Cancer Risk Prediction Models Research Articles

A machine learning prediction model for cancer risk in patients with type 2 diabetes based on clinical tests.

The incidence of type 2 diabetes is rapidly increasing worldwide. Studies have shown that it is also associated with cancer-related morbidities. Early detection of cancer in patients with type 2 diabetes is crucial. This study aimed to construct a model to predict cancer risk in patients with type 2 diabetes. This study collected clinical data from a total of 5198 patients. A cancer risk prediction model was established by analyzing 261 items from routine laboratory tests. We screened 107 risk factors from 261 clinical tests based on the importance of the characteristic variables, significance of differences between groups (P< 0.05), and minimum description length algorithm. Compared with 16 machine learning classifiers, five classifiers based on the decision tree algorithm (CatBoost, light gradient boosting, random forest, XGBoost, and gradient boosting) had an area under the receiver operating characteristic curve (AUC) of > 0.80. The AUC for CatBoost was 0.852 (sensitivity: 79.6%; specificity: 83.2%). The constructed model can predict the risk of cancer in patients with type 2 diabetes based on tumor biomarkers and routine tests using machine learning algorithms. This is helpful for early cancer risk screening and prevention to improve patient outcomes.

Abstract PS10-09: Development of an absolute risk prediction model for premenopausal breast cancer in an international consortium

Abstract Risk prediction models that have been developed for overall breast cancer risk are based on a limited number of premenopausal cases in individual cohorts. As some risk factors differ in their associations with pre- versus postmenopausal breast cancer, a distinct risk prediction model is needed for premenopausal breast cancer. We developed a risk prediction model for premenopausal breast cancer using 779,601 participants and 9,665 incident cases from 18 prospective studies within the Premenopausal Breast Cancer Collaborative Group (PBCCG), across North America (N=9), Europe (N=6), Australia (N=1), and Asia (N=2). Data were split, within each cohort, into training (2/3) and testing (1/3) datasets. Individual risk was assessed in five-year intervals, using variables reported at the start of the time interval. Cox proportional hazards regression was used to model risk factors in a backwards-selection method, stratified by cohort: age at menarche, age at first birth, parity, breastfeeding (months), height (cm), BMI (kg/m2), BMI at age 18, recent weight change (kg), alcohol consumption (drinks/week), first-degree family history of breast cancer, and personal history of benign breast disease. To enable the use of information from all cohorts despite differences in missing variables by design, cohorts were grouped by available variables and risk models were fit for each group. In cohorts with incomplete data, model coefficients were adjusted based on the correlation of covariates with the missing variables in the complete case dataset. Coefficients were meta-analyzed with inverse variance weighting to obtain final coefficients. Discrimination was evaluated in the testing dataset by calculation of the c-index. Work is ongoing to calibrate the model based on five-year absolute risk using GLOBOCAN continent- and age-specific incidence rates to represent baseline risk. The final model included age at menarche, parity, height, BMI, BMI at age 18, first-degree family history of breast cancer, and history of benign breast disease. Young adulthood BMI and BMI (at start of the 5-year risk interval) were associated with a decreased risk (Hazard Ratio (HR) (95% confidence interval (CI)) per 5 kg/m2 = 0.87 (0.81-0.93) and 0.90 (0.86-0.95), respectively) as was parity (HR (95% CI) = 0.92 (0.90-0.94)). Height was associated with increased risk (HR (95% CI) per 10 cm = 1.14 (1.07-1.21)), while history of benign breast disease and family history were associated with larger increases in risk (HR (95% CI) = 1.64 (1.30-2.06) and 1.76 (1.63-1.90), respectively). Model discrimination was higher than that reported for women under 50 years in existing breast cancer risk prediction models considering clinical factors (AUC (95% CI) = 0.61 (0.59-0.62)). Calibration of absolute 5-year risk is ongoing. Several factors driving risk prediction of postmenopausal breast cancer have similar influence on risk of premenopausal breast cancer, while family history has a stronger influence on premenopausal breast cancer risk. Our model demonstrates acceptable discrimination and will enable individual 5-year absolute risk prediction for premenopausal breast cancer. Citation Format: Kristen Brantley, Michael Jones, Minouk Schoemaker, Hazel Nichols, Anthony Swerdlow, Robert MacInnis, Roger Milne, Tess Clenenden, Yu Chen, Xiao-Ou Shu, Wei Zheng, Woon-Puay Koh, Jian-Min Yuan, Cari Kitahara, Martha Linet, Dale Sandler, Bernard A Rosner, Peter Kraft, A. Heather Eliassen. Development of an absolute risk prediction model for premenopausal breast cancer in an international consortium [abstract]. In: Proceedings of the 2023 San Antonio Breast Cancer Symposium; 2023 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2024;84(9 Suppl):Abstract nr PS10-09.

Highly specific vaginal microbiome signature for gynecological cancers.

To investigate the vaginal microbiota signature of patients with gynecologic cancer and evaluate its diagnostic biomarker potential. We incorporated vaginal 16S rRNA-seq data from 529 women and utilized VSEARCH to analyze the raw data. α-Diversity was evaluated utilizing the Chao1, Shannon, and Simpson indices, and β-diversity was evaluated through principal component analysis using Bray-Curtis distances. Linear discriminant analysis effect size (LEfSe) was utilized to determine species differences between groups. A bacterial co-abundance network was constructed utilizing Spearman correlation analysis. A random forest model of gynecologic tumor risk based on genus was constructed and validated to test its diagnostic efficacy. In gynecologic cancer patients, vaginal α-diversity was significantly greater than in controls, and vaginal β-diversity was significantly separated from that of controls; there was no correlation between these characteristics and menopause status among the subject women. Women diagnosed with gynecological cancer exhibited a reduction in the abundance of vaginal Firmicutes and Lactobacillus, while an increase was observed in the proportions of Bacteroidetes, Proteobacteria, Prevotella, Streptococcus, and Anaerococcus. A random forest model constructed based on 56 genus achieved high accuracy (area under the curve = 84.96%) in gynecological cancer risk prediction. Furthermore, there were discrepancies observed in the community complexity of co-abundance networks between gynecologic cancer patients and the control group. Our study provides evidence that women with gynecologic cancer have a unique vaginal flora structure and microorganisms may be involved in the gynecologic carcinogenesis process. A gynecological cancer risk prediction model based on characteristic genera has good diagnostic value.

Risk-based lung cancer screening performance in a universal healthcare setting.

Globally, lung cancer is the leading cause of cancer death. Previous trials demonstrated that low-dose computed tomography lung cancer screening of high-risk individuals can reduce lung cancer mortality by 20% or more. Lung cancer screening has been approved by major guidelines in the United States, and over 4,000 sites offer screening. Adoption of lung screening outside the United States has, until recently, been slow. Between June 2017 and May 2019, the Ontario Lung Cancer Screening Pilot successfully recruited 7,768 individuals at high risk identified by using the PLCOm2012noRace lung cancer risk prediction model. In total, 4,451 participants were successfully screened, retained and provided with high-quality follow-up, including appropriate treatment. In the Ontario Lung Cancer Screening Pilot, the lung cancer detection rate and the proportion of early-stage cancers were 2.4% and 79.2%, respectively; serious harms were infrequent; and sensitivity to detect lung cancers was 95.3% or more. With abnormal scans defined as ones leading to diagnostic investigation, specificity was 95.5% (positive predictive value, 35.1%), and adherence to annual recall and early surveillance scans and clinical investigations were high (>85%). The Ontario Lung Cancer Screening Pilot provides insights into how a risk-based organized lung screening program can be implemented in a large, diverse, populous geographic area within a universal healthcare system.

Risk prediction model for gastric cancer within 5 years in healthy Korean adults.

Although endoscopy is commonly used for gastric cancer screening in South Korea, predictive models that integrate endoscopy results are scarce. We aimed to develop a 5-year gastric cancer risk prediction model using endoscopy results as a predictor. We developed a predictive model using the cohort data of the Kangbuk Samsung Health Study from 2011 to 2019. Among the 260,407 participants aged ≥20years who did not have any previous history of cancer, 435 cases of gastric cancer were observed. A Cox proportional hazard regression model was used to evaluate the predictors and calculate the 5-year risk of gastric cancer. Harrell's C-statistics and Nam-D'Agostino χ2 test were used to measure the quality of discrimination and calibration ability, respectively. We included age, sex, smoking status, alcohol consumption, family history of cancer, and previous results for endoscopy in the risk prediction model. This model showed sufficient discrimination ability [development cohort: C-Statistics: 0.800, 95% confidence interval (CI) 0.770-0.829; validation cohort: C-Statistics: 0.799, 95% CI 0.743-0.856]. It also performed well with effective calibration (development cohort: χ2=13.65, P=0.135; validation cohort: χ2=15.57, P=0.056). Our prediction model, including young adults, showed good discrimination and calibration. Furthermore, this model considered a fixed time interval of 5years to predict the risk of developing gastric cancer, considering endoscopic results. Thus, it could be clinically useful, especially for adults with endoscopic results.

Abstract 7411: CAMStool: Cancer-associated somatic mutation selection tool for cancer risk prediction in whole-genome data

Abstract Predicting cancer risk plays a useful role in presenting guidelines to prevent and manage cancers of individuals. Genomic variants have recently been widely used as a genetic factor, and the utilization of the genetic factor consisting of somatic mutations has shown remarkable predictive performance in many cancer prediction studies. Typically, the accuracy of such prediction studies is significantly affected by the composition of somatic mutations used as the genetic factor; however, in practice, choosing the optimal set of mutations with high prediction accuracy is a challenging task in cancer prediction research. From this perspective, we present a new analysis tool for selecting an optimal set of somatic mutations associated with cancers of interest, named CAMStool. The CAMStool is based on the statistical values, which are the effect size and its p-value of each somatic mutation for the cancers of interest. It heuristically searches for the only somatic mutations that directly increase the prediction accuracy of cancer, through statistical and machine-learning models. Then, our tool presents the finally selected somatic mutations as the final optimal mutation set, and additionally presents an optimal cancer risk prediction model through various machine/deep learning modeling using the optimal set as a genetic factor. For the evaluation, the CAMStool was applied to our real whole-genome dataset which are sequenced from normal-tumor tissues of 459 individuals with thyroid cancer. The CAMStool selected 36 and 55 optimal somatic mutations associated with the benign thyroid and non-invasive follicular thyroid cancer, and presented risk prediction models with sensitivity of 88.04% and 90.32%, respectively. Although the CAMStools has some limitations, it is expected to contribute to various cancer prediction studies that require selecting optimal mutation sets in the future. The CAMStools is written in R and will be released soon on our github (https://github.com/JaeYoonKim72). Citation Format: Jae-Yoon Kim, Seung-Jin Park, Seong-Hwan Park, Seon-Kyu Kim. CAMStool: Cancer-associated somatic mutation selection tool for cancer risk prediction in whole-genome data [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 7411.

The Prognostic Quality of Risk Prediction Models to Assess the Individual Breast Cancer Risk in Women: An Overview of Reviews

Purpose. Breast cancer is the most common cancer among women globally, with an incidence of approximately two million cases in 2018. Organised age-based breast cancer screening programs were established worldwide to detect breast cancer earlier and to reduce mortality. Currently, there is substantial anticipation regarding risk-adjusted screening programs, considering various risk factors in addition to age. The present study investigated the discriminatory accuracy of breast cancer risk prediction models and whether they suit risk-based screening programs. Methods. Following the PICO scheme, we conducted an overview of reviews and systematically searched four databases. All methodological steps, including the literature selection, data extraction and synthesis, and the quality appraisal were conducted following the 4-eyes principle. For the quality assessment, the AMSTAR 2 tool was used. Results. We included eight systematic reviews out of 833 hits based on the prespecified inclusion criteria. The eight systematic reviews comprised ninety-nine primary studies that were also considered for the data analysis. Three systematic reviews were assessed as having a high risk of bias, while the others were rated with a moderate or low risk of bias. Most identified breast cancer risk prediction models showed a low prognostic quality. Adding breast density and genetic information as risk factors only moderately improved the models’ discriminatory accuracy. Conclusion. All breast cancer risk prediction models published to date show a limited ability to predict the individual breast cancer risk in women. Hence, it is too early to implement them in national breast cancer screening programs. Relevant randomised controlled trials about the benefit-harm ratio of risk-adjusted breast cancer screening programs compared to conventional age-based programs need to be awaited.

Abstract B052: Ovarian cancer risk prediction: a systematic assessment of the state of the art and development of a new model

Abstract B052: Ovarian cancer risk prediction: a systematic assessment of the state of the art and development of a new model

Association between Family History of Breast Cancer and Breast Density in Saudi Premenopausal Women Participating in Mammography Screening.

Mammographic density and family history of breast cancer (FHBC) are well-established independent factors affecting breast cancer risk; however, the association between these two risk factors in premenopausal-screened women remains unclear. The aim of this study is to investigate the relationship between mammographic density and FHBC among Saudi premenopausal-screened women. A total of 446 eligible participants were included in the study. Mammographic density was assessed qualitatively using the Breast Imaging Reporting and Data System (BIRADS 4th edition). Logistic regression models were built to investigate the relationship between mammographic density and FHBC. Women with a family history of breast cancer demonstrated an 87% greater chance of having dense tissue than women without a family history of breast cancer (95% CI: 1.14-3.08; p = 0.01). Having a positive family history for breast cancer in mothers was significantly associated with dense tissue (adjusted odds ratio (OR): 5.6; 95% CI: 1.3-24.1; p = 0.02). Dense breast tissue in Saudi premenopausal women undergoing screening may be linked to FHBC. If this conclusion is replicated in larger studies, then breast cancer risk prediction models must carefully consider these breast cancer risk factors.

Addition of polygenic risk score to a risk calculator for prediction of breast cancer in US Black women

BackgroundPrevious work in European ancestry populations has shown that adding a polygenic risk score (PRS) to breast cancer risk prediction models based on epidemiologic factors results in better discriminatory performance as measured by the AUC (area under the curve). Following publication of the first PRS to perform well in women of African ancestry (AA-PRS), we conducted an external validation of the AA-PRS and then evaluated the addition of the AA-PRS to a risk calculator for incident breast cancer in Black women based on epidemiologic factors (BWHS model).MethodsData from the Black Women’s Health Study, an ongoing prospective cohort study of 59,000 US Black women followed by biennial questionnaire since 1995, were used to calculate AUCs and 95% confidence intervals (CIs) for discriminatory accuracy of the BWHS model, the AA-PRS alone, and a new model that combined them. Analyses were based on data from 922 women with invasive breast cancer and 1844 age-matched controls.ResultsAUCs were 0.577 (95% CI 0.556–0.598) for the BWHS model and 0.584 (95% CI 0.563–0.605) for the AA-PRS. For a model that combined estimates from the questionnaire-based BWHS model with the PRS, the AUC increased to 0.623 (95% CI 0.603–0.644).ConclusionsThis combined model represents a step forward for personalized breast cancer preventive care for US Black women, as its performance metrics are similar to those from models in other populations. Use of this new model may mitigate exacerbation of breast cancer disparities if and when it becomes feasible to include a PRS in routine health care decision-making.

Predicting risk of endometrial cancer in asymptomatic women (PRECISION): Model development and external validation.

Develop an endometrial cancer risk prediction model and externally validate it for UK primary care use. Cohort study. The UK Biobank was used for model development and a linked primary (Clinical Practice Research Datalink, CPRD) and secondary care (HES), mortality (ONS) and cancer register (NRCAS) dataset was used for external validation. Women aged 45-60 years with no history of endometrial cancer or hysterectomy. Model development was performed using a flexible parametric survival model and stepwise backward selection aiming to minimise the Akaike information criterion. Model performance on external validation was assessed through flexible calibration plots, calculation of the expected to observed ratio and C-statistic and decision curve analysis. Endometrial cancer diagnosis within 1-10 years of the index date. Model development included 222 031 women (902 incident endometrial cancer cases) and external validation 3 094 371 women (8585 endometrial cancer cases). The final model (with equation provided) incorporated age, body mass index, waist circumference, age at menarche, menopause and last birth, hormone replacement, tamoxifen and oral contraceptive pill use, type 2 diabetes, smoking and family history of bowel cancer. It was well calibrated on external validation (calibration slope 1.14, 95% confidence interval [CI] 1.11-1.17, E/O 1.03, 95% CI 1.01-1.05), with moderate/good discrimination (C-statistic 0.70, 95% CI 0.69-0.70) and had improved net benefit compared with previously developed models. The Predicting risk of endometrial cancer in asymptomatic women model (PRECISION), using easily measurable anthropometric, reproductive, personal and family history, accurately quantifies a woman's 10-year risk of endometrial cancer. Its use could determine eligibility for primary endometrial cancer prevention trials and for targeted resource allocation in UK general practices.

Abstract B145: Validation of a colorectal cancer risk prediction model in US Black women

Abstract Background: The National Cancer Institute’s Colorectal Cancer Risk Assessment Tool (NCI-CCRAT) estimates an individual's five-year risk of developing colorectal cancer. It was developed by estimating relative risks from a population of US White women and men and incorporating race, gender, and age-specific incidence data from population-based cancer registries. However, the model performance has not been assessed for Black Americans, who have the highest colorectal cancer incidence and mortality rates of any US racial/ethnic group. Methods: The NCI-CCRAT estimates the absolute risk of colorectal cancer based on an individual’s age, sex, family history of colorectal cancer, sigmoidoscopy/colonoscopy, polyps, smoking, physical activity, regular aspirin/nonsteroidal anti-inflammatory drug (NSAID) use, vegetable intake, body mass index, and hormone replacement therapy use. Validation of the NCI-CCRAT was performed using prospective data from 57,558 participants ages 40 to 84 years in the Black Women's Health Study (BWHS). We predicted 5-year absolute risk using data from 2001 through 2015 in three 5-year prediction periods: 2001-2005, 2006-2010, and 2011-2015. To assess model performance, we computed calibration by comparing the expected number of colorectal cancer cases predicted by the model to the observed number of cases (E/O) in the BWHS overall and in subgroups of the population. We assessed discriminatory accuracy by calculating the area under the curve (AUC). Results: During follow-up from 2001 through 2015 (634,512 person-years), 433 BWHS participants developed invasive colorectal cancer. Based on the NCI-CCRAT, 543 cases were expected. The NCI-CCRAT was well-calibrated in the first 5-year prediction period (2001-2005 E/O=1.03, 95% CI: 0.88–1.21), but the model overestimated risk in the subsequent 5-year prediction periods, with E/O=1.30 for 2006-2010 and E/O=1.44 for 2011-2015. The overall E/O ratio was 1.25 (95% confidence interval: 1.14–1.38), indicating that NCI-CCRAT overestimated the number of cases in this population of Black women. Of note, associations with colorectal cancer were weaker in the BWHS than in the development cohort for vigorous physical activity, aspirin/NSAID use, vegetable intake, and body mass index. We estimated discriminatory accuracy only for the first 5-year prediction period (2001-2005, N cases=151) for which the model was well calibrated, and obtained an age-adjusted AUC of 0.61 (95% CI: 0.57–0.66). Conclusions: The NCI-CCRAT overestimated the overall number of colorectal cancer cases by 25% in this prospective cohort of Black women, possibly due to population-specific differences in the relative risks of key factors driving colorectal cancer incidence. Discriminatory accuracy was similar to measures from the original validation cohort of White women and men. Impact: This study suggests that further research is needed to determine the most important predictors of colorectal cancer etiology in Black women and to develop and test a colorectal cancer risk prediction model for Black Americans. Citation Format: Jessica L. Petrick, Nelsy Castro-Webb, Gary R. Zirpoli, Kerrie P. Nelson, Julie R. Palmer, Ruth M. Pfeiffer. Validation of a colorectal cancer risk prediction model in US Black women [abstract]. In: Proceedings of the 16th AACR Conference on the Science of Cancer Health Disparities in Racial/Ethnic Minorities and the Medically Underserved; 2023 Sep 29-Oct 2;Orlando, FL. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2023;32(12 Suppl):Abstract nr B145.

An Integrative Pancreatic Cancer Risk Prediction Model in the UK Biobank.

Pancreatic cancer (PaCa) is a lethal cancer with an increasing incidence, highlighting the need for early prevention strategies. There is a lack of a comprehensive PaCa predictive model derived from large prospective cohorts. Therefore, we have developed an integrated PaCa risk prediction model for PaCa using data from the UK Biobank, incorporating lifestyle-related, genetic-related, and medical history-related variables for application in healthcare settings. We used a machine learning-based random forest approach and a traditional multivariable logistic regression method to develop a PaCa predictive model for different purposes. Additionally, we employed dynamic nomograms to visualize the probability of PaCa risk in the prediction model. The top five influential features in the random forest model were age, PRS, pancreatitis, DM, and smoking. The significant risk variables in the logistic regression model included male gender (OR = 1.17), age (OR = 1.10), non-O blood type (OR = 1.29), higher polygenic score (PRS) (Q5 vs. Q1, OR = 2.03), smoking (OR = 1.82), alcohol consumption (OR = 1.27), pancreatitis (OR = 3.99), diabetes (DM) (OR = 2.57), and gallbladder-related disease (OR = 2.07). The area under the receiver operating curve (AUC) of the logistic regression model is 0.78. Internal validation and calibration performed well in both models. Our integrative PaCa risk prediction model with the PRS effectively stratifies individuals at future risk of PaCa, aiding targeted prevention efforts and supporting community-based cancer prevention initiatives.

Clinical implications of incorporating genetic and non-genetic risk factors in CanRisk-based breast cancer risk prediction

BackgroundBreast cancer (BC) risk prediction models consider cancer family history (FH) and germline pathogenic variants (PVs) in risk genes. It remains elusive to what extent complementation with polygenic risk score (PRS) and non-genetic risk factor (NGRFs) data affects individual intensified breast surveillance (IBS) recommendations according to European guidelines. MethodsFor 425 cancer-free women with cancer FH (mean age 40·6 years, range 21–74), recruited in France, Germany and the Netherlands, germline PV status, NGRFs, and a 306 variant-based PRS (PRS306) were assessed to calculate estimated lifetime risks (eLTR) and estimated 10-year risks (e10YR) using CanRisk. The proportions of women changing country-specific European risk categories for IBS recommendations, i.e. ≥20 % and ≥30 % eLTR, or ≥5 % e10YR were determined. FindingsOf the women with non-informative PV status, including PRS306 and NGRFs changed clinical recommendations for 31·0 %, (57/184, 20 % eLTR), 15·8 % (29/184, 30 % eLTR) and 22·4 % (41/183, 5 % e10YR), respectively whereas of the women tested negative for a PV observed in their family, clinical recommendations changed for 16·7 % (25/150), 1·3 % (2/150) and 9·5 % (14/147). No change was observed for 82 women with PVs in high-risk genes (BRCA1/2, PALB2). Combined consideration of eLTRs and e10YRs identified BRCA1/2 PV carriers benefitting from IBS <30 years, and women tested non-informative/negative for whom IBS may be postponed. InterpretationFor women who tested non-informative/negative, PRS and NGRFs have a considerable impact on IBS recommendations. Combined consideration of eLTRs and e10YRs allows personalizing IBS starting age. FundingHorizon 2020, German Cancer Aid, Federal Ministry of Education and Research, Köln Fortune.

A pancreatic cancer risk prediction model (Prism) developed and validated on large-scale US clinical data

Pancreatic Duct Adenocarcinoma (PDAC) screening can enable early-stage disease detection and long-term survival. Current guidelines use inherited predisposition, with about 10% of PDAC cases eligible for screening. Using Electronic Health Record (EHR) data from a multi-institutional federated network, we developed and validated a PDAC RISk Model (Prism) for the general US population to extend early PDAC detection. Neural Network (PrismNN) and Logistic Regression (PrismLR) were developed using EHR data from 55 US Health Care Organisations (HCOs) to predict PDAC risk 6-18 months before diagnosis for patients 40 years or older. Model performance was assessed using Area Under the Curve (AUC) and calibration plots. Models were internal-externally validated by geographic location, race, and time. Simulated model deployment evaluated Standardised Incidence Ratio (SIR) and other metrics. With 35,387 PDAC cases, 1,500,081 controls, and 87 features per patient, PrismNN obtained a test AUC of 0.826 (95% CI: 0.824-0.828) (PrismLR: 0.800 (95% CI: 0.798-0.802)). PrismNN's average internal-external validation AUCs were 0.740 for locations, 0.828 for races, and 0.789 (95% CI: 0.762-0.816) for time. At SIR=5.10 (exceeding the current screening inclusion threshold) in simulated model deployment, PrismNN sensitivity was 35.9% (specificity 95.3%). Prism models demonstrated good accuracy and generalizability across diverse populations. PrismNN could find 3.5 times more cases at comparable risk than current screening guidelines. The small number of features provided a basis for model interpretation. Integration with the federated network provided data from a large, heterogeneous patient population and a pathway to future clinical deployment. Prevent Cancer Foundation, TriNetX, Boeing, DARPA, NSF, and Aarno Labs.

Lung Cancer Risk Prediction Models for Asian Ever-Smokers

IntroductionAlthough lung cancer prediction models are widely used to support risk-based screening, their performance outside Western populations remains uncertain. This study aims to evaluate the performance of 11 existing risk prediction models in multiple Asian populations and to refit prediction models for Asians. MethodsIn a pooled analysis of 186,458 Asian ever-smokers from 19 prospective cohorts, we assessed calibration (expected-to-observed ratio) and discrimination (area under the receiver operating characteristic curve [AUC]) for each model. In addition, we developed the “Shanghai models” to better refine risk models for Asians on the basis of two well-characterized population-based prospective cohorts and externally validated them in other Asian cohorts. ResultsAmong the 11 models, the Lung Cancer Death Risk Assessment Tool yielded the highest AUC (AUC [95% confidence interval (CI)] = 0.71 [0.67–0.74] for lung cancer death and 0.69 [0.67–0.72] for lung cancer incidence) and the Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial Model had good calibration overall (expected-to-observed ratio [95% CI] = 1.06 [0.90–1.25]). Nevertheless, these models substantially underestimated lung cancer risk among Asians who reported less than 10 smoking pack-years or stopped smoking more than or equal to 20 years ago. The Shanghai models were found to have marginal improvement overall in discrimination (AUC [95% CI] = 0.72 [0.69–0.74] for lung cancer death and 0.70 [0.67–0.72] for lung cancer incidence) but consistently outperformed the selected Western models among low-intensity smokers and long-term quitters. ConclusionsThe Shanghai models had comparable performance overall to the best existing models, but they improved much in predicting the lung cancer risk of low-intensity smokers and long-term quitters in Asia.

43P Performance evaluation of a combined risk model for breast cancer risk prediction in Indonesian population (TRIP Study)

43P Performance evaluation of a combined risk model for breast cancer risk prediction in Indonesian population (TRIP Study)

OA07.06 Development and Validation of a Protein-Based Lung Cancer Risk Prediction Model - Initial Results from the Lung Cancer Cohort Consortium (LC3)

OA07.06 Development and Validation of a Protein-Based Lung Cancer Risk Prediction Model - Initial Results from the Lung Cancer Cohort Consortium (LC3)

P1.14-02 Validation of Lung Cancer Risk Prediction Models in an Asbestos Exposed Population

P1.14-02 Validation of Lung Cancer Risk Prediction Models in an Asbestos Exposed Population

Machine learning endometrial cancer risk prediction model: integrating guidelines of European Society for Medical Oncology with the tumor immune framework

ObjectiveCurrent prognostic factors for endometrial cancer are not sufficient to predict recurrence in early stages. Treatment choices are based on the prognostic factors included in the risk classes defined by...