5525 Background: Nearly one in five new cervical cancer cases occur in individuals aged ≥ 65, with higher incidence among Black women compared to White women. Despite these disparities, evidence guiding screening and surveillance in older women is limited. This study evaluates guideline concordance for follow-up after abnormal cervical cancer screening and examines variation by age, race, and ethnicity. Methods: This retrospective cohort study assessed 1,861,906 cervical cancer screening results (2009-2024) from the TrinetX database. Included patients were female, ≥ 21 years old, and had an HPV test or pap smear. Guideline concordance (i.e., colposcopy within one year of high-risk cervical cancer screenings result) was assessed by running logistic regression models by age group, race, and ethnicity. Logistic regression assessed odds of concordant follow-up by age, race, and ethnicity, using women aged 30–39 and non-Hispanic White women as reference groups, respectively. Results: Guideline-concordant follow-up varied by age: 56.87% aged 21-29, 58.84% aged 30-39, 56.36% aged 40-49, 53.93% aged 50-65, and 45.80% aged 65+. Compared with women aged 30-39, both younger and older women were significantly less likely to receive guideline concordant follow-up. Women 65+ had 44% lower odds of concordant follow-up (OR= 0.56, 95% CI 0.53 to 0.60, p<.001). Guideline-concordant care also varied by race and ethnicity: 51.11% of non-Hispanic Black, 56.65% of non-Hispanic Asian, 60.36% of non-Hispanic White, 58.43% of non-Hispanic Other, and 51.71% of Hispanic patients. Compared to non-Hispanic White women, all other groups had significantly lower odds of follow-up, with non-Hispanic Black women having 30% lower odds (OR = 0.70, 95% CI 0.68 to 0.72, p<.001) and Hispanic women having 28% lower odds (OR= 0.72, 95% CI 0.71 to 0.74, p<.001) of concordant follow-up. Conclusions: Overall, the rate of guideline concordant follow-up after high-risk cervical cancer screening results was low (57.83%) and declined with age, with lowest rates among women ≥ 65 (45.80%). All other age groups had significantly lower odds of concordance than those 30-39, with those 65+ having the lowest odds of concordance. Significant disparities were also observed by race and ethnicity, particularly among non-Hispanic Black and Hispanic women. Further research is needed to assess the decreased surveillance rates after high-risk cervical cancer screening results at the population level and specific attention should be focused on groups who have been under-surveilled.

Abstract Objective To develop and evaluate an automated large language model (LLM)-based framework for conducting meta-analyses of nutrition-related exposures and the risk of breast, ovarian, and uterine cancers. Design We developed M eta F emina , an automated evidence-synthesis pipeline for women’s cancers that integrates keyword-based literature retrieval, LLM-assisted evidence extraction, and random-effects meta-analysis. We evaluated its performance against two recently published peer-reviewed meta-analyses and compared exposure–outcome associations across the three cancer types. Data sources PubMed articles identified through keyword-based searches of titles and abstracts. Methods M eta F emina was developed as a web platform that identifies relevant scientific articles, automatically extracts relevant information using LLMs, and synthesizes extracted evidence using random-effects meta-analysis. Additional analyses included assessment of heterogeneity, publication bias, and leave-one-out sensitivity analyses. The platform also provides sample size calculations based on synthesized effect sizes and generates visual summaries and plain-language interpretations. Results Compared with two recent peer-reviewed meta-analyses of folate and vitamin E intake in relation to breast cancer risk, M eta F emina demonstrated high sensitivity (81.82% and 80% respectively) in identifying eligible studies and additionally retrieved relevant articles that had been missed by manual screening (27 and 13 respectively). Among 226 exposures considered, lutein and β -carotene were significantly associated with lower risks of breast, ovarian, and uterine cancers. Vitamin D, antioxidants, and soy were significantly associated with lower risks of both breast and ovarian cancers, whereas calcium and folic acid were significantly associated with lower risks of both breast and uterine cancers. In contrast, iron, red meat, and copper were significantly associated with higher risks of both breast and uterine cancers. ω -6 fatty acids showed contrasting associations, being significantly associated with higher breast cancer risk but lower ovarian cancer risk. After restriction to dietary-intake studies, these cross-cancer significant associations remained statistically significant except for copper, which no longer met the two-study threshold for either breast or uterine cancer. Additionally, calcium became significantly associated with lower ovarian cancer risk, resulting in significant negative associations across all three cancer types, while vitamin E became significantly associated with lower breast cancer risk and remained significantly associated with lower ovarian cancer risk. Conclusions M eta F emina demonstrated high sensitivity for identifying relevant scientific literature, extracts key evidence, and performs statistically rigorous automated meta-analyses. The framework may facilitate more rapid evidence synthesis in nutritional epidemiology and may support researchers in study design, hypothesis generation, and interpretation of emerging evidence.

Abstract Introduction: Persistent infection with high-risk human papillomavirus (HPV) is the primary cause of cervical cancer and can lead to high-grade cervical lesions (CIN2/3) requiring treatment and long-term surveillance. Current guidelines recommend at least 25 years of follow-up, including early post-treatment testing and screening every three years. However, age-related differences in adherence to these guidelines remain unclear. This study evaluates the impact of age on surveillance adherence and its implications for recurrence risk. Methods: This retrospective cohort study used U.S.- based TriNetX electronic health record data to evaluate age-related differences in adherence to post-treatment surveillance among women with high-grade cervical lesions. Women aged ≥21 years with histologically confirmed CIN2, CIN3, or high-grade squamous intraepithelial lesions were identified using ICD-9 and ICD-10 codes and stratified into decade-based age groups. Adherence to immediate treatment and short- and long-term surveillance was assessed as binary outcomes and compared across age groups using chi-square testing. Results of Data Characterization: Among all women with abnormal cytology results (641,172 patients), high-grade squamous intraepithelial lesions (HSIL) account for 5.32% (34,155 patients). Notably, HSIL prevalence was highest among younger patients. Among with abnormal cytology, 35% of women aged 21-29 years 30-39 years had HSIL, while prevalence of HSIL was lower in older age groups(15.36% in ages 40-49, 12.11% in ages 50-59, and 5.32% in ages ≥65). Within HSIL lesions, CIN2 and CIN3 were nearly equally represented (50.32% vs 49.68%), and the prevalence by age group showed a similar pattern for both lesions as with HSIL overall. Similarly to HSIL, CIN2 lesions were the most frequent in younger populations: 21-30 years (36.76%) and 31-40 years (36.14%). Cases with moderate numbers persist as age declines: 41-50 years (15.25%), 51-60 years (10.62%), and ≥65 years (1.22%). CIN3 lesions follow a similar distribution as CIN2 lesions across age groups: 21-30 years (33.85%), 31-40 years (36.57%), 41-50 years (17.10%), 51-60 years (10.50%), and ≥65 years (1.99%). Treatment Results: Within 2 years of diagnosis, 103,538 patients were followed. The majority of patients had no treatment (85.33%) whilst a small proportion of patients underwent a loop electrosurgical excision procedure (LEEP) (13.55%) and underwent ablative procedures (1.15%). Conclusions: In this large cohort of women with abnormal cytology, while HSIL represents a smaller proportion of abnormal pap smears, the majority of these lesions are disproportionately concentrated in women who are 40 years old and younger. CIN2 and CIN3 lesions are nearly equally distributed and follow a similar trend by age. Most people did not receive procedural treatment within 2 years of diagnosis. By the time of poster presentation, further analysis will be conducted to determine whether guideline concordance differs significantly across age groups. Citation Format: Shifa Hossain, Maeve Givens, Edward Gemson, Rulla Tamimi. The impact of age on the prevalence and management of high-grade cervical lesions: A retrospective cohort study [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 2 (Late-Breaking, Clinical Trial, and Invited Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(8_Suppl):Abstract nr LB396.

Childhood and adolescence may represent critical time windows for shaping future breast cancer risk. The association between early-life diet and breast cancer risk has been investigated, but few studies have examined the relation between adolescent diet and benign breast disease (BBD), an established breast cancer risk factor. Among 11,422 female Growing Up Today Study participants followed from 1996 to 2016 who completed food frequency questionnaires, we investigated the associations between adherence to three dietary patterns (Alternative Healthy Eating Index [AHEI], the Empirical Dietary Inflammatory Pattern [EDIP], and the Empirical Dietary Index for Hyperinsulinemia [EDIH]) at ages 10 and 14 years and self-reported BBD diagnosis. Cox proportional hazards models were used to estimates hazard ratios (HRs) and 95

Social determinants of health are known drivers of a broad range of cancer outcomes. These determinants influence factors at the structural and institutional levels of society, which disproportionately affects those living in persistent poverty. Americans living in persistent poverty areas experience higher cancer incidence, delayed cancer diagnosis and treatment, and greater cancer morbidity and mortality. This report describes the aims of the NCI’s Persistent Poverty Initiative, a novel research program developed to study the association of structural and institutional factors that affect cancer outcomes in persistent poverty areas. The goals of the five funded centers located across the country are described. A detailed multilevel conceptual framework was developed to incorporate the perspectives of the five centers. The framework serves as a model for other researchers studying the complex social factors and pathways that contribute to cancer outcomes. Future research can build upon this work to integrate multilevel intervention models that can examine the association of upstream social determinants of health factors and cancer outcomes. The Persistent Poverty Initiative can serve as a model for addressing cancer outcomes at the structural and institutional levels, which will hopefully lead to improved local, state, and national strategies in an effort to achieve optimal health for all.

Experimental studies have shown anti-carcinogenic properties of vitamin D, but meta-analyses of randomized trials of vitamin D supplementation show reduction in cancer mortality but not cancer incidence. This study evaluated the effects of randomized moderate-dose vitamin D3 supplementation on mammographic features associated with breast cancer, and gene expression pathways in breast tumors. As part of an ancillary study to the VITamin D and OmegA-3 TriaL (VITAL) trial, we collected digital mammograms from women ≥ 55 years old, randomized in a 2X2 factorial design to vitamin D3 supplementation (2, 000 IU/day of cholecalciferol; n = 1317) or placebo (n = 1280). Digital mammograms were collected prior to randomization, and at 1, 3, and 4 years post randomization. Formalin fixed paraffin embedded tissue tumor blocks were collected for women who reported breast cancer post-randomization. RNASeq data were obtained from 25 tumors in the vitamin D arm and 21 from the placebo arm. The primary outcomes were (a) changes in percent mammographic density (PMD), and V, a mammographic feature that captures gray-scale pixel variation, and (b) tumor gene-expression patterns. A gene expression-based vitamin D score was developed in VITAL and evaluated in 3 independent data sets (Nurses’ Health Studies, TCGA, METABRIC) to determine if it was associated with long-term survival. The mean age at randomization was 64 years of age; 27% of women self-reported their race as Black and 71% White. Both arms experienced significant declines in PMD and V from baseline to year 4. For PMD, those in the vitamin D arm had a 19.6% decline vs. 18.1% in the placebo arm (p=0.41 for difference between arms). For V, those in the vitamin D arm had a 6.5% decline compared with 4.0% in the placebo arm (p=0.02). There was a significant p-trend in the net effect of vitamin D3 over time for V (p=0.01). However, after adjustment for time-varying BMI the net effect was attenuated (p=0.11).Tumors in the vitamin D arm demonstrated down-regulation of multiple cancer related pathways (FDR ≤0.05), including proliferation and inflammation. The vitamin D gene expression score was associated with improved prognosis among stage I to III breast cancer cases in TCGA (p-trend=0.07), METABRIC (p-trend=0.005), and NHS (p-trend=0.04). Overall, randomized moderate dose vitamin D did not have a significant effect on PMD. Vitamin D had a stronger impact on V compared with placebo that appeared to be mediated through changes in BMI. Pathway analyses suggested that supplementation may have an important effect on breast cancer survival, consistent with the reduction in advanced cancers observed in VITAL. In conclusion, there was evidence that vitamin D impacted changes in breast tissue architecture on mammograms and influenced gene-expression patterns that were associated with improved breast cancer prognosis. Rulla M. Tamimi, Franco Giulianini, Cheng Peng, Nancy Cook, Yujin Heng, Vanessa Bret-Mounet, Liza Quintana, Julie Buring, JoAnn Manson, Erin Fowler, John Heine, Kathryn Rexrode. Changes in mammographic features and breast tissue gene expression associated with vitamin D supplementation [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 6428.

Epidemiologic data, supported by experiments, suggest aspirin may improve survival in breast cancer patients. However, recent trials reported a lack of protection, though the length of intervention was limited. Among 10,705 stages I-III breast cancer patients in the Nurses' Health Studies (NHS/NHSII), we examined the associations between post-diagnostic aspirin use and long-term breast cancer survival. During up to 34 years of follow-up, regular post-diagnostic aspirin use was associated with a 38% and 28% lower risk of breast cancer-specific and total mortality. Associations were more evident with longer duration of post-diagnostic aspirin use but attenuated with higher stage and older age at diagnosis. Pre-diagnostic long-term aspirin use was associated with the downregulation of tumor proliferation pathways in NHS/NHSII and the aspirin-gene-expression-signature predicted better survival in METABRIC. Our study highlighted the need for trials with longer duration and suggested that aspirin use before diagnosis may alter the tumor-microenvironment towards a less proliferative type.

Background:An individual's metabolic state plays a critical role in breast cancer (BC) risk, influenced by factors such as obesity and insulin signaling. Hypocaloric diets induce metabolic changes that influence these metabolic factors, thereby potentially influencing BC risk. However, it remains unclear whether metabolic profiles like those induced by such beneficial diets are associated with BC risk. Methods:We compared the impact of a hypocaloric low-carbohydrate ketogenic diet (KD) and a low-fat diet (LFD) on BC risk in two stages. First, we developed metabolomics-based scores representing the metabolic states resulting from these two hypocaloric diets. Plasma metabolomics data of 43 individuals from two controlled dietary interventions were analyzed (N = 31 KD, N = 12 LFD) and a metabolite-based score was generated for both KD and LFD using diet-induced fold-changes. Second, these scores were applied to metabolomics data from a nested case-control study of participants from the Nurses' Health Study II (NHSII, 1,058 BC cases, 1,054 controls, predominantly premenopausal women). Using multivariable-adjusted models, we assessed the association between the metabolomic scores and BC risk. Results:KD and LFD had similar but distinct metabolic signatures. Both metabolomics scores were positively associated with breast cancer risk in NHSII. Women in the highest quartile of the KD metabolomic score had a 37% increased risk of BC compared to women in the lowest quartile (p=0.021). Similarly, women in the highest quartile of the LFD metabolomic score had a 32% increased BC risk compared to women in the lowest quartile (p=0.008). Similar increases in risk were seen when further adjusting for BMI at age 18 and weight change since age 18. Increased levels of cholesterol esters (CE), particularly CE 22:6, and long-chain polyunsaturated triglycerides were associated with higher risk in both diet scores, while increases in short-chain, more saturated triglycerides were associated with lower risk. Conclusion:Metabolomic profiles resembling those induced by hypocaloric ketogenic and low-fat diets were unexpectedly associated with an increased risk of breast cancer in a predominantly premenopausal cohort. These associations were independent of BMI, highlighting the complex relationship between metabolic states and cancer risk, independent of actual dietary interventions.

Identification of healthcare access-related factors linked to lower use of cancer screening can inform targeted policies and interventions to mitigate these barriers and ameliorate screening disparities. We evaluated how barriers to healthcare influence cancer screening rates for five cancer types in a diverse population. We identified cohorts of participants within the All of Us Research Program (2017-2022) meeting U.S. Preventive Services Task Force (USPSTF) guideline eligibility criteria for breast, colorectal, cervical, lung, or prostate cancer screening. Participants self-reported whether nine potential barriers had led them to delay seeking medical care in the past year. Using these items, a barrier burden was calculated by summing the number of affirmative responses (range 0-9), and exploratory factor analysis was performed to identify latent classes of barriers. USPSTF guideline-concordant receipt of cancer screening was assessed in the linked participant electronic health record. Multivariable-adjusted odds ratios (OR) were estimated for the association between each barrier, the barrier burden, and factor scores and receipt of cancer screening, accounting for participant sociodemographic characteristics. Compliance with screening guidelines varied by cancer site: breast (42%, n=31, 827 eligible), cervical (28%, n=37, 770), colorectal (42%, n=68, 895), lung (11%, n=2, 737), and prostate (36%, n=8, 807). The most cited barriers were concerns about out-of-pocket costs, nervousness about seeing a provider, and inability to get time off work, though barrier frequency varied across cancer sites. Participants who were younger, female, had lower income or educational attainment, or identified as Hispanic or non-Hispanic Black were more likely to report 3+ barriers to care. Participants reporting 3+ barriers to care had lower screening rates relative to those who reported no barriers for breast (OR 0.66, 95% CI 0.60-0.72), cervical (OR 0.77, 95% CI 0.72-0.83), colorectal (OR 0.77, 95% CI 0.72-0.82), lung (OR 0.65, 95% CI 0.37-1.09), and prostate cancer (OR 0.78, 95% CI 0.64-0.95). Three latent factors were consistently identified across cancer sites reflecting cost concerns, logistical barriers (e.g. transportation), and competing obligations (e.g. time off work). In multivariable analyses, cost concerns were associated with lower screening rates for breast, cervical, and colorectal cancer, logistical barriers for breast, cervical, colorectal, and prostate cancer, and competing obligations for breast and prostate cancer. The cumulative burden of barriers to access and barriers related to cost concerns, logistics, and competing obligations were associated with lower cancer screening rates. Policies and interventions are needed to target multiple dimensions of access simultaneously to increase cancer screening uptake. Kevin H. Kensler, Anjile An, Aaron Gurayah, Faith Morley, Meenakshi Davuluri, David M. Nanus, Bashir Al Hussein Al Awamlh, Rulla M. Tamimi. Self-reported barriers to healthcare and cancer screening rates: results from the All of Us Research Program [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 7369.

Reproductive factors and sex hormones are tightly linked to systemic immunity. However, no studies have examined whether reproductive factors and hormone use modulate the immune microenvironment of breast tissue. We prospectively evaluated the associations of reproductive factors and exogenous hormone use with breast tumor and normal-adjacent tissue immune cell markers among 935 breast cancer cases in the Nurses' Health Studies. We deconvoluted immune cell abundance using CIBERSORTx and derived gene expression signatures of markers for immune checkpoint (PD1, PDL1, and CTLA4), co-regulatory signal and antigen presentation (MHC class I/ II and T cell receptor), and mammary cytokine signaling. Linear regression was used adjusting for confounders. Patterns of associations between reproductive factors and immune profile differed between tumor and normal-adjacent tissues and by estrogen receptor (ER) status. Tumors from postmenopausal women had significantly higher pro-inflammatory cytokine signaling and antigen presentation compared with tumors from premenopausal women (FDR ≤ 0.05). Several reproductive factors were nominally associated with immune profiles of normal-adjacent tissues. For example, parous women had higher CD8 T cell infiltration, higher PDL1 expression, and lower cytokine signaling (P ≤ .05); women who ever breastfed showed higher infiltration of NK cells and T helper cells in normal-adjacent tissue of ER+ tumors but lower infiltration of CD8 T cell and monocyte, and higher expression of cytokine signaling in normal-adjacent tissue of ER- tumors (P ≤ .05). Our study demonstrates for the first time in a large epidemiologic study that reproductive factors may influence breast tumor immune microenvironment and sheds light on potential immune regulation for breast cancer prevention.

Supplementary Figure S2: Pearson correlation between our algorithm and Volpara shows strong positive linear relationship of 0.81.

Background: Increased TILs are associated with higher pathologic complete response (pCR) rates after NACT for breast cancer, but scant data exist in patients ≤ 40 years old, in whom age-related differences in host and tumor immune microenvironment may exist. We assessed the extent and composition of immune infiltration in breast tumors of young women undergoing NACT and correlated with clinicopathologic features and treatment response. Methods: Patients with stage I-III breast cancer who had NACT and available pre-treatment tumor tissue were identified from a prospective cohort study of women with breast cancer diagnosed at age ≤ 40 years. Multiplexed immunofluorescence was used to quantify cytotoxic T (CD8+), T helper (CD3+CD8-), T regulatory (Treg, FOXP3+CD3+), exhausted T (PD1+CD8+), and PDL1+ cells in stroma and tumor as a percentage value of positive cells. Univariate analyses tested associations of TIL levels (high vs. low, divided based on median) with clinicopathologic variables and compared TIL levels (continuous) between pCR and non-pCR cases. Logistic regression tested associations between TIL subtypes (continuous, per 10% increase) and pCR. Results: Among 194 patients, median age was 36 years (range 22-40 years), 14% were positive for BRCA1/2 mutations, and most had grade 3 (71%), > 2 cm (87%), node-positive (79%) tumors. Forty-one percent of tumors were hormone receptor (HR)-positive/HER2-negative, 33% HER2-positive/HR-positive or -negative, and 27% triple-negative. Sixty-one patients (31%) experienced pCR after NACT. Patients with recent pregnancies (≤ 5 years prior, n=80) had higher stromal infiltration of T helper (P=.020) and cytotoxic T (P=.002) cells as well as higher intratumoral infiltration of cytotoxic T (P=.001), Treg (P=.018) and exhausted T (P=.049) cells compared to those who were nulliparous (n=52), pregnant at diagnosis (n=7) or pregnant > 5 years prior (n=24). BRCA1 mutations (n=19) were associated with high levels of exhausted T cells in stroma (P=.021) and tumor (P=.048). Grade 3 tumors (n=137) had higher levels of T helper cells in stroma (P=.030) and tumor (P=.010). Triple-negative (n=52) and HER2-positive (n=63) tumors had higher stromal Treg infiltration (P=.046). Triple-negative and stage III tumors (n=66) were associated with high levels of PDL1+ cells in stroma (P=.033) and tumor (P=.005), respectively. No differences in TILs were seen according to age, race, ethnicity or histological subtype. Patients with pCR had greater stromal Treg (P=.0189), intratumoral T helper (P=.0043) and intratumoral PDL1+ cell infiltration (P=.0372). Increased stromal Tregs was predictive of pCR, controlling for breast cancer stage, grade and subtype (OR 1.63, 95% CI 1.12-2.28, P=.011). Increased levels of all intratumoral TIL subtypes was predictive of pCR, with odds ratios ranging from 1.14 (95% CI 1.02-1.29, P=.026) for PDL1+ cells to 2.02 (95% CI 1.31-3.12, P=.002) for Tregs for each 10% increment, controlling for breast cancer stage, grade and subtype. Conclusions: The distribution of TIL subtypes in young patients’ breast tumors pre-treatment varied according to recency of pregnancy, BRCA1/2 mutation status and tumor characteristics. High levels of most TIL subtypes was associated with improved response to NACT, independent of breast cancer subtype. The relationship between TILs and pCR in young women appears similar to that in older women, suggesting that tumors of younger and older patients may not be fundamentally different when analyzed according to clinically relevant biologic features. Characterization of immune subpopulations could help refine the predictive value of TILs in young patients with breast cancer, who may benefit from individualized escalated and de-escalated treatment strategies. Citation Format: Megan Tesch, Yaileen D. Guzmán-Arocho, Laura C. Collins, Yujing J. Heng, Shoshana M. Rosenberg, Kathryn J. Ruddy, Rulla Tamimi, Lidia Schapira, Jeffrey Peppercorn, Virginia Borges, Steven E. Come, Craig Snow, Eric P. Winer, Elizabeth A. Mittendorf, Ann H. Partridge. Tumor-infiltrating lymphocytes (TILs) and response to neoadjuvant chemotherapy in young patients with breast cancer [abstract]. In: Proceedings of the San Antonio Breast Cancer Symposium 2024; 2024 Dec 10-13; San Antonio, TX. Philadelphia (PA): AACR; Clin Cancer Res 2025;31(12 Suppl):Abstract nr PS1-04.

Existing evidence indicates that colorectal cancer (CRC) patients from high-poverty areas (HPA), defined as areas with 20% or more residents living below federal poverty level, are more likely to receive surgery at low-volume, non-accredited hospitals than those from low-poverty areas (LPA). The systematic difference in surgical care location between patients from HPAs and LPAs, healthcare segregation, could represent a structural equity through which disparities in health outcomes arise. We sought to determine the existence of healthcare segregation and assess its impact on patient health outcomes. We identified patients undergoing surgery for non-metastatic CRC diagnosed in 2009-2019 from SEER-Medicare. Patients living in a county which had been an HPA for at least 30 years were classified as living in a persistent-poverty area (PPA). Patients living in a county which was an HPA during the year of cancer diagnosis but not a PPA were classified as living in a current-poverty area (CPA). Otherwise, they were classified as living in an LPA. Experiencing healthcare segregation was defined as being treated at a poverty-area-serving (PAS) hospital where ≥50% of the patients were from HPAs. We examined the association between area-level poverty, treatment at PAS hospitals, and health outcomes (postoperative adverse events, 30-day readmission, long-term all-cause mortality (ACM), and cancer-specific mortality (CSM)) using logistic and Cox regression. We performed a subgroup analysis for metropolitan, urban, and rural areas. We included 81,767 patients (7.5% from PPAs, 8.4% from CPAs) with a median age of 78 (IQR: 73-84) years. 53.9% of patients were females, 14% were non-White, and 1.5% lived in rural areas. 81% of the hospitals (180/991) were PAS hospitals, which treated 64% of patients from PPAs and CPAs vs. 3% from LPAs. Compared with patients from LPAs treated at non-PAS hospitals, those treated at PAS hospitals who were from PPAs (ACM HR=1.17(1.08-1.26), CSM HR=1.23(1.11-1.37)) and CPAs (ACM HR=1.15(1.08-1.22), CSM HR=1.23(1.13-1.35)) had worse short and long-term outcomes. Patients from CPAs treated at non-PAS hospitals also had worse outcomes, to a lesser degree (ACM HR=1.12(1.05-1.19), CSM HR=1.16(1.06-1.28)). Among rural residents, being treated at PAS-serving hospitals but not area-level poverty was independently associated with worse outcomes. Patients living in HPAs experience healthcare segregation in CRC treatment, contributing to their worse health outcomes. Healthcare segregation reflects systemic inequities and structural barriers limiting access to resources that patients from HPAs may deploy to mitigate consequences of cancer. Understanding healthcare segregation and factors shaping it can help identify areas for intervention to reduce associated health disparities. Xinyan Zheng, Laura C. Pinheiro, Parisa Tehranifar, Erica Phillips, Rulla M. Tamimi, Steven Y. Chao, Maria Pisu, Chuxuan Gao, Andrew G. Rundle, Jialin Mao. Healthcare segregation experienced by colorectal cancer patients residing in high-poverty areas and its impact on health outcomes [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 3610.

Supplementary Table S4: Hazards ratio (95% Confidence Interval) for association between baseline risk factors and breast cancer risk among pre-menopausal women using CC-VPD only.

Supplementary Table S2: Hazards ratio for association between baseline risk factors and breast cancer risk among postmenopausal women using MLO-VPD only.

Supplementary Figure S5: The proportional hazards assumption is deemed reasonable by checking the Schoenfeld residuals plot for CC-VPD.

Incidence of premenopausal breast cancer (BC) has risen in recent years, though most existing BC prediction models are not generalizable to young women due to underrepresentation of this age group in model development. Using questionnaire-based data from 19 prospective studies harmonized within the Premenopausal Breast Cancer Collaborative Group (PBCCG), representing 783,830 women, we developed a premenopausal BC risk prediction model. The data were split into training (2/3) and validation (1/3) datasets with equal distribution of cohorts in each. In the training dataset variables were chosen from known and hypothesized risk factors: age, age at menarche, age at first birth, parity, breastfeeding, height, BMI, young adulthood BMI, recent weight change, alcohol consumption, first-degree family history of BC, and personal history of benign breast disease (BBD). Hazard ratios (HR) and 95

Importance:Women diagnosed with breast cancer at a young age are felt to have a higher risk for locoregional recurrence (LRR) regardless of type of local therapy. Objective:To assess the long-term incidence of isolated LRR by molecular subtype in a modern multicenter cohort of young women. Design, Setting, and Participants:This cohort study, a multicenter prospective study named the Young Women's Breast Cancer Study, enrolled 1302 women diagnosed with breast cancer at 40 years or younger from 2006 to 2016. Treatment information and incident LRR (ipsilateral breast/chest or lymph node recurrence) were self-reported on study surveys and confirmed with medical record review; molecular subtype was determined by record review. Analysis was reported from February 2023 to May 2025. Main Outcomes and Measures:Cumulative incidence of isolated LRR was calculated using the Kaplan-Meier method; hazard ratios were estimated by Cox proportional hazards regression. Results:The cohort included 1135 women with stage I through III breast cancer who had a median follow-up of 10.1 years (range, 0.4-16.3 years). The age at diagnosis was younger than 30 years for 145 patients (12.8%), 31 to 35 years for 318 patients (28.0%), and 36 to 40 years for 672 patients (59.2%). There were 59 isolated local recurrences (5.2%) and 4 isolated regional recurrences (0.4%). Among patients with local therapy and subtype data available (n = 1128), 366 (32%) had luminal A-like tumors; 240 (21%), luminal B-like tumors; 231 (20%) luminal ERBB2 positive (+)-like (formerly HER2 positive); 90 (8%) ERBB2+-like; and 201 (18%) triple negative. A total of 346 women (30%) had breast-conserving therapy (BCT) (98% of whom had radiation), 296 (26%) unilateral mastectomy, and 487 (43%) bilateral mastectomy. Of women who had mastectomy, 425 (54%) had radiation. The cumulative incidence of LRR at 10.1 years by subtype was as follows: luminal A, 4.4% (range, 1.0%-6.9%); luminal B, 4.7% (range 1.8%-7.7%); luminal ERBB2+, 6.1% (range, 3.1%-8.3%); ERBB2+, 2.2% (range, 0%-6.3%); and triple negative, 6.5% (range, 4.2%-10.1%). The cumulative incidence of LRR by locoregional treatment type at 10.1 years was 6.7% after BCT (range, 4.3%-10.1%), 6.5% after mastectomy without radiation (range, 0%-7.7%), and 2.4% after mastectomy with radiation (range, 1%-4.2%). Although mastectomy with radiation was associated with the lowest risk of LRR on multivariable analysis, when examined within molecular subtype, there were no differences seen. Conclusions and Relevance:In this contemporary cohort of women diagnosed with breast cancer at age 40 years or younger, risk of isolated LRR was relatively low (5.6%) at a median follow-up of 10.1 years, and significant differences were not seen by tumor subtype. Concerns for long-term risk of LRR should not influence surgical decision-making with young women, irrespective of molecular subtype.

BACKGROUND:Significant progress has been made in reducing lung cancer mortality in the United States, largely due to decreased smoking rates and advancements in treatment, particularly for non-small cell lung cancer (NSCLC). However, lung cancer remains the leading cause of cancer-related deaths, with an estimated 127,010 deaths in 2023. Disparities persist in lung cancer incidence, screening, and treatment, influenced by factors such as race, ethnicity, socioeconomic status, and geography. METHODS:This study analyzed NSCLC treatment patterns in 302,744 patients from the National Cancer Database (NCDB) diagnosed between 2015-2018, focusing on adherence to National Comprehensive Cancer Network (NCCN) guidelines and disparities in immunotherapy receipt. RESULTS:Overall, 62% of patients received guideline-concordant treatment, though this varied significantly by stage, with only 54% of stage IV patients receiving appropriate care. Disparities in guideline adherence were observed, particularly among non-Hispanic Black patients, elderly individuals, and uninsured or Medicaid-covered patients. Additionally, patients at academic institutions and high-volume facilities were more likely to receive concordant treatment, whereas those at community cancer programs and minority-serving hospitals (MSH) had lower adherence rates. Immunotherapy, increasingly recommended since 2017, was also less accessible to non-Hispanic Black, Hispanic, and Asian patients, those without private insurance, elderly individuals, and patients receiving treatment at MSH institutions. CONCLUSIONS:These findings highlight the need for targeted interventions to address persistent disparities in lung cancer treatment, present from individual to institutional levels, in order to ensure equitable access to recommended treatments and advanced therapies like immunotherapy.