Ovarian Cancer Screening Trial Research Articles

A Linear Relationship between the Number of Cancers among First-Degree Relatives and the Risk of Multiple Primary Cancers.

With advances in the early detection and treatment of cancer, the incidence of multiple primary cancers (MPC) or second primary cancers has increased over time. Characterization of etiologic risk factors, including family history of cancer, within the general population is critical for assessing MPC risk in patients. We examined the association between family history of cancer among first-degree relatives and MPC risk in a prospective study of 139,958 participants from the Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial. Cox proportional hazard models were used to calculate HRs and 95% confidence intervals (95% CI), adjusting for potential confounders. Over a median follow-up of 16 years (IQR: 11-19 years), 6,170 participants were diagnosed with MPC. Having a family history of cancer increased the risk of MPC by 18% (HR, 1.18; 95% CI, 1.12-1.24). A positive linear trend was observed between the reported number of cancers in the family history and MPC risk with HRs (95% CI) of 1.13 (1.07-1.20), 1.23 (1.14-1.33), 1.29 (1.15-1.45), and 1.42 (1.20-1.70) for one, two, three, and four or more cancers among first-degree relatives, respectively (Ptrend = 2.36 × 10-13). No significant differences were observed by cancer histology or specific cancer types reported in the family history. Our study demonstrates that the family history of cancer is an important risk factor for the development of MPCs and that a comprehensive assessment of the number of cancers reported among first-degree relatives may identify those at higher risk who may benefit from targeted cancer prevention and screening strategies. Prevention Relevance: Our study makes a substantial contribution to the understanding of risk factors for MPCs in the general population. It demonstrates that individuals with a strong family history of cancer are at higher risk for MPCs and may benefit from more targeted cancer prevention and screening interventions.

Multivitamin Use and Mortality Risk in 3 Prospective US Cohorts

One in 3 US adults uses multivitamins (MV), with a primary motivation being disease prevention. In 2022, the US Preventive Services Task Force reviewed data on MV supplementation and mortality from randomized clinical trials and found insufficient evidence for determining benefits or harms owing, in part, to limited follow-up time and external validity. To estimate the association of MV use with mortality risk, accounting for confounding by healthy lifestyle and reverse causation whereby individuals in poor health initiate MV use. This cohort study used data from 3 prospective cohort studies in the US, each with baseline MV use (assessed from 1993 to 2001), and follow-up MV use (assessed from 1998 to 2004), extended duration of follow-up up to 27 years, and extensive characterization of potential confounders. Participants were adults, without a history of cancer or other chronic diseases, who participated in National Institutes of Health-AARP Diet and Health Study (327 732 participants); Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial (42 732 participants); or Agricultural Health Study (19 660 participants). Data were analyzed from June 2022 to April 2024. Self-reported MV use. The main outcome was mortality. Cox proportional hazard models were used to estimate hazard ratios (HRs) and 95% CIs. Among 390 124 participants (median [IQR] age, 61.5 [56.7-66.0] years; 216 202 [55.4%] male), 164 762 deaths occurred during follow-up; 159 692 participants (40.9%) were never smokers, and 157 319 participants (40.3%) were college educated. Among daily MV users, 49.3% and 42.0% were female and college educated, compared with 39.3% and 37.9% among nonusers, respectively. In contrast, 11.0% of daily users, compared with 13.0% of nonusers, were current smokers. MV use was not associated with lower all-cause mortality risk in the first (multivariable-adjusted HR, 1.04; 95% CI, 1.02-1.07) or second (multivariable-adjusted HR, 1.04; 95% CI, 0.99-1.08) halves of follow-up. HRs were similar for major causes of death and time-varying analyses. In this cohort study of US adults, MV use was not associated with a mortality benefit. Still, many US adults report using MV to maintain or improve health.

Developing an optimal stratification model for colorectal cancer screening and reducing racial disparities in multi-center population-based studies

BackgroundEarly detection of colorectal neoplasms can reduce the colorectal cancer (CRC) burden by timely intervention for high-risk individuals. However, effective risk prediction models are lacking for personalized CRC early screening in East Asian (EAS) population. We aimed to develop, validate, and optimize a comprehensive risk prediction model across all stages of the dynamic adenoma-carcinoma sequence in EAS population.MethodsTo develop precision risk-stratification and intervention strategies, we developed three trans-ancestry PRSs targeting colorectal neoplasms: (1) using 148 previously identified CRC risk loci (PRS148); (2) SNPs selection from large-scale meta-analysis data by clumping and thresholding (PRS183); (3) PRS-CSx, a Bayesian approach for genome-wide risk prediction (PRSGenomewide). Then, the performance of each PRS was assessed and validated in two independent cross-sectional screening sets, including 4600 patients with advanced colorectal neoplasm, 4495 patients with non-advanced adenoma, and 21,199 normal individuals from the ZJCRC (Zhejiang colorectal cancer set; EAS) and PLCO (the Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial; European, EUR) studies. The optimal PRS was further incorporated with lifestyle factors to stratify individual risk and ultimately tested in the PLCO and UK Biobank prospective cohorts, totaling 350,013 participants.ResultsThree trans-ancestry PRSs achieved moderately improved predictive performance in EAS compared to EUR populations. Remarkably, the PRSs effectively facilitated a thorough risk assessment across all stages of the dynamic adenoma-carcinoma sequence. Among these models, PRS183 demonstrated the optimal discriminatory ability in both EAS and EUR validation datasets, particularly for individuals at risk of colorectal neoplasms. Using two large-scale and independent prospective cohorts, we further confirmed a significant dose–response effect of PRS183 on incident colorectal neoplasms. Incorporating PRS183 with lifestyle factors into a comprehensive strategy improves risk stratification and discriminatory accuracy compared to using PRS or lifestyle factors separately. This comprehensive risk-stratified model shows potential in addressing missed diagnoses in screening tests (best NPV = 0.93), while moderately reducing unnecessary screening (best PPV = 0.32).ConclusionsOur comprehensive risk-stratified model in population-based CRC screening trials represents a promising advancement in personalized risk assessment, facilitating tailored CRC screening in the EAS population. This approach enhances the transferability of PRSs across ancestries and thereby helps address health disparity.Graphical

Association of a polygenic risk score with risk of abnormal ultrasound findings and ovarian cancer in the Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial.

10542 Background: Several susceptibility loci have been identified in genome-wide association studies (GWAS) of single nucleotide polymorphisms (SNPs) that are associated with ovarian cancers. Polygenic risk scores (PRS) combining certain loci have shown limited performance at predicting ovarian cancer risk. However, the utilization of PRS with ovarian cancer screening tests, like transvaginal ultrasound (TVU), warrants further evaluation. This study investigated the association between a polygenic risk score and ovarian cancer risk and abnormal TVU screening in a large-scale randomized controlled ovarian cancer screening trial. Methods: A polygenic risk score was constructed using weighted sums of risk alleles of 15 ovarian cancer SNPs identified in previous GWAS to evaluate the association with abnormal TVU screening results and ovarian cancer outcomes in female participants of the Prostate, Lung, Colorectal, and Ovarian (PLCO) Cancer Screening Trial. Complete genotyping information, TVU results, and ovarian cancer outcomes were available for 15722 females (mean age 62.2 years +/- 5.3) with European ancestry, including 116 ovarian cancer cases. Abnormal TVU was defined as ovarian volume or cyst volume > 10 cm3, any solid/cystic component within a cystic ovarian tumor, or any solid area or papillary projection extending into the cavity of a cystic ovarian tumor. Polygenic risk scores were categorized as quartiles and odds ratios (OR) were calculated to assess risk of ovarian cancer or abnormal TVU result using the lowest PRS quartile as reference category. Results: TVU screening results were abnormal for 732 (732/15722, 4.66%) participants and normal for 14990 (14990/15722, 95.3%) participants at trial baseline. PRS within the fourth quartile demonstrated increased risk of ovarian cancer with reference to the lowest quartile (OR 1.77, 95% confidence interval: 1.04-3.08). Similarly, women in the highest quartile of the PRS had increased odds of an abnormal TVU screening result (OR 1.46, 95% confidence interval: 1.17-1.83) compared to women in the lowest PRS quartile. Conclusions: A polygenic risk score including 15 loci previously shown to be associated with ovarian cancer risk showed an association with increased ovarian cancer risk in the PLCO trial. Furthermore, the PRS was also associated with abnormal ultrasound results. Our findings support further evaluation of polygenic risk scores and transvaginal ultrasound for ovarian cancer risk stratification and screening approaches.

Deep Survival Models Can Improve Long-Term Mortality Risk Estimates from Chest Radiographs

Deep learning has recently demonstrated the ability to predict long-term patient risk and its stratification when trained on imaging data such as chest radiographs. However, existing methods formulate estimating patient risk as a binary classification, typically ignoring or limiting the use of temporal information, and not accounting for the loss of patient follow-up, which reduces the fidelity of estimation and limits the prediction to a certain time horizon. In this paper, we demonstrate that deep survival and time-to-event prediction models can outperform binary classifiers at predicting mortality and risk of adverse health events. In our study, deep survival models were trained to predict risk scores from chest radiographs and patient demographic information in the Prostate, Lung, Colorectal, and Ovarian (PLCO) cancer screening trial (25,433 patient data points used in this paper) for 2-, 5-, and 10-year time horizons. Binary classification models that predict mortality at these time horizons were built as baselines. Compared to the considered alternative, deep survival models improve the Brier score (5-year: 0.0455 [95% CI, 0.0427–0.0482] vs. 0.0555 [95% CI, (0.0535–0.0575)], p < 0.05) and expected calibration error (ECE) (5-year: 0.0110 [95% CI, 0.0080–0.0141] vs. 0.0747 [95% CI, 0.0718–0.0776], p < 0.05) for those fixed time horizons and are able to generate predictions for any time horizon, without the need to retrain the models. Our study suggests that deep survival analysis tools can outperform binary classification in terms of both discriminative performance and calibration, offering a potentially plausible solution for forecasting risk in clinical practice.

Association between coffee and tea consumption and ovarian cancer incidence: A prospective analysis in the PLCO dataset.

Epidemiological evidence regarding the relationship between coffee and tea consumption and the risk of ovarian cancer (OC) is inconsistent. Therefore, we aimed to quantitatively investigate this topic in a large prospective cohort study. This cohort study included 24,715 individuals recruited from the Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trials between 1993 and 2001. The data used for our analysis included the latest follow-up information collected up to 2015. Coffee intake of ≥4 cups/day (hazard ratio [HR], 0.586; 95% confidence interval [CI]: 0.356-0.966) or caffeine intake of 458.787 mg/day (HR, 0.607; 95% CI: 0.411-0.895) were associated with the lowest HR of incident OC in the fully adjusted model. Participants who consumed varying amounts of tea did not exhibit a statistically significant reduction in the risk of OC. Our findings suggest that a higher consumption of coffee or caffeine is associated with a reduced risk of OC. However, no statistically significant association was observed between tea consumption and the risk of OC.

Adherence to the low-fat diet pattern reduces the risk of lung cancer in American adults aged 55 years and above: a prospective cohort study

ObjectivesThere is little evidence on the association between low-fat dietary patterns and lung cancer risk among middle-aged and older adults. To fill this gap, we comprehensively investigated the association of adherence to a low-fat diet (LFD) and intake of different fat components including saturated, monounsaturated, and polyunsaturated fatty acids with incidence of lung cancer and its subtypes [non-small cell lung cancer (NSCLC) and small cell lung cancer (SCLC)] among adults aged 55 years and older. DesignA prospective cohort study with a mean follow-up time of 8.8 years. Setting and participantsThis study used data from the Prostate, Lung, Colorectal, and Ovarian (PLCO) Cancer Screening Trial. The study population included 98,459 PLCO participants age 55 and over at baseline who completed food frequency questionnaires providing detailed dietary information and had no history of cancer. MethodsDietary intake was assessed using a validated food frequency questionnaire at baseline. A LFD score was calculated based on fat, protein, and carbohydrate intake as a percentage of total calories. Cox proportional hazards regression models were used to estimate hazard ratios (HRs) and 95% confidence intervals (CIs) for the association between LFD score and intake of fat components (in quartiles) and incident lung cancer and its subtypes over follow-up. Restricted cubic spline analyses were conducted to examine possible nonlinear relationships. Subgroup analyses were performed to evaluate potential effect modifiers, and several sensitivity analyses were conducted to assess the stability of the findings. ResultsDuring a follow-up of 869,807.9 person-years, 1,642 cases of lung cancer were observed, consisting of 1,408 (85.75%) cases of NSCLC and 234 (14.25%) cases of SCLC. The highest versus the lowest quartiles of the LFD score were found to be associated with a reduced risk of lung cancer (HR, 0.76; 95% CI, 0.66−0.89), NSCLC (HR, 0.79; 95% CI, 0.67−0.93), and SCLC (HR, 0.59; 95% CI, 0.38−0.92). The restricted cubic spline plots demonstrated a linear dose-response relationship between the LFD score and the risk of lung cancer as well as its subtypes. This risk reduction association for overall lung cancer was more pronounced in smokers (HR, 0.71; 95% CI, 0.60−0.84; P for interaction = 0.003). For fat components, high consumption of saturated fatty acids was associated with an increased lung cancer risk (HR, 1.35; 95% CI, 1.10–1.66), especially for SCLC (HR, 2.05; 95% CI, 1.20–3.53). No significant association was found between consumption of monounsaturated or polyunsaturated fatty acids and incident lung cancer and its subtypes. ConclusionsOur findings suggest that adherence to LFD may reduce the lung cancer risk, particularly in smokers; while high saturated fatty acids consumption may increase lung cancer risk, especially for SCLC, among middle-aged and older adults in the US population.

Circulating microRNAs in association with pancreatic cancer risk within 5 years.

Early detection is critical for improving pancreatic cancer prognosis. Our study aims to identify circulating microRNAs (miRNAs) associated with pancreatic cancer risk. The two-stage study used plasma samples collected ≤5 years prior to cancer diagnosis, from case-control studies nested in five prospective cohort studies. The discovery stage included 185 case-control pairs from the Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial. Replication stage samples comprised 277 pairs from Shanghai Women's Health Study/Shanghai Men's Health Study, Southern Community Cohort Study, and Multiethnic Cohort Study. Seven hundred and ninety-eight miRNAs were measured using the NanoString nCounter Analysis System. Odds ratios (OR) and 95% confidence intervals (CI) for per 10% change in miRNAs in association with pancreatic cancer risk were derived from conditional logistic regression analysis in discovery and replication studies, separately, and then meta-analyzed. Stratified analysis was conducted by age at diagnosis (<65/≥65 years) and time interval between sample collection and diagnosis (≤2/>2 years). In the discovery stage, 120 risk associated miRNAs were identified at p < .05. Three were validated in the replication stage: hsa-miR-199a-3p/hsa-miR-199b-3p, hsa-miR-767-5p, and hsa-miR-191-5p, with respective ORs (95% CI) being 0.89 (0.84-0.95), 1.08 (1.02-1.13), and 0.90 (0.85-0.95). Five additional miRNAs, hsa-miR-640, hsa-miR-874-5p, hsa-miR-1299, hsa-miR-22-3p, and hsa-miR-449b-5p, were validated among patients diagnosed at ≥65 years, with OR (95% CI) of 1.23 (1.09-1.39), 1.33 (1.16-1.52), 1.25 (1.09-1.43), 1.28 (1.12-1.46), 0.76 (0.65-0.89), and 1.22 (1.07-1.39), respectively. The miRNA targets were enriched in pancreatic carcinogenesis/progression-related pathways. Our study suggests that circulating miRNAs may identify individuals at high risk for pancreatic cancer ≤5 years prior to diagnosis, indicating its potential utility in cancer screening and surveillance.

Bayesian and deep-learning models applied to the early detection of ovarian cancer using multiple longitudinal biomarkers.

Ovarian cancer is the most lethal of all gynecological cancers. Cancer Antigen 125 (CA125) is the best-performing ovarian cancer biomarker which however is still not effective as a screening test in the general population. Recent literature reports additional biomarkers with the potential to improve on CA125 for early detection when using longitudinal multimarker models. Our data comprised 180 controls and 44 cases with serum samples sourced from the multimodal arm of UK Collaborative Trial of Ovarian Cancer Screening (UKCTOCS). Our models were based on Bayesian change-point detection and recurrent neural networks. We obtained a significantly higher performance for CA125-HE4 model using both methodologies (AUC 0.971, sensitivity 96.7% and AUC 0.987, sensitivity 96.7%) with respect to CA125 (AUC 0.949, sensitivity 90.8% and AUC 0.953, sensitivity 92.1%) for Bayesian change-point model (BCP) and recurrent neural networks (RNN) approaches, respectively. One year before diagnosis, the CA125-HE4 model also ranked as the best, whereas at 2 years before diagnosis no multimarker model outperformed CA125. Our study identified and tested different combination of biomarkers using longitudinal multivariable models that outperformed CA125 alone. We showed the potential of multivariable models and candidate biomarkers to increase the detection rate of ovarian cancer.

Safety net hospital risk model demonstrates stronger, population-specific applicability in characterizing lung cancer risk.

Determining lung cancer (LC) risk using personalized risk stratification may improve screening effectiveness. While the Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial (PLCO) is a well-established stratification model for LC screening, it was derived from a predominantly Caucasian population and its effectiveness in a safety net hospital (SNH) population is unknown. We have developed a model more tailored to the SNH population and compared its performance to the PLCO model in a SNH setting. Retrospective dataset was compiled from patients screened for LC at SNH from 2015 to 2019. Descriptive statistics were calculated using the following variables: age, sex, race, education, body mass index (BMI), smoking history, personal cancer history, family LC history, chronic obstructive pulmonary disease (COPD), and emphysema. Variables distribution was compared using t- and chi-square tests. LC risk scores were calculated using SNH and PLCO models and categorized as low (scores <0.65%), moderate (0.65-1.49%), and high (>1.5%). Linear regression was applied to evaluate the relationship between models and covariates. Of 896 individuals, 38 were diagnosed with LC. Data reflected the SNH patient demographics, which predominantly were African American (53.5%), current smokers (69.9%), and with emphysema (70.1%). Among the non-LC cohort, SNH model most frequently categorized patients as low risk, while PLCO model most frequently classified patients as moderate risk. Among the LC cohort, there was no significant difference between mean scores or risk stratification. SNH model showed 92.1% sensitivity and 96.8% specificity while PLCO model showed 89.4% sensitivity and 26.1% specificity. Emphysema demonstrated a strong association in SNH model (P<0.001) while race showed no relation. SNH model demonstrated greater specificity for characterizing LC risk in a SNH population. The results demonstrated the importance of study sample representation when identifying risk factors in a stratification model.

Abstract 468: Associations of circulating miRNAs with pancreatic cancer risk differ by years between blood collection &amp; and cancer diagnosis

Abstract Pancreatic adenocarcinoma (PDAC) is one of the most fatal cancers and currently used biomarkers for PDAC diagnosis are not adequate for early detection. Therefore, new biomarkers for PDAC early detection are urgently needed. We conducted a large-scale study to prospectively evaluate the association of circulating miRNAs with PDAC risk. Specially, we aimed to examine whether the associations of miRNAs expression levels and changes of those levels from blood collection to cancer diagnosis (BCCD) with PDAC risk. We used pre-diagnostic plasma samples of 1307 PDAC cases from 5 cohort studies (290 from the Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial, 395 from the Shanghai Women’s and Men’s Health Studies, 154 from the Southern Community Cohort Study, and 468 from the Multiethnic Cohort). Cancer free controls (n=1307) were selected and individually matched to index cases in each cohort. All miRNAs were measured by the NanoString nCounter Analysis System using the Human v3 miRNA Expression panel (798 miRNAs total). Data were pooled and quantile normalized prior to analysis. Associations of circulating miRNAs with PDAC risk were assessed by odds ratio per decile change in miRNA levels using conditional logistic regression. Fold-change of miRNA for cases versus controls by BCCD was analyzed via linear regression. Median time of BCCD was 7.13 years (range: 0.03-19.61 years). We first conducted the association analysis stratified by BCCD: &amp;lt; 5 years, 5-10 years, and &amp;gt;10 years. We found 20 miRNAs were significantly associated with PDAC risk in the &amp;lt;5 years after FDR correction. In the analysis of regressing fold-change of these 20 miRNAs on BCCD, we found the fold change of miR-155-5p was inversely associated with BCCD (β=-0.04, P&amp;lt;0.01), i.e., the case-control difference was larger when blood was drawn closer to PDAC diagnosis, suggesting its potential utility as disease biomarkers for early detection. Conversely, the fold-change of miR-93-5p (β=0.04, P=0.01), miR-223-3p (β=0.03, P=0.02), let-7i-5p (β=0.03, P&amp;lt;0.01), and miR-191-5p (β=0.05, P&amp;lt;0.01) were positively associated with BCCD, i.e., the case-control difference increased with an increasing the time interval of BCCD, indicating they are more likely biomarkers for PDAC risk. Ingenuity pathway analyses of these five significant miRNAs revealed that they were significantly enriched in ribonucleotide reductase signaling pathways at p&amp;lt;0.05. Notably, the silencing of ribonucleotide reductase M2 subunit suppresses tumorigenesis in pancreatic cancer by deactivating the PI3K/AKT/mTOR pathway, further supporting our findings. Findings of this large-scale multiethnic study suggest that changes in circulating levels of let-7i-5p, miR-155-5p, miR-93-5p, miR-191-5p, and miR-223-3p may help identify individuals at an elevated risk of developing PDAC, allowing for screening tests and close surveillance. Citation Format: Hui Cai, Veronica Wendy Setiawan, Xingyi Guo, Jie Wu, Rachael Z Stolzenberg-Solomon, Claire Zhu, Yu-Tang Gao, Jordan Berlin, Fei Ye, Qiuyin Cai, Wei Zheng, Xiao-ou Shu. Associations of circulating miRNAs with pancreatic cancer risk differ by years between blood collection &amp; and cancer diagnosis [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 468.

Abstract 3427: Serum metabolites and pancreatic ductal adenocarcinoma in two prospective cohorts

Abstract Objective: Metabolomics may provide insight to understand the etiology and potential mechanisms for pancreatic ductal adenocarcinoma (PDAC) development beyond established risk factors of smoking, diabetes, and adiposity. We extended our previous metabolomic study by including additional participants from the Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial (PLCO) with the aim to examine prospective associations between serum metabolites and incident PDAC. Method: We measured 1483 known metabolites in pre-diagnostic serum (up to 24 years) in two PDAC nested case-control studies within the PLCO (360 cases-control pairs) of mostly non-smoker men and women and Alpha-Tocopherol, Beta-Carotene Cancer Prevention Study (ATBC, 372 cases-control pairs) male smokers using liquid chromatography, high resolution/accurate mass spectrometry. Controls were incidence-density sampled and matched to cases by age, blood draw date, sex, race/ethnicity, and cohort. We used conditional logistic regression to calculate odds ratios and 95% confidence intervals for PDAC per standard deviation increase in log10-metabolite levels in each cohort and combined using fixed-effect meta-analyses and stratified by follow-up time. We also performed ElasticNet regression in each cohort to identify metabolites. We assessed associations of 72 predefined metabolic pathways with PDAC by comparing Fisher’s statistic of metabolite p-values per pathway with a null distribution based on 500,000 permutations. Results: Sixty-six metabolites were associated with PDAC at a false discovery rate (FDR)&amp;lt;0.05 with 26 below the Bonferroni threshold (P&amp;lt;3.4 × 10−5). Among the 66 metabolites, 14 and 15 were selected in ElasticNet regressions in ATBC and PLCO, respectively, with two in both cohorts. Notable findings include: fibrinopeptide B(1-9); seven modified, di- or poly-peptides; eight xenobiotics related to tobacco smoke; amino acids, carbohydrates, and TCA cycle metabolites known to be regulated by mutant KRAS (aspartate, glutamate, lactate, pyruvate, and α-ketoglutarate); four secondary bile acids and three aromatic amino acids (tryptophan, phenylalanine, and tyrosine) that were positively, while four fibrinogen cleavage peptides were inversely associated with PDAC. Among the top identified metabolites, aspartate, α-glutamyltyrosine, pyroglutamylglutamine, and Glu-Gly-Asn-Val showed time-varying associations overall (P≤0.05) with greater magnitudes within the first 5 years after blood draw. Thirteen metabolic pathways were associated with PDAC (Bonferroni P≤0.05). Conclusion: Our study suggests pre-diagnostic circulating metabolites, particularly those potentially related to subclinical disease, microbiome, and tobacco metabolism are associated with PDAC. The identified metabolites along with other known risk factors might be useful in algorithms to model disease risk and progression. Citation Format: Ting Zhang, Steven C. Moore, Sheng Fu, Kevin Wang, Demetrius Albanes, Stephanie J. Weinstein, Kai Yu, Rachael Z. Stolzenberg-Solomon. Serum metabolites and pancreatic ductal adenocarcinoma in two prospective cohorts [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 3427.

Abstract 840: Serum concentrations of per- and polyfluorinated substances and risk of non-Hodgkin lymphoma

Abstract Per- and polyfluoroalkyl substances (PFAS) are persistent organic pollutants, some immunotoxic, that are detectable in the serum of most U.S. adults. Some studies of highly-exposed individuals have suggested positive associations between PFAS and non-Hodgkin lymphoma (NHL); however, it is unknown whether associations exist at lower exposure levels that are more common in the general population. To address this knowledge gap, we conducted a nested case-control study investigating serum PFAS concentrations and NHL within the Prostate, Lung, Colorectal and Ovarian Cancer Screening Trial. We measured pre-diagnostic serum concentrations (collected between 1993-2001) of five PFAS among 706 cases and 706 controls individually matched on the basis of age at baseline, sex, self-reported race and ethnicity, study center, calendar year of blood collection, and number of prior serum thaws. We estimated odds ratios (ORs) and 95% confidence intervals (CIs) for PFAS concentrations in relation to NHL, both overall and for selected histologic subtypes [diffuse large B cell lymphoma (DLBCL), follicular lymphoma (FL), lymphoplasmacytic lymphoma (LPL), and marginal zone lymphoma (MZL)] using conditional logistic regression. We found no evidence of a positive association with NHL for any of the five PFAS in analyses adjusting for all measured PFAS. In analyses of histologic subtypes, we observed a positive association between perfluorohexane sulfonate (PFHxS) and DLBCL (OR per doubling in concentration=1.27, 95% CI=1.07, 1.50) which remained for cases diagnosed ≥10 years after blood collection (OR=1.34, 95% CI=1.09, 1.64). We also observed inverse associations with PFNA for all subtypes (DLBCL, OR=0.84, 95% CI=0.72, 0.98; FL, OR=0.86, 95% CI=0.69, 1.08; LPL/MZL, OR=0.86, 95% CI=0.66, 1.12), although these associations were null among participants with blood drawn prior to 1997 (DLBCL: OR&amp;lt;1997=0.95, OR≥1997=0.67, Pinteraction=0.001; FL: OR&amp;lt;1997=1.01, OR≥1997=0.77, Pinteraction =0.08; LPL/MZL: OR&amp;lt;1997=1.26, OR≥1997=0.88, Pinteraction =0.03). In conclusion, our findings from a cohort study with PFAS serum concentrations comparable to that of the general population do not support an association with increased risk of NHL overall. The suggestive evidence of a positive association between PFHxS and DLBCL warrants further investigation. Citation Format: Jongeun Rhee, Jani Koponen, Joshua N. Sampson, Alexander P. Keil, Mary H. Ward, Jonathan N. Hofmann, Wen-Yi Huang, Debra T. Silverman, Panu Rantakokko, Mark P. Purdue. Serum concentrations of per- and polyfluorinated substances and risk of non-Hodgkin lymphoma [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 840.

Abstract 3430: Mosaic loss of the Y chromosome is associated with prostate cancer risk

Abstract Mosaic loss of the Y chromosome (mLOY), a type of clonal hematopoiesis, is the most frequent chromosomal alteration observed in leukocytes of aging men. mLOY is a putative biomarker of genomic instability with evidence of associations with risk of some solid tumors, but more studies are needed to refine associations and identify potential mechanisms for cancer risk. Prostate cancer (PCa), the most common non-cutaneous cancer in males, is also associated with aging. To further investigate potential relationships between mLOY and PCa, we examined DNA derived from leukocytes of male participants in two large biobanks: the UK Biobank (UKBB; N = 210,103) and the Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial (PLCO, N = 26,795). Of the 236,898 male participants without previous cancer diagnosis, 14,253 (6.0%) were diagnosed with incident PCa after DNA collection. High-density genotyping array data collected during the studies were used to create virtual karyotypes for detecting mLOY in whole blood samples utilizing a phase-based detection method in MoChA software. Analyses were restricted to men without cancer at genotyping and multivariable regression models were adjusted for age, smoking status, and genetic similarity. Men with detectable mLOY were older (mean = 62 vs 56 years) and more likely to be smokers (14.8% vs 9.7%) than men without mLOY. A total of 3,848 (27.0%) men with PCa and 42,835 (19.2%) men free of PCa had detectable mLOY. mLOY clonal fraction ranged from 0.01% to 72.1% of total leukocytes. 18.7% of PCa cases and 13.6% of men free of PCa had high clonal fraction mLOY, defined as greater than 10% of leukocytes with mLOY. Fixed effect meta-analysis of multivariable models from UKBB and PLCO produced evidence for a positive association between mLOY in leukocytes and incident PCa; (Odds Ratio (OR) = 1.062, 95% Confidence Interval (CI) [1.020, 1.107], P-Value (p) = 0.004). The effect estimate was the same in men with low clonal fraction of mLOY (OR = 1.062, 95% CI [1.012, 1.113], p = 0.014) and men with high clonal fraction of mLOY (OR = 1.065, 95% CI [0.998, 1.137], p = 0.057). Age at PCa diagnosis was not associated with mLOY or the clonal fraction of mLOY. A sub-analysis of separate participants from PLCO with mLOY detected in buccal cells (n = 18,011) also showed a small positive effect between mLOY and PCa risk, but the relationship did not reach statistical significance (OR = 1.036, 95% CI [0.877, 1.224], p = 0.673)). Our investigation of the relationship between mLOY and PCa in two large prospective studies provides additional evidence for an association between mLOY detected in the blood and increased PCa risk, meriting additional studies of potential biologic mechanisms underlying the relationship. Citation Format: Rebecca L. Kelly, Weiyin Zhou, Kara M. Barnao, Corey D. Young, Aubrey K. Hubbard, Sairah M. Khan, Wen-Yi Huang, Stephen J. Chanock, Mitchell J. Machiela. Mosaic loss of the Y chromosome is associated with prostate cancer risk [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 3430.

Abstract 6617: Oral microbiome and subsequent risk for head and neck squamous cell cancer development

Abstract Background: The oral microbiota may be involved in head and neck squamous cell cancer (HNSCC) etiology, yet current evidence is largely limited to bacterial 16S amplicon sequencing or small retrospective case-control studies. We tested whether any oral bacteria and/or fungi influence the subsequent risk of HNSCC development. Methods: We conducted a prospective case-control study nested within three epidemiological cohorts: the ACS-Cancer Prevention Study-II Nutrition Cohort (ACS-CPS-II), the Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial (PLCO), and the Southern Community Cohort Study (SCCS). Among 159,840 cohort participants, 236 individuals developed HNSCC during an average of 5.1 years of follow-up, and 485 controls who remained HNSCC-free were selected by 2:1 frequency matching, by cohort, age, sex, race/ethnicity, and time since mouthwash collection. All participants provided mouthwash samples and were cancer-free at baseline. We characterized the oral bacterial microbiome, using whole-genome shotgun sequencing, and the oral fungal microbiome, using ITS sequencing. We assessed relative abundance of microbiome-wide bacterial and fungal taxa for association with HNSCC by ANCOM-BC analysis. We also evaluated selected oral pathogen complexes for association with HNSCC by logistic regression. A microbial risk score (MRS) for HNSCC risk was calculated from all risk-associated microbes. Results: Overall microbiome diversity at baseline was not related to the subsequent risk of HNSCC; however, we found that 13 oral bacterial species were differentially associated with development of this disease (P&amp;lt;0.05), including newly identified Prevotella salivae, Streptococcus sanguinis, Leptotrichia sp, and several species belonging to Beta- and Gamma-proteobacteria. Furthermore, a priori periodontal pathogen complexes were associated with a higher risk of HNSCC (“red/orange” complex, OR=1.06, 95% CI=1.00-1.12). We did not identify any fungal taxa associated with HNSCC risk. Together, an MRS constructed for risk-associated bacteria and bacterial periodontal complexes conferred a 50% increase in HNSCC risk, per standard deviation increase in the MRS (multivariate OR=1.50, 95% CI=1.21-1.85). Conclusions: Our research yields compelling evidence that oral bacteria are a risk factor for the development of HNSCC. The identified bacteria and bacterial complexes hold promise for identifying high-risk individuals and may further lead to personalized prevention approaches for HNSCC. Citation Format: Soyoung Kwak, Mykhaylo Usyk, Feng Wu, Neal D. Freedman, Wen-Yi Huang, Marji L. Mccullough, Caroline Y. Um, Martha J. Shrubsole, Qiuyin Cai, Chan Wang, Huilin Li, Jiyoung Ahn, Richard B. Hayes. Oral microbiome and subsequent risk for head and neck squamous cell cancer development [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 6617.

Abstract 3896: Time to diagnosis analysis using miRNA-based ovarian cancer prediction models

Abstract Background: The recent literature proposes several diagnostic Machine Learning (ML) models for ovarian cancer based on miRNA expression profiling. These ML models are trained and validated on subjects whose miRNA sample was drawn proximate to cancer diagnosis (e.g., within one month). Whether these models remain useful remote from the time of diagnosis, when early detection or prevention would be most relevant, is unclear. Therefore, we examine the effect of time between blood draw and cancer diagnosis on miRNA-based ML model accuracy and the relevance of a cancer probability score, Pc, for estimating long-term cancer risk. Methods: The study is based on the miRNA expression profiles of 2983 total subjects, which is comprised of 1829 subjects collected as part of the Biobank at Mass General Brigham (MGB), 110 samples from the Pelvic Mass Protocol at Brigham and Women's Hospital (Cramer), and 1044 samples which were obtained from the Prostate, Lung, Colorectal, and Ovarian (PLCO) cancer screening trial. miRNA expression was measured using a pre-specified panel of 179 miRNAs optimized for serum detection using the Fireplex® circulating miRNA assay. We trained ML models using 1865 controls and 74 ovarian cancer subjects (1829 Biobank + 55 Cramer + 55 PLCO). The training cancer samples were drawn between 1 day and 14 years from diagnosis, most within 30 days. The models were then validated on 769 control and 275 ovarian cancer subjects (989 PLCO + 55 Cramer) whose time to cancer diagnosis ranges between 1 and 1814 days (up to 5 years) after blood draw. Performance was reported as Area Under the receiver operator characteristic Curve (AUC). Results: Among the validation set cases, we observe a decreasing trend in predicted cancer probability (Pc) with increasing log time (R = -0.34, p &amp;lt; 0.0001), and we see a similar trend in AUC score with time. On samples drawn within 21 days of cancer diagnosis, the ML model offers an AUC = 0.88, which decreases to AUC = 0.72 on samples drawn between 21 days and one year from diagnosis, and later plateaus at AUC ~ 0.72 up to five years from diagnosis. Out of 2928 total subjects considered in the study, 286 had multiple blood draws taken over a 5-year time period. Using this data, we analyze the change in Pc over time per subject. The results show that Pc, for the average case subject, increased by 7% per year (p = 0.02). In contrast, when we applied the same analysis to control subjects, Pc increased by only 1% per year on average (p = 0.62). Thus, monitoring changes in Pc at regular intervals could be an informative cancer diagnostic. In terms of relative risk, subjects with Pc &amp;lt; 0.5 had a relative 5-year cancer risk of 0.74, whereas subjects with Pc &amp;gt; 0.5 had relative risk 7.4 (i.e., an order of magnitude higher). Conclusion: The results indicate that miRNA-based ML models can be used to identify individuals at increased long-term risk of ovarian cancer and provide a tool for profiling at regular intervals to allow earlier diagnosis of disease. Citation Format: James Webber, Laura Wollborn, Sudhanshu Mishra, Stephanie Alimena, Bryanna Testino, Allison Vitonis, Daniel Cramer, Dipanjan Chowdhury, Kevin Elias. Time to diagnosis analysis using miRNA-based ovarian cancer prediction models [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 3896.

Abstract 6151: Relationship between polygenic risk score, lifestyle factors, and colorectal cancer risk

Abstract Although heritable factors are estimated to explain up to 35% of colorectal cancers (CRC), most are diagnosed in people without a family history of CRC suggesting that inherited genetic variants, environmental factors, or a combination of both, contribute to CRC risk. We explored the relationship between a polygenic risk score (PRS) for CRC risk and lifestyle risk factors to understand whether individuals with higher genetic risk, who do not develop CRC, engage in healthier lifestyle behaviors, or conversely, whether those with lower genetic risk, who do develop CRC, engage in riskier behaviors in the Prostate, Lung, Colorectal, and Ovarian (PLCO) Cancer Screening Trial. Using summary statistics obtained from a large independent GWAS meta-analysis of CRC by the GECCO consortium, we developed a genome-wide PRS for CRC using SBayesR and applied it to participants of European ancestry genotyped in the PLCO Trial (n=74,396). For all subjects, we computed an environmental (E) score based on 19 lifestyle risk factors for CRC, a WCRF/AICR score that measured adherence to the 2018 Cancer Prevention recommendations, and a 2015 healthy eating index (HEI) with higher values indicating healthier lifestyle profiles. We used t-tests to evaluate differences in lifestyle scores between CRC-free participants in the highest quartile of PRS and incident CRC cases in the lowest quartile of PRS. To examine the correlations between the PRS and lifestyle scores, we performed multivariable linear regression models regressing PRS on each lifestyle score. We used multivariable Cox hazards regression models to compute hazard ratios and 95% confidence intervals for CRC risk; multiplicative interactions between PRS and lifestyle scores were evaluated using the Wald test for the interaction terms. During the 15 years follow-up period, a total of 1,052 CRC cases were diagnosed. Individuals in the highest quartile of PRS who remained cancer-free were more likely to engage in healthier behaviors than those in the lowest quartile of PRS who were diagnosed with incident CRC (E-score: -0.01 vs. -0.18, WCRF/AICR score: -0.05 vs. -0.32, HEI score: 0.06 vs. -0.13); however, differences were not statistically significant (P value: 0.1 to 0.4). PRS and lifestyle scores were significantly (all P &amp;lt;0.0001), albeit weakly (β: -0.02 to -0.03), associated among participants who remained CRC-free but not among CRC cases (P: 0.1 to 0.3). PRS-lifestyle score interactions were not statistically significant (P: 0.2 to 0.4). Although the differences and multiplicative interaction models were not statistically significant, we found that individuals with a higher PRS who remained CRC-free were slightly more likely to engage in healthier behaviors than those at a lower genetic score who developed CRC. Additional studies with larger sample size are warranted to further explore the interplay between genetic risk and lifestyle factors related to CRC risk. Citation Format: Wen-Yi Huang, Erikka Loftfield, Rebecca Landy, Difei Wang, Hormuzd A. Katki, Jianxin Shi, Lingxiao Wang, Mitchell J. Machiela, Neal D. Freedman, Sonja I. Berndt. Relationship between polygenic risk score, lifestyle factors, and colorectal cancer risk [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 6151.

Fiber and whole grain intakes in relation to liver cancer risk: An analysis in 2 prospective cohorts and systematic review and meta-analysis of prospective studies.

The association between fiber or whole grain intakes and the risk of liver cancer remains unclear. We assessed the associations between fiber or whole grain intakes and liver cancer risk among 2 prospective studies, and systematically reviewed and meta-analyzed these results with published prospective studies. A total of 111,396 participants from the Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial (PLCO) and 26,085 men from the Alpha-Tocopherol, Beta-Carotene Cancer Prevention Study were included. Intakes of total fiber and whole grains were estimated from validated food frequency questionnaires. Study-specific HRs and 95% CI with liver cancer risk were estimated using multivariable-adjusted Cox regression. We systematically reviewed existing literature, and studies were combined in a dose-response meta-analysis. A total of 277 (median follow-up = 15.6 y) and 165 (median follow-up = 16.0 y) cases of liver cancer were observed in Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial and Alpha-Tocopherol, Beta-Carotene Cancer Prevention Study, respectively. Dietary fiber was inversely associated with liver cancer risk in Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial (HR 10g/day : 0.69; 95% CI: 0.55-0.86). No significant associations were observed between whole grain intakes and liver cancer risk in either study. Our meta-analysis included 2383 incident liver cancer cases (7 prospective cohorts) for fiber intake and 1523 cases (5 prospective cohorts) for whole grain intake; combined HRs for liver cancer risk were 0.83 (0.76-0.91) per 10 g/day of fiber and 0.92 (0.85-0.99) per 16 g/day (1 serving) of whole grains. Dietary fiber and whole grains were inversely associated with liver cancer risk. Further research exploring potential mechanisms and different fiber types is needed.

Carbohydrate quality, not quantity, linked to reduced colorectal cancer incidence and mortality in US populations: evidence from a prospective study

BackgroundCarbohydrates have been implicated in colorectal cancer (CRC) risk, but the specific impact of carbohydrate quality and quantity on CRC susceptibility in US populations remains unclear.MethodsWe followed 101,694 participants from Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial. The carbohydrate quality index (CQI) and low-carbohydrate diet score (LCDs) were used to evaluate the daily carbohydrate quality and quantity separately, where higher scores indicated greater adherence. Cox proportional hazards regression was used to compute HRs and 95% CIs for incident CRC and related death. Subgroup analyses were conducted to identify potential effect modifiers.ResultsDuring follow-up, we documented 1085 incident cases of CRC, of whom 311 died from CRC. Individuals in the highest compared with the lowest quartiles of CQI had a lower CRC incidence (Q4 vs Q1: HR 0.80, 95% CI 0.67–0.96, Ptrend = 0.012) and mortality (Q4 vs Q1: HR 0.61, 95% CI 0.44–0.86, Ptrend = 0.004). The inverse association between CQI and CRC risk was observed for distal colon and rectum but not for proximal colon cancer. Regarding mortality, this association was only significant for rectum cancer. Subgroup analyses indicated this inverse association of CQI with CRC risk was only observed in participants with lower LCDs. No significant associations were found between LCDs and CRC incidence or mortality.ConclusionsOur findings suggest focusing on higher quality, rather than restricting the quantity, of carbohydrate consumption may be an effective approach to reduce the risk of CRC in the US population, particularly for distal colon and rectal cancers.

Abstract B051: Genome-wide association studies of 26,187 women without ovarian cancer identifies 12 genetic loci of blood CA125

Abstract Background: CA125 is elevated in women with ovarian cancer and is the single best ovarian cancer screening biomarker candidate to date, although two randomized screening trials showed no clinically significant difference in ovarian cancer mortality. Consideration of genetic variants that influence blood CA125 levels in healthy women could improve its performance as an ovarian cancer screening biomarker. Furthermore, while studies have suggested CA125 is associated with ovarian carcinogenesis and progression, its biological role remains unclear. Thus, we aimed to identify genetic variants of CA125 in women without ovarian cancer. Methods: We conducted a genome-wide association analysis to identify common genetic variants associated with CA125 using data on 26,187 women without ovarian cancer, leveraging existing blood CA125 data and genetic data from Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial (PLCO), European Prospective Investigation into Cancer and Nutrition (EPIC), Nurses’ Health Studies (NHS/NHSII), and New England Case Control Study (NEC). All genotype data were inputted using TOPMed. For blood CA125 measurement, PLCO, NHS/NHSII, and NEC used the CA125II assay and EPIC used the MSD assay. We used multivariable linear regression models adjusting for age, study, and population ancestry (top 4 principal components). Analyses were conducted separately by CA125 assay, genotyping platforms, and menopausal status at blood draw. Because epidemiological predictors of CA125 differ by menopausal status, we conducted analyses stratified by menopausal status. When we meta-analyzed these results, there were no significant heterogeneity by menopausal status in majority of the genome-wide significant SNPs. Therefore, we performed a fixed-effects meta-analysis to combine summary statistics across all the results. Results: We identified 12 independent loci associated with CA125 at a genome-wide significance level (p&amp;lt;5 × 10−8), of which one SNP (rs73005869) had been previously linked with CA125 and is located in the MUC16 gene. In our meta-analysis, 7 SNPs were associated with lower and 5 were associated with higher levels of CA125. We discovered SNPs on chromosome 16 in the TET2 gene that were significantly associated with increased CA125 levels. We also discovered SNPs in MSLN, a gene reported to be overexpressed in ovarian tumors that contributes to metastasis in the peritoneal microenvironment, were associated with lower CA125 levels. SNPs in the HIC1 gene on chromosome 17, a tumor suppressor gene whose aberrant methylation has been observed in ovarian tumors, was associated with lower CA125 levels. Conclusion: We identified 12 independent loci associated with blood CA125 levels in women without ovarian cancer, providing novel biological insights on the role of CA125. Accounting for these genetic factors that influence blood CA125 levels could improve the performance of CA125 as ovarian cancer screening biomarker in average risk women, substantially improving survival through earlier detection of this deadly disease. Citation Format: Naoko Sasamoto, Jihye Kim, Constance Turman, Shelley S. Tworoger, Rudolf Kaaks, Renée T. Fortner, Daniel W. Cramer, Kathryn L. Terry, Nicolas Wentzensen, Peter Kraft. Genome-wide association studies of 26,187 women without ovarian cancer identifies 12 genetic loci of blood CA125 [abstract]. In: Proceedings of the AACR Special Conference on Ovarian Cancer; 2023 Oct 5-7; Boston, Massachusetts. Philadelphia (PA): AACR; Cancer Res 2024;84(5 Suppl_2):Abstract nr B051.