100 % appropriate. iPrevent® successfully integrates the IBIS and BOADICEA risk assessment models into a decision support tool that provides evidence-based, risk-adapted risk management advice. This may help to facilitate precision breast cancer prevention discussions between women and their healthcare providers.Electronic supplementary materialThe online version of this article (doi:10.1007/s10549-016-3726-y) contains supplementary material, which is available to authorized users.