Risk of first recurrence after treatment in a population-based cohort of young women with breast cancer.

PURPOSE: Breast cancer (BC) in women under 45 is rare yet often aggressive. We aim to analyze loco-regional recurrences (LR), distant recurrences (DR), second breast cancers, and mortality in young BC patients. METHODS: We enrolled 776 women with non-metastatic BC ≤45 years diagnosed from 1970 to 2012. Variables included age, family history, tumor stage/grade, and treatment. We used multivariate Cox regression and competing risk models. RESULTS: Among the participants, 37.0% were diagnosed before the age of 40. Most had stage I or II, grade II, ER- and PR-positive, HER2-negative tumors. Over a median follow-up of 8.7 years, 10.1% experienced LR, 13.7% developed DR, and 10.8% died, primarily due to BC. The majority of recurrences occurred within the first five years. Older age (>40) significantly reduced the risk of LR and DR. Advanced disease stage, certain surgical strategies, and positive margins increased DR risk. In the cohort diagnosed between 2001 and 2012, recent diagnosis, triple-negative cancer, and hormonal therapy were associated with reduced LR risk. Breast-conserving surgery appeared to offer protective effects against DR. CONCLUSION: This study highlights that BC in young women carries a significant risk of early recurrence, with age, tumor characteristics, and treatment modalities influencing outcomes. The findings emphasize the need for tailored treatment strategies for young BC patients, focusing on surgical precision and aggressive adjuvant therapy for high-risk cases. This research contributes valuable insights into managing BC in younger patients, aiding in improving long-term outcomes.

Acceptability and Usability of the Family Gene Toolkit for Swiss and Korean Families Harboring BRCA1/BRAC2 Pathogenic Variants: A Web-Based Platform for Cascade Genetic Testing.

The study adapted the Family Gene Toolkit and developed a customized web application for Swiss and Korean families harboring BRCA1 or BRCA2 pathogenic variants to support family communication of genetic testing results and promote cascade genetic testing among at-risk relatives. In the first step, narrative data from 68 women with BRCA1/BRCA2 pathogenic variants and clinician feedback informed a culturally sensitive adaptation of the content consistent with current risk management guidelines. In the second step, the Information Technology team developed the functions and the interface of the web application that will host the intervention. In the third step, a new sample of 18 women from families harboring BRCA1/BRCA2 pathogenic variants tested the acceptability and usability of the intervention using "think-aloud" interviews and a questionnaire. Participants expressed high levels of satisfaction with the intervention. They provided positive feedback for the information regarding active coping, strategies to enhance family communication, interactive elements, and illustrative stories. They reported that the information was useful and the web application was easy to navigate. Findings suggest that the Family Gene Toolkit is well-designed and can increase rates of cascade testing among at-risk relatives. Its efficacy will be tested in a subsequent randomized trial.

Chemotherapy-related agranulocytosis as a predictive factor for germline BRCA1 pathogenic variants in breast cancer patients: a retrospective cohort study.

BACKGROUND: Carriers of germline pathogenic variants of the BRCA1 gene (gBRCA1) tend to have a higher incidence of haematological toxicity upon exposure to chemotherapy. We hypothesised that the occurrence of agranulocytosis during the first cycle of (neo-)adjuvant chemotherapy (C1) in breast cancer (BC) patients could predict gBRCA1 pathogenic variants. PATIENTS AND METHODS: The study population included non-metastatic BC patients selected for genetic counselling at Hôpitaux Universitaires de Genève (Jan. 1998 to Dec. 2017) with available mid-cycle blood counts performed during C1. The BOADICEA and Manchester scoring system risk-prediction models were applied. The primary outcome was the predicted likelihood of harbouring gBRCA1 pathogenic variants among patients presenting agranulocytosis during C1. RESULTS: Three hundred seven BC patients were included: 32 (10.4%) gBRCA1, 27 (8.8%) gBRCA2, and 248 (81.1%) non-heterozygotes. Mean age at diagnosis was 40 years. Compared with non-heterozygotes, gBRCA1 heterozygotes more frequently had grade 3 BC (78.1%; p = 0.014), triple-negative subtype (68.8%; p <0.001), bilateral BC (25%; p = 0.004), and agranulocytosis following the first cycle of (neo-)adjuvant chemotherapy (45.8%; p = 0.002). Agranulocytosis and febrile neutropenia that developed following the first cycle of chemotherapy were independently predictive for gBRCA1 pathogenic variants (odds ratio: 6.1; p = 0.002). The sensitivity, specificity, positive predictive value, and negative predictive value for agranulocytosis predicting gBRCA1 were 45.8% (25.6-67.2%), 82.8% (77.5-87.3%), 22.9% (6.1-37.3%), and 93.4% (88.9-96.4%), respectively. Agranulocytosis substantially improved the positive predictive value of the risk-prediction models used for gBRCA1 evaluation. CONCLUSION: Agranulocytosis following the first cycle of (neo-)adjuvant chemotherapy is an independent predictive factor for gBRCA1 detection in non-metastatic BC patients.

Hematologic toxicities of chemotherapy in breast and ovarian cancer patients carrying BRCA1/BRCA2 germline pathogenic variants. A single center experience and review of the literature.

BRCA1 and BRCA2 play a central role in DNA repair and their germline pathogenic variants (gBRCA) confer a high risk for developing breast and ovarian cancer. Standard chemotherapy regimens for these cancers include DNA-damaging agents. We hypothesized that gBRCA carriers might be at higher risk of developing chemotherapy-related hematologic toxicity and therapy-related myeloid neoplasms (t-MN). We conducted a retrospective study of women newly diagnosed with invasive breast or ovarian cancer who were screened for gBRCA1/gBRCA2 at Geneva University Hospitals. All patients were treated with (neo-)adjuvant chemotherapy. We evaluated acute hematologic toxicities by analyzing the occurrence of febrile neutropenia and severe neutropenia (grade 4) at day 7-14 of the first cycle of chemotherapy and G-CSF use during the entire chemotherapy regimen. Characteristics of t-MN were collected. We reviewed medical records from 447 patients: 58 gBRCA1 and 40 gBRCA2 carriers and 349 non-carriers. gBRCA1 carriers were at higher risk of developing severe neutropenia (32% vs. 14.5%, p = 0.007; OR = 3.3, 95% CI [1.6-7], p = 0.001) and of requiring G-CSF for secondary prophylaxis (58.3% vs. 38.2%, p = 0.011; OR = 2.5, 95% CI [1.4-4.8], p = 0.004). gBRCA2 carriers did not show increased acute hematologic toxicities. t-MN were observed in 2 patients (1 gBRCA1 and one non-carrier). Our results suggested an increased acute hematologic toxicity upon exposure to chemotherapy for breast and ovarian cancer among gBRCA1 but not gBRCA2 carriers. A deeper characterization of t-MN is warranted with the recent development of PARP inhibitors in frontline therapy in gBRCA breast and ovarian cancer.

Side Effects of Cancer Immunotherapies: Dermatologic and Rheumatologic Toxicities.

Dermatologic and rheumatologic toxicities are frequent adverse events during treatment with immune checkpoint inhibitors in oncology. Timely identification of these events and the referral to the oncologist or dermatologist are important in daily practice, such an organ damage involvements can be severe and sometimes life-threatening. The diagnosis and management of cutaneous toxicities, such as maculopapular rash and bullous dermatitis in particular, must be based on the body surface affected by skin lesions. Corticosteroids are basic treatment in moderate to severe cases. Rheumatologic toxicities are rarer and more heterogeneous, and they are often underestimated. They can occur in the absence of autoantibodies, and myositis can be life-threatening.

Relatives from Hereditary Breast and Ovarian Cancer and Lynch Syndrome Families Forgoing Genetic Testing: Findings from the Swiss CASCADE Cohort.

Cascade genetic testing of relatives from families with pathogenic variants associated with hereditary breast and ovarian cancer (HBOC) or Lynch syndrome (LS) has important implications for cancer prevention. We compared the characteristics of relatives from HBOC or LS families who did not have genetic testing (GT (-) group) with those who had genetic testing (GT (+) group), regardless of the outcome. Self-administered surveys collected cross-sectional data between September 2017 and December 2021 from relatives participating in the CASCADE cohort. We used multivariable logistic regression with LASSO variable selection. Among n = 115 relatives who completed the baseline survey, 38% (n = 44) were in the GT (-) group. Being male (OR: 2.79, 95% CI: 1.10-7.10) and without a previous cancer diagnosis (OR: 4.47, 95% CI: 1.03-19.42) increased the odds of being untested by almost three times. Individuals from families with fewer tested relatives had 29% higher odds of being untested (OR: 0.71, 95% CI: 0.55-0.92). Reasons for forgoing cascade testing were: lack of provider recommendation, lack of time and interest in testing, being afraid of discrimination, and high out-of-pocket costs. Multilevel interventions designed to increase awareness about clinical implications of HBOC and LS in males, referrals from non-specialists, and support for testing multiple family members could improve the uptake of cascade testing.

Cohort profile: population-based cohorts of patients with colorectal cancer and of their relatives in Geneva, Switzerland.

PURPOSE: Colorectal cancer (CRC) is the third leading cause of cancer death worldwide. Variability between patients in prognosis and treatment response is partially explained by traditional clinicopathological factors. We established a large population-based cohort of patients with CRC and their first-degree and second-degree relatives registered in the Canton of Geneva, to evaluate the role of family history and tumour biomarkers on patient outcomes. PARTICIPANTS: The cohort includes all patients with CRC diagnosed between 1985 and 2013. Detailed information on patient and tumour characteristics, treatment and outcomes were extracted from the Geneva Cancer Registry database, completed by medical records review and linkage with administrative and oncogenetics databases. Next-generation tissue microarrays were constructed from tissue samples of the primary tumour. A prospective follow-up of the cohort is realised annually to collect data on outcomes. First-degree and second-degree relatives of patients are identified through linkage with the Cantonal Population Office database and information about cancer among relatives is retrieved from the Geneva Cancer Registry database. The cohort of relatives is updated annually. FINDINGS TO DATE: The cohort includes 5499 patients (4244 patients with colon cancer and 1255 patients with rectal cancer). The great majority of patients were diagnosed because of occurrence of symptoms and almost half of the cases were diagnosed with an advanced disease. At the end of 2019, 337 local recurrences, 1143 distant recurrences and 4035 deaths were reported. At the same date, the cohort of first-degree relatives included 344 fathers, 538 mothers, 3485 children and 375 siblings. Among them, we identified 28 fathers, 31 mothers, 18 siblings and 53 children who had a diagnosis of CRC. FUTURE PLANS: The cohort will be used for long-term studies of CRC epidemiology with focus on clinicopathological factors and molecular markers. These data will be correlated with the most up-to-date follow-up data.

The Communication Chain of Genetic Risk: Analyses of Narrative Data Exploring Proband-Provider and Proband-Family Communication in Hereditary Breast and Ovarian Cancer.

Low uptake of genetic services among members of families with hereditary breast and ovarian cancer (HBOC) suggests limitations of proband-mediated communication of genetic risk. This study explored how genetic information proceeds from healthcare providers to probands and from probands to relatives, from the probands' perspectives. Using a grounded-theory approach, we analyzed narrative data collected with individual interviews and focus groups from a sample of 48 women identified as carriers of HBOC-associated pathogenic variants from three linguistic regions of Switzerland. The findings describe the "communication chain", confirming the difficulties of proband-mediated communication. Provider-proband communication is impacted by a three-level complexity in the way information about family communication is approached by providers, received by probands, and followed-up by the healthcare system. Probands' decisions regarding disclosure of genetic risk are governed by dynamic and often contradictory logics of action, interconnected with individual and family characteristics, eventually compelling probands to engage in an arbitrating process. The findings highlight the relevance of probands' involvement in the communication of genetic risk to relatives, suggesting the need to support them in navigating the complexity of family communication rather than replacing them in this process. Concrete actions at the clinical and health system levels are needed to improve proband-mediated communication.

Intention to Inform Relatives, Rates of Cascade Testing, and Preference for Patient-Mediated Communication in Families Concerned with Hereditary Breast and Ovarian Cancer and Lynch Syndrome: The Swiss CASCADE Cohort.

Cascade screening for Tier 1 cancer genetic conditions is a significant public health intervention because it identifies untested relatives of individuals known to carry pathogenic variants associated with hereditary breast and ovarian cancer (HBOC) and Lynch syndrome (LS). The Swiss CASCADE is a family-based, open-ended cohort, including carriers of HBOC- and LS-associated pathogenic variants and their relatives. This paper describes rates of cascade screening in relatives from HBOC- and LS- harboring families, examines carriers' preferences for communication of testing results, and describes theory-based predictors of intention to invite relatives to a cascade screening program. Information has been provided by 304 index cases and 115 relatives recruited from September 2017 to December 2021. On average, 10 relatives per index case were potentially eligible for cascade screening. Approximately 65% of respondents wanted to invite relatives to the cohort, and approximately 50% indicated a preference for patient-mediated communication of testing results, possibly with the assistance of digital technology. Intention to invite relatives was higher for first- compared to second- and third-degree relatives, but was not different between syndromes or based on relatives' gender. The family environment and carrying pathogenic variants predicts intention to invite relatives. Information helps optimize delivery of tailored genetic services.

Swiss cost-effectiveness analysis of universal screening for Lynch syndrome of patients with colorectal cancer followed by cascade genetic testing of relatives.

BACKGROUND: We estimated the cost-effectiveness of universal DNA screening for Lynch syndrome (LS) among newly diagnosed patients with colorectal cancer (CRC) followed by cascade screening of relatives from the Swiss healthcare system perspective. METHODS: We integrated decision trees with Markov models to calculate incremental cost per quality-adjusted life-year saved by screening all patients with CRC (alternative strategy) compared with CRC tumour-based testing followed by DNA sequencing (current strategy). RESULTS: The alternative strategy has an incremental cost-effectiveness ratio of CHF65 058 compared with the current strategy, which is cost-effective according to Swiss standards. Based on annual incidence of CRC in Switzerland, universal DNA screening correctly identifies all 123 patients with CRC with LS, prevents 17 LS deaths and avoids 19 CRC cases, while the current strategy leads to 32 false negative results and 253 LS cases lost to follow-up. One way and probabilistic sensitivity analyses showed that universal DNA testing is cost-effective in around 80% of scenarios, and that the cost of DNA testing and the number of invited relatives per LS case determine the cost-effectiveness ratio. CONCLUSION: Results can inform policymakers, healthcare providers and insurance companies about the costs and benefits associated with universal screening for LS and cascade genetic testing of relatives.

Genetic Literacy and Communication of Genetic Information in Families Concerned with Hereditary Breast and Ovarian Cancer: A Cross-Study Comparison in Two Countries and within a Timeframe of More Than 10 Years.

Examining genetic literacy in families concerned with hereditary breast and ovarian cancer (HBOC) helps understand how genetic information is passed on from individuals who had genetic counseling to their at-risk relatives. This cross-study comparison explored genetic literacy both at the individual and the family level using data collected from three sequential studies conducted in the U.S. and Switzerland over ≥10 years. Participants were primarily females, at-risk or confirmed carriers of HBOC-associated pathogenic variants, who had genetic counselling, and ≥1 of their relatives who did not. Fifteen items assessed genetic literacy. Among 1933 individuals from 518 families, 38.5% had genetic counselling and 61.5% did not. Although genetic literacy was higher among participants who had counselling, some risk factors were poorly understood. At the individual level, genetic literacy was associated with having counselling, ≤5 years ago, higher education, and family history of cancer. At the family level, genetic literacy was associated with having counselling, higher education, and a cancer diagnosis. The findings suggest that specific genetic information should be emphasized during consultations, and that at-risk relatives feel less informed about inherited cancer risk, even if information is shared within families. There is a need to increase access to genetic information among at-risk individuals.

Update Swiss guideline for counselling and testing for predisposition to breast, ovarian, pancreatic and prostate cancer.

This paper presents the Swiss guideline for genetic counselling and testing of individuals with an increased probability for carrying mutations in high risk cancer predisposition genes, particularly BRCA1 and BRCA2. It aims to help providers of genetic counselling to identify valuable candidates for testing and serves as a basis for reimbursement claims to Swiss insurance companies.

Ovarian cancer with high-level focal ERBB2 amplification responds to trastuzumab and pertuzumab.

Epithelial ovarian cancer (EOC) is usually diagnosed at an advanced stage and significantly contributes to cancer mortality in women. Despite multimodal treatment associating chemotherapy and surgery, most patients ultimately progress and require palliative systemic therapy. In EOC, the efficacy of anti-HER2 agents is minimal even after selecting patients for HER2 expression. ERBB2 gene amplification is observed in 3-10% of patients, depending on the specific method of detection and cutoffs. We report the case of a young woman with a FIGO stage IV high-grade serous ovarian cancer with an amplification of ERBB2. She was treated with the association of trastuzumab - pertuzumab after two lines of standard treatment and presented an excellent long-lasting partial response after 36 months of treatment. The association of trastuzumab and pertuzumab, without chemotherapy, has not been previously tested in this context and could be more efficacious than monotherapy with either agent. In addition, the significant benefit observed in this case could be attributed to the presence of a high-level focal amplification that is relatively rare and probably more specific than an increase in HER2 expression. In conclusion, prospective trials of the trastuzumab and pertuzumab combination should be considered in an appropriately selected EOC patient population.

Using a Tailored Digital Health Intervention for Family Communication and Cascade Genetic Testing in Swiss and Korean Families With Hereditary Breast and Ovarian Cancer: Protocol for the DIALOGUE Study.

BACKGROUND: In hereditary breast and ovarian cancer (HBOC), family communication of genetic test results is essential for cascade genetic screening, that is, identifying and testing blood relatives of known mutation carriers to determine whether they also carry the pathogenic variant, and to propose preventive and clinical management options. However, up to 50% of blood relatives are unaware of relevant genetic information, suggesting that potential benefits of genetic testing are not communicated effectively within family networks. Technology can facilitate communication and genetic education within HBOC families. OBJECTIVE: The aims of this study are to develop the K-CASCADE (Korean-Cancer Predisposition Cascade Genetic Testing) cohort in Korea by expanding an infrastructure developed by the CASCADE (Cancer Predisposition Cascade Genetic Testing) Consortium in Switzerland; develop a digital health intervention to support the communication of cancer predisposition for Swiss and Korean HBOC families, based on linguistic and cultural adaptation of the Family Gene Toolkit; evaluate its efficacy on primary (family communication of genetic results and cascade testing) and secondary (psychological distress, genetic literacy, active coping, and decision making) outcomes; and explore its translatability using the reach, effectiveness, adoption, implementation, and maintenance framework. METHODS: The digital health intervention will be available in French, German, Italian, Korean, and English and can be accessed via the web, mobile phone, or tablet (ie, device-agnostic). K-CASCADE cohort of Korean HBOC mutation carriers and relatives will be based on the CASCADE infrastructure. Narrative data collected through individual interviews or mini focus groups from 20 to 24 HBOC family members per linguistic region and 6-10 health care providers involved in genetic services will identify the local cultures and context, and inform the content of the tailored messages. The efficacy of the digital health intervention against a comparison website will be assessed in a randomized trial with 104 HBOC mutation carriers (52 in each study arm). The translatability of the digital health intervention will be assessed using survey data collected from HBOC families and health care providers. RESULTS: Funding was received in October 2019. It is projected that data collection will be completed by January 2023 and results will be published in fall 2023. CONCLUSIONS: This study addresses the continuum of translational research, from developing an international research infrastructure and adapting an existing digital health intervention to testing its efficacy in a randomized controlled trial and exploring its translatability using an established framework. Adapting existing interventions, rather than developing new ones, takes advantage of previous valid experiences without duplicating efforts. Culturally sensitive web-based interventions that enhance family communication and understanding of genetic cancer risk are timely. This collaboration creates a research infrastructure between Switzerland and Korea that can be scaled up to cover other hereditary cancer syndromes. TRIAL REGISTRATION: ClinicalTrials.gov NCT04214210; https://clinicaltrials.gov/ct2/show/NCT04214210 and CRiS KCT0005643; https://cris.nih.go.kr/cris/. INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID): PRR1-10.2196/26264.

Citizens' views on sharing their health data: the role of competence, reliability and pursuing the common good.

BACKGROUND: In this article, we address questions regarding how people consider what they do or do not consent to and the reasons why. This article presents the findings of a citizen forum study conducted by the University of Geneva in partnership with the Geneva University Hospitals to explore the opinions and concerns of members of the public regarding predictive oncology, genetic sequencing, and cancer. METHODS: This paper presents the results of a citizen forum that included 73 participants. A research tool titled "the mechanics of consent" was designed for this study. This tool is a table encouraging participants to reflect on social and research actors, types of data, and desired levels of control while sharing different types of data with different actors. Participants' discussion that led to the completion of each table were audio-recorded, transcribed, and analyzed using thematic analysis. RESULTS: The results are a compilation of responses from the mechanics of consent tool divided into two sections; the first presents quantitative results of collective responses regarding attitudes to consent to donate their data. The second section present qualitative findings emerged from the discussion amongst participants. DISCUSSION: Choice and control of personal data is crucial for the public to be able to decide who and how to trust. Key information to be disclosed to potential research participants shall include information about potential risks and benefits; who will be accessing and using their data; as well as assurances that their choice will be respected. Furthermore, researchers ought to make sure they are trustworthy, by acting in a competent, reliable, and honest manner. Governance systems ought to be better equipped to address ethical issues raise by the growing presence of non-traditional research actors, consent of exchanges of data via digital devices and online activity such as social media and fairness of data trading. Finally, informed consent is one of the various elements that contribute to conducting ethical research. More needs to be done to strengthen governance and ensure adequate protection of research participants, particularly to address issues related to predictive health analytics.

Publisher Correction: Clinical outcome of breast cancer in carriers of BRCA1 and BRCA2 mutations according to molecular subtypes.

An amendment to this paper has been published and can be accessed via a link at the top of the paper.

Concordance of tumour characteristics and survival clustering among pairs of first-degree relatives with breast cancer.

BACKGROUND: Family history is a known risk factor for breast cancer, but its prognostic value and the prognostic value of tumour characteristics in relation to family history has not been clearly established. In addition, studies of intra-familial tumour characteristics and prognosis in population-based settings are very rare. Two previous studies have suggested that breast cancer prognosis clusters within families. However, both studies lack information on HER2 expression status, which is a strong prognostic factor and could contribute to the observed results. METHODS: We conducted a population-based study on 145 mother-daughter and sister-sister affected pairs using data extracted from the Geneva Cancer Registry. Histopathological characteristics were determined in archived tumour blocks by immunochemistry techniques. Breast cancer survival among family members was studied according to patient and tumour characteristics. RESULTS: No significant intra-familial agreement of pathological characteristic features was observed. We found that relatives of breast cancer patients experienced a much higher risk of breast cancer death compared to the general population. However, we did not find significant concordance in good and poor breast cancer-specific survival between pairs. The small number of family pairs and deaths from breast cancer may partly explain our results. CONCLUSIONS: Large-scale studies with accurate data on strong prognosticators are still needed to confirm the possibility of familial inheritance of breast cancer prognosis.  .

Physicians communicating with women at genetic risk of breast and ovarian cancer: Are we in the middle of the ford between contradictory messages and unshared decision making?

BRCA1/2 genetic testing offers tremendous opportunities for prevention, diagnosis and treatment of breast and ovarian cancer. Women acquire valuable information that can help them to make informed decisions about their health. However, knowing one's susceptibility to developing cancer may be burdensome for several women, as this risk needs to be managed over time through a continuous dialogue with multiple healthcare professionals. We explored how communication between physicians and unaffected women carrying BRCA1/2 germline pathogenic variants was experienced by women in relation to their genetic risk. Data came from qualitative interviews conducted in Switzerland with 32 unaffected women carrying BRCA1/2 pathogenic variants and aware of their genetic status for at least 3 years. We identified three different types of message as conveyed by physicians to women: (1) a normative message, (2) an over-empowering message, and (3) a minimizing message. On one hand, we found that women are exposed to contradictory messages, often simultaneously, in their interactions with healthcare professionals during their post-genetic testing journey. On the other hand, women's reports highlighted the absence of shared decision-making in such interactions. The combination of these two findings resulted in a strong sense of disorientation, frustration, and powerlessness among participants. Healthcare professionals interacting with high cancer risk women are urged to align in favor of a both concerted and shared decision-making approach when discussing options for managing genetic risk.

Machine learning-based lifetime breast cancer risk reclassification compared with the BOADICEA model: impact on screening recommendations.

BACKGROUND: The clinical utility of machine-learning (ML) algorithms for breast cancer risk prediction and screening practices is unknown. We compared classification of lifetime breast cancer risk based on ML and the BOADICEA model. We explored the differences in risk classification and their clinical impact on screening practices. METHODS: We used three different ML algorithms and the BOADICEA model to estimate lifetime breast cancer risk in a sample of 112,587 individuals from 2481 families from the Oncogenetic Unit, Geneva University Hospitals. Performance of algorithms was evaluated using the area under the receiver operating characteristic (AU-ROC) curve. Risk reclassification was compared for 36,146 breast cancer-free women of ages 20-80. The impact on recommendations for mammography surveillance was based on the Swiss Surveillance Protocol. RESULTS: The predictive accuracy of ML-based algorithms (0.843 ≤ AU-ROC ≤ 0.889) was superior to BOADICEA (AU-ROC = 0.639) and reclassified 35.3% of women in different risk categories. The largest reclassification (20.8%) was observed in women characterised as 'near population' risk by BOADICEA. Reclassification had the largest impact on screening practices of women younger than 50. CONCLUSION: ML-based reclassification of lifetime breast cancer risk occurred in approximately one in three women. Reclassification is important for younger women because it impacts clinical decision- making for the initiation of screening.