10-year Absolute Risk Research Articles

#1804 Long-term outcomes of CKD patients with lupus nephritis: a nationwide cohort study

Abstract Background and Aims Health trajectories of patients with lupus nephritis (LN) are poorly understood in the setting of moderate and advanced chronic kidney disease (CKD). Furthermore, data on comparison of outcomes of LN to other forms of CKD are scarce. Therefore, the aim of this study was to compare long-term outcomes between LN and CKD from other etiologies. Method Using data from the Swedish Renal Registry (2006-2021), we compared clinical outcomes between patients with moderate and advanced CKD related to LN (LN-CKD), and patients with CKD due to primary glomerular diseases (PGD-CKD, i.e. IgA nephropathy, focal segmental glomerulosclerosis, membranous nephropathy), and due to other common CKD etiologies with high cardiovascular risk (Other-CKD, mostly diabetes and nephrosclerosis). Cox proportional hazard models were used to estimate the adjusted hazard ratios of mortality, major cardiovascular events (MACE), heart failure (HF), kidney replacement therapy (KRT) and acute kidney injury (AKI). Models were adjusted for age, sex, comorbidities, clinical and biological data including kidney function, medications and healthcare utilization. Results At baseline, patients with LN-CKD (N = 317, 61 years, 76% women, median eGFR 30 mL/min per 1.73 m²) and PGD-CKD (N = 2296, 57 years, 30% women, eGFR 29 mL/min per 1.73 m²), were younger and had lower prevalence of cardiovascular disease than patients with Other-CKD (N = 34778, 75 years, 36% women, eGFR 25 mL/min per 1.73 m²). Over a median follow-up of 6.2 [3.3; 9.8] years, 19029 deaths (51%), 15768 (42%) MACE and 8390 (22%) KRT events occurred. The unadjusted 5-year absolute risks of death and MACE were high both in patients with LN-CKD (27% and 25%) and Other-CKD (50% and 44%), but lower in those with PGD-CKD (16% and 14%), whereas the 5-year absolute KRT risk was higher in patients with PGD-CKD (PGD-CKD: 37%, LN-CKD: 23%, Other-CKD: 23%). In multivariable analyses and as compared to PGD-CKD, patients with LN-CKD had a higher risk of death (adjusted HR: 1.63 [95% CI: 1.32-2.02]), MACE (1.65 [1.31-2.08]), HF (1.68 [1.31-2.14]) and AKI (1.45 [1.02-2.05]) and showed a lower risk of KRT (HR: 0.81 [0.64-1.02]), although the 95% CI slightly overlapped (Figure, panel A). In contrast, the risks of adverse events were similar between patients with LN-CKD and those with Other-CKD (Figure, panel B). Conclusion While LN-CKD had a lower risk of KRT than PGD-CKD, they exhibited a higher risk of death, MACE, KRT, HF and AKI, reaching the risk magnitude of patients with high cardiovascular burden (Other-CKD). Our findings may inform decisions for care planning in patients with advanced CKD and LN.

#886 Prevalence and prognostic relevance of electrocardiographic abnormalities among patients with ANCA-associated vasculitis

Abstract Background and Aims Cardiovascular disease (CVD) is one of the leading causes of death in patients with ANCA-associated vasculitis (AAV). However, in current clinical guidelines, limited evidence is available to support the recommendation of AAV-specific screening for cardiovascular comorbidities. This study reported the prevalence of electrocardiogram (ECG) abnormalities and investigated associations between ECG abnormalities and fatal CVD in a validated cohort of patients with AAV compared to matched controls. Method Based on a retrospective matched cohort design, patients with granulomatosis with polyangiitis [ICD-10: DM31.3] or microscopic polyangiitis [ICD-10: DM3.17] and accessible ECGs were identified in nationwide Danish health care registers from 2000 to 2021. Patients with AAV were matched 1:3 on age, sex, and calendar year of ECG measurement with controls without AAV. Index date was defined as the date of ECG measurement. ECGs were hierarchically categorized as either no, minor, or major ECG abnormalities according to contemporary literature. The associations between ECG abnormalities and fatal CVD were assessed in multiple Cox regression models adjusted for age, sex, and comorbidities. Counterfactual G-estimation of hazard ratios (HRs) standardized to age and sex was performed to estimate the 5-year absolute risk (AR) of fatal CVD according to ECG-abnormalities. Results A total of 1431 patients with AAV were matched to 4293 controls. The median age of patients with AAV and matched controls was 69 (IQR 58-77) years, and 52.3% were males. Median study follow-up time was 4.8 (IQR, 2.7-7.8) years. AAV was associated with increased prevalence of left ventricular hypertrophy (17.5% vs. 12.5%, p < 0.001), ST-T deviations (10.1% vs. 7.1%, p < 0.001), atrial fibrillation (9.6% vs. 7.5%, p = 0.017) and corrected QT prolongation (5.9% vs. 3.6%, p < 0.001) compared to controls. When examining associations of fatal CVD across no, minor or major ECG abnormalities, only the patients with AAV and major ECG abnormalities had a higher risk of fatal CVD [HR 1.99 (95% CI, 1.49-2.65)], compared to matched controls. This corresponded to a significantly higher standardized 5-year AR of fatal CVD compared to controls: 19.14% (95% CI, 16-22%) vs. 9.41% (95% CI, 8-11%). Conclusion AAV was associated with a higher prevalence of major ECG abnormalities such as left ventricular hypertrophy, atrial fibrillation, and ST-T deviations compared to matched controls. Moreover, the patients with AAV demonstrating major ECG abnormalities had a particularly elevated risk of fatal CVD. This indicates that AAV patients with major ECG abnormalities are a distinct high-risk group within AAV patients prone to cardiovascular-related mortality.

#2576 Effect of dialysis versus conservative care on all-cause mortality in older patients with advanced CKD: nationwide Swedish cohort study

Abstract Background and Aims Conservative care is emerging as an alternative to dialysis in older patients with advanced chronic kidney disease (CKD). However, high-quality evidence comparing the effect of dialysis versus conservative care on patient survival is lacking. Previous observational studies suffered from avoidable biases, such as immortal time bias and lead time bias, and adjusted for limited confounders. This study aimed to assess the effects of dialysis versus conservative care on mortality in older patients with advanced CKD. Method The target trial emulation framework was used to design and analyze this observational study. We conducted a nationwide observational cohort study using data from the Swedish Renal Registry between 2007-2021. We included all nephrologist-referred patients with eGFR < 20 ml/min/1.73 m2 who had a registered decision for dialysis or conservative care. Eligible patients were required to be ≥ 65 years with a Davies Comorbidity Score ≥ 2 or ≥ 80 years. The primary endpoint was 5-year all-cause mortality. Inverse probability of treatment weighting was used to adjust for 50 baseline confounders, including demographics, comorbidities, medication use, blood pressure, laboratory measurements, hospitalizations, and calendar year. Patients in the non-overlapping region of the propensity score distribution were excluded since they had an absolute indication for dialysis or conservative care. Cox proportional hazards regression was used to estimate intention-to-treat hazard ratios (HRs) with 95% confidence intervals (CIs), and the Kaplan-Meier estimator was used to estimate absolute risks. Additionally, we calculated the 5-year restricted mean survival times (RMST) and the 5-year RMST difference between dialysis and conservative care arms. To test the robustness and consistency of our main results, we performed a subgroup analysis in patients aged ≥80. Results Of 2834 eligible patients, 1903 had a registered choice for dialysis and 931 for conservative care. The median age was 81 years, 36.1% were female and median eGFR was 13.1 ml/min/1.73 m2. During the 5-year follow-up period, 29 conservative care patients started dialysis and 430 patients died before starting dialysis. After weighting, all confounders were balanced (standardized mean difference < 0.10). Compared to dialysis, conservative care was associated with a higher absolute 5-year risk of death (91.8% vs. 74.1%), corresponding with an absolute risk difference of 17.6% (95% CI 13.3%-21.9%) (Fig. 1). The adjusted hazard ratio for conservative care vs. dialysis was 2.3 (95% CI 2.0-2.6). The 5-year restricted mean survival times were 34.8 (95% CI 33.2-37.3) months for dialysis and 21.5 (95% CI 19.4-23.6) months for conservative care groups. The 5-year RMST difference was 13.3 months (95% CI 10.8-16.6), meaning that dialysis patients would live on average 13.3 months longer over a 5-year follow-up period than conservative care patients (Table 1). The results were consistent among people ≥ 80 years (adjusted HR 2.2, 95% CI 1.9-2.5; 5-year RMST difference 9.4 months, 95% CI 7.2-13.9). Conclusion In older patients with advanced CKD under nephrologist care, conservative care was associated with a higher risk of mortality than dialysis. These findings could improve the shared decision-making process with patients on the choice of kidney failure treatment.

Cardiovascular risk factors and aortic valve stenosis: towards 10-year absolute risk charts for primary prevention.

Due to aging populations the incidence of aortic valve stenosis (AVS) is increasing steeply. Since no medical therapy is available but only surgical interventions, it is highly warranted to identify modifiable risk factors for early prevention. The aim of the study was to investigate the associations of cardiovascular risk factors with AVS and to create 10-year absolute risk scores for use in primary prevention. In the Copenhagen General Population Study (N=93,979) lifestyle data, biochemical measures, and confounders were assessed at baseline. Risk factors with the strongest association with aortic valve stenosis from Cox regression analyses were included in ten-year risk prediction models. Ten-year absolute risk scores were conducted using the method of Fine-Gray proportional sub-hazards models, accounting for competing events. 1,132 individuals developed AVS during follow-up. Of well-known cardiovascular risk factors, those that associated with AVS included increasing levels of remnant cholesterol, triglycerides, lipoprotein(a), systolic blood pressure, and body mass index, low adherence to Danish dietary guidelines, current smoking, high alcohol consumption, lipid-lowering therapy and diabetes mellitus. Ten-year absolute risk scores increased when compiling the most important risk factors for AVS; age, sex, body mass index, systolic blood pressure, lipoprotein(a), and diabetes. Ten-year absolute risk increased from <1% to 19%. The presence of cardiovascular risk factors is associated with AVS, supporting that this disease, at least partly, may be modifiable through lifestyle changes. Risk charts combining cardiovascular risk factors have the potential to identify high-risk individuals, offering opportunities for preventive strategies. (Word count 245).

Polygenic risk-stratified screening for nasopharyngeal carcinoma in high-risk endemic areas of China: a cost-effectiveness study.

Nasopharyngeal carcinoma (NPC) has an extremely high incidence rate in Southern China, resulting in a severe disease burden for the local population. Current EBV serologic screening is limited by false positives, and there is opportunity to integrate polygenic risk scores for personalized screening which may enhance cost-effectiveness and resource utilization. A Markov model was developed based on epidemiological and genetic data specific to endemic areas of China, and further compared polygenic risk-stratified screening [subjects with a 10-year absolute risk (AR) greater than a threshold risk underwent EBV serological screening] to age-based screening (EBV serological screening for all subjects). For each initial screening age (30-34, 35-39, 40-44, 45-49, 50-54, 55-59, 60-64, and 65-69 years), a modeled cohort of 100,000 participants was screened until age 69, and then followed until age 79. Among subjects aged 30 to 54 years, polygenic risk-stratified screening strategies were more cost-effective than age-based screening strategies, and almost comprised the cost-effectiveness efficiency frontier. For men, screening strategies with a 1-year frequency and a 10-year absolute risk (AR) threshold of 0.7% or higher were cost-effective, with an incremental cost-effectiveness ratio (ICER) below the willingness to pay (¥203,810, twice the local per capita GDP). Specifically, the strategies with a 10-year AR threshold of 0.7% or 0.8% are the most cost-effective strategies, with an ICER ranging from ¥159,752 to ¥201,738 compared to lower-cost non-dominated strategies on the cost-effectiveness frontiers. The optimal strategies have a higher probability (29.4-35.8%) of being cost-effective compared to other strategies on the frontier. Additionally, they reduce the need for nasopharyngoscopies by 5.1-27.7% compared to optimal age-based strategies. Likewise, for women aged 30-54 years, the optimal strategy with a 0.3% threshold showed similar results. Among subjects aged 55 to 69 years, age-based screening strategies were more cost-effective for men, while no screening may be preferred for women. Our economic evaluation found that the polygenic risk-stratified screening could improve the cost-effectiveness among individuals aged 30-54, providing valuable guidance for NPC prevention and control policies in endemic areas of China.

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.

318 Discovering Subgroups with Supervised Machine Learning Models for Heterogeneity of Treatment Effect Analysis

OBJECTIVES/GOALS: The goal of the study is to provide insights into the use of machine learning methods as a means to predict heterogeneity of treatment effect (HTE) in participants of randomized clinical trials. METHODS/STUDY POPULATION: Using data from 2,441 participants enrolled in the ASPirin in Reducing Events in the Elderly (ASPREE) randomized controlled trial of daily low-dose aspirin vs placebo in the United States, we developed multivariable risk prediction models for the composite outcome of dementia, disability, or death. We used two machine learning techniques, decision trees and random forests, to develop novel non-parametric outcomes classifiers and generate risk-based subgroups. The comparator method was an extant semi-parametric proportional hazards predictive risk model. We then assessed HTE by examining the 5-year absolute risk reduction (ARR) of aspirin vs placebo in each risk subgroup. RESULTS/ANTICIPATED RESULTS: In the random forest classifier, the ARR at 5 years in the highest risk quintile was 13.7% (95% CI 3.1% to 24.4%). For the semi-parametric proportional hazards model, the ARR in the highest risk quintile was 15.1% (95% CI 4.0% to 26.3%). These results were comparable and provide evidence of the viability of internally developed parsimonious non-parametric machine learning models for HTE analysis. The decision tree model results (5-year ARR = 17.0%, 95% CI= -5.4% to 39.4% in the highest risk subgroup) exhibited more uncertainty in the results. DISCUSSION/SIGNIFICANCE: None of the models detected significant HTE on the relative scale; there was substantial HTE on the absolute scale in three of the models. Treatment benefit on the absolute scale may be regarded as bearing greater clinical importance and may be present even in the absence of benefit on the relative scale.

Abstract 6088: Predicting the risk of any-type cancer in asymptomatic adults using noninvasive glycosaminoglycan profiling

Abstract Background: Early detection of cancers through screening can reduce mortality. However, screening excludes individuals not considered at an increased risk of developing cancer, as determined by age, sex, or tobacco use. No molecular biomarker is routinely used to predict risk. Here, we investigated urine glycosaminoglycan profiles (GAGomes) as noninvasive biomarkers for any-cancer risk prediction. Methods: In this population-based case-control study, we included adults aged over 18 years from the Lifelines Cohort Study, Netherlands who were presumed healthy at baseline. All cases who self-reported any-type cancer or died by the 5-year study visit were confirmed in the Dutch Cancer Registry and matched 1:3 to randomly selected controls (1:1 for the interim analysis). We first developed a logistic regression reference model, including age, sex, and clinical variables to predict any-type cancer. Next, we added urine GAGomes (as a single aggregate score) to the model and assessed model improvement using the likelihood ratio test. We categorized the output of the saturated model (clinical variables plus urine GAGome score) into four groups based on bespoke specificity cut-offs: low-, moderate-, intermediate-, and high-risk. We estimated odds-ratios (ORs) for their association with any-type cancer compared to the moderate-risk group, as well as their sensitivity across subsets. Using the so-estimated OR in the high-risk group, we calculated the minimum pre-test 5-year absolute risk needed to reclassify an adult into a post-test 5-year absolute risk &amp;gt;3.00%. Results: In the interim analysis, we included 2054 adults (median age = 51 years, 57% females). Of them, 924 were diagnosed with any-type cancer over a 6 year-follow-up period (median time to diagnosis = 1.4 years). Urine GAGomes were found to be independent risk factors when added to the reference model, explaining 27% of the variance (Nagelkerke R2 = 0.26 vs. 0.21, p &amp;lt; 0.0001). The ORs in the low-, intermediate-, and high-risk groups were 0.05 (95% CI: 0.02-0.13 p &amp;lt; 0.0001), 5.31 (95% CI: 3.84-7.48, p &amp;lt; 0.0001) and 27.6 (95% CI: 14.3 - 51.6, p &amp;lt; 0.0001), respectively. The high-risk group detected +80% more cancers than the reference model at the same specificity (99.29%), particularly in situ and stage I (+204% and +64%) and among those aged 50-69 (+247%). Adults with a pre-test 5-year absolute risk &amp;gt; 0.25% returning a high-risk group would be reclassified to a post-test 5-year absolute risk &amp;gt; 3.00% for any-type cancer. Conclusion: Implementing urine GAGomes as a strategy for risk-stratified targeted screening could identify adults with increased risk of any-type cancer who are currently excluded from screening. Citation Format: Francesco Gatto. Predicting the risk of any-type cancer in asymptomatic adults using noninvasive glycosaminoglycan profiling [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 6088.

Abstract 7321: Second primary cancer risks following breast cancer in BRCA1/2 pathogenic variant carriers

Abstract Background: Second primary cancer (SPC) risk estimates in breast cancer (BC) survivors carrying germline BRCA1/2 pathogenic variants (PVs) remain uncertain. We estimated relative and absolute SPC risks following BC in BRCA1/2 PV carriers using genetic testing laboratory data in England, linked for the first time to population-scale electronic health records data from the National Disease Registration Service and Hospital Episode Statistics. Material and methods: The cohort contained 27,154 women diagnosed with primary BC and tested for germline BRCA1/2 PVs between 1995-2019. Follow-up began at the latest of the genetic test dates and one year following the BC diagnosis and continued until the first SPC diagnosis, death, migration, relevant surgery, and the end of 2020, whichever came first. We estimated overall and site-specific standardized incidence ratios (SIRs) by comparing observed and expected SPC counts, with expected counts calculated using incidences for the general English population accounting for calendar year, SPC site, age, and gender. We stratified the SIRs by age, estrogen-receptor status, and cancer site. We further assessed associations using Cox regression adjusted for age at BC diagnosis and estrogen-receptor status of the first BC, and estimated 10-year absolute SPC risks with Kaplan-Meier analysis. Results: There were 1909 BRCA1 (138 with SPCs) and 1926 BRCA2 (101 with SPCs) PV carriers, with mean follow-up lengths of 4.2 and 4.4y and median follow-up lengths of 3.4 and 3.6y. In BRCA1 PV carriers, there were elevated contralateral BC (SIR:15.3, 95%CI:11.5-19.8), ovarian (SIR:43.3, 95%CI:30.9-58.9), non-breast and non-ovarian combined (SIR:2.23, 95%CI:1.63-2.98), colorectal (SIR:4.72, 95%CI:2.58-7.92) and uterine (SIR:3.42, 95%CI:1.25-7.44) SPC risks. In BRCA2 PV carriers, there were increased contralateral BC (SIR:7.53, 95%CI:5.33-10.3), ovarian (SIR:16.5, 95%CI:10.1-25.5), non-breast and non-ovarian (SIR:1.72, 95%CI:1.27-2.27) and pancreatic (SIR:6.57, 95%CI:2.63-13.5) SPC risks. Consistently, when we compared BC survivors carrying a BRCA1 PV to BC survivors without a PV in either gene, they were at elevated contralateral BC (HR:3.94, 95%CI:2.91-5.33), ovarian (HR:35.7, 95%CI:21.3-60.0), non-breast and non-ovarian (HR:1.53, 95%CI:1.11-2.12) and colorectal (HR:2.95, 95%CI:1.56-5.55) SPC risks, whereas BRCA2 PV carriers were at increased contralateral BC (HR:2.59, 95%CI:1.83-3.65), ovarian (HR:12.3, 95%CI:6.93-21.9) and pancreatic (HR:5.32, 95%CI:2.30-12.3) SPC risks. We estimated 10-year cumulative contralateral BC, ovarian, and non-breast and non-ovarian SPC risks of 15%, 6.3%, and 7.8% in BRCA1 PV carriers and 12%, 2.9%, and 6.2% in BRCA2 PV carriers. Conclusions: Cancer risks were increased, at combined and specific sites, following BC in BRCA1/2 PV carriers. These findings could aid the clinical management of BC survivors carrying BRCA1/2 PVs. Citation Format: Isaac Allen, Hend Hassan, Yvonne Walburga, Catherine Huntley, Lucy Loong, Tameera Rahman, Beth Torr, Andrew Bacon, Craig Knott, Sophie Jose, Sally Vernon, Margreet Lüchtenborg, Joanna Pethick, Francesco Santaniello, Shilpi Goel, Sophie Allen, Katrina Lavelle, Fiona McRonald, Diana M. Eccles, Eva Morris, Steven Hardy, Clare Turnbull, Marc Tischkowitz, Paul Pharoah, Antonis C. Antoniou. Second primary cancer risks following breast cancer in BRCA1/2 pathogenic variant carriers [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 7321.

Abstract 3464: Development and prospective validation of an estrogen receptor positive breast cancer risk model to identify women who could benefit for risk-reducing therapies

Abstract Background: Breast cancer risk-reducing therapies such as tamoxifen and aromatase inhibitors are effective for ER-positive (ER+) but not for ER-negative disease. As these therapies have side effects it is important to identify women at risk of ER+ disease. We developed and validated ER-specific risk prediction models in a prospective cohort and evaluated the expected gains of predicting ER+ vs overall disease. Methods: The iCARE-Lit model integrates established questionnaire-based risk factors and a 313 variant polygenic risk score. We reparametrized the iCARE-Lit model to estimate risk of ER+ and ER- disease separately by obtaining relative risks from published literature for ages at menarche, menopause, first childbirth, and parity to account for their heterogeneous associations with risk of ER-defined subtypes. We estimated 5-year (yr) absolute risk (AR) for ER+ disease using an overall and ER+ model. We evaluated model performance in the Prostate, Lung, Colorectal and Ovarian (PLCO) Cancer Screening Trial prospective cohort. This included evaluation of calibration (expected-to-observed (E/O) ratio) and discrimination (area under the receiver operating characteristic curve (AUC)), among 46,571 self-reported non-Hispanic White women aged 50-75 yrs (1,052 ER+ cases with 5 yrs of follow-up from enrollment). We compared projected risk stratification of the overall and ER+ models among eligible US women in terms of proportions of women and future ER+ cases identified above the 3% 5-yr AR threshold recommended for risk-reducing interventions in the US. Results: In predicting ER+ disease in PLCO, on average the overall and ER+ AR were well calibrated; however, models overpredicted risk at the ≥3% 5-yr AR threshold, (E/O = 1.31 (1.20, 1. 42) and 1.23 (1.13, 1.34), respectively. Risk discrimination was similar from the overall and ER+ models (AUC=0.65 (0.64-0.67) and 0.66 (0.64-0.67), respectively. The overall model projected that 16% of women in the US could be eligible for risk-reducing therapies, compared to 11% when using ER+ model, resulting in ~1.4 million less women crossing the risk threshold when using the ER+ score. However, the percentage of ER+ cases in the US population projected to occur within 5-yrs among eligible women was larger for the overall (38%) than ER+ (32%) risk score. Conclusion: Findings suggest that ER-specific risk scores might not offer a substantial advantage compared to overall risk scores to identify women eligible for risk-reducing therapies. Further risk-benefit analyses are needed to quantify the potential gains of predicting ER+ vs overall breast cancer. Citation Format: Thomas U. Ahearn, Srijon Mukhopadhyay, Jeya Balasubramanian, Nilanjan Chatterjee, Montserrat García-Closas, Parichoy Pal Choudhury. Development and prospective validation of an estrogen receptor positive breast cancer risk model to identify women who could benefit for risk-reducing therapies [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 3464.

Abstract P428: Risk-Based Approach to Defining Prediabetes Among Middle-Aged Adults in the Atherosclerosis Risk in Communities (ARIC) Study

Introduction: Current prediabetes definitions, based on fasting blood glucose (FG) or HbA1c, are broad and capture many individuals who will never develop diabetes. Other risk factors including age, sex, body mass index (BMI), systolic blood pressure (SBP), and family history of diabetes, are not part of current definitions but are highly predictive of progression from prediabetes to diabetes. We evaluated whether approaches to defining prediabetes that incorporate standard clinical measures can improve 10-year diabetes risk prediction. Methods: We conducted a prospective analysis of 11,322 participants in the Atherosclerosis Risk in Communities (ARIC) Study aged 46-70 years (55% women, ~20% Black adults) without diabetes at baseline (1990-1992). We used logistic regression to develop risk equations for incident diabetes over ~10-year based on combinations of HbA1c, FG, and/or clinical measures (age, sex, BMI, SBP, family history of diabetes). We calculated area under the curve (AUC) for 10-year diabetes risk, comparing American Diabetes Association (ADA) prediabetes definitions (HbA1c 5.7-6.4%, FG 100-125 mg/dL, either, or both) with our new risk equations. Results: Over ~10 years of follow-up, there were 1230 incident cases of diabetes. The ~10-year absolute diabetes risk for persons with HbA1c 5.7-6.4% was 27% (AUC=0.66) and for persons with an FG of 100-125 mg/dL was 16% (AUC=0.64). Risk equations combining FG and HbA1c yielded an AUC of 0.76. The additional inclusion of all 5 clinical measures increasing the AUC to 0.78 (ΔAUC=0.02; p&lt;0.001). Simple clinical measures, primarily BMI, led to modest improvements in risk discrimination when added to HbA1c alone (ΔAUC=0.03; p&lt;0.001) or FG alone (ΔAUC=0.03; p&lt;0.001). Conclusions: Defining prediabetes based on both an elevated FG and elevated HbA1c (confirmatory definition) or using clinical variables (age, sex, BMI, SBP, family history) added to either FG or HbA1c can substantially improve our ability to identify middle-aged adults who will progress to diabetes.

Evaluating different low-density lipoprotein cholesterol thresholds to initiate statin for prevention of cardiovascular diseases in patients with type 2 diabetes mellitus: A target trial emulation study.

The present study aimed to evaluate the effect of statin therapy for primary prevention of cardiovascular diseases (CVDs) when initiating therapy at different baseline low-density lipoprotein cholesterol (LDL-C) levels in patients with type 2 diabetes mellitus (T2DM). Using territory-wide public electronic medical records in Hong Kong, we emulated a sequence of trials on patients with T2DM with elevated LDL-C levels in every calendar month from January 2008 to December 2014. Pooled logistic regression was applied to obtain the hazard ratios for the major CVDs (stroke, myocardial infarction, heart failure), all-cause mortality and major adverse events (myopathies and liver dysfunction) of statin therapy. The estimated hazard ratios (95% confidence intervals) of CVD incidence for statin initiation were 0.78 (0.72, 0.84) in patients with baseline LDL-C of 1.8-2.5 mmol/L (i.e., 70-99 mg/dL) and 0.90 (0.88, 0.92) in patients with baseline LDL-C ≥2.6 mmol/L (i.e., ≥100 mg/dL) in intention-to-treat analysis, which was 0.59 (0.51, 0.68) and 0.77 (0.74, 0.81) in per-protocol analysis, respectively. No significant increased risks were observed for the major adverse events. The absolute 10-year risk difference of overall CVD in per-protocol analysis was -7.1% (-10.7%, -3.6%) and -3.9% (-5.1%, -2.7%) in patients with baseline LDL-C 1.8-2.5 and ≥2.6 mmol/L, respectively. The effectiveness and safety were consistently observed in patients aged >75 years initiating statin at both LDL-C thresholds. Compared with the threshold of 2.6 mmol/L, initiating statin in patients with a lower baseline LDL-C level at 1.8-2.5 mmol/L can further reduce the risks of CVD and all-cause mortality without significantly increasing the risk of major adverse events in patients with T2DM, including patients aged >75 years.

Long-Term Risk of VTE in Sarcoidosis: A Nationwide Cohort Study

BackgroundChronic inflammation is increasingly recognised as a risk factor for venous thromboembolism (VTE), but unlike other inflammatory diseases including systemic lupus erythematosus and rheumatoid arthritis, data on the risk of VTE in patients with sarcoidosis are sparse. Research questionDo patients with sarcoidosis have a higher long-term risk of VTE (pulmonary embolism or deep venous thrombosis, and each of these individually) compared with the background population? Study designand methods: Using Danish nationwide registries, patients >18 years with newly diagnosed sarcoidosis (>2 in-/outpatient visits, 1996-2020) without prior VTE were matched 1:4 by age, sex, and comorbidities with individuals from the background population. The primary outcome was venous thromboembolism. ResultsWe included 14,742 patients with sarcoidosis and 58,968 matched individuals (median age 44.7 years; 57.2% males). The median follow-up was 8.8 years. Absolute 10-year risks of outcomes for patients with sarcoidosis vs. the background population were: VTE, 2.9% vs. 1.6% (P<0.0001); pulmonary embolism, 1.5% vs. 0.7% (P<0.0001); and deep venous thrombosis, 1.6% vs. 1.0% (P<0.0001). In multivariable Cox regression, sarcoidosis was associated with an increased rate of all outcomes in the first year after diagnosis (VTE: HR 4.94 [95% CI, 3.61-6.75]) and after the first year (VTE: HR 1.65 [95% CI, 1.45-1.87]) compared with the background population. These associations persisted when excluding patients with a history of cancer and censoring patients with incident cancer during follow-up. Three-month mortality was not significantly different between patients with VTE with and without sarcoidosis (adjusted HR 0.84 [95% CI, 0.61-1.15]). InterpretationIn this nationwide cohort study, sarcoidosis was associated with a higher long-term risk of VTE compared with a matched background population.

Estimation of age of onset and progression of breast cancer by absolute risk dependent on polygenic risk score and other risk factors.

Genetic, lifestyle, reproductive, and anthropometric factors are associated with the risk of developing breast cancer. However, it is not yet known whether polygenic risk score (PRS) and absolute risk based on a combination of risk factors are associated with the risk of progression of breast cancer. This study aims to estimate the distribution of sojourn time (pre-clinical screen-detectable period) and mammographic sensitivity by absolute breast cancer risk derived from polygenic profile and the other risk factors. The authors used data from a population-based case-control study. Six categories of 10-year absolute risk based on different combinations of risk factors were derived using the Breast and Ovarian Analysis of Disease Incidence and Carrier Estimation Algorithm. Women were classified into low, medium, and high-risk groups. The authors constructed a continuous-time multistate model. To calculate the sojourn time, they simulated the trajectories of subjects through the disease states. There was little difference in sojourn time with a large overlap in the 95% confidence interval (CI) between the risk groups across the six risk categories and PRS studied. However, the age of entry into the screen-detectable state varied by risk category, with the mean age of entry of 53.4years (95% CI, 52.2-54.1) and 57.0years (95% CI, 55.1-57.7) in the high-risk and low-risk women, respectively. In risk-stratified breast screening, the age at the start of screening, but not necessarily the frequency of screening, should be tailored to a woman's risk level. The optimal risk-stratified screening strategy that would improve the benefit-to-harm balance and the cost-effectiveness of the screening programs needs to be studied.

Characterisation of Symptom and Polysomnographic Profiles Associated with Cardiovascular Risk in a Sleep Clinic Population with Obstructive Sleep Apnoea.

Recent data have identified specific symptom and polysomnographic profiles associated with cardiovascular disease (CVD) in patients with obstructive sleep apnoea (OSA). Our aim was to determine whether these profiles were present at diagnosis of OSA in patients with established CVD and in those with high cardiovascular risk. Participants in the Sydney Sleep Biobank (SSB) database, aged 30-74 years, self-reported presence of CVD (coronary artery disease, cerebrovascular disease, or heart failure). In those without established CVD, the Framingham Risk Score (FRS) estimated 10-year absolute CVD risk, categorised as "low" (<6%), "intermediate" (6-20%), or "high" (>20%). Groups were compared on symptom and polysomnographic variables. 629 patients (68% male; mean age 54.3 years, SD 11.6; mean BMI 32.3 kg/m2, SD 8.2) were included. CVD was reported in 12.2%. A further 14.3% had a low risk FRS, 38.8% had an intermediate risk FRS, and 34.7% had a high risk FRS. Groups differed with respect to age, sex and BMI. OSA severity increased with established CVD and increasing FRS. The symptom of waking too early was more prevalent in the higher FRS groups (p=0.004). CVD and FRS groups differed on multiple polysomnographic variables; however, none of these differences remained significant after adjusting for age, sex, and BMI. Higher CVD risk was associated with waking too early in patients with OSA. Polysomnographic variations between groups were explained by demographic differences. Further work is required to explore the influence of OSA phenotypic characteristics on susceptibility to CVD.

Risk of alcohol-associated liver disease in the offspring of parents with alcohol-associated liver disease: A nationwide cohort study.

Offspring of patients with alcohol-associated liver disease (ALD) may have a higher risk of ALD. We examined their risk of ALD and survival with ALD. We used Danish nationwide registries to identify the offspring of patients diagnosed with ALD in 1996-2018 and 20:1 matched comparators from the general population. They were followed for ALD diagnosis through 2018. We used landmark competing risk analysis to estimate the age-specific absolute and relative 10-year risks of ALD. ALD was diagnosed in 385 of 60,707 offspring and 2842 of 1,213,357 comparators during 0.7 and 14.0 million person-years of follow-up, respectively, yielding an incidence rate ratio of 2.73 (95% CI: 2.44-3.03). The risk of being diagnosed with ALD within the next 10 years peaked at age 55 years for offspring and age 57years for comparators with 10-year risks of 1.66% (95% CI: 1.16-2.30) in offspring and 0.81% (95% CI: 0.68-0.97) in comparators at these ages. Offspring were younger at ALD diagnosis than comparators (median age of 47.4 vs. 48.9years), yet slightly more of them had developed cirrhosis (60.3% vs. 58.7%). Survival after ALD diagnosis was similar in offspring and comparators, adjusted hazard ratio=1.03 (95% CI: 0.88-1.21), so on average offspring died younger due to their younger age at diagnosis. Offspring of patients with ALD had a low but increased risk of ALD. Screening offspring for chronic liver disease may be unnecessary, but other interventions to mitigate alcohol-associated harm should be considered.

Chronic cough associated with COPD exacerbation, pneumonia and death in the general population

BackgroundChronic cough affects up to 10% of the general population and was previously perceived as a comorbidity of underlying conditions, but is nowadays classified as a disease in its own entity that could confer increased risk of morbidity and mortality. We tested the hypothesis that chronic cough is associated with increased risk of COPD exacerbation, pneumonia and all-cause mortality in the general population.MethodsWe identified 2801 individuals with chronic cough, defined as cough lasting &gt;8 weeks, among 44 756 randomly selected individuals from the Copenhagen General Population Study, and recorded COPD exacerbations, pneumonia and all-cause mortality during follow-up.ResultsDuring up to 5.9 years of follow-up (median 3.4 years), 173 individuals experienced COPD exacerbation, 767 experienced pneumonia and 894 individuals died. Individuals with chronic coughversusthose without had cumulative incidences at age 80 years of 12%versus3% for COPD exacerbation, 30%versus15% for pneumonia, and 25%versus13% for death from all causes. After adjustment for age, sex and smoking, individuals with chronic coughversusthose without had adjusted hazard ratios of 4.6 (95% CI 2.9–7.2) for COPD exacerbation, 2.2 (1.7–2.7) for pneumonia and 1.7 (1.4–2.0) for all-cause mortality. Among current smokers aged &gt;60 years with airflow limitation, those withversuswithout chronic cough had an absolute 5-year risk of 10%versus4% for COPD exacerbation, 16%versus8% for pneumonia and 19%versus12% for all-cause mortality.ConclusionChronic cough is associated with higher risks of COPD exacerbation, pneumonia and death, independent of airflow limitation and smoking.

Lipoprotein(a) and Risks of Peripheral Artery Disease, Abdominal Aortic Aneurysm, and Major Adverse LimbEvents.

Lp(a) (lipoprotein[a])-lowering therapy to reduce cardiovascular disease is under investigation in phase 3 clinical trials. High Lp(a) may be implicated in peripheral artery disease (PAD), abdominal aortic aneurysms (AAAs), and major adverse limb events (MALE). The authors investigated the association of high Lp(a) levels and corresponding LPA genotypes with risk of PAD, AAA, and MALE. The authors included 108,146 individuals from the Copenhagen General Population Study. During follow-up, 2,450 developed PAD, and 1,251 AAAs. Risk of MALE was assessed in individuals with PAD at baseline and replicated in the Copenhagen City Heart Study. Higher Lp(a) was associated with a stepwise increase in risk of PAD and AAA (P for trend<0.001). For individuals with Lp(a) levels≥99th (≥143mg/dL, ≥307nmol/L) vs<50th percentile (≤9mg/dL,≤17nmol/L), multivariable-adjusted HRs were 2.99 (95%CI: 2.09-4.30) for PAD and 2.22 (95%CI: 1.21-4.07) for AAA. For individuals with PAD, the corresponding incidence rate ratio for MALE was 3.04 (95%CI: 1.55-5.98). Per 50mg/dL (105nmol/L) genetically higher Lp(a) risk ratios were 1.39 (95%CI: 1.24-1.56) for PAD and 1.21 (95%CI: 1.01-1.44) for AAA, consistent with observational risk ratios of 1.33 (95%CI: 1.24-1.43) and 1.27 (95%CI: 1.15-1.41), respectively. In women smokers aged 70 to 79 years with Lp(a)<50th and≥99th percentile, absolute 10-year risks of PAD were 8% and 21%, and equivalent risks in men 11% and 29%, respectively. For AAA, corresponding risks were 2% and 4% in women, and 5% and 12% in men. High Lp(a) levels increased risk of PAD, AAA, and MALE by 2- to 3-fold in the general population, opening opportunities for prevention given future Lp(a)-lowering therapies.

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.

Impact of the metabolic syndrome on cardiopulmonary morbidity and mortality in individuals with lung function impairment: a prospective cohort study of the Danish general population.

Whether the metabolic syndrome plays a role for the prognosis of individuals with lung function impairment (preserved ratio impaired spirometry (PRISm) or airflow limitation) is unclear. We hypothesised that the metabolic syndrome in individuals with lung function impairment is associated with increased cardiopulmonary morbidity and mortality. The Copenhagen General Population Study was initiated in 2003 based on a random sample of white men and women aged 20-100 years drawn from the Danish general population. The risk of ischemic heart disease/heart failure, respiratory disease, and all-cause mortality was analysed with Cox models adjusted for age, sex, current smoking, and asthma during 15 years of follow-up. Among 106,845 adults, 86,159 had normal lung function, 6126 had PRISm, and 14,560 had airflow limitation. We observed 10,448 hospital admissions for ischemic heart disease/heart failure, 21,140 for respiratory disease, and 11,125 deaths. Individuals with versus individuals without the metabolic syndrome generally had higher 5-year absolute risk of all outcomes, including within those with normal lung function, mild-moderate-severe PRISm, and very mild-mild-moderate-severe airflow limitation alike. Compared to individuals without the metabolic syndrome and with normal lung function, those with both the metabolic syndrome and severe PRISm had hazard ratios of 3.74 (95% CI: 2.53-5.55; p<0.0001) for ischemic heart disease/heart failure, 5.02 (3.85-6.55; p<0.0001) for respiratory disease, and 5.32 (3.76-7.54; p<0.0001) for all-cause mortality. Corresponding hazard ratios in those with both the metabolic syndrome and severe airflow limitation were 2.89 (2.34-3.58; p<0.0001) for ischemic heart disease/heart failure, 5.98 (5.28-6.78; p<0.0001) for respiratory disease, and 4.16 (3.50-4.95; p<0.0001) for all-cause mortality, respectively. The metabolic syndrome explained 13% and 27% of the influence of PRISm or airflow limitation on ischemic heart disease/heart failure and all-cause mortality. The metabolic syndrome conferred increased risk of cardiopulmonary morbidity and mortality at all levels of lung function impairment. Danish Lung Foundation, Danish Heart Foundation, Capital Region of Copenhagen, and Boehringer Ingelheim. JV is supported by the NIHR Manchester BRC.