Designing modern oncology trials requires synthesizing evidence from prior studies to inform hypothesis generation and sample size determination. Trial designs based on incomplete or imprecise summaries can lead to misspecified hypotheses and underpowered studies, resulting in false positive or negative conclusions. To address this challenge, we developed LEAD-ONC (Literature to Evidence for Analytics and Design in Oncology), an AI-assisted framework that transforms published clinical trial reports into quantitative, design-relevant evidence. Given expert-curated trial publications that meet prespecified eligibility criteria, LEAD-ONC uses large language models to extract baseline characteristics and reconstruct individual patient data from Kaplan-Meier curves, followed by Bayesian hierarchical modeling to generate predictive survival distributions for a prespecified target trial population. We demonstrate the framework using five phase III trials in first-line non-small-cell lung cancer evaluating PD-1 or PD-L1 inhibitors with or without CTLA-4 blockade. Clustering based on baseline characteristics identified three clinically interpretable populations defined by histology. For a prospective randomized trial in the mixed-histology population comparing mono versus dual immune checkpoint inhibition, LEAD-ONC projected a modest median overall survival difference of 2.8 months (95 percent credible interval -2.0 to 7.6) and an estimated probability of at least a 3-month benefit of approximately 0.45. As LEAD-ONC remains under active development, these results are intended as preliminary demonstrations of the frameworks potential to support evidence-driven oncology trial design rather than definitive clinical conclusions.

107 Background: Despite effective endocrine therapy, the absolute benefit of adjuvant chemotherapy varies widely in node-positive HR+ disease, particularly among patients with 1-3 positive nodes. Prognostic tools for these patients often rely on genomic assays, which are costly, timely, or inaccessible in many clinical settings. We present a multimodal artificial intelligence (MMAI) model that integrates clinical and histopathological data to quickly stratify risk of distant metastasis and inform therapeutic decisions. Developed in six phase III randomized clinical and validated for chemotherapy benefit in N0 patients, MMAI offers an accessible alternative to genomic tools. Here, we validated MMAI for prognosis and prediction of chemotherapy benefit in SWOG S8814 – a randomized phase III trial of tamoxifen ± chemotherapy (CT) in postmenopausal women with node-positive (N+) HR+ breast cancer. Methods: Patients with digitized baseline H&E- stained diagnostic slides and clinical data (age, tumor size, nodal status) were analyzed (N = 413). The locked MMAI generated a continuous risk score and categorical risk groups (low, high). Associations with disease-free survival (DFS; 168 events) and overall survival (OS; 125 events) were assessed using univariable and multivariable Cox Proportional Hazard models. Hazard ratios (HRs) and 95% confidence intervals (CI) were estimated. Differential CT benefit was evaluated by estimating relative risk reduction by MMAI risk groups. Results: MMAI was prognostic for DFS (HR per SD 1.73, 95% CI 1.48–2.03; p < 0.001) and OS (HR per SD 1.93, 95% CI 1.60–2.32; p < 0.001), remaining significant after adjustment for age, tumor size, and nodal burden. In the subset of patients with 1–3 positive nodes (n = 253), MMAI identified differential CT benefit: high-risk patients (56% of the patients) demonstrated a 26.3% relative reduction in 10-year DFS risk with CAF-T+TAM versus TAM alone, while low-risk patients (44% of the patients) derived minimal benefit (1.8% relative reduction in 10-year DFS risk). Additionally, the addition of CT resulted in DFS HRs of 1.21 (95% CI: 0.51-2.83) and 0.85 (95% CI: 0.55-1.23) in low- and high-risk patients, respectively. Conclusions: In SWOG S8814, a locked MMAI model using routinely available pathology and clinical data independently stratified prognosis and identified node-positive HR+ patients most likely to benefit from adjuvant CT, with minimal benefit among MMAI low-risk patients with 1–3 nodes. These findings support the use of MMAI as a fast (hours instead of weeks), scalable, cost-effective, and non-tissue consumptive alternative to genomic testing to inform adjuvant decisions in HR+ N+ EBC patients.

INTRODUCTION:Recent conflicting reports regarding the tolerability of PARP inhibitors administered concurrently with breast radiotherapy motivate analysis of adverse events reported in a large national cooperative group trial. METHODS:From 05/19 to 06/24, we enrolled patients with inflammatory (T4d) non-metastatic breast cancer to a Phase 2 NCI-cooperative group trial, which randomized patients after neoadjuvant systemic therapy selected by the treating physician and modified radical mastectomy to two arms. The control arm was assigned 50 Gy of chest wall and nodal radiotherapy, including bolus, plus 10 Gy boost. The intervention arm was assigned the same radiotherapeutic regimen with 25 mg of olaparib twice daily during radiotherapy. Adverse events were assessed using the CTCAE v5.0 weekly during radiotherapy. This analysis explores the distribution of radiation dermatitis, all acute adverse events in the chest-wall region, and other adverse events through the end of radiotherapy by study arm, where the adverse events were deemed possibly, probably, or definitely treatment-related. Chi-squared tests were used to compare proportions. RESULTS:Among 146 evaluable participants (73 control, 73 intervention), median age was 54.1. No Grade 4 or 5 treatment-related events were reported. Grade 3 radiation dermatitis was reported in 24.7% of patients in the intervention arm and 5.5% in the control arm (p=0.003) during radiotherapy. When considering all acute chest-region adverse events (Table), 24.7% had grade 3 acute adverse events in the intervention arm vs 6.8% in the control arm (p=0.006) during treatment. One patient on the intervention arm had early, confluent telangiectasia throughout the 50 Gy fields with radiation-induced lichenoid dermatitis. Intervention arm patients were more likely to experience Grade 2 or greater gastrointestinal adverse events (17.8% vs 0%, p<0.001) and any grade of laboratory investigation abnormalities (19.2% vs 0%, p<0.001). CONCLUSION:This analysis suggests continued caution if considering concurrent administration of radiotherapy and olaparib outside of the investigational context.

Private pharmacies are ubiquitous yet underutilized for HIV pre- and post-exposure prophylaxis (PrEP and PEP) delivery. In a cluster-randomized trial in Kenya (NCT05842122), we randomized 60 pharmacies 1:1:1:1 to: client-sustained delivery (~$2/visit user fee); implementor-sustained delivery (~$2/visit reimbursement); counselor-supported delivery (task shifting; ~$1/visit reimbursement); or clinic referral (control; ~$1/referral reimbursement). Commodities were supplied free to pharmacies from government stock. Primary outcomes were PrEP initiation and one-month continuation (any dispensing or refilling, respectively), self-reported by clients 60 days post-enrollment (multiple-comparisons threshold: p=0.017). From June 2023-April 2025, 5,808 clients enrolled; 64% were PEP candidates. Compared to referral, the counselor-supported arm had significantly higher PrEP initiation (RR=6.5, 95% CI [2.6, 16], p<0.001) and continuation rates (RR=5.1, 95% CI [1.4, 19], p=0.016); all intervention arms had significantly higher PEP initiation rates. One seroconversion, one social harm, and two provider needlestick injuries occurred. Pharmacy PrEP/PEP delivery outperformed clinic referral, particularly when fully subsidized and counselor-supported.

Background: Poor compliance with medications has been linked to inferior health outcomes. Inability to complete therapy is often attributed to adverse effects from the treatment or its financial burden. However, other less easily measurable factors (e.g. trust in the recommendation, fear of future adverse events, nocebo effect, misinformation, convenience, competing life priorities, etc.) can also influence patients’ willingness and ability to follow medical recommendations. In this study, we compared the survival of patients who completed versus not the everolimus placebo pill assigned in the control arm of the S1207 randomized trial. Methods: The S1207 trial compared 1 year of adjuvant everolimus versus 1 year of placebo added to standard of care adjuvant endocrine therapy. No benefit of everolimus was seen for invasive disease-free (IDFS) or overall survival (OS1. Treatment completion rates were 48% in the everolimus arm and 73% in the placebo arm. In this exploratory analysis, we assessed in the control arm if iDFS and OS differed between patients who could complete 1 year of placebo versus those who could not. The Kaplan-Meier method was used to estimate survival from 1 year (i.e. end of treatment period) and the log rank test to compare the groups. Multivariable Cox regression model included age, performance status, and prognostic risk group1 as covariates. Results: This 1-year OS landmark analysis included 823 patients in the placebo arm. Median age was 54 years, 86% were White, 32% premenopausal, and 80% were in prognostic risk groups 3 (i.e. ≥ 4 positive nodes at diagnosis) and 4 (≥ 1 positive nodes after neoadjuvant chemotherapy). Two hundred and twenty-one patients (18%) did not complete placebo defined as taking < 75% of total planned placebo pills. Grade 1-2 adverse events (AE) were reported in 77% (515 out of 670) of the patients in the completer cohort and in 63% (133 out of 211) (p < 0.0001) of the non-completer cohort. Grade ≥3 AEs were reported in 5% and 13% (p < 0.0001) respectively. After placebo discontinuation, endocrine therapy continued for 89% of patients in the completer and 85% of patients in the non-completer cohort (p=0.30). The 1-year iDFS landmark analysis included 782 patients in the placebo arm. The estimated 5-year iDFS rates from 1-year were similar, 76.4% and 78.2% respectively (HR=0.95, 95% CI: 0.64 –1.41, p=0.81). At median follow-up of 74 months, the 5-year OS rate from 1-year among placebo completers and non-completers was 88.6% and 67.4%, respectively (HR=0.32, 95%CI: 0.22-0.44, p<0.0001). In multivariable analysis, only completer status was associated with improved OS (HR=0.31, 95%CI 0.22-0.44, p<0.0001). Conclusion: Patients who were not able to complete the assigned placebo drug in the control arm of the randomized S1207 trial had significantly lower OS compared to those who completed the assigned placebo treatment. Reference: 1. Chavez-MacGregor, M., et al. Phase III Randomized, Placebo-Controlled Trial of Endocrine Therapy±1 Year of Everolimus in Patients With High-Risk, Hormone Receptor–Positive, Early-Stage Breast Cancer. J. Clin Oncol, JCO.23.02344 DOI:10.1200/JCO.23.02344. Citation Format: Tara Sanft, Jieling Miao, Allison Meisner, Shay Bellasea, William E. Barlow, Mariana Chavez-MacGregor, Lajos Pusztai. Ability to comply with placebo predicts overall survival in randomized trial [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 PS15-06.

Background: Socioeconomic disparities may drive cancer inequities in Hispanic/Latino populations. We examined associations of perceived access to healthy foods (AHF) and food insecurity (FI) with diet and body mass index (BMI) changes in Latina breast cancer (BC) survivors. Methods: Latina BC survivors in a 12-month intervention trial aiming to increase fruit/vegetable intake and physical activity were analyzed. AHF was from a modified, validated neighborhood environment scale and dichotomized (low–medium vs. high). FI was defined as eating less and/or going hungry due to a lack of money. AHF and FI surveys were self-reported. Outcomes included dietary intake, diet quality, and BMI. Fruit/vegetable intake was log-transformed. Relationships between AHF and FI and changes in diet and BMI were evaluated using generalized estimating equations. Results: Of women with AHF data (n = 86), 58% reported low–medium access and 42% reported high access. Fruit/vegetable (FV) intake declined overall from baseline to 12 months, with greater reductions among low–medium AHF women (−32%, 95% CI: −51%, −7%) compared with high AHF women (−17%, 95% CI: −40%, +13%). Statistically significant 12-month decreases in total calories, carbohydrates, sugars, and fat occurred in low–medium AHF women but not high AHF women, and changes in total energy density, carbohydrates, sugars, and BMI at 12 months were statistically significantly different between women with low–medium AHF and women with high AHF, p ≤ 0.05. Among 157 women, 23% reported FI. Reductions in fruit/vegetable intake were larger in women with FI (−39%, 95% CI: −57%, −14%) than in women without FI (−10% reductions, 95% CI: −25%, +8%) and between-group differences were significant at both 6 and 12 months, p ≤ 0.05. Most diet measures decreased for both FI and non-FI women, with greater decreases among those with FI. Conclusions: Latina BC survivors with FI or perceived limited AHF experienced greater declines in indicators of healthy diets including FV intake. Future interventions should integrate strategies to measure AHF and FI to address disparate access to healthy food options.

Rates of triple negative breast cancer (TNBC) are higher in Black women than in non-Hispanic White women. Breastfeeding duration and younger age at first birth are known risk factors for TNBC and vary by race. To quantify the contribution of these risk factors to disparities in TNBC, we calculated the population-attributable fraction (PAF). A PubMed search was performed to identify relevant studies and pooled odds ratios for breastfeeding for < 6 months and age at first birth < 25 years were calculated. PAF was calculated using the Levin formula. PAF of breastfeeding for < 6 months was 12% (95% confidence interval (CI) 5-20%) among White women and 15% (95%CI 3-26%) among Black women. We estimate that up to 15% of annual new TNBC in Black women and 12% in White women might be avoided by supporting breastfeeding. Policies supporting breastfeeding could hence reduce TNBC incidence and lessen racial disparities.

In Kenya, as in many African countries, private pharmacies are ubiquitous, frequently accessed, and underutilized for the delivery of HIV prevention services. Whether enabling private pharmacies to initiate and manage clients on HIV pre- and post-exposure prophylaxis (PrEP and PEP) leads to greater uptake and continuation than the current standard—pharmacy referral to clinic-based PrEP/PEP—is unknown. To address this gap and inform how private pharmacies might partner with the public sector, we are testing several models of pharmacy-delivered PrEP/PEP in comparison to the current standard. The Pharm PrEP cRCT is a 60-pharmacy, four-arm cluster-randomized controlled trial ongoing in Central and Western Kenya (first enrollment: 26 June 2023). Eligible pharmacies were licensed by the government, had a private room, and were willing to complete research activities (including a 3-day provider training). Study pharmacies were randomized 1:1:1:1 to (1) client-sustained delivery, in which clients pay pharmacies 250 KES ( 2 USD) per PrEP/PEP visit; (2) implementor-sustained delivery, in which clients pay nothing and implementors pay pharmacies 250 KES per PrEP/PEP visit; (3) implementor-sustained + counselor-supported delivery, in which clients pay nothing, delivery is supported by an HIV testing services (HTS) counselor, and implementors pay pharmacies 100 KES (1 USD) per PrEP/PEP visit; or 4) referral (control), in which clients pay nothing and implementors pay pharmacies 100 KES per referral to clinic-based PrEP/PEP. Pharmacies delivering PrEP/PEP receive supporting commodities free from government stock. Primary outcomes are PrEP initiation and continuation (any refilling) reported by clients 60 days post-enrollment; secondary outcomes include PEP initiation, PEP-to-PrEP transition, repeat PEP use, PrEP/PEP initiation, and PrEP/PEP continuation at 60 and 270 days post-enrollment. Primary analyses will compare each intervention arm to the control; secondary analyses will compare intervention arms to one another. We will additionally assess implementation outcomes (e.g., acceptability, feasibility, cost) from client and provider perspectives. This trial will generate evidence on the potential benefits of leveraging private pharmacies for delivery of PrEP and PEP and the relative effectiveness of pharmacy delivery when subsidized by clients, implementors, and/or supported by HTS counselors. The findings may inform enabling policy and approaches for scale-up. ClinicalTrials.gov NCT05842122. Registered on April 5, 2023.

BACKGROUND:Latina breast cancer survivors experience health disparities. Effective lifestyle interventions are sparse. OBJECTIVE:This trial tested the effectiveness of a culturally tailored diet and physical activity (PA) intervention in Latina breast cancer survivors. DESIGN:Using a 2 × 2 factorial design, women were randomized to receive 4 weekly in-person group sessions; 11 months of eHealth communications, in-person sessions, and eHealth; or control. Follow-up data were collected at months 6 and 12. PARTICIPANTS:Eligibility criteria were self-identification as Latina, post-treatment for early-stage breast cancer, and consuming <5 daily servings of fruits and vegetables (F/V) and/or engaging in <150 min/wk of moderate-to-vigorous PA (MVPA). A total of 167 women from New York City were enrolled from July 2016 to October 2018, with 93.4% retention at 12-month follow-up (n = 156). INTERVENTION:All participants received a Fitbit for self-monitoring and a 1-on-1 health coaching session. In-person group sessions included nutrition and PA education, cooking classes, fitness classes or a grocery store visit, and social activities. The eHealth communications included motivational text messaging, e-mailed newsletters, and study website access. Activities were conducted in Spanish and English. MAIN OUTCOME MEASURES:Primary outcomes were 12-month change in F/V servings/d and energy density of food. Secondary outcomes were 12-month change in MVPA and anthropometry. STATISTICAL ANALYSIS:Outcomes comparing intervention arms with the control were examined using generalized estimating equations. RESULTS:At baseline (n = 167), mean age was 56.7 years; 82.3% had overweight or obesity. At month 12, daily F/V intake for women in the in-person sessions increased by 10% and decreased by 44% for women in the control group, a +96% group difference (P = .01); no other between-group differences were observed. At month 12, women in the control group had a 53% increase in minutes per week of MVPA, and women in the in-person plus eHealth group had a 34% decrease, a -57% group difference (P = .01); no other between-group differences were noted. No changes in energy density or weight between groups were observed. CONCLUSIONS:Women randomized to the in-person ¡Mi Vida Saludable! classes modestly increased F/V intake at 12 months relative to control. Those receiving eHealth communication did not have diet, MVPA, or weight change relative to control. More research is needed to understand how to support Latina breast cancer survivors in making sustained diet and PA changes.

Breast cancer subtyping is essential for precision oncology, influencing prognosis, treatment selection, and clinical trial design. The Integrative Subtype Classification (IC) categorizes breast tumors into groups with distinct long-term outcomes based on genomic and correlated transcriptomic features. This method relies on sequencing data, which, despite decreasing costs, is not always available in research or clinical settings. Here we introduce PATH-IC, a computational pathology model that predicts ER+ breast cancer IC subtype risk of relapse categories from routine histology data. We enhance the current state-of-the-art computational pathology approach with BERGERON, which leverages generative AI to correct class imbalance and reduce overfitting, showing that synthetic data improves PATH-IC's performance by the equivalent of 41% more real training samples. PATH-IC achieves a testing AUROC of 0.814, with predictions correlating to Oncotype DX scores and long-term relapse risk. Using attention-based model interpretation and CRAWFORD, a novel embedding-to-image foundation model, we demonstrate that PATH-IC identifies expected tumor microenvironment patterns for IC subtypes and highlights heterochromatin condensation as a key feature of high-risk tumors. Matched single-cell spatial transcriptomics confirm IC subtype-specific gene expression patterns identified by PATH-IC, including active metabolic, proliferative, and proteostasis pathways in the high-risk group. PATH-IC advances computational pathology through generative AI, enabling subtype inference from histopathology data. ### Competing Interest Statement Unrelated to this work, the following interests are declared: C.C. has advised Bristol Myers Squibb, DeepCell, Genentech, NanoString, Pfizer and 3T Biosciences and has equity in 3T Biosciences, DeepCell and Illumina. All other authors declare no competing interests.

ABSTRACT Black men who have sex with men (MSM) are disproportionately burdened by the HIV epidemic in the US. The effectiveness of pre-exposure prophylaxis (PrEP) in preventing HIV infection has been demonstrated through randomized placebo-controlled clinical trials in several populations. Importantly, no such trial has been conducted exclusively among Black MSM in the US, and it would be unethical and infeasible to do so now. To estimate the causal effects of PrEP access, initiation, and adherence on HIV risk, we utilized causal inference methods to combine data from two non-randomized studies that exclusively enrolled Black MSM. The estimated relative risks of HIV were: (i) 0.52 (95% confidence interval: 0.21, 1.22) for individuals with versus without PrEP access, (ii) 0.48 (0.12, 0.89) for individuals who initiated PrEP but were not adherent versus those who did not initiate, and (iii) 0.23 (0.02, 0.80) for individuals who were adherent to PrEP versus those who did not initiate. Beyond addressing the knowledge gap around the effect of PrEP in Black MSM in the US, which may have ramifications for public health, we have provided a framework to combine data from multiple non-randomized studies to estimate causal effects, which has broad utility.

10522 Background: Age-adjusted incidence rates of triple negative breast cancer (TNBC) are higher in African American (AA) women than in non-Hispanic White (White) women. The disproportionate incidence of TNBC in AA women partially drives the 40% higher mortality rate from breast cancer in AA women compared to White women. Several studies have identified lack of breastfeeding and younger age at first birth as risk factors for TNBC. There are large differences in breastfeeding uptake and duration in the US: only 44% of AA women breastfeed for 6 months or longer, compared to 60% of White women. To quantify the contribution of differences in breastfeeding patterns and younger age at first birth to disparities in incidence of TNBC, we calculate the population-attributable fraction (PAF), which is the proportion of the incidence of a disease that may be attributable to a particular exposure. Methods: A systematic review was performed to identify relevant case-control studies. Pooled odds ratios (ORs) for breastfeeding for < 6 months and age at first birth < 25 years were calculated from data extracted from studies using a common effect model. PAF was calculated using the Levin formula. Risk factor population prevalences were extracted from national surveys. Pooled ORs were used as estimates of relative risk (RR). A combined PAF was calculated using the polychoric correlation coefficient between breastfeeding duration and age at first birth to account for contributions of multiple risk factors. Results: Using race specific ORs, the PAF of breastfeeding for < 6 months was 12% (95% confidence interval (CI) 5-20%) among White women and 15% (95% CI 3-26%) among AA women. The PAF of having a child before age 25 was not significantly different from zero for White women (2% (CI -6-11%)) but was significantly different from zero for AA women (21% (CI 5-25%)). The combined PAF was 27% for AA women and 17% for all women in the US. Extrapolating to the US population, we estimated that 4,850 annual cases of TNBC (2,421 among White women and 1,533 among AA women) could be attributed to breastfeeding for < 6 months and age at first birth < 25 years. Conclusions: Given the protective role of breastfeeding against TNBC, policy changes aimed at supporting breastfeeding, addressing structural barriers, and promoting a culture shift could reduce racial disparities in TNBC incidence in the US. [Table: see text]

INTRODUCTION: Mailed fecal immunochemical test (FIT) outreach is an effective strategy to increase colorectal cancer (CRC) screening. The aim of this study was to determine the patient-level, clinic-level, and geographic-level factors associated with CRC screening completion in a mailed FIT outreach program. METHODS: This retrospective cohort study was conducted in the integrated healthcare system of University of Washington Medicine and included patients aged 50-75 years, who were due for CRC screening, and had a primary care encounter in the past 3 years. Eligible patients received mailed outreach that included a letter with information about CRC screening, FIT kit, and a prepaid return envelope. CRC screening and factors associated with completion were obtained from electronic health records and the CRC screening program database. RESULTS: Of the 9,719 patients who received mailed outreach, 29.6% completed FIT mailed outreach. The median FIT return time was 27 days (interquartile range 14-54). On multivariate analysis, patients with a higher area deprivation index, insured through Medicaid, living without a partner, and whose last primary care visit was >12 months ago were less likely to complete a FIT compared with their counterparts. Over a 12-month period, overall CRC screening across the health system increased by 2 percentage points (68%-70%). DISCUSSION: Mailed FIT outreach in an integrated academic-community practice was feasible, with 32% of invited patients completing CRC screening by FIT or colonoscopy, on par with published literature. Patient and geographic-level factors were associated with CRC screening completion. These data will inform additional interventions aimed to increase CRC screening participation in this population.

PURPOSE:Clinicopathological factors and the 21-gene Oncotype DX Breast Recurrence Score (RS) test both influence prognosis. Our goal was to develop a new tool, RSClinN+, to individualize recurrence risk and chemotherapy benefit predictions by menopausal status for patients with HR+/human epidermal growth factor receptor 2-negative, lymph node-positive breast cancer by integrating the RS result with clinicopathological factors (grade, tumor size, age). METHODS:We used patient-level data from 5,283 patients treated with chemoendocrine therapy (CET) versus endocrine therapy alone (ET) in the S1007 (N = 4,916) and S8814 (N = 367) trials to develop the tool. Cox proportional hazards regression models stratified by trial were used to estimate 5-year invasive disease-free survival for pre- and postmenopausal woman, respectively. The integrated RSClinN+ model was compared with RS alone and clinicopathological models using likelihood ratio tests. Absolute CET benefit was estimated as the difference between ET and CET risk estimates. Validation of RSClinN+ was performed in 592 patients with node-positive disease in the Clalit Health Services registry. RESULTS:RSClinN+ provides better prognostic information than RS model alone (premenopausal P = .034; postmenopausal P < .001) or clinicopathological model alone (premenopausal P = .002; postmenopausal, P < .001). In postmenopausal women, RS showed interaction with CET benefit (P = .016), with RSClinN+ absolute CET benefit ranging from <0.1% to 21.5% over RS ranges 0-50. In premenopausal patients with RS ≤25, there was no significant interaction between RS and CET benefit. In external validation, RSClinN+ risk estimates were prognostic (hazard ratio, 1.75 [95% CI, 1.38 to 2.20]) and concordant with observed risk (Lin's concordance, 0.92). CONCLUSION:RSClinN+ provides improved estimates of prognosis and absolute CET benefit for individual patients compared with RS or with clinical data alone and could be used in patient counseling.

BACKGROUND:Adherence to daily oral pre-exposure prophylaxis (PrEP) is low among African young women, and layered support strategies are needed to improve PrEP adherence in this population. We aimed to evaluate potentially scalable adherence-support strategies for young women aged 18-25 years who initiated PrEP in Johannesburg, South Africa. METHODS:We conducted a sequential multiple-assignment randomised trial at Ward 21 of the Wits Reproductive Health and HIV Institute clinical research site, affiliated with University of the Witwatersrand, Johannesburg, South Africa. Participants were eligible if they were assigned female sex at birth, aged 18-25 years, not living with HIV, sexually active, newly initiating PrEP, had regular access to a mobile telephone, and could read. Using sequentially numbered, sealed, opaque envelopes containing group allocation, a staff member assigned enrolled participants (1:1) to receive one of two adherence-support interventions: once per week two-way SMS communication or participation in a WhatsApp peer-support group. Participants assigned to WhatsApp were put into groups with approximately 25 participants, during which they were prompted by staff facilitators to discuss any challenges with PrEP use or other events happening in their lives. The allocation sequence was generated by the data manager using random numbers with variable block sizes between 10 and 14. Only trial investigators were masked to participant intervention assignments; participants, people giving interventions, people assessing outcomes, and people analysing data were not masked to group assignment. All enrolled participants were offered PrEP (ie, co-formulated, once per day oral emtricitabine 200 mg and tenofovir disoproxil fumarate 300 mg). The primary outcome was high PrEP adherence at month 9, defined as concentration of tenofovir diphosphate on dried blood sample of 700 fmol per punch or more. At month 3, participants with low PrEP adherence were randomly assigned to a secondary, intensified intervention of issue-focused counselling once per month or drug-level feedback counselling based on PrEP drug concentrations at months 3 and 6. The protocol was registered at ClinicalTrials.gov (NCT04038060) and the trial is complete. FINDINGS:Participants were enrolled and followed up between May 16, 2019, and Jan 25, 2022. From May 16, 2019, to Jan 29, 2021, 401 participants were screened and 360 were enrolled and initiated PrEP. 180 (50%) were randomly assigned to two-way SMS and 180 (50%) were randomly assigned to WhatsApp support groups. At month 9, 34 (20%) of 174 participants in the two-way SMS arm had tenofovir diphosphate 700 fmol per punch or more, compared with 32 (18%) of 174 in the WhatsApp arm (relative risk 1·06, 95% CI 0·69-1·64; p=0·78). At month 9, four (5%) of 76 participants in the drug-level feedback arm had tenofovir diphosphate 700 fmol per punch or more, compared with three (4%) of 76 participants in the monthly counselling arm (1·33, 0·31-5·76; p=0·70). 22 serious adverse events were reported during the trial, but were all deemed unrelated to the trial. INTERPRETATION:PrEP adherence did not differ across interventions among young women in Johannesburg, South Africa. Future research is needed on whether and how to scale-up PrEP support for young women in resource-constrained settings. FUNDING:US National Institutes of Health.

Background Black men who have sex with men (MSM) are disproportionately burdened by the HIV epidemic in the USA. The effectiveness of pre-exposure prophylaxis (PrEP) in preventing HIV infection has been demonstrated through randomized placebo-controlled clinical trials in several populations. Importantly, no such trial has been conducted exclusively among Black MSM in the USA, and it would be unethical and infeasible to do so now.Methods To estimate the causal effects of PrEP access, initiation, and adherence on HIV risk, we utilized causal inference methods to combine data from two non-randomized studies that exclusively enrolled Black MSM.Results The estimated relative risks of HIV were: (i) 0.52 (95% confidence interval: 0.21, 1.22) for individuals with versus without PrEP access, (ii) 0.48 (0.12, 0.89) for individuals who initiated PrEP but were not adherent versus those who did not initiate, and (iii) 0.23 (0.02, 0.80) for individuals who were adherent to PrEP versus those who did not initiate.Conclusion Beyond addressing the knowledge gap around the effect of PrEP in Black MSM in the USA, which may have ramifications for public health, we have provided a framework to combine data from multiple non-randomized studies to estimate causal effects, which has broad utility.

508 Background: The RSClin tool was developed to provide estimates of recurrence risk and absolute chemotherapy benefit for patients with HR+/HER2- node-negative breast cancer using the 21-gene recurrence score (RS) and clinicopathologic factors. For patients with node-positive disease, we developed a new tool (RSClin N+) to increase the prognostic and predictive utility of risk estimates by combining RS with clinicopathologic factors and menopausal status. Methods: We used Cox regression to estimate 5-year risk of invasive disease or death (iDFS) and likelihood ratio (LR) tests to compare fit of RSClin N+ with RS alone and clinicopathologic factors (tumor grade, tumor size, positive lymph nodes [N1/N2], age) alone in 5283 node-positive patients treated with chemoendocrine therapy (CET) vs endocrine therapy alone (ET) in the S8814 (n=367: N1=227, N2=140, all postmenopausal) and RxPONDER (n/N1=4916) trials. Data from both studies were pooled and stratified by study for postmenopausal models; premenopausal models were fit only on RxPONDER data. Absolute CET benefit across combination of covariate values was estimated as the difference between ET and CET risk estimates. Validation of RSClin N+ was performed in 592 Nmic/N1 patients in the Clalit registry real-world dataset. Results: For pre- and postmenopausal patients, RSClin N+ provided significantly more prognostic information for iDFS than RS or clinicopathologic factors alone (LR Chi-square p<0.05). In postmenopausal patients, RS and clinicopathologic factors were independently prognostic, but only RS showed interaction with CET benefit (p=0.016). Absolute CET benefit for iDFS ranged from <0.1% to 21.5% as the RS increased from 0 to 50. Table shows examples of risk stratification and predicted benefit by RS within clinical low and high risk patients. In premenopausal patients, both RS and clinicopathologic factors remained prognostic but no interaction was observed between RS and CET.In external validation, RSClin N+ risk estimates were concordant with (Lin’s concordance=0.92) and prognostic for observed risk (HR 1.75; 95% CI 1.38 to 2.20). Conclusions: The RSClin N+ model provides improved estimates of prognostic risk and absolute CET benefit than clinical or genomic data alone in node-positive, HR+/HER2- breast cancer and could be used in patient counseling. [Table: see text]

Abstract Socioeconomic disparities prevalent in Hispanic populations may cause inequities in cancer outcomes and risk factors. We explored the association of access to healthy foods (AHF) and food insecurity (FI) with changes in fruit and vegetable (FV) intake in Latina breast cancer (BC) survivors living in New York City who participated in the Mi Vida Saludable trial. Mi Vida Saludable was a 2x2 factorial trial (n=167) that tested the effects of an in-person and/or eHealth intervention on FV intake and physical activity (PA). Main study results showed only the in-person intervention had modest increases in FV intake. Eligibility criteria included being female, self-identifying as Latina/Hispanic, ³18 years old, history of stage 0-III BC, >3 months post-treatment, <5 daily serving of FV and/or <150 minutes/week of moderate-vigorous PA. FV intake was assessed via 3 24-hour diet recalls at baseline and 12 months. AHF was from a modified validated Access to Healthy Food scale. FI was assessed via questions about eating less or going hungry due to lack of money. Generalized estimating equations examined whether baseline AHF and FI were associated with change in daily FV intake, adjusting for baseline age, marital status, education, and study arm. Three-way interaction terms (AHF or FI with time and study arm) were included to test if associations of AHF and FI with change in FV intake varied by intervention arms. FV intake was log-transformed due to a non-normal distribution; percent changes from baseline were described. AHF was measured in 86 women with 58% and 42% reporting low-medium and high access, respectively. Women with high AHF were more likely to be older (59 vs. 53 years, p=.002) and scored lower on a 5-point acculturation scale (higher score suggests higher acculturation, 1.8 vs. 2.1, p=.01). FI, measured in 157 participant, was prevalent in 23% of women. Women with FI were more likely to be single (44% vs. 16%, p=.003), renting (100% vs. 86%, p=.02), and have higher mean BMI (31 vs. 29 kg/m2, p=.02). 3-way interactions were not statistically significant, though analyses were likely underpowered. In the control group, the adjusted daily FV intake decreased by 79% (p=.01) from baseline to 12 months for women with low-medium AHF and by 49% (p=.10) for those with high AHF; similarly, daily FV intake decreased by 64% (p=.01) for women with FI and by 58% (p=.01) for women without FI. Most Latina BC survivors in New York City who participated in a 12-month diet and PA intervention trial had low-medium AHF at baseline; nearly 1/4 had FI. Associations of AHF or FI with changes in FV intake did not vary significantly by study arm, although sample sizes across strata were small. In the control arm, women with low-medium AHF and FI had larger decreases in FV intake after 12 months. Future dietary intervention studies in Latina BC survivors, and other underserved or under-resourced populations, should account for AHF and FI in the study design, and develop strategies to connect people to healthy food resources. Citation Format: Zachary O. Kadro, Eileen Rillamas-Sun, Blake Langley, Pam Koch, Isobel Contento, Ann Ogden Gaffney, Allison Meisner, Heather Greenlee. Associations between the food environment and food insecurity on fruit and vegetable intake among urban-dwelling Latina breast cancer survivors participating in the Mi Vida Saludable trial [abstract]. In: Proceedings of the 17th AACR Conference on the Science of Cancer Health Disparities in Racial/Ethnic Minorities and the Medically Underserved; 2024 Sep 21-24; Los Angeles, CA. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2024;33(9 Suppl):Abstract nr B143.